Wettability and Wettability Alteration at The Pore - and Nano - Scales

Imperial College London
Department of Earth Science and Engineering
Wettability and wettability alteration at the

pore- and nano- scales
Maja Rücker
Submitted in part fulfilment of the requirements for the degree of

Doctor of Philosophy in Petroleum Engineering of the Imperial College London July 2018
Declaration of Originality
I herewith certify that all material in this dissertation which is not my own work has been
properly acknowledged.
Maja Rücker
i
ii
Copyright Declaration
The copyright of this thesis rests with the author and is made available under a Creative
Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy,
distribute or transmit the thesis on the condition that they attribute it, that they do not use it
for commercial purposes and that they do not alter, transform or build upon it. For any reuse
or redistribution, researchers must make clear to others the licence terms of this work.
iii
iv
Abstract
It has been well documented that Darcy scale multiphase flow is significantly influenced by
the wettability of the fluid-solid system. So far this impact can only be determined through
costly and time intensive experiments. To predict observations at the core-scale, smaller scale
parameters influencing the results need to be determined and assessed.
In this work core-scale wettability responses were compared to nano-and pore- scale observations
by combining micro-computed tomography and atomic force microscopy.
Amott spontaneous imbibition tests were used to determine the core-scale wettability. A rela-
tionship was found between wettability and the capillary pressure applied during the sample
initialization with crude oil. Samples initialized with higher capillary pressure showed an oil-wet
behaviour, while samples initialized with lower capillary pressure appeared mixed-wet leaning
towards the water-wet side.
µCT waterflood experiments showed that ganglion dynamic behaviour, in which the connectiv-
ity of the oil phase changes via break-up and coalescence, is enhanced in mixed-wet systems.
The oil-filling events appeared larger and more frequent than in water-wet samples. Contact
angle distributions often used to assess wettability at this length scale were found to not reflect
the wettability of the sample correctly. The µCT images showed thin oil-layers, just above the
resolution of the image covering the grains of the rock, while all contact angles obtained from
the same images appeared water-wet.
These apparently contradictory results were targeted with a new approach based on atomic
force microscopy. Surface roughness measured with AFM was used for drainage simulation
assessing the surface coverage by the two fluids present in the rock. The surface coverage at
connate water-saturation varied from 1.5% - 50% depending on the capillary pressure applied.
These results were compared to in situ fluid film measurements on the nano-scale, which showed
a surface coverage of 60 % when the initialization protocol for mixed-wet sample was followed.
This small scale wettability pattern may lead to the different responses observed at larger scales.
The conducted experiments show how the different aspects of wettability across length scales
impact capillary pressure and relative permeability at the core-scale.
v
vi
List of publications and conference proceedings
Rücker M., Bartels W.-B, Berg, S. Mahani, H. Georgiadis, A., Brussee, N. Coorn, A, van der
Linde, H. Ott, H. Singh, K. Bonnin, A. Hassanizadeh, S. M., Blunt M., The effect of wettability
on pore scale flow regimes, submitted to Geophysical Research Letters.
Bartels W.-B., Rücker M., Boone. M, Bultreys, T., Mahani, H., Berg, S., Hassanizadeh, S.M.,
Cnudde, V., Fast imaging of wettability dependent pore scale displacement during spontaneous
imbibition, submitted to Water Resources Research.
Bartels, W.-B., Rücker, M., Berg, S., Mahani, H., Georgiadis, A., Fadili, A., Brussee, N.,
Coorn, A., van der Linde, H., Hinz C. (2017), Fast X-Ray micro-CT Study of the impact of
brine salinity on the pore-scale fluid distribution during waterflooding, Petrophysics, 58(01),
36-47.
Rücker, M., W.-B. Bartels, E. Unsal, S. Berg, N. Brussee, A. Coorn, and A. Bonnin (2017),
The formation of microemulsion at flow conditions in rock, SCA2017-041, paper presented at
International Symposium of the Society of Core Analysts, Vienna, Austria.
Bartels, W.-B., M. Rücker, M. Boone, T. Bultreys, H. Mahani, S. Berg, S. Hassanizadeh, and V.

Cnudde (2017), Pore-scale displacement during fast imaging of spontaneous imbibition, paper
presented at International Symposium of the Society of Core Analysts, Vienna, Austria.
Bartels, W.-B., M. Rücker, S. Berg, H. Mahani, A. Georgiadis, A. Fadili, N. Brussee, A. Coorn,

H. van der Linde, C. Hinz, A. Jacob, C. Wagner, S. Henkel, F. Enzmann, A. Bonnin, M.
Stampanoni, H. Ott, M. Blunt, S.M. Hassanizadeh (2016), Micro-CT study of the impact of
low salinity waterflooding on the pore-scale fluid distribution during flow, paper presented at
International Symposium of the Society of Core Analysts, Snowmass, Colorado, USA.
M. Rücker, W.-B. Bartels, S. Berg, H. Mahani, H. Ott, A. Georgiadis, N. Brussee, A. Coorn,

H. van der Linde, A. Fadili, C. Hinz, A. Jacob, C. Wagner, S. Henkel, F. Enzmann, A. Bonnin,
M. Stampanoni, M. Blunt, Impact of salinity on the pore scale distribution of crude oil in rock.
Interpore, Cincinnati, Ohio, USA (05/2016).
vii
M. Rücker, W.-B. Bartels, S. Berg, H. Mahani, H. Ott, A. Georgiadis, N. Brussee, A. Coorn,
H. Van Der Linde, C. Hinz, A. Jacob, C. Wagner, S. Henkel, F. Enzmann, A. Bonnin, M.
Stampanoni, S.M. Hassanizadeh, M. Blunt, Impact of wettability on two phase flow at the pore
scale. Goldschmidt, Yokohama, Japan (06/2016).
M. Rücker, W.-B. Bartels, S. Berg, H. Mahani, A. Georgiadis, H. Ott; N. Brussee, A. Coorn,

H. van der Linde, C. Hinz; A. Jacob, C. Wagner, F. Enzmann, S. Henkel, A. Bonnin, M.
Stampanoni, S. M. Hassanizadeh, M. Blunt, Impact of wettability at the pore scale flow regime.
Interpore, Rotterdam, The Netherlands (05/2017).
M. Rücker, W.-B. Bartels, M. Boone, T. Bultreys, H. Mahani, S. Berg, S.M. Hassanizadeh,

V. Cnudde, Pore-scale Processes in Amott Spontaneous Imbibition Tests. Workshop 4: Ap-
plications for Pore Scale Experimentation and Modeling for Multiphase Flow and Reactive
Transport, 79th EAGE Conference & Exhibition, Paris, France (05/2017).
M. Rücker, K. Singh, G. Garfi, A. Scanziani, S. Yesufu, W.-B. Bartels, M. Blunt, S.Berg, A.

Georgiadis, P. Luckham, Detection of fluid films and layers in porous rocks. Flow and Transport
in Permeable Media, Gordon Reseach Conference, Newry, Maine, USA (07/2018).
viii
Acknowledgements
The last three years, in which I was working on my PhD degree were extremely exciting for
me. Certainly, a prevailing reason for that were the great people I met, had exchange and
collaborations with. With the following pages I want to express my sincerest gratitude for the
great support and all the inspiring moments.
The person I would like to start with is Willem-Bart Bartels. I thank him very much for all
the discussions, motivational quotes and encouraging coffee breaks. It was a great pleasure to
share experimental experiences and to face all the challenges together.
Furthermore, I would like to acknowledge my supervisors, Apostolos Georgiadis, Steffen Berg,

Holger Ott, Paul Luckham and Martin Blunt. Especially, I would like to thank Apostolos
Georgiadis for his guidance, patience and support and for his help to find all the opportunities
along this journey and to turn them into the great adventure it became. I would like to
thank Steffen Berg for the tremendous amount of scientific conversation we had. Many things
which I have learnt about scientific research, I have learned from him. I thank Holger Ott for
initializing this project and the great guidance at the beginning. Furthermore, I want to thank
Paul Luckham for the warm welcome to Imperial and his guidance during my time in London.
I would like to thank Martin Blunt for his valuable feedback throughout my whole project. It
was a pleasure to work with him and his group.
I consider myself very fortunate for the opportunity to work with such a great team of su-
pervisors. Without their support, advice and encouragement this thesis would not have been
possible.
My sincerest gratitude also goes to the Shell PTI Rock and Fluids team and to Axel Makurat
in particular, for the opportunity to work in their laboratories in Rijswijk. Furthermore, I want
to express my gratitude to Niels Brussee, Ab Coorn, Hilbert van der Linde, Fons Marcelis,
Patricia van den Bos, Hassan Mahani and Ove Wilson for sharing their knowledge and for their
support in planning and/or executing the experiments presented in this thesis. In addition,
I would like to thank Marjan Batenburg for her great administrative support. The time in
Rijswijk was a great experience.
I would also like to thank Evren Unsal, Ivana Sersic and Ali Fadili from Shell PTI RO for the
ix
helpful discussions. Furthermore, I want to gratefully acknowledge the experimental service
team in Shell. In particular Alex Schwing, Rob Neiteler, Maria Arias and Rien Groenewegen for
their help in designing the experimental set-ups. I would also like to thank Frans Korndorffer,
Jos Pureveen, Willie Schermer and Len van Gelder for characterising crude-oils.
I acknowledge the funding of this project provided by Shell Global Solutions International B.V..
Furthermore, I want to express my greatest gratitude to the Digital Rock team and others at
Imperial College London. The discussions we had helped me a lot to gain a better understanding
of wettability across the length scales. In particular, I would like to thank Kamalijt Singh,
Hayley Meek, Sheri Yesufu and Matthew Leivers for their support in conducting the AFM
fluid film measurements in rock and for their patience, trust and resilience. I would like to
thank Gaetano Garfi and Alessio Scanziani for their help in image analysis and Lingru Zheng,
Takashi Akai, Tom Bultreys and Ali Q Raeini for their input from a modelling perspective.
Furthermore, I want to thank Paul Luckham’s lab team for the great working environment.
Special thanks also to Bhavna Patel, for her great support and for organizing all the student
activities.
Furthermore, I want to thank all the external partners for their support. I acknowledge the
Paul Scherrer Institute, Villigen, Switzerland for provision of synchrotron radiation beamtime
at Tomcat beamline of the SLS, our beamline scientist Anne Bonnin and Christian Hinz, Arne
Jacob and Frieder Enzmann from Johannes-Gutenberg University of Mainz, Christian Wagner
from Math2Market and Steven Flesch from Friedrich-Schiller University Jena for their assis-
tance during the beamline. I would like to thank Hannes Schniepp from the College of William
and Mary, Williamsburg, Steve Higgins and Jyothirmayee Aravind from Wright State Univer-
sity, Naveen Kumar and Marcel Bus from TU Delft and Alex Winkel from JPK instruments
for sharing their knowledge on AFM. Special thanks also to Marijn Boone from XRE and Tom
Bultreys for their help in conducting flow experiments at UGCT and to UGCT for providing
the facilities.
I would like to thank the Interpore society for their support in forming the Interpore student
section, Leon Leu and Willem-Bart Bartels for the great time in the student affairs committee
and Marijn Boone, Matthew Andrew and Tom Bultreys for the joint organization of the short
course on µCT. I learned a lot from it.
x
Furthermore, I would like to thank Martin Fernø, Ryan Armstrong, Steffen Schlüter and James
McClure for the helpful discussions at conferences and per e-mail.
I am grateful to my examiners Tannez Pak and Branko Bijeljic for reviewing this work and the
comments they provided.
I want to thank my family for all the inspiration throughout my lifetime, my parents for giving
guidance whenever needed and for the encouragement to find my own path, my brother Marius
Bajorski and sister Maria Bajorski, which always inspired me with their own curiosity, and most
important, my husband Kai Rücker, who always supported me in my aims and encouraged me
to follow my dreams.
Special thanks also to Sabine and Stefan Röhm for all the advice they gave me. I want to thank
Alejandra Castro for the great time in the Netherlands and Minou Foadi for the great time in
UK. Furthermore, I would like to thank my uncle Martin Brandt and aunt Alexandra Bajorski-
Brandt for their advice and all my friends, especially Katharina Schmidt, Lisa Both, Melanie
Lewalter, Katja Grube, Benedikt Hahn and Georg Schorn for their support and encouragement.
xi
xii
‘There is no security in this world. There is only opportunity.’
General Douglas MacArthur
xiii
xiv
Contents
Declaration of originality i
Copyright declaration iii
Abstract v
List of publications and conference proceedings vii
Acknowledgements ix
Nomenclature xxxv
1 Introduction 1
2 Theoretical background 5
2.1 Wettability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.1 Molecular interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.2 Sub-pore scale surface structure . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.3 Pore scale fluid distribution and flow behaviour . . . . . . . . . . . . . . 16
2.1.4 Core scale capillary-pressure and relative-permeability . . . . . . . . . . . 23
2.1.5 Connecting the length scales . . . . . . . . . . . . . . . . . . . . . . . . . 30
xv
2.2 Enhanced oil recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2.1 Low salinity flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2.2 Surfactant flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3 Materials and methods 39
3.1 Fluid and rock systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.2 Micro computed tomography (µCT) . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.2.1 Equipment and facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.2.2 µCT core-flooding experiments . . . . . . . . . . . . . . . . . . . . . . . 51
3.2.3 Spontaneous (Amott) imbibition . . . . . . . . . . . . . . . . . . . . . . 59
3.3 Atomic force microscopy (AFM) . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.1 Force spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.3.2 Topographical imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.3.3 Wettability assessment by AFM . . . . . . . . . . . . . . . . . . . . . . . 70
3.4 Supporting data and experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.4.1 Core flood experiment (S1) for comparison with pore-scale flooding ex-
periment (1A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.4.2 Data: Flooding experiment water-wet sample (S2) for comparison with
mixed-wet systems (1A) . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.4.3 Static tube test (S3) conducted for surfactant flooding assessment (3A) . 80
4 Core-scale wettability characterization 81
4.1 Amott spontaneous imbibition tests . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.1.1 Water-wet systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
xvi
4.1.2 Mixed-wet systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.1.3 Impact of sample handling protocols . . . . . . . . . . . . . . . . . . . . 86
4.2 Steady-state experiments in mixed-wet systems . . . . . . . . . . . . . . . . . . 87
4.2.1 Relative permeability at the core scale . . . . . . . . . . . . . . . . . . . 88
4.2.2 Impact of emulsion on the oil-phase connectivity . . . . . . . . . . . . . . 89
4.2.3 Comparison of core-scale relative permeability with pore-scale data . . . 91
4.3 Impact of sample initialization on wettability . . . . . . . . . . . . . . . . . . . . 95
4.4 Summary: Dependency of core-scale parameters on sample handling protocols . 97
5 Impact of wettability on pore-scale flow regimes 99
5.1 Wettability classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2 Pore scale processes during waterflooding . . . . . . . . . . . . . . . . . . . . . . 104
5.2.1 Physics of oil filling events in water-wet systems . . . . . . . . . . . . . . 104
5.2.2 Physics of oil filling events in mixed-wet systems . . . . . . . . . . . . . . 106
5.3 Summary: Enhanced ganglion dynamics in mixed-wet systems . . . . . . . . . . 109
6 Enhanced oil recovery at the pore scale 111
6.1 Low-salinity flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.2 Surfactant flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6.3 Summary: Changes of oil configuration during enhanced oil recovery . . . . . . . 119
7 Impact of nano-scale structures on wettability 121
7.1 Detection of fluid films . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.2 Impact of surface roughness on wettability . . . . . . . . . . . . . . . . . . . . . 124
7.3 Summary: Relationship between surface roughness and wettability . . . . . . . . 127
xvii
8 Conclusions 129
9 Outlook 133
Literature 137
Appendix A 167
Appendix B 170
Appendix C 172
Appendix D 175
Appendix E 179
Appendix F 185
xviii
List of Tables
3.1 Overview of the different experiments conducted in this work. A more detailed
list including the rock-fluid-fluid systems used can be found in corresponding
Sections 3.3 and 3.4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.2 List of rocks used in this study. . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.3 List of minerals used in this work. . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.4 TAN, TBN, asphaltene content as well as viscosity and density of the crude oils
used in this work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.5 SARA fractions for the crude oils used in this work displayed in the relative
weight of each compound class. . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.6 Composition of formation water (FW1+2), seawater (SW), doped high salinity
(HS) and low salinity brine (LS) and optimum doped brine (DB) used in this
work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.7 List of flooding experiments conducted during this work. . . . . . . . . . . . . . 54
3.8 List of spontaneous imbibition experiments performed in this study. . . . . . . . 60
3.9 List of AFM experiments performed in this study. . . . . . . . . . . . . . . . . . 71
A1 List of flooding experiments conducted during this work. . . . . . . . . . . . . . 168
A2 List of flooding experiments in tertiary mode conducted during this work. . . . . 169
xix
xx
List of Figures
2.1 Wettability depends on various properties such as brine- and crude-oil compo-
sition, reservoir pressure and temperature (P -T ) conditions, surface chemistry
associated with mineralogy, surface structure and fluid-film formation as well as
saturation history. All of these influence wettability and hence the pore-dynamics
during two-phase flow and the relative permeability. . . . . . . . . . . . . . . . . 6
2.2 The description of wettability is length scale dependent. At the molecular scale
wettability is defined as adhesion, which depends on crude oil, brine and mineral
composition as well as P -T . Based on the molecular interactions an intrinsic
contact angle can be computed which considering surface roughness, chemical
heterogeneity and fluid film formation can be translated into an effective contact
angle. Pore architecture, saturation history and mineral distribution are addi-
tional parameters which impact wettability at the core scale, which is character-
ized through capillary-pressure and relative-permeability/saturation functions.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 (a) shows an example of a force distance curve, which is the sum of all forces
acting between two molecules at different distances. (b) illustrates the different
forces acting on molecules when approaching a surface (adapted from Raina, 2013). 8
2.4 A system is called water-wet for a contact angle θi 90 (a), intermediate-wet
for θi 90 (b) and oil-wet for θi ¡ 90 (c), when the contact angle is measured
through the denser brine phase. . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
xxi
2.5 A droplet of oil on a water-wet surfaces forms a characteristic contact angle. On
rough surfaces the measured apparent contact angle θ , for which the surface is
considered smooth as indicated by the red dotted line (a), may deviate from the
sub-resolution intrinsic contact angle θi (b). Furthermore, surface roughness may
lead to pinning: (c) and (d) show two possible menisci configurations resulting
from the same contact angle. At this location all contact angles between the two
configurations may form (e). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.6 In dynamic conditions a droplet on the surface will form an advancing and reced-
ing contact angle (a). the observed contact angle hysteresis for a specific texture
of the surface is depending on the intrinsic contact angle (b, schematic drawing
based on Morrow, 1975). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.7 If the contact area of the drop and the solid is smaller than the oil-wet patch the
contact angle of the drop will reflect the contact angle of the oil-wet surface θow
(a). The opposite will happen if the drop is exceeding the oil-wet patch. The
contact angle will reflect the wettability of the water-wet patch θww (b). When
the contact line is pinned at the transition between these patches any contact
angle between θow and θww may form (c) (adapted from Israelachvili, 2011b). . . 15
2.8 Capillary rise experiment in which a tube of radius r is placed in a beaker filled
with one fluid. The capillary rise h can be derived from the Young-Laplace
equation (Eq. 2.4). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.9 Receding meniscus during drainage into a pore. The highest capillary pressure
occurs at pore throats. Once the meniscus passes through the invading phase
will rapidly fill the consecutive pore body. The receding meniscus will thereby
move along the surface with the receding contact angle θr (adapted from Blunt,
2017). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.10 An oil ganglion in a rock gets mobilized when viscous forces acting along the
cluster length L overcome capillary forces (adapted from Stegemeier, 1977). . . . 18
xxii
2.11 Different flow regimes may occur during two-phase flow in porous media. Dur-
ing connected pathway flow the two phases move solely through the pores the
respective phase occupies (a). In a ganglion dynamics flow regime the pores
experience a change in occupancy and distinct ganglia move through break-up
and coalescence events (b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.12 Two-phase flow in water-wet systems may lead to different event types, oil-filling
events such as Haines jumps (a) and coalescence (b) and the water-filling events
such as break up (c) and piston-like displacement (d). Examples for each event
type are shown before (top) and after the event (bottom), respectively (adapted
from Rücker et al., 2015b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.13 Darcy’s experimental set-up for single phase flow in porous media allows perme-
ability determination (based on Darcy, 1856). . . . . . . . . . . . . . . . . . . . 24
2.14 An unsteady state flow experiment can be described by the Buckley-Leverett

equation (Eq. 2.14). However, capillary effects lead to a dispersion of the shock
front (adapted from Lake, 1989). . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.15 Capillary-pressure/saturation functions for oil- and water invasion cycles. (a)
The wettability of the system determines whether the oil can get produced spon-
taneously (water-wet) or only through forced water injection (oil-wet). Corre-
spondingly, capillary-pressure/saturation functions can be used for wettability
characterizations. The Amott index thereby refers to the spontaneous part of
the curve, while the USBM index is based on the forced parts (b). . . . . . . . . 27
2.16 Schematic drawings of relative permeability-saturation functions for a water-wet

(a), oil-wet (b) and mixed-wet scenario (c) based on examples and descriptions
from Blunt (2017). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.17 The heterogeneity of rock at different scales and the response of permeability
(adapted from Nordahl & Ringrose, 2008). . . . . . . . . . . . . . . . . . . . . . 30
2.18 Enhanced oil recovery methods are often based on molecular processes. How-
ever, the acceleration in oil-production at the core- and field scale may also be
impacted on length scales in between. . . . . . . . . . . . . . . . . . . . . . . . . 32
xxiii
2.19 Depending on the microscopic capillary number Nc, the amount of trapped wet-
ting and non-wetting fluid can be higher or lower (adapted from Lake, 1989). . . 32
3.1 3D illustration of Ketton limestone (a). The round grains build a simple macro-
scopic pore structure which is easy to interpret with µCT. However, as shown in
the MIP curves (b) the grains themselves contain additional microporosity. . . . 41
3.2 µCT image of an Estaillades limestone (a). Consisting of different recrystallized

fossils the rock shows a complex pore-structure and a dual porosity as shown
in the MIP-curve (b). Therefore, image segmentation and processing are more
difficult than for Ketton and the sandstones studied. The rock consists predomi-
nantly of calcite (97.9%) with minor quartz (2.1%) components (Andrew, 2015).
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3 The µCT image of the Gildehauser sandstone illustrates the pore-structure of
the rock (a). The MIP-curve is homogenous with a peak pore diameter of 33.45
µm (b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.4 3D illustration (a) and MIP-curve (b) of a Berea sandstone. . . . . . . . . . . . 43
3.5 Topographical AFM images of a calcite (a) and a biotite mica sample (b). . . . 45
3.6 Basic illustration of a microtomography apparatus with a parallel beam source

and detector monitoring a rotating sample (adapted from Wildenschild et al.,
2004). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.7 Schematic drawing of the Paul-Scherrer Institute Swiss Light Source (PSI SLS).
synchrotron. Electrons are first accelerated in a linear accelerator, boosted within
the booster ring before they are directed into the storage ring where they main-
tain in orbit. Bending wigglers and magnets deflect the electron trajectories and
generate brilliant bremsstrahlung, which can be used at different beamlines, such
as TOMCAT for e.g. fast µCT experiments (schematic drawing adapted from
”SLS Beamlines”, 2018) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
xxiv
3.8 µCT is not capable of visualizing macroscopic phenomena such as a saturation
front (a) however, it is capable to resolve pore-scale phenomena (b). For flooding
experiments, the sample is mounted on top of a flow cell (c). . . . . . . . . . . 52
3.9 First wet and dry images were filtered. In a second step the wet image was
registered to the dry image. The segmented images were then combined to
reflect the 3 phase system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.10 In an Amott test an oil saturated core sample is placed into a vessel filled with
brine. The oil produced to spontaneous imbibition accumulates at the top of the
sample (a; adapted from Bartels et al., 2017c) . . . . . . . . . . . . . . . . . . . 59
3.11 Experimental protocols for core-scale and pore-scale spontaneous imbibition tests.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.12 Force distance curve as obtained during force spectroscopy. Initially the can-
tilever is in a relaxed shape (1) until surface forces of the scanned solid start to
attract or repel the tip. In the presented case the tip is attracted to the surface
and therefore jumps into contact (2). When the cantilever is moved closer to the
surface the cantilever is pushed back into the relaxed shape (3). This point is
used as a height reference. The tip is then pushed further into the sample until a
chosen force (set point) is detected (4) and the cantilever starts to retract. The
tip experienced adhesion forces and gets released, once the spring forces of the
cantilever exceed the adhesion (5). . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.13 Force distance curves obtained with an atomic force microscope based on the
calibrated deflection sensitivity and spring constant (a) do not reflect actual
force distance curves of approaching molecules (b). The approach curve gives us
the force vs. distance for a chosen system up to the snap-in point. . . . . . . . 67
3.14 In QITM mode the tip does not move in x-y direction while a force distance curve
is obtained. Yet, the movement algorithm allows an overshoot above the set z-
distance, which enables a fast data acquisition (adapted from , ”QITM mode -
Quantitative imaging with the NanoWizard 3 AFM” (2018)). . . . . . . . . . . 69
xxv
3.15 Processing of force distance curves. First the zero point was found by determining
the zero level within the no-contact distance (a). In a second step the deflection
d [V] was converted to the deflection d [nm] with the help of the deflection
sensitivity ds [nm/V] (b). In the third step, the true distance r was calculated
by subtracting the deflection d from the measured distance z (c). In a last step
the force was computed by multiplying the deflection d with the spring constant
k (d). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.16 3D fluid film visualization workflow. . . . . . . . . . . . . . . . . . . . . . . . . 78
4.1 In Amott tests, an oil-saturated core-plug is placed into a brine filled vessel. If
the brine spontaneously imbibes into the rock, the oil is produced. Image (a)
shows the formation of an n-decane-bubble during experiment 5A (water-wet
system). (b) shows the formation of an oil-bubble during experiment 5B. The
cumulative production vs. time curves for all SCAL-plug Amott experiments are
shown in (c; adapted from Bartels et al., 2017c). . . . . . . . . . . . . . . . . . . 82
4.2 Spontaneous imbibition within a water-wet rock (experiment 5A). The difference
between the initial image (a) and the last image of the scanning sequence (b)
shows an asymmetric imbibition front. The differential image (c) highlights the
locations at which changes occurred. This asymmetric behaviour may be a result
of preferential production sites, at which n-decane drops repeatedly emerged at
the same location (d; adapted from Bartels et al., 2017c). . . . . . . . . . . . . . 83
4.3 Spontaneous imbibition in mixed-wet rock (experiment 5B). The difference be-
tween the initial (a) and the final (b) image of the scanning sequence indicates
only small changes in fluid distribution. (c) illustrates the water-filling events
within the rock in 3D (adapted from Bartels et al., 2017c). . . . . . . . . . . . . 85
4.4 The final scan of the whole sample (experiment 5B) (a) shows a homogeneous
distribution of the oil phase. The initial (b) and final (c) fluid distribution within
the field of view monitored throughout the experiment illustrates that 1% of the
oil-phase was produced from the inside of the rock (adapted from Bartels et al.,
2017c). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
xxvi
4.5 Relative permeability curves obtained from a steady-state SCAL experiment (S1)
displayed on an arithmetic (a) and a semi-logarithmic scale (b). The results are
typical for a mixed-wet system leaning towards water-wet. . . . . . . . . . . . . 88
4.6 In the grey scale images obtained by fast µCT we see the rock (grey spheres),
crude oil (black) and brine (bright phase). Furthermore, a fourth phase with an
intermediate grey value appears in the pore space, which is an emulsion (adapted
from Bartels et al., 2017b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.7 Connectivity indicates the possibility of connected pathway flow. The distri-
bution of the connected (red) and disconnected (yellow) oil phase is displayed
for both, the oil phase only (a) and the oil phase including the emulsion (b).
In (c) the overall saturation vs. time is displayed as well as the proportion of
connected oil, oil and emulsion and total oil. The oil phase only gets discon-
nected and reconnected with time independent of the behaviour of the overall
saturation, which indicates either a ganglion dynamic behaviour or connected
pathway flow through the emulsion phase. The Euler characteristic, a measure
of connectivity, displayed in (d) shows a highly disconnected oil phase. However,
with the emulsion phase the Euler characteristic shows a hysteretic behaviour as
known from water wet systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.8 The relative permeability curves of µCT-experiment 1A were obtained by single-

phase flow simulation through the space occupied by the water phase and the
oil and emulsion phase, respectively. The graphs show the original data ob-
tained from this simulation with the absolute (a,b) and mobile saturation (c,d)
in comparison with results of SCAL experiments. . . . . . . . . . . . . . . . . . 92
4.9 The relative permeability curves of µCT-experiment 1A after emulsion correction

shown with the absolute (a,b) and mobile saturation (c,d) in comparison with
results of SCAL experiments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
xxvii
4.10 Droplet formation during Amott test. During experiment 5B the oil drop spreads
over the rock surface indicating an oil-wet behaviour (a). In experiment 6B the
droplets emerging during spontaneous imbibition appeared water-wet (b). In
both experiments the same COBR system was used. The difference between
these two experiments is (beside size) the initialization protocol. Experiment 5B
was saturated with oil by centrifugation and experiment 6B was initialized by
flooding (adapted from Bartels et al., 2017c). . . . . . . . . . . . . . . . . . . . . 96
4.11 MPI curve of Ketton rock. Based on the oil-saturations obtained from the ex-
periments, initialization by flooding filled pores down to a diameter of 20 µm
and centrifugation down to 0.3 µm (adapted from Bartels et al., 2017c). . . . . . 97
5.1 Visual assessment of wettability for experiments 1A, which was initialized by
flooding (a,b) and 4B, which was initialized with centrifugation (c,d). At some
locations with presence of oil in the initial state residual oil films were detected
after flooding (b,d). This observation is consistent with the contact angle distri-
bution for experiment 4B. However, the contact angle distribution in experiment
1A (100 contact angles) indicates a water-wet state even though more oil-wet
than the water-wet reference obtained from Scanziani et al. (2017). Due to the
low number of contact angles obtained for experiment 4B (21 contact angles)
solely the median value, the maximum and minimum are plotted. . . . . . . . . 102
5.2 Fluid distribution for water-wet experiment S2 (a) and mixed-wet experiment
1A (b) sample at around 40% oil-saturation. . . . . . . . . . . . . . . . . . . . 103
5.3 (a) Number of oil- and water filling events vs. time detected in the unsteady
state waterflood experiment in a water-wet Ketton sample (supporting data S2;
Singh et al., 2017a). The event-size distribution in comparison with the pore-
size distribution of the rock reveals that water-filling events cover up to multiple
pore-bodies, while no oil-filling event larger than 106 µm3 was detected (b).
(c-e) show few examples of the oil-filling events (blue) next to the oil-phase not
affected by changes (red). Oil-filling events may cause reconnection of previously
disconnected clusters (f; adapted from Rücker et al., 2015a). . . . . . . . . . . . 105
xxviii
5.4 The oil-filling events observed in this work can cover full pore bodies (a) The oil
cluster (red) moves with time more and more into the pore body until it connects
with the surrounding oil phase. These oil-filling events fill a pore previously filled
with brine (b). The displacement mechanisms taking place within the mixed-wet
system of experiment 1A is illustrated schematically in (c). . . . . . . . . . . . . 107
5.5 The frequency of the three event types, water-filling events, oil-filling and emulsion-
filling events for each time step (a) shows that the oil-and emulsion redistribution
is of significance in mixed-wet cases. The pore-size distribution and the initial
fluid distribution (b) illustrate the initial situation before events occur and indi-
cate the impact of the wetting state of the surface on the flow behaviour. Much
more oil-filling events are detected on surfaces initially covered with oil compared
to surfaces covered with emulsion or water (c). . . . . . . . . . . . . . . . . . . 108
6.1 µCT grey value images obtained after a waterflood (a) and at the beginning (b)
and end of LS flood (c). During the LS-flood experiment (2A) the change in
salinity within the pore-space was tracked. Pore-geometry and oil-configuration
affect the mixing and dilution process of the HS-brine. As the LS-brine ap-
peared with the same grey value as the emulsion, these two phases could not be
differentiated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.2 Final distribution of the fluids after the waterflood experiment 1A in Ketton
(a) and experiment 1E in Berea (b) and after the associated low salinity flood
experiments 2A and 2E (c, d), as well as a comparison of the final distribution
of oil (e, f; adapted from Bartels et al., 2017b). . . . . . . . . . . . . . . . . . . 114
6.3 Tube tests, conducted to determine the optimum formula for surfactant flooding
revealed, that the grey values of the emulsion phase measured with µCT reflect its
oil- and water- content, when the oleic phase is doped with iodo-decane (adapted
from Rücker et al., 2017). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
xxix
6.4 Grey scale image of the rock without any fluid (a), final scan of the waterflood
performed prior to the surfactant flood in experiment 3A (b) as well as first (c)
and final(d) image of the surfactant-flood itself. Subtle changes in the grey level
can be observed when comparing images (c) and (d)(adapted from Rücker et al.,
2017). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
6.5 The mean grey value of the fluids within the pore-space correlates linearly with
saturation (a), which allows to estimate the oil content of the whole sample, once
emulsification takes place (b). The oil-content within the sample drops sharply
with the start of the surfactant injection resulting in an additional production
of 55% of the remaining oil (adapted from Rücker et al., 2017). . . . . . . . . . 117
6.6 Last scan of the water to surfactant flooding of oil (red) and water (blue) within
the pore space prior to surfactant flooding (a). The first image obtained 7 s after
the surfactant-flood started showed a change in grey values indicating immediate
emulsification within pores initially filled with oil and pores initially filled with
water. In some pores the oil was immediately produced (b). (c) shows further
production of oil via solubilisation. The pore-scale processes in rock confirm
observations in micromodels from Unsal et al. (2016) and Broens and Unsal
(2018) indicating a faster emulsification under dynamic conditions than expected
from phase-behaviour studies (adapted from Rücker et al., 2017). . . . . . . . . 118
7.1 Representative force distance curves for measurements in brine (a- experiment
9B) and oil (b- experiment 9C). . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.2 Adhesion map of experiment 10A, in which a calcite surface was first placed in
brine and then submerged in decane (a). The image shows round patches with
high adhesion, which gets larger towards its centre. This is also reflected by
the three force distance curves taken from this image (c – e). In addition, a
snap-in can be detected during the approach of the tip. This snap-in marks the
transition of the tip through the fluid-fluid interface and can be used for a 3D
illustration (b) of the wetting film (blue) on top of the sample surface (grey). A
cross-section through the droplet can be used for contact angle measurements (f).123
xxx
7.3 Topographical images of Ketton (a,b,c) and Estaillades (d,e,f) rock taken from
within pores. The optical images (a,d) show the scanning location for each rock
respectively, (b), (e) the corresponding topographical images, and (c), (f) the
associated roughness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.4 Water films (blue) on top of a rock surface (grey) obtained from a drainage
simulation based on sphere fitting (a: sphere diameter 1.6 µm, b: sphere diameter
5.6 µm, c: sphere diameter 6.4 µm). The rock surface is in direct contact with
oil where no water is visible. d shows the percentage of surface area covered with
brine as a function of sphere diameter. . . . . . . . . . . . . . . . . . . . . . . . 126
7.5 (a) shows the water film (blue) on top of a rock surface (grey) measured in n-
decane (experiment 11A). The water layer shows a film thickness ranging between
0 nm and 200 nm with a peak film thickness of 125 nm (b). . . . . . . . . . . . 127
9.1 Workflow for upscaling wettability from the molecular scale to the core scale. . . 134
B1 The contrast agent within oil was found to cause photochemical reactions like
bubble formation (a) and disintegration (b) when exposed to the high intense
light from a beamline (adapted from Bartels et al., 2017b). The images were
taken in 10s intervals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
B2 a shows the image of the muscovite surface at a tip velocity of 120 µm/s and
z-length of 1000 nm. The noise level is 800 pm, which is 4x larger than the
molecular structure of muscovite. b shows the noise level for different tip veloc-
ities and z-lengths. Only with a tip velocity of 2.5 µm/s atomic resolution may
be achieved. A higher z-length results also in a higher noise level. . . . . . . . . 172
B3 Rough surfaces may lead to image artefacts like stripes (a). Furthermore, surface
structures as shown in the height map (b) may impact adhesion forces obtained.
Due to the larger contact area of the tip the adhesion is higher in dents than on
top of hills. The adhesion map corresponding to (b) is shown in (c). . . . . . . . 173
xxxi
B4 3D reconstruction of fluid films measured with AFM (a) can be strongly affected
by noise. The reason is the sensitivity of the automatic fluid / fluid-interface
detection, which is based on a sudden deflection when approaching the surface.
Such deflection may also be caused by noise or vibrations. b shows a force dis-
tance curve affected by noise. Such image artefacts can be reduced by decreasing
the sensitivity, which also decreases the resolution of the fluid / fluid interface
or by applying a median filter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
B5 Next to brine and oil a third phase, emulsion, was detected in all studied COBR
systems with HS as brine. (a) shows Estaillades rock with crude A, (b) Estail-
lades with crude B, (c) Berea sandstone with crude C and (d) Ketton rock with
crude A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
B6 Test tubes before (a) and after (b) rigorous shaking show the ability of different
fluid-fluid system to form emulsion. The test tube containing crude B and LS
brine was investigated further by µCT (c), which revealed a water-in-oil emul-
sion. The water droplets observed range from a size of 500 µm down to image
resolution (adapted from Bartels et al., 2017b). . . . . . . . . . . . . . . . . . . 176
B7 The histogram obtained from an image of experiment 1A shows that the Ketton
rock separates brine from crude oil. Partial volume effects, which may be related
to the presence of emulsion spread in between brine and oil. . . . . . . . . . . . 177
B8 Grey-scale images obtained with in 1 min intervals during experiment 1A. The
red circles highlight an emulsion filling event replacing brine. The emulsion
appears at the same location in consecutive images and the grey value of the
emulsion remains constant. The blue circles highlight a brine filling event: Due
to a partial volume effect the grey value of the brine appears lower during the
event. In the consecutive image the partial volume effect disappears. . . . . . . . 177
B9 To ensure absence of emulsion in the examples displayed in Figure 5.4 the 3D

grey scale images for all times steps were assessed within the FOV. . . . . . . . 178
xxxii
B10 Time sequence of experiment 1A showing the fluid distribution of the oil- and
emulsion phase. The yellow phase represents oil and emulsion disconnected to
the outlet and the red phase connected oil and emulsion. Circles highlight re-
connection events associated with ganglion dynamics. . . . . . . . . . . . . . . . 181
B11 Time sequence of experiment 1A showing the fluid distribution of the emulsion
phase only. Circles highlight the change in fluid distribution. . . . . . . . . . . . 183
B12 Time sequence showing the dilution of the high salinity brine (HS; dark blue-
green) during displacement by LS brine (experiment 2F). . . . . . . . . . . . . . 184
B13 The adhesion map of a Ketton surface at residual oil-saturation reveals a crude
oil layer covering the full 4.5 µm x 4.5 µm image. The force distance curves
obtained during the measurement show a repelling behaviour of the tip to the
fluid-fluid interface. Once the tip passes through the interface it gets further
deflected before coming into contact with the surface (b). In some cases, the
deflection gets attractive once the fluid/fluid interface is transgressed (c). The
cause of this behaviour is unknown. . . . . . . . . . . . . . . . . . . . . . . . . . 186
B14 The deflection measured after the tip passed through the fluid/fluid interface
could be caused by meniscus forces. A negative deflection would occur for a
receding contact angle (a) and a positive for an advancing contact angle (b).
Advancing and receding contact angles are depending on the surface roughness
of the tip. c shows the movement of a droplet over a rough surface and the
relationship of the advancing/receding contact angle to flow direction. . . . . . . 187
xxxiii
xxxiv
Nomenclature
A area
b width
d deflection
e exponential
f fluid fraction
F Force
g acceleration due to gravity
h height
I intensity
IU SBM {Amott wettability index
k constant
kb Boltzman constant
kr relative permeability
K permeability
L length
Nc capillary number
P pressure
Pc capillary pressure
q Darcy velocity
Q flow rate
Qf quality factor
r radius
xxxv
R roughness
S saturation
t time
T temperature
V volume
W energy
β Betti number
γ fluid content
θ angle/ contact angle
µ viscosity
µa X-ray linear attenuation coefficient
Π disjoining pressure
ρ density
σ interfacial tension
φ porosity
χ Euler characteristic
ωf resonant frequency
ø diameter
AFM atomic force microscopy

cEOR chemical enhanced oil recovery
COBR crude oil/ brine/ rock system
DLVO Derjaguin-Landau-Verwey-Overbeek theory
EDL electrical double layer
EOR enhanced oil recovery
FOV field of view
HS high salinity brine
LS low salinity brine
PSD pore size distribution
PV pore volume
xxxvi
REV representative elementary volume
SARA saturates/ aromates/ resins/ asphaltenes
SCAL special core analysis
TAN total acid number
TBN total basic number/ total basic nitrogen
µCT micro-computed tomography
xxxvii
xxxviii
Chapter 1
Introduction
The origin of wettability, the preference of a fluid to be in contact with a solid in presence of
another immiscible fluid, are the molecular interactions between these three phases. However,
wettability of a specific fluid/fluid/solid system can be manipulated at the larger scale e.g. by
the surface structure. A rough surface leads to entrapment of the wetting phase in dents and
hence to a smaller contact area of the non-wetting phase with the solid causing a more non-
wetting behaviour. This effect, commonly known as the ”Lotus” effect, is applied in various
features we use in our day to day life as the superhydrophobic coating of our car, or our jackets.
The concept of wettability gets even more complex when we consider porous material. The
large surface area of such media enhances solid/fluid interactions. Capillary effects start to play
a role. Our jackets remain water resistant, only until the water column exceeds the capillary
pressure preventing the water to penetrate our clothes. Once the water breaks through, capillary
pressure may cause entrapment of the fluid and our clothes get heavy. Similarly, wettability
influences the flow behaviour of the oil and brine within a reservoir. The crude oil, brine and
mineral composition determine molecular interactions, but also surface texture, occurring at the
nano- and micrometer scale and mineral distributions and pore-geometry of the rock at the pore-
scale may impact flow behaviour. In order to assess uncertainties in oil-recovery scenarios, the
flow behaviour within the rock is characterized by capillary-pressure and relative-permeability
saturation functions. However, so far these functions can only be determined through time
1
2
intense and costly core flooding experiments. A better understanding of wettability is crucial
for computational models aiming to predict these functions.
But, what defines wettability within this context? Which parameters need to be considered
and how do they relate to the macroscopic flow behaviour? This is the scope of this work.
Objective
The definition of wettability is length scale dependent. On the molecular scale wettability
depends solely on molecular interactions, while at a larger scale other parameters as the rock
structure need to be considered. This leads to the main question of this work, which is the
identification of wettability parameters and length scales that determine core-scale character-
istics.
Computational simulations, which aim to predict core-scale capillary-pressure and relative-

permeability saturation functions, often use pore-scale parameters such as contact angle mea-
surements as input. These simulations have been proven effective for model systems, but fail
to predict the impact of wettability in realistic scenarios. In this work, the relationship be-
tween core- and pore-scale processes is assessed for different wettability stages to bridge those
two length scales. Furthermore, the impact of pore-scale wettability parameters as contact
angle measurements on fluid propagation is evaluated. The question how pore-scale wettability
parameters depend on sub-pore scale features as surface structure is addressed as well.
Wettability is not the only property originating from the molecular scale that impacts core-scale
processes of petroleum engineering applications. In so called enhanced oil-recovery techniques
fluids replacing the oil in the ground are manipulated to increase oil production. The deter-
mination of fluids optimal for these applications is often based on nano-scale concepts. Low
salinity flooding is one example. This technique is based on the change of wettability of the
system. Surfactant flooding is another. Here the interaction between the two fluids, brine and
oil is changed.
Correspondingly, another question addressed in this work is how changes in wettability or

3
fluid/fluid-interactions impact the pore-scale flow regime in rock.
The focus of this work lies on pore- and nano-scale aspects of wettability and enhanced oil-
recovery concepts. The pore scale refers to the range from µm to mm, where the pore structure
of a rock and the flow behaviour of the two fluids within are identified and analysed. The nano
scale represents the range just below the pore scale, also referred to as sub-pore scale. At this
scale the focus lies on the rock surface and interactions with the fluids in contact.
Thesis layout
In Chapter 2 the current understanding of wettability will be discussed for all length scales.
Starting from the general concepts of wettability, specifics in more detail are then presented
with respect to petroleum applications.
Micro computed tomography (µCT) and atomic force microscopy (AFM) were used to study
wettability and enhanced oil recovery at different length scales. The applied techniques are
discussed in detail in Chapter 3. To ensure consistency across the length scales the different
experiments were conducted with the same set of fluids and rocks, which are also introduced
in this chapter.
In Chapter 4 the core-scale aspects of the obtained results are discussed. The main objective was
to identify and characterize the response obtained in Amott spontaneous imbibition and core-
flood tests for the chosen rock/brine/oil systems. The relationship of pore-scale flow behaviour
and wettability assessed through flooding experiments monitored with fast µCT is discussed in
Chapter 5. In Chapter 6 pore-scale aspects of enhanced oil recovery techniques are presented.
The impact of nano-scale structures within rock measured with AFM on the macroscale wet-
tability observed at the pore scale is covered in Chapter 7. The relationship between sub-pore
scale features, pore-scale flow behaviour and core-scale observations obtained from the chosen
systems are the subject of Chapter 8. How these findings can be used and what else is needed to
be able to predict core-scale relative permeability and capillary-pressure/saturation functions
is covered in Chapter 9.
4
Chapter 2
Theoretical background
To assess the impact of uncertainties on a reservoir scenario and to optimize this scenario to
be robust against these uncertainties, commonly reservoir models are constructed. For this
information on relative permeability and capillary pressure for the different rock types within
the reservoir is required. These properties determine the flow behaviour of two fluid phases
— in case of an oil reservoir of oil and water — within the porous medium, which in this
case is rock. Within such a three-phase system wettability is a dominating property. The flow
behaviour is different for a system in which the oil tends to adhere to the rock surface from one
in which the rock is predominantly in contact with water.
Commonly, the development of an oil field is separated into three productions stages. During
primary recovery the oil is produced from the sub-surface rock by the expansion of the fluids
inside as the pressure drops and buoyancy forces. Thereafter, during secondary recovery, com-
monly brine is injected into the reservoir to maintain reservoir pressure and to displace oil.
This procedure is referred to as waterflooding. For these two stages the flow behaviour within
the field is sufficiently described by the phenomenological extension of Darcy’s law to multi-
phase flow, which requires relative permeability and capillary pressure as a function of brine
saturation as input. During tertiary production, various techniques are utilized to enhance the
production of a field, predominantly by adding additives to the injected brine. The aim is often
to change the flow behaviour or wettability of the system.
5
6 Chapter 2. Theoretical background
2.1 Wettability
”Wettability describes the preference of a solid phase to be in contact with one fluid rather
than another” (Abdallah et al., 2007). In an oil reservoir, which includes rock, oil and water we
find exactly the described situation. A rock surface is called water-wet when the water tends
to cover it and is called oil-wet when it prefers to be in contact with the oil phase (Anderson,
1986a). Furthermore, wettability may vary from location to location. In this case the rock is
called mixed-wet (Anderson, 1986a; Brown & Fatt, 1956; Donaldson & Alam, 2008a; Masalmeh,
2003; Salathiel, 1973).
However, wettability by itself is not a property used as direct input parameter for reservoir
models. Yet, it is known to significantly impact input values such as relative permeability
and capillary pressure (Anderson, 1987a, 1987b; Donaldson & Alam, 2008b). So far, relative
permeability and capillary pressure can only be determined with core-scale experiments. To
predict these parameters for a specific reservoir a better understanding of the principles behind
wettability is crucial. In a reservoir the wettability depends on various properties such as the
exact brine and crude oil composition, reservoir pressure and temperature (P -T ) conditions,
structure of the rock surface, as well as the surface chemistry associated with the mineralogy
of the system and the saturation history, as illustrated in Figure 2.1.
Figure 2.1: Wettability depends on various properties such as brine- and crude-oil composi-
tion, reservoir pressure and temperature (P -T ) conditions, surface chemistry associated with
mineralogy, surface structure and fluid-film formation as well as saturation history. All of
these influence wettability and hence the pore-dynamics during two-phase flow and the relative
permeability.
2.1. Wettability 7
The definitions and concepts used to describe wettability are length scale dependent. Brine and
crude-oil composition, surface chemistry and the P -T conditions directly relate to molecular
interactions. At the molecular scale, wettability is given through the adhesion force between
the solid and one fluid in presence of another fluid. At the sub-pore scale, roughness, which
may facilitate the formation of thin fluid films and layers, becomes an additional influencing
factor. At this length scale the contact angle forming along three-phase contact lines may be
observed and measured (Schmatz et al., 2015). At the larger pore and pore-network scale, the
confined space within a porous medium is taken into account. Wettability is impacting flow
regimes and fluid distributions through capillary pressure (Bultreys et al., 2018; Rücker et al.,
2015b; Ying et al., 2017). Mineralogy and mineral distribution as well as the saturation history
needs to be considered. At this length scale wettability can be characterized through contact
angle distributions (AlRatrout et al., 2017; Andrew et al., 2014; Scanziani et al., 2017), fluid
distributions (Herring et al., 2016) or by defining local capillary pressure (Armstrong et al.,
2012) and small scale relative permeability (Berg et al., 2016). An overview of the different
wettability effects over increasing length scales is shown in Figure 2.2.
Figure 2.2: The description of wettability is length scale dependent. At the molecular scale
wettability is defined as adhesion, which depends on crude oil, brine and mineral composition
as well as P -T . Based on the molecular interactions an intrinsic contact angle can be computed
which considering surface roughness, chemical heterogeneity and fluid film formation can be
translated into an effective contact angle. Pore architecture, saturation history and mineral
distribution are additional parameters which impact wettability at the core scale, which is
characterized through capillary-pressure and relative-permeability/saturation functions.
To predict wettability at the core scale the contribution of each constitutive scale to the overall
wettability of the system needs to be considered. In the following sections the wettability
definitions at each length scale and the relation to the corresponding properties will be discussed
in detail.
2.1.1 Molecular interaction
At the molecular scale wettability is a result of intermolecular interactions between the solid
and two fluid phases. The nature of the interaction between a solid and one fluid, the solid and
the other fluid or between the two fluids depends on the forces acting between the respective
molecules (Israelachvili, 2011a). As illustrated in Figure 2.3, these forces depend on the distance
between the molecules and may be attractive or repulsive (Israelachvili, 2011a; Raina, 2013).
Figure 2.3: (a) shows an example of a force distance curve, which is the sum of all forces
acting between two molecules at different distances. (b) illustrates the different forces acting
on molecules when approaching a surface (adapted from Raina, 2013).
The sum of all forces acting between the molecules for given distances result in a force distance
curve as shown in Figure 2.3a. The minimum in a force distance curve represents the adhesion
2.1. Wettability 9
force. This is the force required to pull apart one molecule from another. In a fluid/fluid/solid
system, the adhesion force represents a measure of the ability of a molecule of one fluid phase
to remain in contact with the surface in presence of another fluid. As such, adhesion is a
measure for wettability. The stronger the adhesion, the more wetting the associated fluid phase
(Israelachvili, 2011c).
Impact of fluid components and rock chemistry on wettability
The strength of the different acting forces depends on the specific components of the system
(rock/crude/brine).
Only few minerals such as graphite or talc are naturally hydrophobic (Bailey & Gray, 1958;
Schrader & Yariv, 1990). Most minerals present in reservoir rock are originally strongly water-
wet (Anderson, 1986b; Benner et al., 1938; Hassenkam et al., 2009). However, the wettability
of a surface may be altered towards more oil-wet by adsorption of surface active components
(Buckley, 2001; Drummond & Israelachvili, 2004; Fernø et al., 2010; Kumar et al., 2005b).
Carbonate rock tends to become more oil-wet than sandstone (Anderson, 1986a; Chilingar
& Yen, 1983; Treiber & Owens, 1972). The reason for this could be the difference in surface
charge between the two rocks. While silica, which are dominating sandstone rocks, are normally
negatively charged and weakly acidic, calcite, dominating carbonates, are positively charged
and weakly basic (Benner et al., 1938).
Correspondingly, these surfaces would be expected to predominantly adsorb components of

opposite polarity or acidity. This was found to be the case for simple crude-oil compounds
(Anderson, 1986a). However, Denekas et al. (1959) detected the adsorption of acidic and basic
components in sandstone and the domination of basic components in limestone when the full
range of crude-oil molecules was used. Polar interactions represent one reason for adsorption
of crude oil (Buckley et al., 1998). This is supported by the finding that adsorbed material is
enriched in nitrogen, oxygen and sulphur when compared to the native oil (Akhlaq et al., 1996).
Adsorbed crude-oil material was also found to be of higher molecular weight than the bulk crude
oil, leading to the conclusion that the more complex compounds dominate wettability alteration
(Akhlaq et al., 1996; Anderson, 1986a). Furthermore, asphaltenes may alter wettability of a
surface by precipitation due to changes in crude composition, temperature or pressure (Akhlaq
et al., 1996; Buckley et al., 1997; Tang & Morrow, 1997). Correspondingly, crude oils are usually
classified by assessing their acid (TAN) and base (TBN) numbers as well as saturate, aromatics,
resins and asphaltene (SARA) fractions, which indicate the stability of asphaltenes within that
crude (Fan et al., 2002; Por, 1992; White et al., 2017). The ratio saturates/aromatics indicates
the ability of the oil to dissolve asphaltenes and the asphaltenes/resins ratio relate to the
colloidal stability of asphaltenes in the crude. The lower these ratios are, the more stable the
oil (Por, 1992; White et al., 2017).
In addition to the chemistry of the crude oil and surface chemistry, brine composition also
impacts wettability. On a mineral surface, the balance between electro-static forces and van der
Waals forces leads to a double layer of counter ions in correspondence with Derjaguin-Landau-
Verwey-Overbeek (DLVO theory; Derjaguin and Landau (1993) Overbeek and Verwey (1948)).
This further impacts the repulsive/attractive forces of approaching molecules (Hirasaki, 1991;
Israelachvili, 2011d). In combination with structural forces of the fluid/fluid interface these
lead to the disjoining pressure Π, defined as the sum of changes in interfacial tensions with
respect to the film thickness h given by:
δ pσ ob σ bs q
Π , (2.1)
δh
where σ ob represents the interfacial tension between the oil and the brine phase and σ bs the
interfacial tension between brine and solid phase (Basu & Sharma, 1996). The disjoining
pressure determines the pressure required to break the water film, and hence to get the non-
wetting phase in contact with the solid. Only if this happens the wettability of the surface can
get altered. Furthermore, the pH of the brine may change surface charge of the rock through
dissolution (Morrow et al., 1973; Thomas et al., 1993) or could cause ionization of surface-
active components leading to a change of the solubility of crude-oil surfactants (Cuiec, 1975).
Correspondingly, a change in brine composition would result in a change in wettability. This
effect is often used in enhanced oil recovery methods, further discussed in Section 2.2.
2.1. Wettability 11
Assessment of wettability with atomic force microscopy
Molecular interactions can be assessed with various methods (Israelachvili, 2011c). One tool
which is capable of resolving the full fluid/fluid/solid system down to the atomic level is the
Atomic Force Microscope (AFM). This method provides force distance curves between a solid
and specific compounds in the presence of a surrounding fluid. A cantilever with a functionalized
atomically sharp tip is moved towards the solid until contact and is then retracted. The bending
of the cantilever reflects the forces at different distances (Israelachvili, 2011c), which can be
translated into a force distance curve as shown in Figure 2.3a.
Atomic force spectroscopy for assessing wettability in a Crude-Oil Brine Rock (COBR) system
was introduced by Basu and Sharma (1996). They attached a sphere to the cantilever and
immersed it in crude oil. The experiments were conducted with planar glass and mica surfaces
inside a fluid cell filled with brine. While they focussed on the determination of disjoining
pressure required to bring the oil phase in contact with the solid, various other researchers
investigated the adsorption of crude oil on a sample surface (Buckley & Lord, 2003; Freer et
al., 2003; Toulhoat et al., 1994). Kumar et al. (2005a) treated surfaces with different crude
oil components and found that of all SARA components minerals treated with resins show the
strongest adhesion between the silicon tip and the sample. However, surfaces treated just with
the asphaltene fraction resulted in adhesion similar to surfaces treated with bulk crude. Van der
Vegte and Hadziioannou (1997) introduced the functionalization of tips with different functional
groups, which allowed to measure force distance curves and the corresponding adhesion force
per group for surfaces in different brines (Eastman & Zhu, 1996; Teschke & de Souza, 2003).
Dickinson et al. (2016a) presented a method to functionalize the tip with crude by drying.
The described methods allow wettability classification on the molecular scale. However, these
measurements cannot be considered representative for the entire rock as they miss larger scale
phenomena.
2.1.2 Sub-pore scale surface structure
At the sub-pore scale the measure for wettability is the contact angle, which can be obtained by
direct imaging (Deng et al., 2018; Giro et al., 2017; Hjelmeland & Larrondo, 1986; Leach et al.,
1962; McCaffery & Mungan, 1970; Mennella et al., 1995; Neumann & Good, 1979; Treiber &
Owens, 1972). As illustrated in Figure 2.4 for a brine/oil/rock system, three different states are
distinguished. The system is called water-wet for a contact angle θi 90 , intermediate-wet
for a contact angle θi 90 and oil-wet for θi ¡ 90 , measured through the denser brine phase
(Abdallah et al., 2007; Anderson, 1986b).
Figure 2.4: A system is called water-wet for a contact angle θi 90 (a), intermediate-wet
for θi 90 (b) and oil-wet for θi ¡ 90 (c), when the contact angle is measured through the

denser brine phase.
However, at this length scale the measured wettability, does not necessarily correspond to
the intrinsic system wettability. Here, wettability is not only influenced by the brine/rock/oil
composition and pressure and temperature conditions, but surface roughness and chemical
heterogeneity also impact the results obtained (Cassie & Baxter, 1944; Dettre & Johnson,
1965; Israelachvili, 2011b; Morrow, 1975; Quéré, 2008; Wenzel, 1936; Xie et al., 2002).
Roughness
Roughness is a measure of surface texture, defined as the deviation of a surface from its ideal
smooth form. When contact angle is measured an angle is fitted to match the observed brine-oil
interface and fluid-rock interfaces. These are considered to be flat. But when imaged at higher
resolution, this is not necessarily the case. For this reason, a theoretical intrinsic system contact
2.1. Wettability 13
angle θi may deviate significantly from the apparent contact angle θ observed in real systems
as illustrated in Figure 2.5a,b (Andrew et al., 2014; Blunt, 2017).
Figure 2.5: A droplet of oil on a water-wet surfaces forms a characteristic contact angle. On
rough surfaces the measured apparent contact angle θ , for which the surface is considered
smooth as indicated by the red dotted line (a), may deviate from the sub-resolution intrinsic
contact angle θi (b). Furthermore, surface roughness may lead to pinning: (c) and (d) show
two possible menisci configurations resulting from the same contact angle. At this location all
contact angles between the two configurations may form (e).
To characterize the deviation in contact angle Wenzel (1936) introduced the roughness factor
given by:
R
Areal
, (2.2)
Aprojection
with Areal being the real surface area and Aprojection being the idealized surface, not considering
the roughness and found the relationship:
cos θ R cos θi (2.3)

Correspondingly, a rough surface causes water-wet systems to appear more water-wet and oil-
wet systems to appear more oil-wet. This equation is limited to various conditions, including
the droplet being larger than the defects of the surface (Quéré, 2008; Swain & Lipowsky, 1998;
Wolansky & Marmur, 1998, 1999).
On a rough surface, texture allows a full range of contact angles to occur. In Figure 2.5 both
cases c and d, show the same contact angle. Yet, the fluid-fluid interface shows a different
configuration. This ‘pinning point’ allows the interface also to form at any other location
between these two states as illustrated in Figure 2.5e (Israelachvili, 2011b; Quéré, 2008). Such
pinning points also occur on model surfaces such as mica, which are commonly considered
atomically smooth (Broseta et al., 2012).
Under dynamic conditions pinning causes contact angle hysteresis as illustrated in Figure 2.6a
(Morrow, 1975; Snoeijer & Andreotti, 2013; Valvatne & Blunt, 2004). The tail of a displacing
drop forms receding contact angles θr and advancing menisci advancing contact angles θa . The
contact angle hysteresis is dependent on the roughness of the surface, the velocity of the moving
contact line and also on the intrinsic contact angle as shown in Figure 2.6b.
Figure 2.6: In dynamic conditions a droplet on the surface will form an advancing and receding
contact angle (a). the observed contact angle hysteresis for a specific texture of the surface is
depending on the intrinsic contact angle (b, schematic drawing based on Morrow, 1975).
Furthermore, grooves on a rough surface lead to entrapment of the non-wetting phase (Her-
minghaus, 2000; Quéré, 2008). This can have large effects on the wettability alteration process.
During drainage, in most cases the aqueous phase is wetting and would form water-films within
2.1. Wettability 15
such grooves. These water-films prevent the contact between the oil and the solid and hence
lead to mixed wettability effects (Buckley, 2001; Buckley et al., 1996; Salathiel, 1973).
Chemical heterogeneity
Chemical heterogeneity in COBR systems is associated with mineral heterogeneity, but may
also be a result of inhomogeneous wettability alteration. Similar to surface roughness, chemical
heterogeneity also leads to pinning effects (Brandon et al., 2003; Dettre & Johnson, 1965;
Extrand, 2003). At the transition between a water-wet surface with contact angle θow and an
oil-wet surface with contact angle θww all contact angles between these two extremes (θow and
θww ) may occur. The contact angle is thereby dependent on the fluid volume placed on the
surface as illustrated in Figure 2.7. Similar to roughness this pinning effect leads to contact
angle hysteresis (Israelachvili, 2011b; Liimatainen et al., 2017; Robin et al., 1997).
Figure 2.7: If the contact area of the drop and the solid is smaller than the oil-wet patch
the contact angle of the drop will reflect the contact angle of the oil-wet surface θow (a). The
opposite will happen if the drop is exceeding the oil-wet patch. The contact angle will reflect the
wettability of the water-wet patch θww (b). When the contact line is pinned at the transition
between these patches any contact angle between θow and θww may form (c) (adapted from
Israelachvili, 2011b).
In reservoir rock a combination of both, surface roughness and chemical heterogeneity is ex-
pected to occur (Buckley, 2001; Buckley et al., 1996).
2.1.3 Pore scale fluid distribution and flow behaviour
Within the confined space of porous media wettability is impacting the fluid distribution and
flow behaviour through a phenomenon called capillary rise. Consider a cylindrical tube of
radius r placed into a vessel of one fluid with one end, while in contact with another fluid on
the other end as shown in Figure 2.8.
Figure 2.8: Capillary rise experiment in which a tube of radius r is placed in a beaker filled
with one fluid. The capillary rise h can be derived from the Young-Laplace equation (Eq. 2.4).
The wettability of the fluid-fluid-solid system, given by the contact angle θ, would control the
spontaneous fluid displacement within the tube following the equation:
h
2σcosθ
, (2.4)
∆ρgr
with h being the height of the fluid column, σ the interfacial tension, ∆ρ the density difference
between the two fluids and g the acceleration due to gravity (Blunt, 2017; Leverett, 1941; Young,
1805).
The pressure in a fluid with no motion is being defined by dP

dz
ρg , with the depth z. In the
given example, the pressure related to the capillary rise dz h is called capillary pressure Pc.
In correspondence with Eq. 2.4, this Pc can be derived from:
Pc ∆ρgh 2σcospθ q . (2.5)

r
2.1. Wettability 17
The described model is based on one single capillary and can be conceptually extended to
systems of vertical capillaries with varying diameters reflecting the pore size distribution of
the rock. This network can be characterised through the maximal-ball approach (Arns et
al., 2005; Hazlett, 1995; Hilpert & Miller, 2001; Silin & Patzek, 2006). Thereby, spheres are
fitted into each voxel of void space in a 3D structure usually obtained through micro-computed
tomography. These spheres are then grown until they touch the solid. Spheres which entirely
overlap with a larger sphere are ignored. The radii associated to the remaining spheres are
analysed further. Local maxima are referred to as pore bodies and local minima as pore
throats. When one fluid displaces another immiscible fluid in a porous medium, the flow is
controlled by this network of pore bodies and throats. Figure 2.9 illustrates the invasion of a
non-wetting fluid into such a system. Following Eq. 2.5 the capillary pressure is highest at the
pore throats, reaching the maximum meniscus curvature equal to pore throat radius. As in this
case the meniscus is receding the contact angle θr needs to be used.
Figure 2.9: Receding meniscus during drainage into a pore. The highest capillary pressure
occurs at pore throats. Once the meniscus passes through the invading phase will rapidly fill
the consecutive pore body. The receding meniscus will thereby move along the surface with
the receding contact angle θr (adapted from Blunt, 2017).
A pore body following a pore throat will only be filled if viscous forces of an oil ganglion as
shown in Figure 2.10 exceed the capillary forces (Stegemeier, 1977). These forces are given by
the equations:
Fcap AporePc Apore r σ , (2.6)

pore
Figure 2.10: An oil ganglion in a rock gets mobilized when viscous forces acting along the
cluster length L overcome capillary forces (adapted from Stegemeier, 1977).
Fvis Ablob∆P Ablob∆L kµw q, (2.7)

rw
where Apore is the cross-sectional area of the pores the ganglion has to cross and Ablob is the
active ganglion surface contributing to the flow. Capillary pressure is P c , pressure difference
along the ganglion ∆P, interfacial tension σ , pore radius r pore , the ganglion length ∆L, viscosity
of the injected (water) phase µw , relative permeability krw and Darcy velocity q. The balance
between these two forces is referred to as capillary number Nc given by the equation:
Ncmacro FFvis AAblob∆Lµ

k P
wq
(2.8)
cap pore rw c
Based on this equation, viscous forces dominate the flow regime for Ncmacro ¡ 1 and conversely
capillary forces dominate for Ncmacro 1 (Dullien, 1992a, 1992b).
However, this expression contains values, which are not easily accessible. Armstrong et al.
(2014) and Berg et al. (2014) introduced a few additional simplifications of this equation to be
able to determine Ncmacro from µCT images.
As µCT is not always available often the following approximation is used:

2.1. Wettability 19
Ncmicro µσw q , (2.9)
Observations showed, that this expression is not capable to consistently predict the flow regime
(viscous or capillary dominated). The mobilization of oil occurs, depending on the rock type,
for Ncmicro between 10- 5 and 10- 7 (Dullien, 1992a, 1992b).
Flow behaviour in water-wet systems
Pore-scale flow regimes have been studied in 2D in micro-models since the 1950s (Chatenever
& Calhoun, 1952). Micro-model experiments allow direct observation of the dynamic flow
behaviour in pores under a transmission microscope at high spatial and temporal resolution.
For the water-wet case, this method provides great insight into two-phase flow behaviour at
different flow rates and saturations.
For instance, Avraam and Payatakes (1995) observed different flow regimes when conducting
micro-model experiments with varying Nc . During connected pathway flow, the two phases
move solely through the (parts of) pores the respective phases occupy. Throughout time the
pathways of each phase remain the same. Under ganglion dynamics, the flow regime of the
non-wetting phase includes clusters with the size of several pores. Illustrations of the different
flow regimes are shown in Figure 2.11. These clusters or ganglia move through the entire pore
space by break-up and reconnection, irrespective of what phase occupies which pore.
Although these models can provide quick insight to flow mechanism (or behaviour), they can be
misleading. In many cases micro-models are chemically homogeneous, have a significantly sim-
plified structure and smooth surfaces, which means they do not represent the full complexity of
natural systems such as wettability, surface roughness and corner flow (Blunt, 1997; Lenormand
& Zarcone, 1984). Also micro-models are only 2-dimensional, which limits the connectivity of
corner flow of the wetting phase, which in most natural (3D) porous media is long-range and
has important impact on pore scale displacement processes and flow regimes (Mohanty et al.,
1987).
Figure 2.11: Different flow regimes may occur during two-phase flow in porous media. During
connected pathway flow the two phases move solely through the pores the respective phase
occupies (a). In a ganglion dynamics flow regime the pores experience a change in occupancy
and distinct ganglia move through break-up and coalescence events (b).
X-ray micro-computed tomography does provide 3D insight into the configuration of wetting
and non-wetting phase fluids in the pore space at a resolution that allows the assessment of
local wetting conditions via the measurement of in situ contact angle (Andrew et al., 2014). A
temporal resolution of only several seconds, which may be obtained at synchrotron beamlines,
allows to study flow behaviour in both model and natural systems (Armstrong et al., 2016;
Berg et al., 2013; Bultreys et al., 2015; Dobson et al., 2016; Schlüter et al., 2016; Singh et
al., 2017a). In addition, pore-scale flow dynamics has been observed using confocal microscopy
(Datta et al., 2014).
From such experiments, we know that during two-phase flow in water-wet systems different
types of events occur. Examples are shown in Figure 2.12. Water-filling events occur when
the pore space initially occupied with oil is filled with water. Examples of this are snap-off
(Lenormand et al., 1983) or piston-like displacement (Dixit et al., 1998b; Lenormand et al.,
1983). Oil-filling events occur when the pore space initially filled with water is replaced by oil.
Under strongly water-wet conditions Haines jumps, in which multiple pores are filled rapidly,
are observed during drainage or co-injection (Berg et al., 2013; Rücker et al., 2015b). These
events lead to significant rearrangement of the phases in the pore space (Andrew et al., 2015;
Berg et al., 2013). During imbibition, smaller pore-space rearrangements are observed due to
fluctuations of the fluid/fluid-interfaces associated with larger water-filling events (Rücker et
al., 2015a). These events induce reconnection of previously trapped clusters and therefore are
2.1. Wettability 21
considered as ganglion dynamics behaviour. Nevertheless, the observed ganglion dynamics do

not contribute significantly to the overall transport of oil, but rather enhance pathway flow by
increasing the connectivity (Berg et al., 2016).
Figure 2.12: Two-phase flow in water-wet systems may lead to different event types, oil-filling
events such as Haines jumps (a) and coalescence (b) and the water-filling events such as break
up (c) and piston-like displacement (d). Examples for each event type are shown before (top)
and after the event (bottom), respectively (adapted from Rücker et al., 2015b).
Natural crude oil/ rock systems however are often mixed-wet. Yet, only little is known on
the effect of wettability on flow dynamics in these systems (Murison et al., 2014; Singh et al.,
2017b; Zhao et al., 2016).
Wettability determination at the pore scale
So far 3D studies on mixed-wet and oil-wet systems have focussed on wettability characteriza-
tion. As shown in Section 2.1.2, one way is the measurement of contact angles. Andrew et al.
(2014) introduced in situ measurements of contact angles in rock and Scanziani et al. (2017)
and AlRatrout et al. (2017) automated the process with different methods.
In contrast to model systems, one does not obtain one single contact angle value, but a distribu-
tion. This may be due to various reasons: First, heterogeneity in surface roughness, mineralogy
or saturation history leads to different wettability at different locations in the rock (mixed wet-
tability). Furthermore, the contact angles may be in different states of movement, in which the
contact angle at any location may vary from advancing to receding. Yet, the contact angle dis-
tribution gives an indication of the wettability of the system. Alhammadi et al. (2017) showed
a shift of the contact angle distribution when the wettability of the rock is altered by ageing.
However, this contact angles are not sufficient for prediction of core-scale relative permeability
curves (Bultreys et al., 2018).
Another way for characterizing wettability is through fluid distribution. In oil-wet and mixed-
wet systems the oil clusters appear sheet-like and flat compared to the more spherical geometry
in water-wet systems (Al-Raoush, 2009; Iglauer et al., 2012). Singh et al. (2016) observed
oil layer formation on the rock surface and between two water interfaces, which provides a
conductive flow path for the oil phase in mixed-wet systems.
The fluid phase distribution or geometry may be fully characterized through morphological
descriptors known as the four Minkowski functionals: volume, surface area, integrated mean
curvature and Euler characteristic (Herring et al., 2013; Joekar-Niasar et al., 2008; Lehmann
et al., 2008; Schlüter et al., 2016; Vogel et al., 2010). The volume is commonly given through
the saturation of the system. Following the Young-Laplace equation
Pc Po Pw σow κ, (2.10)
capillary pressure Pc is a function of the mean curvature κ of the oil-brine interface, with σ ow
being the interfacial tension between oil and water and Po / Pw being the pressure of the oil
and water phases respectively.
The interfacial area and Euler characteristic can be determined by pore-scale imaging. Euler
characteristic χ is a measure of connectivity of a system defined by,
χ β0 β1 β2 , (2.11)
with β0 being the Betti number representing the number of objects, β1 representing the number
2.1. Wettability 23
of loops and β2 representing the number of inclusion. However, given the structure of a con-
solidated rock, β2 is redundant (Herring et al., 2013). The system has a low connectivity if the
number of objects is larger than the number of loops resulting in a positive Euler characteristic.
Correspondingly a negative Euler characteristic represents a highly connected system.
Schlüter et al. (2016) found for water-wet systems that the Euler characteristic appears hys-
teretic within a drainage/imbibition cycle, indicating a higher connectivity of the oil-phase
during imbibition than drainage. Herring et al. (2016) encountered only slight deviation in the
Euler characteristic-saturation relationship when comparing water-wet to intermediate-wet con-
ditions for imbibition. But the final Euler characteristic after drainage indicated a significantly
lower connectivity for the oil-phase in the intermediate-wet than the water-wet scenario, which
they attributed to the absence of connected water-layers leading to trapping of the water-phase.
Furthermore, fluid distribution and configurations at the pore-network scale can be used to
derive relative permeability or capillary pressure estimations (Armstrong et al., 2012; Berg
et al., 2016). The information these properties reveal about wettability relate to analysis
commonly conducted at the core scale.
2.1.4 Core scale capillary-pressure and relative-permeability
The core scale ranging from 5 cm upwards, refers to the length scale at which flooding experi-
ments are conducted. In such experiments one (single-phase flow) or multiple fluids (multiphase
flow) are injected into a porous medium. Generally, flow through a porous medium is charac-
terized by permeability.
Single phase flow in porous media
The absolute permeability of a rock is measured by single-phase flow experiments. First perme-
ability measurements were based on Darcy’s experimental apparatus from 1856 (Darcy, 1856),
shown in Figure 2.13.
Figure 2.13: Darcy’s experimental set-up for single phase flow in porous media allows perme-
ability determination (based on Darcy, 1856).
The permeability K of the rock is thereby given through the following equation
Q
A
q Kµ p dP
dx
ρgq, (2.12)
with Q being the flow rate and A the cross-sectional area of the sample, with the cross-pressure
gradient across the sample dP , the length of the sample dx , the fluid parameters viscosity µ
and density ρ and gravity g (Nutting, 1930). However, Darcy’s experiment considers solely the
presence of one phase.
Two phase-flow in porous media
Multiphase flow in porous media can be characterized by the phenomenological extension of

Darcy’s law given by
qi kriµK p dP
dx
i
ρigq, (2.13)
i
where qi is the Darcy velocity of the respective brine or oil phase i=o,w, dPi is the pressure drop
2.1. Wettability 25
along the core, µi as the viscosity and ρi as the density of the fluid and kri is the permeability
of the phase relative to K, the permeability of the rock (Muskat, 1949).
Relative-permeability-saturation functions are determined experimentally e.g. through steady-

state flow experiments in which the two immiscible phases are co-injected into a rock at different
oil-water fractions fw Qw { pQw Q0 q while the saturation and pressure drop within the core
are monitored. However, these experiments are often very time consuming, as saturation and
pressure must reach equilibrium at each fw (Bear, 1988; Brownscombe & Collins, 1950; Dake,
1978; McPhee et al., 2015; Osoba et al., 1951).
Another way to determine relative permeability are unsteady-state experiments in which one
phase, commonly the brine, replaces the other fluid present in the rock, which is commonly
the oil phase. An example apparatus is shown in 2.14. However, this requires an accurate
measurement of the oil production and continuous monitoring of the saturation across the core
(Batycky et al., 1981; Bear, 1988; Dake, 1978; Johnson et al., 1959; Welge, 1952).
The two-phase displacement process is described through the Buckley Leverett equation
φ
δSw
δt
qi
δfw δSw
δSw δx
0, (2.14)
with the water saturation being Sw (x,t) for time t and position x, qi being the total flow rate,
φ porosity and A the cross-sectional area of the sample and fw pSw q being the fractional flow
(Buckley & Leverett, 1942).
According to this solution the injected fluid propagates through the rock with a shock front
appearing as a sudden change in a saturation vs distance diagram as shown in Figure 2.14.
However, experimental observations show that capillary effects lead to spreading of the water
saturation wave around the shock wave, referred to as capillary dispersion zone (Lake, 1989).
Figure 2.14: An unsteady state flow experiment can be described by the Buckley-Leverett
equation (Eq. 2.14). However, capillary effects lead to a dispersion of the shock front (adapted
from Lake, 1989).
Capillary pressure
Capillary effects within the rock are commonly assessed through capillary-pressure/saturation
relationships as shown in Figure 2.15 (Bear, 1988; Blunt, 2017). These can be obtained through
water and oil injection cycles. In drainage the non-wetting phase replaces the wetting phase.
During imbibition the wetting phase replaces the non-wetting phase.
Most rocks are water-wet prior to the first (primary) drainage. Correspondingly pressure needs
to be applied, e.g. through centrifugation to push the oil phase into the rock. For measurements
of capillary-pressure/saturation functions the pressure needs to be gradually increased. First
only the few large pore bodies at the outer space of the rock get filled until the pressure
is sufficient to pass through the pore-throats of the rock. At this point the water saturation
starts to decrease substantially until connate water saturation is reached. Thereafter the water-
saturation stays constant as the pressure increases (Abdallah et al., 2007; Bear, 1988; Blunt,
2017).
Afterwards, a brine invasion cycle is initiated by placing the oil-saturated sample into brine.
This procedure is called the Amott spontaneous imbibition test (Amott, 1959). Depending on
the wettability after ageing water can spontaneously imbibe into the sample while the capillary
pressure in the system decreases to zero. From this point onwards a further increase in water-
saturation can only be achieved through forced water injection leading to residual oil saturation
2.1. Wettability 27
(Donaldson et al., 1969).
Figure 2.15: Capillary-pressure/saturation functions for oil- and water invasion cycles. (a)
The wettability of the system determines whether the oil can get produced spontaneously
(water-wet) or only through forced water injection (oil-wet). Correspondingly, capillary-
pressure/saturation functions can be used for wettability characterizations. The Amott index
thereby refers to the spontaneous part of the curve, while the USBM index is based on the
forced parts (b).
Like the contact angle hysteresis at the sub-pore scale, the drainage and imbibition cycles also
appear hysteretic. This is also the case for a secondary drainage/imbibition cycle (Anderson,
1987a; Bear, 1988; Blunt, 2017; Morrow, 1976).
Wettability at the core scale
In accordance with the Young Laplace Eq. 2.5 capillary pressure, and therewith, capillary-
pressure/saturation functions, depend on the wettability of a system. As illustrated in Figure
2.15, spontaneous imbibition would cause larger oil production for water-wet samples than
for an oil-wet sample. However, due to oil-trapping the residual oil-saturation is higher for
water-wet samples than for oil-wet samples (Donaldson et al., 1969). Various indices have been
introduced to characterize wettability based on the capillary-pressure/saturation relationship
with the Amott and USBM index being the most common (Amott, 1959; Donaldson et al.,
1969).
The Amott index for oil Io or water Iw defined by
Iw S S1 S and Io S1 SS4 , (2.15)

2 1 4 3
is based on values obtained from the spontaneous part of the capillary-pressure/saturation

function. S1 is the connate-water saturation and S2 the water saturation after spontaneous
imbibition, S4 is the residual-oil saturation and S3 the water saturation after spontaneous oil
invasion as illustrated in Figure 2.15.
The USBM index IUSBM given by
IU SBM logp AA1 q (2.16)

2
is based on the forced part of the capillary-pressure/saturation function. A1 and A2 are given
through the integral area of the forced part of the capillary-pressure saturation function for
each fluid, respectively as illustrated in Figure 2.15. Yet, the comparison between different
COBR system is challenging as initial water saturation, viscosity, and interfacial tension of the
fluids may impact the results. Different scaling groups have been introduced to eliminate these
effects (Ma et al., 1995; Mattax & Kyte, 1962; Schmid & Geiger, 2013). It has been shown that
the scaling groups are effective in cases of uniform wettability (Ma et al., 1995), but natural
systems are often mixed-wet (Salathiel, 1973). Furthermore, Amott and USBM indices are
often misleading when it comes to predicting sample wettability and associated flow behaviour
(Dixit et al., 1998a). One reason may be, that the index represents a non-unique measure of
wettability since different combinations may lead to the same result.
Another way to assess wettability at the core scale is through relative-permeability/saturation

functions.
Blunt (2017) proposed to assess relative permeability curves based on end-point water relative
2.1. Wettability 29
permeability, oil-layer drainage and shape of the water relative permeability. Furthermore, he
considered a mixed-wet scenario in addition to the oil-wet and water-wet system. An end point
max max
permeability krw 0.2 suggests a water-wet system, 0.3 krw 0.6 to a mixed wet and
max
krw ¡ 0.6 to an oil-wet system. Information about oil-layer drainage can be obtained from kro
near the end-point. If kro 0.05, while the saturation is still decreasing, oil is moving through
a continuous oil layer indicative for a mixed- or oil-wet system. If the shape of krw shows a
sharp rise with a cross over saturation Swcross 0.5 the system is often oil-wet. A low krw 0.2
for Sw Swi 0.2 is indicative of a water-wet system (Blunt, 2017).
A schematic example for each scenario is shown in Figure 2.16.
Figure 2.16: Schematic drawings of relative permeability-saturation functions for a water-wet

(a), oil-wet (b) and mixed-wet scenario (c) based on examples and descriptions from Blunt
(2017).
The exact relationship between relative permeability and wettability, however, is still the sub-
ject of investigation. Valvatne and Blunt (2004) studied the relationship with pore network
models, in which contact angle and initial water saturation, assumed to control the wettability
pattern, were matched to experimental work at the core scale (Jadhunandan & Morrow, 1995).
They found that the spatial correlation of oil-wet pores plays an important role. This finding
was confirmed in various other studies (Bultreys et al., 2018; Landry et al., 2014; Zhao et al.,
2010). Berg et al. (2016) used fluid distributions imaged with µCT at different saturations as
input for single phase flow simulations to compute relative-permeability/saturation functions.
The water-wet sandstone sample showed a good match with core-scale experiments (Berg et
al., 2016). However, experiments on mixed-wet systems did not match (Zou et al., 2018). Zou
et al. (2018) concluded that mixed-wettability systems facilitate dynamic conductivity and
ganglion dynamics, which is not reflected by this relative permeability simulation.
2.1.5 Connecting the length scales
In rock heterogeneity is often length scale dependent. A popular example is the graph by
Nordahl and Ringrose (2008) plotting the permeability as a function of sample size, as illustrated
in Figure 2.17. While at the small scale mainly the pore sizes and shapes are affecting the result,
lamina and lithofacies influence the results at lager scales. The different heterogeneity length
scales can be determined through a representative elementary volume (REV) analysis. The
REV is the minimum volume of a rock from which on a specific parameter becomes independent
of the sample size. Correspondingly, the REV may vary for different rock properties (Bear &
Bachmat, 1990).
Upscaling is a process in which a measure at a smaller length scale is translated to a higher

scale. For this process, all REV levels of a specific property need to be taken into account.
In order to build a proper upscaling workflow for wettability all essential parameters and the
corresponding length scales need to be identified and characterized.
Figure 2.17: The heterogeneity of rock at different scales and the response of permeability
(adapted from Nordahl & Ringrose, 2008).
2.2. Enhanced oil recovery 31
Suijkerbuijk et al. (2013) proposed such a workflow to predict the larger scale response to the
low salinity effect. The low salinity effect is based on a change in wettability due to lowering
the salinity of the brine phase, which is often used to enhance oil recovery (Chapter 2.2.1).
However, as the low salinity effect represents a wettability alteration phenomenon, a similar
approach may be used for wettability itself.
Suijkerbuijk et al. (2013) proposed to characterize the wettability alteration process on the
molecular level. The parameters determined at this length scale may be used as input for
computational models based on extended DLVO theory to determine sub-pore scale properties,
which again could be further processed with pore-network models. These results can then be
validated against SCAL experiments.
In this work a more general upscaling workflow for wettability is developed, by assessing the
impact of wettability at each length scale from molecular scale to the core scale.
2.2 Enhanced oil recovery
After conventional waterfloods more than half of the oil remains trapped in the pore-space of the
rock (Lake, 1989; Warren & Calhoun, 1955). To increase oil-recovery various techniques have
been developed. Some of them enhance oil production by changing the rock-fluid interactions
and therewith the wettability of the system, others impact flow behaviour by changing fluid-
fluid interactions. These approaches often originate in the molecular-scale. The acceleration
of oil-production at the field scale, however, may also be impacted by properties of the length
scales in between (Figure 2.18).
Low salinity flooding, already introduced in chapter 2.1.5 is one technique aiming to enhance oil-
recovery (Bernard, 1967; Jadhunandan & Morrow, 1991; Sheng, 2014; Yildiz & Morrow, 1996).
This is achieved by reducing the salinity of the brine which leads to a change in wettability.
However, the underlying wettability alteration mechanism, commonly referred to as low salinity
effect is not yet fully understood. Different studies addressing this effect and theories regarding
the mechanism are elucidated in Section 2.2.1.
Figure 2.18: Enhanced oil recovery methods are often based on molecular processes. However,
the acceleration in oil-production at the core- and field scale may also be impacted on length
scales in between.
Another way to decrease residual oil saturation after conventional waterflood is by changing
the capillary number Nc in favour of the viscous mobilization of the (capillary) trapped oil
(Hirasaki et al., 2008; Lake, 1989; Putz et al., 1981). As illustrated in Figure 2.19, an increase
of Nc by 3-4 orders of magnitude is required to significantly reduce the residual oil-saturation
(Lake, 1989; Lenormand et al., 2006).
Figure 2.19: Depending on the microscopic capillary number Nc, the amount of trapped wetting
and non-wetting fluid can be higher or lower (adapted from Lake, 1989).
Given Eq. 2.9 for Ncmicro this can be achieved either by an increase in viscosity µw of the injected
fluid or by decreasing the interfacial tension σwnw . Based on this concept various injection
strategies have been developed. Polymers or emulsions are used to increase the viscosity of the
injected fluid (Lake, 1989) and surfactants are added to reduce the interfacial tension between
the oleic and aqueous phases. To significantly reduce the residual oil saturation, a surfactant
with ultra-low interfacial tension ( 10-2 mN/m) is required. Surfactants of this low interfacial
tension also lead to the formation of emulsion, which in return increases the effective viscosity.
For optimal conditions for oil-recovery it is crucial to identify the optimal composition of the
surfactant solution, which will be expanded in Section 2.2.2.
2.2.1 Low salinity flooding
Various studies demonstrated an increase in oil production when brine with low salinity was
injected into an oil-saturated rock instead of, or consecutively to a waterflood (Bernard, 1967;
Jadhunandan & Morrow, 1991; Yildiz & Morrow, 1996). Yet, the cause for this change is not
fully understood (Bartels, 2018; Sheng, 2014).
Low salinity effect in sandstone
Tang and Morrow (1999) investigated recovery in Berea cores by low salinity flooding. They
found that the cores only showed a low-salinity effect if they fulfilled the following conditions:
presence of connate water, exposure of the rock to crude oil and a large clay fraction within
the rock. Although these conditions were given, laboratory and pilot tests of Skrettingland
et al. (2011) failed to demonstrate an improvement in recovery. Some of the more prevailing
theories regarding the change in wettability due to low salinity injection include fines migration
(Tang & Morrow, 1999), multivalent ion exchange (Lager et al., 2007; Lager et al., 2008) and
pH variation (Austad et al., 2010).
Tang and Morrow (1999) assumed that the mixed-wet clay particles may get stripped off the
rock surface during low salinity flooding and be produced with the crude oil, letting a water-wet
surface behind, which simplifies the oil production. However, Lever and Dawe (1984) showed
that clay can plug thin pore throats and cause formation damage. Therefore, fines migration
would be rather contra-productive.
Lager et al. (2007) found that the clay particles themselves became hydrophilic when exposed
to brine with lower salinity. They proposed that the decrease of ionic strength and divalent
ions could cause an expansion of the electrical double layer (EDL) at the mineral surfaces.
The release of the oil adhering to the surface would be caused by the release of adsorbed
cations. Austad et al. (2010) proposed that the injection of low-salinity brine could lead to a
substitution of Ca+ by H+ on the clay surfaces. The reaction between OH- and the adsorbed
acid and protonated base causes an increase in water wetness. However, these theories are
based on the presence of clay particles in the rock, but, Nasralla et al. (2014) demonstrated
the LS effect for carbonate reservoirs, which mostly do not contain clay.
Low salinity effect in carbonates
Mahani et al. (2015b) stated that the low-salinity mechanism for carbonates differs from that
for sandstone, since besides decreasing the salinity a change of the brine composition also
shows an effect. This effect is attributable to potential determining ions (PDIs) such as Mg2+
and SO4 2- (Chandrasekhar, 2013; Yousef et al., 2012). However, various studies also showed a
change in wettability towards water-wet for carbonates by diluting the brine, even though the
presence of PDIs was low (Chandrasekhar, 2013; Romanuka et al., 2012; Yousef et al., 2011).
The mechanism behind the observations is still not fully understood and the following theories
were proposed: Mineral dissolution (Hiorth et al., 2010) and surface charge change (Mahani et
al., 2015b).
Hiorth et al. (2010) stated based on a chemical model, that the changes in solution are not
sufficient. He found that especially in the temperature range, where an increase of oil production
is detected a dissolution of calcite takes place. However, Mahani et al. (2015b) still observed
a low salinity effect in the absence of mineral dissolution. They concluded that the change of
surface charges is causing an expansion of the electronic double layer (EDL).
Low salinity at different length scales
Though no consensus about the cause of the low-salinity effect could be reached, most studies
agree that it originates at the molecular or sub-pore scale. Atomic force microscopy is a tool
operating at this very length scale and therefore was used in various studies on the low salinity
effect. Some investigated the impact of ions on the EDL (Dishon et al., 2009; Hilner et al.,
2015; Mugele et al., 2015). Furthermore, Liu et al. (2018) investigated the impact of different
components as well as the wettability state of the surface on the low-salinity effect. Hassenkam
et al. (2009) introduced methods for low-salinity measurements on the original rock surface.
The measurements on chalk were conducted on a broken surface of the rock. The sandstone
samples were crushed and the grains were sprinkled onto a sample holder covered with epoxy
(Hassenkam et al., 2011). With an optical microscope suitable grains were chosen. They found
that the low-salinity response is highest at the clay edges. Matthiesen et al. (2014) compared
a preserved and cleaned core and found that the low salinity response was also enhanced on
residual organic matter.
Hassenkam et al. (2015) used adhesion maps obtained with AFM to estimate contact angles,
which were then implemented in pore-network models to assess the impact of the adhesion
change on the mobilization of the oil at the larger scale. However, larger scale experimental
observations in model systems showed that contact angle changes alone do not necessarily lead
to a release of oil, but that processes such as diffusion through wetting layers also need to be
considered (Bartels et al., 2017a; Mahani et al., 2015a). Similarly, Dufour et al. (2016) found
through computational modelling that small scale surface texture also impacts the outcome.
A prevailing question is how the change in contact angle will impact the pore-scale flow regime
in rock, which is addressed in this work.
2.2.2 Surfactant flooding
Surfactant flooding is a chemical enhanced oil recovery (cEOR) technique in which a low con-
centration of surfactant is added to the brine solution injected into the reservoir. The surfactant
reduces the oil/brine interfacial tension which, in return, increases the capillary number leading
to viscous mobilization of the capillary trapped oil within the rock (Lake, 1989). For a signifi-
cant reduction of the residual oil saturation, an ultra-low interfacial tension between crude oil-
and brine-phase ( 10-2 mN/m) is required. (Chou & Shah, 1981; Hirasaki et al., 2008; Salager
et al., 1979). In such surfactant systems microemulsion forms once the injected solution comes
into contact with the oil-phase. (Chou & Shah, 1981; Salager et al., 1979). Commonly, the
emulsification is studied in laboratory under controlled conditions facilitated by some form of
external mechanical energy such as stirring (Sjoblom, 2005).
Static assessment of the effectiveness of surfactants
The ideal composition of a surfactant solution is typically assessed by equilibrium phase-

behaviour tests. In these tests, the surfactant solution and the oil are initially mixed in
test tubes by rigorous shaking, and then left to stagnate while the emulsification takes place
(Sjoblom, 2005; Unsal et al., 2016). The properties of the microemulsion are assessed by vi-
sually inspecting the solubilization, and by measuring the oil/water interfacial tension and the
viscosity of the resulting microemulsion to determine the optimum surfactant solution.
These tests are done under static conditions and do not probe the influence of flow or in
situ mixing. In subsurface application no equivalent energy as stirring persists. In rock the
emulsification occurs in situ solely controlled by the pore scale flow dynamics (Broens & Unsal,
2018; Tagavifar et al., 2017). Within the reservoir interfacial tension reduction will increase
the Nc and promote displacement but, the higher viscosity, resulting from emulsion structures
with feature sizes ranging from a few to several hundred nm may lead to other adverse effects
like trapping and bypassing.
Dynamic experiments on surfactant flooding
Glass micromodels provide crucial information for understanding the fundamental pore-scale
physics during surfactant flooding as they allow direct visualization of the pore space (Broens
& Unsal, 2018; Tagavifar et al., 2017; Unsal et al., 2016; Yadali Jamaloei & Kharrat, 2009).
Studies in micromodels have shown that microemulsions can form in situ under flowing con-
ditions at shorter time scales than under static conditions (Tagavifar et al., 2017; Unsal et
al., 2016). The pore structure of a rock is more complex, and, therefore, may have significant
influence on the emulsification behaviour. Coreflood experiments can be performed where the
aqueous solutions of surfactant are injected in an oil-saturated rock ranging in length from sev-
eral centimetres to 1m (Fletcher et al., 2015; Jawitz et al., 1998). These experiments provide
an estimate on the average oil recovery values but do not give access to microscopic details at
pore scale.
Recent advances at µCT synchrotron beamlines have enabled the study of multiphase flow
through rock samples at a time resolution of a few seconds (Berg et al., 2013; Wildenschild et
al., 2005). With this capability, the effect of surfactant flooding can be also investigated in
situ (Javanbakht et al., 2017; Youssef et al., 2014).
In this work the formation of microemulsion within the confined space of porous rock will be
investigated and compared to micromodel observations to assess the impact of rock structure
on emulsification.
Chapter 3
Materials and methods
For assessing in situ wettability and wettability alteration processes at the pore- and nano-scale
flow experiments were recorded using X-ray micro-computed tomography (µCT). Fluid-film
formation and adhesion forces were further investigated using atomic force microscopy (AFM).
The overview of the different experiments conducted in this work can be found in Table 3.1.
Flooding experiments provided insights into the impact of mixed-wettability on two-phase flow
behaviour. Furthermore, the impact of EOR techniques such as low-salinity injection and
surfactant flooding were investigated. Amott spontaneous imbibition tests were used to link
core-scale wettability with pore-scale phenomena. AFM was used to bridge the pore-network-
to the sub-pore scale, by assessing fluid-film formation and adhesion forces of different crude-oil
components.
In the following section, the different fluids and rock samples used in the experiments are
described (Section 3.1). The methods applied as well as the detailed experimental protocols are
presented in Section 3.2 (µCT) and Section 3.3 (AFM). A description of supporting experiments
and data used for comparison and validation is provided in Section 3.4.
39
40 Chapter 3. Materials and methods
Table 3.1: Overview of the different experiments conducted in this work. A more detailed list
including the rock-fluid-fluid systems used can be found in corresponding Sections 3.3 and 3.4.
Nr Title
µCT - Flooding (Section 3.2.2)
1 Flow behaviour in mixed-wet systems

2 Low-salinity flooding (secondary and tertiary mode)
3 Surfactant flooding (tertiary mode)
4 Waterflood: water-wet and mixed-wet system
µCT - Spontaneous imbibition (Section 3.2.3)
5 Amott-tests on SCAL plugs

6 Dynamics of spontaneous imbibition processes (dynamic imaging)
AFM (Section 3.3.3)
7 Reference – Topographical imaging

8 Roughness on rock surfaces
9 References – Force spectroscopy
10 Fluid films on calcite
11 Fluid films in rock
12 Adhesion
3.1 Fluid and rock systems
Wettability (as discussed in Section 2.1) is a property resulting from interactions of a solid
in contact with the interface between two fluids. Within the petroleum industry this mostly
relates to Crude Oil, Brine and Rock or mineral phases (COBR system). The different solids
and liquids used in this study were characterized as a basis for further describing the nature of
the interactions observed in order to extend the ability to predict and upscale wettability.
3.1. Fluid and rock systems 41
Rocks and minerals
Two types of carbonates were studied: Ketton, with a simple pore structure and the more
complex Estaillades. In addition, two sandstone rocks were used for comparison: Berea and
Gildehauser. A list of all rocks is shown in Table 3.2.
Table 3.2: List of rocks used in this study.
Rock Rock type Porosity φ Permeability [mD]
Ketton Oolithic carbonate 0.23 3000 – 6000

Estaillades Limestone carbonate 0.3 270
Berea sandstone 0.2 700
Gildehauser sandstone 0.21 1000
Ketton rock, obtained from the Ketton quarry in Rutland, England, is a middle Jurassic oolithic
carbonate rock consisting of round grains (Ooids and peloids) ranging from 100 µm – 1 mm
size (Andrew, 2015; Hudson & Clements, 2007). Due to the round grains Ketton rock shows a
macroscopic pore structure with distinct pore bodies and pore throats, which makes this rock
type suitable for multiphase flow studies based on µCT (Andrew et al., 2014; Singh et al.,
2016). The rock structure measured with µCT is shown in Figure 3.1.
Figure 3.1: 3D illustration of Ketton limestone (a). The round grains build a simple macroscopic
pore structure which is easy to interpret with µCT. However, as shown in the MIP curves (b)
the grains themselves contain additional microporosity.
The grains contain additional micropores, which cannot be resolved in the µCT images, but
which may be assessed with Mercury intrusion porosimetry (MIP) (Figure 3.1b). These mi-
cropores with around 100 nm diameter contribute very little to the overall permeability of the
rock (Bijeljic et al., 2013) and hence are considered to have little impact on two-phase flow.
The rock has a typical porosity φ = 23% and permeability of 3 – 6 D. It consists predominantly
of calcite (99.1%) with minor quartz (0.9%) components (Andrew, 2015).
Estaillades is a Cretaceous bioclastic limestone from the Estaillades quarry in south east France.
The rock has a typical porosity φ = 30% and permeability of 270 mD (Le Guen et al., 2007).
As shown in Figure 3.2, Estaillades shows a dual porosity similar to Ketton.
Figure 3.2: µCT image of an Estaillades limestone (a). Consisting of different recrystallized
fossils the rock shows a complex pore-structure and a dual porosity as shown in the MIP-curve
(b). Therefore, image segmentation and processing are more difficult than for Ketton and
the sandstones studied. The rock consists predominantly of calcite (97.9%) with minor quartz
(2.1%) components (Andrew, 2015).
However, due to its bioclastic composition the rock has a more complex pore-structure, con-
taining smaller features, which are more difficult to resolve. Therefore, image segmentation
and processing are more difficult and the results may be less accurate than for the other sam-
ples. The rock consists predominantly of calcite (97.9%) with minor quartz (2.1%) components
(Andrew, 2015).
Gildehauser, shown in Figure 3.3, is a Cretaceous hydrocarbon reservoir sandstone with a

transgression origin from northern Germany (Mutterlose & Bornemann, 2000). It consists of
96.5% quartz, 2% feldspar and 1.5% clay minerals (kaolinite). Due to the high quartz content,
the rock is suitable for experiments reducing effects of mineral variation to local wettability
changes. The grains have a size of around 0.4 mm. The rock has a porosity φ = 21% (Dresen
et al., 2010) and a permeability around 1 D (Rücker et al., 2017).
Figure 3.3: The µCT image of the Gildehauser sandstone illustrates the pore-structure of the
rock (a). The MIP-curve is homogenous with a peak pore diameter of 33.45 µm (b).
Berea, shown in Figure 3.4, is a Devonian sandstone from Ohio, USA with a transgression
origin (Pepper et al., 1954).
Figure 3.4: 3D illustration (a) and MIP-curve (b) of a Berea sandstone.
The rock is easily accessible, relatively homogeneous in mineral composition and grain size
and is therefore well studied (Churcher et al., 1991; Ma & Morrow, 1994; Pepper et al., 1954)
and often used for core-flood experiments (Iglauer et al., 2010; Øren & Bakke, 2003; Tang &
Morrow, 1999). The rock has a porosity of φ = 20% and a permeability of 700 mD. It consists
of 85 – 90% quartz, 1 – 2% dolomite, 3 – 6% feldspar and 6 – 8% clay minerals (Churcher et al.,
1991). The mineral variety within the rock and especially the presence of clay make this rock
suitable for studies on the impact of wettability distribution and respective alteration process,
such as low salinity flooding (Tang & Morrow, 1999).
In addition to the different rock types, minerals listed in Table 3.3 were used for test measure-
ments with AFM. Two types of mica (Muscovite and Biotite) with a negative surface charge
and calcite with a positive surface charge were used. All minerals are perfectly cleavable, which
is beneficial for two reasons: First, cleavage just before an experiment prohibits contamination
of the surface and second, the freshly cleaved surfaces are smooth, which provides imaging
reference and effective lack of surface roughness.
Table 3.3: List of minerals used in this work.
Minerals Chemical composition Surface charge
Muscovite mica KAl2 [(OH,F)2 |AlSi3 O10 ] -

Biotite mica K(Mg,Fe2 ,Mn2 )3 [(OH,F)2 |(Al,Fe3 ,Ti3 )Si3 O10 ] -
Calcite CaCO3 +
Calcite has a perfect cleavage on the morphological coordinates [1 0 1 1] in all 3 directions at an

angle of 74 . When a calcite sample is cleaved precisely, an atomically smooth surface will be
obtained. A slight deviation may result in steps along the surface which, by themselves show
a 74 angle as illustrated in Figure 3.5a.
Mica is a sheet silica which has a perfect cleavage on [0 0 1]. Similar to calcite, an atomically
flat surface is almost granted after cleavage. A deviation can only be observed if measured at
the transition of one mica sheet to another. In this locations exfoliation of the mica can be
observed as shown in Figure 3.5b.
Figure 3.5: Topographical AFM images of a calcite (a) and a biotite mica sample (b).
Crude oils and model oils
Three crude oils were chosen based on their expected wettability alteration potential. As
indicators for the wettability alteration potential of a crude, the total acid number (TAN)
and total base number or total basic nitrogen (TBN) obtained through titration with KOH
or HCIO4 respectively are used, as those components are known to be surface active and to
adsorb on the rock surface (Buckley et al., 1998; Section 2.1.1). Based on 40 different crudes
tested the TAN may vary in between 0.02 - 2.7 mg KOH g1 . TBN in this study is reported
as the amount of nitrogen [mg] per 1 kg oil instead of the amount of KOH [mg] per 1 g oil
following Shell standard protocols. Based on results for 40 different crude oils the TBN may
vary between 33.4 - 654 mg kg1 . Furthermore, asphaltenes are known to alter the wettability
of a rock. These components may adsorb on the rock surface due to their polar nature, but
may also come out of solution by precipitating and adhering on a surface induced by changes
in crude composition, temperature or pressure (Buckley et al., 1998). TAN, TBN, asphaltene
content and physical properties of the selected crude oils are listed in Table 3.4.
Saturates, Aromatics, Resins, Asphaltene (SARA) analysis, which is based on chromatography

provides an additional breakdown of compound classes. The ratio saturates/aromatics indicates
the ability of the oil to dissolve asphaltenes. The asphaltenes/resins ratio on the other hand
relates to the colloidal stability of asphaltenes in the crude. The lower these ratios are, the
more stable the oil (Por, 1992; White et al., 2017). The SARA fractions of the different crude
oils are listed in Table 3.5.
Table 3.4: TAN, TBN, asphaltene content as well as viscosity and density of the crude oils used
in this work.
Oils TAN TBN Asphaltenes Viscosity 20 C Density 20 C

[mg KOH g1 ] [mg kg1 ] [wt %] [mPa s] [g cm3 ]
crude A 0.07 83.9 0.28 4.87 0.83

crude B 0.38 2.86 0.91 17.03 0.85
crude C 0.02 385-405 0.81-1.44 12.02 0.87
Table 3.5: SARA fractions for the crude oils used in this work displayed in the relative weight
of each compound class.
Oils Saturates Aromatics Resins Asphaltenes

[wt %] [wt %] [wt %] [wt %]
crude A 58.45 36.92 4.36 0.28

crude B 52.08 39.06 7.96 0.91
crude C 53.09 – 53.87 35.55 – 38.15 7.17 – 9.93 0.81 – 1.44
In addition to crude oil, n-decane (viscosity at 20 C: 0.85 mPas, density at 20 C: 0.73 g cm-3 )
was used as a model oil. The advantage of using n-decane as a representative model oil is the
simple chemical composition, which prevents complex behaviour such as emulsion formation.
Furthermore, it is transparent, which allows AFM measurements of a sample submerged in it.
n-decane commonly yields strongly water-wet behaviour in most rock-brine-n-decane systems
and therefore in this work is used when a water-wet reference for comparison with mixed-wet
or oil-wet systems is needed. In some µCT experiments the oil phase (n-decane or crude) was
doped with iodo-decane (20 wt%). This concentration was found optimum as the oil-phase
shows approximately the same grey value as the rock. Correspondingly, this concentration
leads to a high contrast between oil and brine while the attenuation remains at the level of the
rock itself.
Aqueous solutions
The brines used in this study are listed in Table 3.6. The formation water recipe corresponding
to a carbonate reservoir (FW1) and the seawater recipe (SW) were taken from Mahani et
al. (2015b). The second formation water recipe (FW2) is based on a brine composition in
a sandstone reservoir. The respective formation brines were selected based on the rock type
they were used in to ensure a rock-brine equilibrium. Furthermore, doped KI-brinse of different
concentrations were used. KI was used as contrast agent for the aqueous phase. The ideal
concentration of a KI-brine for µCT based studies of two-phase flow in carbonate rock is
around 9 wt % (doped brine: DB). At this concentration the brine is easily distinguishable
from the non-doped oil-phase and the rock. For studies addressing the low-salinity effect, a
17 wt % KI-brine was chosen as a high salinity brine (HS) and 2.7 wt % as corresponding
low-salinity brine (LS). With 27,290 wt-ppm the concentration of the chosen low-salinity brine
(LS) is usually considered as too high for the low salinity effect, but based on the ionic strength
and the number of ions corresponding to 9,500 ppm NaCl, which is considered appropriate
(Mahani et al., 2015b). The high salinity brine corresponds to 70400 ppm NaCl.
Table 3.6: Composition of formation water (FW1+2), seawater (SW), doped high salinity (HS)
and low salinity brine (LS) and optimum doped brine (DB) used in this work.
Ion Na K Mg2 Ca2 Cl SO4 2 HCO3 I Ionic pH

[mg/L] strength
[mol/L]
FW1 49898 0 3248 14501 111812 234 162 0 3.659 6.9

FW2 4270 7240 20 300 13750 0 0 0 0.396 7
SW 13404 483 1618 508 24141 3384 176 0 0.869 8.0
HS 0 47106 0 0 0 0 0 152894 1.205 -
DB 0 21198 0 0 0 0 0 68802 0.542 -
LS 0 6359 0 0 0 0 0 20641 0.162 -
A surfactant solution comprising of 1 % Na2 CO3 , 5 % 2-butanol, 1.25 % NaCl and 2 % surfactant
was used in a surfactant flooding experiment. It was chosen based on its capability to form
a microemulsion phase with n-decane at ambient conditions in equilibrium phase-behaviour

tests (Section 2.2.2). The surfactant was an internal olefin sulfonate (IOS) with 20 – 24 carbon
atoms (C20 – 24) and is produced by Shell Chemicals as the ENORDET O series. The solution
corresponds to a recipe previously used in micromodel experiments (Unsal et al., 2016).
3.2 Micro computed tomography (µCT)
Computed tomography is an X-ray based non-destructive imaging technique. The X-rays used
are commonly generated by X-ray sources such as X-ray tubes (creating a cone beam) or
synchrotron facilities (creating a parallel beam). The X-rays pass through a sample and are
partly absorbed. The absorption rate depends on the X-ray energy and material properties of
the imaged object (density and atomic number). The attenuated X-rays are then received by a
detector, capturing transmission images at any given time. Those images, called radiographs,
represent linear attenuation maps of a sample at a certain angular orientation θ. The linear
attenuation is described by the Beer-Lambert law given as
I I0eµ L,
a
(3.1)
where I is the transmitted intensity, I0 is the incident intensity, µ is the X-ray linear attenuation
coefficient and L is the optical length through the sample (ASTM International, 2011). During
CT multiple radiographs of an object are collected at different orientation angles θ, either
by rotating source and detector around the sample, or the sample itself around its axis, as
illustrated in Figure 3.6. These radiograph sets are used to reconstruct a 3D representation
of the scanned object. First, the collected [x,y] projections are converted into [x,θ] sinograms,
which are stacked sequentially along the y direction. Then, the sinograms are back-projected
into a 3D intensity map. Different algorithms (Marone et al., 2008; Marone & Stampanoni,
2012; Paganin et al., 2002; Vlassenbroeck et al., 2007) may be used for this purpose (ASTM
International, 2011).
3.2. Micro computed tomography (µCT) 49
Figure 3.6: Basic illustration of a microtomography apparatus with a parallel beam source and
detector monitoring a rotating sample (adapted from Wildenschild et al., 2004).
µCT relates to CT techniques imaging structures down to a resolution of the order of 1 µm2 per
pixel. More information on possibilities and limitations of this technique especially in the field
of two-phase flow in porous media can be found in the literature (Cnudde & Boone, 2013; Myers
et al., 2012; Pichon & Lynch, 2005; Wildenschild et al., 2013). In Appendix B possibilities and
limitations related to the chosen experimental protocols are discussed.
3.2.1 Equipment and facilities
In this study, various types of micro-CT scanners were used: 3 different benchtop systems and
one synchrotron beamline. In a synchrotron, electrons are accelerated close to light speed and
then directed into a storage ring, where they orbit. The deflection of the electron trajectories,
caused by bending wigglers, magnets or undulators generate electromagnetic radiation which
at relativistic speed is directed in the flight direction of the electrons as shown in Figure 3.7,
and termed Synchrotron radiation. In accelerators built to study particle physics this radiation
is an undesired by-product but for scientific light sources like the SLS this is the main desired
effect. Such high intensity light with an energy range from infrared to hard X-ray enables µCT
measurements with a high spatial and temporal resolution (Pichon & Lynch, 2005).
To image the pore-scale dynamics of two-phase flow in mixed-wet systems, the facilities at the
TOMCAT (TOmographic Microscopy and Coherent rAdiology experimenTs) beamline at the
Swiss Light Source (SLS) of the Paul-Scherrer-Institute (PSI) in Villigen (Switzerland) were
used. The lay-out of the PSI SLS is shown in Figure 3.7.
Figure 3.7: Schematic drawing of the Paul-Scherrer Institute Swiss Light Source (PSI SLS).
synchrotron. Electrons are first accelerated in a linear accelerator, boosted within the booster
ring before they are directed into the storage ring where they maintain in orbit. Bending
wigglers and magnets deflect the electron trajectories and generate brilliant bremsstrahlung,
which can be used at different beamlines, such as TOMCAT for e.g. fast µCT experiments
(schematic drawing adapted from ”SLS Beamlines”, 2018)
The Synchrotron light directed to the TOMCAT beamline is obtained through a super bending
magnet at a magnetic field of 2.9 Tesla, which shifts the critical energy from 5.4 keV generated
by a normal SLS bending magnet (1.4 T) up to 11.1 keV. This causes an increase of flux at
hard X-rays (¡ 20 keV) (Stampanoni et al., 2006). Polychromatic radiation with a 5%, 250
µm aluminium filter, leading to a peak energy around 20 keV and a total integrated energy of
about 44 keV, was used in this work (Bartels et al., 2017b).
However, the highly intense synchrotron light may also lead to photochemical reactions within
the imaged sample, manifesting in bubble formation. Such bubbles were observed within rock
in experiments with a doped oil phase (Appendix B, Bartels et al. (2017b)). For this reason,
the Centre for X-ray Tomography of the Ghent University (UGCT, www.ugct.ugent.be) and
XRE (www.xre.be) were approached for studies requiring a doped oil phase. Their bench-
top Environmental MicroCT (EMCT) is optimized for monitoring dynamic processes, however
with a less intense beam. Unlike other µCT-systems the sample is fixed, while X-ray source
and detector are rotated in a horizontal plane allowing a fast and continuous data acquisition.
(Bultreys et al., 2016)
Furthermore, HECTOR a high energy µCT scanner at UGCT was used for static pore-scale
analysis on core-scale samples. With a 240 kV open type directional source this device is
capable of achieving 4 µm resolution and a maximal power of 280 W, giving sufficient flux
even after strong filtration. The 40x40 cm flat panel with (200 µm)2 pixels covered with CsI
scintillator is capable of covering a field of view of 80x80 cm2 (Dierick et al., 2013).
All experiments performed externally, were tested prior to the experimental session internally
in a benchtop µCT-scanner Xradia (Zeiss). Furthermore, this scanner was used for additional
static experiments.
3.2.2 µCT core-flooding experiments
Core flooding is a common test to determine the permeability of a rock. Thereby, the pressure
gradient along a core sample during fluid injection is measured. Following Darcy’s law (Eq.
2.13) the permeability of the medium can be calculated. These tests are often referred to as
single-phase flow experiments (Section 2.1.4). To predict two-phase flow, where two immiscible
phases flow through a sample, the contribution of each phase to the overall flow (relative
permeability) needs to be determined. As the two fluids influence each other’s pathways within
the porous medium, their relative permeability is saturation dependent (Section 2.1.4).
Relative-permeability can be obtained by steady-state and unsteady-state core-flood experi-

ments (Chapter 2.1.4). In steady-state experiments the two fluids are co-injected with different
water-oil-fractions, starting with 100% oil and moving in steps to 100% brine. In unsteady-
state experiments brine is injected into an oil-saturated rock. The saturation within the rock
is dependent on the exact location within the sample and time after the experiment started,
following the Buckley-Leverett equation (Eq. 2.14). The relative permeability-saturation func-
tions can be obtained by the Johnson-Bossler-Naumann-method (Johnson et al., 1959).
The unsteady-state method is also often used to assess the impact of enhanced oil-recovery
(EOR) methods. First, a conventional waterflood is conducted and consecutively, the modified
brine, e.g. with diluted salinity (low salinity flooding) or with surfactants (surfactant flooding)
is injected. The efficiency of the EOR method applied is assessed based on the additional
oil-recovery achieved. In this case EOR is applied after waterflooding and therefore is called
EOR in tertiary mode. In contrast, EOR in secondary mode refers to an injection directly into
the oil-saturated rock without a prior waterflood. In this case, the efficiency is assessed by
comparison with a waterflood of a twin sample (Larry et al., 2014).
While classical flooding experiments are conducted at the core scale, which covers a large volume
of rock and therefore is considered representative, these experiments give little insight into the
pore-scale flow regime. Pore-scale phenomena in rock can be studied with µCT. Nevertheless,
one needs to keep in mind that due to the sample size µCT-experiments often miss larger-scale
phenomena as a clear flooding front (Figure 3.8).
Over the last years different flow cells have been developed for this purpose, e.g. a flow cell
operating at elevated pressures (Iglauer et al., 2011) or a dual piston flow set-up allowing the
co-injection of two phases (Armstrong et al., 2014). In this study such flow cells were used for
dynamic and static experiments. In dynamic experiments, the fluid phases are monitored in
3D over a certain time interval. The dynamic experiments were conducted with the flow cell
shown in Figure 3.8c.
Figure 3.8: µCT is not capable of visualizing macroscopic phenomena such as a saturation front
(a) however, it is capable to resolve pore-scale phenomena (b). For flooding experiments, the
sample is mounted on top of a flow cell (c).
This flow cell contains two remotely controlled piston pumps, which enables the injection of two
different fluids without dismantling the set-up allowing the study of low salinity flooding (LSF)
or surfactant flooding in tertiary mode. The two micro-pumps with a total volume of 2011 µl
can be filled with fluid through two filling valves prior to the experiment. The stepper motors
inside each pump, with a maximum torque of 0.15 mNm operating at 384 µsteps/revolution
and with a pitch of 0.5 mm/revolution, push the fluid into the sample mounted at the top of
the cell. After passing through the sample, the fluids are directed into a liquid reservoir. The
electronics are powered by a slip ring connection at the sample stage of the TOMCAT beam
line, enabling a continuous rotation of the flow cell and thus a continuous and prompt data
acquisition (Armstrong et al., 2014).
For the static experiments, the flow cell described in Singh et al. (2016) was used. In contrast
to the dual inlet cell described above, this cell contains only one inlet. Similar to the flow cell
used by Iglauer et al. (2011) and Andrew et al. (2014) the pumps for this cell are mounted
externally. Connecting flow lines limit the rotation and data acquisition, however, this flow
set-up is capable of handling larger fluid volumes enabling floods with a duration up to several
days. In this flow cell the sample is placed in a Viton sleeve attached through metal end-pieces
and flow lines to pumps on each side.
List of experiments
Four types of waterflood experiments were conducted in this work. The first type (1A – E)
addressed the flow behaviour in mixed-wet systems: Two rock types were tested in combina-
tion with two different crude oils. Immediately after each experiment on flow in mixed-wet
systems a tertiary mode low salinity experiment (2A – E) was performed by flooding doped
LS-brine succeeding the doped HS-brine. In addition, one LS-experiment in secondary mode
was conducted, in which LS-brine is injected directly (2F).
The third type consists of a surfactant flooding experiment in tertiary mode (3A).
The fourth experiment represents the sample preparation for AFM studies (Chapter 3.4). One
(water-wet) sample was brought to residual water saturation (4A) and another mixed-wet
sample to residual oil saturation (4B). The aim here was the comparison of wettability indices
at different length scales.
All flooding experiments are listed in Table 3.7.
Table 3.7: List of flooding experiments conducted during this work.
Nr sample brine oil
Dynamic imaging of flow behaviour in mixed-wet systems
1A Ketton HS 1 Crude A
1B Estaillades HS 1 Crude A
1C Ketton HS 1 Crude B
1D Estaillades HS 1 Crude B
1E Berea HS 1 Crude C
Low salinity flooding: tertiary mode (A – E) and secondary mode (F)
2A Ketton LS Crude A
2B Estaillades LS Crude A
2C Ketton LS Crude B
2D Estaillades LS Crude B
2E Berea LS Crude C
2F Ketton LS Crude A
Surfactant flooding (tertiary mode)
3A Gildehauser Surfactant solution Doped n-decane
Waterflood: water-wet (A) and mixed-wet (B)
4A (11A) Ketton HS 2 n-decane

4B (11B) Ketton HS 2 Crude A
Sample preparation and experimental protocol
Experiments 1A – 2F: Dynamic imaging of flow behaviour in mixed-wet systems and low salinity
flooding
The experiments were conducted following the corresponding workflows described below:
When handling the core samples (diameter ø = 4 – 5 mm, length L = 20 mm) nitrile gloves
were used to prevent contamination of the rock upon touch, which may lead to wettability
alteration.
The rock samples were first cleaned using isopropanol (99 wt%) and then heat-shrunk in PEEK.
In the next step, the rock samples were saturated with HS-brine under vacuum and desaturated
with crude-oil in the flow cell by applying a flow of 500 µl/min (depending on rock type 6.6-
9.9 PV/min) for approximatly 2 min (in total 13.3-20 PV were injected). Afterwards, the
samples were submerged in crude oil and aged at 30 bars and 70 C for one week. Prior to
the experiments the two pistons of the dual-piston flow cell were filled with HS-brine and LS-
brine. For experiment 2F only one piston was filled with LS-brine. The experiments were
performed at the TOMCAT-SLS beamline at PSI. During the experiment, the fluids were
flooded consecutively through the sample in the named order by applying a flowrate of 30 µl/min
(depending on rock type 0.4-0.6 PV/min) for 2 x 30 min (in total 13.3-20 PV were injected).
A detailed list showing the flow sequence of each experiment can be found in Appendix A. The
experiments were monitored with a scan rate of one 3D-image per min for the HS-waterflood
and one 3D-image per 7 s for the first 5 min of the LS-flood and then again one 3D-image per
min. The resulting images had a voxel length of 3 µm. While particular attention was given
to eliminating dead volumes, switching from one piston to the other may have led to a release
of crude oil trapped in the line to the sample.
Experiment 3A: Surfactant flooding
The sample (diameter ø = 4 mm, length L = 20 mm) was cleaned and embedded into polycar-
bonate by heat shrinking as in experiment 1A-2E. Prior to the flooding, the sample was satu-
rated first with formation water (FW2, Table 3.7)-brine which was then displaced by drainage
with doped n-decane. The two pistons of the dual-piston flow cell were filled with FW2 brine
and surfactant solutions. The experiments were performed at the TOMCAT beamline at the
SLS of the PSI. As in experiments 1A – 2E the cells were emptied consecutively in the named
order by applying a flowrate of 30 µl/min (9.5 PV/min) for 2 x 30 min (2 x 19 PV) while mon-
itored with a scan rate of one 3D-image per min for the FW2-brine flood and one 3D-image per
7 s for the first 5 min followed by one 3D-image per min for the surfactant flood. The resulting
images had a voxel length of 3 µm. The full flow sequence is listed in Appendix A.
Experiments 4A – 4B: Waterflood
The sample 4A (diameter ø = 6 mm, length L = 20 mm) was cleaned and then mounted on
the single inlet flow cell described in 3.3.2 by first, placing it into Viton sleeve, which were
then connected to metal end-pieces connected via flow lines to a pump. After loading the
sample, a confining pressure of 11.2 MPa was applied. The rock was first flushed with CO2
to displace the air and then with brine. Afterwards, the brine was replaced by n-decane with
an injection rate of 1 ml/min (7.7 PV/min) for 25 min (corresponding to 193 PV) followed by
0.5 ml/min (3.8 PV/min) for 10 min (corresponding to 38 PV). This sample has been used in
an AFM experiment (experiment 11A) for further studies. Sample 4B (diameter ø = 6 mm,
length L = 20 mm) was initialized with brine under vacuum and desaturated with crude oil
by centrifugation (URC-628, 129 Coretest Systems Inc., used at 3500 RPM for 24 hours). The
sample was then aged at 30 bars and 70 C for 4 weeks. During shipment, the sample got
contaminated with air. Nevertheless, the sample was brought to residual water-saturation by
flooding brine for 25 min (corresponding to 193 PV) followed by 0.5 ml/min (3.8 PV/min) for
10 min (corresponding to 38 PV). The sample was scanned prior to and after the waterflood.
The resulting images had a voxel-length of 2 µm. The full flow sequence of these experiments
can be taken from Appendix A.
Image analysis
The images acquired during the experiments were processed with AVIZO 9.0 (Thermo Fis-
cher Scientific, Geodict 2015 (Math2Market) and Matlab (R2014a, MathWorks) following the
workflow described below.
Experiments 1A – 2F: Dynamic imaging of flow behaviour in mixed-wet systems and low salinity
flooding
The images were reconstructed using the Paganin method (Marone & Stampanoni, 2012; Pa-
ganin et al., 2002), corrected for beam hardening effects, which appear due to the preferential
absorption of low energy photons (Brooks & Di Chiro, 1976), and filtered with a non-local
means filter (AVIZO 9.0 (Thermo Fischer Scientific)). Consecutively, the images were registered
and segmented by combining threshold (Sezgin & Sankur, 2004) and watershed segmentation
(Vincent & Soille, 1991) as illustrated in Figure 3.9 with AVIZO 9.0.
Figure 3.9: First wet and dry images were filtered. In a second step the wet image was registered
to the dry image. The segmented images were then combined to reflect the 3 phase system.
The wettability of the system was assessed visually and by contact angle measurements follow-
ing the workflow described in Andrew et al. (2014) based on the final image of the flooding
sequence. Only menisci between oil and brine, which appeared within the volume initially oc-
cupied by oil were considered. The 100 measurements were taken through the denser (brine)
phase. In addition, fluid distributions were used for wettability classification. These distribu-
tions can be computed by fitting spheres into the pore space or the space occupied by each fluid
phase (oil, water, emulsion) using the granulometry algorithm from GeoDict (Hilpert & Miller,
2001). The fluid distribution curves were weighted with the corresponding volume fraction of
each phase. Binarized images were used to determine and quantify event types by overlaying
consecutive images as described in Rücker et al. (2015a). Only events larger than 104 µm
(20 Voxel x 20 Voxel x 20 Voxel) were considered as smaller events are often meniscus fluc-
tuations and more likely effected by image artefacts (Berg et al., 2015; Rücker et al., 2015a).
Furthermore, permeability through the respective oil- and water-phase were computed using
a single-phase Navier-Stokes solver provided by Geodict 2015 (Math2Market) in order to ob-
tain relative permeability saturation functions following the workflow described in Berg et al.
(2016).
Experiment 3A: Surfactant flooding
As in experiment 1A –2F the images were reconstructed using the Paganin method (Marone
& Stampanoni, 2012; Paganin et al., 2002). The FW2-brine flood sequence was then filtered
and segmented with Avizo 9.0 (Thermo Fischer Scientific) as illustrated in Figure 3.9. During
the surfactant flood the mixing with the remaining oil led to the formation of emulsions with
a range of composition and association of grey levels ranging between (non-doped) water and
doped oil levels. The degree of emulsification was assessed from the grey level following the
linear blending rule for X-ray attenuation expressed here as
Grey valuepemulsionq γo Grey valuepoilq γw Grey pwaterq (3.2)
where the oil and water content of the emulsion are γo and γw , respectively.
Experiment 4B: Waterflood
The images were reconstructed using the proprietary software provided by Zeiss, with a non-
local means filter (AVIZO 9.0) and segmented with the trainable Weka segmentation tool
(Arganda-Carreras et al., 2017) provided by Fiji (Schindelin et al., 2012). The air-phase could
clearly be distinguished from the oil-phase. However, the contamination impeded automated
analysis of wettability as the air may lead to oil-layer formation along its interface and impact
the results (Scanziani et al., 2018). The contact angle was therefore measured manually, only
at locations where no air was observed in the surrounding volume, following the procedure
described in Andrew et al. (2014). Due to the high oil-wetness of the sample, leading to
fluid/fluid interfaces occurring predominantly in thin throats which are difficult to assess only
21 measurements were obtained.
3.2.3 Spontaneous (Amott) imbibition
Amott spontaneous imbibition tests are simple experiments conducted at the core scale, which
are often used for wettability determination (Amott, 1959; Mason & Morrow, 2013). In this
test, an oil-saturated rock sample is placed into a vessel containing brine as shown in Figure
3.10a. In water-wet and mixed-wet samples the water would start to spontaneously imbibe into
the rock sample replacing the oil, which then gets produced. In an Amott vessel the produced
oil is collected and monitored over time. The production rate and the overall production are
used as a measure for wettability (Ma et al., 1999). However, this characterization is often
insufficient (Chapter 2.1.4). In order to assess possibilities and limits of this method, core-scale
Amott spontaneous imbibition tests were monitored with µCT.
Figure 3.10: In an Amott test an oil saturated core sample is placed into a vessel filled with
brine. The oil produced to spontaneous imbibition accumulates at the top of the sample (a;
adapted from Bartels et al., 2017c)
For this purpose, an Amott cell made of crude-oil-resistant and X-ray-transparent material was
used (Figure 3.10b). It is crucial for µCT measurements to avoid the movement of the sample
during a scan, as this may cause image artefacts. Therefore, the samples were fixed with a
holder. Contact points of the sample with the holder, however, may influence the behaviour of
the oleic phase during the spontaneous imbibition process. For this reason, the contact points
were made of water-wet glass to prevent oil from spreading along the holders (Figure 3.10c).
To get a better understanding of the underlying pore-scale processes of spontaneous imbibition

additional experiments on smaller samples (mini-plugs) were conducted to obtain a higher tem-
poral resolution. As for the core-scale Amott test the cell constructed for these experiments was
made of crude-oil-resistant and X-ray-transparent material able to contain a holder, designed
with a low contact area towards the sample (Figure 3.10d).
List of experiments
Two different types of Amott spontaneous imbibition experiments were conducted in this study:
Core-plug test (5A – C) and mini-plug test (6A – B). The core plug experiments included
one water-wet sample and two mixed-wet samples, which were prepared following a different
experimental protocol. The mini-plug experiment was conducted with the same fluid-fluid-rock
system as the mixed-wet core-plugs. Nevertheless, a different wettability was observed. In
order to exclude the impact of the different experimental protocols followed for the different
sample sizes (such as the elevated temperature during the centrifugation of the core plug) an
additional experiment was performed on the mini-plugs with a sample aged for 12 h at 40 C.
Table 3.8 shows a list of Amott tests conducted.
Table 3.8: List of spontaneous imbibition experiments performed in this study.
Nr description sample brine oil
Core-scale Amott test
5A Water-wet system Ketton (plug) FW1 Decane (doped)

5B Mixed-wet systems (standard) Ketton (plug) FW1 Crude B (doped)
5C Mixed-wet systems (modified) Ketton (plug) FW1 Crude B (doped)
Dynamic imaging of spontaneous imbibition
6A Non-aged Ketton (mini-plug) FW1 Crude B (doped)

6B Aged Ketton (mini-plug) FW1 Crude B (doped)
Sample preparation and experimental protocol
The purpose of these experiments is the evaluation of the sample handling protocol commonly
used in Amott spontaneous imbibition tests. For a proper assessment, a detailed description
of each single step as well as a clarification of the differences between each experiment are
important. Figure 3.11 illustrates the different protocols used in these experiments.
Figure 3.11: Experimental protocols for core-scale and pore-scale spontaneous imbibition tests.
Core-scale experiments represent a standard experiment for the oil industry and therefore,
equipment has been optimized for a high number of repetitions at the corresponding sample
size. Pore-scale experiments however, are often addressing specific questions which can best
be answered by studies dedicated to this particular length scale. For this reason, equipment is
often custom-made and only optimized for a specific scientific question. Hence, experimental
protocols may differ from the larger scale equivalent.
Nevertheless, for each Amott experiment the following steps need to be covered: First, a dry
scan needs to be obtained as a reference. Second, the sample needs to be saturated with an
aqueous phase. Third, the sample needs to be desaturated with an oleic phase. Fourth, the
sample is placed into the spontaneous Amott-cell and the experiment is conducted. Further-
more, some additional steps as ageing or the removal of oil adhering to the peripheral surface
of the rock may be considered between the third and fourth step.
Experiments 5A – C: Core-scale Amott test
For the core-scale plugs (diameter ø = 2.5 cm and length L = 5 cm) first a dry scan was
conducted. The core-scale plugs were then placed for saturation in a desiccator in which a
vacuum (1034 Pa) was created. The FW1-brine (Table 3.8) added to the desiccator was de-
aerated prior to the saturation of the sample. Subsequently, the rock samples were submerged
in de-aerated brine in a bottle sealed off by a rubber stopper, and exposed to 3 MPa of pressure
for two hours to dissolve gas bubbles that may have remained in the pore space after saturation.
Afterwards, the core-scale plugs were desaturated with n-decane (5A) or crude B (5B – C) in
a centrifuge (URC-628, 129 Coretest Systems Inc., used at 3500 RPM for SCAL plugs) for 24
hours to reach initial water saturation Sro . To counter the gradual heating during centrifugation
the temperature was kept constant at 40 C.
Even in this short time, centrifuging at elevated temperature caused ageing of the crude-oil
saturated rock as indicated by the corresponding production curves (Chapter 4). This lead
to two mixed-wet to oil-wet crude-oil saturated samples and one water-wet core-scale plug
saturated with n-decane.
The next step within the standard experimental protocol for core-scale spontaneous imbibition
tests is the rolling of the sample on a tissue saturated with the oil used in the respective
experiment to remove oil attached to the outside of the sample, which may falsify the monitored
production. The oil adhering to the outside of the sample may detach and accumulate at the
top leading to a production not related to the spontaneous imbibition of the aqueous phase into
the sample. To assess the effect of this step, this protocol was only followed in experiment 5A
(n-decane) and B (crude B). For experiment 5C (crude B) this step was skipped. Subsequently,
the sample was placed into a brine filled Amott vessel for µCT scanning as described in Section
3.3.3.
The top part of the core-plugs was scanned immediately (ca. 5 min) after the placement of the
sample into the Amott vessel and several more times at later stages of the experiment. The
scanning time for one full 3D image was 100 min. At the end of each experiment the entire
sample was scanned by obtaining 4 images along the core. The image voxels length achieved
was 14.25 µm.
Experiments 6A – B: Dynamic imaging of spontaneous imbibition
The mini-plug samples (diameter ø= 4 mm, length L = 20 mm) were first fitted into a sleeve.
The sleeve is required for the placement on a core-holder for later desaturation. As for the
core-scale plugs the mini-plugs were then saturated in a desiccator under vacuum by slowly
filling it with de-aerated FW1-brine (Table 3.8). Subsequently, the sample was placed into a
brine filled vessel, which was sealed and pressurized to 1034 Pa for two hours.
The mini-plug samples were then mounted on a core-holder similar to the pump based set-
up described in Chapter 3.3.2. To fix the sample and to avoid by-pass flow the sleeve was
pressurized with distilled water. In this set-up, the sample could be scanned prior and during
desaturation, allowing a quality control of the prepared samples prior to the experiments. The
dry scan was obtained after saturation. Since in these experiments the oleic phase was doped,
the contrast between brine and rock was sufficient for the segmentation of the rock. Afterwards,
the sample was desaturated with crude oil at a flow rate of 4 mL/min (34.8 PV/min) for 5 min.
Subsequently, the core-holder was submerged in crude for dismantling to avoid the rock getting
in contact with air. The mini-plugs were stored in crude at ambient conditions until they were
placed into the mini-Amott cell described in Section 3.3.3 for the experiment 6A. In contrast
to the core-plug experiments, the mini-Amott samples showed a rather water-wet behaviour.
To exclude the possibility of an effect due to the elevated temperature during centrifugation
one sample was aged for 12 h at 40 C in an oven (6B) prior to the spontaneous imbibition
experiment.
The onset of imbibition processes in the small plugs was scanned with a high temporal resolution
of 15 s per 3D-image with a voxel length of 13 µm. Later the resolution was increased to a
voxel length of 6.7 µm with a scanning time of 70 s per 3D-image. The time difference between
the actual start of the Amott-experiment and the scanning was approximately 1 min.
Image analysis
The scans were reconstructed using dedicated reconstruction tools in the Acquila software
package from XRE. Visualization and additional post-processing of the data were done with
Avizo 9.2.0. The images were first filtered with a non-local means filter and subsequently
segmented by comparing the data of the dry scan with the dynamic scans captured during
the spontaneous imbibition experiment, following the procedure described in Experiments 1A
– 2F and as illustrated in Figure 3.9. Pore size distribution and the pore occupancy by oil was
computed using the granulometry algorithm from GeoDict (Hilpert & Miller, 2001).
3.3 Atomic force microscopy (AFM)
Traditional microscopes create a magnified image by focusing electromagnetic radiation, such as

photons or electrons. Atomic force microscopes, in contrast image mechanically. An atomically
sharp tip, attached to the end of a cantilever rasters a surface. The tip interacts with the
underlying surface through various intermolecular forces. These forces lead to a bending of the
cantilever, which is monitored by a laser. This technique allows a very precise 3D visualization
of the topography of a surface and also the detection of various physical properties.
Physical properties, such as sample stiffness or adhesion forces can be obtained by force spec-
troscopy. A piezoelectric transducer moves the tip towards the sample until the deflection signal
encounters a bending of the cantilever due to contact with the sample and is then retracted.
The nature of the deflection of the cantilever right before and after contact gives detailed in-
sights into molecular interactions. The recording of a force distance curve as well as other
properties which can be obtained via this measurement are explained in detail in Section 3.3.1.
While force spectroscopy is a great tool for studying molecular interactions, the AFM was
initially developed as an imaging tool to magnify vertical surface features of an object. The
first AFM introduced by Binnig et al. (1986) was based on the concept of a stylus profiler, in
which a stylus scratches over a sample moving in x-y direction while its deflection is monitored.
With this imaging technique atomic resolutions could be obtained. However, one disadvantage
3.3. Atomic force microscopy (AFM) 65
of this measuring technique nowadays referred to as “contact mode”, is that larger features may
lead to collision with the tip causing probe bending and resulting in image artefacts (Eaton
and West, 2010; Maver et al., 2016). With the introduction of the “tapping mode”, in which
the cantilever is oscillated between imaged points, these artefacts could be avoided (Garcia &
Perez, 2002; Johnson et al., 2012; Magonov et al., 1997; Martin et al., 1987; Müller & Engel,
1997; Stark et al., 2001). Recent developments allow faster scanning rates and hence, to obtain
actual force distance curves for each measured pixel, giving topographical images and detailed
material properties at the same time. These AFM imaging techniques will be introduced in
detail in Chapter 3.3.2.
In summary, AFM allows the investigation of both molecular interactions and surface topog-
raphy effects. For studies in petroleum engineering this tool is often used to investigate the
interaction of crude oil and constituent components with mineral surfaces in the presence of
different aqueous solutions (Matthiesen et al., 2014; Mugele et al., 2016) or to study adsorption
and the presence of organic components in rock (Buckley & Lord, 2003; Kumar et al., 2005a).
Other studies illustrate the capability of the AFM to assess the wetting of a surface by a fluid
(Dickinson et al., 2016b; Giro et al., 2017; Liimatainen et al., 2017). In this work, this approach
is used to investigate the formation of water and oil films within the rock.
3.3.1 Force spectroscopy
Force spectroscopy allows the investigation of sample properties such as stiffness or molecular
interactions e.g. via the detection of adhesion forces. The tip is moved towards the sample
until contact is achieved and then backwards by piezoelectric transducers while the deflection
of the cantilever is recorded.
Figure 3.12 displays the approach and retraction of a cantilever to a surface while recording
a force distance curve. When the tip approaches the sample, initially the cantilever does not
bend and no forces can be detected by the laser deflection (0-point) (Figure 3.12-1). As the
separation distance decreases, forces such as electrostatic, Van der Waals or forces related to the
Derjaguin-Landau-Verwey Overbeek (DLVO) theory of colloids in fluids (Chapter2.1.1) exceed
the spring constant and the tip either jumps in contact with the sample (Figure 3.12-2) or is
repelled.
Figure 3.12: Force distance curve as obtained during force spectroscopy. Initially the cantilever
is in a relaxed shape (1) until surface forces of the scanned solid start to attract or repel the
tip. In the presented case the tip is attracted to the surface and therefore jumps into contact
(2). When the cantilever is moved closer to the surface the cantilever is pushed back into the
relaxed shape (3). This point is used as a height reference. The tip is then pushed further into
the sample until a chosen force (set point) is detected (4) and the cantilever starts to retract.
The tip experienced adhesion forces and gets released, once the spring forces of the cantilever
exceed the adhesion (5).
From the point of contact onwards the cantilever bending towards the sample is progressively
reduced, as the sample continues to move closer, until the laser spot reaches the 0-point (Figure
3.12-3). Further movement then opposes contact, which is indicated by the cantilever bending
in the opposite direction (Figure 3.12-4). When the cantilever is retracted the bending of the
cantilever reverses again, while the intermolecular forces hold the tip in contact with the surface
until the forces applied by the motion control of the cantilever exceed the adhesion forces and
the cantilever jumps back out of contact with the surface (Figure 3.12-5) (”A practical guide
to AFM force spectroscopy and data analysis”, 2018; Kumar et al., 2005a).
The force distance curve recorded with the AFM displays the deflection versus the position
of the z-piezo. To obtain a force distance curve these values need to be corrected by cali-
brating the deflection sensitivity of the detector response to the movement of the piezo and
the spring constant of the cantilever. The deflection sensitivity and the spring constant can
be calibrated using different approaches depending on the accuracy required. An overview of
different approaches and their sensitivities are presented in Ohler (2007). A corrected force
distance curve is shown in Figure 3.13a. Yet, this does not reflect the actual force distance
curve of one molecule approaching a surface, but only specific parts as illustrated in Figure
3.13b. The approach curve gives us the force vs. distance for a chosen system up to the snap-in
point (Figure 3.13-2).
Figure 3.13: Force distance curves obtained with an atomic force microscope based on the
calibrated deflection sensitivity and spring constant (a) do not reflect actual force distance
curves of approaching molecules (b). The approach curve gives us the force vs. distance for a
chosen system up to the snap-in point.
At this point the attraction of the surface towards the cantilever exceeds the spring constant
pulling the tip directly into contact (Figure 3.13-3). The height of the contact point can be
obtained by further moving the tip towards the sample until the deflection d 0. At this
point the position of the piezo represents the sample height. When the tip is further pushed
into the sample to reach a chosen set point (Figure 3.13-4) the stiffness can be detected. The
slope between the contact point and set point is steeper for stiff matter than for soft matter.
The retraction curve on the other hand gives the part of the force distance curve up to the
maximal attraction which correlates with the adhesion force Fadh . Once this point is reached
the cantilever jumps back to the start position with d 0. The force vs. distance for r between
the snap-in point (Figure 3.13-2) during the approach and the snap-off point during retraction
(Figure 3.13-5) cannot be determined.
This illustrates that force distance curves obtained with AFM give insight into the nature
of interaction between two solid substances (tip and sample surface). Different approaches
have been developed to also study the interaction between a fluid and a solid: Dickinson et al.
(2016b) measured the interactions by using a colloid probe dipped into a fluid (crude). However,
with this approach meniscus forces overlay the actual force distance curve and therefore are
more difficult to interpret. In addition, the amount of fluid on the probe gets smaller with
each force distance curve as fluid is progressively left behind at the surface. Liimatainen et al.
(2017) attached the droplet by shooting sub-nl droplets using the piezoelectric microdispenser
onto a disc. The droplet was refilled prior to each measurement and its volume was monitored
through the force given by the sensor. However, this method may not be suitable for more
viscous crude.
Another way to study the attraction between fluids and solids is by focusing on their compo-
nents. In a different approach Dickinson et al. (2016b) attached crude oil to a colloidal probe
by drying. The advantage of this method is that the tip is functionalized with a variety of
crude oil components. However, volatile components will escape during the drying process and
surface-active components may adhere with their surface-active tail pointing to the tip instead
of the sample surface. Another approach is to focus on a model system e.g. by functionalizing
the tip with hydrophobic –CH3 or hydrophilic –COOH groups representing surface active and
non-surface-active components of the crude. These groups can be attached to a tip surface
following the method described in Frisbie et al. (1994). By measuring the adhesion force of
these components to a solid within another e.g. aqueous fluid this method is capable of giving
a local indication of the wettability within this particular system.
For this work, a NanoWizard 4 Nanoscience AFM (JPK instruments) and a Bioscope V (Bruker)
were utilized. Both AFMs were equipped with additional top down optics to determine the
position of the measurements.
3.3.2 Topographical imaging
The AFM was initially developed as an imaging tool. But, it is more similar to a surface profiler
than traditional optical or electron microscopes, as it can measure along the z axis to obtain
topographical information of an object. The largest AFM images have a size around 100 µm x
100 µm. Yet, AFM is also capable of resolving atoms at the smallest scale.
Throughout time different imaging techniques have been developed. ‘Force mapping’ used in
this work, provides an image and full force distance curves at each pixel at the same time allow-
ing a more rigorous analysis of the sample. With this method the distance of the tip approach
and retraction as well as the force applied can be controlled. However, these measurements
take a long time. JPK instruments has developed an enhanced force mapping technique (QITM
mode) with a tip movement algorithm which controls the exact x-y position during a force
distance measurement and allows to overshoot in z when moving to the next point, leading to a
faster imaging as illustrated in Figure 3.14. This technique allows the measurement of height,
stiffness and adhesion directly and in parallel and also to introduce own functions as a full force
distance curve is recorded for each data point (”QITM mode - Quantitative imaging with the
NanoWizard 3 AFM”, 2018).
Figure 3.14: In QITM mode the tip does not move in x-y direction while a force distance curve
is obtained. Yet, the movement algorithm allows an overshoot above the set z-distance, which
enables a fast data acquisition (adapted from , ”QITM mode - Quantitative imaging with the
NanoWizard 3 AFM” (2018)).
The limitations of this technique are discussed in Appendix C.

3.3.3 Wettability assessment by AFM
In this work, atomic force microscopy was used to investigate the formation of water- and
oil-films within the rock. Other studies demonstrated the capability of the AFM to assess the
wetting of a surface by a fluid (Dickinson et al., 2016b; Giro et al., 2017; Liimatainen et al.,
2017). In this work, this approach was developed further for applications on rock samples.
List of experiments
The list of all AFM experiments conducted in this work is shown in Table 3.9.
In a first step the noise level of the AFM was tested on a flat muscovite surface (7A) and on
roughened calcite (7B). Furthermore, AFM reference images of rock surfaces (8A – B )were
obtained. The focus of this study was the detection of fluid films in rock in the presence of
two immiscible fluids (11A and 11B). For this purpose, samples were first submerged in one
fluid and then placed in the other. The procedure was first tested on model surfaces (calcite).
The fluid film left behind on the sample was identified based on the change in the shape of
the corresponding force distance curve (experiments 10A – D). Reference force distance-curves
were obtained by measuring on the same substrate within each respective fluid (9B and 9C)
and air (9A).
To assess the importance of different crude-oil components and different brine composition,
adhesion forces were investigated using probes functionalized with –COOH and –CH3 groups
within distilled water and seawater (experiments 12A – C).
Table 3.9: List of AFM experiments performed in this study.
Nr title sample fluid probe
Reference – Topographical imaging
7A Noise sensitivity Muskovite Air/SW Si

7B Calcite (roughened) Calcite Air Si
Roughness on rock surfaces
8A Ketton roughness Ketton Air Si

8B Estaillades roughness Estaillades Air Si
References – Force spectroscopy
9A Reference calcite in air Calcite Air Si3 N4

9B Reference calcite in SW Calcite SW Si3 N4
9C Reference calcite in n-decane Calcite n-decane Si3 N4
Fluid films on calcite
10A Calcite split in SW submerged in n- Calcite (split in SW) n-decane Si3 N4

decane
10B Calcite split in n-decane submerged in Calcite (split in SW Si3 N4
SW decane)
10C Calcite split in crude submerged in Calcite (split in SW Si3 N4
seawater crude A)
10D Calcite split in seawater submerged in Calcite (split in SW, SW Si3 N4
crude, aged and measured in seawater aged in crude A)
Fluid films in rock
11A Water film thickness in water wet Ket- Ketton after n-decane Si3 N4
(4A) ton (DB and n-decane) drainage
11B Oil film thickness in aged Ketton (DB Ketton after DB Si3 N4
(4B) and crude) waterflood
Adhesion of crude oil-components
12A Adhesion – CH3 in seawater Calcite SW –CH3

12B Adhesion – CH3 in distilled water Calcite dist. water –CH3
12C Adhesion –COOH in seawater Calcite SW –COOH
Sample preparation
Prior to each experiment the deflection sensitivity and spring constant k of the cantilever were
determined using either the thermal noise tuning method (Butt & Jaschke, 1995; Hutter &
Bechhoefer, 1993) for experiments 7A–C, 8A–D and 9A–B or the Sader method (Sader et al.,
1999) for experiments 10A-D.
The thermal noise tuning method requires the determination of the deflection sensitivity ds
through a force distance measurement on a hard substance and a power spectral density analysis
to determine the mean square displacement at the end of the cantilever due to Brownian motion
xz2y. The spring constant k can be determined by
k 0.971kbT xz2y, (3.3)
with the Boltzmann constant kb and temperature T .
For the Sader method the length L and width b of the cantilever and its thermal noise spectrum
need to be measured. Following the equation:
k 0.1906ρbLωf2 Γfi pQf q, (3.4)
in which ρ represents the density of the bulk fluid, ω f and Qf the resonant frequency and
the quality factor of the fundamental resonance peak and Γfi the hydrodynamic part of the
function, k can be determined (Sader, 1998). The advantage of this method is that the tip does
not need to get in contact with the sample prior to the experiment and therefore is preferable
for measurements with functionalized tips. The sensitivity can be obtained from the amplitude
measured at a specific frequency.
After this step, the experimental workflows were varied as described below:
Experiments 7 A – B: Reference – Topographical imaging
For experiment 7A a muscovite sample was cleaved and measured with a silicon nitride tip by
applying contact - and QITM mode. As cleavage of muscovite reveals an atomically flat surface,
this experiment is suitable to assess the noise level.
In experiment 7B a calcite sample was cleaved in air and roughened with sand paper. Consec-
utively, residues were blown away with nitrogen. The surface topography was measured with
QITM -mode and a silicon tip. This sample was used to assess the effect of surface roughness on
the image quality.
Experiments 8 A – B: Roughness on rock surfaces
The rock samples (Ketton and Estaillades) were drilled to I = 4 – 6 mm and 2 cm long sub-
samples and then cleaned using isopropanol (99%). The measurements were either conducted
at approachable pores at the outer space of the sample or at breakages, where special attention
was payed to approach an original pore surface instead of a freshly broken one. The surface
topography was measured within an area of 10 µm x 10 µm with QITM -mode using a silicon
tip.
Experiments 9A – C: References – Force spectroscopy
Properties such as wettability or interfacial tension are sensitive to contaminations (Georgiadis

et al., 2011). Therefore, all glassware was cleaned with a detergent, placed in an ultrasonic bath
for at least 10 min, rinsed with isopropanol and dried in an oven. The AFM chip was treated
with ultraviolet light prior to the experiment to decompose organic molecules (Vig, 1985).
For the reference scan in air (9A) a calcite (island spar) was cleaved to obtain a fresh uncon-
taminated surface. For the reference scans in fluid the calcite sample was placed in a vessel
filled with seawater (SW, Table 3.6) or n-decane respectively and then cleaved within the fluid
to avoid surface modification due to contact with air.
These experiments were conducted with silicon nitride probes.

Experiments 10A – D: Measurement of fluid films on calcite
Prior to the sample preparation all dishes were cleaned as described for experiments 9A–C.
With the objective to create a fluid film at the calcite surface the calcite was first split in one
fluid (10A: SW, Table 6; 10B: n-decane), which was then flushed away with the other fluid
(10A: n-decane; 10B: SW) before submerging the sample into the same fluid (10A: n-decane;
10B: SW) within the measurement vessel in which the force distance curves were obtained.
In order to study the adhesion of oil on a carbonate surface (10C), a calcite sample was cleaved
in crude A. This brought the crude oil into immediate contact with the surface and could then
be considered as aged-equivalent. For comparison, an aged sample was prepared by splitting
calcite in seawater (10D, SW, Table 3.6). Consecutively, the sample was submerged in crude
A and aged for 4 weeks at elevated temperature (70 C and pressure (3 MPa). Images were
obtained in QITM - mode. All experiments were conducted with silicon nitride probes.
Experiments 11A – B: Measurement of fluid films in rock
The samples used for this experiment were previously used in the µCT flooding experiments
4A and B. Afterwards, the sample was dismantled within the respective fluid (11A: n-decane,
11B: DB) and placed in the AFM holder for measurements. The experiments were conducted
with silicon nitride probes in QITM -mode.
Experiments 12A – C: Adhesion of crude oil components
The calcite sample was cleaved in air prior to the experiment. Consecutively, a seawater droplet
was placed on top of the surface. The force distance curves obtained with functionalized tips
(12A: -CH3 ; 12C: -COOH) were measured at different locations. Subsequently, the seawater
droplet was replaced by distilled water by thorough flushing followed by additional force distance
curve measurements (12B: -CH3 ; 12D: -COOH)
Post processing
Experiments 7A – B and 8A – B: Roughness on reference samples and rock surfaces
Roughness is the deviation of the topography of a surface from the ideal approximation. In
the context of this study roughness is the surface topography below or close to µCT-resolution.
As this topography can be resolved by atomic force microscopy, the deviation of the ideal
approximation may be assessed. The ideal for the studied area of 10 µm x 10 µm is considered
flat.
The roughness of the studied system is given by the frequency and the deviation z from the
0-level. The 0-level was defined with JPK Data Processing software (version spm 6.0.63, JPK
n1 °ni1 |zi|,
b°
instruments) by a plane fit through the image. In addition, roughness average Ra
the root mean square average roughness Rq 1
n
n

2
i 1 zi and the peak to valley roughness
Rt |maxpzq| |minpzq| were determined (Gadelmawla et al., 2002).
As muscovite mica used in experiment 7A provides an atomically smooth surface the roughness
detected in this system represents the noise level. Experiment 7B was visually inspected to
assess additional noise and image artefacts introduced by a rough topography.
The topographical image obtained in experiment 8A was further analysed with Geodict 2015
(Math2Market). A morphological drainage simulation, calculating fluid distribution for dif-
ferent capillary pressure by fitting spheres into the system as described in Section 2.1.3 was
applied on the 3D surface obtained (Hilpert & Miller, 2001). The fluid distributions obtained
from the simulation were filtered with a median filter. In this work, this approach was used to
assess the coverage of the surface by water and oil.
Experiments 9A – C: References – Force spectroscopy
The force distance curves obtained in experiments 9A–C were processed by applying the work-
flow shown in Figure 3.15.
Figure 3.15: Processing of force distance curves. First the zero point was found by determining
the zero level within the no-contact distance (a). In a second step the deflection d [V] was
converted to the deflection d [nm] with the help of the deflection sensitivity ds [nm/V] (b).
In the third step, the true distance r was calculated by subtracting the deflection d from the
measured distance z (c). In a last step the force was computed by multiplying the deflection d
with the spring constant k (d).
In the first step, the obtained force distance curve was shifted to the zero position (Figure
3.14a). For this purpose, the zero force-level must be determined based on the region, in which
the cantilever is not yet affected by the surface. The contact point z = 0 was determined as
the point within the region of intermittent contact with F = 0. (In case of a repelling surface
this location was found by extrapolating the slope when in contact.) In the second step (Figure
3.15b), the deflection d [V] displayed on the y-axis was converted to deflection d [m] by using
the sensitivity ds [m/V].
Within the third step, the position of the tip r was computed based on the piezo position z
and the cantilever deflection d [m] from
h z dpz q, (3.5)
where h is the position of the tip relative to the contact point, z represents the position of the
piezo and dpz q the deflection of the cantilever at this point in m (Figure 3.15c). The last step
(Figure 3.15d) was the conversion of the deflection dphq into force F phq following equation:
F phq kdphq, (3.6)
The force distance curves measured in experiments 9A and 9B were used as references for
experiments 10A–D and 11A–B and therefore did not require further processing.
The main objective of these experiments was the determination of the fluid film distribution
and thickness. This was possible as the tip responded with deflection when passing through
the fluid-fluid interface. The distance between this response and the response to the surface
contact represents the film thickness.
Following the analysis for individual force distance curves described for experiment 9A and B
and shown in Figure 3.14, a few force distance curves were selected and processed manually.
Furthermore, an automated workflow was developed and applied on the full QITM -images of
experiment 10A and 11A. First, the zero-level of each force distance curve was identified. As
the water-film caused an attractive response, the fluid-fluid interface was determined using an
algorithm obtaining the largest contiguous negative area of the approach curve provided by
JPK Data Processing software (version spm 6.0.63, JPK instruments). The “upper bound”
value given by this algorithm represented the fluid-fluid interface. The “lower bound” value
represented the surface of the sample. The difference between the two values represents the
film thickness. As the surface not covered by water showed only little attraction the largest
contiguous negative area algorithm computed a false film thickness for these locations. There-
fore, a mask obtained from the adhesion image with a thresholding tool offered by ImageJ was
applied. The topographical image and film thickness image were combined to a voxel based 3D
image with Matlab (R2014a, MathWorks). The result was visualized with AVIZO 9.0 (Thermo
Fischer Scientific). The full workflow is shown in Figure 3.16.
Figure 3.16: 3D fluid film visualization workflow.
To remove noise a median filter was applied on the generated 3D image. The 3D image was used
for contact angle determination following the approach described in Scanziani et al. (2017).
Experiments 12A – D: Adhesion of crude oil components
To obtain the desired adhesion force of specific components towards a surface, it is not sufficient
to measure the adhesion of the tip on the sample as multiple molecules may interact with the
surface. However, based on the JKR theory the adhesion force F adh is related to the surface
energy W given in the form,
Fadh 32 πrW, (3.7)
where r is the radius of the tip. (Kumar et al., 2005a). For experiments 10A–B this value was
computed to compare the effect of the chemical composition of the fluids on wettability.
3.4 Supporting data and experiments
In addition to the experiments listed above additional measurements were conducted for further
integration of the studied systems and respective length scales assessed. A core-flood experiment
3.4. Supporting data and experiments 79
was conducted with the same COBR system as in 1A (Section 3.5.1). The processing of
an external data set on flow dynamics in water-wet systems, analysed in comparison to the
experiment on flow behaviour in a mixed-wet system (1A) is explained in Section 3.5.2. Tube
tests supporting the surfactant experiment (3A) are described in Section 3.5.3.
3.4.1 Core flood experiment (S1) for comparison with pore-scale
flooding experiment (1A)
The SCAL experiment to determine relative permeability was performed in a custom-built X-

ray saturation measurement apparatus at Shell (Berg et al., 2008; Berg et al., 2016; Kokkedee
et al., 1996) following the steady state method (Dake, 1983; McPhee et al., 2015) in which the
two fluids, brine and oil, are co-injected (Section 2.1.4).
First the sample was saturated with brine under vacuum. Then, the sample was mounted inside
an X-ray transparent core holder and placed in the flow apparatus for steady-state relative
permeability measurement. The sample was desaturated with crude A by flooding and aged
at 70 C and 30 MPa for 1 week. The measurements were conducted at a constant flow rate,
where the fractional flow fw was systematically changed from 100% crude oil to 100% brine in
10 steps (fw1 0, 01; fw2 0, 05; fw3 0, 1; fw4 0, 3; fw5 0, 5; fw6 0, 7; fw7 0, 9; fw8
0, 95; fw9 0, 99; fw10 1q. At each step saturation (and spatial profile along the core)
and phase pressures were recorded after steady-state was reached. The SCAL data (pressure
drop over the core and in situ saturation profiles at each fractional flow step) were matched
numerically using the Shell in-house simulator (MoReS) to estimate the relative permeability
as a function of saturation.
3.4.2 Data: Flooding experiment water-wet sample (S2) for com-
parison with mixed-wet systems (1A)
For comparison with the studied mixed-wet system, a water-wet data-set published in Singh et
al. (2017a) has been further analysed. The reference data-set represents a flooding experiment
conducted at the synchrotron beamline I13-2 of Diamond Light Source (UK). The Ketton rock
sample had a diameter of ø = 3.8 mm and a length of L = 10 mm. A 23 wt% KI solution has
been used as the aqueous phase and n-decane as the oleic phase. This rock-fluid-fluid system
shows a strongly water-wet rock behaviour. During the waterflood the brine phase was injected
into the oil-saturated sample with a flow rate of 44.75 nL/min (0.002 PV/min) corresponding
for 4.6 h. The images with a voxel size of 1.64 µm obtained in 38 s have been reconstructed,
filtered and segmented following the workflow described in Singh et al. (2017a). In this work,
segmented images were further processed following the procedures described in Section 3.3.2
(experiment 1A). Due to the low flow rate applied, only every 20th image was used for event
identification.
3.4.3 Static tube test (S3) conducted for surfactant flooding assess-
ment (3A)
The oil recovery efficiency of surfactants was assessed by an equilibrium phase behaviour test.
In this test, the surfactant solution and the oil were mixed in test tubes by rigorous shaking,
and let stagnate while the emulsification took place. This process lasted for approximately 1h
for the surfactant/oil/water system used in this study (Huh, 1979; Unsal et al., 2016). The
equilibrated phase behaviour test tubes were scanned in a benchtop µCT-scanner Xradia (Zeiss)
in order to study the composition of the (fully) equilibrated middle microemulsion phase.
Chapter 4
Core-scale wettability characterization
Wettability at the core scale is usually not assessed directly, but rather indirectly via exper-
iments determining capillary pressure and relative permeability. These properties are crucial
for field-scale modelling of oil reservoirs. However, as these experiments are costly and time
consuming, more effort has been put into predicting capillary pressure and relative permeability
with computational simulations (Section 2.1.4). For these simulations wettability determined at
smaller length scales, e.g. through contact angle measurements, represents an input parameter.
Yet, there is a lack of understanding on how the wettability determined at these length scales
impacts the core-scale output. While relationships between wettability and capillary pressure
as well as relative permeability have been established with various model systems, they fail
to predict the impact of wettability for realistic cases (Section 2.1.4). In this section, the im-
pact of wettability on two classical core-scale experiments, Amott spontaneous imbibition tests
(Section 4.2) and flooding experiments (Section 4.3), were analysed by relating the core-scale
results with pore-scale processes. It was found that the sample initialization has major impact
on the wettability observed at the core scale, which is elucidated in Section 4.4 .
81
82 Chapter 4. Core-scale wettability characterization
4.1 Amott spontaneous imbibition tests
Amott tests are experiments assessing wettability that are straightforward to conduct, but
difficult to interpret (Section 2.1.4). In these tests an oil-saturated sample is placed into a
vessel filled with brine. While the water-spontaneously imbibes the oil-production is measured.
A detailed description of experimental procedure can be found in Section 3.3.3.
Figure 4.1 shows the cumulative production versus time for all SCAL-plugs Amott experiments
(5A-C) conducted in this work. As expected, the water-wet sample shows more production than
either mixed-wet sample. Additionally, the wettability states could be confirmed by the drops
emerging from the edge of the samples. In experiment 5A, drops with a contact angle below 90
(obtained through the water-phase) emerged and snapped-off continuously, whereas the drop
emerging from the mixed-wet sample spread over the rock surface. In the following sections
pore-scale processes related to the macroscopic observations will be discussed for the water-wet
(Section 4.1.1) and mixed-wet case (Section 4.1.2). Furthermore, the sensitivity of the results
to sample handling protocols (Section 4.1.3) as well as the possibilities and limitations of this
method for wettability assessment (Section 4.1.4) will be elucidated.
Figure 4.1: In Amott tests, an oil-saturated core-plug is placed into a brine filled vessel. If the
brine spontaneously imbibes into the rock, the oil is produced. Image (a) shows the formation
of an n-decane-bubble during experiment 5A (water-wet system). (b) shows the formation of an
oil-bubble during experiment 5B. The cumulative production vs. time curves for all SCAL-plug
Amott experiments are shown in (c; adapted from Bartels et al., 2017c).
4.1. Amott spontaneous imbibition tests 83
4.1.1 Water-wet systems
In the water-wet samples (experiment 5A), we observed the formation of preferential production
sites as shown in Figure 4.2.
Figure 4.2: Spontaneous imbibition within a water-wet rock (experiment 5A). The difference
between the initial image (a) and the last image of the scanning sequence (b) shows an asym-
metric imbibition front. The differential image (c) highlights the locations at which changes
occurred. This asymmetric behaviour may be a result of preferential production sites, at which
n-decane drops repeatedly emerged at the same location (d; adapted from Bartels et al., 2017c).
During the measurement droplets formed repeatedly at the same location at the top of the
sample. After this production site stopped a new one emerged just below the field of view.
This behaviour leads to an asymmetric spontaneous imbibition front within the rock. The
asymmetry is an unexpected phenomenon. The imbibition of water into the sample is controlled
by the wetting state and the pore-inlet diameter of the rock. For Ketton rock, both parameters
are expected to be homogenous across the sample (see Chapter 3.2). Hence, an inward-moving
circular front along the length of the core would be anticipated.
From this experiment it can be concluded, that even little differences within the rock structure
can lead to the formation of production sites and an asymmetric displacement within the pore-
network. This causes a misalignment of the macroscopically observed production (33%) and the
observed pore-scale displacement (21%) within the field of view (height h = 1.7 cm, diameter
ø=2.5 cm). Though pore-scale processes affect the imbibition front across the sample, the
macroscopic production curve still reflects the wettability of the sample (strongly water-wet)
in the correct manner.
4.1.2 Mixed-wet systems
The mixed-wet SCAL sample (experiment 5B) showed a small production of 5% within the first
10 h. The oil drop emerging during the experiment at the top of the sample showed a strongly
oil-wetting behaviour. Within the rock isolated changes in fluid occupancy, associated with
small water-filling events replacing oil, were observed throughout the sample as shown in Figure
4.3. However, based on the pore size - and fluid distributions obtained with a granulometry
algorithm (Section 3.2.2) to assess the water-filling locations and associated volumes only a
volume decrease of 1% could be detected within the field of view (Figure 4.3). This contradicts
the production curve shown in Figure 4.1c. The change in oil-volume within the sample detected
through the macroscopic production (5%) does not match the volume-change observed on the
pore scale (1%). One explanation could be that the volume produced is coming from the bottom
part of the sample. In fact, a colour gradient could be observed along the outer surface of the
sample (Figure 4.1) hinting towards a higher oil-content at the top of the sample than at the
bottom.
4.1. Amott spontaneous imbibition tests 85
Figure 4.3: Spontaneous imbibition in mixed-wet rock (experiment 5B). The difference between
the initial (a) and the final (b) image of the scanning sequence indicates only small changes in
fluid distribution. (c) illustrates the water-filling events within the rock in 3D (adapted from
Bartels et al., 2017c).
Buoyancy could potentially explain this behaviour, but is often deemed unimportant for spon-
taneous imbibition processes in porous rock, especially on geologically short time scales as
months or years. The main reason is the (usually) small pore size of those materials, which
leads to a relative domination of capillary forces. However, with a peak pore-throat diameter
at 85 µm and some throats larger than 100 µm in diameter, this assumption may not hold
for Ketton (Section 3.1). The force balance between capillary and gravitational forces for this
specific system was computed:
F orce balance
∆ρghr
, (4.1)
2σcosθ
with ∆ρ being the density difference between the two fluids, g 9.81m{s2 the gravitational
constant, h the sample height, r the pore throat radius, σ the fluid-fluid interfacial tension and
θ being the contact angle. Unless the contact angle of the COBR system is in between 89 and
91 , which based on the images shown in Figure 4.3 is not the case for the presented experiment,
buoyancy does not have a significant impact (F orce balance 0.01). This statement is also
supported by the observations made from the final scan of the full sample shown in Figure 4.4a
(Bartels et al., 2017c).It shows a homogeneous distribution of the oil within the sample, which
hints to the fact that the observed production is not originating from within the pore-space.
Another explanation could be a thick oil-layer covering the surface. As this oil experiences no
capillary forces from the pore-space, buoyancy may lead to its accumulation at the top of the
sample and lead to a false production curve. This hypothesis is discussed next.
Figure 4.4: The final scan of the whole sample (experiment 5B) (a) shows a homogeneous
distribution of the oil phase. The initial (b) and final (c) fluid distribution within the field of
view monitored throughout the experiment illustrates that 1% of the oil-phase was produced
from the inside of the rock (adapted from Bartels et al., 2017c).
4.1.3 Impact of sample handling protocols
Sample handling protocols are known to have an impact on the outcome of Amott tests (Section
2.1.4). For instance, excess oil may affect the result. Usually, the excess oil from the outer
perimeter is removed by rolling the sample over an oil-wetted tissue prior to the Amott test
(Section 3.3.3). It is a delicate step which is difficult to reproduce. If the tissue is too wet, no
or little excess oil is removed. However, removal of fluids within the pore space, which may
occur if the tissue is too dry, must also be avoided. This represents a potential explanation
for the deviation between the high macroscopic production and the low water invasion within
the pore-space in experiment 5B (4.2.2). Following this argument, the rolling procedure would
have a large impact on the cumulative production curves and potentially also on the wettability
indices derived from them.
To investigate the effect of rolling on the production curves, another experiment without rolling
the sample (5C) was conducted. The production of experiment 5B was 50% lower compared
4.2. Steady-state experiments in mixed-wet systems 87
to 5C. In both cases the production did not match the changes of fluid distribution within
the pore-space. Based on the oil-production and assuming a perfect cylindrical shape of the
sample, an outer layer with a thickness of 80 µm may have been left behind after rolling the
sample over the tissue in preparation for experiment 5B. While on a perfectly flat surface this
would be unlikely, rough rock surfaces may facilitate such a layer. The grain size of Ketton
ranges from 100 µm to 1mm. If the grains are perfectly spherical, one can assume a height
variation along adjacent grains within the range of the corresponding radii - from 50 µm to 500
µm. Based on this an 80 µm residual layer seems reasonable.
4.2 Steady-state experiments in mixed-wet systems
Steady-state experiments are rather costly and time consuming. Therefore, a lot of effort is
put into the development of simulators with predictive capability. One method is pore-network
modelling, which simplifies the rock structure by converting it into a network consisting of pore-
bodies and pore throats. This type of simulator mostly captures static pore-scale processes
(quasi steady-state) but may also include dynamic processes (Section 2.1.4). To optimize
the simulator, it is important to identify processes which need to be reflected and processes
which may be neglected. This can be done by comparison with core-scale relative permeability
experiments and µCT-data. For water-wet samples, a good match was found between pore- and
core-scale data and static drainage simulators. However, for imbibition the results deviated.
The simulator did not correctly reflect the connectivity of the oil-phase (Berg et al., 2016). The
predicted connectivity was lower than in the µCT experiments. Ganglion dynamics lead to
reconnection of previous trapped clusters, though pathway flow was the flow regime controlling
the overall flux (see Section 2.1.3 and 2.1.4). In this chapter pore- and core-scale data will be
compared to assess whether this finding can be confirmed for more realistic mixed-wet cases.
First the relative permeability at the core scale will be discussed (Section 4.3.1). In Section
4.3.2, the associated pore-scale data will be introduced followed by the comparison of the pore-
scale information with the core-scale data (Section 4.3.3).
4.2.1 Relative permeability at the core scale
Relative permeability data are crucial for oil-field development plans, but so far, the relative-
permeability/saturation functions can only be obtained experimentally. To predict these a
better understanding of the underlying pore-scale processes is necessary. Various studies bridge
the pore scale and the core scale for water-wet systems, but only little is known on how pore-
scale processes in mixed-wet systems impact the relative permeability-saturation relationship
(Section 2.1.4)
In this work a core-scale steady-state SCAL experiment on a mixed-wet Ketton rock (S1) was
performed for comparison with experiments on smaller samples, which allow to resolve the pore
scale. The corresponding relative permeability curves obtained are shown in Figure 4.5.
Figure 4.5: Relative permeability curves obtained from a steady-state SCAL experiment (S1)
displayed on an arithmetic (a) and a semi-logarithmic scale (b). The results are typical for a
mixed-wet system leaning towards water-wet.
Following the guidelines to assess wettability by Blunt (2017) the water permeability end-point
max
with krw = 0.2 is at the upper limit for a water-wet system, but not yet typical for mixed-
wet systems. However, with a very low residual oil saturation of Sres,o 0.06 and a low water
relative permeability at low water saturations ( Swi 0.2 ) of krw pSwi 0.2q 0.03, the studied
system fulfils two of three criteria proposed to identify a mixed-wet system. From this analysis
we can exclude a strongly water-wet and a strongly oil-wet scenario. Therefore, the system can
be considered mixed-wet leaning to the water-wet side.
4.2.2 Impact of emulsion on the oil-phase connectivity
Experiment 1A was conducted with the same COBR system and with the same initializa-
tion method as the steady state SCAL experiment S1. The pore-scale processes observed in
experiment 1A therefore can be linked to the core-scale observations (S1).
Figure 4.6 shows the fluid distribution within the rock after ageing. Aside from the rock
(spherical grey grains), oil (black) and brine (white) a fourth phase was detected with grey
values ranging between black (oil) and white (brine). A rigorous analysis verified that this
fourth phase is a water-in-oil emulsion (Bartels et al. 2017b, Appendix D).
Figure 4.6: In the grey scale images obtained by fast µCT we see the rock (grey spheres), crude
oil (black) and brine (bright phase). Furthermore, a fourth phase with an intermediate grey
value appears in the pore space, which is an emulsion (adapted from Bartels et al., 2017b).
Within a water-in-oil emulsion the water phase forms micelles within the continuous oil phase.
Under flow conditions, these micelles are expected to be transported along with the oleic phase.
Therefore, the flow-behaviour of the detected emulsion is expected to be more similar to the
oleic phase than to the aqueous phase. However, in the following analysis both extremes, the
emulsion phase behaving in tandem with the oil and brine- phase, respectively are considered.
Figure 4.7 shows the distribution of the oil phase only (a) and the oil phase including emulsion
(b). The connected (red) and disconnected (yellow) fractions are indicated, respectively. (The
full sequences are shown in Appendix E.) Considering solely the oil phase the oil appears
predominantly disconnected throughout the experiment. Adding emulsion to the oil phase
leads to an increase in connectivity (Figure 4.7c). In both cases disconnection and reconnection
events, a behaviour referred to as ganglion dynamics, can be observed.
Figure 4.7: Connectivity indicates the possibility of connected pathway flow. The distribution
of the connected (red) and disconnected (yellow) oil phase is displayed for both, the oil phase
only (a) and the oil phase including the emulsion (b). In (c) the overall saturation vs. time is
displayed as well as the proportion of connected oil, oil and emulsion and total oil. The oil phase
only gets disconnected and reconnected with time independent of the behaviour of the overall
saturation, which indicates either a ganglion dynamic behaviour or connected pathway flow
through the emulsion phase. The Euler characteristic, a measure of connectivity, displayed
in (d) shows a highly disconnected oil phase. However, with the emulsion phase the Euler
characteristic shows a hysteretic behaviour as known from water wet systems.
The saturation vs. time graph displayed in Figure 4.7 shows that the volume of emulsion and
oil within the field of view is decreasing faster then of the oil alone. This indicates movement
of the emulsion phase. The Euler characteristic, one of the four Minkowski functionals, is a
measure for connectivity: The more negative the values are, the higher the connectivity which
facilitates pathway flow (see Chapter 2.1.4). In Figure 4.7, the Euler characteristic is plotted
for the oil phase (with and without emulsion). Furthermore, a drainage simulation considering
a water-wet system, computed on the pore structure within the FOV is displayed. From this
figure, the following observations become apparent: Compared to the drainage simulation, the
connectivity at a given saturation during imbibition appears lower taking solely the oil phase
into account and higher including the emulsion.
A hysteresis towards higher connectivity during imbibition is also known for water-wet model
systems (Schlüter et al., 2016). A lower Euler characteristic for a mixed-wet system compared
to a water-wet system, however, is unreasonable. Oil layers would rather prevent further
breakdown of the clusters than decrease the overall connectivity. If the emulsion phase would
behave like the brine phase this would result in a higher connectivity than that observed in
this work. The lower Euler-characteristic for the oil phase including the emulsion supports the
assumption that the emulsion rather behaves like the oil phase than the water phase.
4.2.3 Comparison of core-scale relative permeability with pore-scale
data
Fluid distributions obtained at the pore-scale can be used to simulate connected pathway flow
through the respective oil- and water-phase. The fluid distributions of each phase obtained at
different time steps during experiment 1A are used as input to single-phase flow simulations
(Section 3.2.2).
The relative permeability for a specific saturation (time step) can then be obtained by dividing
the permeability through a single phase by the permeability of the total rock (kr oil/brine =
koil/brine / krock ). In Figure 4.8, the result of such a single-phase flow simulation through the
space occupied by the water phase and the oil and emulsion phase is compared to the relative
permeability obtained by the SCAL experiment S1 (see Chapter 4.3.1). A simulation through
the oil-phase only would result is a relative permeability of kroil pSw q = 0 as the oil phase was
disconnected throughout the experiment (Section 4.2.2).
Figure 4.8: The relative permeability curves of µCT-experiment 1A were obtained by single-
phase flow simulation through the space occupied by the water phase and the oil and emulsion
phase, respectively. The graphs show the original data obtained from this simulation with the
absolute (a,b) and mobile saturation (c,d) in comparison with results of SCAL experiments.
For this comparison, the heterogeneity of the rock and the resulting structural differences
between the samples need to be considered. Furthermore, as µCT images are limited by
resolution, micro-porosity is not captured. Micro-pores in Ketton contribute only little to
the overall flow; however, they may contain a large amount of water and hence impact the
initial water saturation. The simulated relative permeability shown in Figure 4.9a,b has been
corrected accordingly by adding the volume-proportion of 40% of micro-pores in Ketton to the
water saturation.
Also, the difference in sample size may affect the comparison. Berg et al. (2016) proposed to
evaluate the mobile saturation only by normalization of the endpoints in the following way:
S 1 SwS Sw,cS , (4.2)

w,c o,r
where Sw,c and So,r are the irreducible wetting and non-wetting phase saturations. The re-
sults displayed in Figure 4.8c, d show the relative permeability as a function of mobile phase
saturation.
In addition, the emulsion phase may impact the resulting permeability. Therefore, saturation
and flux obtained from the simulation need to be corrected.
The correct saturation can be estimated through the water content in the emulsion. Based on
Eq. 3.2 the water content can be obtained from the average emulsion grey value in combination
with the grey value of the water and the oil phase. The average water content of the emulsion
phase was 37.5% .
To correct the flux, the computed permeability of the oil and emulsion phase was weighted by
the proportion of water in the emulsion, which was then used to calculate the permeability of
the total oil phase koil p
oil emulsion q and water phase kwater p
water emulsion q:
q koil Vwater in emulsion

koil p
oil emulsion emulsion
V
koil emulsion (4.3)
oil emulsion
q kwater
Vwater in emulsion
kwater p
water emulsion
Voil
koil emulsion (4.4)
emulsion
with the computed permeability of the water phase kwater and the oil and emulsion phase
koil emulsion , the volume of water in the emulsion Vwater in emulsion and the volume of emulsion
and oil Voil emulsion . This calculation follows the assumption that the water of the water-in-oil
emulsion is produced with the same permeability as the oil.
In Figure 4.9, the relative permeability curves after emulsion correction are displayed.
Figure 4.9: The relative permeability curves of µCT-experiment 1A after emulsion correction
shown with the absolute (a,b) and mobile saturation (c,d) in comparison with results of SCAL
experiments.
Despite the corrections, the water-phase permeability and the oil-phase permeability deviate
significantly (before and after the correction) from the SCAL experiment (S1). This indicates
that treating the emulsion phase as oil - even with a correction of permeability and saturation
due to the proportion of water in emulsion - is not sufficient for computing permeabilities based
on pore-scale fluid distributions. In the discussed example, the permeability of the emulsion
4.3. Impact of sample initialization on wettability 95
phase is systematically overestimated. A reason may be a higher viscosity of the emulsion

compared to the oil phase (Rezaei & Firoozabadi, 2014). However, this aspect cannot be
simulated with the solver used.
This finding has implications on common SCAL experiments. Due to the large amount of
fluid necessary experiments are often conducted with model-oils. Potential in situ formation of
emulsion is disregarded. The results of this work confirm that emulsion can form in the confined
space of the rock, in particular in the presence of crudes with high wettability potential and high
amount of surface active components. The emulsion phase which occupied 10% of the pore-space
was found to have large impact on the connectivity of the oil phase and hence on the relative-
permeability/saturation functions. The fluid distribution and the associated connectivity of
the oil phase indicates that both pathway flow and ganglion dynamics occur. However, it was
not possible to determine which flow regime (pathway flow or ganglion dynamics), dominates
flux.
4.3 Impact of sample initialization on wettability
Macroscopic observation of the samples revealed one surprising aspect: Though all chosen
crudes have a high wettability alteration potential, only the Amott tests 5B and 5C showed
oil-wet behaviour. In contrast, the core-scale flooding experiment (S1) appeared mixed-wet
leaning towards the water-wet side. Different crudes may lead to different wettability. Yet,
when visually comparing the Amott-experiments performed on the SCAL-plug (5B and 5C)
with Amott tests performed on a mini-plug (6A and 6B) with the same COBR system, a
different wetting behaviour was observed, too. Figure 4.10 shows the oil-phase drops on the
rock surface for experiments 5B and 6B. The oil-drops in experiment 5B spread over the sample
surface, while the oil in experiment 6B forms water-wet drops.
The only difference between the samples appearing more oil-wet (5B and 5C) and the samples
appearing more water-wet (6A and 6B) is the sample initialization. While the samples 5B and
5C were initialized with crude oil via centrifugation, sample 6A and 6B as well as the core-
flooding experiment (supporting experiment for 1A, Section 3.5.1) were initialized by flooding.
Figure 4.10: Droplet formation during Amott test. During experiment 5B the oil drop spreads
over the rock surface indicating an oil-wet behaviour (a). In experiment 6B the droplets emerg-
ing during spontaneous imbibition appeared water-wet (b). In both experiments the same
COBR system was used. The difference between these two experiments is (beside size) the
initialization protocol. Experiment 5B was saturated with oil by centrifugation and experiment
6B was initialized by flooding (adapted from Bartels et al., 2017c).
The initialization protocol may lead to a difference in wettability for two reasons: The tem-
perature is higher during centrifugation (40 C for 12 h) and due to the difference in capillary
pressure applied. The effect of temperature was ruled out as the water-wet behaviour did not
change when a flooded mini-plug was exposed to elevated temperature (40 C) for 12 h (exper-
iment 6B). Hence, the difference in capillary pressure must cause this effect. The higher the
capillary pressure, the more and progressively smaller pores are filled with oil (Section 2.1.4).
Figure 4.11 shows the MPI curve of Ketton and respective pore diameter that would be filled
with the different initialization methods. Based on the experimental setting, centrifugation
filled pores down to a pore diameter of 0.3 µm, while flooding, which was estimated based
on the detected saturation filled pores down to a pore diameter of 20 µm. This leads to the
conclusion that the rock structure within the range from 0.3 µm - 20 µm must have a significant
impact on the wettability at the larger scale.
4.4. Summary: Dependency of core-scale parameters on sample handling protocols 97
Figure 4.11: MPI curve of Ketton rock. Based on the oil-saturations obtained from the exper-
iments, initialization by flooding filled pores down to a diameter of 20 µm and centrifugation
down to 0.3 µm (adapted from Bartels et al., 2017c).
4.4 Summary: Dependency of core-scale parameters on
sample handling protocols
From the above discussions, it becomes apparent that wettability measures at different length
scales not always reflect the wettability of a system correctly. An example is the cumulative
oil production over time curves, a key measure obtained in Amott spontaneous imbibition test.
For the water-wet sample (5A) the core-scale observation obtained did reflect the wettability
of the studied system correctly, though, the macroscopically obtained production curve did not
match the observed pore scale displacement. The reason is heterogeneities of the rock, which
lead to preferential production sites and hence to an asymmetrical imbibition front. Such an
asymmetric imbibition front may lead to a large REV, which cannot be monitored continuously
with µCT. However, the mismatch for the mixed-wet to oil-wet experiment (5B) could not be
explained with this asymmetric process, or gravitational effects as µCT images across the entire
core showed only minor brine intrusion. The reason for the macroscopic production obtained
was an oil-layer at the outside of the sample, which may have been caused by roughness during
the standard procedure of rolling the sample over a wetted tissue. This standard procedure
leads to inaccuracy in the determined wettability state especially for oil-wet and mixed-wet
cases at the core scale.
Furthermore, the steady-state experiments at the pore- and core-scale revealed a mismatch of
relative permeability. Fluid distribution and the associated connectivity obtained at the pore-
scale show that the oil phase moves through pathway flow and ganglion dynamics. However,
emulsion also was found to impact the flow behaviour. The formation of emulsion is disregarded
in SCAL experiments, which are often conducted with model-oils.
It was found that the wettability of a system is sensitive to sample initialization protocols.
While a sample, in which crude oil was injected by flooding appeared water-wet, a sample
drained with crude oil by centrifugation appeared oil-wet. The difference between those two
methods is the capillary pressure applied. With centrifugation, smaller cavities get invaded than
by flooding. To assess the impact of the rock structure on these length scales smaller pore-scale
investigations are required. The results for this length scale are presented in Chapter 5.
Chapter 5
Impact of wettability on pore-scale

flow regimes
At the pore scale wettability is commonly assessed either by contact angle measurements or
indirectly by studying the fluid distribution within different pore sizes. These can be obtained
directly from µCT images of sufficient resolution. The ‘apparent’ contact angle measured at
this scale is distinguished from the intrinsic contact angle as it is also affected by sub pore-scale
chemical heterogeneity and surface roughness. Such sub pore-scale properties are often con-
sidered irrelevant at the pore- and pore-network scale, as the main objective is not wettability
characterization, but the prediction of fluid movement through the pore space (Blunt, 2017).
The apparent contact angle determined from µCT images for rough/heterogeneous surfaces
results in the same capillary curvature as that for the same contact angle of a smooth surface.
Based on the Young-Laplace equation (Eq. 2.10) the movement of menisci, controlled by the
pressure gradient imposed due to curvature, would be equivalent for both cases (Section 2.1.3).
This hypothesis, however, has not yet been experimentally validated, and is the subject of the
following sections.
In Section 1, the characterization of pore-scale wettability is presented following the frame

described above. The effect of wettability on the fluid propagation will be discussed in Section
5.2.
99
100 Chapter 5. Impact of wettability on pore-scale flow regimes
5.1 Wettability classification
The core scale data obtained in this work showed two types of mixed wettability. Depending
on the initialisation procedure the result appeared mixed-wet leaning towards water-wet or
mixed-wet leaning towards oil-wet (Chapter 4.4). In this section, both types will be discussed
in more detail by analysing the corresponding µCT images obtained from flooding experiments
1A (initialized by flooding) and 4B (initialized by centrifugation).
The experiments were conducted with the same crude oil (crude A) and rock (Ketton). Based
on the DLVO theory, the difference in brine composition (experiment 1A: HS with 17 wt - %
KI, experiment 4B: DB with 9 wt - % KI) has minor impact on the wettability state of the
studied systems (Chapter 2.2.1). However, the different ageing times (experiment 1A: 1 week,
experiment 4B: 4 weeks) need to be considered when comparing the different initialization
methods. The wettability of the two systems was evaluated by visual inspection of the images,
contact angle determination, and assessment of fluid distribution.
Visual inspection of µCT images
Figure 5.1 shows representative images of the pore space after waterflooding for experiments
1A and 4B. Contrary to expectations for water-wet systems (Chapter 2.1.3), in both cases some
oil was left on the surface of the grains or within small pores and pore throats. Such oil-patches
were also observed in 2D-models (Jung et al., 2016) and indicate oil-wet surfaces.
Contact angle distributions
Though in both experiments (1A and 4B) the same COBR system was used, different contact
angle distributions were measured.
The contact angles of experiment 4B range between 44 and 130 with a median at 114
indicating a mixed-wet system leaning towards oil-wet. However, one needs to consider that in
total only 21 contact angles could be obtained (Section 3.2.2). The contact angles of experiment
1A, show a distribution between 20 and 90 (Figure 5.1e). Compared to water-wet systems the
5.1. Wettability classification 101
sample 1A peaks at a higher contact angle indicating a more oil-wet state, but still a water-wet
system.
The contact-angle distribution relating to experiment 1A is inconsistent with the oil-patches

observed (Figure 48). A sub pore-scale wettability pattern would explain these observations.
Oil can advance more easily over oil-wet surfaces than water-wet surfaces, which have a larger
contact angle to oil. Therefore, oil clusters, that move freely over oil-wet parts, would come
to rest at points where oil-wet rock changes to water-wet rock, resulting in a domination of
water-wet contact angles (Naidich et al., 1995). In oil-wet pores which got filled with brine
during the waterflood, some remaining oil is likely to form a layer in form of the oil patches
detected at some grain surfaces. The thickness of these oil layers, however, is close to the image
resolution. Therefore, these are not suitable for contact angle measurements.
In other words, the lack of oil-wet contact angles in the mixed-wet system leaning towards
the water-wet side prevails because the oil-wet surfaces are either completely covered by oil
clusters, causing the liquid-liquid interfaces always to form adjacent to water-wet rock or by
thin oil-films below the resolution of the images.
But why does this behaviour not occur in experiment 4B? In a sample initialized by centrifuga-
tion the oil penetrates smaller pores (down to 0.3 µm diameter) than those reached by flooding
(down to 30 µm diameter) (Chapter 4.4). Based on the PSD shown in Figure 4.11, Ketton has
only a small number of pores in the range between 0.3 µm and 30 µm in diameter. Hence the
aged area should be similar for these two samples. Yet, the wettability differs significantly.
Alyafei and Blunt (2016) hypothesized that in some cases surface roughness may prevent oil
coming into contact with the surface. In fact, on a surface with dents ranging between 0.3 µm
and 30 µm in diameter, the oil phase would be expected to enter the dents via centrifugation
(experiment 4B) but not via flooding (experiment 1A), due to the difference in capillary pressure
applied (Section 4.1.4). Based on this one would assume that when a sample is initialized by
centrifugation and a lower water saturation is reached the coverage of rock surface by oil is
higher, with a small number of water-wet patches, and hence the sample becomes more oil-wet.
Correspondingly, the contact angle distribution is leaning to the oil-wet side. By initialization
with flooding the oil would not get to, nor age such narrow areas, leading to sub pore-scale
mixed-wettability patterns as observed in experiment 1A.
Figure 5.1: Visual assessment of wettability for experiments 1A, which was initialized by flood-
ing (a,b) and 4B, which was initialized with centrifugation (c,d). At some locations with
presence of oil in the initial state residual oil films were detected after flooding (b,d). This
observation is consistent with the contact angle distribution for experiment 4B. However, the
contact angle distribution in experiment 1A (100 contact angles) indicates a water-wet state
even though more oil-wet than the water-wet reference obtained from Scanziani et al. (2017).
Due to the low number of contact angles obtained for experiment 4B (21 contact angles) solely
the median value, the maximum and minimum are plotted.
5.1. Wettability classification 103
Fluid distribution
In addition to contact-angle measurements, fluid distributions may be used for wettability

assessment at the pore scale. In a water-wet sample the oil at a given saturation would pre-
dominantly occupy large pores, while smaller pores would be occupied by oil in oil-wet systems
(Chapter 2.1.3).
Figure 5.2 displays the fluid distributions of a water-wet system (S1) and the mixed-wet system
leaning towards the water wet side (experiment 1A) at similar saturation. (As experiment 4B
was contaminated with air, the images were not comparable; Section 3.2.2) In both cases, the
oil phase is mainly found in larger pores.
This may be a result of the fluid distribution within the sample after initialisation. During
primary drainage the oil penetrates a water-wet rock. Therefore, the oil resides predominantly
in large pore bodies. Only mineral-surfaces in direct contact with crude oil experience ageing
(Section 2.1.1). During the ageing procedure only those pores containing crude oil turn more
oil-wet, hence the sample remains water-wet in smaller pores and pore throats without oil
leading to a large (pore) scale wettability pattern, which is reflected in the fluid distribution
after the waterflood.
Figure 5.2: Fluid distribution for water-wet experiment S2 (a) and mixed-wet experiment 1A
(b) sample at around 40% oil-saturation.
It has been shown that ageing leads to wettability alteration. However, surface roughness
and sample initialisation may impact the pore-scale wettability, even with the same COBR
system. These findings at the pore scale are in line with core-scale observations, in which
initialisation by centrifugation results in a mixed-wet system leaning towards the oil-wet side,
while initialisation by flooding results in a mixed-wet system leaning towards the water-wet
side (Chapter 4.4). It is crucial to not only evaluate contact angle distribution, but also to
assess fluid distributions and potential fluid films. For the sample initialised by flooding, in
addition to a pore- scale wettability pattern also a small sub-pore-scale wettability pattern may
occur. How the wettability types observed impact the flow behaviour will be discussed in the
following section.
5.2 Pore scale processes during waterflooding
Most studies of flow in porous media are based on water-wet systems (Section 2.1.3). Yet, most
natural COBR systems are considered mixed-wet. In the following two sections the behaviour
of oil in a water-wet (Supporting data S2; Section 3.4.2; Singh et al., 2017a) and a mixed-wet
system (experiment 1A) will be presented.
5.2.1 Physics of oil filling events in water-wet systems
The different pore-scale flow regimes during two-phase flow in porous media were subject of
various studies (Section 2.1.3). Most focus on water-wet systems. For steady-state experiments
it was shown, that the flow regime may switch from pathway flow to ganglion dynamics, de-
pending on the water-fraction and injection rate (Section 2.1.3). Different flow regimes were
also found in unsteady state waterflood experiment in Gildehauser sandstone. While pathway
flow dominated the flux of the oil-phase, coalescence events, commonly associated with gan-
glion dynamics, were also observed (Rücker et al., 2015a). These coalescence events have a
large impact on the relative permeability, as they cause an increase in connectivity of the oil
phase (Berg et al., 2016).
In this work, the main objective is to determine how the flow behaviour in mixed-wet systems
5.2. Pore scale processes during waterflooding 105
differs from the behaviour observed in water-wet systems. For better comparison with the re-
sults obtained from the waterflood in the mixed-wet Ketton sample (experiment 1A), additional
analysis was conducted on data provided by Singh et al. (2017a)(Supporting data S2; Section
3.4.2). In this experiment, unsteady-state waterflood was conducted on a water-wet Ketton
sample with an injection rate of 44.75 nL/min (0.002 PV/min and Nc = 1.08 x 109 ). The
results obtained from this experiment are shown in Figure 5.3a, where the number of oil- and
water-filling events within the field of view (3.8 mm x 3.8 mm x 3.3 mm) is plotted vs. time.
Figure 5.3: (a) Number of oil- and water filling events vs. time detected in the unsteady state
waterflood experiment in a water-wet Ketton sample (supporting data S2; Singh et al., 2017a).
The event-size distribution in comparison with the pore-size distribution of the rock reveals
that water-filling events cover up to multiple pore-bodies, while no oil-filling event larger than
106 µm3 was detected (b). (c-e) show few examples of the oil-filling events (blue) next to the
oil-phase not affected by changes (red). Oil-filling events may cause reconnection of previously
disconnected clusters (f; adapted from Rücker et al., 2015a).
A filling event represents a change in occupancy of the pore-space. In a water-filling event the
oleic phase is replaced by the aqueous phase and in an oil-filling event, the aqueous phase is
replaced by the oleic phase (Section 2.1.3; Section 3.2.2). During a waterflood the number of
water-filling events is significantly larger than the number of oil-filling events. In total only 43
oil-filling events were detected. The size of oil-filling events did not exceed 106 µm3 , while the
water-filling events often filled multiple pores (Figure 5.3b). Figure 5.3c-e shows examples of
the detected oil-filling events. Unlike the observations in Rücker et al. (2015a), none of the
oil-filling events caused reconnection of disconnected oil clusters. A reason for this may be the
different rock structure. Rücker et al. (2015a) hypothesised that the reconnection events are a
result of meniscus fluctuation. These fluctuations may fill small pore throats as illustrated in
Figure 5.3f. However, if these pore throats are too long these fluctuations may not reach the
neighbouring clusters and hence not coalesce.
Even though no reconnection events were observed, these findings do confirm the existence of oil-
filling events in water-wet systems. These are low in frequency and small in volume. In unsteady
state waterflood in water-wet samples water filling events dominate the fluid displacement.
5.2.2 Physics of oil filling events in mixed-wet systems
The images obtained with fast µCT during the waterflood experiment 1A show pore-scale
processes and help to determine and describe event types occurring in a mixed-wet system.
This unsteady state waterflood experiment was performed with an aged Ketton sample. As
discussed in Section 4.4 and 5.2, this particular system appears mixed-wet leaning towards the
water-wet side at the core and pore scales with a combination of a pore-scale and sub pore-
scale wettability pattern. In addition to the oil- and brine phases also an emulsion phase was
observed (Section 4.3.2).
Within the investigated system all three phases, oil, emulsion and water have been observed to
displace one another (Appendix D). The behaviour of the oil phase is the most important as it
determines the oil recovery.
Compared to water-wet systems (S1) the oil-filling events observed in the mixed-wet case (1A)
5.2. Pore scale processes during waterflooding 107
appear to be different. Figure 5.4 a and b shows such events. An oil cluster connected to the
inlet starts to fill a pore body. With time (13-min /15 min) the oil cluster grows until it comes
into contact with another oil cluster. Thereby, the oil-phase displaces brine within the pore
(Appendix D). To highlight the difference to the water-wet case (Chapter 5.3.1), the observed
behaviour is displayed in a schematic drawing in Figure 5.4c. In the mixed-wet case full pore-
bodies are involved in the displacement of water by oil (Figure 5.4), while in the water-wet case
the reconnection of disconnected ganglia appears only in pore throats (Section 5.2.1).
Figure 5.4: The oil-filling events observed in this work can cover full pore bodies (a) The
oil cluster (red) moves with time more and more into the pore body until it connects with
the surrounding oil phase. These oil-filling events fill a pore previously filled with brine (b).
The displacement mechanisms taking place within the mixed-wet system of experiment 1A is
illustrated schematically in (c).
Not only the event type differs in mixed-wet systems, but the frequency and volume of oil-filling
events are also significantly higher. Figure 5.5a shows the number of water-, oil- and emulsion-
filling events counted for each time step. In addition, Figure 5.5b shows the corresponding
event-size distribution of the oil phase on aged (initially oil- and emulsion-covered) and non-
aged (initially water-covered) surfaces. The largest events observed have the size of the largest
pores.
Figure 5.5: The frequency of the three event types, water-filling events, oil-filling and emulsion-
filling events for each time step (a) shows that the oil-and emulsion redistribution is of sig-
nificance in mixed-wet cases. The pore-size distribution and the initial fluid distribution (b)
illustrate the initial situation before events occur and indicate the impact of the wetting state
of the surface on the flow behaviour. Much more oil-filling events are detected on surfaces
initially covered with oil compared to surfaces covered with emulsion or water (c).
As observed in the fluid distribution after ageing (Figure 5.5c) oil fills most of the larger pores,
while the water-phase and the emulsion occupy predominantly the smaller pores. Nevertheless,
the volume of oil involved in larger sized oil-filling events is higher for pores initially occupied
with emulsion than with water, which hints to a dependence of oil-filling events on the ageing
state of the rock surface.
These findings show the relationship between the event type and the wetting state of the rock
surface. The assumption that apparent contact angle would sufficiently indicate the wettability
as input for the prediction of two phase flow, could not be verified in this case. The sub-pore
scale mixed-wet pattern, which was not reflected in the contact angle distribution obtained for
5.3. Summary: Enhanced ganglion dynamics in mixed-wet systems 109
this experiment (Figure 5.1) lead to a different behaviour than known for water-wet systems.
These observations confirm the hypothesis that mixed wettability eases the movement of oil-
clusters leading to larger oil-filling events covering up to full pore-bodies with a higher frequency
than known from water-wet systems.
The high frequency of oil-filling events and the large volumes involved hint to a high impact
of ganglion dynamics on the overall flow. Even if oil-filling events not necessarily contribute
directly to the flux, the impact on relative permeability may be significant due to their ability
to change the connectivity (Section 2.1.3; Section 2.1.4). Based on the discussed results, this
effect is likely to be more pronounced in mixed-wet systems.
5.3 Summary: Enhanced ganglion dynamics in mixed-
wet systems
The pore-scale analysis confirmed the observations at the core scale (Chapter 4) indicating
that sample initialisation impacts the observed wettability, even with the same COBR system.
Thereby both, a large wettability pattern, in which pores filled with brine get altered while
others filled with brine do not, and a small wettability pattern along the pore surface below
µCT resolution, which leads to a water-wet contact angle distribution though oil-layers were
present, were identified. The origin of the small-scale wettability pattern will be discussed
further in Chapter 7.
The flow regime in a mixed-wet rock-crude oil-brine system showed significantly more ganglion
dynamics than in a water-wet system. Oil-wet surfaces reduce the energy required for an oil-
cluster to move and cause more frequent and larger oil-filling events covering full pore-bodies.
These differ significantly from oil-filling events in a water-wet system, in which oil-filling events
are rare and do not exceed the size of pore throats. As such oil-filling events are known to affect
the oil-configuration in the rock and the connectivity of the oil phase, this behaviour is likely to
impact the overall relative permeability and must be considered in pore-scale flow simulations.
A simple extrapolation of the flow regime from water-wet conditions based on contact angle
adjustment appears insufficient to capture the dynamics and connectivity of the oil phase.
In the following Chapter 6, the impact of enhanced oil-recovery on the flow behaviour will be
discussed.
Chapter 6
Enhanced oil recovery at the pore scale
As wettability is known to impact flow in porous media, wettability alteration is often a target
for enhanced oil-recovery methods. One of those is low salinity flooding, in which an alteration
of the wettability from oil-wet towards more water-wet is induced by reducing the salinity of
the injected brine (Section 2.2.1). While various studies investigated the low salinity effect at
the nano and core scale, only little is known about the impact of LS-injection on pore-scale
processes (Section 2.2.1). In this work, this aspect was investigated by fast µCT. The impact
of wettability alteration occurring during low-salinity flooding on pore-scale processes will be
discussed in Section 6.1.
Another enhanced oil-recovery technique is surfactant flooding (Section 2.2.2). Compared to

low-salinity flooding the effect is not based on a change in wettability, but predominantly in
interfacial tension. Therefore, it is commonly assessed by static experiments involving only the
fluid phases. However, recent studies on micromodels have shown that confined pore-space may
also impact this technique. In Section 6.2, pore-scale observation during surfactant flooding is
discussed.
111
112 Chapter 6. Enhanced oil recovery at the pore scale
6.1 Low-salinity flooding
As wettability is known to affect relative permeability, mechanisms that change wettability

may result in an enhancement of oil-recovery (EOR). One technique targeting this effect is
low salinity flooding. To assess the impact of this EOR technique, a low salinity waterflood
(experiment 2A: Ketton limestone sample; 2E Berea sandstone sample) was performed right
after a conventional waterflood (1A and 1E).
Figure 6.1 shows the last image obtained during the waterflood of experiment 1A (Figure 6.1a)
and the first (Figure 6.1b) and last image (Figure 6.1c) of the LS-flood of experiment 2A.
During the experiment the change in salinity could be tracked as the concentration of the
contrast agent decreased (Appendix E). The decrease in salinity was not homogenous. Pore-
geometry and oil-configuration seemed to impact the mixing and dilution of high salinity (HS)
brine, leading to a HS-brine content of ¡ 5% after 5 min (2.6 PV) of LS-brine injection.
Figure 6.1: µCT grey value images obtained after a waterflood (a) and at the beginning (b)
and end of LS flood (c). During the LS-flood experiment (2A) the change in salinity within
the pore-space was tracked. Pore-geometry and oil-configuration affect the mixing and dilution
process of the HS-brine. As the LS-brine appeared with the same grey value as the emulsion,
these two phases could not be differentiated.
As the low salinity brine injected (LS) yielded the same grey value as the emulsion phase
(Figure 6.1a and Figure 6.1b), these two phases could not be differentiated after dilution.
However, a change in occupancy by the pure oleic phase could be observed ((Figure 6.1c)).
The redistribution is not necessarily linked to the low-salinity effect, but might also be a
consequence of oil being produced from below the field of view. The release of trapped oil
from dead volumes in the set-up also could not be excluded. Yet, the final distributions of oil
6.1. Low-salinity flooding 113
inside the rock provide insights into the low salinity effect.
The distribution of oil within the pore-space depends on the wettability pattern of the rock
(Section 5.2). In general, the oil is expected to occupy smaller pores in an oil-wet system and
larger pores in a water-wet system. In a mixed-wet system this may deviate corresponding
to the ageing history (Section 5.2). A change in wettability would be expected to cause a
change in fluid distribution within the rock. The low-salinity effect known to cause a change in
wettability towards more water-wet would lead to a change in oil occupancy towards the larger
pores.
Figure 6.2 shows the final distribution of the fluids after the waterflood (Figure 6.2a,b) and
the low-salinity flood (Figure 6.2c,d), as well as a comparison of the distribution of oil (Figure
6.2e,f), in both cases for the low-salinity experiment in Ketton (2A) and Berea (2E).
The Ketton sample (experiment 2A) showed a higher oil saturation at the end of LS than at the
end of HS (41% vs. 34%). Also, no redistribution of the oil phase from smaller to larger pores
was observed. In fact, the amount of oil in smaller pores at the end of the LS-flood seemed
to remain constant compared to end of the HS-flood. Considering the large increase in oil-
saturation at the end of the HS-flood and LS-flood, other factors besides the low-salinity effect,
as the release of trapped oil from dead volumes of the set-up, are likely to have contributed to
the observed response.
For the Berea sandstone the behaviour was different: The oil saturation was higher at the end
of the HS-flood than at the end of LS-flood (8% vs. 5%). In addition, the oil-saturation in
the small pores decreased whereas the oil in the larger pores remained constant. The result of
experiment 2E may be explained by wettability alteration towards less oil-wet and subsequent
mobilisation of oil. However, the potential release of oil from the dead volumes of the set-up,
as observed in experiment 2A needs to be considered in this case, too.
In summary, the two samples with different rock types and crude oils, behave differently even
though the samples have been subjected to the same experimental protocols.
Figure 6.2: Final distribution of the fluids after the waterflood experiment 1A in Ketton (a)
and experiment 1E in Berea (b) and after the associated low salinity flood experiments 2A and
2E (c, d), as well as a comparison of the final distribution of oil (e, f; adapted from Bartels et
al., 2017b).
6.2 Surfactant flooding
Another method to improve oil-recovery is surfactant flooding, in which surfactants, reducing

the interfacial tension and therewith increasing the capillary number (Nc ) are added to the
aqueous phase. Surfactants, which produce an ultra low interfacial tension ( 102 mN m1 )
6.2. Surfactant flooding 115
between the oleic and the brine phase typically also form emulsions.
To assess emulsification, commonly static tube-tests are performed, in which the aqueous and
oleic phases are mixed by rigorous shaking (Section 2.2.2). A surfactant formulation is consid-
ered optimal, if the emulsion contains 50% oil and 50% emulsion as illustrated in Figure 6.3.
In this work, the oil-phase was doped with iodo-decane and imaged by µCT. Following the
equation
Grey valuepemulsionq γo Grey valuepoilq γw Grey pwaterq (6.1)
the grey value of the emulsion was found to reflect the oil- and water- content as expected from
the tube test (experiment S3).
Figure 6.3: Tube tests, conducted to determine the optimum formula for surfactant flooding
revealed, that the grey values of the emulsion phase measured with µCT reflect its oil- and
water- content, when the oleic phase is doped with iodo-decane (adapted from Rücker et al.,
2017).
Studies in micromodels have shown that flow in porous space impacts the formation of mi-
croemulsion (Chapter 2.2.2). In this work, fast µCT (7s/ image) was employed to study the
impact of flow dynamics in rock on the emulsification process (experiment 3A). In experiment
3A, first a waterflood (brine: FW2) was performed on an oil-saturated (doped n-decane) rock
sample (Gildehauser sandstone) consecutively followed by a surfactant flood which was contin-
uously monitored to resolve pore-scale processes during surfactant flooding.
In situ formation of emulsion in rock
Figure 6.4 shows grey-scale images of experiment 3A. As in these images the oleic phase is
doped with contrast agent, the oil appears white and the water phase black.
Figure 6.4: Grey scale image of the rock without any fluid (a), final scan of the waterflood
performed prior to the surfactant flood in experiment 3A (b) as well as first (c) and final(d)
image of the surfactant-flood itself. Subtle changes in the grey level can be observed when
comparing images (c) and (d)(adapted from Rücker et al., 2017).
Even in the first scan of the surfactant injection, obtained 7s after the injection started, no
oil was detected within the field of view. Although, the total injected surfactant solution
volume was less than 1 PV ( 0.8 PV), it is likely that surfactant breakthrough occurred
through preferential pathways and/or quasi-miscible flow and associated fingering took place
(Lenormand et al., 2006).
Figure 6.4d shows the final scan of the experiment which was taken approximately 45 min later.
At this point the pore space was flooded by the surfactant for multiple PVs. Therefore, it was
assumed that any quasi-miscible displacement was completed. Compared to the first scan of
6.2. Surfactant flooding 117
the surfactant flood, subtle differences in grey levels within the bigger pores were directly visible
(compare Figure 6.4c and d).
Based on Eq. 6.1, the average grey value of the pore space indicates the composition of the fluids
inside. This averaged grey value for each image is displayed as a function of time in Figure 6.5.
The averaged grey value correlated with the oil saturation obtained by segmentation from the
waterflood results (Figure 6.5a) and therefore, could be used to estimate the oil content in the
pore space during the surfactant flood. Instantly after surfactant injection, the oil saturation
within the field of view decreased from 40% to 18%. This sharp reduction suggested a quick
transport of surfactant, leading to an immediate emulsification. The process was too quick to
infer the emulsification rate from these images. However, it is likely that the interfacial tension
reduction was sufficient to mobilize a significant amount of oil within seconds. Successively,
the oil saturation gradually decreased a further 2%. The residual oil saturation reached 16%
(Figure 6.5b, at 95 mins).
Figure 6.5: The mean grey value of the fluids within the pore-space correlates linearly with
saturation (a), which allows to estimate the oil content of the whole sample, once emulsification
takes place (b). The oil-content within the sample drops sharply with the start of the surfactant
injection resulting in an additional production of 55% of the remaining oil (adapted from Rücker
et al., 2017).
Pore-scale processes during surfactant flooding
The grey value analysis of the images can also be used for investigation of the pore-scale be-
haviour during the emulsification process and the associated quasi-miscible flow. In Figure 6.6,
images from the water- and surfactant-flood with a colour scheme from blue to red representing
the difference in grey values inside the pore space, are shown. Blue represents 0% oil- and red
100% oil-saturation. Figure 6.6a was taken at the end of the waterflood, where oil and water
were not yet mixed. Figure 6.6b and c show the first scan (7 s) and second scan (14 s) after the
surfactant flood started. The initially oil-filled (red) pores immediately changed composition.
With the first scan the colour transformed from light to dark blue suggesting the presence of an
aqueous phase, but possibly with different water content, i.e. emulsion with oil solubilized. The
fraction of the dark blue within individual pores increased with time, which hints to further
reduction of emulsion within the system (Figure 6.6c).
Figure 6.6: Last scan of the water to surfactant flooding of oil (red) and water (blue) within the
pore space prior to surfactant flooding (a). The first image obtained 7 s after the surfactant-
flood started showed a change in grey values indicating immediate emulsification within pores
initially filled with oil and pores initially filled with water. In some pores the oil was immediately
produced (b). (c) shows further production of oil via solubilisation. The pore-scale processes
in rock confirm observations in micromodels from Unsal et al. (2016) and Broens and Unsal
(2018) indicating a faster emulsification under dynamic conditions than expected from phase-
behaviour studies (adapted from Rücker et al., 2017).
The fact that the surfactant retrieved and mobilized a substantial fraction of the remaining
oil during the first few seconds suggests that the emulsification rate was much faster com-
pared to rates observed at static conditions. Unsal et al. (2016) observed in micromodels
6.3. Summary: Changes of oil configuration during enhanced oil recovery 119
that microemulsion formed quicker under dynamic conditions within confined pore space than
expected from static phase behaviour studies. As illustrated in diffusion controlled processes
appearing e.g. in dead-end pores and turbulent mixture during co-injection of two fluids (Fig-
ure 6.6d) provided the means for stirring and faster emulsification. Their effects may be even
more pronounced in rock as the connectivity in a 3D porous medium is higher than in 2D
micromodels.
6.3 Summary: Changes of oil configuration during en-
hanced oil recovery
The results obtained from the surfactant flooding experiment (3A) suggested that the emulsi-
fication during flow took place over shorter time scales (seconds) than what would be expected
for static conditions. It is likely that mixing and the diffusion of surfactant occurring with
flow in the confined pore space of the rock facilitated faster emulsification rates. Therefore, the
pore-scale rock structure needs to be considered for predictions of surfactant flooding efficiency
(Rücker et al., 2017).
The two samples of the low salinity experiment (2A, 2E), with different rock types and crude
oils, discussed in this work behaved differently, even though the samples have been subjected
to the same experimental protocols. One (2E) showed a decrease, the other (2A) an increase in
saturation. Various parameters and processes may have caused this difference: pore or sub-pore
geometry, the different range of pore sizes within the rocks, chemical composition of the crudes
and mineral composition of the rocks. In addition, there are more generic issues in studies
on low salinity flooding that could influence the response of a particular COBR system. The
salinity levels, the ageing period and protocol or the time scale of the processes studied in
the experiment may also influence the occurrence and/or magnitude of a response. Therefore,
despite interesting observations when brine salinity was changed, definitive conclusions on the
effect of salinity change cannot be made (Bartels et al., 2017b).
Chapter 7
Impact of nano-scale structures on

wettability
Wettability at the molecular scale is defined as the relative attraction at contact of one fluid
with a solid in presence of another immiscible fluid (Section 2.1). However, studies at larger
length-scales have proven that, even with the same COBR system, different wettability states
may be achieved (Section 4.3). This variation is largely controlled by sub-pore scale geometrical
features (Section 5.1).
Surface roughness impacts the pore- and core-scale wettability in two ways. In correspondence
with the Wenzel model (Wenzel, 1936) the apparent contact angle is shifted away from the
intrinsic contact angle, exaggerating the effect of more oil-wet for intrinsic oil-wet and more
water-wet for intrinsic water-wet systems. In addition, sub-pore scale wettability patterns
impact the apparent contact angle observed at the pore scale (Chapter 2.1.2). These mixed-
wettability patterns are considered to form because of a water-layer prevalence, facilitated by
surface roughness, which prevents contact of the oil phase with the rock surface (Section 5.1).
What determines the formation of such water layers remains unclear.
For addressing this question, the impact of surface roughness was studied with atomic force
microscopy. How this technique can be used for detection of water-layers is discussed in Section
7.1. The impact of surface roughness on fluid formation is discussed in Section 7.2.
121
122 Chapter 7. Impact of nano-scale structures on wettability
7.1 Detection of fluid films
Atomic force microscopy provides insights into wettability and wettability alteration, as it can
easily be modified to reflect a 3-phase (fluid/fluid/solid) system. Functionalised tips may be
used to assess the adhesion of specific components against a surface. Alternatively, the adsorp-
tion of components on a surface can be investigated (Section 2.1.1). Recent studies demon-
strated the measurement of contact angle with atomic force microscopy (Deng et al., 2018;
Giro et al., 2017). In this work, this approach is further extended to detect and characterise
water-films on mineral surfaces in the presence of oil (n-decane).
In experiment 10A a calcite sample was cleaved in brine, which was then displaced by n-decane
via flushing and measured at static conditions in n-decane. As in this two-phase system water
is the wetting phase, a water film is expected to form. Reference measurements were obtained
for both respective single-phase systems, brine (9B) and n-decane (9C). Representative force
distance curves of these reference experiments are shown in Figure 7.1. In both cases no, or
little, adhesion was observed along the whole sample surface.
Figure 7.1: Representative force distance curves for measurements in brine (a- experiment 9B)
and oil (b- experiment 9C).
In contrast, experiment 10A revealed an adhesion pattern. Figure 7.2a shows the adhesion map
of the scanned area, which reveals round patches with increasing adhesion towards their centre.
Three force distance curves from different data points on the adhesion map are displayed in
Figure 7.2c-e. One is representative of the area outside and two were taken from within a patch
7.1. Detection of fluid films 123
at different distances from the centre. The retrace curve reflects the observations from the
adhesion map. The adhesion gets stronger towards the centre of a patch.
Figure 7.2: Adhesion map of experiment 10A, in which a calcite surface was first placed in
brine and then submerged in decane (a). The image shows round patches with high adhesion,
which gets larger towards its centre. This is also reflected by the three force distance curves
taken from this image (c – e). In addition, a snap-in can be detected during the approach of
the tip. This snap-in marks the transition of the tip through the fluid-fluid interface and can
be used for a 3D illustration (b) of the wetting film (blue) on top of the sample surface (grey).
A cross-section through the droplet can be used for contact angle measurements (f).
A deflection towards the surface was also detected during the approach of the tip (trace). Such
a snap-in behaviour is known from systems in which the tip is attracted by the surface (Section
3.4.1). But, as illustrated in Figure 7.1, this is neither the case in oil, nor in brine. Furthermore,
this attraction did not lead to an immediate contact of the tip to the surface (h = 0) – the
snap-in height h varies along the patch. In Figure 7.1f the snap-in distance along a cross-section
of an adhesion patch is plotted. The closer to the centre, the larger the snap-in height.
It is therefore concluded, that this snap-in behaviour is caused by the tip crossing a fluid-fluid
(decane-brine) interface whilst approaching the surface. The tip shows a water-wet behaviour.
If the tip would be oil-wet it would act repulsive to the fluid-film or -layer it is measuring. The
cantilever would bend towards the opposite direction and might deform the fluid-fluid interface
when moving through. Such behaviour was observed when measuring crude-oil residues in
brine (Appendix F). For a water-wet tip measuring a water-film, however, the snap-in - surface
distance represents the water film thickness. Through an automated analysis, this snap-in
distance was obtained for all data points of the AFM image and translated into a 3D image.
Figure 7.2b shows the resulting 3D image with the snap-in distance (blue) on top of the calcite
surface (grey). It reveals a drop shape of the water phase. Correspondingly, the adhesion
patches in Figure 7.2a do not represent an inhomogeneous wettability distribution as the chosen
fluid/fluid/rock composition is pure and homogeneous and the components do not alter the
wettability of the solid (Section 3.1). The location of the droplets is likely to be related with
surface defects of the calcite surface leading to contact line pinning (Section 2.1.2).
The cross-section as shown in Figure 7.2f can also be used to obtain the contact angle. The
contact angle detected for this brine-decane-calcite system showed a value around 21 {24
indicating a strongly water-wet behaviour. These values are within the lower range of the
contact angle distribution obtained by µCT from Ketton rock-brine-n-decane system (Scanziani
et al., 2017).
These patches do not represent an inhomogeneous wettability distribution as the chosen fluid/fluid/rock
composition is pure and homogeneous (Section 3.1), but are more likely to be caused by surface
defects of the calcite and contact line pinning (Section 2.1.2).
7.2 Impact of surface roughness on wettability
The approach of measuring water-film thickness in presence of oil can also be used within a rock.
However, surface roughness introduces additional image artefacts, which need to be considered
(Appendix C).
7.2. Impact of surface roughness on wettability 125
Figure 7.3 shows the surface topography of a Ketton surface (experiment 8A) and an Estaillades
surface (experiment 8B). During the measurement special attention was given to the selection of
the scanning area, so it would represent the surface of a pore, exposed to ambient fluids. With
the help of an optical image of the rock sample, the AFM cantilever was moved towards a pore
close to the sample border (Figure 7.3a and e). In rocks, in which pores cannot be identified,
the tip may land on broken or cut areas, which would not represent the surface roughness of
the rock correctly.
The surface topographies of the Ketton and Estaillades rocks are shown in Figure 7.3a and e.
The surface structure of Estaillades was dominated by crystals, while the nano-scale features
of Ketton rock appeared round.
Figure 7.3: Topographical images of Ketton (a,b,c) and Estaillades (d,e,f) rock taken from
within pores. The optical images (a,d) show the scanning location for each rock respectively,
(b), (e) the corresponding topographical images, and (c), (f) the associated roughness.
The Ketton sample showed a Gaussian distribution of roughness with average roughness Ra =
375 nm and root mean square roughness Rq = 459.6 nm, given by the deviation from a flat
surface and the average peak-to-valley roughness given by the distance between local minima
and local maxima Rt = 2.457 µm (Figure 7.3c). The Estaillades rock sample had an asymmetric
roughness distribution with an average roughness Ra = 494 nm, a root mean square roughness
Rq = 636.8 nm and the average peak-to-valley roughness Rt = 4.134 µm (Figure 7.3f).
The surface roughness obtained with AFM was used to assess the relationship of capillary
pressure and the observed wetting films. A sphere-fitting drainage simulation, was applied,
which based on Eq. 2.5 represents different capillary pressures (Section 3.3.3). The larger the
sphere, the lower the corresponding capillary pressure. Figure 7.4a-c shows the results of this
simulation. The higher the diameter – and the lower the corresponding capillary pressure – the
larger the surface area covered with brine (Figure 7.4d).
Figure 7.4: Water films (blue) on top of a rock surface (grey) obtained from a drainage simula-
tion based on sphere fitting (a: sphere diameter 1.6 µm, b: sphere diameter 5.6 µm, c: sphere
diameter 6.4 µm). The rock surface is in direct contact with oil where no water is visible. d
shows the percentage of surface area covered with brine as a function of sphere diameter.
50% of the surface was covered with water for a sphere diameter of 7 µm. For sphere diameters
below 1.5 µm only 1.5% of the surface remained in contact with the aqueous phase.
These findings give further insight into the relationship between roughness and wettability
alteration. Assuming that ageing occurs predominantly when the oil is in direct contact with
the surface, the surface coverage by a water film would reflect the sub-pore scale wettability
pattern. However, Lowe et al. (1973) found that wettability alteration through diffusion of
certain crude-oil components may also be possible. This behaviour would be controlled by the
nature of the crude oil components and the film thickness of the wetting layer. As shown in
Section 7.2, the latter can be determined by AFM. Figure 7.5a shows the water film on a rock
7.3. Summary: Relationship between surface roughness and wettability 127
surface measured after primary drainage of oil (n-decane) into a water saturated Ketton sample
with flooding (experiment 11A). In this experiment 60% of the surface was covered with brine.
Only 40% of the surface was directly in contact with n-decane. The 3D measurement of the
wetting film also allows to quantify fluid film thickness. Figure 7.5b shows, that the water-film
thickness ranges between 0 nm and 200 nm with a peak film thickness of 125 nm.
Figure 7.5: (a) shows the water film (blue) on top of a rock surface (grey) measured in n-decane
(experiment 11A). The water layer shows a film thickness ranging between 0 nm and 200 nm
with a peak film thickness of 125 nm (b).
7.3 Summary: Relationship between surface roughness
and wettability
In this work wetting films were measured on the nm-scale and complemented by simulation
results on same geometry, corresponding to a range of capillary pressures. The results show,
that surface roughness of rock can facilitate the formation of water-films. Experiment 11A, in
which a rock sample was drained with n-decane by flooding showed a brine coverage of 60% with
a peak film thickness of 125 nm. Drainage simulation showed that the coverage is depending
on the capillary pressures applied. Considering that wettability alteration is depending on the
oil surface contact, the difference in coverage results in a sub-pore-scale wettability pattern,
which could impact larger scale flow behaviour.
Chapter 8
Conclusions
Wettability is a multi-scale property. At the molecular scale, it is defined by interactions be-

tween molecules of a fluid/fluid/solid system. At the sub-pore scale, contact angles are measured
along the interfaces of the three phases, and surface roughness and small-scale chemical het-
erogeneities are added as wettability defining measures. At the pore-network scale, mineralogy
and fluid distributions need to be considered.
The key question addressed in this work is how these aspects of wettability obtained at differ-
ent length scales impact core-scale flow properties such as relative-permeability and capillary-
pressure/saturation functions. For this purpose, various experiments for wettability determi-
nation were conducted. While the length-scales varied, the same COBR systems were used. A
comparison with core-scale relative permeability data showed that the difference in viscosity
between oil and emulsion also needs to be considered.
For sub-pore-scale investigations, an approach for in situ and 3D measurement of connate

water films was developed allowing the quantification of fluid film distribution, film thickness
and contact angle measurements with AFM. In the studied n-decane/brine/rock system a
connate water film thickness of 120 nm was determined. This film covered around 60% of the
surface. Drainage simulations on the sub-pore-scale surface showed that the surface coverage
is depending on nano- to micro -scale surface features and the capillary pressure applied. The
higher the capillary pressure, the more oil contacts the surface.
129
130 Chapter 8. Conclusions
This finding is consistent with the pore-scale observations obtained in this work. Two samples
initialised with two different preparation methods associated with variance in capillary pressure
showed clear differences in contact angle and fluid distributions. One sample, initialised with
lower capillary pressure, showed a mixed wettability pattern with a contact angle distribution
leaning towards the water-wet side. The other sample, initialised with higher capillary pressure,
showed a contact angle distribution leaning towards the oil-wet side. In unsteady-state flow
experiments at the pore-scale it was shown that the mixed-wet system leaning towards the
water-wet side had larger and more frequent oil-filling events compared to a water-wet scenario.
Correspondingly, for pore-scale models, it is not sufficient to account for the mixed-wet state
by adjusting the contact angles. The difference in flow regime, which shows more ganglion
dynamics behaviour in mixed-wet systems, must also be considered.
The experiments conducted with realistic COBR systems appeared more complex than model
systems. Besides brine and oil, an emulsion phase was found in the void space of the rock. This
brine-in-oil emulsion impacted the flow behaviour by increasing the connectivity of the oil.
Furthermore, Amott tests were conducted. The water-wet sample showed a high oil production
and an irregular imbibition front. In contrast, a low number of small water-filling events
resulting in a production of only 1% were observed in the mixed-wet sample initialised with
high capillary pressure. This observation indicates a mixed-wet system leaning towards the oil-
wet side. Smaller samples, which were prepared with a lower capillary pressure, again, showed
a mixed-wet leaning towards water-wet behaviour. This illustrates that the impact of capillary
pressure at the sub-pore scale wettability gets reflected at larger scales, too.
In summary, this work shows a consistency of the observed wettability across the length scales.
It was shown, that sub-pore scale roughness has a strong impact on the wettability alteration
process which impacts larger scale flow behaviour and wettability measures. To predict relative
permeability curves in simulations this aspect needs to be considered.
Consistent studies across length scales are important to upscale and predict the impact of small-
scale phenomena and therefore should also be considered in enhanced oil-recovery techniques,
which are often solely investigated at the sub-pore scale and the core scale. In this work, a
131
method was introduced to track salinity during low salinity flooding or to assess emulsification
during surfactant flooding through estimations based on the partial volume effect.
During a low-salinity flood in sandstone additional oil production was observed. However, an
underlying mechanism could not be confirmed. For surfactant flooding, faster emulsification
rates associated with flow in confined pore space could also be ratified for rock. This leads to
the question of whether common static tube tests are sufficient for the determination of the
optimal surfactant solution.
132 Chapter 8. Conclusions
Chapter 9
Outlook
While the importance of wettability parameters for each length scale and their impact on the
macroscopic outcome was proven, a remaining question is how the different parameters could
be used to predict the core-scale relative permeability-saturation function. In Figure 9.1 an
adapted workflow on upscaling wettability is presented.
Pore network models could be used to bridge the pore scale to the core scale. However, at the
current stage, these models are not capable of capturing flow dynamics sufficiently. One issue
is the assumption that the contact angle distribution would be an adequate representation of
wettability. The contact angles observed at the pore scale may appear water-wet, though the
system is mixed-wet at the smaller scale. Nevertheless, a sophisticated analysis at the pore
scale would be able to identify those cases e.g., through the presence of oil layers along the
rock surface. More studies on linking different flow regimes with wettability patterns would be
necessary to explore how pore-network models need to be adapted to fit the purpose.
These flow regimes could be studied experimentally through µCT flooding experiments while
the pressure drop along the core is monitored to directly link core-scale relative permeability to
pore-scale processes as done by Gao et al. (2017) for water-wet systems. Furthermore, direct
simulations could be used. Direct simulation can predict flow dynamics in confined rock, but,
as they are computationally expensive, they can only be used for small volumes. Nevertheless,
these simulations could be used to assess the impact of different sub-pore scale wettability
133
134 Chapter 9. Outlook
Figure 9.1: Workflow for upscaling wettability from the molecular scale to the core scale.
patterns and surface structures on the propagation of a fluid, similar to the sensitivity studies
Bultreys et al. (2018) performed with pore-network models with respect to contact angle
distributions. With surface structures and wettability patterns as input and fluid distributions
as output, these simulations could also be used to upscale sub-pore scale to the pore scale.
It has been shown that wettability patterns depend on surface structures on the sub-pore scale.
In this work solely Ketton was investigated in detail. However, the special origin of this rock
leads to the question of whether similar artefacts could be observed in other rock types, too.
Furthermore, an image size of 10µm x 10µm is rather small. Especially for less homogenous
rock a workflow for the determination of a representative elementary volume for roughness
needs to be developed. A combination of roughness analysis based on µCT and AFM analysis
could be a solution. Also, the relationship between nano-scale roughness and mineralogy would
be of interest for this purpose.
Simulation on the sub-pore scale indicated that the capillary pressure controls the formation
of fluid films within the surface structure. However, the impact of the ageing processes and the
intrinsic contact angle were neglected in this work. Additional experiments on water-films at
connate water saturation could also be conducted on aged samples to elucidate this question.
Furthermore, AFM could be used to assess oil-films at residual oil saturation.
In situ AFM measurements on rock surfaces could be further extended to EOR application,
135
too.
Moreover, intrinsic contact angles obtained from AFM results could be compared with molecular
modelling approaches based on the molecular forces of the different components to bridge
molecular and sub-pore scale. While adhesion forces of different solid/fluid/fluid components
were already investigated in detail the impact of temperature and pressure is still a pending
question.
In summary, all parameters necessary to determine the impact of wettability on the core scale
have been identified. To predict the impact of wettability on relative permeability and capillary
pressure curve, more detailed knowledge on the exact relationship between those parameters
and macroscopic wettability is necessary. Sensitivities need to be assessed to create simplified
but sufficient models.
136 Chapter 9. Outlook
Literature
”A practical guide to AFM force spectroscopy and data analysis”. (2018, 27.06.2018). JPK in-
struments Technical Note. Retrieved from http://www.jpk.com/index.230.en.html?src=ws1115
”QITM mode - Quantitative imaging with the NanoWizard 3 AFM”. (2018, 11.04.2018). JPK
instruments Technical Note. Retrieved from https://www.jpk.com/app-technotes-img/AFM/
pdf/jpk-tech-quantitative-imaging.14-1.pdf
”SLS Beamlines”. (2018, 04.03.2018). Retrieved from https://www.psi.ch/sls/beamlines-at-sls
Abdallah, W., Buckley, J. S., Carnegie, A., Edwards, J., Herold, B., Fordham, E., Graue, A.,
Habashy, T., Seleznev, N., & Signer, C. (2007). Fundamentals of wettability. Schlumberger
Oilfield Review, 19/2, 44-61.
Akhlaq, M. S., Kessel, D., & Dornow, W. (1996). Separation and chemical characterization
of wetting crude oil compounds. Journal of Colloid and Interface Science, 180 (2), 309-314.
doi:10.1006/jcis.1996.0308
Al-Raoush, R. I. (2009). Impact of wettability on pore-scale characteristics of residual nonaque-

ous phase liquids. Environmental Science & Technology, 43 (13), 4796-4801.
doi:10.1021/es802566s
Alhammadi, A. M., AlRatrout, A., Singh, K., Bijeljic, B., & Blunt, M. J. (2017). In situ
characterization of mixed-wettability in a reservoir rock at subsurface conditions. Scientific
Reports, 7 (1), 10753. doi:10.1038/s41598-017-10992-w
AlRatrout, A., Raeini, A. Q., Bijeljic, B., & Blunt, M. J. (2017). Automatic measurement of
137
contact angle in pore-space images. Advances in Water Resources, 109, 158-169.
doi:10.1016/j.advwatres.2017.07.018
Alyafei, N., & Blunt, M. J. (2016). The effect of wettability on capillary trapping in carbonates.
Advances in Water Resources, 90 (Supplement C), 36-50. doi:10.1016/j.advwatres.2016.02.001
Amott, E. (1959). Observations relating to the wettability of porous rock. Petroleum Transac-
tions, AIME, 216, 156-162.
Anderson, W. (1986a). Wettability literature survey- part 1: Rock/oil/brine interactions and

the effects of core handling on wettability. Journal of Petroleum Technology, 38 (10), 1125-1144.
doi:10.2118/13932-PA
Anderson, W. (1986b). Wettability literature survey- part 2: Wettability measurement. Jour-

nal of Petroleum Technology, 38 (11), 1.246 - 241.262. doi:/10.2118/13933-PA
Anderson, W. (1987a). Wettability literature survey- part 4: Effects of wettability on capillary

pressure. Journal of Petroleum Technology, 39 (10), 1.283 - 281.300. doi:/10.2118/15271-PA
Anderson, W. (1987b). Wettability literature survey part 5: The effects of wettability on

relative permeability. Journal of Petroleum Technology, 39 (11), 1453-1468. doi:10.2118/16323-
PA
Andrew, M. (2015). Reservoir-condition pore-scale imaging of multiphase flow. (PhD thesis),

Imperial College London, London.
Andrew, M., Bijeljic, B., & Blunt, M. J. (2014). Pore-scale contact angle measurements at
reservoir conditions using x-ray microtomography. Advances in Water Resources, 68, 24-31.
doi:10.1016/j.advwatres.2014.02.014
Andrew, M., Menke, H., Blunt, M. J., & Bijeljic, B. (2015). The imaging of dynamic multiphase
fluid flow using synchrotron-based x-ray microtomography at reservoir conditions. Transport
in Porous Media, 110 (1), 1-24. doi:10.1007/s11242-015-0553-2
Arganda-Carreras, I., Kaynig, V., Rueden, C., Eliceiri, K. W., Schindelin, J., Cardona, A., & Se-
138
bastian Seung, H. (2017). Trainable weka segmentation: A machine learning tool for microscopy
pixel classification. Bioinformatics, 33 (15), 2424-2426. doi:10.1093/bioinformatics/btx180
Armstrong, R. T., McClure, J. E., Berrill, M. A., Rücker, M., Schlüter, S., & Berg, S. (2016).
Beyond Darcy’s law: The role of phase topology and ganglion dynamics for two-fluid flow.
Physical Review E, 94 (4), 043113. doi:10.1103/PhysRevE.94.043113
Armstrong, R. T., Ott, H., Georgiadis, A., Rücker, M., Schwing, A., & Berg, S. (2014). Sub-
second pore-scale displacement processes and relaxation dynamics in multiphase flow. Water
Resources Research, 50 (12), 9162-9176. doi:10.1002/2014WR015858
Armstrong, R. T., Porter, M. L., & Wildenschild, D. (2012). Linking pore-scale interfa-
cial curvature to column-scale capillary pressure. Advances in Water Resources, 46, 55-62.
doi:10.1016/j.advwatres.2012.05.009
Arns, C. H., Bauget, F., Limaye, A., Sakellariou, A., Senden, T., Sheppard, A., Sok, R. M.,
Pinczewski, V., Bakke, S., Berge, L. I., Oren, P. E., & Knackstedt, M. A. (2005). Pore scale
characterization of carbonates using x-ray microtomography. SPE Journal, 10 (04), 475 - 484.
doi:10.2118/90368-PA
ASTM International. (2011). Astm e1441-11 standard guide for computed tomography (ct)
imaging. In. West Conshohocken, PA.
Austad, T., Rezaeidoust, A., & Puntervold, T. (2010). Chemical mechanism of low salinity
water flooding in sandstone reservoirs. Paper presented at the SPE Improved Oil Recovery
Symposium, Tulsa, Oklahoma, USA doi:10.2118/129767-MS
Avraam, D. G., & Payatakes, A. C. (1995). Flow regimes and relative permeabilities dur-
ing steady-state two-phase flow in porous media. Journal of Fluid Mechanics, 293, 207-236.
doi:10.1017/S0022112095001698
Bailey, R., & Gray, V. R. (1958). Contact angle measurements of water on coal. Journal of
Applied Chemistry, 8 (4), 197-202. doi:10.1002/jctb.5010080401
Bartels, W.-B., Mahani, H., Berg, S., Menezes, R., van der Hoeven, J. A., & Fadili, A. (2017a).
139
Oil configuration under high-salinity and low-salinity conditions at pore scale: A paramet-
ric investigation by use of a single-channel micromodel. SPE Journal, 22 (05), 1362-1373.
doi:10.2118/181386-PA
Bartels, W.-B., Rücker, M., Berg, S., Mahani, H., Georgiadis, A., Fadili, A., Brussee, N., Coorn,
A., van der Linde, H., & Hinz, C. (2017b). Fast x-ray micro-ct study of the impact of brine
salinity on the pore-scale fluid distribution during waterflooding. Petrophysics, 58 (01), 36-47.
Bartels, W. (2018). Pore scale processes in mixed-wet systems with application to low salinity
waterflooding. (PhD thesis) UU Dept. of Earth Sciences.
Bartels, W., Rücker, M., Boone, M., Bultreys, T., Mahani, H., Berg, S., Hassanizadeh, S., &
Cnudde, V. (2017c). Pore-scale displacement during fast imaging of spontaneous imbibition
Paper presented at the International Symposium of the Society of Core Analysts, Vienna,
Austria.
Basu, S., & Sharma, M. M. (1996). Measurement of critical disjoining pressure for dewetting of
solid surfaces. Journal of Colloid and Interface Science, 181 (2), 443-455.
doi:10.1006/jcis.1996.0401
Batycky, J. P., McCaffery, F. G., Hodgins, P. K., & Fisher, D. B. (1981). Interpreting rela-
tive permeability and wettability from unsteady-state displacement measurements. Society of
Petroleum Engineers Journal, 21 (03), 296 - 308. doi:10.2118/9403-PA
Bear, J. (1988). Dynamics of fluids in porous media: New York (N.Y.) : Dover.
Bear, J. & Bachmat, Y. (1990). Introduction to Modeling of Transport Phenomena in Porous

Media: Dordrecht, The Netherlands : Kluwer Academic Publishers.
Benner, F. C., Riches, W. W., & Bartell, F. E. (1938). Nature and importance of surface forces
in production of petroleum. Paper presented at the Drilling and Production Practice, Amarillo,
Texas.
Berg, S., Armstrong, R., Ott, H., Georgiadis, A., Klapp, S., Schwing, A., Neiteler, R., Brussee,
N., Makurat, A., & Leu, L. (2014). Multiphase flow in porous rock imaged under dynamic flow
140
conditions with fast x-ray computed microtomography. Petrophysics, 55 (04), 304-312.
Berg, S., Armstrong, R. T., Georgiadis, A., Ott, H., Schwing, A., Neiteler, R., Brussee, N.,
Makurat, A., Rücker, M., Leu, L., Wolf, M., Khan, F., Enzmann, F., & Kersten, M. (2015).
Onset of oil mobilization and nonwetting-phase cluster-size distribution. Petrophysics, 56 (01),
15-22.
Berg, S., Ott, H., Klapp, S. A., Schwing, A., Neiteler, R., Brussee, N., Makurat, A., Leu, L.,
Enzmann, F., & Schwarz, J.-O. (2013). Real-time 3D imaging of haines jumps in porous media
flow. Proceedings of the National Academy of Sciences, 110 (10), 3755-3759.
doi:10.1073/pnas.1221373110
Berg, S., Rücker, M., Ott, H., Georgiadis, A., van der Linde, H., Enzmann, F., Kersten,
M., Armstrong, R., de With, S., & Becker, J. (2016). Connected pathway relative per-
meability from pore-scale imaging of imbibition. Advances in Water Resources, 90, 24-35.
doi:10.1016/j.advwatres.2016.01.010
Bernard, G. G. (1967). Effect of floodwater salinity on recovery of oil from cores containing
clays. Paper presented at the SPE California Regional Meeting, Los Angeles, California
Bijeljic, B., Mostaghimi, P., & Blunt, M. J. (2013). Insights into non-fickian solute transport
in carbonates. Water Resources Research, 49 (5), 2714-2728. doi:10.1002/wrcr.20238
Binnig, G., Quate, C. F., & Gerber, C. (1986). Atomic force microscope. Physical Review
Letters, 56 (9), 930-933. doi:10.1103/PhysRevLett.56.930
Blunt, M. J. (1997). Effects of heterogeneity and wetting on relative permeability using pore
level modeling. SPE Journal, 2 (1), 70 - 87. doi:10.2118/36762-PA
Blunt, M. J. (2017). Multiphase flow in permeable media: A pore-scale perspective. Cambridge,

UK: Cambridge University Press.
Brandon, S., Haimovich, N., Yeger, E., & Marmur, A. (2003). Partial wetting of chemically
patterned surfaces: The effect of drop size. Journal of Colloid and Interface Science, 263 (1),
237-243. doi:10.1016/S0021-9797(03)00285-6
141
Broens, M., & Unsal, E. (2018). Emulsification kinetics during quasi-miscible flow in dead-end
pores. Advances in Water Resources, 113, 13-22. doi:10.1016/j.advwatres.2018.01.001
Brooks, R. A., & Di Chiro, G. (1976). Beam hardening in x-ray reconstructive tomography.
Physics in medicine biology, 21 (3), 390.
Broseta, D., Tonnet, N., & Shah, V. (2012). Are rocks still water-wet in the presence of dense
CO2 or H2 S? Geofluids, 12 (4), 280-294. doi:10.1111/j.1468-8123.2012.00369.x
Brown, R. J. S., & Fatt, I. (1956). Measurements of fractional wettability of oil fields’ rocks by
the nuclear magnetic relaxation method. Paper presented at the Fall Meeting of the Petroleum
Branch of AIME, Los Angeles, California. doi:10.2118/743-G
Brownscombe, E. R., & Collins, F. (1950). Pressure distribution in unsaturated oil reservoirs.
Journal of Petroleum Technology, 2 (10), 9-10. doi:/10.2118/950371-G
Buckley, J. S. (2001). Effective wettability of minerals exposed to crude oil. Current Opinion
in Colloid & Interface Science, 6 (3), 191-196. doi:10.1016/S1359-0294(01)00083-8
Buckley, J. S., Bousseau, C., & Liu, Y. (1996). Wetting alteration by brine and crude oil: From
contact angles to cores. SPE Journal, 1 (03), 341-350. doi:10.2118/30765-PA
Buckley, J. S., Liu, Y., & Monsterleet, S. (1998). Mechanisms of wetting alteration by crude
oils. SPE Journal, 3 (1), 54-61. doi:10.2118/37230-PA
Buckley, J. S., Liu, Y., Xie, X., & Morrow, N. R. (1997). Asphaltenes and crude oil wetting -
the effect of oil composition. SPE Journal, 2 (02), 107-119. doi:10.2118/35366-PA
Buckley, J. S., & Lord, D. L. (2003). Wettability and morphology of mica surfaces after exposure
to crude oil. Journal of Petroleum Science and Engineering, 39 (3), 261-273. doi:10.1016/S0920-
4105(03)00067-6
Buckley, S. E., & Leverett, M. C. (1942). Mechanism of fluid displacement in sands. Transac-
tions of the AIME, 146 (01), 107-116. doi:10.2118/942107-G
Bultreys, T., Boone, M. A., Boone, M. N., De Schryver, T., Masschaele, B., Van Loo, D., Van
142
Hoorebeke, L., & Cnudde, V. (2015). Real-time visualization of haines jumps in sandstone with
laboratory-based microcomputed tomography. Water Resources Research, 51 (10), 8668-8676.
doi:10.1002/2015WR017502
Bultreys, T., De Boever, W., & Cnudde, V. (2016). Imaging and image-based fluid transport
modeling at the pore scale in geological materials: A practical introduction to the current
state-of-the-art. Earth-Science Reviews, 155, 93-128. doi:10.1016/j.earscirev.2016.02.001
Bultreys, T., Lin, Q., Gao, Y., Raeini, A. Q., AlRatrout, A., Bijeljic, B., & Blunt, M. J. (2018).
Validation of model predictions of pore-scale fluid distributions during two-phase flow. Physical
Review E, 97 (5), 053104. doi:10.1103/PhysRevE.97.053104
Butt, H. J., & Jaschke, M. (1995). Calculation of thermal noise in atomic force microscopy.
Nanotechnology, 6 (1), 1.
Cassie, A. B. D., & Baxter, S. (1944). Wettability of porous surfaces. Transactions of the
Faraday Society, 40 (0), 546-551. doi:10.1039/TF9444000546
Chandrasekhar, S. (2013). Wettability alteration with brine composition in high temperature

carbonate reservoirs. University of Texas in Austin Petroleum and Geosystems Engineering,
Chatenever, A., & Calhoun Jr, J. C. (1952). Visual examinations of fluid behavior in porous
media-part i. Journal of Petroleum Technology, 4 (06), 149-156. doi:10.2118/135-G
Chauhan, S., Rühaak, W., Khan, F., Enzmann, F., Mielke, P., Kersten, M., & Sass, I. (2016).
Processing of rock core microtomography images: Using seven different machine learning algo-
rithms. Computers & Geosciences, 86, 120-128. doi:10.1016/j.cageo.2015.10.013
Chilingar, G. V., & Yen, T. F. (1983). Some notes on wettability and relative permeabilities
of carbonate reservoir rocks, ii. Energy Sources, 7 (1), 67-75. doi:10.1080/00908318308908076
Chou, S. I., & Shah, D. O. (1981). The optimal salinity concept for oil displacement by oil-
external microemulsions and graded salinity slugs. Journal of Canadian Petroleum Technology,
20 (03). doi:10.2118/81-03-08
143
Churcher, P. L., French, P. R., Shaw, J. C., & Schramm, L. L. (1991). Rock properties of berea
sandstone, baker dolomite, and indiana limestone. Paper presented at the SPE International
Symposium on Oilfield Chemistry, Anaheim, California. doi:10.2118/21044-MS
Cnudde, V., & Boone, M. N. (2013). High-resolution x-ray computed tomography in geo-
sciences: A review of the current technology and applications. Earth-Science Reviews, 123,
1-17. doi:10.1016/j.earscirev.2013.04.003
Cuiec, L. E. (1975). Restoration of the natural state of core samples. Paper presented at the Fall
Meeting of the Society of Petroleum Engineers of AIME, Dallas, Texas. doi:10.2118/5634-MS
Dake, L. (1978). Fundamentals of reservoir engineering. USA: Elsevier, New York.
Darcy, H. (1856). Les fontaines publiques de la ville de dijon: Exposition et application. Paris,
France: Victor Dalmont.
Datta, S. S., Dupin, J.-B., & Weitz, D. A. (2014). Fluid breakup during simultaneous two-
phase flow through a three-dimensional porous medium. Physics of Fluids, 26 (6), 062004.
doi:10.1063/1.4884955
Denekas, M. O., Mattax, C. C., & Davis, G. T. (1959). Effects of crude oil components on rock
wettability. Petroleum Transactions, AIME, 216, 330-333.
Deng, Y., Xu, L., Lu, H., Wang, H., & Shi, Y. (2018). Direct measurement of the contact
angle of water droplet on quartz in a reservoir rock with atomic force microscopy. Chemical
Engineering Science, 177, 445-454. doi:10.1016/j.ces.2017.12.002
Derjaguin, B., & Landau, L. (1993). Theory of the stability of strongly charged lyophobic sols
and of the adhesion of strongly charged particles in solutions of electrolytes. Progress in Surface
Science, 43 (1), 30-59. doi:10.1016/0079-6816(93)90013-L
Dettre, R. H., & Johnson, R. E. (1965). Contact angle hysteresis. Iv. Contact angle mea-
surements on heterogeneous surfaces1. The Journal of Physical Chemistry, 69 (5), 1507-1515.
doi:10.1021/j100889a012
144
Dickinson, L. R., Suijkerbuijk, B. M., Berg, S., Marcelis, F. H., & Schniepp, H. C. (2016a).
Atomic force spectroscopy using colloidal tips functionalized with dried crude oil: A versatile
tool to investigate oil–mineral interactions. Energy & Fuels, 30 (11), 9193-9202.
Dickinson, L. R., Suijkerbuijk, B. M. J. M., Berg, S., Marcelis, F. H. M., & Schniepp, H. C.
(2016b). Atomic force spectroscopy using colloidal tips functionalized with dried crude oil:
A versatile tool to investigate oil–mineral interactions. Energy & Fuels, 30 (11), 9193-9202.
doi:10.1021/acs.energyfuels.6b01862
Dierick, M., Van Loo, D., Masschaele, B., Boone, M., Van Hoorebeke, L., & Cnudde, V. (2013).
Recent micro-ct scanner developments at ugct. Paper presented at the International Conference
on Tomography of Materials and Structures, Ghent, Belgium.
Dishon, M., Zohar, O., & Sivan, U. (2009). From repulsion to attraction and back to repul-
sion: The effect of nacl, kcl, and cscl on the force between silica surfaces in aqueous solution.
Langmuir, 25 (5), 2831-2836. doi:doi.org/10.1021/la803022b
Dixit, A. B., Buckley, J., McDougall, S., & Sorbie, K. (1998a). Core wettability: Should
IAH equal IUSBM . Paper presented at the International Symposium of the SCA, , the Hague,
Netherlands.
Dixit, A. B., McDougall, S. R., & Sorbie, K. S. (1998b). A pore-level investigation of relative
permeability hysteresis in water-wet systems. SPE Journal, 3 (2), 115 - 123. doi:10.2118/37233-
PA
Dobson, K. J., Coban, S. B., McDonald, S. A., Walsh, J. N., Atwood, R. C., & Withers, P. J.
(2016). 4-d imaging of sub-second dynamics in pore-scale processes using real-time synchrotron
x-ray tomography. Solid Earth, 7 (4), 1059-1073. doi:10.5194/se-7-1059-2016
Donaldson, E. C., & Alam, W. (2008a). Chapter 1 - wettability. In E. C. Donaldson & W.

Alam (Eds.), Wettability (pp. 1-55). Houston, Texas, USA: Gulf Publishing Company.
Donaldson, E. C., & Alam, W. (2008b). Chapter 3 - wettability and production. In E. C. Don-
aldson & W. Alam (Eds.), Wettability (pp. 121-172). Houston, Texas, USA: Gulf Publishing
145
Company.
Donaldson, E. C., Thomas, R. D., & Lorenz, P. B. (1969). Wettability determination and
its effect on recovery efficiency. Society of Petroleum Engineers Journal, 9 (01), 13 - 20.
doi:10.2118/2338-PA
Dresen, G., Stanchits, S., & Rybacki, E. (2010). Borehole breakout evolution through acoustic
emission location analysis. International Journal of Rock Mechanics and Mining Sciences,
47 (3), 426-435. doi:10.1016/j.ijrmms.2009.12.010
Drummond, C., & Israelachvili, J. (2004). Fundamental studies of crude oil–surface water
interactions and its relationship to reservoir wettability. Journal of Petroleum Science and
Engineering, 45 (1-2), 61-81. doi:10.1016/j.petrol.2004.04.007
Dufour, R., Semprebon, C., & Herminghaus, S. (2016). Filling transitions on rough surfaces: In-
adequacy of gaussian surface models. Physical Review E, 93 (3), 032802.
doi:10.1103/PhysRevE.93.032802
Dullien, F. A. L. (1992a). 2 - capillarity in porous media. In Porous media (second edition)

(pp. 117-236). San Diego, USA: Academic Press.
Dullien, F. A. L. (1992b). 5 - multiphase flow of immiscible fluids in porous media. In Porous

media (second edition) (pp. 333-485). San Diego, USA: Academic Press.
Eastman, T., & Zhu, D.-M. (1996). Adhesion forces between surface-modified afm tips and a
mica surface. Langmuir, 12 (11), 2859-2862. doi:10.1021/la9504220
Eaton, P., & West, P. (2010). Atomic force microscopy. UK: Oxford University Press.
Extrand, C. W. (2003). Contact angles and hysteresis on surfaces with chemically heterogeneous
islands. Langmuir, 19 (9), 3793-3796. doi:10.1021/la0268350
Fan, T., Wang, J., & Buckley, J. S. (2002). Evaluating crude oils by sara analysis. Paper pre-
sented at the SPE/DOE Improved Oil Recovery Symposium, Tulsa, Oklahoma.
doi:10.2118/75228-MS
146
Fernø, M. A., Torsvik, M., Haugland, S., & Graue, A. (2010). Dynamic laboratory wettability
alteration. Energy & Fuels, 24 (7), 3950-3958. doi:10.1021/ef1001716
Fletcher, P. D. I., Savory, L. D., Woods, F., Clarke, A., & Howe, A. M. (2015). Model study
of enhanced oil recovery by flooding with aqueous surfactant solution and comparison with
theory. Langmuir, 31 (10), 3076-3085. doi:10.1021/la5049612
Freer, E. M., Svitova, T., & Radke, C. J. (2003). The role of interfacial rheology in reservoir
mixed wettability. Journal of Petroleum Science and Engineering, 39 (1), 137-158.
doi:10.1016/S0920-4105(03)00045-7
Frisbie, C. D., Rozsnyai, L. F., Noy, A., Wrighton, M. S., & Lieber, C. M. (1994). Functional
group imaging by chemical force microscopy. Science, 265 (5181), 2071-2074.
doi:10.1126/science.265.5181.2071
Gadelmawla, E.S., Koura, M.M., Maksoud, T.M.A., Elewa, I.M, & Soliman, H.H. (2002)
Roughness parameters. Journal of Materials Processing Technology, 123 (1), 133-145.
doi:10.1016/S0924-0136(02)00060-2
Gao, Y., Lin, Q., Bijeljic, B., & Blunt, M. J. (2017). X-ray microtomography of intermittency in
multiphase flow at steady state using a differential imaging method. Water Resources Research,
53 10,274–10,292.
doi:10.1002/2017WR021736
Garcı́a, R., & Pérez, R. (2002). Dynamic atomic force microscopy methods. Surface Science
Reports, 47 (6), 197-301. doi:10.1016/S0167-5729(02)00077-8
Georgiadis, A., Maitland, G., Trusler, J. P. M., & Bismarck, A. (2011). Interfacial tension
measurements of the (h2o + n-decane + co2) ternary system at elevated pressures and temper-
atures. Journal of Chemical & Engineering Data, 56 (12), 4900-4908. doi:10.1021/je200825j
Giro, R., Bryant, P. W., Engel, M., Neumann, R. F., & Steiner, M. B. (2017). Adsorption en-
ergy as a metric for wettability at the nanoscale. Scientific Reports, 7, 46317.
doi:10.1038/srep46317
147
Hassenkam, T., Andersson, M. P., Hilner, E., Matthiesen, J., Dobberschütz, S., Dalby, K.
N., Bovet, N., Stipp, S. L., Salino, P., & Reddick, C. (2015). Could atomic-force microscopy
force mapping be a fast alternative to core-plug tests for optimizing injection-water salinity for
enhanced oil recovery in sandstone? SPE Journal, 21 (03), 720-729. doi:10.2118/169136-PA
Hassenkam, T., Mathiesen, J., Pedersen, C., Dalby, K., Stipp, S., & Collins, I. R. (2012).
Observation of the low salinity effect by atomic force adhesion mapping on reservoir sandstones.
Paper presented at the SPE Improved Oil Recovery Symposium, Oklahoma, USA.
Hassenkam, T., Pedersen, C. S., Dalby, K., Austad, T., & Stipp, S. L. S. (2011). Pore scale ob-
servation of low salinity effects on outcrop and oil reservoir sandstone. Colloids and Surfaces A:
Physicochemical and Engineering Aspects, 390 (1), 179-188. doi:10.1016/j.colsurfa.2011.09.025
Hassenkam, T., Skovbjerg, L. L., & Stipp, S. L. S. (2009). Probing the intrinsically oil-wet
surfaces of pores in north sea chalk at subpore resolution. Proceedings of the National Academy
of Sciences, 106 (15), 6071-6076. doi:10.1073/pnas.0901051106
Hazlett, R. D. (1995). Simulation of capillary-dominated displacements in microtomographic

images of reservoir rocks. Transport in Porous Media, 20 (1), 21-35. doi:10.1007/bf00616924
Herminghaus, S. (2000). Roughness-induced non-wetting. EPL (Europhysics Letters), 52 (2),

165-170.
Herring, A. L., Harper, E. J., Andersson, L., Sheppard, A., Bay, B. K., & Wildenschild, D.
(2013). Effect of fluid topology on residual nonwetting phase trapping: Implications for geologic
CO2 sequestration. Advances in Water Resources, 62, 47-58. doi:10.1016/j.advwatres.2013.09.015
Herring, A. L., Sheppard, A., Andersson, L., & Wildenschild, D. (2016). Impact of wettability
alteration on 3D nonwetting phase trapping and transport. International Journal of Greenhouse
Gas Control, 46, 175-186. doi:10.1016/j.ijggc.2015.12.026
Hilner, E., Andersson, M. P., Hassenkam, T., Matthiesen, J., Salino, P. A., & Stipp, S. L. S.
(2015). The effect of ionic strength on oil adhesion in sandstone – the search for the low salinity
mechanism. Scientific Reports, 5, 9933. doi:10.1038/srep09933
148
Hilpert, M., & Miller, C. T. (2001). Pore-morphology-based simulation of drainage in totally
wetting porous media. Advances in Water Resources, 24 (3), 243-255. doi:10.1016/S0309-
1708(00)00056-7
Hiorth, A., Cathles, L., & Madland, M. (2010). The impact of pore water chemistry on carbon-
ate surface charge and oil wettability. Transport in Porous Media, 85 (1), 1-21.
doi:10.1007/s11242-010-9543-6
Hirasaki, G. J. (1991). Wettability: Fundamentals and surface forces. SPE Formation Evalu-
ation, 6 (02), 217-226. doi:10.2118/17367-PA
Hirasaki, G. J., Miller, C. A., & Puerto, M. (2008). Recent advances in surfactant eor. Paper
presented at the SPE Annual Technical Conference and Exhibition, Denver, Colorado, USA.
doi:10.2118/115386-MS
Hjelmeland, O. S., & Larrondo, L. E. (1986). Experimental investigation of the effects of

temperature, pressure, and crude oil composition on interfacial properties. SPE Reservoir
Engineering, 1 (04), 321-328. doi:10.2118/12124-PA
Hudson, J. D., & Clements, R. G. (2007). The middle jurassic succession at Ketton, Rutland.
Proceedings of the Geologists’ Association, 118 (3), 239-264. doi:10.1016/S0016-7878(07)80026-
5
Huh, C. (1979). Interfacial tensions and solubilizing ability of a microemulsion phase that coex-
ists with oil and brine. Journal of Colloid and Interface Science, 71 (2), 408-426.
doi:10.1016/0021-9797(79)90249-2
Hutter, J. L., & Bechhoefer, J. (1993). Calibration of atomic-force microscope tips. Review of
Scientific Instruments, 64 (7), 1868-1873. doi:10.1063/1.1143970
Iglauer, S., Fernø, M., Shearing, P., & Blunt, M. (2012). Comparison of residual oil cluster
size distribution, morphology and saturation in oil-wet and water-wet sandstone. Journal of
Colloid and Interface Science, 375 (1), 187-192. doi:10.1016/j.jcis.2012.02.025
149
Iglauer, S., Paluszny, A., Pentland, C. H., & Blunt, M. J. (2011). Residual CO2 imaged with x-
ray micro-tomography. Geophysical Research Letters, 38 (21), n/a-n/a. doi:10.1029/2011gl049680
Iglauer, S., Wu, Y., Shuler, P., Tang, Y., & Goddard, W. A. (2010). New surfactant classes for
enhanced oil recovery and their tertiary oil recovery potential. Journal of Petroleum Science
and Engineering, 71 (1), 23-29. doi:10.1016/j.petrol.2009.12.009
Israelachvili, J. N. (2011a). 1 - historical perspective. In Intermolecular and surface forces

(third edition) (pp. 3-22). San Diego, USA: Academic Press.
Israelachvili, J. N. (2011b). 17 - adhesion and wetting phenomena. In Intermolecular and

surface forces (third edition) (pp. 415-467). San Diego, USA: Academic Press.
Israelachvili, J. N. (2011c). Chapter 12 - force-measuring techniques. In Intermolecular and

surface forces (third edition) (pp. 223-252). San Diego, USA: Academic Press.
Israelachvili, J. N. (2011d). Chapter 14 - electrostatic forces between surfaces in liquids. In

Intermolecular and surface forces (third edition) (pp. 291-340). San Diego, USA: Academic
Press.
Jadhunandan, P. P., & Morrow, N. R. (1991). Spontaneous imbibition of water by crude-oil

brine rock systems. In Situ, 15 (4), 319-345.
Jadhunandan, P. P., & Morrow, N. R. (1995). Effect of wettability on waterflood recovery for
crude-oil/brine/rock systems. SPE Reservoir Engineering, 10 (1), 40-46. doi:10.2118/22597-PA
Javanbakht, G., Arshadi, M., Qin, T., & Goual, L. (2017). Micro-scale displacement of napl by
surfactant and microemulsion in heterogeneous porous media. Advances in Water Resources,
105, 173-187. doi:10.1016/j.advwatres.2017.05.006
Jawitz, J. W., Annable, M. D., Rao, P. S. C., & Rhue, R. D. (1998). Field implemen-
tation of a winsor type i surfactant/alcohol mixture for in situ solubilization of a complex
lnapl as a single-phase microemulsion. Environmental Science & Technology, 32 (4), 523-530.
doi:10.1021/es970507i
150
Joekar-Niasar, V., Hassanizadeh, S. M., & Leijnse, A. (2008). Insights into the relationships
among capillary pressure, saturation, interfacial area and relative permeability using pore-
network modeling. Transport in Porous Media, 74 (2), 201-219. doi:10.1007/s11242-007-9191-7
Johnson, D. J., Al Malek, S. A., Al-Rashdi, B. A. M., & Hilal, N. (2012). Atomic force
microscopy of nanofiltration membranes: Effect of imaging mode and environment. Journal of
Membrane Science, 389, 486-498. doi:10.1016/j.memsci.2011.11.023
Johnson, E. F., Bossler, D. P., & Bossler, V. O. N. (1959). Calculation of relative permeability
from displacement experiments. Petroleum Transactions, AIME, 216, 370-372.
Jung, M., Brinkmann, M., Seemann, R., Hiller, T., Sanchez de La Lama, M., & Herminghaus, S.
(2016). Wettability controls slow immiscible displacement through local interfacial instabilities.
Physical Review Fluids, 1 (7), 074202. doi:10.1103/PhysRevFluids.1.074202
Klise, K. A., Moriarty, D., Yoon, H., & Karpyn, Z. (2016). Automated contact angle estimation
for three-dimensional X-ray microtomography data. Advances in water resources, 95, 152-160.
doi:10.1016/j.advwatres.2015.11.006
Kokal, S. L. (2002). Crude oil emulsions: A state-of-the-art review. Paper presented at the SPE
Annual Technical Conference and Exhibitio, San Antonio, Texas, USA. doi:10.2118/77497-MS
Kumar, K., Dao, E., & Mohanty, K. (2005a). Afm study of mineral wettability with reservoir
oils. Journal of Colloid and Interface Science, 289 (1), 206-217. doi:10.1016/j.jcis.2005.03.030
Kumar, K., Dao, E., & Mohanty, K. (2005b). Atomic force microscopy study of wettability
alteration. Paper presented at the SPE International Symposium on Oilfield Chemistry, The
Woodlands, Texas, USA.
Lager, A., Webb, K., & Black, C. (2007). Impact of brine chemistry on oil recovery. Paper
presented at the 14th European Symposium on Improved Oil Recovery, Cairo, Egypt.
Lager, A., Webb, K., Black, C., Singleton, M., & Sorbie, K. (2008). Low salinity oil recovery-an
experimental investigation1. Petrophysics, 49 (01).
151
Lake, W. L. (1989). Enhanced oil recovery. New Jersey, USA: Prentice-Hall Inc.
Landry, C. J., Karpyn, Z. T., & Ayala, O. (2014). Relative permeability of homogenous-wet
and mixed-wet porous media as determined by pore-scale lattice boltzmann modeling. Water
Resources Research, 50 (5), 3672-3689. doi:10.1002/2013WR015148
Larry, W. L., Russell, J., Bill, R., & Garry, P. (2014). Fundamentals of enhanced oil recovery.
Richardson, USA: SPE.
Le Guen, Y., Renard, F., Hellmann, R., Brosse, E., Collombet, M., Tisserand, D., & Gratier,
J. P. (2007). Enhanced deformation of limestone and sandstone in the presence of high fluids.
Journal of Geophysical Research: Solid Earth, 112 (B5), n/a-n/a. doi:10.1029/2006JB004637
Leach, R. O., Wagner, O. R., Wood, H. W., & Harpke, C. F. (1962). A laboratory and field
study of wettability adjustment in water flooding. Journal of Petroleum Technology, 14 (02),
206-212. doi:10.2118/119-PA
Lehmann, P., Berchtold, M., Ahrenholz, B., Tölke, J., Kaestner, A., Krafczyk, M., Flühler, H.,
& Künsch, H. R. (2008). Impact of geometrical properties on permeability and fluid phase distri-
bution in porous media. Advances in Water Resources, 31 (9), 1188-1204.
doi:10.1016/j.advwatres.2008.01.019
Lenormand, R., Touboul, E., & Zarcone, C. (2006). Numerical models and experiments
on immiscible displacements in porous media. Journal of Fluid Mechanics, 189, 165-187.
doi:10.1017/S0022112088000953
Lenormand, R., & Zarcone, C. (1984). Role of roughness and edges during imbibition in square
capillaries. Paper presented at the SPE Annual Technical Conference and Exhibition, Houston,
Texas, USA. doi:10.2118/13264-MS
Lenormand, R., Zarcone, C., & Sarr, A. (1983). Mechanisms of the displacement of one
fluid by another in a network of capillary ducts. Journal of Fluid Mechanics, 135, 337-353.
doi:10.1017/S0022112083003110
Leu, L., Berg, S., Enzmann, F., Armstrong, R. T., & Kersten, M. (2014). Fast x-ray micro-
152
tomography of multiphase flow in berea sandstone: A sensitivity study on image processing.
Transport in Porous Media, 105 (2), 451-469. doi:10.1007/s11242-014-0378-4
Lever, A., & Dawe, R. A. (1984). Water-sensitivity and migration of fines in the hopeman sand-
stone. Journal of Petroleum Geology, 7 (1), 97-107. doi:10.1111/j.1747-5457.1984.tb00165.x
Leverett, M. C. (1941). Capillary behavior in porous solids. Transactions of the AIME,

142 (01), 152-169. doi:10.2118/941152-G
Liimatainen, V., Vuckovac, M., Jokinen, V., Sariola, V., Hokkanen, M. J., Zhou, Q., & Ras, R.
H. A. (2017). Mapping microscale wetting variations on biological and synthetic water-repellent
surfaces. Nature Communications, 8 (1), 1798. doi:10.1038/s41467-017-01510-7
Liu, F., Yang, H., Wang, J., Zhang, M., Chen, T., Hu, G., Zhang, W., Wu, J., Xu, S., Wu,
X., & Wang, J. (2018). Salinity-dependent adhesion of model molecules of crude oil at quartz
surface with different wettability. Fuel, 223, 401-407. doi:10.1016/j.fuel.2018.03.040
Lowe, A. C., Phillips, M. C., & Riddiford, A. C. (1973). On the wetting of carbonate surfaces
by oil and water. Journal of Canadian Petroleum Technology, 12 (02). doi:10.2118/73-02-04
Ma, S., Morrow, N. R., & Zhang, X. (1995). Generalized scaling of spontaneous imbibition
data for strongly water-wet systems. Paper presented at the Technical Meeting / Petroleum
Conference of The South Saskatchewan Section, Regina, Canada. doi:10.2118/73-02-04
Ma, S. M., & Morrow, N. R. (1994). Effect of firing on petrophysical properties of berea
sandstone. SPE Formation Evaluation, 9 (03), 213-218. doi:10.2118/21045-PA
Ma, S. M., Zhang, X., Morrow, N. R., & Zhou, X. (1999). Characterization of wettability
from spontaneous imbibition measurements. Paper presented at the Annual Technical Meeting,
Calgary, Alberta, Canada.
Magonov, S. N., Cleveland, J., Elings, V., Denley, D., & Whangbo, M. H. (1997). Tapping-mode
atomic force microscopy study of the near-surface composition of a styrene-butadiene-styrene
triblock copolymer film. Surface Science, 389 (1), 201-211. doi:10.1016/S0039-6028(97)00412-3
153
Mahani, H., Berg, S., Ilic, D., Bartels, W.-B., & Joekar-Niasar, V. (2015a). Kinetics of low-
salinity-flooding effect. SPE Journal, 20 (01), 8-20. doi:10.2118/165255-PA
Mahani, H., Keya, A. L., Berg, S., Bartels, W.-B., Nasralla, R., & Rossen, W. R. (2015b).
Insights into the mechanism of wettability alteration by low-salinity flooding (lsf) in carbonates.
Energy & Fuels, 29 (3), 1352-1367. doi:10.1021/ef5023847
Marone, F., Hintermuller, C., Geus, R., & Stampanoni, M. (2008). Towards real-time tomog-
raphy: Fast reconstruction algorithms and gpu implementation. Paper presented at the 2008
IEEE Nuclear Science Symposium Conference Record, Dresden, Germany.
Marone, F., & Stampanoni, M. (2012). Regridding reconstruction algorithm for real-time tomo-
graphic imaging. Journal of Synchrotron Radiation, 19 (6), 1029-1037.
doi:10.1107/S0909049512032864
Martin, Y., Williams, C. C., & Wickramasinghe, H. K. (1987). Atomic force microscope–force
mapping and profiling on a sub 100-å scale. Journal of Applied Physics, 61 (10), 4723-4729.
doi:10.1063/1.338807
Masalmeh, S. K., (2003). The effect of wettability on capillary pressure and relative per-
meability. Journal of Petroleum Science and Engineering, 39, 399–408. doi:10.1016/S0920-
4105(03)00078-0
Mason, G., & Morrow, N. R. (2013). Developments in spontaneous imbibition and possibilities
for future work. Journal of Petroleum Science and Engineering, 110, 268-293.
doi:10.1016/j.petrol.2013.08.018
Mattax, C. C., & Kyte, J. R. (1962). Imbibition oil recovery from fractured, water-drive
reservoir. Society of Petroleum Engineers Journal, 2 (02), 177-184. doi:10.2118/187-PA
Matthiesen, J., Bovet, N., Hilner, E., Andersson, M. P., Schmidt, D. A., Webb, K. J., Dalby, K.
N., Hassenkam, T., Crouch, J., Collins, I. R., & Stipp, S. L. S. (2014). How naturally adsorbed
material on minerals affects low salinity enhanced oil recovery. Energy & Fuels, 28 (8), 4849-
4858. doi:10.1021/ef500218x
154
Maver, U., Velnar, T., Gaberšček, M., Planinšek, O., & Finšgar, M. (2016). Recent progres-
sive use of atomic force microscopy in biomedical applications. TrAC Trends in Analytical
Chemistry, 80, 96-111. doi:10.1016/j.trac.2016.03.014
McCaffery, F. G., & Mungan, N. (1970). Contact angle and interfacial tension studies of
some hydrocarbon-water-solid systems. Journal of Canadian Petroleum Technology, 9 (03).
doi:10.2118/70-03-04
McLean, J. D., & Kilpatrick, P. K. (1997). Effects of asphaltene solvency on stability of water-
in-crude-oil emulsions. Journal of Colloid and Interface Science, 189 (2), 242-253.
doi:10.1006/jcis.1997.4807
McPhee, C., Reed, J., & Zubizarreta, I. (2015). Core analysis: A best practice guide (Vol. 64).
Amsterdam, The Netherlands: Elsevier.
Mennella, A., Morrow, N. R., & Xie, X. (1995). Application of the dynamic wilhelmy plate
to identification of slippage at a liquid-liquid-solid three-phase line of contact. Journal of
Petroleum Science and Engineering, 13 (3), 179-192. doi:10.1016/0920-4105(95)00009-7
Mohanty, K., Davis, H., & Scriven, L. (1987). Physics of oil entrapment in water-wet rock.
SPE Reservoir Engineering, 2 (01), 113-128. doi:10.2118/9406-PA
Morrow, N. R. (1975). The effects of surface roughness on contact: Angle with special reference
to petroleum recovery. Journal of Canadian Petroleum Technology, 14 (04). doi:10.2118/75-04-
04
Morrow, N. R. (1976). Capillary pressure correlations for uniformly wetted porous media.
Journal of Canadian Petroleum Technology, 15 (04). doi:10.2118/76-04-05
Morrow, N. R., Cram, P. J., & McCaffery, F. G. (1973). Displacement studies in dolomite
with wettability control by octanoic acid. Society of Petroleum Engineers Journal, 13 (04).
doi:10.2118/3993-PA
Mugele, F., Bera, B., Cavalli, A., Siretanu, I., Maestro, A., Duits, M., Cohen-Stuart, M., van
den Ende, D., Stocker, I., & Collins, I. (2015). Ion adsorption-induced wetting transition in
155
oil-water-mineral systems. Scientific Reports, 5, 10519. doi:10.1038/srep10519
Mugele, F., Siretanu, I., Kumar, N., Bera, B., Wang, L., de Ruiter, R., Maestro, A., Duits,
M., van den Ende, D., & Collins, I. (2016). Insights from ion adsorption and contact-angle
alteration at mineral surfaces for low-salinity waterflooding. SPE Journal, 21 (04), 1203-1213.
doi:10.2118/169143-PA
Müller, D. J., & Engel, A. (1997). The height of biomolecules measured with the atomic
force microscope depends on electrostatic interactions. Biophysical Journal, 73 (3), 1633-1644.
doi:10.1016/S0006-3495(97)78195-5
Murison, J., Semin, B., Baret, J.-C., Herminghaus, S., Schröter, M., & Brinkmann, M. (2014).
Wetting heterogeneities in porous media control flow dissipation. Physical Review Applied,
2 (3), 034002. doi:10.1103/PhysRevApplied.2.034002
Muskat, M. (1949). The physical properties and behavior of petroleum fluids (1949 ppop
chapter 2). In Physical principles of oil production. (pp. 29-113). New York, USA: McGraw-
Hill Book Co.
Mutterlose, J., & Bornemann, A. (2000). Distribution and facies patterns of lower cretaceous
sediments in northern germany: A review. Cretaceous Research, 21 (6), 733-759.
doi:10.1006/cres.2000.0232
Myers, G., Varslot, T., Kingston, A., Herring, A., & Sheppard, A. (2012). Ground-truth
verification of dynamic x-ray micro-tomography images of fluid displacement. Paper presented
at the SPIE Optical Engineering + Applications, San Diego, USA.
Naidich, Y. V., Voitovich, R. P., & Zabuga, V. V. (1995). Wetting and spreading in hetero-
geneous solid surface-metal melt systems. Journal of Colloid and Interface Science, 174 (1),
104-111. doi:10.1006/jcis.1995.1370
Nasralla, R. A., Sergienko, E., Masalmeh, S. K., van der Linde, H. A., Brussee, N. J., Mahani,
H., Suijkerbuijk, B., & Alqarshubi, I. (2014). Demonstrating the potential of low-salinity water-
flood to improve oil recovery in carbonate reservoirs by qualitative coreflood. Paper presented
156
at the Abu Dhabi International Petroleum Exhibition and Conference, Abu Dhabi, UAE.
Neumann, A. W., & Good, R. J. (1979). Techniques of measuring contact angles. In R. J. Good
& R. R. Stromberg (Eds.), Surface and colloid science: Volume 11: Experimental methods (pp.
31-91). Boston, MA: Springer US.
Nordahl, K., & Ringrose, P. S. (2008). Identifying the representative elementary volume for
permeability in heterolithic deposits using numerical rock models. Mathematical geosciences,
40 (7), 753-771. doi:10.1007/s11004-008-9182-4
Nutting, P. G. (1930). Physical analysis of oil sands. AAPG Bulletin, 14 (10), 1337-1349.
Ohler, B. (2007, 27.06.2018). Practical advice on the determination of cantilever spring con-
stants. Bruker Application Note# AN94. Retrieved from http://nanoscaleworld.bruker-axs.com/
nanoscaleworld/media/p/143.aspx
Øren, P.-E., & Bakke, S. (2003). Reconstruction of berea sandstone and pore-scale mod-
elling of wettability effects. Journal of Petroleum Science and Engineering, 39 (3), 177-199.
doi:10.1016/S0920-4105(03)00062-7
Osoba, J. S., Richardson, J. G., Kerver, J. K., Hafford, J. A., & Blair, P. M. (1951). Labo-
ratory measurements of relative permeability. Journal of Petroleum Technology, 3 (02), 47-56.
doi:10.2118/951047-G
Overbeek, J. T. G., & Verwey, E. (1948). Theory of the stability of lyophobic colloids: The
interaction of sol particles having an electric double layer. Mineola, New York, USA: Dover
Publication Inc.
Paganin, D., Mayo, S., Gureyev, T. E., Miller, P. R., & Wilkins, S. W. (2002). Simultaneous
phase and amplitude extraction from a single defocused image of a homogeneous object. Journal
of Microscopy, 206 (1), 33-40. doi:10.1046/j.1365-2818.2002.01010.x
Pak, T. (2015). Saturation tracking and identification of residual oil saturation (PhD thesis),
University of Edinburgh.
157
Pak, T., Archilha, N. L., Mantovani, I. F., Moreira, A. C., & Butler, I. B. (2018). The Dynamics
of Nanoparticle-enhanced Fluid Displacement in Porous Media-A Pore-scale Study. Scientific
reports, 8 (1), 11148. doi:10.1038/s41598-018-29569-2
Pepper, J. F., De Witt, W., & Demarest, D. F. (1954). Geology of the bedford shale and berea
sandstone in the appalachian basin. Science, 119 (3094), 512-513.
doi:10.1126/science.119.3094.512-a
Pichon, C., & Lynch, J. (2005). Synchrotron radiation and oil industry research. Oil & Gas
Science and Technology - Rev. IFP, 60 (5), 735-746. doi:10.2516/ogst:2005052
Por, N. (1992). Stability properties of petroleum products. Tel Aviv, Israel: Israel Institute of
Petroleum and Energy.
Putz, A., Chevalier, J. P., Stock, G., & Philippot, J. (1981). A field test of microemul-
sion flooding, chateaurenard field, france. Journal of Petroleum Technology, 33 (04), 710-718.
doi:10.2118/8198-PA
Quéré, D. (2008). Wetting and roughness. Annual Review of Materials Research, 38 (1), 71-99.
doi:10.1146/annurev.matsci.38.060407.132434
Raina, G. (2013). Atomic force microscopy as a nanometrology tool: Some issues and future
targets. MAPAN, 28 (4), 311-319. doi:10.1007/s12647-013-0085-6
Rezaei, N., & Firoozabadi, A. (2014). Macro- and microscale waterflooding performances of
crudes which form w/o emulsions upon mixing with brines. Energy & Fuels, 28 (3), 2092-2103.
doi:10.1021/ef402223d
Robin, M., Paterson, A., Cuiec, L., & Yang, C. (1997). Effects of chemical heterogeneity and
roughness on contact angle hysteresis: Experimental study and modeling. Paper presented at the
International Symposium on Oilfield Chemistry, Houston, Texas, USA. doi:10.2118/37291-MS
Romanuka, J., Hofman, J., Ligthelm, D. J., Suijkerbuijk, B., Marcelis, F., Oedai, S., Brussee,
N., van der Linde, H., Aksulu, H., & Austad, T. (2012). Low salinity eor in carbonates. Paper
presented at the SPE Improved Oil Recovery Symposium, Oklohoma, USA.
158
Rücker, M., Bartels, W.-B., Unsal, E., Berg, S., Brussee, N., Coorn, A., & A, B. (2017). The
formation of microemulsion at flow conditions in rock. Paper presented at the International
Symposium of theSociety of Core Analysts, Vienna, Austria.
Rücker, M., Berg, S., Armstrong, R., Georgiadis, A., Ott, H., Schwing, A., Neiteler, R., Brussee,
N., Makurat, A., & Leu, L. (2015a). From connected pathway flow to ganglion dynamics.
Geophysical Research Letters, 42 (10), 3888–3894. doi:10.1002/2015GL064007
Rücker, M., Berg, S., Armstrong, R., Georgiadis, A., Ott, H., Simon, L., Enzmann, F., Kersten,
M., & de With, S. (2015b). The fate of oil clusters during fractional flow: Trajectories in
the saturation-capillary number space Paper presented at the International Symposium of the
Society of Core Analysts, St. John’s Newfoundland and Labrador, Canada.
Sader, J. E. (1998). Frequency response of cantilever beams immersed in viscous fluids with ap-
plications to the atomic force microscope. Journal of Applied Physics, 84 (1), 64-76.
doi:10.1063/1.368002
Sader, J. E., Chon, J. W. M., & Mulvaney, P. (1999). Calibration of rectangular atomic force mi-
croscope cantilevers. Review of Scientific Instruments, 70 (10), 3967-3969. doi:10.1063/1.1150021
Salager, J. L., Bourrel, M., Schechter, R. S., & Wade, W. H. (1979). Mixing rules for optimum
phase-behavior formulations of surfactant/oil/water systems. Society of Petroleum Engineers
Journal, 19 (05), 271-278. doi:10.2118/7584-PA
Salathiel, R. A. (1973). Oil recovery by surface film drainage in mixed-wettability rocks. Jour-
nal of Petroleum Technology, 25 (10), 1216-1224. doi:10.2118/4104-PA
Saxena, N., Hofmann, R., Alpak, F. O., Berg, S., Dietderich, J., Agarwal, U., Tandon, K.,
Hunter, S., Freeman, J., & Wilson, O. B. (2017). References and benchmarks for pore-scale
flow simulated using micro-ct images of porous media and digital rocks. Advances in Water
Resources, 109, 211-235. doi:10.1016/j.advwatres.2017.09.007
Scanziani, A., Singh, K., Blunt, M. J., & Guadagnini, A. (2017). Automatic method for
estimation of in situ effective contact angle from x-ray micro tomography images of two-phase
159
flow in porous media. Journal of Colloid and Interface Science, 496 (Supplement C), 51-59.
doi:10.1016/j.jcis.2017.02.005
Scanziani, A., Singh, K., Bultreys, T., Bijeljic, B., Blunt, M. J. (2018). In situ characterization
of immiscible three-phase flow at the pore scale for a water-wet carbonate rock. Advances in
Water Resources, 121, 446-455. doi:10.1016/j.advwatres.2018.09.010
Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., Preibisch,
S., Rueden, C., Saalfeld, S., Schmid, B., Tinevez, J.-Y., White, D. J., Hartenstein, V., Eliceiri,
K., Tomancak, P., & Cardona, A. (2012). Fiji: An open-source platform for biological-image
analysis. Nature Methods, 9, 676. doi:10.1038/nmeth.2019
Schlüter, S., Berg, S., Rücker, M., Armstrong, R. T., Vogel, H. J., Hilfer, R., & Wildenschild,
D. (2016). Pore-scale displacement mechanisms as a source of hysteresis for two-phase flow in
porous media. Water Resources Research, 52 (3), 2194-2205. doi:10.1002/2015WR018254
Schmatz, J., Urai, J. L., Berg, S., & Ott, H. (2015). Nanoscale imaging of pore-scale fluid-fluid-
solid contacts in sandstone. Geophysical research letters, 42 (7), 2189-2195. doi:10.1002/2015GL063354
Schmid, K. S., & Geiger, S. (2013). Universal scaling of spontaneous imbibition for arbitrary
petrophysical properties: Water-wet and mixed-wet states and handy’s conjecture. Journal of
Petroleum Science and Engineering, 101, 44-61. doi:10.1016/j.petrol.2012.11.015
Schrader, M. E., & Yariv, S. (1990). Wettability of clay minerals. Journal of Colloid and
Interface Science, 136 (1), 85-94. doi:10.1016/0021-9797(90)90080-8
Sezgin, M., & Sankur, B. (2004). Survey over image thresholding techniques and quantitative
performance evaluation. Journal of Electronic Imaging, 13 (1), 20. doi:10.1117/1.1631315
Sheng, J. J. (2014). Critical review of low-salinity waterflooding. Journal of Petroleum Science

and Engineering, 120, 216-224. doi:10.1016/j.petrol.2014.05.026
Silin, D., & Patzek, T. (2006). Pore space morphology analysis using maximal inscribed spheres.
Physica A: Statistical Mechanics and its Applications, 371 (2), 336-360.
doi:10.1016/j.physa.2006.04.048
160
Singh, K., Bijeljic, B., & Blunt, M. J. (2016). Imaging of oil layers, curvature and contact angle
in a mixed-wet and a water-wet carbonate rock. Water Resources Research, 52 (3), 1716-1728.
doi:10.1002/2015WR018072
Singh, K., Menke, H., Andrew, M., Lin, Q., Rau, C., Blunt, M. J., & Bijeljic, B. (2017a).
Dynamics of snap-off and pore-filling events during two-phase fluid flow in permeable media.
Scientific Reports, 7 (1), 5192. doi:10.1038/s41598-017-05204-4
Singh, K., Scholl, H., Brinkmann, M., Michiel, M. D., Scheel, M., Herminghaus, S., & Seemann,
R. (2017b). The role of local instabilities in fluid invasion into permeable media. Scientific
Reports, 7, 444. doi:10.1038/s41598-017-00191-y
Sjoblom, J. (2005). Emulsions and emulsion stability: Surfactant science series 132 (Vol. 132).
Boca Raton, USA: CRC Press.
Skrettingland, K., Holt, T., Tweheyo, M. T., & Skjevrak, I. (2011). Snorre low-salinity-water
injection–coreflooding experiments and single-well field pilot. SPE Reservoir Evaluation &
Engineering, 14 (02), 182-192. doi:10.2118/129877-PA
Snoeijer, J. H., & Andreotti, B. (2013). Moving contact lines: scales, regimes, and dynamical
transitions.Annual review of fluid mechanics, 45, 269-292. doi:10.1146/annurev-fluid-011212-
140734
Stampanoni, M., Groso, A., Isenegger, A., Mikuljan, G., Chen, Q., Bertrand, A., Henein, S.,
Betemps, R., Frommherz, U., & Böhler, P. (2006). Trends in synchrotron-based tomographic
imaging: The sls experience. Paper presented at the SPIE Optics+ Photonics, San Diego, USA.
Stark, M., Möller, C., Müller, D. J., & Guckenberger, R. (2001). From images to interactions:
High-resolution phase imaging in tapping-mode atomic force microscopy. Biophysical Journal,
80 (6), 3009-3018. doi:10.1016/S0006-3495(01)76266-2
Stegemeier, G. L. (1977). Mechanisms of entrapment and mobilization of oil in porous media.

In D. O. Shah & R. S. Schechter (Eds.), Improved oil recovery by surfactant and polymer flooding
(pp. 55-91). New York, USA: Academic Press.
161
Suijkerbuijk, B., Kuipers, H., Van Kruijsdijk, C., Berg, S., Van Winden, J., Ligthelm, D.,
Mahani, H., Almada, M. P., Van den Pol, E., & Niasar, V. J. (2013). The development of a
workflow to improve predictive capability of low salinity response. Paper presented at the IPTC
2013: International Petroleum Technology Conference, Beijing, China.
Sun, M., Mogensen, K., Bennetzen, M., & Firoozabadi, A. (2016). Demulsifier in Injected
Water for Improved Recovery of Crudes That Form Water/Oil Emulsions. SPE Reservoir
Evaluation Engineering, 19 (04), 664-672. doi:10.2118/180914-PA
Swain, P. S., & Lipowsky, R. (1998). Contact angles on heterogeneous surfaces: A new look at
cassie’s and wenzel’s laws. Langmuir, 14 (23), 6772-6780. doi:10.1021/la980602k
Tagavifar, M., Xu, K., Jang, S. H., Balhoff, M. T., & Pope, G. A. (2017). Spontaneous and
flow-driven interfacial phase change: Dynamics of microemulsion formation at the pore scale.
Langmuir, 33 (45), 13077-13086. doi:10.1021/acs.langmuir.7b02856
Tang, G.-Q., & Morrow, N. R. (1999). Influence of brine composition and fines migration on
crude oil/brine/rock interactions and oil recovery. Journal of Petroleum Science and Engineer-
ing, 24 (2), 99-111. doi:10.1016/S0920-4105(99)00034-0
Tang, G. Q., & Morrow, N. R. (1997). Salinity, temperature, oil composition, and oil recovery
by waterflooding. SPE Reservoir Engineering, 12 (04), 269-276. doi:10.2118/36680-PA
Teschke, O., & de Souza, E. F. (2003). Hydrophobic surfaces probed by atomic force microscopy.
Langmuir, 19 (13), 5357-5365. doi:10.1021/la0340450
Thomas, M. M., Clouse, J. A., & Longo, J. M. (1993). Adsorption of organic compounds on
carbonate minerals: 1. Model compounds and their influence on mineral wettability. Chemical
Geology, 109 (1), 201-213. doi:10.1016/0009-2541(93)90070-Y
Toulhoat, H., Prayer, C., & Rouquet, G. (1994). Characterization by atomic force microscopy
of adsorbed asphaltenes. Colloids and Surfaces A: Physicochemical and Engineering Aspects,
91, 267-283. doi:10.1016/0927-7757(94)02956-3
Treiber, L. E., & Owens, W. W. (1972). A laboratory evaluation of the wettability of fifty oil-
162
producing reservoirs. Society of Petroleum Engineers Journal, 12 (06), 531-540. doi:10.2118/3526-
PA
Unsal, E., Broens, M., & Armstrong, R. T. (2016). Pore scale dynamics of microemulsion
formation. Langmuir, 32 (28), 7096-7108. doi:10.1021/acs.langmuir.6b00821
Valvatne, P. H., & Blunt, M. J. (2004). Predictive pore-scale modeling of two-phase flow in
mixed wet media. Water Resources Research, 40 (7), W07406. doi:10.1029/2003WR002627
van der Vegte, E. W., & Hadziioannou, G. (1997). Scanning force microscopy with chemical
specificity: An extensive study of chemically specific tip- surface interactions and the chemical
imaging of surface functional groups. Langmuir, 13 (16), 4357-4368. doi:10.1021/la970025k
Verri, I., Della Torre, A., Montenegro, G., Onorati, A., Duca, S., Mora, C. A., Radaelli,
F., & Trombin, G. (2017). Development of a digital rock physics workflow for the analysis
of sandstones and tight rocks. Journal of Petroleum Science and Engineering, 156, 790-800.
doi:10.1016/j.petrol.2017.06.053
Vig, J. R. (1985). UV/ozone cleaning of surfaces. Journal of Vacuum Science Technology A:

Vacuum, Surfaces, and Films, 3 (3), 1027-1034. doi:10.1116/1.573115
Vincent, L., & Soille, P. (1991). Watersheds in digital spaces: An efficient algorithm based
on immersion simulations. IEEE Transactions on Pattern Analysis and Machine Intelligence,
13 (6), 583-598. doi:10.1109/34.87344
Vlassenbroeck, J., Dierick, M., Masschaele, B., Cnudde, V., Van Hoorebeke, L., & Jacobs,
P. (2007). Software tools for quantification of x-ray microtomography at the ugct. Nuclear
Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors
and Associated Equipment, 580 (1), 442-445. doi:10.1016/j.nima.2007.05.073
Vogel, H. J., Weller, U., & Schlüter, S. (2010). Quantification of soil structure based on
minkowski functions. Computers & Geosciences, 36 (10), 1236-1245.
doi:10.1016/j.cageo.2010.03.007
Warren, J. E., & Calhoun, J. C., Jr. (1955). A study of waterflood efficiency in oil-wet systems.
163
Petroleum Transactions, AIME, 204, 22-29.
Welge, H. J. (1952). A simplified method for computing oil recovery by gas or water drive.
Journal of Petroleum Technology, 4 (04), 91-98. doi:10.2118/124-G
Wenzel, R. N. (1936). Resistance of solid surfaces to wetting by water. Industrial & Engineering
Chemistry, 28 (8), 988-994. doi:10.1021/ie50320a024
White, M., Pierce, K., & Acharya, T. (2017). A review of wax-formation/mitigation technolo-
gies in the petroleum industry. SPE Production & Operations, Preprint. doi:10.2118/189447-
PA
Wildenschild, D., Culligan, K. A., & Christensen, B. S. B. (2004). Application of x-ray micro-
tomography to environmental fluid flow problems. Paper presented at the Optical Science and
Technology, the SPIE 49th Annual Meeting, Denver, USA.
Wildenschild, D., Hopmans, J. W., Rivers, M. L., & Kent, A. J. R. (2005). Quantitative
analysis of flow processes in a sand using synchrotron-based x-ray microtomography. Vadose
Zone Journal, 4 (1), 112-126. doi: 10.2136/vzj2005.0112
Wildenschild, D., & Sheppard, A. P. (2013). X-ray imaging and analysis techniques for quan-
tifying pore-scale structure and processes in subsurface porous medium systems Advances in
Water Resources, 51, 217-246. doi:10.1016/S0022-1694(02)00157-9
Wolansky, G., & Marmur, A. (1998). The actual contact angle on a heterogeneous rough surface
in three dimensions. Langmuir, 14 (18), 5292-5297. doi:10.1021/la960723p
Wolansky, G., & Marmur, A. (1999). Apparent contact angles on rough surfaces: The wenzel
equation revisited. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 156 (1),
381-388. doi:10.1016/S0927-7757(99)00098-9
Xie, X., Morrow, N. R., & Buckley, J. S. (2002). Contact angle hysteresis and the stability
of wetting changes induced by adsorption from crude oil. Journal of Petroleum Science and
Engineering, 33 (1), 147-159. doi:10.1016/S0920-4105(01)00182-6
164
Yadali Jamaloei, B., & Kharrat, R. (2009). Fundamental study of pore morphology effect in low
tension polymer flooding or polymer–assisted dilute surfactant flooding. Transport in Porous
Media, 76 (2), 199-218. doi:10.1007/s11242-008-9243-7
Yildiz, H. O., & Morrow, N. R. (1996). Effect of brine composition on recovery of moutray
crude oil by waterflooding. Journal of Petroleum Science and Engineering, 14 (3), 159-168.
doi:10.1016/0920-4105(95)00041-0
Ying, G., Qingyang, L., Branko, B., & J., B. M. (2017). X-ray microtomography of intermit-
tency in multiphase flow at steady state using a differential imaging method. Water Resources
Research, 53 (12), 10274-10292. doi:10.1002/2017WR021736
Young, T. (1805). Iii. An essay on the cohesion of fluids. Philosophical Transactions of the
Royal Society of London, 95, 65-87. doi:10.1098/rstl.1805.0005
Yousef, A. A., Al-Saleh, S., & Al-Jawfi, M. S. (2012). Improved/enhanced oil recovery from
carbonate reservoirs by tuning injection water salinity and ionic content. Paper presented at
the SPE Improved Oil Recovery Symposium, Tulsa, USA.
Yousef, A. A., Al-Saleh, S. H., Al-Kaabi, A., & Al-Jawfi, M. S. (2011). Laboratory investiga-
tion of the impact of injection-water salinity and ionic content on oil recovery from carbonate
reservoirs. SPE Reservoir Evaluation & Engineering, 14 (05), 578-593. doi:10.2118/137634-PA
Youssef, S., Rosenberg, E., Deschamps, H., Oughanem, R., Maire, E., & Mokso, R. (2014).
Oil ganglia dynamics in natural porous media during surfactant flooding captured by ultra-fast
x-ray microtomography. Paper presented at the Symposium of the Society of Core Analysts,
Avignon, France.
Zhao, B., MacMinn, C. W., & Juanes, R. (2016). Wettability control on multiphase flow in
patterned microfluidics. Proceedings of the National Academy of Sciences, 113 (37), 10251-
10256. doi:10.1073/pnas.1603387113
Zhao, X., Blunt, M. J., & Yao, J. (2010). Pore-scale modeling: Effects of wettability on water-
flood oil recovery. Journal of Petroleum Science and Engineering, 71 (3), 169-178.
165
doi:10.1016/j.petrol.2010.01.011
Zou, S., Armstrong, R. T., Arns, J.-Y., Arns, C. H., & Hussain, F. (2018). Experimental and
theoretical evidence for increased ganglion dynamics during fractional flow in mixed-wet porous
media. Water Resources Research, 54 (5), 3277-3289. doi:10.1029/2017WR022433
166
Appendix A: Detailed list of flow
experiments
Table A1 shows the the complete flow sequence of experiments 1A-1E, 2F, 3A, 4A and B.
Table A2 shows the flow sequence of the experiments 2A-E which were contducted right after
experiment 1A-E.
167
Table A1: List of flooding experiments conducted during this work.
Nr Rock Initialization Desaturation Aging Flood (secondary mode)

(Brine) time
Fluid Flow rate PV 30 bars Fluid Flow rate PV
injected 70 C injected
µl/min PV/min µl/min PV/min
1A Ketton HS 1 Crude A 500 8.7 17.4 1 week HS 1 30 0.52 17.4

1B Estaillades HS 1 Crude A 500 6.6 13.3 1 week HS 1 30 0.40 13.3
1C Ketton HS 1 Crude B 500 8.7 17.4 1 week HS 1 30 0.52 17.4
1D Estaillades HS 1 Crude B 500 6.6 13.3 1 week HS 1 30 0.40 13.3
168
1E Berea HS 1 Crude C 500 9.94 20.0 1 week HS 1 30 0.60 20

2F Ketton LS Crude A 500 8.7 17.4 1 week HS 1 30 0.52 17.4
3A Gildehauser FW 2 doped 500 9.5 19.0 - FW 2 30 0.57 19.0
n-decane
4A Ketton HS 2 doped 1000 7.7 193.0 - - - - -
n-decane 500 3.8 38.0
4B Ketton HS 2 crude A Initialized by centrifugation 4 weeks HS 2 1000 7.7 193
3500 RPM for 24 h 500 3.8 38
Table A2: List of flooding experiments in tertiary mode conducted during this work.
Nr previous Rock Initialization Desaturation Aging Flood Flood (tertiary mode)

experiment time (secondary
Nr mode)
Fluid Fluid 30 bars Fluid Fluid µl/min PV/min PV
70 C injected
2A 1A Ketton HS 1 Crude A 1 week HS 1 LS 30 0.52 17.4

2B 1B Estaillades HS 1 Crude A 1 week HS 1 LS 30 0.40 13.3
2C 1C Ketton HS 1 Crude B 1 week HS 1 LS 30 0.52 17.4
2D 1D Estaillades HS 1 Crude B 1 week HS 1 LS 30 0.40 13.3
169
2E 1E Berea HS 1 Crude C 1 week HS 1 LS 30 0.60 20

3A 3A Gildehauser FW 2 doped - FW 2 surfactant 30 0.57 19.0
n-decane solution
Appendix B:
µCT for assessment of wettability
µCT used within the context of multiphase flow in rock is often considered challenging as it
is significantly impacted by the image analysis workflow. Various studies addressed this issue
(Chauhan et al., 2016; Leu et al., 2014; Saxena et al., 2017; Verri et al., 2017). To ensure
sufficient quality, the best approach is to perform the analysis with different methods and
boundary conditions to check whether the observed trend or effect is robust. In this work
methods and boundaries were chosen based on the analysis of previous studies (Armstrong et
al., 2014; Berg et al., 2015; Leu et al., 2014) and based on results of Bartels (2018) obtained
for the data set discussed in this work.
However, an additional aspect to be considered for µCT investigations on wettability, a property

depending on specific fluid-fluid-rock compositions, is the choice of contrast agent. In order to
distinguish the two fluid phases their absorption coefficients need to differ. This can be achieved
through the addition of a substance with high density and atomic number to one of the fluids.
Often ions such as Cs or I are dissolved in aqueous phases. In this study, KI was chosen
as its ions are within the same group in the periodic table as NaCl, which natively occurs in
formation- and seawater. However, the high amount of KI required for sufficient contrast limits
the suitability of this agent for studies on low salinity flooding. Based on the ionic strength
2.7 wt% KI correspond to an ionic strength of 9500 ppm NaCl, which is considered the upper
boundary for the low salinity effect to occur (Mahani et al., 2015b).
While this is sufficient to study general concepts related to low salinity flooding, this contrast
170
agent is not suitable to assess the impact of different brine composition.
An alternative is to dope the oleic phase with organic substances containing compounds of
high density and atomic number. In this work, iodo-decane was used for Amott spontaneous
imbibition tests. Contrast agent within brine, which in Amott tests is surrounding the sample
would lead to a significant reduction of image quality. A crude-oil/iodo-decane mixture of
20 wt-% was found to be optimal (Bartels, 2018). The disadvantage of this approach is the
significant dilution of surface-active components within the oleic phase as these are crucial for
wettability alteration processes. Furthermore, the contrast agent within oil was found to cause
photochemical reactions like bubble formation and disintegration when exposed to the high
intense light from a beamline. Figure B1 shows examples of such processes. This effect was not
observed in experiments conducted in bench-top µCT scanners.
Figure B1: The contrast agent within oil was found to cause photochemical reactions like bubble
formation (a) and disintegration (b) when exposed to the high intense light from a beamline
(adapted from Bartels et al., 2017b). The images were taken in 10s intervals.
171
Appendix C
Sensitivity of results to image
processing (AFM)
In this work, AFM was used for the first time as a tool for fluid film detection within rock.
Additional experiments (7A,7B) were conducted in order to assess the sensitivity of this method.
To evaluate the impact of background noise and imaging velocity an atomically flat surface
of a freshly cleaved muscovite was measured (7A). An example for such an image is shown in
Figure B2a. The noise ranges between 0 pm and 800 pm.
Figure C1: a shows the image of the muscovite surface at a tip velocity of 120 µm/s and z-
length of 1000 nm. The noise level is 800 pm, which is 4x larger than the molecular structure
of muscovite. b shows the noise level for different tip velocities and z-lengths. Only with a tip
velocity of 2.5 µm/s atomic resolution may be achieved. A higher z-length results also in a
higher noise level.
With this image quality features larger 800 pm would appear well recognisable. Smaller features
might be difficult to assess. In QITM -mode the imaging velocity can be controlled through two
172
parameters: z- length, which corresponds to the distance the tip is moving when approaching
and disengaging from the surface, and pixel time, defining the time the approach and disen-
gaging process takes. The tip velocity can be increased by reducing the z-length and pixel
time. While different z-distance/ pixel-time configurations may result in the same imaging
time, the noise level might differ. The noise level obtained for different z-distance/ pixel time
configurations are shown in Fig B2b.
In general, the image quality becomes better with lower tip velocity. The smallest noise level
achieved was 200 pm, which corresponds to atomic resolution (”QITM mode - Quantitative
imaging with the NanoWizard 3 AFM”, 2018). The increase in z-distance showed a larger
impact on the noise level than the pixel time. However, to avoid crashing the tip into the
rough surfaces of the sample a high z-distance is required. Rough surfaces may also lead to
image artefacts as shown in Figure B3a measured on an artificially roughened calcite sample
(experiment 7B). The tip rasters the surface line by line along one direction (fast axis). If the
height difference between one pixel and the next pixel is smaller than the chosen z-distance the
frequency of the cantilever gets disturbed leading to artificial stripes along the fast axis.
Figure C2: Rough surfaces may lead to image artefacts like stripes (a). Furthermore, surface
structures as shown in the height map (b) may impact adhesion forces obtained. Due to the
larger contact area of the tip the adhesion is higher in dents than on top of hills. The adhesion
map corresponding to (b) is shown in (c).
An additional artefact caused by roughness is the false adhesion detection as shown in Figure
B3b,c. The adhesion map (Figure B3 c) corresponds to the height map shown in Figure B3 b.
The measurements were taken on a cleaned Ketton rock. Surface chemistry and the adhesion
should be homogenous. However, the dents of the surface appear more adherent than the hills.
173
The reason for that is the larger surface area the tip interacts within a dent than at the top of
a hill. Correspondingly, this artefact can be reduced by using sharper and longer tips.
A rough surface may impact a force distance curve during retrace, but, it does not impact
the trace curve. Correspondingly, this part of the force distance curve is suitable for the
determination of the height of fluid/fluid interfaces. However, other artefacts may impact the
detection of fluid films. One AFM image may contain up to 1024 x 1024 data points which can
only be processed through automatic approaches. The fluid/fluid interface is detected through
the sudden deflection occurring. Such a deflection may also be caused by noise or vibration
as shown in Figure B4. There are two ways to resolve this issue: Either by decreasing the
sensitivity of the automatic detection, which leads to a lower resolution of the fluid film, or by
applying a median filter removing the false pixels.
Figure C3: 3D reconstruction of fluid films measured with AFM (a) can be strongly affected by
noise. The reason is the sensitivity of the automatic fluid / fluid-interface detection, which is
based on a sudden deflection when approaching the surface. Such deflection may also be caused
by noise or vibrations. b shows a force distance curve affected by noise. Such image artefacts
can be reduced by decreasing the sensitivity, which also decreases the resolution of the fluid /
fluid interface or by applying a median filter.
174
Appendix D
Determination of emulsion in rock
All flooding experiments (experiments 1A-E), revealed the presence of a third phase next to
the doped brine and the undoped oleic phase (Figure B5).
Figure D1: Next to brine and oil a third phase, emulsion, was detected in all studied COBR
systems with HS as brine. (a) shows Estaillades rock with crude A, (b) Estaillades with crude
B, (c) Berea sandstone with crude C and (d) Ketton rock with crude A.
To disclose image artefacts related to reconstruction, filter and beam-hardening, different re-
175
construction methods and filters were tested (Bartels et al., 2017c). To test the ability of the
chosen fluid-fluid systems to form emulsion tube tests were performed. In such tests two fluids
are filled into a test tube and mixed by rigorous shaking. The fluids are then left to stagnate
for several hours. Tube tests before and after the experiment conducted for different fluid-brine
systems are shown in Figure B6. All crudes, except the combination of crude B with low salinity
brine (LS), were found to form emulsion. Such emulsion forms when surface-active components
of the crude oil interact with the aqueous phase (Kokal, 2002; McLean & Kilpatrick, 1997;
Pak, 2015; Rezaei & Firoozabadi, 2014; Sun et al., 2016) or may be induced by nanoparticles
(Pak et al., 2018). To assess how emulsion appears in µCT images one tube was scanned. The
images revealed that the emulsion phase was built of water droplets ranging from 500 µm down
to image resolution. Such water-droplets may also appear in a size below resolution, leading
to partial volume effects as observed within the rock (Figure B5). Larger water droplets were
only observed in the oil-wet system (Figure 5.1d).
Figure D2: Test tubes before (a) and after (b) rigorous shaking show the ability of different fluid-
fluid system to form emulsion. The test tube containing crude B and LS brine was investigated
further by µCT (c), which revealed a water-in-oil emulsion. The water droplets observed range
from a size of 500 µm down to image resolution (adapted from Bartels et al., 2017b).
The emulsion phase represents a challenge for segmentation. Figure B7 shows that the emulsion
phase does not cause a distinguishable peak in an image histogram. Grey values ranging from
100% brine to 100% oil were detected. However, these intermediate grey values may also
originate from partial volume effects related to movement.
176
Figure D3: The histogram obtained from an image of experiment 1A shows that the Ketton
rock separates brine from crude oil. Partial volume effects, which may be related to the presence
of emulsion spread in between brine and oil.
The difference can be assessed manually on a case to case basis. Partial volume effects due
to movement occur solely during the event and correspondingly in one image only, while the
presence of emulsion persists through multiple consecutive images as illustrated in Figure B8.
Corresponding assessment of all images leads to the conclusion that grey values close to the
grey value of the rock and above represent brine.
Figure D4: Grey-scale images obtained with in 1 min intervals during experiment 1A. The red
circles highlight an emulsion filling event replacing brine. The emulsion appears at the same
location in consecutive images and the grey value of the emulsion remains constant. The blue
circles highlight a brine filling event: Due to a partial volume effect the grey value of the brine
appears lower during the event. In the consecutive image the partial volume effect disappears.
The emulsion and oleic phase showed only a subtle difference in grey value though, the experi-
ments were conducted with a high amount of contrast agent (17 wt% KI-brine) to mimic high
salinity brine. At lower concentration (10 wt% KI-brine) the emulsion phase might remain
177
invisible. The high oil-content of the water-in-oil emulsion justifies the assumption that the
emulsion behaves similar to the oil-phase. Yet, the increase in viscosity may impact the overall
core-scale flow behaviour (Section 4.2.2). To ensure that the findings at the pore-scale were not
compromised by the presence of emulsion, contact angle measurements were obtained solely
manually at brine-oil interfaces from the original grey scale images and global measures as fluid
distribution and event counts were displayed for both phases separately. The oil-filling event
examples for mixed-wet systems shown in Figure 5.4 occurred in the absence of emulsion. The
absence of emulsion was ensured by assessing the grey scale image in 3D for the FOV for all
time steps as illustrated in Figure B9.
Figure D5: To ensure absence of emulsion in the examples displayed in Figure 5.4 the 3D grey
scale images for all times steps were assessed within the FOV.
178
Appendix E
Fluid distribution as a function of time
The obtained µCT images allow the tracking of fluid phases within the rock.
Figure B10 shows the fluid distribution of the oil- and emulsion phase for all time steps recorded
during the HS 1 flood in Ketton (experiment 1A). The yellow phase represents the oil and
emulsion connected to the outlet of the flow cell. The red phase shows the disconnected oil and
emulsion. The time sequence contains various disconnection and reconnection events, which
are associated with ganglion dynamics. Figure B11 shows the emulsion-phase only. Though
the phase represented a water-in-oil emulsion, it cannot be treated as oil. Considering emulsion
as oil leads to an overestimation of the oil-phase permeability. The reason may be an increased
viscosity . However, the contribution of the emulsion-phase to the oil- permeability cannot be
disregarded either (Section 4.2.2). The difference in emulsion distribution between various time
steps (e.g.: 1 min and 2 min, 10 min and 11 min) shows that this phase keeps redistributing
and therefore behaves mobile.
Similarly, the brine phase was tracked. As the high salinity brine (HS) and low salinity brine
(LS) contain different amounts of contrast agents, experiments in which the brine is switched
can be also used to study diffusion and mixing processes during two phase flow within rock. In
experiment 2F the sample was prepared and aged with HS brine and crude A. Solely LS brine
was injected during the waterflood. The corresponding images shown in Figure B12 illustrate
how formation water may get diluted or replaced during a waterflood.
179
180
Figure E1: Time sequence of experiment 1A showing the fluid distribution of the oil- and
emulsion phase. The yellow phase represents oil and emulsion disconnected to the outlet and
the red phase connected oil and emulsion. Circles highlight reconnection events associated with
ganglion dynamics.
181
182
Figure E2: Time sequence of experiment 1A showing the fluid distribution of the emulsion
phase only. Circles highlight the change in fluid distribution.
183
Figure E3: Time sequence showing the dilution of the high salinity brine (HS; dark blue-green)
during displacement by LS brine (experiment 2F).
184
Appendix F
Measurement of an oil-layer
Water films at connate water saturation can be detected by AFM (Section 7.1). Oil layers at
residual oil saturation can be imaged, too. However, test experiments (10B) in which a calcite
was first split within model oil (n-decane) and then flushed with seawater did not show any
fluid-film formation. The reason for that may be the water-wetness of the sample. The low
adhesion of the oil towards the surface may have let the oil to disconnect when in contact with
the tip. Similarly, test experiments (10C) in which a calcite sample was split in crude A did
not show any oil-layer. Unlike n-decane, the crude contained surface active components. As
the calcite surface was immediately in contact with the crude, one would expect an immediate
ageing. Yet, oil layers were only detected after exposure to elevated temperature and pressure
(experiments 10D, 11B). An adhesion map and force distance curves related to such an oil layer
are shown in Figure B13.
In contrast to water films, the oil-layer appeared repulsive towards the tip. Once the tip passed
through the fluid-fluid interface the deflection decreased (Figure B13b). In some cases, the tip
even became attracted (Figure B13c). Further approach caused again a repelling behaviour.
Once the tip got in contact with the surface, the distance remained constant. The disengage-
ment process showed an adhesive response at the rock surface while the tip was still in oil,
though, still with a repelling deflection. Once the tip snaps-off the surface, it follows the same
track as during the engaging process, starting with a repelling deflection towards an attractive
deflection. Meniscus forces lead to an additional strong adhesion signal before the tip moves
185
back into zero position.
Figure F1: The adhesion map of a Ketton surface at residual oil-saturation reveals a crude oil
layer covering the full 4.5 µm x 4.5 µm image. The force distance curves obtained during the
measurement show a repelling behaviour of the tip to the fluid-fluid interface. Once the tip
passes through the interface it gets further deflected before coming into contact with the surface
(b). In some cases, the deflection gets attractive once the fluid/fluid interface is transgressed
(c). The cause of this behaviour is unknown.
One intriguing question is the origin of the deflection after the tip passes through the fluid-fluid
interface, especially as the same path is followed during the disengagement process. Three
theories have been developed to explain this behaviour:
1. Viscosity of the oil-phase
One theory is, that the viscosity of the oil prevents the tip to reach equilibrium. The force
distance curve recorded would not display the actual force acting at position z. Fluid dynamics
overlay this signal. While this theory would explain why the tip jumps from attraction to
deflection once it passes through the interface, this aspect would not explain why the tip
follows the same track when disengaging. If viscosity would dominate the tip position, this
effect would be opposed during disengagement.
However, the impact of viscosity on the force distance curve could easily be assessed by de-
creasing the tip velocity.
2. Meniscus forces at the fluid-fluid-interface
Another theory is, that the movement of the fluid/fluid contact line along the tip surface causes
the deflection. Tip roughness allows any contact angles ranging between the receding and
186
advancing extreme to form. When the tip pushes through the fluid-fluid interface the meniscus
of the surface may come to rest at a receding contact angle position. Further movement of the
tip leads to a switch from receding to advancing, which is reflected in the change of deflection
as illustrated in Figure B14. During disengagement the tip would follow the same path.
Figure F2: The deflection measured after the tip passed through the fluid/fluid interface could
be caused by meniscus forces. A negative deflection would occur for a receding contact angle
(a) and a positive for an advancing contact angle (b). Advancing and receding contact angles
are depending on the surface roughness of the tip. c shows the movement of a droplet over a
rough surface and the relationship of the advancing/receding contact angle to flow direction.
Meniscus forces would be able to explain the full force distance curve shown in Figure B13c.
However, if this theory would hold, the deflection after the tip pushes through the fluid-fluid
interface should be independent of the layer thickness. Figure B13b shows, that this is not the
case. In fact, the initial deflection after the tip passes through the fluid-fluid interface seems to
depend on the layer thickness.
3. Chemical gradient within the oil phase
A third theory is, that during aging the crude oil develops a chemical gradient towards the
surface. The tip passes first through attractive components and then through repulsive com-
ponents before it gets in contact with the actual rock surface. However, more experiments, e.g.
with different tip charges, would be needed to confirm this view.
187

Wettability and Wettability Alteration at The Pore - and Nano - Scales

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wettability and Wettability Alteration at The Pore - and Nano - Scales

Uploaded by

Copyright:

Available Formats

Imperial College London

Department of Earth Science and Engineering

Wettability and wettability alteration at the

Submitted in part fulfilment of the requirements for the degree of

Bartels, W.-B., M. Rücker, M. Boone, T. Bultreys, H. Mahani, S. Berg, S. Hassanizadeh, and V.

Bartels, W.-B., M. Rücker, S. Berg, H. Mahani, A. Georgiadis, A. Fadili, N. Brussee, A. Coorn,

M. Rücker, W.-B. Bartels, S. Berg, H. Mahani, H. Ott, A. Georgiadis, N. Brussee, A. Coorn,

M. Rücker, W.-B. Bartels, S. Berg, H. Mahani, A. Georgiadis, H. Ott; N. Brussee, A. Coorn,

M. Rücker, W.-B. Bartels, M. Boone, T. Bultreys, H. Mahani, S. Berg, S.M. Hassanizadeh,

M. Rücker, K. Singh, G. Garfi, A. Scanziani, S. Yesufu, W.-B. Bartels, M. Blunt, S.Berg, A.

Furthermore, I would like to acknowledge my supervisors, Apostolos Georgiadis, Steffen Berg,

General Douglas MacArthur

Copyright declaration iii

List of publications and conference proceedings vii

2.1.1 Molecular interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.1.2 Sub-pore scale surface structure . . . . . . . . . . . . . . . . . . . . . . . 12

2.1.3 Pore scale fluid distribution and flow behaviour . . . . . . . . . . . . . . 16

2.1.4 Core scale capillary-pressure and relative-permeability . . . . . . . . . . . 23

2.1.5 Connecting the length scales . . . . . . . . . . . . . . . . . . . . . . . . . 30

2.2.1 Low salinity flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2.2.2 Surfactant flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

3 Materials and methods 39

3.1 Fluid and rock systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

3.2 Micro computed tomography (µCT) . . . . . . . . . . . . . . . . . . . . . . . . . 48

3.2.1 Equipment and facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.2.2 µCT core-flooding experiments . . . . . . . . . . . . . . . . . . . . . . . 51

3.2.3 Spontaneous (Amott) imbibition . . . . . . . . . . . . . . . . . . . . . . 59

3.3 Atomic force microscopy (AFM) . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

3.3.1 Force spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

3.3.2 Topographical imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

3.3.3 Wettability assessment by AFM . . . . . . . . . . . . . . . . . . . . . . . 70

3.4 Supporting data and experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 78

4 Core-scale wettability characterization 81

4.1 Amott spontaneous imbibition tests . . . . . . . . . . . . . . . . . . . . . . . . . 82

4.1.1 Water-wet systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

4.1.3 Impact of sample handling protocols . . . . . . . . . . . . . . . . . . . . 86

4.2 Steady-state experiments in mixed-wet systems . . . . . . . . . . . . . . . . . . 87

4.2.1 Relative permeability at the core scale . . . . . . . . . . . . . . . . . . . 88

4.2.2 Impact of emulsion on the oil-phase connectivity . . . . . . . . . . . . . . 89

4.2.3 Comparison of core-scale relative permeability with pore-scale data . . . 91

4.3 Impact of sample initialization on wettability . . . . . . . . . . . . . . . . . . . . 95

4.4 Summary: Dependency of core-scale parameters on sample handling protocols . 97

5 Impact of wettability on pore-scale flow regimes 99

5.1 Wettability classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

5.2 Pore scale processes during waterflooding . . . . . . . . . . . . . . . . . . . . . . 104

5.2.1 Physics of oil filling events in water-wet systems . . . . . . . . . . . . . . 104

5.2.2 Physics of oil filling events in mixed-wet systems . . . . . . . . . . . . . . 106

5.3 Summary: Enhanced ganglion dynamics in mixed-wet systems . . . . . . . . . . 109

6 Enhanced oil recovery at the pore scale 111

6.1 Low-salinity flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

6.2 Surfactant flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

7 Impact of nano-scale structures on wettability 121

7.1 Detection of fluid films . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

7.2 Impact of surface roughness on wettability . . . . . . . . . . . . . . . . . . . . . 124

7.3 Summary: Relationship between surface roughness and wettability . . . . . . . . 127

3.2 List of rocks used in this study. . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3.3 List of minerals used in this work. . . . . . . . . . . . . . . . . . . . . . . . . . . 44

3.7 List of flooding experiments conducted during this work. . . . . . . . . . . . . . 54

3.8 List of spontaneous imbibition experiments performed in this study. . . . . . . . 60

3.9 List of AFM experiments performed in this study. . . . . . . . . . . . . . . . . . 71

A1 List of flooding experiments conducted during this work. . . . . . . . . . . . . . 168

2.14 An unsteady state flow experiment can be described by the Buckley-Leverett

2.16 Schematic drawings of relative permeability-saturation functions for a water-wet

cos θ R cos θi (2.3)

Pc ∆ρgh 2σcospθ q . (2.5)

Fcap AporePc Apore r σ , (2.6)

Fvis Ablob∆P Ablob∆L kµw q, (2.7)

Ncmacro FFvis AAblob∆Lµ

Ncmicro µσw q , (2.9)

Iw S S1 S and Io S1 SS4 , (2.15)

IU SBM logp AA1 q (2.16)