Conference PaperPDF Available

In-situ, real time micro-CT imaging of pore scale processes, the next frontier for laboratory based micro-CT scanning


Abstract and Figures

Over the past decade, laboratory based X-ray computed micro-tomography (micro-CT) has given unique insights in the internal structure of complex reservoir rocks, improving the understanding of pore scale processes and providing crucial information for pore scale modelling. Especially in-situ imaging using X-ray optimized Hassler type cells has enabled the direct visualization of fluid distributions at the pore scale under reservoir conditions. While sub-micrometre spatial resolutions are achievable in lab-based micro-CT, the temporal resolutions are still limited to minutes or hours. This time restriction is often a bottleneck for imaging dynamic in-situ processes, thus limiting the applicability to relatively slow pore scale processes occurring in the order of hours to days, or to end points in drainage-imbibition cycles. To overcome this issue, X-ray Engineering (XRE) and Ghent University's Centre for X-ray Tomography (UGCT) have jointly developed a gantry-based micro-CT system. This system's X-ray tube and detector rotate continuously in a horizontal plane around the fixed sample. The setup still allows to tune the geometrical magnification, with spatial resolutions down to 5 µm. This fixed sample setup is also ideal for in-situ imaging, as the flow cells can be directly connected to high pressure flow tubing and sensor lines, without the need to allow rotational movement relative to the X-ray source and detector. An efficient hardware design with a fast flat panel detector, combined with custom X-ray transparent flow cells to increase X-ray flux and dedicated 4D software tools in acquisition, reconstruction and analysis, allows to reach temporal resolutions in the order of seconds. The possibilities of this new approach in dynamic in-situ imaging are illustrated with flow tests on a carbonate sample. We discuss the challenges in dynamic imaging and present methods to improve X-ray flux and optimize image quality by means of this experiment. Furthermore, we show that the integration of fast imaging experiments with other information from peripheral sensors or from imaging data at different resolutions can help to link behaviour at the pore scale to the effective properties at the core scale, but also facilitates the experimental workflow.
Content may be subject to copyright.
In-situ, real time micro-CT imaging of pore scale processes, the
next frontier for laboratory based micro-CT scanning
Boone, Marijn1,2, Bultreys, Tom2, Masschaele, Bert1, Van Loo, Denis1, Van Hoorebeke,
Luc3, Cnudde, Veerle2
1X-Ray Engineering bvba
2UGCT/PProGRess, Dept. of Geology, Ghent University
3UGCT, Dept. of physics and astronomy, Ghent University
This paper was prepared for presentation at the International Symposium of the Society of Core
Analysts held in Snowmass, Colorado, USA, 21-26 August 2016
Over the past decade, laboratory based X-ray computed micro-tomography (micro-CT)
has given unique insights in the internal structure of complex reservoir rocks, improving
the understanding of pore scale processes and providing crucial information for pore
scale modelling. Especially in-situ imaging using X-ray optimized Hassler type cells has
enabled the direct visualization of fluid distributions at the pore scale under reservoir
conditions. While sub-micrometre spatial resolutions are achievable in lab-based micro-
CT, the temporal resolutions are still limited to minutes or hours. This time restriction is
often a bottleneck for imaging dynamic in-situ processes, thus limiting the applicability
to relatively slow pore scale processes occurring in the order of hours to days, or to end
points in drainage-imbibition cycles.
To overcome this issue, X-ray Engineering (XRE) and Ghent University’s Centre for X-
ray Tomography (UGCT) have jointly developed a gantry-based micro-CT system. This
system’s X-ray tube and detector rotate continuously in a horizontal plane around the
fixed sample. The setup still allows to tune the geometrical magnification, with spatial
resolutions down to 5 µm. This fixed sample setup is also ideal for in-situ imaging, as the
flow cells can be directly connected to high pressure flow tubing and sensor lines,
without the need to allow rotational movement relative to the X-ray source and detector.
An efficient hardware design with a fast flat panel detector, combined with custom X-ray
transparent flow cells to increase X-ray flux and dedicated 4D software tools in
acquisition, reconstruction and analysis, allows to reach temporal resolutions in the order
of seconds.
The possibilities of this new approach in dynamic in-situ imaging are illustrated with
flow tests on a carbonate sample. We discuss the challenges in dynamic imaging and
present methods to improve X-ray flux and optimize image quality by means of this
experiment. Furthermore, we show that the integration of fast imaging experiments with
other information from peripheral sensors or from imaging data at different resolutions
can help to link behaviour at the pore scale to the effective properties at the core scale,
but also facilitates the experimental workflow.
X-ray computed microtomography (micro-CT) has the unique ability to obtain reliable
high resolution 3D information inside otherwise non-transparent objects in a non-
destructive manner. Over the past 2 decades micro-CT evolved from a synchrotron
technique to a standard laboratory microscopy technique. In this period, the achievable
spatial resolution of the laboratory based micro-CT systems has improved drastically,
reaching resolutions up to 50nm for some systems. This evolution has had important
implications in the field of geosciences, as it enabled detailed microscopic studies of the
internal structure of geological samples in 3D, while before this was limited to 2D
techniques like optical or scanning electron microscopy. It was especially challenging to
describe the pore space of a rock, which is essentially a 3D property of a rock, and
therefore required statistical or process based modelling to extrapolate the 2D
information to a 3D model [1]. The evolution in micro-CT imaging enabled the direct
visualization and characterization of the pore space in 3D at scales below the micrometer,
even for complex carbonate pore systems.
Besides the static pore structure characterization, the non-destructive nature of micro-CT
also makes it possible to visualize different fluids inside the pore space and to monitor
how these fluids migrate through the rock. This is in recent years often referred to as in-
situ micro-CT and implies the imaging of a sample under certain well-constrained
external conditions [2-8]. In pore scale studies these conditions are usually increased
pressure and temperature in order to obtain an insight in for example reactive flow,
multiphase flow and rock mechanics under reservoir conditions. Traditional flow studies
at reservoir conditions are performed by placing reservoir samples in large stainless steel
Hassler type flow cells and monitoring the fluid transport through the samples using
different sensors (e.g. pressure, pH and electrical conductivity) and analyzing the
chemical composition of the fluid coming in and going out the cell. Based on the
information of these different sensors and chemical data, assumptions are made on the
pore scale processes occurring in the sample. However, what happens inside individual
pores in the sample remains inaccessible. In-situ micro-CT imaging makes it possible to
visualize fluid distribution in the pore space and if and how the pore structure of the rock
changes throughout the experiment.
Traditional Hassler type flow cells are however not practical for in-situ imaging due to
their size and constitution. An in-situ set-up for X-ray imaging requires a custom built,
miniature version of a traditional Hassler type cell. The first requirement of such a
custom built is the diameter of the cell. This diameter has to be kept as small as possible,
as most laboratory based micro-CT systems use a geometrical magnification to obtain
high resolutions. The second requirement is the composition of the cell. The cell needs to
be made from X-ray transparent materials, in order to ensure that a sufficient X-ray flux
reaches the detector. These in-situ cells are therefore usually constructed out of weakly
X-ray attenuating metals like aluminium [9] or strong polymers like PEEK[9] or carbon
fiber [8,11,12] fiber instead of stainless steel.
When a fluid is injected into a reservoir sample in an in-situ setup, the fluid moves trough
the pore space and can interact with other fluids in the pore space or with the rock itself.
By monitoring these interactions over time using micro-CT imaging it is possible to
capture the dynamics of pore scale processes. In-situ imaging is therefore often related to
dynamic or 4D (3D + time) imaging. However, to acquire reliable pore scale 3D images,
the sample needs to remain unchanged during the micro-CT acquisition, otherwise
motion artefacts and image blurring can occur. To avoid these motion artefacts, the
temporal resolution of the micro-CT system should be high enough to capture the
dynamic changes occurring inside the sample while maintaining e.g. flow conditions.
While the spatial resolution of laboratory micro-CT system has improved during the last
years, the temporal resolution has remained in the order of minutes to hours. This limits
the type of pore scale processes that can be visualized using traditional laboratory micro-
CT systems to slow processes (hours to days), like mineral-fluid interactions in CCS
studies [13], or to quasi-static fluid distributions during drainage-imbibition cycles [14].
Monitoring fast (seconds to minutes) pore scale processes mostly remained restricted to
synchrotron facilities where temporal resolutions below 1 second can be obtained.
To tackle this issue, we have developed a benchtop scale gantry-based micro-CT system
which is optimized for dynamic in-situ imaging of pore scale processes (figure 1). The
temporal resolution of the system is around 12 seconds, which is an order of magnitude
higher compared to standard micro-CT systems. This is illustrated in figure 2, where an
overview of the spatial and temporal resolution of laboratory and synchrotron based
micro-CT systems is given.
In this manuscript the possibilities of this gantry-based micro-CT system are illustrated
by a solute transport experiment using a tracer salt inside a carbonate sample.
Traditionally, a tracer salt would be pumped through the sample (while mounted in a
flow cell) and the conductivity at the outlet would be monitored. Based on these
measurements one can investigate the dispersive behavior of the porous sample, and
determine if the transport of a solute through the sample is dominated by advective or
diffusive processes. Here we augment this data by directly visualizing – spatially
resolved - how the tracer salt is moving through the pore space of the carbonate.
While this is an example of single phase flow visualization, dynamic in-situ imaging is
also applicable to multiphase flow. As an example of fast pore scale imaging of
multiphase flow, Bultreys et al. (2015) presented the visualization of Haines jumps in a
sandstone sample with the same laboratory setup as used this work[16].
Experimental setup and optimizations for dynamic imaging
For the in-situ solute transport experiment a simple flow cell with confining pressure was
used (figure 1). Because this experiment is conducted at low pressure conditions, the used
flow cell was constructed out of polymethylmethacrylate (PMMA), which can be
considered as quasi transparent for X-rays, especially in comparison to the carbonate
sample. In the flow cell, a carbonate sample of 6 mm diameter and 16 mm in height was
mounted in a viton sleeve. A confining pressure of 10 bar was placed around the sleeve
using a manual syringe pump. A MilliGAT high-precision continuous flow pump
controlled the flow through the sample. To the outlet of the cell an electrical conductivity
sensor was placed as an indication to the brine salinity. The diameter of the entire flow
cell is 25 mm, allowing to obtain a spatial resolution of 7 µm and a field of view which
covers the entire diameter of the carbonate sample.
The carbonate sample investigated in this experiment was the Savonnières limestone,
which has a porosity ranging from 22% to 40% and a permeability from 115 to more than
2000 mD, depending on local variations [20]. It is a grain supported oolithic limestone
consisting of ooids and shell fragments which are overgrown by sparite. During
diagenesis, some grain fragments were partially dissolved, resulting in a pore network
with well-connected pores between the grains (intergranular porosity) and secondary
porosity inside the dissolved grains (intragranular porosity or vuggy porosity), which is
connected to the rest of the pore network through micropores [17]. A high-quality micro-
CT 2D slice through the sample with a resolution of 7.3µm is given in figure 3. The
resolution is sufficient to capture the macroporosity in the sample, but unable to capture
the microporous connections between some larger pore bodies.
The dried limestone sample was flushed with CO2 to remove the air phase in the pores.
Afterwards the sample was flushed with distilled water for a period of 2 hours to obtain a
complete water saturation. Scans before and after water flushing were used to evaluate
the water saturation degree and check for potential dissolution effects in the sample due
to the CO2 dissolution in the pore fluid. No dissolution effects were apparent in the CT
images at a resolution of 7.3µm. To investigate dispersive solute transport in the sample,
a salt solution of 10 wt.% CsCl was injected, because Cs acts as a tracer due to its high
X-ray attenuation coefficient. The change of the salt concentration in the outgoing fluid
was measured using the conductivity sensor on the in-situ flow cell, while the changes in
the distribution of the salt solution in the pore space of the carbonate sample were
continuously monitored by dynamic micro-CT imaging. The experiment was performed
twice on the same carbonate sample at volumetric injection rates of 0.25 µl/s in the first
run and 1 µl/s in the second run (with sufficient clean water flushing in between the
In order to obtain a 3D image of the sample with X-ray micro-CT, radiographs of the
sample have to be taken at different angles, which requires a rotation of the sample
relatively to the source and detector. The self-developed in-situ cell setup in the described
experiment had 2 flowlines going towards the cell (inlet and confining pressure), 1
flowline going out of the sample and 1 sensor wire coming from the sample. A more
complex setup often has even more (high pressure) flow and sensor lines connected to the
cell, which makes a rotational movements of the cell challenging. For time lapse micro-
CT imaging, where scans are acquired at a time interval of typically several hours or days
it is possible to perform a full rotation and then return to the original position, causing
less problems with fluid or sensor line tangling. Monitoring fast dynamic processes
requires a continuous rotation of the in-situ cell setup, which in turn requires very
complex in-situ cell designs with slip ring hydraulic/sensor contacts or with integration of
pumps in the in-situ cell [18].
The fixed sample configuration in our gantry based setup is ideal as (high pressure)
tubing or sensor lines remain immobile, thus avoiding entanglement, vibrations during
scanning and possible flow instabilities in the fluids going towards the sample.
The micro-CT systems design is optimized for fast dynamic in-situ imaging[19]. The X-
ray source and detector are mounted on a gantry, which can be continuously rotated at a
maximum speed of 30°/s or 12 seconds for 360° rotation. Apart from the rotational
movement, the gantry can also perform a translational movement to change the distance
between the X-ray source and the detector allowing to tune the geometrical magnification
of the sample (figure 1). The X-ray source is a compact closed type transmission source
with a maximum tube voltage of 130kV and a maximum power of 39W. The tube has a
focal spot of 5 µm, which is also the highest achievable spatial resolution with the
system. The detector is a CMOS flat panel with a thick CsI scintillator and a readout
speed of 30 frames per second at a full resolution and 60 frames per second in 2 x 2
binned mode. The sample stage can be moved vertically with a travel of 0.75 m. This
offers flexibility to mount different types of in-situ equipment, allows to perform stacked
scans of elongated core samples for a more representative overview and even follow slow
moving fluid fronts through the sample.
Before performing dynamic acquisition, a high-quality 3D image of the pore structure
was obtained in a normal acquisition of 30 minutes at a voxel size of 7.3 µm, a 100 kV
tube voltage and 6W tube power (figure 3). For the dynamic scan the detector was used
in a 2 x 2 binned mode resulting in a voxel size of 14.6 µm and an increase of the signal
to noise in the micro-CT image. The tube voltage and tube power were increased to 130
kV and 16 W respectively, to increase the X-ray flux reaching the detector and therefore
decrease the noise level in the micro-CT image. The total acquisition time was 15
minutes and each 360 degree rotation took 12 seconds, resulting in a total of 45000
projection images. This data was recorded and processed with the proprietary 4D tools
(XRE, Ghent, Belgium) in the ACQUILA software (UGCT/XRE, Ghent, Belgium).
Dynamic acquisitions generate a massive amount of data and require dedicated smart
reconstruction and analysis tools to the desired and useful information of the pore scale
process under investigation. The acquired projections were automatically analysed and
differences in radiographies were used to pinpoint changes in the pore space of the
sample. Data from external sources like sensor data can also be incorporated and synced
with the continuous stream of X-ray projections to augment the data and avoid redundant
data from being reconstructed. Tomographic reconstruction was performed with the FDK
algorithm, implemented on the GPU.
Because the acquisition was continuous, projections acquired during any full rotation of
the system could be reconstructed regardless of the starting angle. This is very useful
when discrete events like fracture formation or sudden pore filling events like Haines
jumps are visualized, as a reconstruction can be done just before and just after the event.
This avoids image blurring and motion artefacts. In these experiments the pore scale
process is a continuous process without discrete events. Therefore the angle between
every consecutive reconstruction equalled 360 degrees, resulting in a full 3D image every
12 seconds.
The effluent salt concentration was calculated based on the electrical conductivity
measurements, resulting in breakthrough curves for both experiments. The two curves for
the 0.25 µl/s and the 1 µl/s experiments are shown in figure 4. In the breakthrough curve
of the 1 µl/s flow experiment, we can see an almost instantaneous increase in salt
concentration of the effluent fluid, indicating that the dispersion in the system is mainly
controlled by advective processes.. We used a non-linear least squares analysis
implemented in STANMOD to roughly estimate the dispersion coefficient based on the
breakthrough curves. This yielded an effective dispersion coefficient of 3.89E-3 mm2/s
for the 1 µl/s experiment. Assuming a tortuosity of 24.4 and a porosity of 26% in
Savonnières [20], the effective diffusion coefficient of Cs in this rock is estimated at
1.83E-5 mm2/s. This confirms that the behavior in the well-connected macropores is
In the 0.25 µl/s constant flow experiment the diffusive processes play a larger role in the
dispersion coefficient, but generally it is also an advective dominated system (estimated
dispersion coefficient 1.44E-3 mm2/s). From this curve it is however also clear that the
solute transport process takes longer than the scanned timeframe of 15 minutes, so only a
part of the process was captured with in-situ dynamic imaging.
Vertical slices trough the reconstructed volumes of the dynamic acquisition are given in
figure 5 and figure 6. In the raw vertical slices the Cs-concentration rise within the pore
space can be clearly seen. In the 1 µl/s experiment the fluid front reaches the bottom of
the sample after about 1 minute and reaches the top about 1 minute later. The
heterogeneity of the advective flow field is clearly visible: the Cs-concentration in some
well-connected macropores lags behind. After about 3 minutes however, the well-
connected pore space is fully saturated with the salt. Some pores in the system behave
rather differently and show a significantly slower concentration increase. Most of these
pores are ooids that were dissolved during diagenesis of the carbonate and which are only
connected to the rest of the pore network through micropores. In these microporous
connections, the advective flow of the solute is limited, and the Cs-transport to these
pores is thus likely dominated by diffusion. In figure 6 at 168s, the indicated ooid pore is
still filled with distilled water, while the rest of the pore space already contains solute. It
takes more then 10 minutes for its Cs-concentration starts to rise. It should be noted that
in the 1 µl/s experiment, some pores contain trapped air. During the experiment, this can
be considered as an immobile phase that does not interact with the solute transport.
In the 0.25 µl/s experiment, the Cs-concentration rise is much slower and moves more
gradually towards the top of the sample. Contrary to the 1 µl/s experiment, the well-
connected macropores have not yet reached a constant Cs-saturation before the grey
values of the ooid pores start increasing. This could indicate that diffusive transport may
start to play a role in some parts of the pore space with lower advection rates (other than
the ooid pores). Advective transport is however still dominant for the solute distribution
at 0.25 µl/s.
It is possible to obtain a good idea about the general type of transport in the carbonate
based on the fast scans. The noise level in these images makes it however difficult to
obtain reliable quantitative information from the fast scans. Especially segmenting the
pore space is a challenging endeavour due to the noise in the data and the changes inside
the pores. One of the options to improve the fast micro-CT data is to apply image
filtering. For example, 4D filtering seems to be a promising method to deal with the
higher noise levels associated with fast-acquired, dynamic data [2,16]. A second method
uses high quality, pre-acquired data on the sample to augment the 4D data. For example,
in the present case the pore space can be segmented from a high-quality micro-CT scan,
which can then be used as a mask to analyze the dynamic data (figure 3). By analyzing
the different phases on the high quality data and combining this information with the fast
scans, a dynamic map of the changing pore space can be obtained. Research into the
extraction of the velocity field from these experiments is ongoing.
Given the fact that at short ranges in the pore space, the Cs-concentration can be assumed
to vary little, we applied a median filter to average the greyscale values in the pore space.
The resulting 3D distribution map of the CsCl concentration in the sample in function of
time is shown in figure 7 for the 0.25 µl/s experiment. This concentration map shows
pores where the CsCl concentration change more slowly in time compared to the
surrounding pores. These pores remain blue through time and are the more isolated vuggy
ooid porosity, in which the concentration is controlled by mainly diffusive processes. The
concentration in the surrounding pores increases more rapidly which is illustrated by the
more rapid change from blue to orange-red. These pores are well connected and are the
preferred pathways along which the solute is transported.
The possibilities of laboratory based in-situ fast dynamic imaging are illustrated by
visualizing solute transport inside a porous carbonate rock. By maximizing the X-ray
transparency in-situ flow cell; optimizing the X-ray flux from the source; choosing the
right detector with a maximum efficiency and high read out speed; optimizing dynamic
reconstruction and integrating other micro-CT data to augment the dynamic scans, it is
possible to obtain temporal resolutions of 12 seconds. Thus allowing to continuously
monitor and quantify dynamic processes pore scale processes through time. In the applied
example a 3D map of the solute concentration in function of time could be generated,
which allows to visualize advection controlled preferential flow paths and more isolated
pore bodies controlled by diffusive transport.
These results provide a direct insight in fluid transport in complex porous media and
provide vital information to predict slower processes like reactive fluid flow and provide
input and validation for pore scale modeling.
The Research Foundation – Flanders (FWO) is acknowledged for the FWO research
grants 1521815N and 3G004115. Tom Bultreys is funded by the Flemish agency for
Innovation by Science and Technology.
1. Cnudde, V., & Boone, M. N. (2013). High-resolution X-ray computed
tomography in geosciences: A review of the current technology and applications.
Earth-Science Reviews, 123(0), 1–17.
2. Bultreys, T., Boone, M. A., Boone, M. N., De Schryver, T., Masschaele, B., Van
Hoorebeke, L., & Cnudde, V. (2015). Fast laboratory-based micro-computed
tomography for pore-scale research: illustrative experiments and perspectives on
the future. Advances in Water Resources, in press.
3. Rücker M, Berg S, Armstrong RT, Georgiadis A, Ott H, Schwing A, et al. (2015)
From connected pathway flow to ganglion dynamics. Geophys Res Lett, 42,3888–
4. Armstrong RT, Ott H, Georgiadis A, Rücker M, Schwing A, Berg S. (2014)
Subsecond pore-scale displacement processes and relaxation dynamics in
multiphase flow. Water Resour Res,50,9162–76.
5. Youssef S, Deschamps H, Dautriat J, Rosenberg E, Oughanem R, Maire E, et al.
(2013) 4D Imaging of fluid flow dynamics in natural porous media with ultra-fast
X-ray microtomography. Int. Symp. Soc. Core Anal., Napa Valley, USA: Society
of Core Analysts.
6. Andrew, M. G., Bijeljic, B., Blunt, M. J. (2013) Pore-scale imaging of geological
carbon dioxide storage under in situ conditions. Geophysical Research Letters, 40,
7. Andrew M, Menke H, Blunt MJ, Bijeljic B. (2015) The Imaging of Dynamic
Multiphase Fluid Flow Using Synchrotron-Based X-ray Microtomography at
Reservoir Conditions. Transp Porous Media, 110, 1–24.
8. Iglauer S, Paluszny A, Pentland CH, Blunt MJ. (2011) Residual CO2 imaged with
X-ray micro-tomography. Geophys Res Lett, 38.
9. Fusseis F, Steeb H, Xiao X, Zhu W, Butler IB, Elphick S, et al. A low-cost Xray-
transparent experimental cell for synchrotron-based X-ray microtomography
studies under geological reservoir conditions. J Synchrotron Radiat 2014;21:251–
10. Boone MA, Nielsen P, De Kock T, Boone MN, Quaghebeur M, Cnudde V.
Monitoring of stainless-steel slag carbonation using X-ray computed
microtomography. Environ Sci Technol 2014;48:674–80.
11. Andrew M, Bijeljic B, Blunt MJ. Pore-scale imaging of trapped supercritical
carbon dioxide in sandstones and carbonates. Int J Greenh Gas Control
12. Ott H, de Kloe K, van Bakel M, Vos F, van Pelt A, Legerstee P, et al. Core-flood
experiment for transport of reactive fluids in rocks. Rev Sci Instrum
13. Herring AL, Harper EJ, Andersson L, Sheppard A, Bay BK, Wildenschild D.
(2013) Effect of fluid topology on residual nonwetting phase trapping:
Implications for geologic CO2 sequestration. Adv Water Resour, 62, 47–58.
14. Myers GR, Kingston AM, Varslot TK, Turner ML, Sheppard AP. (2011)
Dynamic X-ray micro-tomography for real time imaging of drainage and
imbibition processes at the pore scale. Int. Symp. Soc. Core Anal., Society of
Core Analysts.
15. Maire E, Withers PJ. Quantitative X-ray tomography. Int Mater Rev 2014;59:1–
16. Bultreys, T., Boone, M., Boone, M., De Schryver, T., Masschaele, B., Van Loo,
D., Van Hoorebeke, L., et al. (2015). Real-time visualization of Haines jumps in
sandstone with laboratory-based microcomputed tomography. Water Resources
Research, 51(10), 8668–8676.
17. Roels S, Elsene J, Carmeliet J, Hens H. Characterisation of pore structure by
combining mercury porosimetry and micrography. Mater Struct 2001;34:76–82.
18. Berg, S., Ott, H., Klapp, S. A., Schwing, A., Neiteler, R., Brussee, N.,
Stampanoni, M. (2013). Real-time 3D imaging of Haines jumps in porous media
flow. Proceedings of the National Academy of Sciences, 110(10), 3755–3759.
19. Dierick M, Van Loo D, Masschaele B, Van den Bulcke J, Van Acker J, Cnudde
V, et al. Recent micro-CT scanner developments at UGCT. Nucl Instrum Methods
Phys Res Sect B Beam Interact Mater Atoms 2014;324:35–40.
20. Derluyn, H. (2012) Salt transport and crystallization in porous limestone: neutron
– X-ray imaging and poromechanical modeling, Dissertation ETH No. 20673,
ETH Zurich, Switzerland.
Figure 1. Scanner setup with arrows illustrating the magnification and rotational movement of source and
detector (left). Detail of the flow cell (right).
Figure 2. (adapted from [15]) The temporal and spatial resolutions attained at different synchrotron beamlines and lab set-ups.
Open symbols denote synchrotron sources, while filled ones represent laboratory sources, squares denote polychromatic
(“white”) beam and circles denote monochromatic scanners. The reported time is the time needed to gather 1000 projections
Figure 3. Vertical slice through a high quality scan of the Savonnières carbonate sample showing the inter- and
intragranular porosity in the sample. Scan time 30 minutes and resolution 7.3 µm.
Figure 4. Breakthrough curve of the salt solution determined by conductivity measurement in the flow cell
Figure 5. Raw vertical slices through selected reconstructed volume of the continuous acquisition.
Experiment at a constant flow speed of 0.25 µl/s. Dotted circle indicating more isolated, diffusion
controlled vuggy porosity.
Figure 6. Raw vertical slices through selected reconstructed volume of the continuous acquisition.
Experiment at a constant flow speed of 1 µl/s. Dotted circle indicating more isolated, diffusion controlled
vuggy porosity.
Figure 7. 3D rendering of the CsCl concentration map at different points in time. The concentration in the
majority of the pores increases rapidly (rapid change from blue to red), indicating that these pores are well
connected and are a part of the preferential flow path of the solute. Other pores remain blue and these are
isolated pores controlled by diffusion. Image at 0 seconds showing a rendering of the rock sample without
... In addition, 4D X-ray CT extends X-ray imaging to assess and visualize dynamic processes, such as multiple phase flow and solute transport in pores structures, with sufficient spatial and temporal resolutions, especially on the scale of milli-to microseconds [200][201][202][203][204]. The result of 4D imaging is a series of uninterrupted 3D images of the internal structure of the material during a dynamic process of interest as a function of time [204,205]. ...
... Moreover, the sample holder should be as transparent as possible to X-rays. A gantry-based 4D µCT system reaching 5 µm spatial-resolutions and 12 s temporal resolutions was jointly developed by X-ray Engineering (XRE) and Ghent University's Centre for X-ray Tomography (UGCT) [200,201,204]. They investigated several complicated fluid flow and transport processes in complex porous media. ...
Full-text available
Understanding transport phenomena and governing mechanisms of different physical and chemical processes in porous media has been a critical research area for decades. Correlating fluid flow behaviour at the micro-scale with macro-scale parameters, such as relative permeability and capillary pressure, is key to understanding the processes governing subsurface systems, and this in turn allows us to improve the accuracy of modelling and simulations of transport phenomena at a large scale. Over the last two decades, there have been significant developments in our understanding of pore-scale processes and modelling of complex underground systems. Microfluidic devices (micromodels) and imaging techniques, as facilitators to link experimental observations to simulation, have greatly contributed to these achievements. Although several reviews exist covering separately advances in one of these two areas, we present here a detailed review integrating recent advances and applications in both micromodels and imaging techniques. This includes a comprehensive analysis of critical aspects of fabrication techniques of micromodels, and the most recent advances such as embedding fibre optic sensors in micromodels for research applications. To complete the analysis of visualization techniques, we have thoroughly reviewed the most applicable imaging techniques in the area of geoscience and geo-energy. Moreover, the integration of microfluidic devices and imaging techniques was highlighted as appropriate. In this review, we focus particularly on four prominent yet very wide application areas, namely “fluid flow in porous media”, “flow in heterogeneous rocks and fractures”, “reactive transport, solute and colloid transport”, and finally “porous media characterization”. In summary, this review provides an in-depth analysis of micromodels and imaging techniques that can help to guide future research in the in-situ visualization of fluid flow in porous media.
... The high-resolution time-lapse 3D images not only successfully image the progressive degradation of Portland cement and the precipitation/dissolution of CaCO 3 in leakage paths present in the cement when exposed to supercritical CO 2 , but also monitor the progress of reaction fronts in Portland cement, as well as density changes and sample deformation. Besides, in-situ micro-CT imaging, which is related to dynamic or 4D (3D + time) imaging, makes it possible to visualize fluid distribution in the pore space and how the pore structure of the rock sample changes throughout the experiment (Boone et al., 2016). In summary, the information obtained by real-time CT imaging enables construction of a dynamic map of the changing pore space given very short time intervals, and certain short-lived pore space change phenomena can be observed. ...
Full-text available
Computed tomography (CT) is a useful sample characterization and analysis technique to better understand complicated reactive transport processes in geologic CO2 utilization and storage (GCUS) conditions. According to previous studies, we have identified four major challenges that hinder the application of CT scanning in GCUS-related sample characterization: (1) lack of registration, segmentation, noise/artifact-reducing and model selection algorithms; (2) great uncertainty in mineral composition characterization; (3) low resolution to characterize caprock with nanopores, and (4) limited real-time CT imaging capacity. To tackle these challenges, future R&D directions regarding CT applications in GCUS research are proposed.
... Currently, micro-CT is likely the only technique that can be used for in situ imaging experiments of rock samples at these time resolutions while achieving spatial resolutions in the micrometer range. Bultreys et al. (2016) and Boone et al. (2016) suggested that fast (12-15 s time resolution) laboratory-based micro-CT can be used to image tracer dispersion experiments. Where typical laboratory-based micro-CT scans are taken in about 15 min to 24 h (Wildenschild & Sheppard, 2013), a time resolution up to 15 s was reached in these studies. ...
Full-text available
Solute transport is important in a variety of applications regarding flow in porous media, such as contaminant groundwater remediation. Most recent experimental studies on this process focus on field‐scale or centimeter‐scale data. However, solute spreading and mixing are strongly influenced by pore‐scale heterogeneity. To study this, we developed a novel methodology to quantify transient solute concentration fields at the pore scale using fast laboratory‐based microcomputed tomography. Tracer injection experiments in samples with different degrees of pore‐scale heterogeneity (porous sintered glass and Bentheimer sandstone) were imaged in 3D by continuous scanning at a time resolution of 15 s and a spatial resolution of 13.4 μm. While our calibration experiments indicated a high uncertainty (1σ) on the concentration in single voxels due to imaging noise (± 27% of the total concentration range), we show that coarse gridding these values per individual pore significantly lowers the uncertainty (± 1.2%). The resulting pore‐based tracer concentrations were used to characterize the transport by calculating the solute's arrival time and transient (filling) time in each pore. The average velocities estimated from the arrival times correspond well to the interstitial velocities calculated from the flow rate. This suggests that the temporal resolution of the experiment was sufficient. Finally, the pore‐based transient filling times, the global concentration moment and the global scalar dissipation rate calculated from our experiments, indicated more dispersion in the sandstone sample than in the more homogeneous sintered glass. The developed method can thus provide more insight in the influence of pore‐scale heterogeneity on solute transport.
... PMMA was chosen considering the mechanical strength needed for the cyclic action of the flow cell and because of its relatively low X-ray attenuation. The low X-ray attenuation of the flow cell is necessary to make sure that sufficient X-ray flux reaches the detector [42]. ...
Full-text available
Recent advances in high resolution X-ray tomography (μCT) technology have enabled in-situ dynamic μCT imaging (4D-μCT) of time-dependent processes inside 3D structures, non-destructively and non-invasively. This paper illustrates the application of 4D-μCT for visualizing the removal of fatty liquids from kitchen sponges made of polyurethane after rinsing (absorption), squeezing (desorption) and cleaning (adding detergents). For the first time, time-dependent imaging of this type of system was established with sufficiently large contrast gradient between water (with/without detergent) and olive oil (model fat) by the application of suitable fat-sensitive X-ray contrast agents. Thus, contrasted olive oil filled sponges were rinsed and squeezed in a unique laboratory loading device with a fluid flow channel designed to fit inside a rotating gantry-based X-ray μCT system. Results suggest the use of brominated vegetable oil as a preferred contrast agent over magnetite powder for enhancing the attenuation coefficient of olive oil in a multi fluid filled kitchen sponge. The contrast agent (brominated vegetable oil) and olive oil were mixed and subsequently added on to the sponge. There was no disintegration seen in the mixture of contrast agent and olive oil during the cleaning process by detergents. The application of contrast agents also helped in accurately tracking the movement and volume changes of soils in compressed open cell structures. With the in house-built cleaning device, it was quantified that almost 99% of cleaning was possible for contrasted olive oil (brominated vegetable oil with olive oil) dispersed in the sponge. This novel approach allowed for realistic mimicking of the cleaning process and provided closer evaluation of the effectiveness of cleaning by detergents to minimize bacterial growth.
Full-text available
Solute transport processes are influenced by pore-scale heterogeneity. To study this, transient micron-scale solute concentration fields were imaged by fast laboratory-based X-ray micro-computed tomography. We performed tracer injection experiments in three types of porous material with increasing levels of heterogeneity (sintered glass, Bentheimer sandstone and Savonnières limestone). Different Peclet numbers were used during the experiments. For each sample and Peclet number, datasets of 40 to 74 3D images were acquired by continuous scanning with a voxel size of 13.4 to 14.6 µm and a temporal resolution of 15 to 12 seconds. To determine the measurement uncertainty on the obtained concentration fields, we performed calibration experiments under similar circumstances (temporal resolution of 12 seconds and voxel size of 13.0 µm). Here, we provide a systematic description of the data acquisition and processing and make all data, a total of 464 tomograms, publicly available. The combined dataset offers new opportunities to study the influence of pore-scale heterogeneity on solute transport, and to validate pore-scale simulations of this process in increasingly complex samples.
Full-text available
In this work, we present a novel laboratory-based micro-computed tomography (micro-CT) experiment designed to investigate the pore scale drainage behavior of natural sandstone under dynamic conditions. The fluid distribution in a Bentheimer sandstone was visualized every 4 seconds with a 12 second measurement time, allowing the investigation of single- and few-pore filling events. To our knowledge, this is the first time that such measurements were performed outside of synchrotron facilities, illustrating the growing application potential of lab-based micro-CT with sub-minute temporal resolutions for geological research at the pore scale. To illustrate how the workflow can lead to an improved understanding of drainage behavior, the experiment was analyzed using a decomposition of the pore space into individual geometrical pores. Preliminary results from this analysis suggest that the distribution of drainage event sizes follows a power law scaling, as expected from percolation theory.
Full-text available
Fast synchrotron-based X-ray microtomography was used to image the injection of super-critical CO2 under subsurface conditions into a brine-saturated carbonate sample at the pore-scale with a voxel size of 3.64μm and a temporal resolution of 45 s. Capillary pressure was measured from the images by finding the curvature of terminal menisci of both connected and disconnected CO2 clusters. We provide an analysis of three individual dynamic drainage events at elevated temperatures and pressures on the tens of seconds timescale, showing non-local interface recession due to capillary pressure change, and both local and distal (non-local) snap-off. The measured capillary pressure change is not sufficient to explain snap-off in this system, as the disconnected CO2 has a much lower capillary pressure than the connected CO2 both before and after the event. Disconnected regions instead preserve extremely low dynamic capillary pressures generated during the event. Snap-off due to these dynamic effects is not only controlled by the pore topography and throat radius, but also by the local fluid arrangement. Whereas disconnected fluid configurations produced by local snap-off were rapidly reconnected with the connected CO2 region, distal snap-off produced much more long-lasting fluid configurations, showing that dynamic forces can have a persistent impact on the pattern and sequence of drainage events.
Full-text available
During imbibition, initially connected oil is displaced until it is trapped as immobile clusters. While initial and final states have been well described before, here we image the dynamic transient process in a sandstone rock using fast synchrotron-based X-ray computed microtomography. Wetting film swelling and subsequent snap off, at unusually high saturation, decreases nonwetting phase connectivity, which leads to nonwetting phase fragmentation into mobile ganglia, i.e., ganglion dynamics regime. We find that in addition to pressure-driven connected pathway flow, mass transfer in the oil phase also occurs by a sequence of correlated breakup and coalescence processes. For example, meniscus oscillations caused by snap-off events trigger coalescence of adjacent clusters. The ganglion dynamics occurs at the length scale of oil clusters and thus represents an intermediate flow regime between pore and Darcy scale that is so far dismissed in most upscaling attempts.
Full-text available
With recent advances at X-ray micro-computed tomography (μCT) synchrotron beam lines, it is now possible to study pore-scale flow in porous rock under dynamic flow conditions. The collection of 4 dimensional data allows for the direct 3D visualization of fluid-fluid displacement in porous rock as a function of time. However, even state-of-the-art fast-μCT scans require between one and a few seconds to complete and the much faster fluid movement occurring during that time interval is manifested as imaging artifacts in the reconstructed 3D volume. We present an approach to analyze the 2D radiograph data collected during fast-μCT to study the pore-scale displacement dynamics on the time scale of 40 milliseconds which is near the intrinsic time scale of individual Haines jumps. We present a methodology to identify the time intervals at which pore scale displacement events in the observed field of view occur and hence, how reconstruction intervals can be chosen to avoid fluid-movement induced reconstruction artifacts. We further quantify the size, order, frequency, and location of fluid-fluid displacement at the millisecond time scale. We observe that after a displacement event, the pore scale fluid distribution relaxes to (quasi-) equilibrium in cascades of pore-scale fluid re-arrangements with an average relaxation time for the whole cascade between 0.5 and 2.0 seconds. These findings help to identify the flow regimes and intrinsic time and length scales relevant to fractional flow. While the focus of the work is in the context of multiphase flow, the approach could be applied to many different μCT applications where morphological changes occur at a time scale less than that required for collecting a μCT scan.
Full-text available
The pore-scale imaging of fluid distributions during two-phase fluid displacements, such as drainage and imbibition, is of inestimable value for the validation of both experiment and modelling. Previous studies by several groups have shown that fluid distributions can be directly imaged at the pore scale with X-ray micro-CT. However, this work suffers the fundamental constraint that the sample must remain unchanging while each complete set of projections is acquired, a period from several minutes to several hours, depending on the instrument and the desired image quality. Any movement of the fluid-fluid interface that occurs during the data acquisition results in inconsistent projection data and degraded images. To achieve a static system during a fluid displacement, one must halt the flood then wait for any transient movement to dissipate before acquiring image data and finally re-starting the experiment. This procedure significantly devalues the experiment since (a) one does not know how much interface relaxation has occurred during the waiting period, (b) one cannot study the effect of displacement rate with any confidence and (c) dynamic capillary effects such as contact angle hysteresis are likely to be lost.
Full-text available
X-ray computer tomography (CT) is fast becoming an accepted tool within the materials science community for the acquisition of 3D images. Here the authors review the current state of the art as CT transforms from a qualitative diagnostic tool to a quantitative one. Our review considers first the image acquisition process, including the use of iterative reconstruction strategies suited to specific segmentation tasks and emerging methods that provide more insight (e.g. fast and high resolution imaging, crystallite (grain) imaging) than conventional attenuation based tomography. Methods and shortcomings of CT are examined for the quantification of 3D volumetric data to extract key topological parameters such as phase fractions, phase contiguity, and damage levels as well as density variations. As a non-destructive technique, CT is an ideal means of following structural development over time via time lapse sequences of 3D images (sometimes called 3D movies or 4D imaging). This includes information needed to optimise manufacturing processes, for example sintering or solidification, or to highlight the proclivity of specific degradation processes under service conditions, such as intergranular corrosion or fatigue crack growth. Besides the repeated application of static 3D image quantification to track such changes, digital volume correlation (DVC) and particle tracking (PT) methods are enabling the mapping of deformation in 3D over time. Finally the use of CT images is considered as the starting point for numerical modelling based on realistic microstructures, for example to predict flow through porous materials, the crystalline deformation of polycrystalline aggregates or the mechanical properties of composite materials. © 2014 Institute of Materials, Minerals and Mining and ASM International.
Full-text available
Geological carbon dioxide storage must be designed such that the CO2 cannot escape from the rock formation into which it is injected, and often simple stratigraphic trapping is insufficient. CO2 can be trapped in the pore space as droplets surrounded by water through capillary trapping. X-ray microtomography was used to image, at a resolution of 6.6 μm, the pore-scale arrangement of these droplets in three carbonates and two sandstones. The pressures and temperatures in the pore space were representative of typical storage formations, while chemical equilibrium was maintained between the CO2, brine and rock phases to replicate conditions far away from the injection site. In each sample substantial amounts of CO2 were trapped, with the efficiency of trapping being insensitive to pore-morphology and chemistry. Apart from in one extremely well connected sample, the size distribution of residual ganglia larger than 105 voxel3 obey power law distributions with exponents broadly consistent with percolation theory over two orders of magnitude. This work shows that residual trapping can be used to locally immobilise CO2 in a wide range of rock types.
Carbon capture and storage (CCS), where CO2 is injected into geological formations, has been identified as an important way to reduce CO2 emissions to the atmosphere. While there are several aquifers worldwide into which CO2 has been injected, there is still uncertainty in terms of the long-term fate of CO2. Simulation studies have proposed capillary trapping - where the CO2 is stranded as pore-space droplets surrounded by water - as a rapid way to secure safe storage. However, there has been no direct evidence of pore-scale trapping. We imaged trapped super-critical CO2 clusters in a sandstone at elevated temperatures and pressures, representative of storage conditions using computed micro-tomography (μ-CT) and measured the distribution of trapped cluster size. The clusters occupy 25% of the pore space. This work suggests that locally capillary trapping is an effective, safe storage mechanism in quartz-rich sandstones.
Over the past decade, the wide-spread implementation of laboratory-based X-ray micro-computed tomography (micro-CT) scanners has revolutionized both the experimental and numerical research on pore-scale transport in geological materials. The availability of these scanners has opened up the possibility to image a rock's pore space in 3D almost routinely to many researchers. While challenges do persist in this field, we treat the next frontier in laboratory-based micro-CT scanning: in-situ, time-resolved imaging of dynamic processes. Extremely fast (even sub-second) micro-CT imaging has become possible at synchrotron facilities over the last few years, however, the restricted accessibility of synchrotrons limits the amount of experiments which can be performed. The much smaller X-ray flux in laboratory-based systems bounds the time resolution which can be attained at these facilities. Nevertheless, progress is being made to improve the quality of measurements performed on the sub-minute time scale. We illustrate this by presenting cutting-edge pore scale experiments visualizing two-phase flow and solute transport in real-time with a lab-based environmental micro-CT set-up. To outline the current state of this young field and its relevance to pore-scale transport research, we critically examine its current bottlenecks and their possible solutions, both on the hardware and the software level. Further developments in laboratory-based, time-resolved imaging could prove greatly beneficial to our understanding of transport behavior in geological materials and to the improvement of pore-scale modeling by providing valuable validation.
While geological carbon dioxide (CO2) storage could contribute to reducing global emissions, it must be designed such that the CO2 cannot escape from the porous rock into which it is injected. An important mechanism to immobilize the CO2, preventing escape, is capillary trapping, where CO2 is stranded as disconnected pore-scale droplets (ganglia) in the rock, surrounded by water. We used X-Ray microtomography to image, at a resolution of 6.4 µm, the pore-scale arrangement and distribution of trapped CO2 clusters in a limestone. We applied high pressures and temperatures typical of a storage formation, while maintaining chemical equilibrium between the CO2, brine, and rock. Substantial amounts of CO2 were trapped, with an average saturation of 0.18. The cluster sizes obeyed a power law distribution, with an exponent of approximately -2.1, consistent with predictions from percolation theory. This work confirms that residual trapping could aid storage security in carbonate aquifers.