AllAboutShale.com

 

Flow visualization of CO2 in tight shale formations at reservoir conditions.

M. A. Fernø1, L. P. Hauge1, A. Uno Rognmo1, J. Gauteplass1, and  A. Graue1

 

1Department of Physics and Technology, University of Bergen, Bergen, Norway

Abstract

 

 

The flow of CO2  in porous  media  is fundamental   to  many  engineering   applications  and geophysical processes. Yet detailed CO2 flow visualization remains challenging. We address this problem via positron emission tomography using11C nuclides and apply it to tight formations—a difficult but relevant rock type to investigate. The results represent  an important  technical advancement for visualization and quantification  of flow properties  in ultratight rocks and allowed us to observe that local rock structure in a layered, reservoir shale (K = 0.74 μdarcy) sample dictated  the CO2 flow path by the presence  of high-density layers. Diffusive transport  of CO2 in a fractured sample (high-permeable sandstone)  was also visualized, and an effective diffusion coefficient (Di=2.2•10-8 m2/s) was derived directly from the dynamic distribution of CO2. During CO2 injection tests for oil recovery from a reservoir shale sample we observed a recovery factor of RF=55% of oil in place without fracturing the sample.

1. Introduction

 

Easily accessible energy sources are a prerequisite for a sustainable future for human kind. Synergy between the need  for increased  energy production  and the needed reduction  in anthropogenic CO2 emissions has been  suggested through  CO2 storage  in mature  oil fields, with associated  incremental  oil recovery [see, e.g., Falcone and  Harrison, 2013]. This approach  has been  termed  carbon  capture  utilization and  storage (CCUS) where CO2 emissions from energy production are captured and injected into the subsurface to reduce the carbon footprint associated with fossil fuels in a transitional phase to a more sustainable energy outlook [Chu and Majumdar, 2012]. Specifically for CCUS, research on transport and trapping mechanisms in storage sites is needed to minimize costs and ensure safe long-term CO2 storage.

1.1. Oil Recovery and  Diffusivity in Shales

 

 

Shale formations are considered impermeable layers that restrict upward migration of hydrocarbons  and CO2 in sedimentary formations in the  subsurface  [Eiken et al., 2011]. Recently, shale has also become  a target  for hydrocarbon  exploration  and  is rapidly becoming a major  energy  resource  worldwide but especially true in the U.S. Economic hydrocarbon  production  from such reserves was until recently unfeasible mainly related to the very low to ultralow rock permeability, a parameter that  determines the connectivity and flow between pores where hydrocarbons  are stored. Harvesting the energy stored in a shale formation today relies on creating conduits  for flow through  high-pressure injection of water to hydraulically fracture the near-well regions. Although hitherto a major economic success, using data from 65,000 shale wells in 30 shale gas and 21 tight oil fields in the U.S., Hughes argued  that the shale revolution will be hard to sustain because  well production  rates decline rapidly within a few years [Hughes, 2013]. Indeed, production  generally falls as the square root of time, indicative of diffusive drive [Patzek et al., 2013]. Molecular diffusion is the mixing of fluids due to random  motion of molecules and can be expressed  by the following equation derived from Fick’s second law of diffusion in bulk fluids:

where Ci is the concentration of phase i, C0  is the surface concentration, t is time, x is distance, and Di is the molecular diffusion coeffici

where Ci is the concentration of phase i, C0  is the surface concentration,  t is time, x is distance, and Di is the molecular diffusion coefficient. Diffusion lengths  are determined by tortuosity and are generally longer in porous media compared  with bulk systems. Effective diffusion coefficients based  on Fickian diffusion may not  apply in ultratight  formations  [Webb and  Pruess, 2003], although  the  error introduced  by using  an incorrect  diffusion model  decreases   at  elevated  pressures.  We reserve  a  full investigation  of diffusive models in shale samples for future work and use here a fractured sandstone core rather than a shale sample as we here wish to emphasize  the use of local CO2 tracking in the determination of Di.

 

 

1.2. CO2 Injection for Oil Recovery in Tight Shales

 

 

Current production  behavior from fractured, tight gas reserves suggests  a diffusive drive and similar behavior is expected in tight oil formations during CO2 injection. Although steeply declining production rates and low overall recoveries are observed in shale formations—largely a result from challenging microscopic characteristics such as pore sizes (in the  nanometer range), pore connectivity (permeability in microdarcy to nanodarcy  range), and surface properties  of the  rock (to a large degree  unknown)—the  number  of scientific investigations into the underlying mechanisms is still low. Other possible fracking fluids exist, but water is cheap and (still) readily available, so a switch is unlikely before  the  increased  costs of other  fluids are justified or policies are changed. Advantages using CO2 as a fracking fluid were recently discussed by Middleton et al. [2015], in which large volumes of CO2 could be used for energy production from shale, combined with a significant reduction of water usage for fracturing and large-scale storage  of CO2. Specifically, improved CO2 technology  must be developed through research on transport  and trapping  mechanisms  in storage  sites to minimize costs and ensure safe long-term CO2 storage. Indeed, improved knowledge about flow in unconventional rocks also provides the necessary basis to improve current production rates. In this context, access to detailed flow information is vital. Reactivity between dry supercritical CO2 and the shale is generally low but may potentially extract organic matter [Busch et al., 2008] and may be a beneficial, combined  effect during CO2 injection for fracking as suggested by others  [see, e.g., Middleton et al., 2015] for CCUS. The oil recovery in oil-bearing U.S. shale reservoirs like the Bakken or Eagleford formation is believed to be less than 10%, and the potential for enhanced oil recovery (EOR) is therefore huge. We present  the first CCUS experimental results of explicit CO2 flow visualization in porous media using positron emission  tomography (PET)  and  report  high  oil recoveries  during  CO2 injection  using samples  from an oil-producing unit in the U.S. We also use the CO2 tracking data to gain insight to local flows in a layered shale sample and to calculate a diffusion coefficient directly from visualization data in a fractured sandstone core to demonstrate the use of a new imaging tool for explicit CO2 flow tracking in unconventional and fractured formations.

2. Materials and  Methods

 

 

2.1. Positron Emission Tomography (PET)

 

 

Although primarily used as a clinical diagnostic tool, PET has previously been used to visualize fluids in porous structures [see, e.g., Boutchko et al., 2012; Kulenkampff et al., 2008]. PET is based on positron-emitting radionuclides where a positron is emitted from the nucleus accompanied by an electron to balance atomic charge. The positron loses kinetic energy by interactions  with the surroundings,  and at near-zero momentum the positron  combines  with an electron  and annihilates. The physics of nucleus  decay and annihilation limits the spatial resolution of PET, and the achieved resolution depends on the distance to the detectors. A detector array registers the electromagnetic radiation in the form of two 511 keV photons  emitted  in opposite directions  to conserve  momentum. For practical purposes,  the  beta  decay  is insensitive to temperature and pressure [Emery, 1972], which, combined  with high photon  energy, makes making PET particularly suitable for visualization of flow in porous rocks because the photons penetrate the aluminum confinement vessel holding the rock sample at elevated  pressures.

 

Throughout this article, we will also use the phrase explicit imaging when discussing PET imaging. We use the term explicit imaging to emphasize  that PET provides a direct measurement of the labeled fluid saturation, which is CO2 in this work. In contrast, attenuation methods,  such as X-ray and the more common computed tomography (CT), measure fluid saturation indirectly, through the gradual loss in X-ray flux intensity through the medium  that produces  a time-averaged density distribution  image of the rock, if fluids with sufficient density difference are used. Comparison and  use of PET and  CT for flow visualization in porous  rocks is detailed elsewhere [Fernø et al., 2015].

 

 

2.2. Experimental Setup  for CO2 Injection and Explicit CO2 Tracking

 

 

Cylindrical core plugs were installed in an aluminum biaxial core holder (CoreLab Hassler Core Holder) with a rubber  sleeve  to  apply  a radial confinement pressure  to ensure  that the  injected  fluid was  transported through  the  pore  space.  The core holder  with  the  rock samples  was placed  in  the  center  of the  PET/CT (Siemens   Biograph   Truepoint   PET-CT) bore   (diameter 700 mm).  A   CT image (voxel   size   0.156 mm3: 0.51 × 0.51 × 0.6 mm3) was obtained to ensure that the rock sample was positioned correctly in the PET detector array and adjusted  if needed. Unlike normal diagnostic operations, the rock system was stationary positioned within detector  array, with an axial field of view of 169 mm. This allowed for dynamic scans with extended PET recording  times  (up to 17 h continuous  scanning  was successfully  tested)  with  a spatial  voxel size of 8.49 mm3  (2.04 × 2.04 × 2.04 mm3). Signals were continuously recorded, and temporal resolution was set during postprocessing and determined based  on a balance between image quality, expressed  as signal-to-noise ratio (SNR), and temporal resolution: the higher temporal resolution (shorter time between each image), the lower SNR. An excellent SNR of 200:1 was achieved using temporal  resolutions of 10–30 s.

Visualization of rock characteristics through  CT imaging (grey scale), coupled with explicit CO2 signal through  PET imaging in Core A.

Figure 1. Oil recovery by CO2 injection in ultratight  unconventional stacked  core system. Graph: Average oil recovery versus time (pore volumes injected) resulting in final oil recovery of 55% OOIP during 3.7 PV CO2 injected using Cores A, B, and C. Inset: Visualization of rock characteristics through  CT imaging (grey scale), coupled with explicit CO2 signal through  PET imaging in Core A. Aligning a threshold CT image (i) and CO2 PET image (ii), obtained  after 1 h injection, we observe that the emerged CO2 flow pattern  correlated to local rock structure and layered high-low density bands. The injected CO2 flowed in the lower density regions of the core sample, indicative of a layered permeability system, leading to viscous fingers and a highly irregular displacement front.

Positron-emitting  radionuclides were produced  using particle accelerators on site due to the relatively short half-life (approximately  20 min). Reduction in signal intensity by radioactive  decay during  flow tests was correctly compensated for using algorithms imbedded in the  standard  PET/CT software provided  by the manufacturer.  The use  of 11C as  a  radionuclide  tag  for methane (CH4) has  previously  been  proposed [Maucec, 2013] but experimentally verified in this work, for the first time, to characterize CO2 flow in porous systems. The 11CO2  phase was produced  in a cyclotron by bombarding the target media (N2+1% O2) with 16.5 MeV protons. A batch of 78 ml 11CO2  (and traces of nitrogen) was mixed with CO2 in a 1dm3  injection pump (ST Stigma 1000) and pressurized to experimental conditions. Each injection test started approximately one half-life after initial 11C delivery. Injected radioactive CO2 was collected at the outlet  in a production pump set to maintain a constant  pressure.

 

 

3. Experimental Results and  Discussion

 

 

3.1. Description of CO2  Flow and  CO2  EOR in Tight Shale

 

 

With nanodarcy  level permeability,  properties   like effective  diffusion  coefficients, CO2  capillary entry pressure, and CO2 flow description in the shale are generally very difficult to measure accurately in the laboratory [Liu et al., 2012]. In this context, alternative  approaches to measure  these  properties  are useful, and we report  here  the  first experimental  demonstration of CO2 tracking for flow characterization  in shale using PET/CT imaging. We also evaluate  the oil recovery by CO2 injection (see Figure 1), without fracking, in  ultratight,  unconventional  shale  core  plugs  using  three  stacked 1.5 in diameter cores (Core A: K = 0.74 μdarcy, L = 3.92 cm;   Core B:  K = 1.7 μdarcy, L= 3.80 cm; Core C: K = 0.12 μdarcy,  L = 2.45 cm). Injection conditions (ΔP=7.09MPa; Pinlet= 22.1MPa and Poutlet= 15.0 MPa; T = 60°C) were above minimum miscibility pressure (MMP) between CO2 and crude oil (American Petroleum Institute gravity 38). The initial oil saturation was SO = 0.80. Oil recovery was determined from volumetric measurements downstream of a back pressure regulator  (Equilibar HC276-5) at ambient  conditions. The injected CO2 was exposed  to the inlet end face for 5 days before the injection rate gradually increased for the subsequent 3 days, with an average  rate of 6°•10-3 cm3/min.  Injection conditions were not changed  during the entire test. Final oil recovery factor was RF=55.0 ± 9.2% Original oil in place (OOIP), and oil was still produced (albeit at a very low rate) when the test was terminated.

 

Coupled fluid-rock interactions during CO2 injection (Ppore= 10MPa, T ambient; injection rate 0.5 cm3/min) in Core A were studied in detail through aligned CO2 flow PET data and rock structures CT data (see Figure 1, inset). The imaging results demonstrated that (1) the layered nature of the sample dictated the preferred flow pattern of the injected CO2 and (2) there is a potential for CO2 to displace oil without fracturing the tight rock. Using dynamic explicit imaging, we observed  the development of a dispersed CO2 front and accurately pinpoint the underlying cause for this behavior. The observed shape is indicative of a combination of viscous displacement and molecular diffusion, where local high-density horizontal layers reduce transverse flux. Furthermore, with access to local CO2 flow paths, we learn that the injected CO2 does not fracture the formation when entering the pore space to produce oil. The high oil recovery reported  in the stacked system, with RF=55% OOIP, corroborate  the second point.

(left) Dynamic longitudinal CO2 profiles showing increased CO2 saturation over time. Slight intensity dips along the length correlate to support

Figure 2. Visualization of diffusive CO2 transport  and mixing in a fractured (1 mm constant  fracture aperture  held open with a spacer) oil-saturated (n-decane) Bentheim core plug. (left) Dynamic longitudinal 11CO2 profiles showing increased CO2 saturation over time. Slight intensity dips along the length correlate to support columns in spacer. (right) Symmetric, transverse diffusive CO2 transport  from the CO2 saturated  fracture (RD=0.0) into the oil-saturated matrix at location XD=0.5. Analytical profiles (dashed lines) using equation (1) were fitted to dynamic imaging data with an effective diffusion coefficient of 2.2•108m2/s.

3.2. Calculating the Diffusion Coefficient With PET

 

 

We use a fractured sandstone core rather than a shale sample as we here wish to emphasize  the use of CO2 tracking in the determination of Di and not attempt an investigation of validity of Fickian diffusion in shale. Explicit CO2 tracking was utilized in fractured, high-permeable (ϕ=0.22 and K=1.2D) Bentheim sandstone to determine an effective diffusion coefficient directly from PET CO2 tracking data (Figure 2) during miscible CO2 flow (P=8.3MPa, T = 25°C, and Q = 0.15 cm3/min). The fracture was held open with a constant aperture of 0.5mm using a spacer to assure a high conduit flow path to limit viscous forces. Transverse CO2 transport from the CO2 saturated  longitudinal fracture to the completely oil-saturated (n-decane) matrix occurred by molecular diffusion only. An effective diffusion coefficient (Di) was estimated  using equation (1), with Di as a fitting parameter. With boundary conditions Ci(0, t)=C0 for t>0 (i.e., constant SCO2 at RD=0.0) and Ci(∞, t)=0 for all t (i.e., SCO2=0) at RD=[-1, 1] and the initial condition Ci(x,0)=0 for all x, we derived an effective CO2 diffusion coefficient of 2.2•10-8 m2/s (slightly overestimated due to decreasing volume in the transverse  direction of a cylindrical core plug). The diffusion coefficient varies both  with temperature and pressure, in addition to rock type (due to variations in pore sizes and distribution, i.e., diffusion path tortuosity), and the reported coefficient agrees reasonably well to other CO2-decane diffusion coefficient ranging between 0.83 and 5.05•10-9 m2/s [Eide et al., 2015; Renner, 1988; Tenga et al., 2014; Trivedi and Babadagli, 2006], although the literature did not use the same temperature and pressure conditions and the rock type as studied in this work. The measured 11CO2 intensity profiles deviate from equation (1) over time as the boundary condition is violated, as expected, when the CO2 reach the outer end of the core.

 

 

4. Concluding Remarks

 

 

We demonstrate the potential to evaluate CO2 flow and diffusion coefficient with direct, dynamic, and explicit CO2 tracking, rather than using indirect methods,  through  scouting experiments  with combined  PET/CT imaging. In particular, access to CO2 flow in challenging tight formations represents a scientific advancement with potentially large impact. The main advantage with PET is its high sensitivity, requiring a tracer activity as low as 10-12mol/l [Kulenkampff et al., 2008], which enables accurate determination of flow, even in the ultra-tight samples used in this work. Indeed, separate CT imaging cannot provide the same high-quality imaging, especially in low porous rocks, although recent advances are promising [Vega et al., 2014]. Combined PET/CT imaging, however,  provides  complementary information  that  exceeds  the  imaging  capability from each method  separately. This approach is utilized here to study the fluid-rock interactions relevant for flow in tight formations but can be applied to a larger range of rock types and displacement processes.

 

Due to the short half-life of 11C (20 min), injection tests must be carefully designed  and planned, and 11CO2 cannot  be used  to evaluate, e.g., long-term  carbon  capture  and  storage  processes  like cap rock integrity [Iglauer et al., 2015] or geochemical  effects [Liu et al., 2012]. For these  processes,  we propose  the  use of 22Na (half-life 2.6 years and NaCl occurs in most brines), which enables long-term evaluation CO2-brine-shale interaction through  direct PET visualization. Based on the experimental  results presented herein, we report the following key observations:

  1. We show for the first time explicit CO2 flow characterization  using 11C nuclides to visualize and quantify dynamic, spatial CO2 distribution in porous media. We experimentally  demonstrate the  benefits  of a robust, decoupled imaging approach and highlight the potential of combined PET/CT imaging. In particular, access to CO2 flow paths in ultratight  rocks represents  an important  technical advancement, with potentially large impact to the scientific community on transport  in porous media.
  2. CO2 injection for oil recovery from unconventional, ultratight formations should be considered  a viable technique for the future, and we observe recovery of RF=55% OOIP within 4 pore volume (PV) injected in the laboratory. The oil is produced  without fracturing the formation and by developing miscibility with the crude oil saturating  the pore system. The substantial oil production, compared  to currently reported recovery factors, coupled with capillary trapping of CO2, provides an economical basis for CCUS in shale formations.
  3.  A link between local rock structures and CO2 flow was determined by explicit CO2 tracking in a layered, ultratight reservoir shale (K=0.74 μdarcy) sample, where the flow profile was dictated by the presence of high-density  layers. Diffusive transport  of CO2 in a fractured (high-permeable)  sandstone sample was visualized, and an effective diffusion coefficient (Di =2.2•10-8 m2/s) was calculated directly from the PET images. These imaging results, along with the demonstrated applicability in tight formations, show the  benefits  of this imaging  technique for visualization and quantification  of important flow properties.

 

 

 

Abbreviations

 

 

API       American Petroleum Institute

CCUS   Carbon capture utilization and storage

CT        Computed  tomography

EOR     Enhanced oil recovery

MMP    Minimum miscibility pressure

PET     Positron emission tomography

PV       Pore volume

 

 

Nomenclature

 

 

Ci          concentration of phase i

C0         surface concentration

darcy    darcy (unit for permeability: 1 darcy = 0.9863•10-12 m2)

Di          molecular diffusion coefficient for phase i

K          absolute  permeability

Pinlet       absolute  pressure at inlet (MPa)

Poutlet     absolute  pressure at outlet (MPa)

Ppore      pore pressure (MPa)

Q          injection rate (cm3/min)

RD         dimensionless radius

SCO2      CO2 saturation

Sg         gas saturation

So         oil saturation

Sor        residual oil saturation

Sw        water saturation

Swi        initial water saturation

t           time

x          distance

XD        dimensionless length

ϕ          porosity

 

 

Acknowledgments

 

 

The authors are indebted to the Norwegian Research Council under Climit project 200032 “In- situ imaging of CO2 flow, storage and entrapment in subsurface aquifers and hydrocarbon,” Petromaks project 200538 “Integrated Enhanced Oil Recovery in Fractured and Heterogeneous Reservoirs,” and Statoil. We also acknowledge  Geir-Espen Abell and Tom Christian Holm Adamsen at Centre for Nuclear Medicine and PET, Department  of Radiology, Haukeland University Hospital for the operation  of PET/CT scanner. The experimental  data are available upon request by contacting the corresponding author.

 

The Editor thanks Stefan Iglauer and an anonymous  reviewer for their assistance evaluating this manuscript.

 

 

Key Points:

 

  • CO2 injection in tight shale effectively produces  oil without fracturing the formation
  • Positron emission tomography successfully used to explicitly image CO2 flow in shale
  • Diffusion  coefficient derived exclusively from PET imaging in fractured media during CO2 injection

 

 

 

Correspondence to: M. A. Fernø, Martin.Ferno@uib.no

 

 

Citation:

 

Fernø, M. A., L. P. Hauge, A. Uno Rognmo, J. Gauteplass, and A. Graue (2015), Flow visualization of CO2 in tight shale formations at reservoir conditions, Geophys. Res. Lett., 42, 7414–7419, doi:10.1002/2015GL065100.

 

Received 6 JUL 2015

Accepted 25 AUG 2015

Accepted article online 29 AUG 2015

Published online 18 SEP 2015

 

©2015. The Authors. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

 

 

References

 

 

Boutchko, R., V. L. Rayz,  N. T. Vandehey, J. P. O’Neil, T. F. Budinger, P. S. Nico,  J. L. Druhan, D. A. Saloner, G. T. Gullberg, and W. W. Moses (2012), Imaging and modeling of flow in porous media using clinical nuclear emission tomography systems and computational fluid dynamics, J. Appl. Geophys., 76, 74–81, doi:10.1016/j.jappgeo.2011.10.003.

 

Busch, A., S. Alles,  Y. Gensterblum, D. Prinz, D. N. Dewhurst, M. D. Raven, H. Stanjek, and B. M. Krooss  (2008),  Carbon dioxide storage potential of shales, Int. J. Greenhouse Gas Control, 2(3), 297–308, doi:10.1016/j.ijggc.2008.03.003.

 

Chu, S., and A. Majumdar (2012), Opportunities  and challenges for a sustainable energy future, Nature, 488, 294–303.

 

Eide, Ø., M. A. Fernø, Z. Alcorn, and A. Graue (2015), Visualization of carbon dioxide enhanced oil recovery by diffusion in fractured chalk, SPE J., doi:10.2118/170920-PA.

 

Eiken, O., P. Ringrose, C. Hermanrud, B. Nazarian, T. A. Torp,  and L. Høier (2011), Lessons learned from 14 years of CCS operations: Sleipner, In Salah and Snøhvit, Energy Procedia, 4(0), 5541–5548, doi:10.1016/j.egypro.2011.02.541.

 

Emery, G. T. (1972),  Perturbation  of nuclear decay rates, Annu. Rev. Nucl. Sci., 22(1), 165–202, doi:10.1146/annurev.ns.22.120172.001121.

 

Falcone, G., and R. Harrison (2013), Deciding whether  to fund either CCS or CCUS offshore projects: Are we comparing apples and pears in the North Sea? in SPE Annual Technical Conference and Exhibition, Society of Petroleum Engineers, New Orleans, Louisiana, doi:10.2118/ 166388-ms.

 

Fernø, M. A., J. Gauteplass, L. P. Hauge, G. E. Abell,  T. C. H. Adamsen, and A. Graue (2015), Combined positron emission tomography and computed tomography to visualize and quantify fluid flow in sedimentary  rocks, Water Resour. Res., 51, doi:10.1002/2015WR017130.

 

Hughes, J. D. (2013), Energy: A reality check on the shale revolution, Nature, 494(7437), 307–308.

 

Iglauer, S., C. H. Pentland, and A. Busch (2015), CO2 wettability of seal and reservoir rocks and the implications for carbon geo-sequestration, Water Resour. Res., 51, 729–774, doi:10.1002/2014WR015553.

 

Kulenkampff, J., M. Gründig, M. Richter, and F. Enzmann (2008), Evaluation of positron-emission-tomography for visualisation of migration processes in geomaterials, Phys. Chem. Earth Parts A/B/C, 33(14–16), 937–942, doi:10.1016/j.pce.2008.05.005.

 

Liu, F., P. Lu, C. Griffith,  S. W. Hedges, Y. Soong, H. Hellevang, and C. Zhu (2012), CO2–brine–caprock interaction: Reactivity experiments  on Eau Claire shale and a review of relevant literature, Int. J. Greenhouse Gas Control, 7(0), 153–167, doi:10.1016/j.ijggc.2012.01.012. Maucec, M. (2013), Systems and methods for determining fluid mobility in rock samples.

 

Middleton, R. S., J. W. Carey,  R. P. Currier, J. D. Hyman, Q. Kang, S. Karra, J. Jiménez-Martínez, M. L. Porter, and H. S. Viswanathan (2015), Shale gas and non-aqueous fracturing fluids: Opportunities  and challenges for supercritical CO2, Appl. Energ., 147, 500–509, doi:10.1016/j.apenergy.2015.03.023.

 

Patzek, T. W., F. Male, and M. Marder (2013), Gas production  in the Barnett Shale obeys a simple scaling theory, Proc. Natl. Acad. Sci. U.S.A., 110(49), 19,731–19,736.

 

Renner, T. A. (1988),  Measurement  and correlation of diffusion coefficients for CO2 and rich-gas applications, SPE J., 3(2), 517–523.

 

Tenga, Y., Y. Liua,  Y. Songa, L. Jianga, Y. Zhaoa, X. Zhoua, H. Zhenga, and J. Chena (2014),  A study on CO2 diffusion coefficient in n-decane saturated  porous media by MRI, Energy  Procedia, 61, 603–606.

 

Trivedi, J., and T. Babadagli (2006), Efficiency of diffusion controlled miscible displacement in fractured porous media, Transport Porous Media.

 

Vega, B., A. Dutta, and A. Kovscek  (2014),  CT imaging of low-permeability, dual-porosity systems using high X-ray contrast gas, Transport Porous Media, 101(1), 81–97, doi:10.1007/s11242-013-0232-0.

 

Webb, S., and K. Pruess (2003), The use of Fick’s law for modeling trace gas diffusion in porous media, Transport Porous Media, 51(3), 327–341, doi:10.1023/A:1022379016613.

Flow visualization of CO2 in tight shale formations at reservoir conditions.

where Ci is the concentration of phase i, C0  is the surface concentration, t is time, x is distance, and Di is the molecular diffusion coeffici
Visualization of rock characteristics through  CT imaging (grey scale), coupled with explicit CO2 signal through  PET imaging in Core A.
(left) Dynamic longitudinal CO2 profiles showing increased CO2 saturation over time. Slight intensity dips along the length correlate to support

where Ci is the concentration of phase i, C0  is the surface concentration, t is time, x is distance, and Di is the molecular diffusion coeffici
Visualization of rock characteristics through  CT imaging (grey scale), coupled with explicit CO2 signal through  PET imaging in Core A.
(left) Dynamic longitudinal CO2 profiles showing increased CO2 saturation over time. Slight intensity dips along the length correlate to support

Visualization of rock characteristics through  CT imaging (grey scale), coupled with explicit CO2 signal through  PET imaging in Core A.
(left) Dynamic longitudinal CO2 profiles showing increased CO2 saturation over time. Slight intensity dips along the length correlate to support

where Ci is the concentration of phase i, C0  is the surface concentration, t is time, x is distance, and Di is the molecular diffusion coeffici
Visualization of rock characteristics through  CT imaging (grey scale), coupled with explicit CO2 signal through  PET imaging in Core A.
(left) Dynamic longitudinal CO2 profiles showing increased CO2 saturation over time. Slight intensity dips along the length correlate to support

where Ci is the concentration of phase i, C0  is the surface concentration, t is time, x is distance, and Di is the molecular diffusion coeffici
Visualization of rock characteristics through  CT imaging (grey scale), coupled with explicit CO2 signal through  PET imaging in Core A.
(left) Dynamic longitudinal CO2 profiles showing increased CO2 saturation over time. Slight intensity dips along the length correlate to support