The impact of gradational contact at the reservoir-seal interface on geological CO2 storage capacity and security Onoja, M & Shariatipour, SM Author post-print (accepted) deposited by Coventry University’s Repository Original citation & hyperlink: Onoja, M & Shariatipour, SM 2018, 'The impact of gradational contact at the reservoir-seal interface on geological CO2 storage capacity and security' International Journal of Greenhouse Gas Control, vol 72, pp. 1-13 https://dx.doi.org/10.1016/J.IJGGC.2018.03.007 DOI ISSN 10.1016/J.IJGGC.2018.03.007 1750-5836 Publisher: Elsevier NOTICE: this is the author’s version of a work that was accepted for publication in International Journal of Greenhouse Gas Control. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in International Journal of Greenhouse Gas Control, [72, (2018)] DOI: 10.1016/J.IJGGC.2018.03.007 © 2018, Elsevier. Licensed under the Creative Commons AttributionNonCommercial-NoDerivatives 4.0 International http://creativecommons.org/licenses/by-nc-nd/4.0/ Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders. This document is the author’s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it. The Impact of Gradational Contact at the Reservoir-Seal Interface on Geological CO2 Storage Capacity and Security Michael U. Onojaa*, Seyed M. Shariatipoura a Centre for Flow Measurement and Fluid Mechanics, Maudslay House, Coventry University. CV1 2NL. Coventry, United Kingdom. * Corresponding author: onojau@coventry.ac.uk 1 1 Abstract 2 The implementation of CO2 storage in sub-surface sedimentary formations can 3 involve decision making using relevant numerical modelling. These models are often 4 represented by 2D or 3D grids that show an abrupt boundary between the reservoir 5 and the seal lithologies. However, in an actual geological formation, an abrupt contact 6 does not always exist at the interface between distinct clastic lithologies such as 7 sandstone and shale. This article presents a numerical investigation of the effect of 8 sediment-size variation on CO2 transport processes in saline aquifers. Using the 9 Triassic Bunter Sandstone Formation (BSF) of the Southern North Sea (SNS), this 10 study investigates the impact a gradation change at the reservoir-seal interface on CO2 11 sequestration. This is of great interest due to the importance of enhanced geological 12 detail in reservoir models used to predict CO2 plume migration and the integrity of 13 trapping mechanisms within the storage formation. The simplified strategy was to 14 apply the Van Genutchen formulation to establish constitutive relationships for pore 15 geometric properties, which include capillary pressure (Pc) and relative permeability 16 (kr), as a function of brine saturation in the porous media. The results show that the 17 existence of sediment gradation at the reservoir-seal interface and within the reservoir 18 has an important effect on CO2 migration and pressure diffusion in the formation. The 19 modelling exercise shows that these features can lead to an increase in residual gas 20 trapping in the reservoir and localised pore pressures at the caprock’s injection point. 21 22 Keywords: 23 24 CO2 Sequestration; Capillary Pressure; Relative Permeability; Physical Trapping; Clastic Sediments. 25 26 27 2 28 1. Introduction 29 The geological storage of carbon dioxide (CO2) serves as an option to sequester CO2 30 emissions from the atmosphere. Carbon Capture and Storage (CCS) is a three-step 31 process that involves CO2 capture, its transport, and subsequent underground storage. 32 It was inspired by the utilisation of CO2 in enhanced oil or gas recovery (EOR or 33 EGR) which offers potential economic gain from the increased production of 34 hydrocarbons (Bondor, 1992; Martin and Taber, 1992; Kovscek and Cakici, 2005; 35 Grigg, 2005; Gozalpour et al., 2005). Various studies on this approach, duly 36 summarised in the Intergovernmental Panel on Climate Change (IPCC) special report 37 on carbon capture and storage, have elaborated on the feasibility of mitigating the 38 adverse effects of this greenhouse gas while enhancing the recovery of fossil fuels for 39 future energy production (IPCC, 2005). The report outlines sedimentary rocks as 40 naturally ideal media for geologic CO2 sequestration (GCS), and deep saline aquifers 41 as subsurface formations which possess the largest storage capacity. One critical issue 42 in GCS, however, is demonstrating the long-term safety and security of subsurface 43 CO2 storage. This entails assessing the potential for CO2 leakage from deep 44 formations into shallow groundwater aquifer zones (Zheng et al., 2013; Lawter et al., 45 2017) as well as any possibility of injection-induced seismicity (Nicol et al., 2011; 46 Dempsey et al., 2014). Undoubtedly, this has necessitated a broad scientific approach 47 that elucidates the geological processes influencing the estimation of CO2 storage 48 capacity and the integrity of caprocks overlying the storage aquifers (Bachu, 2015). 49 This approach usually involves the use of numerical models which incorporate 50 reservoir parameters such as porosity, permeability, and saturation functions to solve 51 the governing equations for subsurface fluid flow and transport. The models simulate 52 complex geological processes which aid in the design of injection schemes as well as 53 the assessment of storage capacities in target locations. 54 In petroleum literature, numerical models are commonly applied to the quantitative 55 analysis of heterogeneic effects in subsurface storage media (e.g. Pruess et al., 2003; 56 Doughty and Pruess, 2004; Kumar et al., 2005; Mo et al., 2005). A number of studies 57 using numerical models acknowledge the importance of two constitutive functions: 58 capillary pressure (Pc) and relative permeability (kr), on multiphase fluid flow during 59 GCS (Fleet et al., 2004; Ennis-King and Paterson, 2005; Juanes et al., 2006; Obi and 60 Blunt, 2006; Burton et al., 2009; Kopp et al., 2009). The main aim of this study is to 3 61 investigate the variability in transport and flow processes of injected CO2 resulting 62 from a gradational contact at the reservoir-seal interface and the gradual change in 63 clast-size within an aquifer. This variability is described in consitutive functions using 64 an empirical correlation that is based on grain-size variation i.e Van Genuchten’s 65 (1980) formulation. Through this contribution, we intend to encourage the 66 representation of heterogeneity in capillary pressure (Pc) and relative permeability (kr) 67 functions during reservoir simulation. To the best of our knowledge, no large-scale 68 study on GCS has incorporated such small-scale variability in both Pc– and kr– 69 saturation curves. Although Saadatpoor et al. (2010) and Meckel et al. (2015) showed 70 the influence of grain-scale heterogeneity on a reservoir-scale, their studies only 71 emphasised the influence of capillary heterogeneity on CO2 storage performance. The 72 former scaled the variability flow processes using intrinsic permeability heterogeneity 73 with a spatially constant capillary pressure curve, while the latter introduced capillary 74 heterogeneity by generating a capillary threshold pressure distribution based on 75 defined median grain size. In this study, we scale capillary heterogeneity from a 76 spatially constant threshold pressure and describe variation in relative permeability 77 curves through the grain size. The contribution of other effects such as the wettability 78 of the porous medium and the interfacial tension between the fluids in contact are not 79 considered here. 80 2. 81 Reservoir heterogeneity is dominated by depositional and diagenetic processes. The 82 sedimentology of the formation primarily influences the reservoir quality by 83 regulating its pore system (Pettijohn et al., 1972). This dictates the porosity and 84 permeability of the media and in turn controls the storage capacity and the efficiency 85 of physical trapping mechanisms during CO2 sequestration (Benson and Cole, 2008). 86 The physical and chemical changes that alter the characteristics of sediments after 87 deposition are reffered to as diagenesis. Volumetrically, siliciclastic rocks are the 88 most important variety of sedimentary rocks for GCS (Boggs, 2009). Clasts, i.e. rock 89 fragments, vary in size ranging from fine-textured clay and silt, to medium-textured 90 sand (see Table 1), up to coarse-textured pebble, cobble and boulder sized materials 91 (Wentworth, 1922). During transportation and deposition, the clasts are sorted 92 according to their average grain-size diameter and deposited in a geological sequence 93 of interleaved rocks known sedimentary beds or strata (Hiscott, 2003). Consequently, Problem statement 4 94 sedimentary structures such as gradational contacts or graded beds are formed. 95 Gradational contact describes the gradual transition in the average size of deposited 96 clasts between conformable strata while graded bedding refers to the vertical 97 evolution of grain size in a stratum. These structures are reservoir-scale 98 heterogeneities which can influence injected CO2 flow patterns due to distinct 99 hydraulic conductivities arising from grain-scale heterogeneities. Grain-scale 100 heterogeneity dictates the capillary effect that governs two physical traps: 101 stratigraphic and residual gas trapping (Bjørlykke, 2010). This capillarity effect 102 emanates from fluid and interfacial physics at the pore-scale. Hence, the effective 103 hydraulic behaviour on any practical field-scale is dominated by the large scale 104 spatial-arrangement of small-scale variability (Krevor et al., 2015). The reader is 105 referred to Pettijohn (1957) and Haldorsen (1986) for the basics of sedimentary 106 structures and scales of heterogeneity, respectively. Geological Size Range (mm) 2.0 - 1.0 Sediment Texture General term for Consolidated Rock Very coarse sand 1.0 - 0.5 Coarse sand 0.5 - 0.25 Medium sand 0.25 - 0.125 Fine sand 0.125 - 0.0625 Very fine sand 0.0625 - 0.0313 Coarse silt 0.0313 - 0.0156 Medium silt 0.0156 - 0.0078 Fine silt 0.0078 - 0.0039 Very fine silt Sandstone Siltstone Mudstone (Shale) < 0.0039 Clay Claystone Table 1: The Wentworth scale for clastic sediments (Wentworth, 1922) 107 108 2.1 The reservoir-seal interface 109 The lithostratigraphic units of many generic reservoir models are usually interpreted 110 from wireline logs of representative geologies, such as the Gamma Ray (GR) tool 111 (Doveton, 1991; Darling, 2005). However, the GR log may be considered to fall short 112 of its capabilities when distinguishing between types of mudstones, i.e. siltstone and 113 claystone. This is because the primary radioactive isotopes in rocks, i.e. potassium, 114 thorium and uranium, are more common in clay minerals than in sand and silt 115 (Bigelow, 1992), hence the GR log is often used as a measure of shale content 116 (Katahara, 1995; Fabricius et al., 2003; Nazeer et al., 2016). Since clay distribution 5 117 alone cannot account for the fine-grained sediments in an actual reservoir, it is 118 important to assess the impact of sediment-size gradation on GCS, particularly at the 119 reservoir-seal interface. Most models that simulate CO2 plume distribution are built 120 under the assumption that the stratigraphic contact between the reservoir rock and the 121 caprock is abrupt (i.e. a sudden distinctive change in the lithology). This may not 122 always be the case in geological formations because the bedding contact between 123 sandstone and mudstone can show a gradation in particle sizes at the interface. For 124 example, the Sherwood Sandstone Group shows an upward grading of sediments from 125 coarse sandstones to siltstones, and then to the Mercia Mudstone Group (Benton et 126 al., 2002; Newell 2017). Nevertheless, a number of contemporary studies performed 127 using reservoir models have included geological details such as top-surface 128 morphologies and transition zone heterogeneities (e.g. Shariatipour et al., 2014; 2016; 129 Newell and Shariatipour, 2016). These studies demonstrated that such geological 130 detail can affect various trapping mechanisms within the reservoir as well as influence 131 CO2 plume migration, the estimation of storage capacity, and the volume of the 132 aquifer. Generally, increasing the level of detail in geological modeling for simulation 133 models is essential for producing meaningful and accurate results (Van De Graaff and 134 Ealey, 1989). 135 2.2 136 Due to the scarcity of experimental data on Pc and kr, the common practice in 137 reservoir modelling is the use of empirical formulations to describe flow 138 characteristics. Many GCS studies have adopted the constitutive functions by either 139 Brooks and Corey (1966) or Van Genuchten (1980) to describe the capillary pressure 140 (Pc), saturation (S), and relative permeability (kr) relationship (Pc–S–kr relationship) in 141 the flow model (e.g. Class et al., 2009; Oldenburg et al., 2001; Cameron and 142 Durlofsky, 2012). A comprehensive review by Oostrom et al. (2016) highlights the 143 van Genuchten, VG, function to be much more efficient in describing the dynamic 144 fluid model in GCS. This is usually coupled to Mualem’s (1976) and Corey’s (1954) 145 formulations to give the integrated Van Genutchten-Mualem-Corey (VGMC) flow 146 model for Pc–S–kr relationships: 147 148 Describing flow characteristics in reservoir models 𝑆𝑤 + 𝑆𝑛𝑤 = 1 𝑆𝑒𝑤 = (1) 𝑆𝑤 − 𝑆𝑤,𝑚𝑖𝑛 𝑆𝑤,𝑚𝑎𝑥 − 𝑆𝑤,𝑚𝑖𝑛 6 (2) − 𝑃𝑐 = 𝑃𝑒 [(𝑆𝑒𝑤 ) 149 1 1 𝑚 − 1] 1 𝑚 (3) 𝑘𝑟𝑤 = (𝑆𝑒𝑤 )2 [1 − (1 − (𝑆𝑒𝑤 ) 150 𝑚 1⁄ 𝑚) ] 2 𝑘𝑟𝑛𝑤 = (1 − 𝑆𝑒𝑤 )2 (1 − (𝑆𝑒𝑤 )2 ) 151 (4) (5) 152 where Sew is the effective wetting fluid saturation; Sw,min and Sw,max represent the 153 minimum and maximum saturation for the wetting fluid which occurs for a given 154 problem at an actual wetting fluid saturation of Sw; Snw is the saturation of the non- 155 wetting fluid; Pe is the capillary entry pressure; krw and krnw are relative permeability 156 values for the wetting fluid and the non-wetting fluid respectively, at an effective 157 wetting fluid saturation, Sew; n and m correspond to pore geometry/model parameters 158 related by the assumption that 𝑚 = 1 − 1⁄𝑛. 159 A large number of numerical models that have used this flow model assumed a 160 generic parameter value of 0.457 for the pore size index, m (Oostrom et al., 2016). 161 According to Birkholzer et al. (2009) this value is typical of sedimentary formations 162 suitable for CO2 storage. A vast number of studies have used a constant value for the 163 fitting parameter, m, to generate Pc–S–kr relationships irrespective of the geological 164 heterogeneity of the model (e.g. Gor et al., 2013; Zhou et al., 2010; Al-Khdheeawi et 165 al., 2017; Espinet et al., 2013). Utilising a fixed value to represent the pore size 166 distribution index of an entire storage formation fails to account for the differences in 167 the average pore size of rock lithologies within strata in the reservoir. Additionally, 168 predictions from such reservoir models may fall short of precision because the 169 accuracy of flow processes in a porous medium is highly dependent on the description 170 of the Pc–S–kr relationship (Mori et al., 2015). 171 3. 172 In this paper, we employ Carsel and Parrish’s (1988) descriptive statistics for the pore 173 size distribution index, n, and introduce a parameterisation scheme that describes the 174 fluid flow behaviour of various clastic rocks (Table 2): Methodology General term for consolidated rock Sedimentary components (%) Sand Silt Clay 7 Van Genuchten Parameter (n) Term for Consolidated Rock as used in this study Sandstone > 85 Silt + (1.5*Clay) < 15 2.68 Coarse Sandstone Sandstone 70 – 90 Silt + (1.5*Clay) ≥ 15; and Silt + (2*Clay) < 30 2.28 Sandstone Sandstone > 52 Silt + (2*Clay) ≥ 30; if a) Clay is between 7 – 20, or b) Clay < 7, and Silt < 50 1.89 Silty Sandstone Sandstone < 52 28 - 50 7 - 27 1.56 Muddy Sandstone Sandstone > 45 < 28 20 - 35 1.48 Clayey Sandstone Mudstone 20 – 50 50 - 80 12 - 27 1.41 Sandy Siltstone Mudstone < 20 > 80 < 12 1.37 Siltstone Mudstone < 45 < 40 > 40 1.09 Claystone 175 176 Table 2: Sedimentary components and the terminology for clastic sedimentary rocks (USDA, 1987; Folk, 1974), along with the associated VG parameter from Carsel and Parrish (1988). 177 3.1. 178 The study is patterned after the Triassic Bunter Sandstone Formation (BSF) of the 179 Southern North Sea (SNS) in the United Kingdom (UK) sector (Williams et al., 180 2013). The BSF is a reservoir unit composed of predominantly medium- to coarse- 181 grained sandstone units of metre-scale upward coarsening regime interbedded with 182 fine-grained sediments (Rhys, 1974). It is described as the major gas producing 183 reservoir in the SNS. Most of the BSF is filled with saline water and considered to 184 have significant CO2 storage potential. In the UK sector, it overlies the Triassic 185 Bunter shale formation and is sealed by mudstones and evaporites of the upper 186 Triassic Haisborugh Group (Brook et al., 2003). Structurally, Bunter sandstones 187 contain several periclines commonly referred to as Bunter domes (Williams et al., 188 2013). Based on previous investigations and studies, one such Bunter dome in the UK 189 sector was recently identified by the Energy Technologies Institute’s UK CO2 Storage 190 Appraisal Project (UKSAP) as a promising candidate for CO2 storage (James et al., 191 2016). This dome is penetrated by Well 44/26-01, a deep exploration well completed 192 in 1968 with interpreted log data identifying the strata within the dome (see Fig. 1). 193 The BSF within this dome is interpreted as having five intra-reservoir sandstone zones 194 possessing interbedded shale and cemented sandstone layers. A detailed description of 195 the sedimentology and lithostratigraphy of this dome, hereafter referred to as Bunter 196 aquifer, was given by Williams et al. (2013). Here we present only a brief overview 197 of Bunter aquifer in order to justify the lithological modelling approach used to Model Development 8 198 investigate the multiphase fluid flow regime resulting from pore scale variation in the 199 reservoir-seal interface. The reservoir-seal interface is assumed to be Zone 1 (Fig. 1) 200 and henceforth referred to as the transition zone. Because this is a generic study of 201 CO2 storage in deep sandstone aquifers overlain by mudstones, rather than the study 202 of a specific aquifer, the goal was to select representative characteristics for the 203 aquifer as a base case for systematic parameter study. As such, the thickness and other 204 aquifer characteristics were based on the log data from Well 44/26-01. 205 206 207 Fig. 1: Lithostratighraphical correlation of Bunter Well 44/26-01 from logging data (Williams et al., 2013). 9 208 3.2 209 A simplified 3D static geological model with an areal size of 2 km x 2 km and a 210 thickness of 300 m was developed and discretised into a total of 544,000 active cells 211 (ni = 80, nj = 80, nk = 85) using Schlumberger’s PETREL software (Schlumberger, 212 2016). Although the study is based on a dome-like structure, the geological layering 213 in this model is horizontal (Fig. 2). 214 Numerical Modelling Zones Top depth (m) Number of layers Rot Halite 1200 4 Claystone 1225 12 R.Zone 1 1237 8 R.Zone 2 1245 10 R.Zone 3 1264 15 R.Zone 4 1350 22 R.Zone 5 1438 5 Base Shale 1455 9 Table 3: Vertical grid discretisation for the modelling domain. 215 216 Fig. 2: Reservoir model used for ECLIPSE simulations. 217 An average horizontal permeability (Kh) value of 6.5 x 10-3 mD was assigned to the 218 top and base seal lithologies, after Spain and Conrad (1997), while the average Kh for 10 219 the reservoir was assumed to be 233 mD. The top seal capacities of the Solling 220 Claystone and the Rot Halite were assigned porosity values of 4% and 1% 221 respectively. This was based on the range of porosity values in the Solling, Rot, and 222 Muschelkalk caprocks above the BSF in the southern Dutch North Sea (Spain and 223 Conrad, 1997). The base seal and reservoir formation in the model were assigned 224 average porosity values of 4% and 22% respectively. Permeability anisotropy was 225 assumed to be 0.3 since the average vertical permeabilities of the Bunter sandstone 226 are reported to be typically some 30 % lower than the horizontal permeabilities (Noy 227 et al., 2012). Pore fluid in the domain was modelled under an isothermal condition of 228 42ºC and an initial pressure of 12 MPa with a brine pore fluid gradient of 10.7 229 MPa/km. This implies a pore fluid density of 1.09 g/cc at a salinity of 133,000 ppm. 230 Pressure control consideration for dynamic modelling is 75% of a lithostatic pressure 231 gradient of 22.5 MPa/km (after Noy et al., 2012). 232 Pc-S-kr relationships were generated under the assumption of a strongly water wet 233 system with a CO2/brine interfacial tension of 30 mN/m, following published results 234 by Hebach et al. (2002), Chiquet et al. (2007), and Perrin and Benson (2010). 235 Draining and imbibition curves were included allowing for the residual trapping of 236 CO2 to be modelled (Fig 3): 237 238 239 Fig. 3: Pc-S-kr functions for (a) drainage relative permeability, (b) drainage capillary pressure, (c) imbibition relative permeability, and (d) imbibition capillary pressure 11 240 CO2 saturation end points for the reservoir and seal were based on published results 241 for the Captain formation in the North Sea Goldeneye Field (Shell, 2011) and the 242 Colorado Shale (Bennion and Bachu, 2008) respectively. The capillary displacement 243 pressure of shale was assumed to be 4.7 MPa after Spain and Conrad’s (1997) 244 experimental investigation on the Solling Claystone in the southern Dutch North Sea. 245 In the absence of closely related data, the maximum pore throat size in the reservoir 246 was assumed to be 37 microns. This value falls within the range of dominant pore 247 throat sizes of Permo-Triassic sandstones in the United Kingdom (Bloomfield et al., 248 2001). Numerical simulations were conducted using ECLIPSE E300 (Schlumberger, 249 2015) which adopts Darcy’s law description for immiscible two-phase flows in 250 porous media (Bear, 1972). This study assumes no conductive faults, nor cemented 251 sand layers, interbedded shale or leaky wellbores in the formation. 252 3.3 253 Simulation studies were conducted in aquifer systems idealised as “closed” and 254 “open” to observe the impact of the two sedimentary structures identified in Section 1 255 on CO2 storage. The closed aquifer system was identified as Aquifer-1 while the open 256 aquifer system was identified as Aquifer-2. The concept of graded bedding was 257 investigated using normal grading where the strata coarsens downwards, and inverse 258 grading where the strata coarsens upwards. Five reservoir lithologies were identified 259 from Table 2. In the order of decreasing particle size, these reservoir lithologies are 260 sandstone, silty sandstone, muddy sandstone, clayey sandstone, and sandy siltstone, 261 respectively. The spatial porosity value of 22% remained the same for all the reservoir 262 lithologies. However, permeability data for the varying lithologies were extrapolated 263 from rock permeability values used in UKSAP’s 2016 report for the intra-reservoir 264 zones (James et al., 2016): 265 266 Sensitivity Design Rock lithology Rock permeability [mD] Sandstone (S) 233 Silty Sandstone (SiS) 223 Muddy Sandstone (MS) 219 Clayey Sandstone (CS) 195 Sandy Siltstone (SSi) 162 Table 4: Permeability data for reservoir rock lithologies The plot of the sensitivity study was outlined in three phases: 12 267  Phase I focused on the effect of varying the dynamic properties of the rock 268 geometry (i.e. the Pc–S–kr functions) in the reservoir model at a constant 269 permeability within the reservoir. 270  Phase II focused on the effect of varying the permeability values and the Pc– 271 S–kr functions in the reservoir. Simulation cases in this phase were identified 272 by the suffix “A”. 273  Phase III cases, identified by the suffix “B”, were modelled with variable 274 permeability values and a single Pc–S–kr function within the reservoir. This 275 was to compare, in magnitude, the “stand-alone” effect of Pc–S–kr functions 276 over intrinsic permeability functions in the modelled domain. 277 For this study, permeability and porosity data are henceforth regarded as the static 278 functions while Pc–S–kr functions are regarded as the dynamic functions. The base 279 case for the simulation regarded all five reservoir zones as sandstone and was 280 identified as CASE 1. Sensitivity cases were then labelled according to the description 281 in Table 5: Reservoir Zone Case 1 2 3 4 5 1 Sandstone Sandstone Sandstone Sandstone Sandstone 2 Sandstone Silty Sandstone Clayey Sandstone Muddy Sandstone Muddy Sandstone Clayey Sandstone Silty Sandstone Sandy Siltstone 3 4 5 6 7 282 Sandy Sandstone Siltstone Silty Sandstone Muddy Sandstone Sandstone Clayey Sandstone Sandy Siltstone Table 5: Pore geometric parameters for the reservoir simulation. 283 3.3.1 Aquifer-1 284 Aquifer-1 was confined vertically and laterally within the modelled domain (Fig 2) 285 and had a reservoir pore volume of 1.93 x 108 m3. This aquifer was used to investigate 286 the impact of gradational contact and graded bedding on the reservoir’s injectivity and 13 287 physical trapping mechanisms. In this aquifer, a numerical simulation was initiated at 288 an annual CO2 injection rate of 100,000 metric tonnes through an injection well 289 peforated in R.Zone 4 and 5. The sensitivity plots of Phase I, II and III were 290 investigated in this aquifer. 291 3.3.2 Aquifer-2 292 Aquifer-2 was confined in the vertical boundaries of the modelled domain but was 293 assumed to have lateral aquifer connection. This aquifer was used to investigate the 294 impact of gradational contact and graded bedding on overpressure at the reservoir-seal 295 interface. The concept of an open aquifer was introduced in the study because closed 296 aquifers do not communicate with other reservoirs, laterally, and as a result, may be 297 under- or over-pressured following the CO2 injection (Elewaut et al., 1996). For two- 298 phase flow in porous media, one important role the aqueous phase plays in affecting 299 the evolution of CO2 plume is that it serves as a pressure transmission medium within 300 the porous media (Pruess and Nordbotten, 2011). As a result, the ease with which the 301 migrating CO2 plume evacuates brine from the pore space will influence the pressure 302 evolution within the formation. 303 4. Results and Discussion 304 4.1 Reservoir Injectivity 305 For the simulation of gas injection, CO2 plume was observed to rise vertically to the 306 superjacent impermeable barrier. This buoyant migration of the plume was due to the 307 density difference between the supercritical CO2 and brine. Simulation results for 308 reservoir injectivity for all three phases of the analysis showed an equivalence in CO2 309 injection with time for the first 13 years of injection before reaching the limiting field 310 pressure in the 14th year of injection (Fig. 4): 14 311 312 Fig. 4: CO2 injection rates for all sensitivity cases. 313 The illustrations in Fig. 4 show the importance of the Pc–S–kr functions on a 314 reservoir’s injectivity. Incorporating this dynamic relationship for heterogeneity at the 315 transition zone was a major influence on the reservoir injectivity as the pore fluid 316 pressure approached the well control pressure. For gradational contact at the 317 reservoir-seal interface, the rate of CO2 injection into the lower part of the reservoir 318 increases with the decrease in size of clastic sediments at the top of the reservoir. For 319 the graded reservoir, the rate of CO2 injection favors normal grading over reverse 320 grading. This can be seen from the 14th year of injection. The results indicate that the 321 relative permeability functions predominate over permeability and porosity data when 322 describing sedimentary heterogeneity. This is further emphasised in the comparison 323 between the total CO2 injected for all cases investigated. Fig. 5 shows neglible 324 differences in the total amount of CO2 injected between the base case and other 325 sensitivity cases at the end of simulation for Phase III as opposed to Phases I and II: 15 326 327 Fig. 5: Total volume of CO2 injected into the reservoir during Phase III analysis 328 At the end of the simulation, all cases that used the Pc–S–kr functions to describe 329 hetergeneity within the model allowed for more CO2 injection than the base case. In 330 Fig. 5, Cases 4 and 4A show the smallest margin in total CO2 injection and this 331 accounts for an additional 23,000 tonnes of CO2 being injected into the reservoir. 332 4.2 333 At the end of the injection period, the upward migration of CO2 plume was restrained 334 by the caprock layer in all the cases simulated. When graded bedding was 335 incorporated into the pore geometric analysis, the impact of the dynamic functions 336 followed the trend identified in Section 4.1 and amplified the supporting role of the 337 static parameter thereby decreasing the effective permeability to the non-wetting 338 phase. This was attributed to the impact of the irreducible aqueous phase on the 339 relative permeability to CO2 within the reservoir. Fig. 6 illustrates this impact based 340 on the VGMC-model (section 2.2) which described a constant KrCO2-S curve for all the 341 reservoir lithologies in this study: Physical Trapping 16 342 343 344 Fig. 6: Relative permeability curves showing the value for kr(CO2) at the intercept of kr(brine) in various reservoir lithologies 345 Fig. 6 shows a decreasing value of the relative permeability to CO2 at the intercept 346 between the non-wetting krnw-S curve and the variable wetting krw-S curves for the 347 reservoir rocks. This resulted in a lower degree of mobile CO2 in models that 348 incorporated smaller clasts within the reservoir, particularly at the transition zone. The 349 relative drag in plume movement within the constricting rock matrix led to an 350 increase in the local capillary trapping, a trapping mechanism resulting from intrinsic 351 capillary heterogeneity (Saadatpoor et al., 2010). This was duly represented by the 352 Pc–S relationships in the model and explains why the impact of the static parameter 353 on capillary trapping is not noticeable for the gradational changes investigated. 354 Retention of CO2 within the pore spaces is enhanced by the capillary forces acting at 355 the pore throats. Due to the larger distribution of the capillary processes, graded 356 bedding in the reservoir accounted for a higher degree of capillary trapping when the 357 static and dynamic parameters were integrated. Normally graded reservoirs were seen 358 to residually trap more CO2 than their inversely graded counterparts. This was 359 attributed to the gradual rise in the magnitude of capillary forces acting within 360 normally graded stratum, as opposed to the fall in magnitude for the inversely graded 361 stratum. Fig. 7 shows the quantification of CO2 trapping for all cases modelled in 362 Aquifer-1: 17 363 364 365 Fig. 7: a) Quantification of the physical trapping mechanisms and b) the total CO2 trapped, in order of increasing total trapping from top to bottom, at the end of the injection period. 366 We observe in Fig. 7a that more gas is trapped residually as we proceed from the top, 367 Case 1, to the bottom, Case 3A, of the chart. The prominence of capillary trapping 368 within the reservoir serves to reduce the rate of CO2 spreading at the base of the 369 caprock, as well as increasing brine contact which is beneficial for CO2 dissolution 370 (Golding et al., 2011). This was noted through the lateral extent of plume migration 371 beneath the caprock for all simulated cases which followed the trend 1 > 4 > 2> 5 > 6 372 > 7 > 3 in the order Phase III > Phase I > Phase II. It suggests that the failure to 373 include a variance in the Pc–S–kr functions within the reservoir domain will lead to an 374 over estimation of bouyant drive to- and the gravity current at- the transition zone. 375 Following this observation, the open aquifer, i.e. Aquifer-2, became only an extension 376 of Phase II for an analysis on overpressure. 18 377 4.3 Pressure Evolution 378 To simulate the pressure evolution in the reservoir we first assumed an infinite lateral 379 communication at both ends of the modelled domain. This was undertaken to identify 380 which of the cases of gradational contact at the transition zone and gradation in the 381 reservoir would have the least impact on the structural integrity of the caprock. The 382 assumption of an infinite-acting aquifer was reasonably based on the vast lateral 383 extent of the Bunter sandstone rock unit which crops up onshore in Eastern England 384 as the Sherwood Sandstone Group (Brook et al., 2003). To assess the structrual 385 trapping mechanism, we quantified the volume of mobile CO2 lodged at the transition 386 zone after 20 years of CO2 injection (Fig. 8). 387 388 389 Fig. 8: Mobile CO2 in the reservoir-seal gradation zone of an aquifer with infinite lateral communication. 390 As illustrated in section 4.2, the proportion of mobile CO2 at the transition zone is 391 influenced by the average particle-size within the rock matrix. This can be seen in a 392 comparison between Case 3A and 2A where a decreasing particle-size, from the base 393 to the top of Case 3A aquifer, progressively reduced the amount of free gas migrating 394 vertically. On the other hand, the increasing particle-size from the base to the top of 395 Case 2A propelled the vertical migration of CO2 plume. Following the observations in 396 Fig. 7, Case 3A and 7A were chosen for respective analysis on the impact of a graded 397 reservoir and a gradational contact at the reservoir-seal interface on the pressure 398 distribution in the domain. These cases showed the lowest magnitude of buoyant force 19 399 in the transition zone. Consequently, Case 1, 3A, and 7A were simulated in a version 400 of Aquifer-2 that reflected the probable pore volume of the Bunter Sandstone 401 Formation (Table 6). Reservoir Formation Domain Reservoir Pore Volume (rm3) Reference Aquifer-2 (Infinite-acting) 4.83E+14 Current study Aquifer-2 (Bunter-estimation) 1.45E+12 Current study Bunter Sandstone 1.52E+12 (Brook et al., 2003) Bunter Sandstone 1.396 E+12 (Holloway et al., 2006) 402 Table 6: Provisional figures for reservoir pore volume used in this study 403 In the Bunter-estimate version of Aquifer-2, overpressure in the transition zone varied 404 directly with the mass of free CO2 in the strata. However, pressure evolution in the 405 overlying caprock did not show such correlation (Fig 9): 20 406 407 408 409 Fig. 9: Time plots showing: a) mobile CO2, b) pressure evolution, and c) CO2 dissolution through the injection period; as well as d) the percentage volume of total dissolved and mobile CO2 at the 20th year of injection, in the transition zone and the caprock respectively. 410 This disparity was accounted for by the measure of capillary trapped CO2 within each 411 strata. This is because in pore spaces, the incumbent aqueous phase will further 412 dissolve immobilised CO2 ganglia which can account for the pressure drop (Peters et 413 al., 2015). In other words, the degree of CO2 dissolution through residual trapping 414 within a strata serves to counteract the impact of mobile CO2 saturation on the pore 415 fluid pressure (Fig. 9d). Notwithstanding, the results suggest that pore pressure within 21 416 caprocks superjacent to graded strata at the reservoir/seal interface will show a lower 417 evolution profile in comparison to those that are further removed. This assumption, 418 however, is mostly valid for a field-scale determination of pressure evolution within 419 the caprock. Generally, higher capillary forces resulting from smaller pore geometry 420 tend to thicken the horizontal gravity current as a result of the reduced effective 421 permeability of the intruding CO2 (Fig. 10): 422 423 424 Fig. 10: A depiction of CO2 saturation in the 20th year of gas injection. NB: The curves X and Y in Case1 illustrate the trend for gravity current in Case7A and Case3A respectively. 425 This usually results in a larger capillary fringe, i.e the region occupied by both phases. 426 With constant CO2 flux, the partial saturation of the non-wetting phase within the 427 capillary fringe increases and thicker horizontal currents contact a greater region of 428 the reservoir (Golding et al., 2013). This has an immediate effect on pressure 429 evolution within the contact area, as localised pore pressures increase while the 430 capillary forces within the matrix immobilise the CO2 ganglia. This phenomenon was 431 notably observed around the injection well within the reservoir and the caprock a t the 432 end of the numerical simulation (Fig. 11): 22 433 434 Fig. 11: 2D illustration of pressure change in the 20th year of CO2 injection. 435 The reasoning is that within a thinning pore matrix, higher capillary forces correlate to 436 a higher saturation of irreducible brine. The lateral continuity of such a matrix will 437 result in little or no path being available for the migrating CO2 plume to bypass the 438 constricted strata, hence the gravity current expands beneath it. A consequence is the 439 increased local capillary trapping of the gas within the strata, while the continous flux 440 of the buoyant CO2 plume results in CO2 permeability though the region of highest 441 gas concentration. The significance of a laterally continous reservoir-seal gradation 442 zone within a semi-finite aquifer is a higher overpressure around the injection point, 443 thus increasing the magnitude of pressure transmitted in the lower part of the caprock. 23 444 5 Summary and Conclusions 445 Numerical modelling of CO2 geosequestration is to a large extent dependent on the 446 quality of the quantitative knowledge of the geological descriptions that is used in the 447 construction of the reservoir model. Through relating fluid and transport processes to 448 primary sedimentary structures in siliciclastic formations, we employed numerical 449 simulation to probe the heterogeneic effects of dynamic flow parameters on CO2 450 storage performance. The results emphasise the significance of enhancing geological 451 details in reservoir-specific models. Specifically, we identified the importance of 452 modelling heterogeneity in the capillary pressure and relative permeability functions. 453 We have demonsrated that for CO2 storage in geological formations, the reservoir 454 injectivity and trapping mechanisms are sensitive to gradational changes at the 455 reservoir-seal interface as well as within the reservoir. Clast-size gradation from 456 coarser- to finer- sediments within the reservoir leads to more favorable capillary 457 trapping scenarios for CO2 sequestration, irrespective of the boundary conditions. 458 Gradation further increases the opportunity for CO2 dissolution during the injection 459 phase. Hence, the presence of these structures is vital in numerical models that 460 investigate the post-injection sequestration processes. We also showed that the 461 measure of how such sedimentary structures influence CO2 storage will not be 462 adequately determined if their description is based on the permeability and porosity 463 data alone. This is based on the observation that the relative permeability data 464 essentially dictates the effective permeability of fluids in a porous media. 465 The presence of a gradational contact at the reservoir-seal interface can also impact on 466 the storage security. The study showed that for an open aquifer, the lateral continuity 467 of such structures will likely reduce the field-scale overpressure in the caprock by 468 mitigating brine migration into the seal. However, this could also increase localised 469 pore pressures centred on the injection point within the caprock. Such scenarios can 470 lead to the hydraulic fracturing of structural traps within the injection point, especially 471 at the base of the trapping unit (Rozhko et al., 2007). Gradation at the reservoir-seal 472 interface may then be said to improve field-scale CO2 storage security while also 473 diminishing local-scale caprock integrity. This creates a paradoxical impact of 474 gradation on structural trap integrity and further goes to highlight the importance of 475 including such geological detail in numerical simulation studies. 24 476 In summary, we conclude that numerical models which disregard the sensitivity of 477 geological detail to multi-phase fluid transport processes will fail to sufficiently 478 account for CO2 storage performance. This is specifically with respect to various Pc– 479 S–kr relationships that may arise from the variance in pore geometry. We 480 acknowledge that the present results were obtained under the simplifying assumption 481 that variations in these constitutive functions only depend on the average grain size. 482 There is room for further investigation by considering the effects of additional factors 483 such as cemented sand layers, impermeable faults, leaky well bores, etc. on changes 484 in hydraulic properties. For instance, faults are important in compartmentalising 485 reservoirs and modifying the depositional continuity (Bouvier et al., 1989). The 486 increased 487 communication can influence how primary sedimentary structures define reservoir 488 flow processes. Also, subsequent diagenetically precipitated materials post particle 489 deposition can form tightly cemented flow barriers within the reservoir. Laterally 490 continous cementation not only constitutes barriers to flow but may also form 491 pressure seals which can impact on the reservoir injectivity (Bjørkum and 492 Walderhaug 1990). Hence, detailed sedimentary and petrographic analyses, including 493 the fine-scale examination of well data and reservoir-specific models, are required to 494 adequately predict CO2 storage performance. Our future work will include other 495 sedimentary features including faults with different transmissibility and cemented 496 sand in the model in order to study their influence on the results. knowledge of fault-induced reservoir compartmentalisation and 497 498 499 500 501 Acknowledgments 502 This material is based upon work supported by Coventry University’s Flow 503 Measurement and Fluid Mechanics Research Centre. The authors would like to thank 504 Schlumberger for the use of ECLIPSE and Petrel Software. We appreciate the 505 constructive input of Dr. Adrian Wood and Dr. Philip Costen, including valuable 506 comments from three anonymous reviewers. 25 507 References 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 Al-khdheeawi, E.A., Vialle, S., Barifcani, A. et al. 2017. Impact of reservoir wettability and heterogeneity on CO2 plume migration and trapping capacity. International Journal of Greenhouse Gas Control 58: 142-158. Bachu, S. 2015. Review of CO2 storage efficiency in deep saline aquifers. International Journal of Greenhouse Gas Control 40: 188-202. Bear, J. 1972. Dynamics of Fluids in Porous Media. New York: American Elsevier Publishing Company. Bennion, D.B. and Bachu, S. 2008. Drainage and imbibition relative permeability relationships for supercritical CO2/brine and H2S/brine systems in intergranular sandstone, carbonate, shale, and anhydrite rocks. SPE Reservoir Evaluation and Engineering 11: 487-96. Benson, S.M. and Cole, D.R. 2008. CO2 Sequestration in Deep Sedimentary Formations. Elements 4 (5): 325-31. doi: 10.2113/gselements.4.5.325. Benton, M.J., Cook, E. and Turner, P. 2002. Permian and Triassic Red Beds and the Penarth Group of Great Britain, Geological Conservation Review Series, No. 24 edition. Peterborough: Joint Nature Conservation Committee. Bigelow, E.L. 1992. Introduction to Wireline Log Analysis. Texas: Western Atlas International. Birkholzer, J.T., Zhou, Q. and Tsang, C. 2009. Large-scale impact of CO2 storage in deep saline aquifers: A sensitivity study on pressure response in stratified systems. International Journal of Greenhouse Gas Control 3 (2): 181-94. Bjørkum, P.A., and Walderhaug, O. 1990. Lateral extents of calcite-cemented zones in shallow marine sandstones. In North Sea Oil and Gas Reservoirs – II, ed. A. T. Buller, E. Berg, O. Hjelmeland et al., pp. 331-336. London: Graham and Trotman. Bjørlykke, K. 2010. Petroleum Geoscience: From Sedimentary Environments to Rock Physics. Berlin: Sringer-Verlag. Bloomfield, J.P., Gooddy, D.C., Bright, M.I. et al. 2001. Pore-throat size distributions in Permo-Triassic sandstones from the United Kingdom and some implications for contaminant hydrogeology. Hydrogeology Journal 9 (3): 219-30. Boggs, S. 2009. Petrology of sedimentary rocks, 2nd edition. Cambridge: Cambridge University Press. Bondor, P.L. 1992. Applications of carbon dioxide in enhanced oil recovery. Energy Conversion and Management 33 (5): 579-86. Bouvier, J.D., Kaars-Sijpesteijn, C.H., Kluesner, D.F. et al. 1989: Three-dimensional seismic interpretation and fault sealing investigations, Nun River Field, Nigeria. AAPG Bulletin 73 (11): 1397-1414. Brook, M., Shaw, K., Vincent, C. et al. 2003. Storage potential of the Bunter Sandstone in the UK sector of the Southern North Sea and the adjacent onshore area of eastern England. CR/03/154. Brooks, R.H. and Corey, A.T. 1966. Properties of Porous Media Affecting Fluid Flow. Journal of the Irrigation and Drainage Division 92 (2): 61-90. Burton, M., Kumar, N. and Bryant, S.L. 2009. CO2 injectivity into brine aquifers: Why relative permeability matters as much as absolute permeability. Energy Procedia 1 (1): 3091-8. Cameron, D.A. and Dyrlofsky, L.J. 2012. Optimization of well placement, CO2 injection rates, and brine cycling for geological carbon. International Journal of Greenhouse Gas Control 10: 100-112. Carsel, R.F., Parrish, R.S. 1988. Developing joint probability distributions of soil water retention characteristics. Water Resources Research 24 (5): 755-769. Chiquet, P., Daridon, J., Broseta, D. et al. 2007. CO2/water interfacial tensions under pressure and temperature conditions of CO2 geological storage. Energy Conversion and Management 48 (3): 736-44. 26 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 Class, H., Edigbo, A., Helmog, R. et al. 2009. A benchmark study on problems related to CO2 storage in geologic formations. Computational Geosciences 13 (4): 409. Corey, A.T. 1954. The interellation between gas and oil relative permeabilites. Producers Monthly 19: 38-41. Darling, T. 2005. Well Logging and Formation Evaluation. Houston, Texas: Elsevier, Gulf Professional Publishing. Dempsey, D., Kelkar, S., Pawar, R. et al. 2014. Modeling caprock bending stresses and their potential for induced seismicity during CO2 injection. International Journal of Greenhouse Gas Control 22: 223-36. Doughty, C. and Pruess, K. 2004. Modeling supercritical carbon dioxide injection in heterogeneous porous media. Vadose Zone Journal 3: 837-47. Doveton, J.H. 1991. Lithofacies and Geochernical-Facies Profiles from Modern Wireline Logs - New Subsurface Templates for Sedimentary Modeling. In Sedimentary Modelling: Computer Simulations and methods for improved parameter definition, ed. E.K. Franseen, W.L. Watney, C.G. St. Kendall et al., pp. 101-110. Kansas: Kansas Geological Survey Bulletin 233. Elewaut, E., Koelewijin, D., van der Straaten, R. et al. 1996. Inventory of the theoretical CO2 storage capacty of the European Union and Norway. In The Underground Disposal of Carbon Dioxide. Final Report of Joule II Project No. CT92-0031, ed. S. Holloway, pp. 16-115. Keyworth, Nottingham: British Geological Survey. Ennis-King, J.P. and Paterson, L. 2005. Role of Convective Mixing in the Long-Term Storage of Carbon Dioxide in Deep Saline Formations. SPE Journal 10 (03): 349-56. Espinet, A., Shoemaker, C. and Doughty, C. 2013. Estimation of plume distribution for carbon sequestration using parameter estimation with limited monitoring data. Water Resources Research 49 (7): 4442-4464. Fabricius, I.L., Fazladic, L.D., Steinholm, A. et al. 2003. The use of spectral natural gammaray analysis in reservoir evaluation of siliciclastic sediments: a case study from the Middle Jurassic of the Harald Field, Danish Central Graben. Geological Survey of Denmark and Greenland Bulletin 1: 349-66. Fleet, M., Gurton, R. and Taggart, I. 2004. The Function of Gas-Water Relative Permeability Hysteresis in the Sequestration of Carbon Dioxide in Saline Formations. Presented at the SPE Asia Pacific Oil and Gas Conference and Exhibition, Perth, Australia, 18-20 October. Folk, R.L. 1974. The Petrology of Sedimentary Rocks. Texas: Hemphill Publishing Company. Golding, M.J., Huppert, H.E. and Neufeld, J.A. 2013. The effects of capillary forces on the axisymmetric propagation of two-phase, constant-flux gravity currents in porous media. Physics of Fluids 25 (3): 036602. doi: 10.1063/1.4793748. Golding, M.J., Neufeld, J.A., Hesse, M.A. et al. 2011. Two-phase gravity currents in porous media. Journal of Fluid Mechanics 678: 248-70. doi: 10.1017/jfm.2011.110. Gor, G.Y., Elliot, T.R. and Prevost, J.H. 2013. Effects of thermal stresses on caprock integrity during CO2 storage. International Journal of Greenhouse Gas Control 12: 300-309. Gozalpour, F., Ren, S.R. and Tohidi, B. 2005. CO2 EOR and Storage in Oil Reservoir. Oil & Gas Science and Technology - Rev.IFP 60 (3): 537-46. Grigg, R.B. 2005. Long-Term CO2 Storage: Using Petroleum Industry Experience. In Carbon Dioxide Capture for Storage in Deep Geologic Formations, ed. D.C. Thomas, pp. 853865. Amsterdam: Elsevier Science. Hebach, A., Oberhof, A., Dahmen, N. et al. 2002. Interfacial Tension at Elevated PressuresMeasurements and Correlations in the Water + Carbon Dioxide System. Journal of Chemical & Engineering Data 47 (6): 1540-6. Hiscott, R.N. 2003. Grading, graded bedding. In Sedimentology, Chap. 535-538. Dordrecht: Springer Netherlands. Haldorsen, H.H. 1986. Simulation parameter assignment and the problem of scale in reservoir engineering. In Reservoir Characterisation, ed. L.W. Lake and H.B. Carroll, Jr., pp. 293-340. London: Academic Press. 27 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 Holloway, S., Vincent, C.J., Bentham, M.S. et al. 2006. Top-down and bottom-up estimates of CO2 storage capacity in the United Kingdom sector of the southern North Sea basin. Environmental Geosciences 13 (2): 71-84. doi: 10.1306/eg.11080505015. IPCC 2005. IPCC Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group III of the Intergovernmental Panel on Climate Change. James, A., Baines, S. and McCollough, S. 2016. D10: WP5A – Bunter Storage Development Plan. 10113ETIS-Rep-13-03: 1-212. Juanes, R., Spiteri, E.J., Orr, F.M. et al. 2006. Impact of relative permeability hysteresis on geological CO2 storage. Water Resources Research 42 (12): W12418. Katahara, K.W. 1995. Gamma Ray Log Response in Shaly Sands. The Log Analyst 36 (4): 50-71. Kopp, A., Class, H. and Helmig, R. 2009. Investigations on CO2 storage capacity in saline aquifers. International Journal of Greenhouse Gas Control 3 (3): 263-76. Kovscek, A.R. and Cakici, M.D. 2005. Geologic storage of carbon dioxide and enhanced oil recovery. II. Cooptimization of storage and recovery. Energy Conversion and Management 46 (11): 1941-56. Krevor, S., Blunt, M.J., Benson, S.M. et al. 2015. Capillary trapping for geologic carbon dioxide storage – From pore scale physics to field scale implications. International Journal of Greenhouse Gas Control 40: 221-237. Kumar, A., Ozah, R., Noh, M. et al. 2005. Reservoir Simulation of CO2 Storage in Aquifers. SPE Journal 10 (03): 336-48. Lawter, A.R., Qafoku, N.P., Asmussen, R.M. et al. 2017. Risk of Geologic Sequestration of CO2 to Groundwater Aquifers: Current Knowledge and Remaining Questions. Energy Procedia 114 (Supplement C): 3052-9. Martin, D.F. and Taber, J.J. 1992. Carbon Dioxide Flooding. Journal of Petroleum Technology 44 (04): 396-400. Meckel, T.A., Bryant, S.L., Ganesh, P.R., 2015. Characterization and prediction of CO2 saturation resulting from modelling buoyant fluid migration in 2D heterogeneous geologic fabrics. International Journal of Greenhouse Gas Control 34: 85–96. Mo, S., Zweigel, P., Lindeberg, E. et al. 2005. Effect of Geologic Parameters on CO2 Storage in Deep Saline Aquifers. Presented at the SPE Europec/EAGE Annual Conference, Madrid, Spain, 13-16 June. Mori, H., Trevisan, L. and Illangasekare, T.H. 2015. Evaluation of relative permeability functions as inputs to multiphase flow models simulating supercritical CO2 behavior in deep geologic formations. International Journal of Greenhouse Gas Control 41: 328335. Mualem, Y. 1976. A new model for predicting the hydraulic conductivity of unsaturated porous media. Water Resources Research 12 (3): 513-22. Nazeer, A., Abbasi, S.A. and Solangi, S.H. 2016. Sedimentary facies interpretation of Gamma Ray (GR) log as basic well logs in Central and Lower Indus Basin of Pakistan. Geodesy and Geodynamics 7 (6): 432-43. Newell, A. and Shariatipour, S.M. 2016. Linking outcrop analogue with flow simulation to reduce uncertainty in sub-surface carbon capture and storage: an example from the Sherwood Sandstone Group of the Wessex Basin, UK. In The value of outcrop studies in reducing subsurface uncertainty and risk in hydrocarbon exploration and production, ed. M. Bowman, H.R. Smyth, T.R. Good et al., pp. 231-246. London: Geological Society Special Publications 436. Newell, A.J. 2017. Evolving stratigraphy of a Middle Triassic fluvial-dominated sheet sandstone: The Otter Sandstone Formation of the Wessex Basin (UK). Geological Journal: 1-19. Nicol, A., Carne, R., Gerstenberger, M. et al. 2011. Induced seismicity and its implications for CO2 storage risk. Energy Procedia 4 (Supplement C): 3699-706. Noy, D.J., Holloway, S., Chadwick, R.A. et al. 2012. Modelling large-scale carbon dioxide injection into the Bunter Sandstone in the UK Southern North Sea. International Journal of Greenhouse Gas Control 9: 220-33. 28 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 Obi, E.I. and Blunt, M.J. 2006. Streamline-based simulation of carbon dioxide storage in a North Sea aquifer. Water Resources Research 42, (3): 1-13. doi: 10.1029/2004WR003347. Oldenburg, C.M., Pruess, K. and Benson, S.M. 2001. Process modeling of CO2 injection into natural gas reservoirs for carbon sequestration and enhanced gas recovery. Energy Fuels 15 (2): 293-298. Oostrom, M., White, M.D., Porse, S.L. et al. 2016. Comparison of relative permeability– saturation–capillary pressure models for simulation of reservoir CO2 injection. International Journal of Greenhouse Gas Control 45: 70-85. Perrin, J. and Benson, S. 2010. An Experimental Study on the Influence of Sub-Core Scale Heterogeneities on CO2 Distribution in Reservoir Rocks. Transport in Porous Media 82 (1): 93-109. doi: 10.1007/s11242-009-9426-x. Peters, E., Egberts, P.J.P., Loeve, D. et al. 2015. CO2 dissolution and its impact on reservoir pressure behavior. International Journal of Greenhouse Gas Control 43: 115-23. Pettijohn, F.J. 1957. Sedimentary rocks. 2nd Edn. New York: Harper. Pettijohn, F.J., Potter, P.E. and Siever, R. 1972. Sand and sandstones. New York: SpringerVerlag. Pruess, K., Xu, T., Apps, J. et al. 2003. Numerical Modeling of Aquifer Disposal of CO2. SPE Journal 8 (01): 49-60. Pruess, K. and Nordbotten, J. 2011. Numerical Simulation Studies of the Long-term Evolution of a CO2 Plume in a Saline Aquifer with a Sloping Caprock. Transport in Porous Media 90 (1): 135-51. doi: 10.1007/s11242-011-9729-6. Rhys, G.H. 1974. A proposed standard lithostratigraphic nomenclature for the southern North Sea and an outline structural nomenclature for the whole of the (UK) North Sea. Rozhko, A.Y., Podladchikov, Y.Y. and Renard, F. 2007. Failure patterns caused by localized rise in pore-fluid overpressure and effective strength of rocks. Geophysical Research Letters 34 (22): L22304. doi: 10.1029/2007GL031696. Saadatpoor, E., Bryant, S.L. and Sepehrnoori, K. 2010. New Trapping Mechanism in Carbon Sequestration. Transport in Porous Media 82 (1): 3-17. Schlumberger. 2015. ECLIPSE SimLauncher, Version 2015.2.0.0. Schlumberger. 2016. Petrel E&P Software Platform, Version 2016. Shariatipour, S.M., Pickup, G.E. and Mackay, E.J. 2014. The effect of aquifer/caprock interface on geological storage of CO2. Energy Procedia 63: 5544-5555. Sharaitipour, S.M., Pickup, G.E. and Mackay, E.J. 2016. Simulations of CO2 storage in aquifer models with top surface morphology and transition zones. International Journal of Greenhouse Gas Control 54 (1): 117-128. Shell 2011. UK Carbon Capture and Storage Demonstration Competition. UKCCS-KTS7.19-Shell-002, SCAL Report. Spain, D.R. and Conrad, C.P. 1997. Quantitative analysis of top-seal capacity: offshore Netherlands, southern North Sea. Geologie en Mijnbouw 76 (3): 217-26. USDA 1987. Soil Mechanics Level I, Module 3: USDA Textural Soil Classification. Van De Graaff, W.J.E. and Ealey, P.J. 1989. Geological modelling for simulation studies. AAPG Bulletin 73 (11): 1436-1444 Van Genuchten, M.T. 1980. A Closed Form Equation for Predicting the Hydraulic Conductivity of Unsaturated Soils. Soil Science Society of America Journal 44: 892-8. Wentworth, C.K. 1922. A Scale of Grade and Class Terms for Clastic Sediments. The Journal of geology 30 (5): 377-92. doi: 10.1086/622910. Williams, J.D.O., Jin, M., Bentham, M. et al. 2013. Modelling carbon dioxide storage within closed structures in the UK Bunter Sandstone Formation. International Journal of Greenhouse Gas Control 18: 38-50. Zheng, L., Spycher, N., Birkholzer, J. et al. 2013. On modeling the potential impacts of CO2 sequestration on shallow groundwater: Transport of organics and co-injected H2S by supercritical CO2 to shallow aquifers. International Journal of Greenhouse Gas Control 14: 113-27. 29 723 724 Zhou, Q., Birkholzer, J.T., Mehnert, E. et al. 2010. Modeling basin- and plume-scale processes of CO2 storage for full scale deployment. Ground Water 48 (4): 494-514. 30