Storage Capacity Estimation, Site Selection and Characterisation for CO2 Storage Projects

CO2CRC
March 2008, CO2CRC Report No: RPT08-1001

Storage Capacity Estimation, Site Selection and Characterisation for CO2 Storage Projects
CO2CRC
March, 2008 CO2CRC Report No: RPT08-1001

Edited by: J. G. Kaldi1 and C. M. Gibson-Poole1 CO2CRC

With significant contributions from the following CO2CRC researchers: B. Ainsworth1, G. Allinson2, T. Berly3, R. Causebrook4, A. Chirinos4, Y. Cinar2, P. Cook3, T. Dance5, R. Daniel1, K. Dodds5, J. Esterle5, M. Faiz5, R. Funnell6, C. Gibson-Poole1, A. Golab3, L. Gurba2, A. Hennig5, J. Kaldi1, D. Kirste7, L. Paterson5, D. Sherlock5, P. Tingate1, J. Underschultz5, B. Vakarelov1, P. van Ruth1, M. Watson1 and M. Werner1
1

University of Adelaide; 2 University of New South Wales; 3CO2CRC, Canberra; 4 Geoscience Australia; 5CSIRO; 6 Geological and Nuclear Science, NZ; 7Simon Fraser University

Cooperative Research Centre for Greenhouse Gas Technologies (CO2CRC) GPO Box 463 CANBERRA ACT 2601 AUSTRALIA Ground Floor, NFF House 14-16 Brisbane Avenue, Barton, Canberra. CANBERRA ACT 2601 Phone: +61 2 6120 1601 Fax: +61 2 6273 7181 Email: pjcook@co2crc.com.au Web: www.co2crc.com.au Reference: CO2CRC, 2008. Storage Capacity Estimation, Site Selection and Characterisation for CO2 Storage Projects. Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. CO2CRC Report No. RPT08-1001. 52pp. CO2CRC 2008 Unless otherwise specified, the Cooperative Research Centre for Greenhouse Gas Technologies (CO2CRC) retains copyright over this publication through its commercial arm, CO2CRC Technologies Pty Ltd. You must not reproduce, distribute, publish, copy, transfer or commercially exploit any information contained in this publication that would be an infringement of any copyright, patent, trademark, design or other intellectual property right. Requests and inquiries concerning copyright should be addressed to the Communication Manager, CO2CRC, GPO Box 463, CANBERRA, ACT, 2601. Telephone: +61 2 6120 1601.

Table of Contents
Preface............................................................................................................................................................ 1 Executive Summary ...................................................................................................................................... 2 Introduction ................................................................................................................................................... 3 Geological Storage of Carbon Dioxide ........................................................................................................ 3 CO2 Subsurface Behaviour and Trapping Mechanisms........................................................................... 4 Geological Storage Systems for CO2 ....................................................................................................... 6 Saline Formations .............................................................................................................................. 6 Oil and Gas Fields Depleted or Enhanced Recovery...................................................................... 7 Coal Seams Deep Unmineable or Enhanced Methane Recovery ................................................... 7 Storage Capacity ........................................................................................................................................... 8 Previous Studies on Storage Capacity Estimation: DOE & CSLF .......................................................... 9 CO2CRC Storage Capacity Classification (modified from SPE, 2007) ................................................ 10 Methodology for Storage Capacity Estimation...................................................................................... 13 Storage Capacity Estimation in Saline Formations ......................................................................... 13 Storage Capacity Estimation in Depleted Oil and Gas Fields ......................................................... 14 Volumetrics-Based CO2 Storage Capacity Estimate for Depleted Oil and Gas Fields ............... 15 Production-Based CO2 Storage Capacity Estimate for Depleted Oil and Gas Fields ................. 16 Storage Capacity Estimation in Coal Seams.................................................................................... 16 Site Selection and Characterisation........................................................................................................... 18 Country/State-Scale Screening .............................................................................................................. 18 Screening Criteria ............................................................................................................................ 20 Ranking of Sedimentary Basins....................................................................................................... 25 Basin-Scale Assessment ........................................................................................................................ 25 Ranking of Prospective Sites ........................................................................................................... 26 Site Characterisation: Saline Formations and Petroleum Reservoirs..................................................... 28 Geoscience Characterisation: Injectivity ......................................................................................... 29 Injectivity in Low Permeability Reservoirs ................................................................................ 30 Geoscience Characterisation: Containment ..................................................................................... 30 Geoscience Characterisation: Capacity ........................................................................................... 31 Engineering Characterisation........................................................................................................... 32 Socio-Economic Characterisation ................................................................................................... 32 Site Characterisation: Coal Seams ......................................................................................................... 32 Data Needs for Site Characterisation..................................................................................................... 33 Seismic Data.................................................................................................................................... 35 Well Log and Core Data.................................................................................................................. 35 Use of Analogues............................................................................................................................. 36 Formation Pressure .......................................................................................................................... 36 Site Deployment .................................................................................................................................... 37 Conclusions .................................................................................................................................................. 38 References .................................................................................................................................................... 40 Appendix A: CO2 Trapping Mechanisms ................................................................................................. 47 Structural/Stratigraphic Trapping .......................................................................................................... 47 Hydrodynamic Trapping........................................................................................................................ 47 Residual Trapping.................................................................................................................................. 49 Solubility Trapping................................................................................................................................ 49 Mineral Trapping ................................................................................................................................... 50 Storage Security for Saline/Hydrocarbon Formations ........................................................................... 51 Adsorption Trapping on Coal ................................................................................................................ 51

Figures
Figure 1: Estimated depths to critical temperature and pressure for CO2 (from Holliday et al., 1991) .......... 4 Figure 2: Increasing storage effectiveness for CO2 with depth. Note that above critical depth, CO2 is in gaseous state (balloons); below critical depth it is in liquid-like state (droplets). Volumetric relationship shown by blue numbers (eg 100 m3 of CO2 at surface would occupy 0.32 m3 at a depth of 1 km) (CO2CRC; modified from IPCC, 2005).......................................................................................................... 5 Figure 3: Options for the geological storage of carbon dioxide (from CO2CRC; IPCC, 2005). .................... 6 Figure 4: Total storage density as a function of depth, highlighting how the storage density of CO2 adsorbed onto coal at subcritical depths is comparable to the storage density of CO2 captured in pore space at supercritical depths (hydrostatic pressure gradient is 10.5 MPa, mean surface temperature is 15C and geothermal gradient is 25C/km unless noted) (from Ennis-King & Paterson, 2001). ................... 8 Figure 5: CO2 storage volume classification system proposed by CO2CRC (modified from SPE, 2007). . 12 Figure 6: CO2 storage capacity pyramid (modified from CSLF, 2007). ....................................................... 12 Figure 7: Site characterisation workflow for geological storage of CO2 (modified from Gibson-Poole, 2008). .. 19 Figure 8: Integrated scales of site assessment and storage capacity pyramid for geological storage of CO2 (modified from CSLF, 2007). The storage volume level assigned to the scale of characterisation represents the maximum likely assessment level achievable given the likely database available. ............... 20 Figure 9: Australia, with its intracratonic and passive margin shelf basins, exhibits low to moderate earthquake activity, in contrast to the high frequency of earthquakes in Indonesia and New Zealand (from USGS, 2007). ...................................................................................................................................... 22 Figure 10: Map of Australia showing the estimated temperature distribution at 5 km depth. The hot area in central-eastern Australia comprises the Galilee and Cooper-Eromanga basins (image courtesy of the Hot Rock Energy Program, Australian National University) (from Murdoch University, 2006)....... 24 Figure 11: Examples of (a) structural and (b) stratigraphic physical traps for CO2 (modified from Biddle & Wielchowsky, 1994)...................................................................................................................... 47 Figure 12: Hydrodynamic trapping of CO2, where the CO2 migration pathway is 10s to 100s km long allowing for a long residence time. ............................................................................................................... 48 Figure 13: Residual trapping of CO2............................................................................................................. 49 Figure 14: Convective mixing of CO2: example of a numerical simulation showing the high-density plumes of CO2-saturated brine (grey colours) sinking into the brine column below (white colour) (kv/kh = 0.01, after 14400 years) (Ennis-King & Paterson, 2005a).............................................................. 50 Figure 15: Schematic representation of the change of dominant trapping mechanisms and increasing CO2 storage security with time (from IPCC, 2005). ..................................................................................... 51 Figure 16: Pure gas absolute adsorption in standard cubic meters per tonne on Tiffany coals at 55 C (from IPCC, 2005). ....................................................................................................................................... 52

Tables
Table 1: Volumetric equation parameters for capacity calculation in saline formations (from DOE, 2006).13 Table 2: Terms included in storage efficiency factor for saline formations (from DOE, 2006). .................. 13 Table 3: Volumetric equation parameters for storage capacity in oil and gas reservoirs (from DOE, 2006). .. 15 Table 4: Volumetric equation for storage volume estimation for CO2 storage in coal seams (from DOE, 2006). ............................................................................................................................................................ 16 Table 5: Terms included in storage efficiency factor for coal seams (from DOE, 2006).............................. 17 Table 6: Criteria for screening sedimentary basins within a country or state for geological storage of CO2 (modified from Bachu, 2003)........................................................................................................................ 21 Table 7: Ranking factors for saline formations and petroleum reservoirs as prospective CO2 storage sites (modified from Bradshaw & Rigg, 2001; Rigg et al., 2001; Bradshaw et al., 2002). ................................... 26 Table 8: Ranking factors for coal seams as potential CO2 storage sites (modified from Bradshaw et al., 2001).. 27 Table 9: Summary of the main data needs (required as well as desirable) for the various levels of site characterisation and storage volume assessment........................................................................................... 33

ii

Preface
This report addresses an increasingly important issue: how do we determine the capacity of a geological formation at a particular site to store carbon dioxide safely and securely for thousands of years and longer? Also, given that we are dealing with a subsurface system with inherent uncertainties, how do we express our level of confidence in that capacity and of the storage site? Whilst these are interesting scientific questions they are also crucially important commercial, economic, environmental and public policy issues as we contemplate how to tackle the need to decrease emissions of carbon dioxide to the atmosphere. In seeking to address these issues it is important to develop an acceptable and consistent scheme that will provide a uniform language that is understandable to (and usable by) the scientific community but also by industry and the financial community, for if we are to deploy geological storage on a large scale it is essential that we are able to establish whether projects are bankable or not. Through a process involving a wide range of Australian and New Zealand stakeholders from industry, government, federal and state geological surveys, universities and research bodies, CO2CRC has sought to develop a scheme for storage capacity and site characterisation that builds not only on the extensive work of CO2CRC but also, and crucially (for we had absolutely no wish to re invent any wheels!), upon the 2007 schemes of the CSLF and the US DoE. This report extends those schemes in the light of the widely accepted SPE methodology for expressing oil and gas reserves and resources that are accepted by the financial sector, in order to provide the basis for bankable storage capacity. It is recognised that this report is a step along the way, but at the same time, it is an important step in that it provides a storage capacity scheme that we believe will work for Australia and New Zealand. A great many people have contributed to developing this Report and the underlying ideas. I thank them for so freely providing those ideas and for taking the time to participate in the workshops and subsequent discussions. I thank in particular, John Kaldi and Catherine Gibson-Poole for taking on the vital task of compiling those ideas and this report. We hope you find this report useful and welcome your feedback.

Peter J Cook Chief Executive Cooperative Research Centre for Greenhouse Gas Technologies

Executive Summary
Carbon dioxide (CO2) capture and geological storage (CCS) is a means of reducing greenhouse gas emissions into the atmosphere. The determination of carbon dioxide storage capacity and the selection and characterisation of potential sites for CO2 storage are key issues in taking CCS forward. Most current storage capacity estimates are imperfect and there is a need for more understanding of the issues and more general agreement on assessment methodologies for site selection. This report provides an overview of the key issues and challenges in storage capacity estimation and storage site selection and suggests methodologies to address the issues and to meet those challenges. A distinction is made between available pore volume (a theoretical estimate of the amount of pore space that can be used to store CO2 in subsurface geologic formations) and storage capacity (the pore volume constrained by economic or engineering feasibility limitations). The term total pore volume is introduced to define the entire volume which is estimated to exist originally in sedimentary basins, naturally occurring storage systems or potential storage sites, plus the estimated pore volume yet to be identified. Total pore volume is subdivided into discovered and undiscovered pore volume. Storage capacity is considered a resource, and as in petroleum accumulations and mineral deposits, categorised based on levels of certainty of resource availability. Three categories are proposed: operational storage capacity is discovered pore volume that is considered commercially viable i.e. CO2 is injectable under existing technical, economic and regulatory conditions. Operational storage capacity, which is further classified as proved (1P), proved plus probable (2P) and proved plus probable plus possible (3P) categories following standard petroleum industry nomenclature. Contingent storage capacity is discovered pore volume which will be commercial in the future when technical, economic and regulatory conditions are in place and prospective storage capacity is undiscovered pore volume which might become commercial at some future time. Because of uncertainties inherent in subsurface evaluation, exact quantification of geological properties is not possible and therefore storage capacity is always, at best, an approximation of the amount of pore space into which CO2 can be injected. Hence, the likelihood of contingent and prospective storage volumes achieving commerciality is determined probabilistically, utilising high, low and best estimates. The selection of storage sites suitable for significant volumes of CO2 comprises mainly geological evaluation of the applicable storage system (e.g. saline formations, depleted or near depleted oil and gas reservoirs and/or coal systems) on various levels of detail. These correspond to degrees of confidence in estimating storage capacity. Country or state screening involves identification of appropriate sedimentary basins which can then be screened and ranked as to their overall suitability for CO2 storage. Evaluation of the size and thickness of the basin gives an indication of total pore volume it may hold. More detailed basin assessment allows the estimation of the prospective storage capacity in each of the identified storage systems and trap types. Site characterisation is the most time-consuming and costly part of the CO2 storage site selection process! Site characterisation typically involves collection and analysis of more detailed information than basin assessment investigations and may involve re-evaluation of regional geology and generation of new data and/or updating of existing data such as static (geologic and seismic) and dynamic (flow simulation and injection) data. The level of detail used in site characterisation allows the estimation of contingent pore capacity. The ultimate goal of a storage project is commercial site deployment, which requires all the geological, engineering, economic and regulatory considerations of a site being taken into account. Site deployment therefore requires estimation of operational storage capacity. The various scales of site selection and the different levels of storage capacity estimation need to be integrated and the processes for these analyses need to be iterative. The natural variability and geological, engineering and economic complexity of any potential CO2 storage site means that each site needs to be assessed individually; however, a similar workflow can be applied to all site evaluations. This report attempts to provide a consistent and systematic set of methodologies that can be used in assessing and classifying CO2 storage volumes of potential storage sites in Australia, New Zealand, and elsewhere.

Introduction
The determination of carbon dioxide (CO2) storage capacity and the selection and characterisation of potential sites for CO2 storage are key issues in taking carbon capture and storage (CCS) forward. The CO2CRC (and its predecessor APCRC GEODISC Program) has been involved in both of these issues for many years, as have a number of other organisations. The Special Report on CCS by the Intergovernmental Panel on Climate Change (IPCC) concluded that current storage capacity estimates are imperfect and that there is a need for more development and agreement on assessment methodologies (IPCC, 2005). The Carbon Sequestration Leadership Forum (CSLF) convened a Task Force to review and identify standards for CO2 storage capacity estimation (CSLF, 2005, 2007). The Society of Petroleum Engineers (SPE), together with the World Petroleum Council (WPC), the American Association of Petroleum Geologists (AAPG) and the Society of Petroleum Evaluation Engineers (SPEE), has recently developed a Petroleum Resources Management System for the classification of petroleum resources and reserves (SPE, 2007), which is obviously relevant to calculating equivalent resources and reserves for CO2 storage capacity. However, as yet there is no single agreed scheme in Australia or internationally on which industry or the financial sector can confidently base commercial (bankable) CO2 storage projects. In Australia, the APCRC GEODISC project evaluated CO2 storage resources from a broad regional perspective, as well as conducting four detailed site assessments. The regional study did not determine storage capacity in the commercial sense (i.e. analogous to reserves in the petroleum and mining industries). Rather, it was designed to assess whether the potential geological storage capacity of Australia was likely to be large enough to be important to Australia as a mitigation option. The answer was a very clear yes at the national level; however, the economic considerations were derived in part from assumptions about the potential CO2 sources and therefore not of sufficient detail to enable true reserve-style estimations of storage capacity. To provide this requires far more detailed work. The work which commenced within the GEODISC project has developed further within the CO2CRC research program, with more detailed assessment and characterisation in specific basins, such as the Gippsland Basin (e.g. Hooper et al., 2005). Such studies are data intensive and require a wide range of skills. This in itself presents a challenge, as there are a great number of basins yet to be assessed but there are a limited number of skilled people and often incomplete information to assess the basins. A well-defined assessment scheme is needed to maximise the efficiency and effectiveness in looking at potential CO2 storage sites. Equally important, there is a need for consistency in data compilation, interpretation and modelling (to the extent that this is possible, given the usual variability in the extent and quality of geological data), to enable comparisons between sites on an equal footing. Related to this is the need to ensure consistency in selecting storage sites and/or determining storage capacity across state boundaries, between offshore and onshore, between Australia and New Zealand, or globally. This comes into even sharper focus when the implications of emissions trading are considered. In this light, CO2CRC convened a workshop of a wide range of stakeholders from industry, government and the research community to consider methodologies for CO2 storage capacity estimation and storage site selection and characterisation, with the objective of developing a system that CO2CRC can apply on a consistent basis to all its studies throughout Australia, New Zealand and elsewhere. It is important to emphasise that CO2CRC has utilised existing schemes if possible (so as to avoid unnecessary duplication of effort) and adapted them where necessary, to develop a consistent and readily useable methodology that will ultimately deliver the basis for commercial storage projects in an economical, credible and timely fashion. This report presents the CO2CRCs approach to estimating CO2 storage capacity and storage site selection and characterisation. In making it widely available to stakeholders, CO2CRC aims to encourage discussion on the scheme (and the issues) presented here.

Geological Storage of Carbon Dioxide

Geological storage of carbon dioxide is the process whereby CO2 captured and separated from a source (such as a high- CO2 natural gas field, liquefied natural gas (LNG) or mineral processing plant, or coal-fired power station) is transported and injected into the geological subsurface for long-term storage (Cook et al.,

2000; IPCC, 2005). Conventional geological constraints on finding the right place to store CO2 include having a porous and permeable reservoir rock (e.g. sandstone) to allow injection and storage of the CO2, overlain by an impermeable seal rock (e.g. claystone) to retain the injected CO2 in the geological subsurface (van der Meer, 1992; Bachu et al., 1994; Rochelle et al., 1999). However, recent studies suggest that due to the behaviour of CO2 in the subsurface, additional mechanisms may also be effective for long term safe storage of CO2. These are discussed below.

CO2 Subsurface Behaviour and Trapping Mechanisms

CO2 storage in a deep saline aquifer or a depleted petroleum reservoir will preferentially be injected as a supercritical fluid. The critical point where CO2 enters the supercritical phase is defined as 31.1C and 7.38 MPa (Holloway & Savage, 1993; van der Meer, 1993; Bachu, 2000). Based on worldwide average geothermal and hydrostatic pressure conditions, this equates to an approximate minimum subsurface depth of about 800 m (Figure 1) (van der Meer, 1992; Holloway & Savage, 1993). In its supercritical form CO2 is much denser than gaseous CO2 and therefore a greater volume of CO2 can be stored in the pore space available (Figure 2); (Holloway & van der Straaten, 1995; Cook et al., 2000) Below this depth (under normal sedimentary basin conditions), supercritical CO2 is 3040 % less dense than a typical saline formation water under the same conditions (Ennis-King & Paterson, 2001, 2002). This means that the lighter CO2 will naturally rise upwards by buoyancy through the reservoir rock until trapped by various physical, hydrodynamic or geochemical trapping mechanisms (although in the longer term hundreds to thousands of yearsthe CO2 will sink as it goes into solution (Ennis-King & Paterson, 2005a).

Figure 1: Estimated depths to critical temperature and pressure for CO2 (from Holliday et al., 1991)

Figure 2: Increasing storage effectiveness for CO2 with depth. Note that above critical depth, CO2 is in gaseous state (balloons); below critical depth it is in liquid-like state (droplets). Volumetric relationship shown by blue numbers (eg 100 m3 of CO2 at surface would occupy 0.32 m3 at a depth of 1 km) (CO2CRC; modified from IPCC, 2005)

Adequate containment of stored CO2 for sufficiently long periods of time (enough that it can be considered to be safely removed from atmospheric circulation) is a critical part of any assessment of potential for CO2 storage in the subsurface. CO2 can be stored by a number of different trapping mechanisms, such as: Physical trapping in structural or stratigraphic traps, where the free-phase CO2 is physically trapped by the geometric arrangement of the reservoir and seal rock units (in a similar manner to hydrocarbon accumulations, e.g. Biddle & Wielchowsky, 1994); Hydrodynamic trapping, where the dissolved and immiscible CO2 travels with the formation water for very long residence (migration) times (Bachu et al., 1994); Residual trapping, where the CO2 becomes trapped in the pore spaces by capillary pressure forces (Ennis-King & Paterson, 2001; Holtz, 2002; Flett et al., 2005); Solubility trapping, where the CO2 dissolves into the formation water (Koide et al., 1992); Mineral trapping, where the CO2 precipitates as new carbonate minerals (Gunter et al., 1993); and Adsorption trapping, where the CO2 adsorbs onto the surface of the coal (Gunter et al., 1997a).

The exact trapping mechanism will depend on the specific geological conditions and typically combines a combination of the above. Appendix A provides a more detailed technical discussion of the trapping mechanisms of CO2 in the subsurface.

Geological Storage Systems for CO2

Carbon dioxide can be stored geologically in a variety of different storage systems (Figure 3). Of these, the three main alternatives are: saline formations, depleted oil and gas fields, and coal systems (Bachu & Gunter, 1999; Cook et al., 2000; IPCC, 2005).

Figure 3: Options for the geological storage of carbon dioxide (from CO2CRC; IPCC, 2005).

Saline Formations
Saline formations are deep porous sedimentary rocks saturated with formation waters that are considered unsuitable for human consumption or agricultural or industrial use. They have been identified by many studies as one of the best potential options for large volume geological storage of CO2 (e.g. Bachu, 2000; Bradshaw et al., 2002). Supercritical CO2 can be effectively stored in deep saline formations because of its high density and high solubility in formation water at the relatively high formation pressures encountered. However, these formations are commonly less understood (in comparison to shallow freshwater aquifers or hydrocarbon bearing reservoirs) and any assessment of their CO2 storage potential typically includes significant uncertainty because of the lack or scarcity of subsurface data. In addition, the containment potential of the seal rock is usually untested and there is uncertainty regarding potential for undiscovered natural resources. However, their main advantages are that they are distributed widely over the world and

their potential storage capacity is large (Koide et al., 1992; Hendriks & Blok, 1993; Rigg et al., 2001; IPCC, 2005). CO2 can be stored in saline formations by a number of different trapping mechanisms, including structural/stratigraphic, hydrodynamic, residual, solubility and mineral trapping (see Appendix A for a detailed technical discussion on CO2 trapping mechanisms).

Oil and Gas Fields Depleted or Enhanced Recovery

CO2 can be geologically stored in oil and gas fields once they have been depleted and are no longer producing, or can be used to enhance oil or gas recovery (EOR/EGR) in fields that are still producing. The main advantages of storage in depleted oil and gas fields over saline formations is that the containment potential of the site has been proven by the retention of hydrocarbons for millions of years and there are typically large amounts of geological and engineering data available for detailed site characterisation (Holloway & Savage, 1993; IPCC, 2005). Possible drawbacks, however, may be the physical size of the structural/stratigraphic trap (i.e. potential storage volume may be limited), the possibility that pore pressure depletion has led to pore collapse (which will reduce the potential storage volume), the presence of existing old wells which may provide potential leak points, and the timing of availability of depleted fields with respect to the source of CO2 (Bradshaw & Rigg, 2001; Bradshaw et al., 2002; Celia & Bachu, 2003; Streit & Siggins, 2005). In EOR or EGR, the CO2 is used to incrementally increase the amount of hydrocarbons extracted by either immiscible or miscible flooding, thus providing an economic benefit whilst also storing CO2. As with depleted oil and gas fields, the potential storage capacity may be limited due to the physical size of the field and also due to EOR operational issues such as the rate at which the CO2 is recycled (Islam & Chakma, 1993; Cook et al., 2000; IPCC, 2005).

Coal Seams Deep Unmineable or Enhanced Methane Recovery

CO2 storage in coal seams is very different to storage in saline formations or oil and gas fields, as the trapping mechanism is by adsorption as opposed to storage in rock pore space. CO2 is preferentially adsorbed onto the coal micropore surfaces, displacing the existing methane (CH4) (Gunter et al., 1997a; Bradshaw & Rigg, 2001; IPCC, 2005). In contrast to saline/hydrocarbon formations, storage density (i.e. storage capacity) is greatest in coals at depths less than 600 m, when CO2 is in the gaseous phase, not supercritical (Figure 4) (Ennis-King & Paterson, 2001). CO2 can be geologically stored in coal seams that are considered economically unmineable or can be used to enhance coal seam methane recovery (ECSM). Since coals have a higher adsorption affinity for CO2 than for CH4 (Killingley, 1990), CO2 injection in coal, coupled with CSM production, is potentially an attractive option for CO2 storage. In reality, all CO2 coal storage projects must be in conjunction with an ECSM recovery program, both to create permeability so that the CO2 can be injected and the space for the CO2 to be stored. It is also vital that the methane released from the coal matrix does not become an emission to the atmosphere as it has a higher greenhouse radiative effect (21 times stronger by weight) than CO2. The CH4 therefore needs to be captured to ensure a net greenhouse emission mitigation outcome (CSLF, 2007). Technical challenges for CO2 storage in coal seams focus around the feasibility of injecting the CO2, due to the typically low permeability characteristics of the coal cleat system (especially with increasing depth). Coal sorption and permeability varies with coals of different character (rank, grade and type). It is also dynamic and changes during both gas extraction (matrix shrinkage) and injection (matrix swelling) as the coal interacts with the CO2. In addition, the economic viability of enhanced methane recovery can potentially be compromised due to the large number of wells that may need to be drilled to overcome injectivity issues relating to low permeabilities (Gunter et al., 1997a; Bradshaw & Rigg, 2001; IPCC, 2005). It can also be compromised if the price of coal increases to such an extent that coal beds previously regarded as uneconomic to mine suddenly become economically mineable. Research into CO2 storage in coal is still at quite an early stage and further work needs to be conducted to fully understand the processes involved and the most suitable coal characteristics for CO2 storage (IPCC, 2005).

Figure 4: Total storage density as a function of depth, highlighting how the storage density of CO2 adsorbed onto coal at subcritical depths is comparable to the storage density of CO2 captured in pore space at supercritical depths (hydrostatic pressure gradient is 10.5 MPa, mean surface temperature is 15C and geothermal gradient is 25C/km unless noted) (from Ennis-King & Paterson, 2001).

Storage Capacity
CO2 storage capacity is an estimate of the amount of CO2 that can be stored in subsurface geologic formations. Because of uncertainties inherent to subsurface evaluation, exact quantification of geological properties is not possible and therefore storage capacity is always at best an approximation of the amount of CO2 that can be stored. Storage capacity estimates therefore rely on the integrity, skill and judgment of the evaluator and are affected by the geological complexity, stage of exploration or development, amount of existing storage and of available data. Use of the definitions should sharpen the distinction between the various classifications and provide more consistent reporting. Factors affecting CO2 storage capacity include the density of the CO2 at subsurface reservoir conditions, the amount of interconnected pore volume of the reservoir rock and the nature of the formation fluids. Due to the flow behaviour of CO2 in the subsurface, not all potentially available pore volume of the reservoir will become occupied during injection and migration, with flow preferentially occurring either upward due to buoyancy forces or laterally below low permeability zones (i.e. spreading out in thin layers beneath intraformational seals or the regional top seal rather than filling the entire pore volume). This can make CO2 storage capacity volumes difficult to calculate, particularly in the reservoir rocks underlying defined structural or stratigraphic closures, where much of the available rock pore volume can be bypassed by CO2 preferentially utilising higher permeability zones (Gibson-Poole, 2008). The potential CO2 storage capacity should therefore be assessed in terms of available interconnected pore space, accounting for factors such as injection rate, rate of CO2 migration, the dip of the reservoir, the heterogeneity of the reservoir and the potential for fill-to-spill structural closures encountered along the migration path. In addition, long-term prospects for storage, including residual trapping, dissolution into the formation water or mineral trapping (formation of new minerals) can also be considered (especially for estimating potential storage volume within deep saline formations) (Gibson-Poole, 2008). Such issues are best addressed by building geological models and running numerical flow simulations to test the importance of the various factors inherent to each specific site. The numerical flow simulations can give a more accurate assessment of how much of the available pore volume is actually used (sweep efficiency) in each particular type of trapping (Gibson-Poole, 2008).

While there have been previous attempts to assess CO2 storage capacity, most have used a range of approaches and methodologies, and data sets of variable size and quality, resulting in widely varying storage capacity estimates of inconsistent quality and reliability. Of the various methods to estimate storage capacity, few are constrained by commercial or engineering feasibility limitations. Most workers, e.g. CSLF (2007) and DOE (2006) have defined storage capacity as including the total pore space (volume) available for CO2 storage, without any economic qualifier.

Previous Studies on Storage Capacity Estimation: DOE & CSLF

There are two major works providing methodologies for the estimation of storage capacity of CO2 in geological formations. These are the DOE (2006) Methodology for Development of Carbon Sequestration Capacity Estimates prepared for the Carbon Sequestration Program of the National Energy Technology Laboratory, U.S. Department of Energy (DOE), by the Capacity and Fairways Subgroup of the Geologic Working Group of the DOE Regional Carbon Sequestration Program in December, 2006, and the CSLF (2007) Estimation of CO2 Storage Capacity in Geological Media Phase II prepared by the Task Force on CO2 Storage Capacity Estimation for the Technical Group (TG) of the Carbon Sequestration Leadership Forum (CSLF) in April, 2007. Both the CSLF (2007) and the DOE (2006) studies use the term storage capacity in its widest sense to cover all the categories referred to within this work. In effect, their use of the term Storage Capacity does not incorporate a commercial perspective and thereby corresponds to this studys broadest category of Total Pore Volume. CSLF (2007) refer to CO2 storage capacity as a geological resource, whose availability can be expressed in the same manner as resources and reserves are classified in other commodities (e.g. oil and gas, gold, uranium, iron, coal, etc.). The CSLF (2007) study provides a Techno-Economic ResourceReserve Pyramid for CO2 storage capacity in depleted oil and gas fields, based on earlier concepts of Masters (1979) and McCabe (1998), whereby the degree of geological and economic uncertainty associated with a capacity estimate is represented by its place on the pyramid. For this reason CSLF (2007) suggest that it is essential that, when an estimate of storage capacity is performed, the type of estimate and its position in the resource pyramid are specified. In addition, both studies include equations for volumetric calculations of pore volume (which is referred to as storage capacity). The DOE (2006) study provides volumetric equations for capacity calculation in each of the three main storage systems (saline formations, oil and gas reservoirs, and coal seams) and also uses a Monte Carlo approach to estimate expected uncertainty in the efficiency of the storage capacity from a combination of trapping mechanisms. CSLF (2007) provides individual equations for each of the trapping mechanisms but the coefficients for storage capacity efficiency have not yet been determined. Where the DOE (2006) use one equation with one storage efficiency factor to account for all the corrections such as net-to-gross, gravity effects, etc. (and has provided values for the efficiency factor), CSLF (2007) use different equations for each trapping type. At present no efficiency factors are provided, so the calculations cannot yet be applied for commercial (bankable) projects. Should a future CSLF methodology resolve the storage efficiency factors, then perhaps it will be the more viable option, but until then the use of the DOE (2006) methodology, which already includes efficiency factors, is recommended. For this study, CO2CRC has incorporated the resource pyramid from CSLF (2007) but adapted it to take account of the SPE Petroleum Resources Management System for the classification of petroleum resources and reserves (SPE, 2007). It is interesting to note that the CSLF (2007) study actually recommends integrating their classifications with established schemes like the SPE. In addition, the current study uses the approach suggested by DOE (2006) for deterministic storage capacity estimates in saline formations, oil and gas fields, and coal seams. Thus, an alternative set of CO2 storage capacity classifications and definitions, based on amalgamations of the DOE (2006) and CSLF (2007) plus simple revisions of the SPE Petroleum Resources Management System (SPE, 2007), are proposed here. Following the SPE definitions, greater emphasis is put on the pore space in which it is economically and technically feasible to store CO2 either under current or anticipated future conditions. Others have also advocated application of the SPE Petroleum Resources Management System to estimate storage capacity (Frailey et al., 2006).

CO2CRC Storage Capacity Classification (modified from SPE, 2007)

The proposed CO2CRC CO2 storage capacity classification system is summarised in Figure 5 and the relevant definitions are given below. Total Pore Volume. Total Pore Volume is defined here as the entire volume which is estimated to exist in sedimentary basins, naturally occurring storage systems or potential storage sites, plus that volume which may already be in use to store CO2, plus the estimated pore volume yet to be identified. Total Pore Volume may be subdivided into Discovered Pore Volume and Undiscovered Pore Volume, with Discovered Pore Volume being limited to known ( i.e. well characterised) storage sites. Discovered Pore Volume. Discovered Pore Volume is defined here as that volume which is estimated, on a given date, to remain in known storage sites, plus that volume already used for storage. Discovered Pore Volume may be subdivided into Commercial and Sub-commercial categories, classified as Commercial Storage Capacity and Contingent Storage Capacity respectively, as defined below. Operational Storage Capacity. Operational Storage Capacity is defined here as an estimate of that volume of pore space which will be technically and commercially available for injecting CO2 into known storage sites from a given date forward. Storage capacity is separated into proved (1P), proved plus probable (2P) and proved plus probable plus possible (3P) categories following petroleum industry standards (SPE, 2007). Proved (1P) Operational Storage Capacity will be limited to that which is commercial under current techno-economic conditions, while Probable (2P) and Possible (3P) Operational Storage Capacity may be based on anticipated future techno-economic conditions. In general, pore volume should not be classified as Operational Storage Capacity unless there is an expectation that the storage site will be developed and used to store CO2 within a reasonable timeframe. The definitions of commercial and technically feasible for a storage site will vary according to local conditions and circumstances and is left to the discretion of the operator or jurisdictional (country/state) authority concerned. In general, Proved (1P) operational storage capacity is the subset of storage capacity applicable to bankable storage projects. Contingent Storage Capacity. Contingent Storage Capacity is defined here as that quantity of pore space which is estimated, on a given date, to be potentially technically and economically feasible for CO2 injection into known storage sites based on anticipated future techno-economic conditions, but which is not currently considered to be commercially viable. In certain circumstances, Operational Storage Capacity may be assigned even though development may not occur for some time. An example of this would be where storage sites are dedicated to a long-term CO2 supply contract and will only be developed as and when they are required to satisfy that contract. Contingent Storage Capacity may include, for example, storage sites for which there is currently no commercially viable CO2 injection project, or where commercial injection is dependent on the development of new technology, or where regulatory frameworks are not yet in place, or where evaluation of the storage site is still at an early stage. It is recognised that some ambiguity may exist between the definitions of Contingent Storage Capacity and Unproved (P2 and P3) Operational Storage Capacity. It is recommended that if the storage site is not expected to be developed and have CO2 injection within a reasonable timeframe, the estimated injectable storage volume for the site be classified as Contingent Storage Capacity. Undiscovered Pore Volume. Undiscovered Pore Volume is defined here as that pore volume which is yet to be discovered, but is estimated to be available for storage at some future date after discovery. The estimated Undiscovered Pore Volume is classified as Prospective Storage Capacity, as defined below. Prospective Storage Capacity. Prospective Storage Capacity is defined here as that quantity of pore space into which it is estimated, on a given date, that CO2 will be technically and economically potentially injectable into as yet undiscovered storage sites.

10

Aggregation. Quantities of pore volume classified as Storage Capacity, (Operational Storage Capacity, Contingent Storage Capacity or Prospective Storage Capacity) should not be added to one another (aggregated) without due consideration of the significant differences in the criteria associated with their classification. In particular, there may be a significant risk that storage sites containing Contingent Storage Capacity or Prospective Storage Capacity will not be commercially viable. Estimated Ultimate Injection. Estimated Ultimate Injection (EUI) is not a category as such, but a term which may be applied to an individual storage site of any status/maturity (discovered or undiscovered). Estimated Ultimate Injection is defined here as that quantity of CO2 which is estimated, on a given date, to be technically and economically injectable into a storage site, plus those quantities of CO2 which may already be stored therein. Range of Uncertainty. The Range of Uncertainty, as shown in Fig. 5 reflects a reasonable range of estimated potentially injectable pore volumes for CO2 at a specific storage site. Any estimation of capacity at a storage site is subject to both technical and commercial uncertainties, and should, in general, be determined in a probabilistic manner and quoted as a range. In the case of Contingent and Prospective Storage Capacity the terms Low Estimate, Best Estimate and High Estimate are recommended. The term Best Estimate is used here as a generic expression for the estimate closest to the quantity that will actually be injected into the storage site between the date of the estimate and the time of abandonment. If probabilistic methods are used, this term would generally be a measure of the central tendency of the uncertainty distribution (most likely/mode, median/P50 or mean). The terms Low Estimate and High Estimate should provide a reasonable assessment of the range of uncertainty in the Best Estimate. For undiscovered pore volume (Prospective Storage Capacity) the range will, in general, be substantially greater than the ranges for discovered pore volume (at specific storage sites). In all cases, however, the actual range will be dependent on the amount and quality of data (both technical and commercial) which is available for that storage site. As more data become available for a specific storage site (e.g. additional wells, reservoir performance data) the range of uncertainty in EUI for that storage site will be reduced. If probabilistic methods are used, these estimated quantities should be based on methodologies analogous to those applicable to the 1P-2P-3P categories of Operational Storage Capacity. Therefore, in general, there should be at least a 90 % probability that, assuming the storage site is actually developed, the quantities actually injected will equal or exceed the Low Estimate. In addition, an equivalent probability value of 10 % should, in general, be used for the High Estimate. As one possible example, consider a storage site that is currently not commercial because there is not yet a viable injection project. The estimated injectable storage volume is classified as Contingent Storage Capacity, with Low, Best and High estimates. Where a project is subsequently developed, and in the absence of any new technical data, the storage site moves up into the Operational Storage Capacity category and the Proved Operational Storage Capacity estimate would be expected to be close to the previous Low Estimate of Contingent Storage Capacity. Figure 6 is a graphical representation of the definitions detailed above. The horizontal axis represents the range of uncertainty in the estimated potentially injectable storage volume for a storage site, whereas the vertical axis represents the level of status/maturity of the storage site. The use of the proposed definitions allows the subdivision of each storage category, further using the vertical axis to classify storage sites on the basis of their commerciality, thereby enabling decisions required to move a storage site towards injection.

11

Figure 5: CO2 storage volume classification system proposed by CO2CRC (modified from SPE, 2007).

Oper ational Stor ag e C ap acity

Contingen Storage Cap t acity

Prospective Storage Cap aci ty

Increasing Certainty Decreasing Storage Volume

Total Pore V olume

Figure 6: CO2 storage capacity pyramid (modified from CSLF, 2007).

12

Methodology for Storage Capacity Estimation

Storage Capacity Estimation in Saline Formations
DOE (2006) provides a relatively simple volumetric equation for the calculation of CO2 storage capacity in saline formations (Equation 1 and Table 1) based on the concept that CO2 occupies the pore space (or parts of it) within a permeable rock: GCO2 = A hg
tot

(1)

Table 1: Volumetric equation parameters for capacity calculation in saline formations (from DOE, 2006).

Parameter GCO2
A hg
tot

Units* M
L2 L L3/L3 M/L3

L3/L3

Description Mass estimate of saline-formation CO2 storage capacity Geographical area that defines the basin or region being assessed for CO2 storagecapacity calculation Gross thickness of saline formations for which CO2 storage is assessed within the basin or region defined by A Average total porosity of entire saline formation over thickness hg Density of CO2 evaluated at pressure and temperature that represents storage conditions anticipated for a specific geologic unit averaged over the depth range associated with hg CO2 storage efficiency factor that reflects a fraction of the total pore volume that is filled, or contacted, by CO2 (see Table 2)

* L is length; M is mass

The storage efficiency factor (E) adjusts total gross thickness to net gross thickness, total area to net area and total porosity to effective (interconnected) porosity actually containing CO2. Without E, Equation 1 presents the Total Pore Volume or maximum upper limit to capacity. Inclusion of E provides a means of estimating storage volume for a basin or region with the level of knowledge (uncertainty) in specific parameters determining the type of CO2 storage capacity estimated. A reasonable maximum Prospective Storage Volume may be estimated by assuming CO2 injection wells may be placed regularly throughout the basin, or region, and multiplying the storage efficiency terms in Table 2 together to determine the storage efficiency factor for the basin or region. In Table 2, the first three terms An/At, hn/hg and e tot are used to define the regional effective pore volume, and the next three terms EA, EI and Eg are used to define the fraction of that volume accessed by CO2 from injection wells. The remaining term Ed in Table 2 accounts for the proportion of the effective pore volume filled with CO2. Ideally, the last four terms would be determined for each potential well site and combined to estimate the total storage volume; however, practically it may be easier to determine the average volume of pore space accessed by injecting CO2 into a single well and then multiplying by the expected number of wells.

Table 2: Terms included in storage efficiency factor for saline formations (from DOE, 2006).

Term

Net to total area Net to gross thickness Effective to total e tot porosity ratio Terms used to define the fraction of pore volume accessed by CO2 from injection wells

Symbol Description Terms used to define the entire basin/region pore volume An/At Fraction of total basin/region area that has a suitable formation present Fraction of total geologic unit that meets minimum porosity and hn/hg permeability requirements for injection Fraction of total porosity that is effective, i.e. interconnected

13

Term
Areal displacement efficiency Vertical displacement efficiency

Symbol
EA

EI

Gravity Microscopic displacement efficiency

Eg

Ed

Description Fraction of immediate area surrounding an injection well that can be contacted by CO2; most likely influenced by areal geologic heterogeneity such as faults or permeability anisotropy Fraction of vertical cross section (thickness), with the volume defined by the area (A) that can be contacted by the CO2 plume from a single well; most likely influenced by variations in porosity and permeability between sub layers in the same geologic unit. If one zone has higher permeability compared with others, the CO2 will fill this one quickly and leave the other zones with less CO2 or no CO2 in them Fraction of net thickness that is contacted by CO2 as a consequence of the density difference between CO2 and in situ water. In other words, (1-Eg) is that portion of the net thickness not contacted by CO2 because the CO2 rises within the geologic unit Portion of the CO2-contacted, water-filled pore volume that can be replaced by CO2. Ed is directly related to irreducible water saturation in the presence of CO2 or residual CO2 saturation, dependent on position within the plume

Saline formations generally have limited data available to assess storage capacity volumes, and additional information and data will inevitably be required. In addition, there are a number of different trapping mechanisms for geological storage of CO2 in saline formations. The specific mechanism needs to be defined at either the basin-scale assessment or site characterisation part of site selection (discussed later), as there will be variations in the storage volume and storage volume assessment method for each trap type. These trap types include: (1) structural/stratigraphic trapping; (2) hydrodynamic trapping; (3) residual trapping; (4) solubility trapping; and (5) mineral trapping (see Appendix A for a detailed technical discussion).

Storage Capacity Estimation in Depleted Oil and Gas Fields

CO2 storage capacity estimations in depleted (or near depleted) oil and gas fields are generally easier than estimates for coal seams or saline formations because there is typically a greater amount of data associated with oil and gas fields and hence they are better characterised. Also, unlike coal seams and saline formations, oil and gas fields are considered as a single discrete system. This means that estimates of CO2 storage volume in oil and gas fields can either be based on the effective pore space volume (as determined for saline formations), the calculated original oil and gas in place or from the volume of oil and gas produced from the field. In general, storage in depleted oil and gas fields is based initially on two primary assumptions: (1) the volume previously occupied by the produced hydrocarbons will become available for CO2 storage; and (2) the existing caprock seal will also contain the CO2 provided the pressure does not increase above the original reservoir pressure prior to production. The first assumption is generally valid for fields that were not in hydrodynamic contact with an aquifer, or that were not flooded during secondary and tertiary oil recovery (i.e. pressure-depleted fields) (CSLF, 2007). The invasion of pore space by formation waters during oil and gas production leads to an increase in the water saturation to balance the residual hydrocarbon saturation. When subsequently introducing CO2, each phase may have a relative permeability such that the space previously occupied by oil or gas may no longer be available for CO2 storage. This may be due to wettability, capillarity, viscous fingering and gravity effects (Stevens et al., 2001). The relative permeability for a non-wetting fluid during drainage is different from that during imbibition (e.g. Bennion & Bachu, 2005; Juanes et al., 2006). It should also be noted that the original hydrocarbon pool was filled over geological time. When re-filling the trap with CO2 over a period of a few years, lower permeability parts of the reservoir might not be accessible if a limit on injection pressure is maintained.

14

The second assumption is reasonable provided that the reservoir behaves elastically during pressure depletion and subsequent re-pressurisation, such that the reservoir stress path is reversible. Nonetheless, pressure is often the limiting factor on rate of injectivity, and the constraint imposed by not increasing pressures above virgin (pre-production) reservoir pressures may limit the volumes able to be injected over the life of the project, or increase costs associated with having to drill a larger number of injection wells. In estimating the CO2 storage capacity for depleted oil and gas fields it is assumed, as stated above, that the volume of hydrocarbons produced is replaced by an equivalent volume of CO2, where both hydrocarbon and CO2 volumes are calculated at initial formation pressure or at a pressure that is considered a maximum for CO2 storage. Two alternative methods to that used for the saline formation estimate (which could also be applied to depleted oil and gas reservoirs) are presented, based on the DOE (2006) approach to estimate the CO2 storage volume: (1) a volumetrics-based CO2 storage volume estimate; and (2) a production-based CO2 storage volume estimate. Any additional storage capacity, such as that associated with residual saturation beneath the original hydrocarbon-water contact, should be determined using Equation 1. Volumetrics-Based CO2 Storage Capacity Estimate for Depleted Oil and Gas Fields The volumetrics-based CO2 storage capacity estimate (Equation 2 and Table 3) uses standard industry methods to calculate original oil in place (OOIP) or original gas in place (OGIP). A bulk rock volume is calculated by multiplying together the reservoir area (A), net oil column height (hn) and a geometric factor which accounts for the geometry of the trap (i.e. 4-way dip structural versus stratigraphic). An average effective porosity ( e) in combination with the bulk rock volume (A hn g) provides an estimate of the pore space available for storage and the storage efficiency factor (E) provides a measure of the fraction of total pore volume from which oil and/or gas has been produced and that can be filled by CO2. GCO2 = A hn g
e

(2)

Table 3: Volumetric equation parameters for storage capacity in oil and gas reservoirs (from DOE, 2006).

Parameter GCO2
A hn g
e

Units* M
L2 L L3/L3 L3/L3 M/L3

L3/L3

Description Mass estimate of hydrocarbon reservoir CO2 storage capacity Area that defines oil or gas reservoir that is assessed for CO2 storage capacity calculation Net hydrocarbon column height in the reservoir Geometric factor based on the trap type Average effective porosity over net thickness hn Density of CO2 evaluated at pressure and temperature that represents storage conditions in the reservoir averaged over hn CO2 storage efficiency factor that reflects a fraction of the total pore volume from which oil and/or gas has been produced and that can be filled by CO2

* L is length; M is mass

The storage efficiency factor (E) should account for irreducible water saturation, as well as an estimate of the remaining irreducible hydrocarbon saturation and hydrocarbon recovery factor. More complex mechanisms, such as dissolution of CO2 into the oil and/or water and saturation hysteresis related to reservoir drive processes, may also be included in the storage efficiency factor. In some situations, compositional simulation models can be used to estimate the volume of CO2 able to be stored per stock tank barrel of original oil in place. In such cases Equation 2 should be modified to Equation 3 as follows: GCO2 = A hn g
e

1/Bo (1 Sw Soirr)

(3)

where Bo is the oil shrinkage factor, Sw is the average water saturation, Soirr is the average irreducible oil saturation within the gross rock volume and V is the volume of CO2 able to be stored per stock tank barrel of original oil in place (scfCO2/stbOOIP).

15

Production-Based CO2 Storage Capacity Estimate for Depleted Oil and Gas Fields Provided suitable records of produced volumes are available for a field, then a production-based CO2 storage estimate is also possible. CSLF (2007) present equations relating to production-based calculations of capacity in oil and gas reservoirs and some discussion on the approach is provided in DOE (2006). In general, where data are available, it is possible to apply efficiency to production data to convert it to CO2 storage volumes; otherwise replacement of produced hydrocarbons by CO2 on a volume-for-volume basis (at reservoir pressure and temperature) is recommended. Consideration should be given to any produced or injected water (due to water-flooding), to the presence of solution gas volumes if gas production in an oil reservoir is included, possible miscibility of CO2 into oil, dissolution of CO2 into residual and associated water, mineral trapping, pressure variations within the formation during production and pressure decline as a result of production. Finally, it is worth noting that a primary containment risk in CO2 storage sites will be the integrity of wells penetrating the storage volume and this will need to be included in any assessment of injection pressures and final CO2 storage volume.

Storage Capacity Estimation in Coal Seams

Gas storage in coal seams is different to storage in oil and gas reservoirs or saline formations, as the trapping mechanism is mainly by adsorption onto the coal medium as opposed to storage in rock pore space (see Appendix A for detailed technical discussion). Hence, the assessment of coal seam storage capacity requires additional knowledge of a coals adsorption capacity at a given depth and temperature, and it will vary depending on the quality (rank, grade and type) of the coal. Moreover, competition for access and utilisation of the coal resource for mining, coal seam gas extraction or in situ gasification must also be considered to ensure that coals are not sterilised for future use. It should be noted that CO2 storage in coal seams is a technology that is only in the demonstration phase (IPCC, 2005), the commercial success (or failure) of which will affect the application and evaluation of its capacity. In summary, the basic parameters that describe CO2 storage capacity and injectivity in coals are seam thickness, adsorption capacity, and permeability.
DOE (2006) provide a volumetric equation for CO2 storage capacity in coal seams (Equation 4 and

Table 4): GCO2 = A hg C E (4)

Table 4: Volumetric equation for storage volume estimation for CO2 storage in coal seams (from DOE, 2006).

Parameter GCO2
A hg C

Units* M
L2 L L3/L3 M/L3

L3/L3

Description Mass estimate of CO2 storage capacity of one or more coal seams Geographical area that outlines the coal basin or region for CO2 storage capacity calculation Gross thickness of coal seam(s) for which CO2 storage is assessed within the basin or region defined by A Concentration of CO2 standard volume per unit of coal volume (adsorption capacity at a given pressure or depth as determined by Langmuir volume or alternative); assumes 100% CO2-saturated coal conditions; if presented on dryash-free (daf) basis, then A and h must be corrected for daf Density of CO2 under (PT) conditions present in coal seam CO2 storage efficiency factor that reflects a fraction of the total coal bulk volume that is contacted by CO2

* L is length; M is mass; P is pressure; T is temperature

16

The value E (CO2 storage efficiency factor) has several components that reflect CO2 not being able to contact 100% of the bulk coal volume of a given basin or site. Depending on the area, thickness, and CO2 concentration (from Langmuir or others), the CO2 storage efficiency factor may reflect the volumetric difference between bulk volume and coal volume (DOE, 2006). For example, if A and h are based on dryash-free (daf) conditions, C must have a daf basis too. Additionally, because gross thickness is used in the equation above, E includes a term that adjusts gross thickness to net thickness (DOE, 2006). Parameters used to estimate E are presented in Table 5 (DOE, 2006). The determination of E is one of the many challenges facing coal storage volume estimations. Underground coal mine drainage and commercial gas production have demonstrated the heterogeneity and high lateral variability in the reservoir properties of coal seams. These same factors will impact directly upon CO2 storage behaviour.
Table 5: Terms included in storage efficiency factor for coal seams (from DOE, 2006).

Term

Net to total area

Net to gross thickness Terms used to define the fraction of pore volume accessed by CO2 from injection wells Fraction of immediate area surrounding an injection well that can be Areal displacement contacted by CO2; most likely influenced by areal geologic EA efficiency heterogeneity such as faults or permeability anisotropy Fraction of the vertical cross section (thickness), with the volume Vertical defined by the area (A) that can be contacted by a single well; most displacement EI likely influenced by variations in the cleat system within the coal. If one efficiency zone has higher permeability than others, the CO2 will fill this one quickly and leave the other zones with less CO2 or no CO2 in them Fraction of the net thickness that is contacted by CO2 as a consequence of the density difference between CO2 and the in situ water. In other Gravity Eg words, (1-Eg) is that portion of the net thickness not contacted by CO2 because the CO2 rises within the coal seam Reflects the degree of saturation achievable for in situ coal compared Microscopic with the theoretical maximum predicted by the CO2 Langmuir Isotherm displacement Ed efficiency CSLF (2007) also suggest that once a site or region of possible coal seam storage has been established, the CO2 storage volume in the respective region can be estimated in a manner similar to that used for coal seam methane (CSM) prospects, taking into account the higher affinity of coal for CO2 than for CH4, (commonly 2:1 ratio, but higher in lower rank coals (Burruss, 2003)). It is this 2:1 exchange ratio that suggests the injection of CO2 into a production field will improve the recovery of CH4 (Gale and Freund, 2001). While conventional CSM is recovered by reservoir pressure depletion, which is an inefficient process that leaves about 2060% of gas in place, CO2 injection has been shown to boost recovery of coal seam methane-inplace by over 80% (van Bergen & Pagnier, 2001). The additional CH4 production would offset the costs of injection in addition to storing the CO2. It is this aspect, as well as the proximity of coal seams to many coal-fired power stations, that contributes to coals consideration as a viable CO2 storage option, but it is important to re emphasise that the technology has yet to be shown to be commercially viable.

Symbol Description Terms used to define the entire basin/region pore volume Fraction of total basin/region area that has a bulk coal present; used if known or suspected locations are within a basin/region outline where a An/At coal seam may be discontinuous. For example, in the Illinois Basin there are subregions within the basin where sand channels have incised and replaced coal. This situation can be handled through this term Fraction of total coal seam thickness that has adsorptive capability hn/hg

17

Site Selection and Characterisation

The selection of a suitable site for the storage of significant volumes of CO2 comprises mainly geological evaluation on progressively more and more detailed scales. The different levels of site assessment that can be undertaken range from an initial regional screening to very detailed site-specific characterisation. A workflow that summarises the different scales of site assessment is presented in Figure 7. Each level of detail progressively reduces uncertainty, but also results, typically, in a decrease of the storage volume. In addition, each level of detail in site selection requires greater effort, and increasing amounts and types of data, time and costs. This is schematically displayed in Figure 8. The correlation from characterisation scale level to storage volume classification level as displayed may not always be a one-to-one correlation, but rather represents the likely maximum level of storage capacity assessment achievable given the likely database available at that level of characterisation.

Country/State-Scale Screening
This level of CO2 storage evaluation represents the coarsest scale of assessment with the least site-specific detail. Such screening focuses on large geographic areas and evaluates the overall suitability of sedimentary basins for CO2 storage within a country or state before specific sites are identified and selected for further investigation. The following discussion is mostly extracted or modified from Gibson-Poole (2008). The methodology for country/state-scale screening is as follows: 1. Identify sedimentary basins. Within sedimentary basins there is potential for CO2 storage via filling of pore space (e.g. in sandstones and limestones) and/or adsorption onto coals. Areas of crystalline rocks or meta-sediments can generally be disregarded for CO2 storage due to their low storage capacity and injectivity potential. 2. Review characteristics of sedimentary basins. The geological, geographical and industrial characteristics of sedimentary basins can be used as screening criteria. An example is the basinscale criteria for CO2 storage developed by Bachu (2003), which includes factors such as tectonic setting, basin size and depth, intensity of faulting, hydrodynamic and geothermal regimes, existing resources and industry maturity. The criteria are discussed in more detail below. 3. Qualitatively or quantitatively rank sedimentary basins in order of suitability. Once data have been compiled on the characteristics of the sedimentary basins, they can be compared, contrasted and ranked for their suitability for CO2 storage. This can be done either qualitatively or quantitatively if scores are given for each criterion (e.g. as per Bachu, 2003). The ranking procedure is discussed in more detail below.

18

Figure 7: Site characterisation workflow for geological storage of CO2 (modified from Gibson-Poole, 2008).

19

Figure 8: Integrated scales of site assessment and storage capacity pyramid for geological storage of CO2 (modified from CSLF, 2007). The storage volume level assigned to the scale of characterisation represents the maximum likely assessment level achievable given the likely database available.

Screening Criteria
Table 6 is a modified version of the basin-scale criteria for CO2 storage developed by Bachu (2003) and documents the criteria that can be used to assess the suitability of sedimentary basins for geological storage of CO2. For each criterion, the classes are arranged from least favourable to most favourable from left to right across the table. The criteria relate to either the containment security, the volume of storage capacity achievable, or consider the economic or technological feasibility. The seismicity of a basin is an important consideration for containment risk. Seismically very active areas are prone to large earthquakes and have a potential risk for catastrophic escape or continuous leakage of CO2 into the atmosphere (Bachu, 2003). Seismic data are commonly readily available from national or state geological survey organisations (e.g. Geoscience Australia, USGS, BGS) and should be examined as part of the initial assessment. If seismic history data are not available for a basin then the basins present-day tectonic setting can be used as a proxy for this information. Globally, it is possible to make broad generalisations on tectonic stability. Tectonically very unstable areas include fore-, intra- and back-arc basins in subduction-related geodynamic environments such as in Southeast Asia along the western margin of the Pacific Ocean. Other areas of pronounced tectonic instability are basins in active rift zones such as in East Africa, and basins affected by major strike-slip movements such as the Western Southland basins in New Zealand or the Californian basins in the vicinity of the San Andreas Fault. The overall tectonic stability of foreland basins or foredeeps on the continental side of orogens can be considered as intermediate but varies from region to region depending on the impacting stresses and the lithospheres response to them. Again seismicity is a pivotal criterion for assessing the suitability of such basins for geological sequestration of CO2 but at the same time it is important to recognise that in some of theses seismically active areas there are large petroleum accumulations, suggesting that effective storage of CO2 can also be possible in such areas. More work is needed to assess the storage potential of such areas

20

Interior cratonic and Atlantic-type passive margin shelf basins are tectonically fairly stable and are considered the most suited for CO2 storage (Bachu, 2003). Minor tectonic activity in such basins probably results from continental-scale warping or thermal subsidence. Figure 9 shows the low to fairly moderate seismicity of Australia, containing only stable intracratonic and passive margin shelf basins, in contrast to the strong seismicity of Indonesia and New Zealand. Again, it is important to emphasise that this does not necessarily mean that seismically active areas should be eliminated from consideration; rather it means that the process of site characterisation has to be undertaken with greater focus on seismic processes and their potential impact.
Table 6: Criteria for screening sedimentary basins within a country or state for geological storage of CO2 (modified from Bachu, 2003).

Increasing CO2 Storage Potential Criterion 1

1 2 3 4 5 Seismicity (tectonic setting) Size Depth Faulting intensity Hydrogeology Very high (e.g. subduction) Very small (<1000 km2) Very shallow (<300 m) Extensive Shallow, short flow systems, or compaction flow Warm basin (>40C/km) Poor None Anthracite None None Unexplored Deep offshore Arctic Inaccessible None Very shallow (<300 m) Lignite Small Exploration Sub-arctic Difficult Minor Domes Medium Developing Shallow offshore Desert

Classes 2
High (e.g. syn-rift, strike-slip) Small (1000 5000 km2) Shallow (300800 m)

3
Intermediate (e.g. foreland) Medium (5000 25000 km2) Moderate Intermediate flow systems

4
Low (e.g. passive margin) Large (25000 50000 km2) Deep (>3500 m)

5
Very low (e.g. cratonic) Very large (>50000 km2) Intermediate (8003500 m) Limited Regional, long-range flow systems; topography or erosional flow Cold basin (<30C/km) Excellent Shallow (300800 m) Bituminous Beds Giant Super-mature Onshore Temperate Easy Extensive

6 7 8 9 10 11 12 13 14 15 16

Geothermal Reservoirseal pairs Coal seams Coal rank Evaporites Hydrocarbon potential Maturity Onshore/ offshore Climate Accessibility Infrastructure

Moderate (30 40C/km) Intermediate Deep (>800 m) Subbituminous Large Mature Tropical Acceptable Moderate

21

Figure 9: Australia, with its intracratonic and passive margin shelf basins, exhibits low to moderate earthquake activity, in contrast to the high frequency of earthquakes in Indonesia and New Zealand (from USGS, 2007).

The basin size and depth reflect the overall storage volume achievable, as the larger the basin the greater the likelihood of having laterally extensive reservoir and seal pairs, possibly in multiple stratigraphic intervals, and therefore the greater the likelihood of injectable pore volume. The depth of the sedimentary fill of the basin is also relevant to the phase state of the CO2. Depths greater than ~800 m allow for storage as dense supercritical CO2 and hence significantly increased storage capacity in porous media (the actual depth to supercritical CO2 is dependent on the pressure and temperature regime of the basin and the depth of the water table, but 800 m is a typical minimum depth for average reservoir conditions). Furthermore, depth also impacts on the likely economic feasibility, as the greater the depth to the injection target the larger the associated costs of drilling, especially when multiple wells are required. In terms of depth, the least favourable conditions for CO2 storage are very shallow depths above 300 m as these preclude both storage into coal seams and saline formations/petroleum reservoirs. Shallow depths between approximately 300 and 800 m are theoretically suitable for CO2 storage into coal seams (discussed in coal seams criterion below) but will be still be too shallow for efficient storage in saline formations or petroleum reservoirs because the CO2 will be subcritical. Also it is possible that at such relatively shallow depths, the coal might be mined at some time in the future. Depths over 3500 m are generally undesirable due to the greater drilling costs involved and common reduction in reservoir permeability and corresponding injectivity. For large-scale storage in saline formations, intermediate depths between about 800 and 3500 m are considered to be the most favourable conditions for CO2 storage.

22

Faulting intensity is both a containment issue and a capacity issue. Depending on the nature of the faulting, the more extensively faulted that an area is, the greater the risk for containment breaches via conductive faults and fractures. Conversely, the greater the amount of faulting, the greater the potential for compartmentalisation of individual reservoirs if faults are sealing, which could reduce storage volume. Compartmentalised reservoirs generally require multiple wells to access the pore volume available within individual fault blocks. Australian examples include the relatively low faulting intensity of the Eocene stratigraphy in the central Gippsland Basin versus the highly faulted and compartmentalised Cretaceous stratigraphy of the onshore Otway Basin. It must be noted that even within the same basin, faulting intensity can differ between various stratigraphic intervals, as determined by the geological history of the basin. An example of this would be the expected difference in faulting intensity between syn-rift and postrift strata in an extensional passive margin basin. Hydrogeology describes the natural dynamic flow system and hence the potential for hydrodynamic trapping within the basin being assessed. Hydrodynamic traps are deep, laterally extensive, confined, saline aquifers which do not terminate with stratigraphic or structural traps but depend on long residence times (of the order of thousands of years) to immobilise free-phase CO2. Long residence times depend on slow flow rates and/or long migration pathways. Shallow, short flow systems therefore do not meet the geological requirements for maintaining supercritical CO2 (i.e. depth, pressure and temperature) and do not have a long enough residence time to immobilise the injected CO2 by one of the other trapping mechanisms (e.g. residual trapping, solution trapping or mineral trapping). The most suitable hydrogeology consists of a deep, confined aquifer with sufficient permeability for injection, but a relatively slow flow rate for both the free-phase and dissolved-phase CO2. Flow rate is controlled by the driving forces on the fluid (i.e. buoyancy and hydraulic gradient) and by the permeability and porosity characteristics of the reservoir rock the fluid is moving through. Geothermal effects on CO2 storage are discussed at length and in some detail in Bachu (2003). The geothermal conditions of a basin impact on the storage volume, as the density of CO2 is higher in colder basins than in warmer basins, allowing more CO2 to be contained within the same unit volume of rock. The temperature within a basin is dependent on both the in situ geothermal gradient and the surface temperature. In central-eastern Australia, the Galilee and Cooper-Eromanga basins represent warm regions with above average geothermal gradients, whereas the Amadeus, Georgina and parts of the Canning Basin have below average geothermal gradients (Figure 10). Reservoirseal pairs are a crucial prerequisite to selection of most CO2 storage sites. Reservoirs (specifically the pore space within them) provide the storage volume, and seals, due to their low permeabilities, provide the vertical containment. The lithology of the basin fill and its stratigraphic architecture are important contributors to the quality of both the reservoirs and seals. For most sedimentary basins a generalised stratigraphic column or a simplified chronostratigraphic chart is readily available and can be easily and quickly screened for a qualitative assessment of the presence, distribution and frequency of reservoir-seal pairs. However it has to be recognised that in areas where reservoirs have relatively low permeabilities, it does not necessarily preclude their use for CO2 storage, but it does have significant cost implications in terms of the need for more injection wells or the horizontal wells.

23

Figure 10: Map of Australia showing the estimated temperature distribution at 5 km depth. The hot area in central-eastern Australia comprises the Galilee and Cooper-Eromanga basins (image courtesy of the Hot Rock Energy Program, Australian National University) (from Murdoch University, 2006).

Coal seams are potential reservoirs due to their capacity to adsorb significant amounts of CO2 and so their presence in sedimentary basins provides another possibility for CO2 storage. In general, coal seams at shallower depths are likely to have higher permeability (and hence easier injection and lower injection cost) than deeper coals. In addition, there may be greater storage efficiency in coals at depths of 300600 m in comparison to saline formations or depleted petroleum reservoirs at the same shallow depths (Ennis-King & Paterson, 2001). Coals at very shallow depths (<300 m) are unlikely to have sufficient overburden pressure to retain the CO2 adsorbed to the coal matrix (Bradshaw et al., 2001). Coal-bearing strata in Australia that may be suitable for CO2 storage have been investigated for their ECSM potential exist, in the Bowen-Gunnedah-Sydney foreland basin system. Preliminary work suggests that the major challenge for CO2 storage in these systems is the low permeability of the coal seams. Target coal seams commonly have permeabilities of less than 10 mD. Despite these low permeability values, large amounts of CSM are produced in some basins. For example, in the southern Sydney basin the CH4 flow rates from some wells drilled in the high production fairway range up to 900 Mcf/day, which suggest possibly favourable permeabilities in these coals for CO2 injection. Coal rank is an important criterion when targeting coals for CO2 storage. Rank can be considered as an indicator of the adsorption capacity and therefore the storage potential of coals. Gas content generally increases with rank from lignites to sub-bituminous, bituminous and anthracite coals, given similar depth and pressure. The downside of deep high rank coals is that commonly they have low permeabilities, resulting in low achievable injection rates. Evaporites generally provide the best caprock seals, and hence the presence of evaporites, particularly in continuous beds, is likely to be beneficial to CO2 containment. In Australia, evaporite accumulations are known to occur in the Amadeus, Canning and other central Australian basins The hydrocarbon potential of a region provides an indication of the suitability of the area for CO2 storage, in that if the rocks are suitable for containing and producing oil and gas then it is likely that they are also suitable for storing CO2. However this also means that consideration has to be given to the potential impact of CO2 storage on petroleum resources.

24

The maturity of the extractive industries in the region reflects the likely database available, i.e. the more developed an area is, the greater amounts of data available for CO2 storage assessment. Furthermore, in mature areas, the majority of hydrocarbon/coal occurrences are probably discovered, lowering possible concerns about contaminating undiscovered occurrences. Mature areas also have the highest potential to contain abandoned, disused or depleted oil and gas reservoirs which could be utilised for CO2 storage. In addition, mature basin areas commonly have a good infrastructure in place with easy road access as well as an established pipeline and well network. The offshore Gippsland and Carnarvon basins represent good examples of relatively mature basins in Australia. Yet, even within these relatively mature basins, there are multiple possible sites that have not been characterised in any detail. Whether a basin is onshore or offshore provides an important economic consideration, as in general it is likely to be cheaper and technically easier to implement a CO2 injection site onshore rather than offshore. On the other hand, public perception and land use issues may dictate that offshore sites are preferential for many CO2 storage projects. Climate affects the likely surface temperatures (and hence the geothermal conditions) as well as the depth of the water table and also impacts on the likely ease of development. Likewise, accessibility and infrastructure reflect the technological feasibility and ease of future developments.

Ranking of Sedimentary Basins

By compiling data on the criteria above, different basins can be compared, contrasted and ranked for their suitability for CO2 storage. This can be done either qualitatively or semi-quantitatively (e.g. as per GibsonPoole et al., 2006) or quantitatively if scores are given for each criterion (as per Bachu, 2003). This allows the sedimentary basins within a country or state to be ranked in order of their suitability for geological storage of CO2. It should be noted that both storage associated with saline formations and with coal seams are dealt with under the one country/state-scale screening, so it is possible for a basin with high coal seam storage potential and low saline formation storage potential to have lower scores and vice versa. In methodology presented in this report, the separate treatment of these storage types is carried out in the next phase of basin-scale assessment. Examples of country and state-scale screening studies conducted by GEODISC/CO2CRC include the Bradshaw et al. (2002) summary of the potential for geological storage of CO2 in Australia and the GibsonPoole et al. (2006) review of CCS potential in Victoria.

Basin-Scale Assessment
Once a sedimentary basin has been identified as potentially suitable for CO2 storage, a basin-scale assessment can be undertaken to locate possible injection and storage sites. This is an intermediate scale of CO2 storage capacity evaluation (Figure 7). Potential sites can then be scored and ranked in order to identify those that have the highest prospect of successful CO2 storage and that warrant further detailed site characterisation. The earliest Australian regional assessment methodologies were developed by researchers such as Bradshaw et al. (2001). The following discussion is mostly extracted or modified from GibsonPoole (2008). The geology of a sedimentary basin is appraised in three basic steps to identify possible CO2 injection and storage sites: 1. Review basin stratigraphy. The stratigraphy of the entire sedimentary basin fill is reviewed to identify suitable rock combinations that may provide reservoirseal pairs and potentially injectable coal seams.

25

2. Determine reservoirseal pair and coal seam distributions. The distribution of reservoirseal pairs and coal seams within the basin are then determined and mapped, using available data such as geological and structural maps, seismic sections and well data. It is also useful to establish the minimum and maximum depth distributions of these media where CO2 storage is likely to be suitable: a. For saline formations or depleted petroleum reservoirs, the minimum depth distribution is considered to be the depth where the top of the target reservoir (or base of the overlying seal) is deep enough for injected CO2 to be in the dense supercritical phase (i.e. ~800 m below surface). The maximum depth distribution may be where permeability of the reservoir is no longer sufficient to allow viable injection rates (e.g. ~35004000 m). The depth window will vary from basin to basin based on factors such as geothermal gradient, pressure, fluid composition and mineralogy. The maximum depth distribution can also be established based on economic constraints, such as the maximum cut-off depth for economic drilling of injection wells or the maximum number of wells that can be economically viable.

b. For coal seams, the minimum depth distribution may be the limit at where the overburden pressure is still sufficient to retain the CO2 adsorbed to the coal, generally assumed to be 300m or greater (Bradshaw et al., 2001), or it could be the current ( or future) economic depth limit for underground coal mining. The maximum depth distribution may again be determined by the permeability characteristics which control the potential injectivity. Coal seams deeper than 13001500 m are commonly expected to be too impermeable to allow sufficient CO2 injection rates (CSLF, 2007). 3. Assess CO2 migration pathways and possible traps. The subsurface geometry of reservoir and seal units must then be assessed for the likely CO2 migration pathways and for physical traps, such as structural closures and stratigraphic pinchouts, to identify suitable injection locations. This can be achieved using structural contour and isopach maps, in combination with well data. Petroleum/coal prospectivity reports and well completion reports are also useful guides in identifying potential CO2 storage sites. Structural or stratigraphic traps containing existing hydrocarbon accumulations demonstrate the presence of viable reservoirseal pairs, but the risks of compromising oil and gas resources will need to be assessed. These locations need not be discounted if it can be assumed that these sites will eventually be available but at a later date (i.e. post-depletion), or can be evaluated as possible candidates for enhanced recovery. Furthermore, a physical structural or stratigraphic trap is not necessary if the expected migration pathway is sufficiently long to residually trap or dissolve the injected CO2 into the formation water before it reaches subcritical depths or the surface (i.e. hydrodynamic trap, Bachu et al., 1994).

Ranking of Prospective Sites

Once prospective sites have been identified, the relative merits of one potential site over another site can be compared and contrasted by utilising a ranking scheme, such as that devised by the GEODISC Program, data are compiled for each potential storage site to assess five key factors fundamental to any potential CO2 storage site: storage capacity, injectivity potential, site logistics, containment and existing natural resources (Table 7):
Table 7: Ranking factors for saline formations and petroleum reservoirs as prospective CO2 storage sites (modified from Bradshaw & Rigg, 2001; Rigg et al., 2001; Bradshaw et al., 2002).

Factor Storage capacity

Injectivity potential

Chance Being Assessed Will meet the volume requirements of neighbouring, currently identified CO2 sources Reservoir conditions viable for injection

Considerations Temperature, pressure, area, pore volume

Porosity, permeability, thickness

26

Factor Site logistics

Containment Existing natural resources

Chance Being Assessed Site is economically and technically viable Seal and trap will work for CO2 No viable natural resources in the site that may be compromised

Considerations Distance from CO2 source, water depth, reservoir depth, overpressure Seal capacity and thickness, trap, faults Proven or potential petroleum system, groundwater, coal or other natural resource (e.g. National Park)

1. Storage Capacity: this evaluates the total pore volume available for CO2 storage at a particular site, and compares it to the likely CO2 source volume that the site will need to accommodate (storage capacity). This is controlled by parameters such as size of the injection site area, the thickness of the reservoir, the pore volume available and the density of the CO2. This is a very simplistic storage volume calculation at this point, as the data available at this semi-regional level of characterisation is unlikely to be sufficient to establish more than Prospective Storage Capacity. 2. Injectivity Potential: this considers the reservoir characteristics, such as permeability and porosity, which will impact on how easily the CO2 can be injected into the reservoir. 3. Site Logistics: these are a reflection of the likely economic and technological feasibility, such as how deep an injection well needs to be drilled (depth to reservoir) or how far a pipeline might need to extend (distance from source). 4. Containment: this considers the seal and trap characteristics, such as the likely effectiveness of the seal based on its thickness, extent and lithology, or the migration distance (the longer the migration distance the greater the probability for secure trapping by residual, hydrodynamic or mineralisation mechanisms). Faulting size and intensity are also important containment considerations. 5. Existing Natural Resources: this assesses the likely presence of another resource that could potentially be compromised by CO2 storage, such as oil and gas, mineable coal, potable water or proximity to population centres or national parks, which could limit surface operations. The storage potential of coal seams can be assessed in a similar manner to saline formations and petroleum reservoirs, focusing on the same five key parameters of storage capacity, injectivity potential, site logistics, containment and existing natural resources. However, the differences in the CO2 trapping mechanism in coals in comparison to saline/hydrocarbon formations means that the characteristics under consideration for each factor are different Table 8).
Table 8: Ranking factors for coal seams as potential CO2 storage sites (modified from Bradshaw et al., 2001).

Factor Storage capacity

Injectivity potential Site logistics Containment Existing natural resources

Chance Being Assessed Will meet the volume requirements of neighbouring, currently identified CO2 sources Reservoir conditions are viable for injection Site is economically and technically viable Seal and trap will work for CO2
No viable natural resources in the site that may be compromised

Considerations Temperature, pressure, rank, ash content, lithotype, seam thickness, continuity, aerial extent Permeability, stress regime, mineralisation, structure, rank, lithotype Distance from CO2 source, coal seam depth, infrastructure, CSM potential Seal type and thickness, hydrology, trap, fault breaches Proven or potential coal resource, groundwater or other natural resource (e.g. National Park)

27

1. Storage Capacity: this is the ability of the site to accommodate the necessary source volume of CO2. In coal seams, this is controlled by parameters such as size of the injection site area (continuity of coal seams), the thickness of the coal seams, the ash and moisture content (low ash and low moisture contents are favourable) and the rank of the coal. For a given depth and pressure, higher ranked coals have greater storage capacity potential due to having greater adsorptive capabilities and CH4 contents (CH4 contents can be considered indicative of the CO2 storage potential, as the adsorptive capability of CO2 is at least double that of CH4). Crucially the depth of the coal will also determine whether the coal is or might in the future be mineable. 2. Injectivity: this is one of the most crucial factors in coal seams, due to the typically low permeabilities (in comparison to saline formations and petroleum reservoirs at similar depths and pressures). Injectivity tends to be more favourable in higher ranked and vitrinite-rich coals, as lower ranked coals have lower permeability due to higher moisture contents and reduced cleat formation. 3. Site Logistics: these are a reflection of the likely economic and technological feasibility, such as infrastructure (e.g. road access) or how far a pipeline might need to extend (distance from source). Coal seam CH4 potential can also be another important economic indicator; especially if net greenhouse gas emissions are to be decreased because the CH4 released through CO2 storage must be captured (due to its greater greenhouse radiative effect). 4. Containment: this considers the possibilities for leakage out of the storage coal seams(s). The CO2 is most likely to be well-contained where coal seams are at depths greater than 300 m (for sufficient overburden pressure), have limited faulting and are overlain by a suitable seal lithology. 5. Existing Natural Resources: this assesses the likely presence of another resource that could potentially be compromised by CO2 storage, such as oil and gas, mineable coal, potable water or national park. The sites can then be qualitatively scored between 0 and 1 (0 = worst, 1 = best) for each of the five factors, and the factors multiplied together to obtain a Chance Factor (as per Bradshaw et al., 2002). The higher the value of the Chance Factor, the more favourable the site characteristics and the greater will be the likelihood of technical and commercial success. This ranking allows each site under analysis to be ranked relative to one another to determine a seriatim of sites within a particular basin in order of preference for further consideration. Examples of basin-scale assessment studies conducted by CO2CRC include an evaluation of storage areas in several eastern Australian basins by Rigg et al. (2005) and the storage potential of various parts of the Bowen Basin by Patchett (2006), Kalinowski (2006) and White (2007). A regional assessment of the Gippsland Basin is provided by Gibson-Poole and Svendsen (2005a), whilst in Western Australia, Spencer (2004) reviews the storage potential of the northwest Canning Basin area and Varma et al. (2007) report on the regional potential for CO2 storage in the Collie and Perth basins.

Site Characterisation: Saline Formations and Petroleum Reservoirs

Once potential storage sites have been identified and ranked during the basin-scale assessment stage of investigation, a prospective site has to be further evaluated through a process of detailed site characterisation (Figure 7). Site characterisation is defined as The collection, analysis and interpretation of subsurface, surface and atmospheric data (geoscientific, spatial, engineering, social, economic, environmental) and the application of that knowledge to judge, with a degree of confidence, if an identified site will geologically store a specific quantity of CO2 for a defined period of time and meet all required health, safety, environmental and regulatory standards (Cook, 2006).

28

Site characterisation is the most time-consuming and costly part of the CO2 storage site selection process. Because CO2 site characterisation taps into a vast array of expertise, and requires skills in reservoir engineering, structural geology, sedimentology, stratigraphy, hydrogeology and geological modelling, this stage of CO2 storage is best done in a multidisciplinary team environment. It is also the stage that clearly goes beyond the pre competitive stage of a study and may therefore be more properly seen as the role of industry or a project proponent, rather than of government. Site characterisation requires greater detail than basin-scale assessment investigations and may involve re-evaluation of regional geology, generation of new data and/or updating of existing static geologic and seismic data and dynamic engineering data and numerical flow simulation models. Data sources can include 2D and 3D seismic surveys, well log and core data, drill cuttings, biostratigraphy, field production and fluid data. Typical steps in site characterisation are structural and stratigraphic interpretation based on available subsurface data, building of geological models with realistic stratigraphic heterogeneity, building of geochemical, geomechanical and hydrogeologic models, and constructing numerical flow simulations to predict CO2 plume migration (Figure 7) (GibsonPoole et al., 2005; Gibson-Poole, 2008). Models have to be continuously updated as additional data become available, a process that must extend to the post-injection phase for monitoring purposes. Depending on data coverage and quality, all of the above steps should incorporate appropriate levels of uncertainty in interpretation, which should be reflected in the various outcomes of multiple flow simulation scenarios. An important aspect of site characterisation is the determination of acceptable versus unacceptable levels of uncertainty in order to determine the amount of risk associated with the site and the amount and type of additional data required to reduce the uncertainty. These additional data have associated costs, both financially and in effort, as well as an additional time component that must be built into the project management. Three key factors that require further detailed evaluation (modified from the five suggested by Bradshaw et al., 2002) are: injectivity, containment and capacity (Figure 7). These three factors encompass the fundamental elements needed to characterise any potential CO2 geological storage site and are described in more detail below. The following discussion is mostly extracted or modified from Gibson-Poole et al. (2005) and Gibson-Poole (2008).

Geoscience Characterisation: Injectivity

Injectivity is the rate at which CO2 can be injected into a given reservoir interval (a volume of CO2 per unit of time) and the ability of the subsequent CO2 plume to migrate away from the injection well. Low injectivity potential for an interval might result in a site with otherwise excellent capacity and containment characteristics turning out to be uneconomic and, therefore, unsuitable for CO2 storage. During CO2 injection into a reservoir, the injectivity and nature of plume migration will depend on parameters such as the viscosity ratio, injection rate and relative permeability. These parameters will in turn depend on variables such as depositional environment and reservoir heterogeneity, stratigraphic architecture, postdepositional diagenetic alteration, structural dip, fault distribution and fault seal capacity, pressure distribution, and the nature of the formation fluids. Injectivity issues that can be assessed through the geoscience characterisation include the geometry and connectivity of individual flow units, the nature of the heterogeneity within those units (i.e., the likely distribution and impact of baffles such as interbedded siltstones and shales) and the physical quality of the reservoir in terms of porosity and permeability characteristics (Figure 7). Reservoir quality can be assessed by evaluating core (porosity and permeability) and wireline logs (petrophysical interpretations) of existing wells. Well data is one-dimensional and away from these points, rock properties have to be inferred through well log correlation and use of analogues, guided by seismic interpretation. Geological (static reservoir) models need to be constructed to provide likely reservoir distribution and horizontal and vertical connectivity of flow units. These are best placed in a sequence stratigraphic framework, which allow improved predictions of heterogeneity and barrier/baffle distribution. Since important levels of heterogeneity can fall below seismic resolution, uncertainties in interpretation increase with distance from the wells. These uncertainties are best handled by allowing for multiple model realisations. Both favourable and unfavourable outcomes must be considered, with the overriding

29

parameter that all realisations honour existing data. The uncertainties will greatly decrease if the site has pre-existing boreholes or is in an area with 3D seismic coverage. Evaluation of injectivity for a site should also incorporate examination of the mineralogical composition of the reservoir for potential post-injection effects. CO2 dissolution into the formation water may result in CO2-water-rock interactions, which may alter the mineralogy and pore system of the rock (Watson et al., 2004). This can have important implications for injectivity, as mineral dissolution may lead to increased porosity and permeability. However, it could also result in mobilisation of fine clay particles or the precipitation of new minerals, either of which can occlude the porosity and permeability of the reservoir rock, thereby decreasing injectivity. Determination of which, if any, of these reactions might occur is dependent on the specific geological properties at each site. Injectivity in Low Permeability Reservoirs CO2 storage in saline formations has several advantages over other geological storage options (i.e. depleted oil and gas reservoirs, coal seams), such as greater storage volume potential and less risk of compromising existing resources. However, many deep saline formations suffer from low permeability, due either to depositional processes (fine grained sediments and corresponding small pore throat sizes) or diagenetic processes (post-depositional mineralogical modification of pores and pore throats). Numerical investigations show that formation permeability is one of the main controlling parameters of CO2 storage in geological formations (van der Meer, 1995; Law & Bachu, 1996; Ennis-King & Paterson, 2001). For low permeability formations, numerical simulations show that there will be large pressure gradients near the wellbore, which will restrict the injectivity considerably. Thus, high permeability formations would be desirable for injection. However, high permeability formations would allow relatively fast migration of CO2, lowering the proportion of CO2 trapped behind the main plume by residual trapping. Slower movement provides greater residence time and thus several advantages such as enhancement of the dissolution potential of CO2 in formation water and high volumes of CO2 trapped by capillarity (Flett et al., 2005). Bachu et al. (1994) suggest that while higher permeability may be required near the wellbore to increase injectivity, lower permeability is desirable outside the radius of influence of the wellbore to increase residence times and encourage the rate of residual trapping, dissolution and mineral trapping.

Geoscience Characterisation: Containment

Containment refers to the retention of injected CO2 within the subsurface site relative to the overall risk of its escape. Containment is an issue in CO2 storage because injected supercritical CO2 is less dense than water and has the tendency to be driven upward due to buoyancy forces. Loss of containment can occur through vertical fluid migration via the top seal, faults/fractures and existing well penetrations, or lateral migration to a point where stratigraphic loss of seal occurs. Possible containment issues, therefore, include the distribution and continuity of the seal, the seal capacity (maximum CO2 column height retention), CO2water-rock interactions (potential for mineral trapping), potential migration pathways (structural orientations and dips), distribution and extent of intraformational seals (acting as localised barriers and baffles to flow), hydraulic gradient (formation water flow direction and rate) and the integrity of the reservoir and seal (rock strength, fault/fracture stability and maximum sustainable pore fluid pressures) (Figure 7). Migration through existing well bores and leaking faults are considered the greatest containment risks in CO2 storage (Celia & Bachu, 2003). CO2 injection into the geological subsurface increases the formation pressure, which can then potentially reactivate pre-existing faults or generate new fractures. Opening of fractures or causing slip (movement) on faults could lead to a loss of containment (Holloway & Savage, 1993; Bergman & Winter, 1995; Streit & Hillis, 2004). Thus, an understanding of the pressure regime and geomechanical modelling needs to be undertaken to estimate maximum sustainable fluid pressures for CO2 injection that will not induce fracturing and faulting. This requires the determination of prevailing stresses, fault geometries and rock strengths. Details of geomechanical modelling techniques are described in Streit and Hillis (2004). The properties of the seal, both in terms of CO2 retention capacity as well as distribution and continuity, can be assessed by a combination of laboratory and sequence stratigraphic analyses. The seal capacity of

30

regional top seals and localised intraformational seals is calculated by determining the capillary pressure properties of the sealing rock and the physio-chemical properties of both the CO2 and formation water (e.g. density, wettability and interfacial tension). Seal capacity is best assessed by Mercury Injection Capillary Pressure (MICP) analysis. MICP tests are a measurement of the pressures required to move mercury through the pore network system of a core sample. The air/mercury capillary pressure data can then be translated to equivalent CO2/brine data at reservoir conditions and converted into seal capacity for CO2, expressed as the column height that the seal rock would be capable of holding (sealing). Standard procedures for MICP analysis are reviewed by Vavra et al. (1992), Dewhurst et al. (2002) and Daniel and Kaldi (in press). CO2 introduced into a reservoir system can also chemically interact with the host rock. Detailed reservoir petrology, water chemistry and pressure-temperature conditions provides information necessary for modelling of potential mineral reactions associated with CO2, including dissolution, alteration and precipitation. Whereas for injectivity, CO2-water-rock interactions could either enhance or reduce the injectivity through mineral dissolution or mobilisation, mineral precipitation can lead to mineral trapping of CO2 and therefore increased containment security as the CO2 is permanently trapped (Bachu et al., 1996; Perkins & Gunter, 1996; Watson et al., 2004; Watson & Gibson-Poole, 2005). Sequence stratigraphic assessment of both the reservoir and the seal can predict facies distribution even in areas of poor data coverage. This is based on predictable occurrence of strata packages with changes in potential space available for sedimentation and sediment supply, which can be inferred from subsurface wireline and seismic data. In a homogeneous reservoir, the buoyancy of the free-phase (immiscible) CO2 will cause it to migrate vertically up to the top of the reservoir. Stratigraphic heterogeneities, such as intraformational siltstones and shales, have the potential to reduce this type of flow and create a more tortuous migration pathway. Once the CO2 plume has reached the top of the reservoir, the structural dip and geometry at the base of the overlying sealing unit will have a strong influence on the subsequent migration direction and rate. The details of the exact geometries and heterogeneities at the reservoir/seal boundary can prove especially important and can result in unexpected plume migration and possible loss of containment. Trapping mechanisms that can be identified through geoscience characterisation include physical structural closures and stratigraphic pinch-outs, and potential hydrodynamic (Bachu et al., 1994) or rate seal traps. An understanding of the existing formation water flow system within a geological reservoir is important for site characterisation as the rate and direction of flow of the formation water system will impact the effectiveness of hydrodynamic trapping, as well as dissolution and residual trapping along the migration pathway. Hydrodynamic modelling assesses the vertical communication between reservoir units, and hence the effectiveness of a seal, and provides information about the horizontal hydraulic continuity (e.g. fault compartments) and the impact of low permeability zones. In areas where there has been fluid removal from the subsurface (e.g. water extraction or oil production), an assessment of the pre-production hydrodynamic regime is used to provide an understanding of the long-term (hundreds to thousands of years) influence of the formation water flow systems on the injected CO2. However, it is also necessary to interpret the present-day hydrodynamic regime, as it may have been affected by hydrocarbon/water production-induced pressure decline. The present-day (post-production) hydrodynamic regime can be used to evaluate the potential short-term (tens to hundreds of years) influence on the predicted migration pathway of CO2 immediately after injection. The past and present formation water flow systems can be characterised from pressure-elevation plots and hydraulic head distribution maps using standard hydrodynamic analysis techniques as presented by Dahlberg (1995), Bachu (1995), Otto et al. (2001) and Bachu and Michael (2002).

Geoscience Characterisation: Capacity

Storage capacity evaluates the commercially available pore volume for CO2 storage at a particular site. This is controlled by parameters such as the size of the containment area, the thickness of the reservoir, the effective porosity and the density of the CO2. Storage capacity is discussed in detail earlier in this report.

31

Engineering Characterisation
The engineering characterisation phase continues on from the geoscience characterisation (Figure 7). Shortterm numerical simulation models of the injection phase are needed to provide data on the injection strategy required to achieve the desired injection rates (e.g. number of wells, well design and injection pattern). Post-injection phase numerical simulations evaluate the long-term storage behaviour, modelling the likely migration, distribution and form of the CO2 in the subsurface. Coupled simulation models, such as geochemical reactive transport, can also be developed to further evaluate the CO2 storage potential of a site. Optimising reservoir engineering via integration of reservoir characterisation with well placement, completion, conformance control, and injection strategies may in fact increase storage capacity.

Socio-Economic Characterisation
The final stage in a detailed site evaluation is the socio-economic characterisation (Figure 7). This includes economic modelling to establish such aspects as the likely capital and operating costs, as well as the cost per tonne of CO2 avoided. Also, the acceptability of a site by the community will be dependant of the communitys perception of the environmental and social impact of geological storage. Risk and uncertainty analysis is crucial to establish whether a selected site can be classed as a safe and effective storage site for thousands of years. The design of a monitoring and verification program is dependent on the geological characteristics of the selected site and needs to be carefully evaluated to produce an optimum program both in terms of efficiency and cost. The current CO2CRC Otway Project will be of great value in enabling optimisation of monitoring.

Site Characterisation: Coal Seams

Coal systems comprise both the coal seams and the sediments that occur above and below coal seams, as well as intercalated within the coal seams. Characterising the interbedded sediments is similar to characterising other clastic sedimentary formations (saline formations and petroleum reservoirs). Coal seams, however, have different properties and characterising them requires parameters to describe the variability in the reservoir that will occur due to coal quality (i.e. its rank, grade and type), in addition to seam thickness and splitting characteristics, and geological and hydrological setting. Basic parameters that describe CO2 storage capacity and injectivity in coals are seam thickness, adsorption capacity and permeability. Seam thickness in coals will vary across basins and coalfields due to the depositional and tectonic setting in which the precursory peat accumulated prior to burial. In this early stage, the coal type (or vitrinite content), the grade (or mineral matter content) and composition will be set, although secondary mineralisation can occur during burial and subsequent uplift episodes. The depth, temperature and duration of burial (i.e. burial history) of the coal formation will determine the thermal maturity or rank. These parameters (rank, grade and type) will control the adsorption capacity and the permeability of the coal. As a result, it is important to understand the variability in these basic parameters across a site. In virgin coal seams, gas content commonly increases with depth of cover but variability can be very high at the scale of a basin, a mine site or even a single well. Gas adsorption potential of a given coal will tend to increase with increasing pore pressure or depth. It will also vary with moisture content and rank of coal. Hence, rank is commonly used as an indicator of the adsorption capacity of a coal, and when combined with thickness or known resource tons can indicate the storage potential within a basin. Storage capacity cannot be decoupled from injectivity, even at a basin-scale. Counter to increasing gas adsorption potential with depth, permeability decreases with depth. Permeability in coal is dependent on the degree of jointing and cleating (which will increase with increasing rank and coal brightness), on the mineralised or unmineralised condition of the cleat, and on the principle horizontal stress magnitude and its orientation relative to the cleats; hence it will decrease with increasing depth of cover. The proximity of igneous intrusions may also be a limiting factor, particularly in the Bowen-Gunnedah-Sydney Basins, due to the possibility of decreased permeability associated with mineralisation of cleats, plasticisation of the

32

coal and native CO2 content in the coal (of thermal origin). With notable exceptions, coals at depth commonly have permeabilities of less than 10 mD. This presents a major challenge for deep CSM extraction, and more so for injectivity and geological storage of CO2. Coal plasticisation due to CO2 injection may also cause reduction of permeability in some coals. However, in a series of recent publications, (Sakurovs et al., 2007), (Sakurovs et al., 2008), (Day et al., 2008) and (Day et al., in press) have found no evidence of irreversible behaviour of the coals on exposure to CO2, even up to 20 MPa exposure. Although this does not completely exclude the possibility that some coals can be plasticised by exposure to CO2, this possibility may be less common than previously believed. New data from Radlinski (unpublished) also suggests that plasticisation may be an isolated phenomenon for some coals but in general is not an issue for Australian coals (Radlinski, pers. comm., 2007). Therefore, further research is required to understand the affect of the micro-structure of the coal on adsorption of CO2.

Data Needs for Site Characterisation

One of the main objectives of site characterisation is to accurately predict CO2 behaviour in the subsurface, with respect to injectivity, containment, and storage capacity at a specific site. It is also essential to evaluate the uncertainty in these predictions. Making reliable predictions about subsurface CO2 behaviour is closely related to data availability and quality, and therefore understanding the uncertainties associated with various types of data used in subsurface work is crucial for a successful CO2 storage project. Poor understanding of data quality can result in unrealistic geological models and unreliable flow simulation results, and hence misleading assessment of the suitability of a CO2 storage site. This section briefly summarises some of the main data types used in site characterisation and highlights the key problems of data resolution with respect to resolving various components of site characterisation. A summary of the main data needs (required as well as desirable) for the various levels of site characterisation and storage volume assessment is presented in Table 9.
Table 9: Summary of the main data needs (required as well as desirable) for the various levels of site characterisation and storage volume assessment.

Data Needs

Country/ State-Scale Screening; Total Pore Volume

Regional geology Detailed/local geology Structural contour Coal/reservoir geometry Coal/reservoir quality Fault Seismicity Hydrologic Surface infrastructure Topographic 2D 3D Gamma ray Porosity

Basin-Scale Assessment; Prospective Storage Capacity

Site Characterisation; Contingent Storage Capacity

Site Deployment; Operational Storage Capacity

Maps

Seismic Well logs

33

Data Needs

Country/ State-Scale Screening; Total Pore Volume

Permeability Sonic Density Image Porosity Permeability Langmuir volume (coal) Ratio vertical/ horizontal permeability Relative permeability Mercury injection capillary pressure Mineralogy Rock strength Oil/gas production CSM reservoir conditions Water chemistry Repeat formation tests; drill stem tests Subsurface fluid properties Leak-off tests; formation integrity tests Sequence stratigraphy Regional tectonic history/model Regional stress analyses Biostratigraphy Analogues Static models Dynamic models

Basin-Scale Assessment; Prospective Storage Capacity

Site Characterisation; Contingent Storage Capacity

Site Deployment; Operational Storage Capacity

Core

Special core analysis

Subsurface history

Pore pressure

Reservoir characterisation

Economics Regulatory framework

= required; = desirable

Apart from actually obtaining new data (e.g. drilling new wells, acquiring new 2D or 3D seismic), which can be expensive, there is the additional cost in money and time for processing and interpretation of the new data. Therefore, in most instances, the early stages of site characterisation will initially rely on data

34

already available from regional geological surveys, or mineral (coal) or petroleum exploration and development activities. Such data will by default be focused around existing fields or potential prospects, which will generally be mainly focused on traditional structural or stratigraphic traps. Dense data coverage (well logs and 3D seismic) will only be available in proven resource basins, and more specifically in presently or formerly producing oil and gas fields or coal seams. In addition, for an injection site near an existing or partially depleted hydrocarbon field, for example, the most likely scenario will be to inject down-dip from the actual accumulation or in the deeper, sub-accumulation saline formation. However, in this location the data are likely to be relatively sparse, even in areas with overall good data coverage (e.g. Gippsland Basin, Gibson-Poole & Svendsen, 2005b). Poor data coverage is one of the leading causes of uncertainty in assessing areas for CO2 storage sites. Obviously if a site is to progress to the stage of being operational, then key data sets will need to be collected and the cost of doing this has to be accepted as part of the total capital cost of developing the site. Various data types can be used for CO2 site characterisation, such as 2D and 3D seismic data, wireline logs, pressure data, core data, biostratigraphy, and production data among others. Different data sources have different resolutions and are capable of capturing different levels of subsurface information. Seismic data, well log, and core data are the most important subsurface data sources and understanding their advantages and limitations is of vital importance in generating robust geological models.

Seismic Data
Seismic data are remotely sensed and based on wave reflections caused by variations in rock (and fluid) density and velocity in the subsurface. Seismic data have long been used for resolving structural deformation in the subsurface. Recent improvements in 3D seismic technology have also made possible the direct imaging of stratigraphic heterogeneity, such as locating palaeo-valleys and channels and reservoir properties (porosity and saturation). Seismic data are routinely used for working out the three-dimensional distribution of faults and anticlines, understanding regional structural and stratigraphic dips, and working out the internal architecture of sedimentary packages. Such knowledge is directly applicable to (and often guides) the generation of geological models and prediction of CO2 migration in the subsurface. Seismic data inherently carry several sources of uncertainty: (1) the data have to be processed, which can be done in different ways and with varying results (i.e. processing can lead to more than one solution); (2) conventional seismic data have poor vertical resolution, which can be tens of metres at typical reservoir depths; (3) any interpretations based solely on seismic data will not include potentially important levels of sub-seismic scale stratigraphic heterogeneity and structure. This in turn can lead to important levels of reservoir heterogeneity (in the extreme, entire reservoir compartments and flow units) not being incorporated in a geological model; (4) seismic data away from areas with well control can become especially unreliable, since migration of the seismic data has to be performed based on rock velocities that are inferred rather than measured, and there is a lack of other means to calibrate the seismic data to real depth; and (5) 2D seismic data, which are more commonly available and much cheaper than 3D seismic data, are much less reliable and associated with higher levels of uncertainty. Models based on interpreted seismic data should allow for suitable levels of uncertainty based on the quality of the seismic data and other aspects of the area of interest. One of the best ways to decrease uncertainty and associated risk is to use seismic data along with other data sources, such as well logs or cores, which have vastly superior vertical resolution. Seismically surveyed areas with poor well control will inherently carry large levels of uncertainty.

Well Log and Core Data

Well log (wireline) data provide continuous measurement of physical parameters of rocks and fluids along a borehole. Availability of well log data is controlled by the number of wells in an area, which can range from sub-kilometre spacing densities over mature fields in prolific resource basins, to wells which are spaced tens to hundreds of kilometres apart in under-explored areas. Commonly used logging tools measure properties such as electric resistivity, spontaneous potential, natural and induced radioactivity and acoustic properties (Zimmerle, 1995). These are commonly used as proxies for various rock and fluid properties

35

such as grain size, porosity, density, stratigraphic dip and composition of formation fluids. Well log data offer vastly superior vertical resolution compared to seismic data (<1 m), which makes such data ideal for seismic calibration and resolving sub-seismic scale heterogeneity in the subsurface. Well log data, calibrated by core data, in conjunction with seismic data can be used to guide both geological interpretation and populate geological models with petrophysical properties. A common difficulty arises from the need to combine data at very different scales, necessitating experienced subsurface specialists, familiar with integration and interpretation of such data. The quality of well log-based geological interpretation is highly dependent on the well spacing (i.e. density of data points). Dense well spacing can result in detailed correlation, which allows mapping of individual reservoir compartments and realistic predictions of subsurface porosity and permeability distribution. Large well spacing can lead to extreme uncertainty in correlation and geological interpretation, and poor predictability of CO2 migration. Generally, for heterogeneity to be resolvable, well spacing has to be smaller than the scales of the feature mapped. Well spacing of tens of kilometres or higher can be used to calibrate seismic data and to evaluate the overall reservoir potential of an interval, but should only be used with extreme care for direct correlation of potential flow units. Many potential CO2 storage sites will have insufficient well spacing for detailed correlation, and will require other methods, such as use of analogues to approximate an appropriate level of heterogeneity within an interval. Well log patterns can be used to infer the sedimentary environment of deposition, which is one of the most important parameters that determines stratigraphic heterogeneity, and choice of analogues. Since log patterns are only a proxy of real rock properties, and can give misleading results due to unusual mineralogies or formation fluids, examination of core from an interval can significantly decrease the uncertainties in geological interpretation. Cores are rock samples extracted during drilling that allow direct observation of rock properties (e.g. sedimentary facies, mineralogy porosity/permeability or capillary pressure measurements), and therefore extremely valuable for site characterisation. Unfortunately, partially because of added cost, core data are rarely available at most wells locations, as most operators view core as only complementary to other data types. As is also true for well log data, the main drawback of core data is that it is representative of a very small area of the reservoir or seal being evaluated and, therefore, may not sample important rock properties beyond the immediate wellbore.

Use of Analogues
Analogues are data compilations or case studies of areas with high quality data (modern, outcrop or subsurface), which can be used to provide levels of expected heterogeneity for a reservoir of a given depositional setting. An example of this would be a database of channel width-to-thickness ratios for braided rivers, which can be used to populate a geomodel with cell properties that approximate channels with appropriate ranges of channel sinuosity, thicknesses and widths. Appropriate use of analogues depends on accurate identification of depositional environments and other factors affecting sedimentation, which can be challenging in data-limited areas. Presence of core data, from which depositional environments can be most accurately inferred, is the most important factor for limiting uncertainties in choosing appropriate analogues.

Formation Pressure
A knowledge of the in situ formation pressure is integral to any site characterisation and underpins the hydrogeological modelling, injectivity studies, geomechanical modelling, seal capacity estimations, flow modelling, and understanding of the horizontal and vertical communication. Formation pressure is sourced either from wireline tests, such as the repeat formation tester (RFT), or from drill stem tests (DSTs). Both types of testing provide information about the pressure, permeability, fluid type and density and the location of fluid contacts. Wireline formation tests consist of a set of measurements taken close to the surface of the wellbore and as such, represent local permeability and pressure. They are also subject to wellbore issues, such as supercharging and formation damage. They are, however, carried out in almost all petroleum exploration bores and are a very reliable estimate of pressure, permeability and fluid properties. DSTs are a much larger scale test and are usually carried out over a several metres of reservoir. The reservoir is allowed to flow for several hours (overcoming borehole effects) and extending for several

36

meters from the borehole. The tool is then shut-in to allow the tool pressure to equilibrate to the formation pressure. This leads to very accurate estimates of formation pressure, permeability and fluid properties. Analysis of the flow and build-up curves can also identify barriers to flow (i.e. sealing faults) and dual permeability systems that were affected by radius of the test. In depleted oil and gas reservoirs, the pressure decline observed in a well or field is used to calibrate the reservoir simulation model and predict the behaviour of injected CO2 in this type of storage model (e.g. the CO2CRC Otway Project in Victoria, Australia).

Site Deployment
Prior to CCS technology being deployed on a commercial scale in specific sites, there is a need for a range of technical and non-technical activities (Figure 7): A detailed volumetric assessment of storage capacity, including dynamic simulation models of injection and migration of CO2 must be completed. The Proved (P1) Operational Storage Capacity must be equal to or greater than the CO2 planned to be transported to the site and injected. Full project economics must be run (to include capital and operating costs of capture, compression, transport and injection, including duration-of-project monitoring and verification and include assessment of total project costs per tonne of CO2 avoided). Monitoring and verification facilities must be in place. This could include direct and remote sensing, geophysical, geochemical and petrophysical technologies for monitoring subsurface, wellbores, near surface (soil and potable aquifers) and atmosphere. Risk assessment and uncertainty analysis should be undertaken, including a qualitative and quantitative risk assessment of potential subsurface CO2 escape or leakage from infrastructure (wellbores, pipelines and facilities). Community consultation should be ongoing from the time a site is selected. Local landowners, neighbouring communities and other tenement holders must be informed and kept involved in the process. The regulatory regime should be in place. This is perhaps the most critical factor. Australia has well-developed and long-standing petroleum and mineral legislative and regulatory regimes. These include frameworks for resource conservation, environmental protection, transport of gases by pipeline (not CO2 specifically), a legislative and/or regulatory regime for storage and injection of gases as part of a hydrocarbon production and/or recovery operation, and considerable technical, legislative and regulatory know-how for petroleum industry operations. Changes to existing oil and gas legislative and regulatory regimes can be made to regulate CCS. At the present time the regulatory regime for geological storage in Australia is for the most part fairly unclear. The CO2CRC Otway project has provided perhaps the best basis to date for regulating a storage site, although the project is actually being conducted under the R&D provisions of the Victorian EPA (Cook & Sharma, 2006), (Sharma et al., 2007) However, the existing regulatory regime does not incorporate a framework that can be adapted to address issues such as long term liability, which is unique to the geological storage of CO2. Appropriate regulatory/legislative frameworks need to be enacted in the State (or Federal, if offshore) jurisdiction to provide certainty for investors by enabling access and providing security of tenure. Currently, there are no uniform guidelines regulating CCS and regulatory issues are addressed in specific projects, such as CO2CRC Otway Project or other demonstration or commercial projects. In these instances, regulations are dealt with on a case by case basis through either the use of existing legislative arrangements or the development of project specific legislation (e.g. Gorgon Project in WA and enactment of the Barrow Island Act 2003). The key focus of a regulatory framework for CCS is to identify the potential health, safety, financial and environmental risks.

37

The quantification of these risks will assist in gaining community confidence as well as providing a sound basis for industry investment. The CO2CRC Otway Project provides several good examples of the detailed assessment for site deployment, including geological characterisation, determination of Proved Operational Storage Capacity and site specific issues including economics, monitoring and verification, community perception, risk assessment, and regulatory matters ((Cook & Sharma, 2006) and (Sharma et al., 2007).

Conclusions
Carbon dioxide capture and geological storage is a means of reducing greenhouse gas emissions into the atmosphere. The estimation of how much storage capacity there is, and where it is located are fundamental issues for the commercial deployment of CCS. Current storage capacity estimates are imperfect and there is a need for more development and more general agreement on assessment methodologies for site selection. Previous attempts to assess CO2 storage capacity have used a range of approaches and methodologies, and data sets of variable size and quality, resulting in widely varying storage estimates of inconsistent quality and reliability. The main efforts to date to develop a scheme have been by the CSLF (2007) and the DOE (2006). The incorporation of components of both the CSLF and DOE methodologies plus the adoption of the SPE (SPE, 2007) classification scheme is an attempt to build on the previous schemes in order to develop one that will be acceptable to industry, the scientific community and regulators, but which is also acceptable to the financial, banking and insurance sectors in much the same way that the SPE scheme (or the Australian JORC scheme of the mining sector) for expressing resources and reserves is both scientifically and financially acceptable . This study distinguishes between available pore volume (a theoretical estimate of the amount of pore space that can be used to store CO2 in subsurface geologic formations) and storage capacity (the pore volume constrained by economic or engineering feasibility limitations). A similar distinction was proposed by CSLF (2007). The term total pore volume refers to the entire volume which is estimated to exist originally in sedimentary basins, naturally occurring storage systems or potential storage sites, plus the estimated pore volume yet to be identified. Total pore volume is subdivided into discovered and undiscovered pore volume. Discovered pore volume is classified as either commercial or sub-commercial. Undiscovered pore volume is pore volume that is as yet unfound, but which, under the right technical, economic or regulatory conditions, might be commercial. Storage capacity comprises 3 categories: operational storage capacity is discovered pore volume that is considered commercial i.e. CO2 is injectable under existing technical, economic and regulatory conditions. Operational storage capacity is separated into proved (1P), proved plus probable (2P) and proved plus probable plus possible (3P) categories following standard petroleum industry nomenclature. Contingent storage capacity is discovered pore volume which will be commercial in the future when technical, economic and regulatory conditions are in place and prospective storage capacity is undiscovered pore volume which might become commercial at some future time. Because of uncertainties inherent to subsurface evaluation, exact quantification of geological properties is not possible and therefore storage capacity is always at best an approximation of the amount of CO2 that can be stored. Storage capacity is considered a resource, and should be categorized based on levels of certainty of resource availability. Hence, the likelihood of contingent and prospective storage capacities achieving commerciality requires a probabilistic approach identifying the high, low and best estimates, in much the same way that a probabilistic approach is taken to oil and gas resources and reserves The prime focus of the selection of storage sites is to adequately estimate storage capacity, model the physical and chemical processes which will take place during and after injection, set in place the appropriate technologies for monitoring the movement of the stored CO2 and assess the risks associated with all phases of the process. A critical consideration for both storage capacity estimation and site selection is the identification of the storage system available in any location being considered for storage. The principal storage systems are saline formations, depleted (or near depleted) oil and gas reservoirs and coal systems.

38

Storage capacity estimates of saline formations are subdivided into physical (structural or stratigraphic) traps and hydrodynamic traps. The hydrodynamic traps are further categorised as solution, residual gas and mineral traps. These last three constitute long term (millennia to millions of years) potential storage. Detailed estimation of storage capacity in depleted oil and gas reservoirs follows the volumetric methods used by the petroleum industry to calculate reserves and resources in the petroleum reservoir. Storage capacity in coal systems comprises storage in the (generally) low permeability sediments above, below and in between coal seams, as well as in the coal seams themselves. Storage capacity estimation in sediments is similar to that for saline formations, with the exception that known coal systems are generally not at sufficient depths to allow CO2 to be stored in a supercritical state. Research into CO2 storage in coal is still at an early stage and further work needs to be conducted to fully understand the factors affecting storage capacity estimation in coals and determine whether or not storage in coals has the potential to be a significant mitigation option. The selection of suitable sites for the storage of significant volumes of CO2 comprises mainly geological evaluation on progressively more detailed scales. These correspond to degrees of confidence in estimating storage capacity, and an integrated storage capacity/site selection pyramid is proposed. Country or state screening involves identification of appropriate sedimentary basins which can then be screened and ranked relatively quickly as to their overall suitability for CO2 storage. First-pass evaluation of the size and thickness of the basin gives an indication of total pore volume it may hold. More detailed basin assessment takes into consideration factors such as tectonic setting, basin size and depth, intensity of faulting, hydrodynamic and geothermal regimes, existing resources and industry maturity. It is possible to identify the main trapping mechanisms and thus estimate the prospective storage capacity in each of the identified potential trap types, e.g. stratigraphic vs. structural traps; hydrodynamic, residual gas saturation, dissolution, mineral precipitation, and/or adsorption on coal. Site characterisation is the most time-consuming and costly part of the CO2 storage site selection process. Site characterisation involves greater detail than basin assessment investigations and may involve reevaluation of regional geology, and generation of new and/or updating of existing data such as geologic and seismic data (static) and engineering data (dynamic), computer models, flow simulation, and ultimately injection. Site characterisation considers factors such as injectivity and containment. Data sources can include 2D and 3D seismic surveys, well log and core data, drill cuttings, biostratigraphy, field production and fluid data. Typical steps in site characterisation are structural and stratigraphic interpretation based on available subsurface data, building of geological models with realistic stratigraphic heterogeneity, and flow simulations to predict CO2 plume migration. The level of detail determined in site characterisation allows the estimation of contingent storage capacity. The ultimate goal of a storage project is to have a commercially and technically viable operational site for deployment. Site deployment thus requires all the geological, engineering, economic and regulatory considerations of a site being taken into account, and operational storage capacity being determined. The various scales of site selection and the different levels of storage capacity estimation need to be integrated and the processes for these analyses should be iterative. Each level of detail progressively reduces uncertainty, but also results, inevitably, in a decrease of the defined storage volume. In addition, each level of detail in site selection requires greater effort, amounts and types of data, time and cost. The natural variability and geological, engineering and economic complexity of any potential CO2 storage site means that each site needs to be assessed individually; however, a similar workflow can be applied to all site evaluations. This report offers a consistent and systematic set of methodologies that can be used to assess and classify the CO2 storage potential of sites in Australia, New Zealand, and elsewhere. However it is important to emphasise that there are still gaps in our knowledge, and therefore in our ability to accurately determine bankable projects, that can only be filled through further studies, and by the actual field experience gained from additional demonstration and commercial projects.

39

References
Bachu, S, 1995. Flow of variable-density formation water in deep sloping aquifers: review of methods of representation with case studies. Journal of Hydrology, vol. 164, pp. 19-38. Bachu, S, 2000. Sequestration of CO2 in geological media: criteria and approach for site selection in response to climate change. Energy Conversion and Management, vol. 41 (9), pp. 953-970. Bachu, S, 2003. Screening and ranking of sedimentary basins for sequestration of CO2 in geological media in response to climate change. Environmental Geology, vol. 44, pp. 277-289. Bachu, S, Bonijoly, D, Bradshaw, J, Burruss, R, Holloway, S, Christensen, N P and Mathiassen, O M, 2007. CO2 storage capacity estimation: methodology and gaps. International Journal of Greenhouse Gas Control, vol. 1 (4), pp. 430-443. Bachu, S and Gunter, W D, 1999. Storage capacity of CO2 in geological media in sedimentary basins with application to the Alberta Basin. In: P Reimer, B Eliasson & A Wokaun (eds.) Greenhouse Gas Control Technologies: Proceedings of the Fourth International Conference on Greenhouse Gas Control Technologies, 30 August - 2 September 1998, Interlaken, Switzerland. Elsevier, pp. 195-200. Bachu, S, Gunter, W D and Perkins, E H, 1994. Aquifer disposal of CO2: hydrodynamic and mineral trapping. Energy Conversion and Management, vol. 35 (4), pp. 269-279. Bachu, S, Gunter, W D and Perkins, E H, 1996. Carbon dioxide disposal. In: B Hitchon (ed.) Aquifer Disposal of Carbon Dioxide. Geoscience Publishing Ltd., Sherwood Park, pp. 11-22. Bachu, S and Michael, K, 2002. Flow of variable-density formation water in deep sloping aquifers: minimizing the error in representation and analysis when using hydraulic-head distributions. Journal of Hydrology, vol. 259, pp. 49-65. Bennion, B and Bachu, S, 2005. Relative permeability characteristics for CO2 displacing water in a variety of potential sequestration zones in the Western Canada Sedimentary Basin. SPE 95547, 15 p. In: SPE Technical Conference and Exhibition SPE, Dallas, TX. Bergman, P D and Winter, E M, 1995. Disposal of carbon dioxide in aquifers in the U. S. Energy Conversion and Management, vol. 36 (6-9), pp. 523-526. Biddle, K T and Wielchowsky, C C, 1994. Hydrocarbon traps. In: L B Magoon & W G Dow (eds.) The Petroleum System - From Source to Trap. AAPG Memoir, 60. American Association of Petroleum Geologists, Tulsa, pp. 219-235. Bradshaw, B E, Bradshaw, J, Simon, G and Mackie, V, 2001. GEODISC research: carbon dioxide sequestration potential of Australia's coal basins. In: Proceedings of the 18th International Pittsburgh Coal Conference, 3-7 December 2001, Newcastle, NSW. Bradshaw, J, Bradshaw, B E, Allinson, G, Rigg, A J, Nguyen, V and Spencer, L, 2002. The potential for geological sequestration of CO2 in Australia: preliminary findings and implications for new gas field development. The APPEA Journal, vol. 42 (1), pp. 25-46. Bradshaw, J and Rigg, A J, 2001. The GEODISC Program: research into geological sequestration of CO2 in Australia. Environmental Geosciences, vol. 8 (3), pp. 166-176. Burruss, R, 2003. CO2 adsorption in coals as a function of rank and composition: a task in USGS research on geologic sequestration of CO2 [online]. The Coal-Seq II Consortium (Advanced Resources International Inc.). Available from: http://www.coal-seq.com/Proceedings2003/Burruss.pdf [Accessed 13.02.08].

40

Celia, M A and Bachu, S, 2003. Geological sequestration of CO2: is leakage unavoidable and acceptable? In: J Gale & Y Kaya (eds.) Greenhouse Gas Control Technologies: Proceedings of the 6th International Conference on Greenhouse Gas Control Technologies, Volume I, 1-4 October 2002, Kyoto, Japan. Elsevier Science, pp. 477-482. Cook, P and Sharma, S, 2006. Detailed geological site characterisation of a CO2 storage site: Otway Basin Pilot Project, Australia. In: Proceedings CO2SC Symposium 2006, 20-22 March, LBNL, Berkley, California, pp. 4. Cook, P J, 2006. Site Characterization. In: International Symposium on Site Characterization for CO2 Geological Storage, 20-22 March 2006, Berkeley, California, USA. Lawrence Berkeley National Laboratory (LBNL), pp. 3-5. Cook, P J, Rigg, A J and Bradshaw, J, 2000. Putting it back from where it came from: is geological disposal of carbon dioxide an option for Australia. The APPEA Journal, vol. 40 (1), pp. 654-666. CSLF, 2005. Phase I Final Report: Task Force for Review and Identification of Standards for CO2 Storage Capacity Measurement [online]. Carbon Sequestration Leadership Forum (CSLF) [Bradshaw, J, Bachu, S, Bonijoly, D, Burruss, R, Christensen, N P, Mathiassen, O M]. Available from: http://www.cslforum.org/documents/PhaseIReportStorageCapacityMeasurementTaskForce.pdf [Accessed 23 October 2007]. CSLF, 2007. Estimation of CO2 Storage Capacity in Geological Media - Phase II [online]. Prepared by the Task Force on CO2 Storage Capacity Estimation for the Technical Group (TG) of the Carbon Sequestration Leadership Forum (CSLF) [Bachu, S, Bonijoly, D, Bradshaw, J, Burruss, R, Christensen, N P, Holloway, S and Mathiassen, O M]. Available from: http://www.cslforum.org/documents/PhaseIIReportStorageCapacityMeasurementTaskForce.pdf [Accessed 1 May 2007]. Dahlberg, E C, 1995. Applied Hydrodynamics in Petroleum Exploration, 2nd Edition. Springer-Verlag, New York. Daniel, R F and Kaldi, J G, in press. Evaluating seal capacity of caprocks and intraformational barriers for CO2 containment. In: M Grobe (ed.) Carbon Dioxide Sequestration in Geological Media - State of the Art. AAPG Special Publication. American Association of Petroleum Geologists, Tulsa. Day, S, Duffy, G, Sakurovs, R and Weir, S, in press. Effect of coal properties on CO2 sorption capacity under supercritical conditions. International Journal of Greenhouse Gas Control. Day, S, Fry, R and Sakurovs, R, 2008. Swelling of Australian coals in supercritical CO2. International Journal of Coal Geology, vol. 74 (1), pp. 41-52. Dewhurst, D N, Jones, R M and Raven, M D, 2002. Microstructural and petrophysical characterization of Muderong Shale: application to top seal risking. Petroleum Geoscience, vol. 8 (4), pp. 371-383. DOE, 2006. Carbon Sequestration Atlas of the United States and Canada: Appendix A - Methodology for Development of Carbon Sequestration Capacity Estimates [online]. National Energy Technology Laboratory, Department of Energy. Available from: http://www.netl.doe.gov/publications/carbon_seq/atlas/Appendix%20A.pdf [Accessed 1 May 2007]. Ennis-King, J and Paterson, L, 2001. Reservoir engineering issues in the geological disposal of carbon dioxide. In: D J Williams, R A Durie, P McMullan, C A J Paulson & A Y Smith (eds.) Greenhouse Gas Control Technologies: Proceedings of the Fifth International Conference on Greenhouse Gas Control Technologies, 1316 August 2000, Cairns, Australia. CSIRO Publishing, pp. 290-295. Ennis-King, J and Paterson, L, 2002. Engineering aspects of geological sequestration of carbon dioxide. In: SPE Asia Pacific Oil and Gas Conference and Exhibition, 8-10 October 2002, Melbourne, Australia. Society of Petroleum Engineers, pp. SPE Paper 77809.

41

Ennis-King, J and Paterson, L, 2005a. Role of convective mixing in the long-term storage of carbon dioxide in deep saline formations. SPE Journal, vol. 10 (3), pp. 349-356. Ennis-King, J and Paterson, L, 2005b. Simulation of Geological Storage of Carbon Dioxide in the Offshore Gippsland Basin: Report for the Latrobe Valley Carbon Storage Assessment, CO2CRC. CSIRO Petroleum, Clayton. CO2CRC Report No. RPT05-0085. Faiz, M M, Saghafi, A, Sherwood, N and Wang, I, 2007. The influence of petrological properties and burial history on coal seam methane reservoir characterisation, Sydney Basin, Australia. International Journal of Coal Geology, vol. 70, pp. 193-208. Flett, M A, Gurton, R M and Taggart, I J, 2005. Heterogeneous saline formations: long-term benefits for geosequestration of greenhouse gases. In: E S Rubin, D W Keith & C F Gilboy (eds.) Greenhouse Gas Control Technologies: Proceedings of the 7th International Conference on Greenhouse Gas Control Technologies, Volume I, 5-9 September 2004, Vancouver, Canada. Elsevier, Oxford, pp. 501-509. Flett, M A, Taggart, I J, Lewis, J and Gurton, R M, 2003. Subsurface sensitivity study of geologic CO2 sequestration in saline formations. In: The Second Annual Conference on Carbon Sequestration, 5-8 May 2003, Alexandria, USA. National Energy Technology Laboratory, United States Department of Energy, pp. [CDRom]. Frailey, S M, Finley, R J and Hickman, T S, 2006. CO2 sequestration: storage capacity guideline needed. Oil & Gas Journal, vol. 104 (30), pp. 44-49. Gibson-Poole, C M, 2008. Site Characterisation for Geological Storage of Carbon Dioxide: Examples of Potential Sites from Northwest Australia. Thesis (PhD). The University of Adelaide. Gibson-Poole, C M, Edwards, S, Langford, R P and Vakarelov, B, 2006. Review of Geological Storage Opportunities for Carbon Capture and Storage (CCS) in Victoria. The Cooperative Research Centre for Greenhouse Gas Technologies (CO2CRC), Australian School of Petroleum, The University of Adelaide, Adelaide. ICTPL Consultancy Report Number ICTPL-RPT06-0506. Gibson-Poole, C M, Root, R S, Lang, S C, Streit, J E, Hennig, A L, Otto, C J and Underschultz, J R, 2005. Conducting comprehensive analyses of potential sites for geological CO2 storage. In: E S Rubin, D W Keith & C F Gilboy (eds.) Greenhouse Gas Control Technologies: Proceedings of the 7th International Conference on Greenhouse Gas Control Technologies, Volume I - Peer Reviewed Papers and Overviews, 5-9 September 2004, Vancouver, Canada. Elsevier, Oxford, pp. 673-681. Gibson-Poole, C M and Svendsen, L, 2005a. Regional Geological Evaluation of the Potential for Geological Storage of CO2 in the Gippsland Basin, Southeast Australia. Australian School of Petroleum, The University of Adelaide, Adelaide. CO2CRC Report No. RPT05-0048. Gibson-Poole, C M and Svendsen, L, 2005b. Reservoir Characterisation and Geological Model, Kingfish Field/Southern Oil Fields Area, Gippsland Basin, SE Australia: Implications for CO2 Storage. Australian School of Petroleum, The University of Adelaide, Adelaide. CO2CRC Report No. RPT05-0041. Gunter, W D, Bachu, S and Benson, S M, 2004. The role of hydrogeological and geochemical trapping in sedimentary basins for secure geological storage for carbon dioxide. In: S J Baines & R H Worden (eds.) Geological Storage of Carbon Dioxide. Geological Society Special Publication, 233. The Geological Society of London, London, pp. 129-145. Gunter, W D, Gentzis, T, Rottenfusser, B A and Richardson, R J H, 1997a. Deep coalbed methane in Alberta, Canada: a fuel resource with the potential of zero greenhouse gas emissions. Energy Conversion and Management, vol. 38 (Supplemental), pp. S217-S222. Gunter, W D, Perkins, E H and McCann, T J, 1993. Aquifer disposal of CO2-rich gases: reaction design for added capacity. Energy Conversion and Management, vol. 34 (9-11), pp. 941-948.

42

Gunter, W D, Wiwchar, B and Perkins, E H, 1997b. Aquifer disposal of CO2-rich greenhouse gases: extension of the time scale of experiment for CO2-sequestering reactions by geochemical modelling. Mineralogy and Petrology, vol. 59, pp. 121-140. Hendriks, C A and Blok, K, 1993. Underground storage of carbon dioxide. Energy Conversion and Management, vol. 34 (9-11), pp. 949-957. Hennig, A L and Otto, C J, 2007. A Hydrodynamic Characterisation of the Offshore Vlaming Sub-Basin (appendix 4 of CO2CRC Report No. RPT06-0162). Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. CO2CRC Report No. RPT05-0223. Hennig, A L, Underschultz, J R, Johnson, L C, Otto, C J and Trey, C, 2006. Hydrodynamic Characterisation of the Triassic Showgrounds Aquifer at the Wunger Ridge Site in Queensland: Assessing Suitability for CO2 Sequestration (Appendix 10.6.3 of CO2CRC Report No. RPT05-0225). Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. CO2CRC Report No. RPT05-0036. Holliday, D W, Williams, G M, Holloway, S, Savage, D and Bannon, M P, 1991. A preliminary feasibility study for the underground disposal of carbon dioxide in the UK. British Geological Survey Technical Report 1991. Holloway, S and Savage, D, 1993. The potential for aquifer disposal of carbon dioxide in the UK. Energy Conversion and Management, vol. 34 (9-11), pp. 925-932. Holloway, S and van der Straaten, R, 1995. The Joule II Project: the underground disposal of carbon dioxide. Energy Conversion and Management, vol. 36 (6-9), pp. 519-522. Holt, T, Jensen, J I and Lindeberg, E, 1995. Underground storage of CO2 in aquifers and oil reservoirs. Energy Conversion and Management, vol. 36 (6-9), pp. 535-538. Holtz, M H, 2002. Residual gas saturation to aquifer influx: a calculation method for 3-D computer reservoir model construction. In: SPE Gas Technology Symposium, 30 April-2 May 2002, Calgary, Alberta, Canada. Society of Petroleum Engineers, pp. SPE Paper 75502. Hooper, B, Murray, L and Gibson-Poole, C M, 2005. Latrobe Valley CO2 Storage Assessment - Final Report. CO2CRC Report No. RPT05-0108 [online]. Available from: http://www.co2crc.com.au/PUBFILES/OTHER05/LVCSA_FinalReport.pdf [Accessed 30 November 2006]. Hovorka, S D, Doughty, C, Benson, S M, Pruess, K and Knox, P R, 2004. The impact of geological heterogeneity on CO2 storage in brine formations: a case study from the Texas Gulf Coast. In: S J Baines & R H Worden (eds.) Geological Storage of Carbon Dioxide. Geological Society Special Publication, 233. The Geological Society of London, London, pp. 147-163. IPCC, 2005. IPCC Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group III of the Intergovernmental Panel on Climate Change [Metz, B, Davidson, O, de Coninck, H C, Loos, M and Meyer, L A (eds.)]. Cambridge University Press, New York. Islam, M R and Chakma, A, 1993. Storage and utilization of CO2 in petroleum reservoirs - a simulation study. Energy Conversion and Management, vol. 34 (9-11), pp. 1205-1212. Juanes, R, Spiteri, E J, Orr, F M and Blunt, M J, 2006. Impact of relative permeability hysteresis on geological CO2 storage. Water Resources Research, vol. 42 (12), pp. 1-13. Kalinowski, A, 2006. CO2 Storage Potential in the Southeast Bowen Basin (Appendix 10.4 of CO2CRC Report No. RPT05-0225. Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. CO2CRC Report No. RPT06-0042. Killingley, J, 1990. Gas desorption from coal. In: L Paterson (ed.) Methane Drainage from Coal. CSIRO, Melbourne, pp. 27-32.

43

Kim, A G and Kissell, F N, 1986. Methane formation and migration in coalbeds. In: M Duel & A G Kim (eds.) Methane Control Research: Summary of Results 1964-80. Bureau of Mines Bulletin, 687. U.S. Department of the Interior, pp. 18-25. Kirste, D M, Watson, M N and Tingate, P R, 2004. Geochemical modelling of CO2-water-rock interaction in the Pretty Hill Formation, Otway Basin. In: P J Boult, D R Johns & S C Lang (eds.) Eastern Australasian Basins Symposium II, Special Publication, 19-22 September 2004, Adelaide. Petroleum Exploration Association of Australia, Adelaide, pp. 403-411. Koide, H, Tazaki, Y, Noguchi, Y, Nakayama, S, Iijima, M, Ito, K and Shindo, Y, 1992. Subterranean containment and long-term storage of carbon dioxide in unused aquifers and in depleted natural gas reservoirs. Energy Conversion and Management, vol. 33 (5-8), pp. 619-626. Law, D H S and Bachu, S, 1996. Hydrogeological and numerical analysis of CO2 disposal in deep aquifers in the Alberta Sedimentary Basin. Energy Conversion and Management, vol. 37 (6-8), pp. 1167-1174. Lindeberg, E, 1997. Escape of CO2 from aquifers. Energy Conversion and Management, vol. 38 (Supplemental), pp. S235-S240. Lindeberg, E and Bergmo, P, 2003. The long-term fate of CO2 injected into an aquifer. In: J Gale & Y Kaya (eds.) Greenhouse Gas Control Technologies: Proceedings of the 6th International Conference on Greenhouse Gas Control Technologies, Volume I, 1-4 October 2002, Kyoto, Japan. Elsevier Science, pp. 489-494. Lindeberg, E and Wessel-Berg, D, 1997. Vertical convection in an aquifer column under a gas cap of CO2. Energy Conversion and Management, vol. 38 (Supplemental), pp. S229-S234. Masters, J A, 1979. Deep-basin gas trap, western Canada. AAPG Bulletin, vol. 63, pp. 152-181. McCabe, P J, 1998. Energy resources: cornucopia or empty barrel? AAPG Bulletin, vol. 82 (11), pp. 2110-2134. McPherson, B J O L and Cole, B S, 2000. Multiphase CO2 flow, transport and sequestration in the Powder River Basin, Wyoming, USA. Journal of Geochemical Exploration, vol. 69-70 (6), pp. 65-70. Murdoch University, 2006. Geothermal Energy Systems [online]. Available from: http://wwwphys.murdoch.edu.au/rise/reslab/resfiles/geo/text_files/image013.png [Accessed 23 October 2007]. Nordbotten, J M, Celia, M A and Bachu, S, 2005. Injection and storage of CO2 in deep saline aquifers: analytical solution for CO2 plume evolution during injection. Transport in Porous Media, vol. 58 (3), pp. 339360. Otto, C J, Underschultz, J R, Hennig, A L and Roy, V J, 2001. Hydrodynamic analysis of flow systems and fault seal integrity in the North West Shelf of Australia. The APPEA Journal, vol. 41 (1), pp. 347-365. Patchett, A, 2006. CO2 Storage Potential of Environmentally Sustainable Sites for Carbon Dioxide Injection (UEI37 and 39), Eastern Bowen Basin, Queensland (Appendix 10.3 of CO2CRC Report No. RPT05-0225). Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. RPT06-0037. Perkins, E H and Gunter, W D, 1995. Aquifer disposal of CO2-rich greenhouse gases: modelling of water-rock reaction paths in a siliciclastic aquifer. In: Y K Kharaka & O V Chudaev (eds.) Proceedings of the 8th International Symposium on Water-Rock Interaction, Rotterdam. Balkema, pp. 895-898. Perkins, E H and Gunter, W D, 1996. Mineral traps for carbon dioxide. In: B Hitchon (ed.) Aquifer Disposal of Carbon Dioxide: Hydrodynamic and Mineral Trapping - Proof of Concept. Geoscience Publishing Ltd, Sherwood Park, pp. 93-114. Rigg, A J, Allinson, G, Bradshaw, J, Ennis-King, J, Gibson-Poole, C M, Hillis, R R, Lang, S C and Streit, J E, 2001. The search for sites for geological sequestration of CO2 in Australia: a progress report on GEODISC. The APPEA Journal, vol. 41 (1), pp. 711-725.

44

Rigg, A J, Royal, B, Payenberg, T H D, Lang, S C, Bradshaw, J, Bradshaw, B E, Simon, G, Mackie, V, Miyazaki, S, Langford, R P, Kernich, A, Dance, T, Hennig, A L and Anderson, C, 2005. Final Report: Assessment of Some Potential CO2 Storage Areas, Eastern Australian Sedimentary Basins. Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. CO2CRC Report No. RPT05-0012. Rochelle, C A, Pearce, J M and Holloway, S, 1999. The underground sequestration of carbon dioxide: containment by chemical reactions in the deep geosphere. In: R Metcalfe & C A Rochelle (eds.) Chemical Containment of Waste in the Geosphere. Geological Society Special Publication, 157. The Geological Society of London, London, pp. 117-129. Saghafi, A, Faiz, M M and Roberts, M, 2007. CO2 storage and gas diffusivity properties of coals from Sydney Basin, Australia. International Journal of Coal Geology, vol. 70, pp. 240-254. Sakurovs, R, Day, S, Weir, S and Duffy, G, 2007. Application of a modified Dubinin-Radushkevich equation to adsorption of gases by coals under supercritical conditions. Energy & Fuels, vol. 21, pp. 992-997. Sakurovs, R, Day, S, Weir, S and Duffy, G, 2008. Temperature dependence of sorption of gases by coals and charcoals. International Journal of Coal Geology, vol. 73 (3-4), pp. 250-258. Sharma, S, Cook, P, Berly, T and Anderson, C, 2007. Australia's first geosequestration demonstration project: the CO2CRC Otway Basin Pilot Project. APPEA Journal, vol. 2007 (Part 1), pp. 259-270. SPE, 2007. Petroleum Resources Management System [online]. Oil and Gas Reserves Committee, Society of Petroleum Engineers. Available from: http://www.spe.org/spesite/spe/spe/industry/reserves/Petroleum_Resources_Management_System_2007.pdf [Accessed 3 December 2007]. Spencer, L K, 2004. Dampierland-Leveque Shelf, Northwest Canning Basin Area, Carbon Dioxide Storage Review. Cooperative Research Centre for Greenhouse Gas Emissions, Canberra. CO2CRC Report No. RPT050002. Stevens, S H, Kuuskra, V A and Gale, J, 2001. Sequestration of CO2 in depleted oil and gas fields: global capacity, costs and barriers. In: D J Williams, R A Durie, P McMullan, C A J Paulson & A Y Smith (eds.) Greenhouse Gas Control Technologies: Proceedings of the Fifth International Conference on Greenhouse Gas Control Technologies, 13-16 August 2000, Cairns, Australia. CSIRO Publishing, pp. 278-283. Streit, J E and Hillis, R R, 2004. Estimating fault stability and sustainable fluid pressures for underground storage of CO2 in porous rock. Energy, vol. 29 (9-10), pp. 1445-1456. Streit, J E and Siggins, A F, 2005. Predicting, monitoring and controlling geomechanical effects of CO2 injection. In: E S Rubin, D W Keith & C F Gilboy (eds.) Greenhouse Gas Control Technologies: Proceedings of the 7th International Conference on Greenhouse Gas Control Technologies, Volume I - Peer Reviewed Papers and Overviews. Elsevier, Oxford, pp. 643-651. USGS, 2007. World Seismicity Maps: Australia, Indonesia and New Zealand [online]. United States Geological Survey. Available from: http://earthquake.usgs.gov/regional/world/seismicity/australia.php [Accessed 23 October 2007]. van Bergen, F and Pagnier, H J M, 2001. A potentially clean energy cycle: CO2 sequestration in coal. InFormation, Netherlands Institute of Applied Geoscience TNO - National Geological Survey, vol. 7, pp. 13-15. van der Meer, L G H, 1992. Investigations regarding the storage of carbon dioxide in aquifers in The Netherlands. Energy Conversion and Management, vol. 33 (5-8), pp. 611-618. van der Meer, L G H, 1993. The conditions limiting CO2 storage in aquifers. Energy Conversion and Management, vol. 34 (9-11), pp. 959-966. van der Meer, L G H, 1995. The CO2 storage efficiency of aquifers. Energy Conversion and Management, vol. 36 (6-9), pp. 513-518.

45

Varma, S, Dance, T, Underschultz, J R, Langford, R P and Dodds, K, 2007. Regional study on Potential CO2 Geosequestration in the Collie Basin and the Perth Basin of Western Australia. Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. RPT07-0529. Vavra, C L, Kaldi, J G and Sneider, R M, 1992. Geological applications of capillary pressure: a review. The American Association of Petroleum Geologists Bulletin, vol. 76 (6), pp. 840-850. Watson, M N, Boreham, C J and Tingate, P R, 2004. Carbon dioxide and carbonate cements in the Otway Basin: implications for geological storage of carbon dioxide. The APPEA Journal, vol. 44 (1), pp. 703-720. Watson, M N and Gibson-Poole, C M, 2005. Reservoir selection for optimised geological injection and storage of carbon dioxide: a combined geochemical and stratigraphic perspective. In: The Fourth Annual Conference on Carbon Capture and Storage, 2-5 May 2005, Alexandria. National Energy Technology Laboratory, United States Department of Energy, pp. [CD-Rom]. White, C M, Smith, D H, Jones, K L, Goodman, A L, Jikich, S A, La Count, R B, DuBose, S B, Ozdemir, E, Morsi, B I and Schroeder, K T, 2005a. Sequestration of carbon dioxide in coal with enhanced coalbed methane recovery - a review. Energy & Fuels, vol. 19 (3), pp. 659-724. White, L, 2007. CO2 Storage Potential of the Northeast Bowen Basin, Queensland. Cooperative Research Centre for Greenhouse Gas Technologies, Canberra. CO2CRC Report No. RPT07-0654. White, S P, Allis, R G, Moore, J, Chidsey, T, Morgan, C, Gwynn, W and Adams, M, 2005b. Simulation of reactive transport of injected CO2 on the Colorado Plateau, Utah, USA. Chemical Geology, vol. 217, pp. 387405. Zimmerle, W, 1995. Petroleum sedimentology. Kluwer, Dordrecht.

46

Appendix A: CO2 Trapping Mechanisms

The subsurface flow behaviour of CO2, as well as the length of time involved, will influence the ultimate trapping mechanism. There are a number of different trapping mechanisms for geological storage of CO2, which include structural/stratigraphic trapping, hydrodynamic trapping, residual trapping, solubility trapping, mineral trapping and adsorption trapping. The following discussion is mostly extracted or modified from Gibson-Poole (2008).

Structural/Stratigraphic Trapping
Structural/stratigraphic trapping relates to the free-phase (immiscible) CO2 that is not dissolved in formation water. When supercritical CO2 rises upwards by buoyancy it can be physically trapped in a structural or stratigraphic trap (as a result of the CO2 being the non-wetting phase) in exactly the same manner as a hydrocarbon accumulation. The nature of the physical trap depends on the geometric arrangement of the reservoir and seal units. Common structural traps include anticlinal folds or tilted fault blocks (Figure 11a) and typical stratigraphic traps include those created by a lateral change in facies up-dip or a depositional pinch-out (Figure 11b). As with hydrocarbon accumulations, there are numerous variations of structural and stratigraphic traps, plus combinations of both structural and stratigraphic elements, that can provide physical traps for geological storage of CO2. In a dipping formation with no defined structural closure, any small bumps in the seal geometry will behave like small anticlinal structural traps and free-phase CO2 will fill these to the spill point (due to buoyancy) before migration continues (Bergman & Winter, 1995; Lindeberg, 1997; Ennis-King & Paterson, 2002).

Figure 11: Examples of (a) structural and (b) stratigraphic physical traps for CO2 (modified from Biddle & Wielchowsky, 1994).

Hydrodynamic Trapping
In a storage project, supercritical CO2 will be injected as a single phase, but once in the geological formation it will partition into free-phase (immiscible) CO2 and a CO2-rich brine. The flow of the freephase CO2 is dependant on the dip of the sealing horizon and the flow velocity and direction of the in situ formation water. In horizontal or gently dipping reservoirs, this can lead to very long residence times, of the order of thousands to millions of years (Figure 12) (Bachu et al., 1994). The dissolved CO2 will travel with the moving formation water. Saline aquifers generally have very low flow velocities, of the order of tens of cm/year. This slow flow velocity leads to residence times of millions of years. This geological timescale trapping of CO2 in deep regional aquifers is called hydrodynamic trapping and avoids the need for

47

structural or stratigraphic traps (Bachu et al., 1994). This trapping mechanism can be considered a rate seal, as opposed to traditional rock seal traps associated with structural and stratigraphic traps.

Figure 12: Hydrodynamic trapping of CO2, where the CO2 migration pathway is 10s to 100s km long allowing for a long residence time.

Limited work on hydrodynamic trapping for CO2 storage has been done in Australia. Notable are the studies by Hennig et al. (2006) on the Showgrounds Aquifer at the Wunger Ridge site in Queensland and Henning & Otto (2007) on the offshore Vlaming Sub-basin of Western Australia During the injection phase of a storage project (of the order of 20 to 40 years), the migration of the CO2 away from the injection well involves both gravity override and viscous fingering, as a result of the CO2 being significantly less dense and less viscous than the saline formation water under typical reservoir conditions (Ennis-King & Paterson, 2001; Nordbotten et al., 2005). Simulations by Holt et al. (1995) determined that the rate at which the CO2 is injected can have an impact on this displacement behaviour. At high injection rates, viscous forces dominate and the CO2 plume develops rapidly along the most permeable paths. At lower injection rates gravity forces dominate and buoyancy causes the CO2 to rise upwards. The heterogeneity in the permeability distribution also has a significant impact on the flow behaviour of injected CO2. If the degree of heterogeneity is large, particularly with respect to the ratio between horizontal and vertical permeability, then channelling of CO2 along the most permeable paths will be the dominant flow behaviour (van der Meer, 1995; Ennis-King & Paterson, 2001). After cessation of injection, migration of the CO2 plume continues during the relaxation of the pressure gradient driving the lateral migration. The CO2 then continues to migrate under the influence of gravity (or buoyancy) and the natural formation water flow (Ennis-King & Paterson, 2001, 2002; Flett et al., 2003; Flett et al., 2005). The dissolved CO2 moves with the formation water. The relationship between these two phases determines which storage mechanism dominates and how far the CO2 will migrate from the injection site before no free-phase, mobile CO2 remains. The longer the migration pathway and the slower the migration rate, the more CO2 will become trapped residually or in solution such that eventually no mobile, free-phase CO2 will exist in the system, as modelled in various numerical simulations, (McPherson & Cole, 2000; Ennis-King & Paterson, 2005a). The length of the migration pathway is controlled by geological parameters such as stratigraphic heterogeneities, for example, intraformational siltstones, shales and coals, which play a crucial role in determining the tortuosity of the CO2 flow path. The intraformational seals act as barriers or baffles to the buoyancy-driven upward flow and induce lateral migration until CO2 is able to breach the barrier or get past it (Ennis-King & Paterson, 2002; Hovorka et al., 2004; Flett et al.,

48

2005). The flow velocities are dependant on a number of factors, including the fluid properties (both freephase CO2 and formation water), the mechanism of transport (diffusion, dispersion, convection), buoyancy, rate of dissolution, the hydrodynamic drive and direction, permeability, tortuosity, the impact of any geochemical reactions, temperature (which affects viscosity and density), structural dip, relative permeability, and so on.

Residual Trapping
Residual trapping occurs when the CO2 becomes trapped in the pore space as a residual immobile phase by capillary forces (Figure 13) (Ennis-King & Paterson, 2001; Flett et al., 2005). At the tail of the migrating CO2 plume, imbibition processes are dominant as the formation water (wetting-phase) imbibes behind the migrating CO2 (non-wetting phase). When the concentration of the CO2 falls below a certain level it becomes trapped by capillary pressure forces and ceases to flow (Ennis-King & Paterson, 2001; Holtz, 2002; Flett et al., 2003; Flett et al., 2005). Therefore, a trail of residual, immobile CO2 is left behind the plume as it migrates upward (Juanes et al., 2006). Residual CO2 saturation values vary between 530 % based on typical relative permeability curves (Ennis-King & Paterson, 2001). Over time, the residually trapped CO2 dissolves into the formation water (Ennis-King & Paterson, 2001; Flett et al., 2005).

Figure 13: Residual trapping of CO2.

Solubility Trapping
Solubility trapping relates to the CO2 dissolved into the formation water (Koide et al., 1992). Carbon dioxide solubility increases with increasing pressure and decreases with increasing temperature and water salinity. Carbon dioxide may mix with, and then dissolve in, formation water through the processes of diffusion, dispersion and convection. Thus, time is an important factor in the solubility trapping of CO2. With increasing time, more CO2 dissolves into the formation water (Bachu et al., 1994; Ennis-King & Paterson, 2001). As the CO2 continues to dissolve into the formation water, it can lead to a phenomenon known as convective mixing. The density of the CO2-saturated water increases to become about 1 % more than that of the unsaturated water when CO2 dissolves into the formation water. The dense CO2-saturated water overlying less dense unsaturated water creates a density instability and plumes of CO2-rich water flow downward (Figure 14) (Lindeberg & Wessel-Berg, 1997; Ennis-King & Paterson, 2002; Lindeberg & Bergmo, 2003; Ennis-King & Paterson, 2005a). The time-scale for complete dissolution is critically dependent on the vertical permeability and the geometry of the top seal, but is predicted to occur on a scale of hundreds to thousands of years (Ennis-King & Paterson, 2002). The success of solubility trapping is aided by slow flow velocities of the in situ formation water, as this will lead to a greater residence time and allow more time for CO2 to

49

dissolve into the formation water (Koide et al., 1992; Bachu et al., 1994). Stratigraphic heterogeneities also improve solubility trapping, as they increase the tortuosity of the CO2 migration path and accordingly the CO2 is able to contact larger volumes of formation water into which it can dissolve (Lindeberg, 1997; Hovorka et al., 2004; Flett et al., 2005). Convective mixing (resulting from density instability) is important for solubility trapping of CO2, as it is orders of magnitude faster than pure diffusion mechanisms, and so accelerates the overall dissolution of the CO2 into the formation water (Ennis-King & Paterson, 2002, 2005b)

Figure 14: Convective mixing of CO2: example of a numerical simulation showing the high-density plumes of CO2-saturated brine (grey colours) sinking into the brine column below (white colour) (kv/kh = 0.01, after 14400 years) (Ennis-King & Paterson, 2005a).

Mineral Trapping
Mineral trapping results from the precipitation of new carbonate minerals (Gunter et al., 1993). This storage mechanism is the most permanent of the trapping types discussed as it renders the CO2 immobile (Bachu et al., 1994, 1996). The time-scale for mineral precipitation is typically long, of the order of tens to thousands of years, depending on the initial minerals present (Perkins & Gunter, 1995; Kirste et al., 2004). Siliciclastic reservoirs are favoured over carbonate reservoirs, in particular calcium-, magnesium- or ironrich siliciclastic reservoirs, as they have the best potential for mineral trapping of CO2 (Gunter et al., 1993; Bachu et al., 1994; Perkins & Gunter, 1995; Gunter et al., 1997b). The potential for mineral trapping is quite variable from formation to formation and thus needs to be examined as part of the site characterisation process for storage site selection. Mineral trapping is a function of the mineralogy of the reservoir rock, the chemical composition of the formation water and the formation temperature and pressure. In addition, potential reactions depend on the contact surface (interface) between the mineral grains and the formation water containing dissolved CO2, and on the flow rate of fluids through the rock (Gunter et al., 2004). Due to the specific information required to model mineral storage of CO2, it is appropriate that it be undertaken as part of site characterisation rather than basin-scale assessment (CSLF, 2007). Detailed modelling of injection of CO2 into the White Rim Sandstone, Utah, using the reactive chemical simulator ChemTOUGH, found that 1000 years after the 30 year injection period began approximately 21 % of the injected CO2 was permanently stored as a mineral and 52 % was free-phase gas or dissolved in the groundwater (White et al., 2005b). As in residual and solubility trapping, mineral trapping is a time-dependent process operating on the scale of millennia (Bachu et al., 2007).

50

Storage Security for Saline/Hydrocarbon Formations

In any geological storage site, the injected CO2 will ultimately be trapped by a number of the mechanisms described above. The type of trapping that occurs, and when, is dependent on the dynamic flow behaviour of the CO2 and the time-scale involved. With increasing time, the dominant storage mechanism will change and typically the storage security also increases. Figure 15 is a simple representation of CO2 storage and how the trapping mechanism might alter over time. For example, the initial storage mechanism will dominantly be physical structural and stratigraphic trapping of the immiscible-phase CO2. With increasing time and migration, more CO2 is trapped residually in the pore space or is dissolved in the formation water, increasing the storage security. Finally, mineral trapping may occur by precipitation of carbonate minerals after geochemical reaction of the dissolved CO2 with the host rock mineralogy, permanently trapping the CO2 (IPCC, 2005).

Figure 15: Schematic representation of the change of dominant trapping mechanisms and increasing CO2 storage security with time (from IPCC, 2005).

Adsorption Trapping on Coal

CO2 can be stored in coal seams as free-phase gas in pore spaces, as gas in solution, or as gas directly adsorbed onto internal surfaces of fractures (cleats) in coal (Kim & Kissell, 1986). Adsorption trapping is the chief gas storage mechanism in coal seams and is dependant upon the gas type, temperature and pressure conditions, the coal rank, moisture content, maceral composition and mineral matter content (White et al., 2005a). In some coals, the sorption capacity for CO2 exceeds 10% by weight (Day et al., 2008). A coals potential to adsorb gas is measured experimentally by a gas adsorption isotherm, that is normally determined either by a gravimetric or volumetric method at a constant temperature but varying pressures (Saghafi et al., 2007). The isotherm describes the equilibrium gas adsorption capacity of a particular coal for varying (pore) pressure conditions. Coal has a much higher gas sorption capacity for CO2 than for CH4 or nitrogen (N2) (Figure 16). Commonly, for bituminous coals the adsorption capacity for CO2 is 1.5 to 2 times that of CH4 (Faiz et al., 2007; Saghafi et al., 2007). Lower rank and geologically younger coals may adsorb up to 10 times the sorption capacity of CO2 over CH4, but they also commonly contain lower initial total gas contents due to their immaturity. Adsorption onto coal can also have a strong time component, since initial adsorption of a given quantity of gas can be followed by absorption (penetration of molecules of gas into the mass) and structural rearrangement of the coal (White et al., 2005a).

51

Figure 16: Pure gas absolute adsorption in standard cubic meters per tonne on Tiffany coals at 55 C (from IPCC, 2005).

52