CO 2 storage potential of basaltic rocks, Mpumalanga: Implications for the Just Transition

targeted reducing CO 2 emissions by more than half in the next 10 years. To meet climate change mitigation scenarios, while alleviating continued emissions, South Africa will look to technologies such as carbon capture, utilisation and storage. Initial assessments of South Africa’s potential for CO 2 storage have focused on deep saline aquifers within volcano-sedimentary sequences along the near and offshore regions. Sustaining the Just Transition will, however, require additional storage capacity. In this study, we make an initial assessment of possible CO 2 storage in basaltic sequences of the Ventersdorp Supergroup. Geological and mineralogical information was ascertained from borehole data. The geological information suggests that the subsurface extent of the Ventersdorp Supergroup is at least 80 000 km 2 larger than previously mapped, extending beneath major point-source CO 2 emitters and active coalfields. Furthermore, petrographic analyses suggest pore space of up to ca 15% with minimal alteration, and preservation of mafic silicate minerals that would enable reactive carbonation of injected CO 2 . Notable metasomatic and hydrothermal alteration is confined to significant contact horizons, such as the lowermost Ventersdorp Contact Reef. These results suggest that basaltic sequences may exponentially increase South Africa’s CO 2 sequestration storage capacity and may have a significant impact on the country’s Just Transition. This study shows that basaltic sequences may support the permanent storage of anthropogenic CO 2 in South Africa, in particular, proximal to significant point-source CO 2 emitters. South Africa has voluminous and widespread basaltic sequences, which, in combination, increase South Africa’s geological CO 2 storage potential by several orders of magnitude. These storage reservoirs can have a direct impact in South Africa by enabling a sustainable Just Transition toward a low-carbon economy while meeting intended climate change mitigation scenarios. rocks are globally voluminous 14 ; have unique trapping mechanisms linked to their multi-phase geodynamic emplacement 15 ; and have a chemical composition that is highly susceptible for mineral carbonation on a large scale and which is several orders of magnitude faster than in classical siliciclastic reservoirs 16 . South Africa has extensive basaltic occurrences across the country 17 ; however, these have not been investigated as potential CO 2 storage reservoirs. We present the findings of geological and mineralogical analyses conducted on Ventersdorp Supergroup samples collected from boreholes proximal to major point-source CO 2 emitters and discuss the implications for support of larger-scale CO 2 sequestration.


Introduction
South Africa is the leading CO 2 emitter on the African continent and has one of the largest rates of CO 2 emissions in the world. 1 These emissions largely stem from the nation's heavy reliance on coal as a primary energy-generation feedstock. 2 South Africa has extensive geological sedimentary basins that contribute toward proven reserves of at least 30 billion tonnes of coal. This makes South Africa one of the largest coal producers globally. 3 South Africa has used these coal resources to effectively impact the country's industrialisation and socio-economic development. The coal industry employs at least 100 000 people and contributes a third of mining's total contribution to the country's GDP. 4 However, the South African government has underscored the need to combat climate change and has a target of reducing CO 2 emissions by at least 50% within the next 10 years (Figure 1). 5 It aims to do this through a drastic reduction in coal-fired energy and shifting toward alternative forms of energy. 6 However, with coal forming such a critical role in South Africa's socio-economic and energy landscape, this shift cannot be immediate and requires a careful Just Transition, i.e. a transition that will enable South Africa's intended climate change mitigation strategies while limiting potential negative socio-economic effects associated with the coal and linked industries. 7 Balancing South Africa's coal industry and climate change mitigation will therefore require the implementation of innovative and novel technologies, such as carbon capture, utilisation and storage. 8 Carbon capture, utilisation and storage (CCUS) technologies aim to reduce atmospheric CO 2 emissions by capturing CO 2 at the source (e.g. point-source emitters such as coal-fired plants), and transporting and storing the captured CO 2 in underground storage reservoirs. Some of the captured CO 2 may be used in additional downstream industries, e.g. various petrochemical processes. 9 CCUS investigations in South Africa have typically looked at deep saline aquifers, relatively deep coal seams, and depleted oil and gas fields as potential storage reservoirs. In combination, these amount to approximately 150 gigatons of potential CO 2 storage, with much of this potential (98%) located within offshore volcano-sedimentary basins. 10 Critically, these large potential storage reservoirs are far removed from South Africa's coalfields and CO 2 emission hotspots. In general, South Africa's coal reserves, current coal utilisation and subsequently most CO 2 emissions occur in the northeast of the country. 4 If South Africa is to enable a successful Just Transition, potential CO 2 storage reservoirs proximal to these coalfields and emission hotspots must be investigated and developed.
Basaltic rocks -that is, rocks rich in iron, calcium, magnesium, and aluminium silicate minerals -are regarded as very promising CO 2 storage reservoirs. [11][12][13] This is largely because basaltic rocks are globally voluminous 14 ; have unique trapping mechanisms linked to their multi-phase geodynamic emplacement 15 ; and have a chemical composition that is highly susceptible for mineral carbonation on a large scale and which is several orders of magnitude faster than in classical siliciclastic reservoirs 16 . South Africa has extensive basaltic occurrences across the country 17 ; however, these have not been investigated as potential CO 2 storage reservoirs. We present the findings of geological and mineralogical analyses conducted on Ventersdorp Supergroup samples collected from boreholes proximal to major point-source CO 2 emitters and discuss the implications for support of larger-scale CO 2 sequestration.

Onshore geological CO 2 storage in South Africa
Several critical geological conditions must be met when defining a safe and permanent prospective onshore CO 2 storage reservoir. These conditions include the existence of an appropriately thick geological sequence that is adequately permeable and porous to enable the injection and controlled movement of CO 2 -bearing fluids throughout the target reservoir. The reservoir should consist of components that are naturally amenable to reacting with the injected CO 2 . This reactivity is critical to ensure that the injected CO 2 converts to a solid form and is permanently stored. Geological sequences should also be relatively homogeneous across the targeted reservoir to accommodate a large volume of injected CO 2 . The targeted reservoir should be under-and overlain by adequately thick impermeable and non-porous geological sequences. These sequences will act as reservoir seals to restrict the movement and possible escape of the injected CO 2 . The region should be relatively tectonically undeformed. This would limit the presence of seismicity and geological structures that may promote undue movement of the injected CO 2 . 19 Environmental baseline investigations are critical. The target region requires an extensive and accurate assessment of the natural conditions, including understanding the biosphere, hydrosphere, and pedosphere. This information is needed to monitor any potential deviation from natural conditions during and after the injection of CO 2 . Once injected, the CO 2 can be stored either within the reservoir pore and/or mineral spaces, various geological structures or within saline fluids present at the injected depth. The injected CO 2 might react with reservoir lithologies and in-situ fluids and be converted to carbonate minerals. 20 Conversion to carbonate minerals is ideal because the CO 2 is rendered immobile and is permanently stored. Globally, the vast majority of current CCUS projects target storage of CO 2 in deep saline aquifers with a significant proportion of the captured CO 2 utilised to enhance oil and gas recovery. 9 Furthermore, research into South Africa's potential onshore CO 2 storage reservoirs has largely focused on conventional deep saline aquifer storage in relatively young (i.e. Palaeozoic and younger) sedimentary sequences. 21,22 Much of these sequences are located along the near-shore and offshore along the South African coastline. 10,23 Consideration of basaltic rocks as a potential CO 2 reservoir is relatively new. 11 Basaltic reservoirs are promising because the mineralogy of basaltic rocks is most amenable to support injected CO 2 being converted to carbonate minerals. Basaltic rocks have minerology that is rich in iron, calcium, magnesium, and aluminium. These react readily with CO 2 and produce carbonate minerals such as calcite, dolomite, and ankerite. 24,25 Furthermore, mineral carbonation in basaltic rocks occurs several orders of magnitude faster than in conventional siliciclastic reservoirs. 11,12 While South Africa does have extensive basaltic sequences, these have not been considered as potential CO 2 storage reservoirs.

The Ventersdorp Supergroup
The Ventersdorp Supergroup is one of the largest and oldest volcanosedimentary successions on earth. These successions are largely located in the northern part of South Africa, including proximal to the large coalfields and major point-source CO 2 emitters ( Figure 2). This makes it a viable option for further investigation as a potential basaltic CO 2 reservoir.
Vast sequences of mafic to ultramafic volcanic rocks interspersed with felsic and siliciclastic sequences dominate the Ventersdorp Supergroup ( Figure 3). These rocks were emplaced as a Large Igneous Province during Mesoarchean intraplate tectonic activity atop the Kaapvaal Craton. 26 The various volcanic assemblages cover a surface area of more than 200 000 km 2 and attain a thickness of more than 5 km. 27 Fg, Mg and Cg refers to fine-grained, medium-grained and coarse-grained, respectively. Geological information gathered from various sources. 26,29,[36][37][38] Images from samples at the Council for Geoscience National Core Library. 28 Stratigraphically, the lowermost part of the Ventersdorp Supergroup is defined by ca 2 km thick continental flood basalt assemblage of the ca 2791-2779 Ma Klipriviersberg Group. 29 This comprises a wide range of volcanic rocks including fine-coarse grained mafic and ultramafic sequences, agglomerates, and vesicular porphyritic lavas. The sequences of the Ventersdorp Supergroup are relatively undeformed and were subjected to low-mid grades of greenschist facies metamorphism, as inferred from the underlying Witwatersrand and overlying Transvaal Supergroups. 31 Brittle structures and associated vertical and lateral offset are recorded. 32 These structures are typically associated with deep crustal features, such as the Colesburg and Makhonjwa Lineaments (Figure 2). These structures formed during the amalgamation of the Kaapvaal Craton during the Neoarchean 33 , and would have undergone several phases of convergent and extensional reactivation linked to continental cycles 34 . Importantly, these structures also form critical zones that enabled stress accumulation and the development of proximal structures where much of the regional deformation would be exhibited. 35

Evaluating a potential reservoir in the Ventersdorp Supergroup
Several important considerations are needed for the sequences of the Ventersdorp Supergroup to be considered as a potential CO 2 storage reservoir. These considerations include determining if these sequences are adequately thick and laterally extensive and whether there is an appropriate amount of stratigraphic heterogeneity to enable the development of reservoir and corresponding seal lithologies. Furthermore, these potential sequences should be relatively devoid of certain kinds of regional structures that may enable the possible escape and/or migration of injected CO 2 and possible undue interaction with other natural systems, e.g. groundwater aquifers. To enable these estimations, several boreholes with a depth range of ca 1000-2000 m were considered in this study (Table 1; Figure 2). These were logged and scanned with a SisuROCK hyperspectral core scanner at the Council for Geoscience National Core Library. Selected samples were also collected along some of the boreholes for various petrographic and mineralogical analyses (Figure 2; Figure 4).

Hyperspectral borehole scanning
Selected geological boreholes were used to develop a ca 20-km long geological profile and to ascertain the subsurface extent of the Ventersdorp Supergroup sequences proximal to South Africa's significant point-source CO 2 emitters ( Figure 2). Furthermore, high-resolution mineral spectral information was gathered from the selected boreholes through hyperspectral borehole scanning. Hyperspectral scanning of the boreholes enables reflectance spectroscopy to be undertaken across different spectral regions, including the visible-near infrared, short-wave infrared, and long-wave infrared. The spectral absorption characteristics of various elements enables the identification of mineralogy and their associated textures. Spectral processing includes the development of dominant mineral maps and various mineral indices. This includes high band ratio ranges of mean depth and wavelength spectral signatures in the long-wave infrared, which specifically responds to the occurrence of mafic silicate minerals, i.e. those critical to support potential reactivity with CO 2 . In addition, albedo reflectance was also considered. This denotes regions with higher concentration of felsic mineralogy. This information combines to provide critical subsurface delineation of the geology and dominant mineralogy. Borehole BH 2068 was considered and scanned as a reference borehole. The results are shown in Figure 5.

Petrography
Petrographic analyses were undertaken on selected samples of sequences of the Ventersdorp Supergroup ( Figure 6). This analysis was conducted to test their characteristics and potential as possible CO 2 storage sequences. Petrographic analyses were conducted in both transmission light microscopy and automated scanning electron microscopy with energy dispersive x-ray spectrometry. These analyses enable the development of micron-scale compositional and mineral maps of the samples collected. The large depth of field range and magnification enables robust characterisation of pore space size and quantification. Quantification of the elements present allows for a more accurate mineral classification based on the stoichiometry to produce a fully quantified and detailed mineral map. This also enables the identification and quantification of the mafic and alteration minerals, including any natural carbonation. In this study, carbon-coated sample mounts were analysed using a Carl Zeiss Sigma 300 VP FEG-SEM with an energy dispersive x-ray spectroscopy and backscatter electron detector.

Subsurface geology
The geological profile suggests extensive occurrences of the Karoo, Transvaal, Ventersdorp and Witwatersrand Supergroups across the study area. Sequences of the Karoo Supergroup are generally homogeneous in thickness and extent across the study area. The topmost sequences of the Transvaal Supergroup are also correlative across the study area; however, there is significant heterogeneity along the basal part. There is also a notable thinning of the Transvaal Supergroup toward the south. The Ventersdorp and underlying Witwatersrand Supergroups are also generally well correlated across the study area. However, there is a significant thinning and thickening of these sequences toward the south and north, respectively. Stratigraphically, the boreholes suggest that, generally, only the lower part of the Ventersdorp Supergroup is preserved around the study area, i.e. sequences of the Klipriviersberg Group, while sequences of the Platberg and Pniel groups are poorly preserved.
The subsurface geological profile is well supported by the hyperspectral borehole scanning. The significant changes across the various supergroups are well delineated through the different spectral signatures. Spectral signatures suggest distinctive zones with high concentrations of mafic silicate minerals throughout the Ventersdorp Supergroup sequence. Furthermore, it also highlights zones with higher felsic mineralogy. Importantly, the hyperspectral signatures also provide constraints on the degree of alteration. The results suggest that alteration of the Ventersdorp Supergroup lavas increases toward the very base of the sequence, with appreciable signatures of metasomatic and hydrothermal minerals along the contact zone with the underlying Witwatersrand Supergroup.

Mineralogy
The results of the petrographic analyses correspond well to the reported lithodemic sequences of the lower part of the Ventersdorp Supergroup. 26,29,[36][37][38] Lithologies underlying the study area are largely comprised of mafic to ultramafic basaltic sequences that vary between fine-grained lavas to coarser porphyritic and amygdaloidal sequences. In general, the volume of mafic minerals and coarsening of texture increases toward the base of the sequence. There is also a correlation between the estimated pore space and the texture, with the coarsergrained sequences highlighting relatively increased pore space. The presence of secondary minerals such as chlorite, epidote, and calcite highlights hydrothermal and metasomatic alteration. In general, alteration is relatively minimal with an increase toward the base of the sequence. Importantly, where alteration is low, iron-, magnesium-, and calcium-rich silicate minerals (i.e. euhedral pyroxenes) are still preserved.

Conclusion and implications for South Africa's Just Transition
Enabling a successful Just Transition in South Africa is not straightforward. Currently, South Africa has a very strong reliance on coal, although the country is targeting significant reductions in CO 2 emissions. The successful implementation of CCUS technologies is crucial to support the Just Transition and a shift toward a low-carbon economy. Existing CCUS studies in South Africa suggest limited onshore CO 2 storage potential, with much of this restricted to relatively young volcano-sedimentary sequences that are far removed from South Africa's significant CO 2 emitters and coalfields. 10,23 Basaltic sequences are showing promise as additional CO 2 storage reservoirs. This is crucial because basaltic sequences are globally extensive and have mineralogy that is highly reactive and may rapidly enable the mineralisation of injected CO 2 into an array of carbonate minerals. [11][12][13] In this study, we considered the volcanic sequences of the Ventersdorp Supergroup as a potential storage reservoir of anthropogenic CO 2 .
The known extent of the Ventersdorp Supergroup has been increased from ca 200 000 km 2 38 to at least 280 000 km 2 . Importantly, the additional surface area extends beneath South Africa's highly developed coalfields and largest point-source CO 2 emitters ( Figure 2). Moreover, subsurface geological profiles highlight volcanic sequences that are adequately thick and at depths appropriate to support potential CO 2 storage. Furthermore, hyperspectral information suggests that these sequences are adequately heterogeneous in developing layers of potential reservoir and accompanying sealing lithologies ( Figure 5). The volcanic sequences thin significantly north of the study area ( Figure 4). This thinning is likely due to the presence of underlying structural and geological controls. This possibly forms a significant structural trap for the storage of CO 2 , but also suggests that any potential development of CCUS targeting the Ventersdorp Supergroup cannot extend further north from the study area.
Despite these lavas being emplaced during the Archean, alteration is limited and not widespread throughout the sequence ( Figure 6). Metasomatic alteration and carbonation of reactive minerals (e.g. iron-, calcium-, magnesium-, and aluminium-rich silicates) occurs toward the base of the sequence, near the contact zone with the underlying Witwatersrand Supergroup. This is likely linked to the occurrence of the Ventersdorp Contact Reef, forming the boundary zone between the Ventersdorp Supergroup and the underlying Witwatersrand Supergroup. In general, more coarse-grained porphyritic lavas have relatively higher porosity estimates (i.e. ideally ranging between ca 15% and 30%) that could support the uptake of injected CO 2 . 39 South Africa has an extensive geological evolution that saw the emplacement of several Large Igneous Provinces (Figure 7). This includes extensive basaltic sequences emplaced within the Makhonjwa (Barberton) Supergroup 40  Before geological storage of CO 2 in basaltic sequences can occur, several aspects need to be addressed and further investigated: • The subsurface extent of these sequences needs to be adequately delineated (Figure 7). This process includes establishing the lithodemic variations, inclusive of mineralogy, texture, and various degrees of alteration. This information is pertinent toward attaining precise volumetric estimations of potential basaltic CO 2 storage and their occurrence relative to other geological sequences of interest, e.g. those hosting various mineral occurrences, and the geographic location relative to present and predicted future CO 2 emission sources.
• A consolidation of legacy and new baseline data is needed. This includes attaining an adequate understanding of the natural conditions around prospective CO 2 storage sites, especially proximal to regions that have been subject to long-standing mining and exploration activities. This kind of information will also assist in developing tangible CO 2 utilisation considerations that may contribute toward environmental remediation.
• Implications for enabling a sustainable and inclusive Just Transition need to be determined. Global development and advancement of CCUS technologies are increasing. With the inclusion of basaltic storage, global volumetric CO 2 storage potential will significantly increase. This will need to be considered within the context of the Just Transition toward a low-carbon economy, especially in developing countries that are heavily reliant on fossil fuels, especially hydrocarbons. These considerations should be used to develop/revise current long-term sustainable energy development strategies, in particular those linked to carbon-reduction measures, e.g. carbon taxing. 46