Introduction

Climate researchers and governments worldwide are not in denial of the reality of climate change, and the main culprit is man’s continual dependence on fossil fuels, with the 2015 Paris Agreement setting the world on course to drastically cut down emissions of anthropogenic carbon dioxide (CO2) and reducing global warmth increment under 1.5 °C (UNFCCC, 2015; IPCC, 2018). As the International Energy Agency (IEA) estimated, the energy demand could escalate by well-nigh 4% by 2030, with the demand mainly satisfied by fossil fuels (Yelebe & Samuel, 2015).

Around 90% of South Africa’s vital energy is satisfied by petroleum derivatives, and coal delivers 92% of power generation countrywide (South Africa Department of Energy, 2009). South Africa has vast coal reservoirs used predominantly in power generation, production of liquid fuels and a direct supply of heat and steam in various industrial processes. Presently, emissions of CO2 are thought to exceed 400 Mt (million tons) per year in South Africa (Cloete, 2010). One of the specialized methodologies that can be utilized to moderate worldwide environmental change in non-renewable energy-centered nations such as South Africa is carbon capture, utilization and storage (CCUS) (Anastassia et al., 2010; Viljoen et al., 2010; Chabangu et al., 2014a, 2014b; Tsuji et al., 2014; Kempka et al., 2017; Bandilla et al., 2019; Yan & Zhang, 2019; Alcalde et al., 2021).

Storage of CO2 in deep geological formations involves capturing and separating from an industrial source, onward conveyance and injection into underground geologic reservoirs for permanent storage. Finally, it is measured, monitored and verified that it stays in the storage formation (Würdemann et al., 2010). From Meer (1995), Kumar et al. (2005), Teletzke & Lu (2013), Ojo & Tse (2016), Bui et al. (2018) and Alcalde et al. (2021), the typical geological indicators for the perfect spot for CO2 storage involve:

  • A reservoir rock or unit (such as hydrocarbon-depleted reservoirs, deep saline aquifers, coal seams, and salt caverns), with sufficient porosity and permeability, allowing injection and persistent storage of CO2.

  • The reservoir units occur at depths exceeding 800 m from mean sea level, ensuring reservoir pressure and temperature conditions (typically 73.7 bar and 31 °C geometric and geothermal gradient) allow CO2 existence in a dense supercritical form.

  • An impermeable rock acts as a seal above the reservoir, preventing upward migration of CO2.

The injection and geological storing of CO2 have been utilized for a considerable time to enhance optimization in oil and gas fields, while storage of gas and other substances in geological reservoirs has likewise been ongoing for decades (Viljoen et al., 2010; Vincent et al., 2013). Equivalent geological factors keeping commercial quantities of hydrocarbon in the subsurface for geologic time are currently being applied for CO2 storage. Depleted petroleum fields can geologically host CO2 because their hydrocarbon retention ability has been demonstrated, as they have held barrels of hydrocarbon for geologic time, with accompanying massive geological and engineering data gathered from the fields for detailed reservoir characterization or while in production to improve oil and gas recovery (Alcalde et al., 2019, 2021; Ghanbari et al., 2020).

The drive to attain accurate resource estimation, efficient production, and improve cost effectiveness and economic viability of subsurface resources has necessitated the modeling of rock properties and fluid characteristics in a 3D space, using seismic, core and well logs data (Khadragy et al., 2017; Ali et al., 2021; Ayodele et al., 2021; Othman et al., 2021; Opuwari et al., 2022). Therefore, an integral aspect of site appraisal before CO2 injection is a methodic retention assessment, i.e., construction of a robust 3D static model of these geologic formations found at great depths (Smith et al., 2012; Ojo & Tse, 2016; Shariatipour et al., 2016; Ampomah et al., 2017; Niri, 2018; Abdullah et al., 2021), because it enables modeling of multiplex reservoirs having lateral and vertical lithologic variations, with an improved knowledge of reservoir properties distribution leading to enhanced volumetric estimation, risk and uncertainty analysis, predictions of fluid flow and field development plans (Abdel-Fattah et al., 2018; Adelu et al., 2019; Rahimi & Riahi, 2020; Radwan et al. 2022a, 2022b; Sarhan et al., 2022).

With CO2 emissions of approximately 430 Mt (million tons) annually (Boden et al., 2011), the South African government has acceded to the global demand for reduction in greenhouse gas emissions via some international accords (Winkler et al., 2002; Hietkamp et al., 2004). The country has further investigated the potentials of CO2 storage in South Africa, identified possible sites (saline aquifers, hydrocarbon-depleted reservoirs, coal seams and basement rocks) and estimated their storage capacities, the study revealed geological formations in South Africa can store an estimated 150 Gt (giga tons) of CO2, but onshore sites in the Zululand and Algoa basins can only hold below 2% of this (Cloete, 2010; Chabangu et al., 2014a, 2014b; Tibane et al., 2021). This has necessitated the drive to assess the potentials of offshore lying basins (Outeniqua and Orange basins) for CO2 storage, with the hydrocarbon-depleted reservoirs of greater interest due to availability of data such as wireline logs, seismic, engineering and production data for adequate reservoir assessment and de-risking.

The E-BD field is a depleted oilfield within the offshore lying Bredasdorp Basin, a sub-basin of the Outeniqua basin, southern South Africa. Wildcat, appraisal, development, and production wells have been drilled for oil extraction in the field, leading to its final abandonment. From available data, there exists no public/published work on a 3D static model of an oilfield in the Bredasdorp Basin. This study presents a detailed production of a 3D geological model, integrating 3D seismic data, well logs and available geological information for a good grasp of the geometric dispersal of continuous petrophysical properties such as effective porosity, water saturation, permeability, and discrete properties such as facies distribution in the oilfield. This model also serves as the primary input for dynamic simulation of the oilfield as a potential CO2 sink.

Geological Setting

The Bredasdorp basin is a prolific hydrocarbon province off the South African coast, and lying beneath the Indian Ocean (Parker, 2014; Acho, 2015; Magoba & Opuwari, 2017; Opuwari et al., 2022). It is 200 km long and 80 km wide, covering approximately 18,000 km2, hosting most of South Africa’s prospects and discoveries. The study area (Fig. 1), located within the Bredasdorp Basin, a sub-basin and part of a series of en echelon sub-basins within the Outeniqua Basin, is a synrift half-graben passive margin basin bounded by the Infanta arch and Agulhas arch in the north and south, respectively, the arches being basement highs composed of the Cape Supergroup sediments, metamorphic rocks, and granites dated to the Precambrian (Davies, 1997; Nfor, 2011; Opuwari et al., 2022), overlain with varying thicknesses of drift sediments.

Figure 1
figure 1

The study area map, with wells and seismic cube offshore South Africa, produced from Petrel 2018.2 software (https://usoftly.com/product/schlumberger-petrel-2018-2-7/) modified (Petroleum Agency of South Africa, 2017)

Full size image

The sedimentary successions and tectonic history of the Bredasdorp basin have been well documented, with the distribution of sediments largely eustasy-dependent (Viljoen et al., 2010; Opuwari et al., 2022) (Fig. 2). The Bredasdorp basin was infilled by sediments derived from the denudation of the shallow, deep and transitional marine environments of the Cape and Karoo Supergroups (Mcmillan et al., 1997). The Cape fold belt, extending both offshore and onshore on South Africa's coast, resulted from the Cape Supergroup folding during the Cape orogeny (Haelbich et al., 1983). Towering on the top of a retro arc foreland basin is the succession of the Karoo Supergroup of alluvial, marine, deltaic and glacial origins deposited from late Carboniferous to early Jurassic at the onset of erosion and subduction (Dingle, 1983; Smith, 1990; Brown et al., 1995; Jungslager, 1999; Broad et al., 2012). At the cessation of the Karoo Supergroup erosion, the Eastern Gondwana showed records of rifting, and the synrift half grabens of the Bredasdorp basin appeared (Mcmillan et al., 1997; Hendricks, 2019). Although the main source rock of the oil in the basin is the deep marine mature shales deposited in the mid-Aptian (Fig. 2), the synrift shelf and drift section deep marine turbidite sandstones form the two major basinal reservoirs, while the drift shales of marine origin act as the primary seals. Stratigraphic traps in the drift section and structural traps as tilted fault blocks, in the synrift, are well represented in the Bredasdorp basin (Jungslager, 1999; Petroleum Agency of South Africa, 2017).

Figure 2
figure 2

Generalized chronostratigraphic map of the Bredasdorp Basin (modified after Brown et al., 1995; Petroleum Agency of South Africa, 2017)

Full size image

Methodology

Materials

Data used for the study were made available by the Petroleum Agency South Africa (PTY) Limited (PASA). The work integrated 3D seismic volume (SEG-Y format), formation tops, check shot data, wireline logs, geological reports from the oilfield and core data. Six wells were made available, namely E-BD1, E-BD2, E-BD3, E-BD4, E-BD5 and E-CE1. Wells E-BD5 and E-CE1 were outside the seismic cube (Fig. 1).

Identification of Charis Reservoir and Hydrocarbon Bearing Zone

Gamma-ray (GR) and deep resistivity logs were used to identify potential hydrocarbon-bearing sands/reservoirs. Using gamma-ray, the identification of facies and lithology correlation from well log between the studied wells. For example, ≤ 40 API represented sandstone, 40–95 API was siltstone, with shale indicated by an API ≥ 95. Hence, the three facies identified were shale, siltstone and sandstone (Fig. 3). This process gave the facies distribution in the model generated for the reservoir and enabled facies correlation between wells. Resistivity log aided the identification of possible oil–water contact (OWC), i.e., a differentiation between oil and water-bearing zones (Saadu & Nwankwo, 2018; Adelu et al., 2019; Ayodele et al., 2021).

Figure 3
figure 3

Stratigraphic correlation of log responses through the hydrocarbon column in the drilled wells in the oilfield

Full size image

The reservoir (named Charis) (Fig. 3) was intersected in well E-BD1 from 2622 to 2653 m (31 m), 2605–2682 m (77 m) in well E-BD3 and 2607–2684 m (77 m) in well E-BD4, with the OWC at 2639 m in E-BD1 and 2640 m in E-BD4. (There was no deep resistivity log for E-BD3.) The E-BD2 well was drilled to test commercial quantities of oil stored in the sandstones intersected in the E-BD1 well, but lying above the OWC in E-BD1 and E-BD4 was a 25 m (2577–2602 m) water-saturated sandstone, suggesting the reservoir rocks encountered in E-BD1 did not continue into E-BD2 (Fig. 3). The E-BD2 well was therefore excluded from the correlation and further analysis because the Charis reservoir did not extend into this well. The Charis reservoir intersected in E-BD1, E-BD3 and E-BD4 was then subjected to further seismic and petrophysical investigations for modeling.

Seismic Interpretation

This started with tying seismic (in the time domain) to well data (measured in depth) (called the seismic-to-well tie) using the Petrel 2018.2 software developed by Schlumberger (Fig. 4). The tie was done by using the provided check shot survey from well EBD-1, and reflection coefficients (RC) and acoustic impedance (AI) were generated using density (RHOB1) and calibrated sonic log (Fig. 4). The well-to-seismic tie is principally jeered toward correlating the top and base of the hydrocarbon-bearing sand with their specific reflections on seismic. Seismic horizons (top and base of sands) were already identified on well logs, and structural discontinuities (fault) mapping over the entire grid of the seismic cube was done immediately after the well-to-seismic tie (Fig. 5). The horizons and faults were mapped in time guardedly (though there were only minor faults with no major impact on the reservoir within the entire grid). Subsequently, time was converted to depth with the 3D velocity model developed by integrating the well velocity data and the seismic horizons. The depth conversion supplied the horizons and faults in depth grids, serving as input for the 3D static model, and enabled the generation of structure maps and an isopach (thickness) map (Figs. 5, 6, 7) (Khadragy et al., 2017; Ali et al., 2021; Ayodele et al., 2021; Okoli et al., 2021; Othman et al., 2021; Sarhan et al., 2022).

Figure 4
figure 4

Well-to-seismic tie from E-BD2

Full size image
Figure 5
figure 5

Seismic section showing a mapped fault and some horizons

Full size image
Figure 6
figure 6

Time structure maps of Charis reservoir: (a) top; (b) base

Full size image
Figure 7
figure 7

Depth structure maps of Charis reservoir: (a) top; (b) base

Full size image

Petrophysical Evaluation

Wireline logs (such as resistivity, neutron, density, gamma-ray, and sonic) were subjected to petrophysical examination using Interactive Petrophysics (IP) software (version 2021) to calculate properties of the reservoir such as volume of shale, net-to-gross (NTG), effective porosity (φeff), permeability, water saturation (Sw), and reservoir thickness. The following linear formula was used to obtain the gamma-ray index (IGR) from the log (Asquith & Gibson, 1982), thus:

$${\text{IGR}} = \frac{Gr\; \log - Gr \; minimum}{{Gr \; maximum - Gr \; minimum}}$$
(1)
$$Vsh = 0.33 (2^{{\left( {2*{\text{IGR}}} \right)}} - 1$$
(2)

where Gr log is target formation gamma-ray log reading, Gr minimum is minimum gamma-ray log reading; and Gr maximum is maximum gamma-ray log reading. Corrections were then performed to Eq. 1 using the nonlinear Larionov method (Larionov, 1969). The IGR value gotten from Eq. 1 was then used in the volume of shale estimation by Eq. 2.

Density log-derived total porosity [\(\phi t\) (%)] was estimated as:

$$\phi t = \frac{\rho ma - \rho b}{{\rho ma - \rho fl}}$$
(3)

where \(\rho ma\) is density (g/cm3) of matrix; \(\rho b\) is density (g/cm3) log reading, and \(\rho fl\) is density (g/cm3) of fluid. An average matrix density of 2.67 g/cm3 from core grain density was used (Asquith & Gibson, 1982; Ayodele et al., 2021). By applying a shale correction to the calculated \(\phi t\), effective porosity was obtained as:

$${\text{Effective porosity}} = \phi t - Vsh$$
(4)

An empirical equation from the core porosity versus core permeability cross-plot was derived to estimate permeability using the hydraulic flow unit concept. The input parameters were core porosity and core permeability data. This method has been used extensively and successfully by many researchers (e.g., Amaefule et al., 1993; Abbaszadeh et al., 1996; Perez et al., 2003; Kadkhodaie-Ilkhchi et al., 2013; Nabawy & Al-Azazi, 2015; Opuwari et al., 2021, 2022; Radwan et al., 2021).

Water saturation was estimated as (Archie, 1942):

$$Sw = \left( {\frac{aRw}{{\phi^{m} Rt}}} \right)^{1/n}$$
(5)

where a is coefficient of formation factor, Rw is water resistivity (ohm), m is cementation exponent, Rt is true resistivity (ohm) of the formation, n is saturation exponent, and \(\phi\) is porosity (dec). Estimated results from the log were calibrated with core measurements as supplied by the Petroleum Agency South Africa (PTY) Limited (PASA), which gave a good level of reliability to the results (Figs. 9, 10).

Result and discussion

Time and Depth Map of the Charis Reservoir

The time contour map of the Charis reservoir shows a time increase in the eastern and southeastern areas, with maximum values of about 1.97 and 1.95 s, respectively, indicative of low structural features (Ali et al., 2021). There was a decrease in the TWTs (two-way travel times) in the northern and western parts with values around 1.9 and 1.88 s, with the southwestern area recording the most negligible value of about 1.83 s, indicative of high structural features (Fig. 6a, b). The basal map of the Charis reservoir records similar characteristics, with the time increase in the eastern and southeastern areas, with values around 0.2 s, a decrease of TWTs in the northern and western parts with values around 1.9 and 1.89 s, and the most negligible value of about 1.84 s in the southwestern portion of the map. The depth maps (Fig. 7a, b) of the Charis reservoir show a general southeast, central to northwest deepening with peak values reaching − 2630 m and shallowing to the southwest and northern areas, having values of − 2450 m and − 2480 m, respectively.

Isopach Map of the Charis Reservoir

The Charis reservoir’s isopach map was generated by subtracting the Charis reservoir base depth map from the top depth map (refer to Fig. 3 for stratigraphic positions). As a result, the generated isopach map (Fig. 8) reflected that the Charis reservoir thickness increased markedly in the northwest and central sections of the reservoir, with the basinal area having sediment thickness of around 125 m in the northwest and down to between 95 and 100 m in the central portion of the reservoir. However, thickness decreased in the northeastern region to as low as 5 m.

Figure 8
figure 8

Isopach map of the Charis reservoir

Full size image

Lithology and Petrophysical Characteristics of the Charis reservoir

The stratigraphy of the Charis reservoir, as deduced from core data and well logging analyses, is composed of three facies; shales, predominantly clean good, quality channel sandstones (coarse to fine-grained); and thin siltstone interbeds. Petrophysical data logs of some selected wells and a summary of the petrophysical properties of the Charis reservoir are presented in Figures 9 and 10 and Table 1.

Figure 9
figure 9

Charis reservoir litho-saturation cross-plots from well E-BD1

Full size image
Figure 10
figure 10

Charis reservoir litho-saturation cross-plots from well E-BD4

Full size image
Table 1 Petrophysical properties of Charis reservoir from well log analysis
Full size table

3D Facies Model

Facies log upscaling in the available wells is the first stage of building a 3D facies model of a reservoir (Ali et al., 2020a, 2020b, 2021; Othman et al., 2021; Radwan et al., 2022a, 2022b), with the assigned facies values in the log from the penetrated wells vertically and horizontally distributed to fill the whole 3D grid. With sequential indicator simulation (SIS), the statistical method in the Schlumberger Petrel software adopted for facies modeling, the proportions of facies used were 4.45% shale, 76.22% sandstone and 19.53% siltstone (Fig. 11a). In addition, to explicitly illustrate the vertical and horizontal facies changes across the reservoir, cross sections of the Charis reservoir were generated (Fig. 11b, c). The interpreted facies from well logs and core reports of the Charis reservoir are interpreted as a channel-fill deposition (massive sandstone beds separated by minor shale and siltstone interbeds), with the sandstones being deposited within the confines of a deep-marine channel, firstly as amalgamated inputs, the product of pulses of high-density turbidity currents, followed by deposition in predominantly low sinuosity channels within the overall channel boundaries (Reading & Richards, 1994; Clark & Pickering, 1996).

Figure 11
figure 11

Facies distribution in the Charis reservoir: (a) 3D model; (b) N–S cross section; (c) E–W cross section

Full size image

3D Petrophysical Model

Petrophysical values derived from the Interactive Petrophysics™ software (IP 2021) were upscaled and modeled with the aid of the petrophysical modeling procedure in the Schlumberger Petrel software. The sequential Gaussian simulation algorithm was the statistical method employed for the distribution of the petrophysical parameters (volume of shale, effective porosity, permeability, and water saturation) in the model, cross sections in the NS and EW directions were also extracted to recognize both vertical and lateral property distributions in the model (Figs. 12, 13, 14, 15).

Figure 12
figure 12

Water saturation distribution in the Charis reservoir: (a) 3D model; (b) N–S cross section; (c) E–W cross section

Full size image
Figure 13
figure 13

Effective porosity distribution in the Charis reservoir: (a) 3D model; (b) N–S cross section; (c) E–W cross section

Full size image
Figure 14
figure 14

Permeability distribution in the Charis reservoir: (a) 3D model; (b) N–S cross section; (c) E–W cross section

Full size image
Figure 15
figure 15

Shale volume distribution in the Charis reservoir: (a) 3D model; (b) N–S cross section; (c) E–W cross section

Full size image

The effective porosity model showed values ranging from < 5 to 22%, with permeability values between 100 and 560 mD dominant in the reservoir. The Charis reservoir thus exhibited medium to high porosity and permeability values, with both properties increasing mainly in the northern, southeastern, and central portions of the reservoir. The petrophysical properties of the reservoir tend to be facies-controlled, with the southwestern parts having reduced porosity and permeability values, aligning well with the facies model having a concentration of the siltstone and shale facies in the same area. The Vsh model showed low shale fractions in the reservoir, with the reservoir being dominantly composed of clean sands, with water saturation ranging between 10 and 90%. Generally, from the petrophysical and facies characteristics of the Charis reservoir, it is classed as a good reservoir rock (Table 2) (Levorsen and Berry, 1967).

Table 2 Classification of reservoir (a) porosity and (b) permeability values (adapted from Levorsen & Berry, 1967)
Full size table

As a potential CO2 storage site based on proposed parameters (Bachu, 2003; Smith et al., 2012), which serve as the basis for suitable site scoring and ranking, the results showed that the Charis reservoir has good petrophysical characteristics, increasing its potential for CO2 storage.

Static Storage Capacity Assessment

In the appraisal of a potential CO2 storage site, estimation of the storage capacity of a reservoir is key, with the two broad methods of estimation (Bachu, 2008; Frailey, 2009; Jin et al., 2012). One method is static capacity estimation (which is simple, straightforward, requires less input data and computational time, and provides a good and useful initial assessment of the reservoir). The other method is numerical simulation (it is dynamic, requires more input data and computational time, but provides more reliable results). The rock and fluid properties, which are the input parameters, are time-dependent in the dynamic estimation but time-independent in the static case (Jin et al., 2012). The most widely used static methods for storage estimation are the compressibility method (van der Meer & Egberts, 2008; Zhou et al., 2008) and volumetric method (Holloway et al., 1996; DOE, 2007; Chadwick et al., 2008). The compressibility method (Eq. 6) which is applied usually for CO2 storage estimation in confined aquifers and single-phase oil-depleted reservoirs (Goodman et al., 2011; Vulin et al., 2012) was adopted for this study, thus:

$$V_{{{\text{CO2}}}} = V_{p} \times \left( {C_{p} + \left( {S_{w} \times C_{w} } \right) + \left( {S_{o} \times C_{o} } \right)} \right) \times \Delta P_{\max }$$
(6)

where VCO2 is calculated volume of CO2 that can be stored in the reservoir, C is compressibility, S is saturation, and subscripts p, w and o refer to pore space, water, and oil, respectively; ΔPmax is maximum allowable pressure increase in the reservoir, which is the difference between the maximum pressure allowed in a system (taken as 90% lithostatic pressure gradient) and the initial pressure of a reservoir. This method assumes that pore space to be filled by the injected CO2 is dependent on the compressibility of the existing reservoir fluids and the rock (Obdam, 2000; van der Meer & Egberts, 2008; Jin et al., 2012). The fluids and rock compressibility, reservoir pressure, and saturation values were sourced from the well engineering report of the field, with the pore volume gotten from the static model built using the Petrel software. The CO2 static reservoir capacity computation of Charis reservoir with the presence of an oil phase (virgin oilfield) is presented in Table 3, with input values collected in field units but reported in metric units.

Table 3 Calculated CO2 storage capacity for Charis reservoir with oil phase
Full size table

The initial state of the reservoir before oil migration into it was also considered. Equation 6 was then rewritten to eliminate the oil component, thus:

$$V_{{{\text{CO2}}}} = V_{p} \times \left( {C_{p} + \left( {S_{w} \times C_{w} } \right)} \right) \times \Delta Pmax$$
(7)

where VCO2 is the calculated volume of CO2 that can be stored in the reservoir, C is compressibility, S is saturation, subscripts p and w refer to pore space and water; ΔPmax is maximum allowable pressure increase in the reservoir, which is the difference between the maximum pressure allowed in the system (taken as 90% lithostatic pressure gradient) and the initial pressure of the reservoir. The reservoir pressure, lithostatic pressure, fluids and rock compressibility values were sourced from the well engineering report of the field, while the pore volume was gotten from the static model built using the Petrel software. The CO2 static reservoir capacity computation of Charis reservoir fully saturated with water is provided in Table 4, with input values collected in field units and reported in metric units:

Table 4 Calculated CO2 storage capacity for Charis reservoir fully water saturated
Full size table

Sensitivity Studies

The sensitivity studies performed on the static capacity estimation of the reservoir included the following:

  1. a.

    Depleting oilfield at constant initial reservoir pressure (Pi) and increasing water saturation. Using Eq. 6, the results are presented in Table 5 (with Vp, Cp, Cw, Co and ΔPmax being fixed parameters, while Sw increases with decreasing So). From Table 5, equal oil and water saturation of 0.5 yielded an estimate of 1.28 Mt CO2 that can be stored in the reservoir, and while 90% of the reservoir was filled with brine, it was estimated to store 0.82 Mt CO2. As the oilfield was being depleted and water saturation increased, the volume of CO2 that can be stored in the field reduced (Fig. 16), this was because water is less compressible than oil.

  2. b.

    Depleting oilfield and depressurization—initial pressure was dropped to bubble point pressure (Pb) while increasing water saturation in the system (ΔPmax is the difference between the maximum pressure allowed in the system (taken as 90% lithostatic pressure gradient) and the bubble point pressure of the reservoir; Vp, Cp, Cw, Co, Pb and ΔPmax are fixed parameters, while Sw increased with decreasing So). Using Eq. 6, the results are presented in Table 6 and Figure 17. In a depleting oilfield at bubble point pressure, the total volume of CO2 that can be stored also reduces with increasing water saturation, though the volumes at each water saturation point are more than twice the corresponding values at initial reservoir pressure. Therefore, reduction in reservoir pressure will give increasing volume of CO2 that can be stored in a reservoir.

  3. c.

    Non-depleting oilfield with continued depressurization—constant water saturation and reservoir pressure at 1000 psi.Footnote 1 At bubble point pressure, the first bubble of natural gas begins to come out of solution, and as the reservoir pressure continues to decrease, more gas comes out of solution to form a gas cap. The gas component was therefore introduced into Eq. 6 to give:

    $$V_{{{\text{CO2}}}} = V_{p} \times \left( {C_{p} + \left( {S_{w} \times C_{w} } \right) + \left( {S_{o} \times C_{o} } \right) + \left( {S_{g} \times C_{g} } \right)} \right) \times \Delta Pmax$$
    (8)

    where VCO2 is the calculated volume of CO2 that can be stored in the reservoir, C is compressibility, S is saturation, and subscripts p, w, o and g refer to pore space, water, oil and gas, respectively; ΔPmax is the maximum allowable pressure increase in the reservoir, which is the difference between the maximum pressure allowed in the system (taken as 90% lithostatic pressure gradient) and 1000 psi (6.895 MPa), with Vp, Cp, Cw, Co, Cg, Pb, Sw, new reservoir pressure (1000 psi) and ΔPmax are fixed parameters, while Sg increases with decreasing So. The results are presented in Table 7 and Figure 18. The depressurization case in Table 7 and Figure 18 shows that as gas cap formed below bubble point pressure, the volume of CO2 that can be stored increased with increasing gas saturation. This is because gas is more compressible than oil and water.

Table 5 Depleting oilfield (increasing Sw and reducing So) at constant initial reservoir pressure (Pi)
Full size table
Figure 16
figure 16

CO2 static capacity estimate for depleting oilfield at initial reservoir pressure (Pi)

Full size image
Table 6 Depleting oilfield (increasing Sw and reducing So) and depressurization (reservoir depressurized to bubble point pressure Pb)
Full size table
Figure 17
figure 17

CO2 static capacity estimate for depleting oilfield at bubble point pressure (Pb)

Full size image
Table 7 Non-depleting oilfield (constant Sw) at 1000 psi (increasing gas saturation (Sg) and decreasing oil saturation (So)
Full size table
Figure 18
figure 18

CO2 static capacity estimate for non-depleting oilfield at 1000 psi (1 psi = 6.8947572932 kPa)

Full size image

Conclusions

Carbon capture, utilization and storage remain an important and essential technology for countries and industries to reduce worldwide environmental change, with underground storage of carbon dioxide in geological formations a core of carbon sequestration, especially in non-renewable energy-centered nations. Developing a robust geological model of underground formations is paramount for CO2 storage site appraisal, to deduce their facies and petrophysical properties.

The investigated Charis reservoir as a potential CO2 storage site is a sandstone reservoir with mean effective porosity of 13.92% and dominant permeability values of 100–560 mD, with both properties increasing in the reservoir's northern, southeastern and central portions. The southwestern area with siltstone and shale facies has reduced porosity and permeability, making the petrophysical properties of the reservoir facies controlled. The volume of shale model shows that the reservoir is composed of clean sands and water saturation ranging between 10 and 90%. The reservoir is suitable for CO2 storage based on its porosity and permeability.

As a virgin oilfield, it is estimated the reservoir can sequester around 1.62 Mt (million tons) of CO2, and 0.71 Mt of CO2 with the reservoir fully water-saturated, based on only the static storage volume estimate of the reservoir. Increasing reservoir water saturation (a depleting oilfield at constant initial reservoir pressure) decreases CO2 storage capacity of the reservoir because water is less compressible than oil. Oil depletion (increasing water saturation) at bubble point pressure is also attended with decreasing CO2 storage capacity, although the total CO2 volume that can be stored is more than twice the volume at each water saturation level while the reservoir was at initial reservoir pressure. The oilfield, not under depletion and at 1000 psi, with a gas cap and water saturation of 50%, significantly increases CO2 storage volume because gas is more compressible than oil and water.

Further credence can be given to the reservoir and seal in the field because it has held and kept hydrocarbon in place in the geologic past, and the field has been produced. The presence of residual oil will also increase the volume of CO2 that can be stored in the reservoir. This study presents a detailed investigation into an oilfield within the Bredasdorp Basin as a potential CO2 storage site and can be used to compare the well-explored gas fields in the basin, it is also the primary input for dynamic simulation of the field.