Defining CO₂ Geological Storage Capacity in Unmineable Coal Seams Through Adsorption Data in 3D: Case Study of the Chico Lomã Deposit, Southern Brazil

Abstract

The concentration of greenhouse gases in the atmosphere has led to irreversible climate changes, emphasizing the need for effective strategies to mitigate emissions. Carbon capture, utilization, and storage (CCUS) technologies, including geological CO₂ storage, have gained recognition worldwide due to their potential for CO₂ emissions abatement. Among potential geological reservoirs, coal seams are significant due to their efficiency in securing CO₂ storage, through their adsorption storage capacity. This study presents an innovative methodology for estimating the theoretical CO₂ storage capacity in unmineable coal seams, focusing on the Chico Lomã deposit in southern Brazil. The methodology integrates a comprehensive drillhole database and adsorption isotherm data to define the coal reservoir zone and calculate its CO₂ storage capacity. The results indicate a total theoretical CO₂ storage capacity of 47.8 Gt in the Chico Lomã deposit, with the potential to mitigate emissions from local thermoelectric plants for over 500 years. The study encourages the application of the proposed methodology to assess CO₂ storage capacity in other unmineable coal deposits worldwide.

1. Introduction

The concentration of greenhouse gases in the atmosphere has undergone an increase, leading to irreversible climate changes [1]. Carbon capture, utilization, and storage (CCUS) technologies, including CO₂ geological storage, have been widely acknowledged as effective strategies for greenhouse gas emission mitigation [2,3,4]. About 80% of long-term low-emission development strategies submitted to the United Nations Framework Convention on Climate Change (UNFCCC) recognize the importance of CCUS technologies. According to the IEA Roadmap to Net Zero by 2050, CCUS is projected to reach an abatement capacity of 7.6 Gt (billion metric tons) of CO₂ per year by 2050 [4]. Global coal burning in 2022 resulted in the emission of 15.22 Gt (billion metric tons) of CO₂, representing a 1.6% rise from the previous year [5]. Among the different options for CO₂ geological reservoirs, coal seams should play a major role in CCUS technologies, due to their high CO₂ storage capacity, as well as high efficiency in ensuring permanent and secure CO₂ geological storage [6]. Both the CO₂ storage capacity and efficiency of traps within coal seams are closely related to the inherent characteristics of coal. The CO₂ is mainly stored in the internal pore surfaces of the organic particles in a condensed form that is very close to a liquid state. This enables the storage of a CO₂ volume that significantly exceeds the pore volume capacity [6]. The efficiency of the trap is enhanced by the high organic matter content of coal (more than 50% by weight) (ISO 11760 2005 [7]), as CO₂ has a high adsorption affinity for organic matter and can therefore be safely stored in a coal seam.

There are currently three coal power plants fitted with CCUS in operation in the world as follows: the Boundary Dam in Canada, and the Jinjie Power and Taizhou Power stations in China. The Taizhou project recently began operating in June 2023 and has a significant capacity to capture approximately 550,000 metric tons of CO₂ per year [8]. Other industrial sectors with CO₂ emissions eligible for carbon capture deployment are the iron and steel industry, cement industry, chemicals, refineries, and pulp and paper [9]. Coal is currently acknowledged as one of the best CO₂ removal solutions, in the CCUS technologies portfolio, for achieving CO₂ neutrality by 2050. However, other studies using coal, such as the traditional oil production system by direct coal liquefaction technology, are currently being replaced by a carbon-neutral technology-based direct coal liquefaction system coupled with green hydrogen, green electricity, or CCUS technologies [10].

CCUS presents significant potential for reducing CO₂ emissions in Brazil, especially due to the increased use of fossil fuels [11,12]. The implementation of CCUS in Brazil is essential, but other solutions are also needed, such as changes in land use and the agricultural sector [13]. However, public acceptance is hindered by the limited information about the technology, and Brazil lacks a robust regulatory framework, putting it at a disadvantage compared to countries like the U.S. and Canada [13]. Brazil may probably only achieve capturing up to 40 million tons of CO₂ annually by 2050, with a high investment requirement of around USD 58 billion for capturing 190 million tons per year [13]. The successful adoption of CCUS depends on significant investments in research, creating a clear regulatory framework, and fostering international collaboration and public engagement to ensure the technology’s implementation [11,12,13,14].

Studies considering geological storage of CO₂ in the Paraná Basin, Brazil, were conducted focusing on the black shales of the Irati Formation [15,16,17,18,19,20], as well as sandstone reservoirs [21,22,23], hydrocarbon reservoirs [24] and saline aquifers of the Rio Bonito Formation [25,26,27]. More specifically, the coal seams of the Rio Bonito Formation were also the subject of studies for underground CO₂ storage [28,29,30,31] and on coal mining waste [32]. Few studies combining 3D geological modeling and CO₂ storage capacity estimates were conducted in Brazil [16,17,33], although this approach has become more frequent in recent decades [34,35,36,37,38,39]. These few studies have the advantage of using 3D implicit modeling, which better represents geological surfaces [40,41,42]; however, they were applied to other types of CO₂ reservoirs and have not been applied to coal seams. The current work is the first application of 3D implicit modeling in CO₂ geological storage in coal deposits in Brazil.

The main methodologies for calculating CO₂ storage capacity in coal [43,44,45,46,47,48,49] initially determine the volume and then apply an adsorption parameter. This work proposes that the adsorption data be used first to determine the volume, and consequently, the CO₂ storage capacity. This approach has advantages, especially when applied at the Site Selection and Initial Characterization stages of the Exploration Phase [46].

Thus, our research introduces an innovative methodology for calculating the theoretical CO₂ storage capacity in unmineable coal seams using the most modern 3D implicit modeling techniques together with analytical adsorption isotherm data. The new methodology introduced here already employs isotherm data to initially define the volume in 3D, thereby ensuring the minimum and maximum CO₂ adsorption values for calculating storage capacity.

2. Geological Setting

2.1. Rio Bonito Formation

The Paraná Basin covers an extensive area exceeding 1,000,000 km² in southern Brazil, comprising six depositional supersequences from Late Ordovician to Cretaceous as follows: Rio Ivaí, Paraná, Gondwana I, Gondwana II, Gondwana III, and Bauru [50,51]. The Chico Lomã coal seams are part of the Rio Bonito Formation, which compose the Sequence Gondwana I. The Rio Bonito Formation was deposited over sedimentary layers of the Itararé Group and is part of the Guatá Group and the Tubarão Supergroup. It extends across the following three Brazilian states: Rio Grande do Sul, Santa Catarina, and Paraná (Figure 1). The Rio Bonito Formation is subdivided into three members, which are, from bottom to top, as follows: (a) the Triunfo Member, consisting mainly of sandstone and conglomerate, with subordinate shale and siltstone, mostly organic-rich, and coal layers; (b) the Paraguaçu Member, composed of shale, siltstone, and subordinate fine-grained sandstone; and (c) the Siderópolis Member, comprising sandstone, black shale, and coal layers [52]. In the state of Rio Grande do Sul, varied-textured sandstones, siltstones, and conglomerates have been essentially described, with carbonate breccia, carbonaceous shale, and coal (Figure 2), corresponding to the Rio Bonito Formation. In this region, there are about seven main coal deposits of greater importance (Figure 1), including the Chico Lomã deposit [53,54].

According to de Oliveira et al. [16], the Paraná Basin is a typical intra-plate region presenting a low level of seismic activity. Only five earthquakes with a magnitude above 5.0 (with two being of large magnitude: 6.3 and 6.8) have occurred in the studied region within the last 220 years, according to the bulletins of Centro de Sismologia da Universidade de São Paulo (USP), Brazil (http://moho.iag.usp.br/eq/bulletin, accessed on 20 June 2024) [56]. This scenario also indicates a favorable location for a CO₂ reservoir according to IEA-GHG recommendations [57].

2.2. Chico Lomã Coal Deposit

The Chico Lomã is an unmined coal deposit located in the southern region of Brazil (Figure 1), whose resources were discovered and defined in the early 1970s and mid-1980s [54]. The deposit is generally formed by six coal layers, named from CL1 to CL6, from the stratigraphic top to bottom, hosted in the Rio Bonito Formation [58]. The main coal seam in terms of thickness and lateral continuity is CL4 (Figure 3). The coal seams at Chico Lomã dip eastward towards the contiguous Santa Terezinha deposit. In the southwestern portion of the deposit, the coal seams are found at depths of 50 m or less; in the central portion, the depths vary between 50 m and 300 m, in the north-central, southeastern, and northeastern parts, with depths equal to or greater than 300 m [54].

The Santa Terezinha deposit covers an area of approximately 1000 km² and the coal layers of interest occur only at depth, with no outcropping. Coal layers in Santa Terezinha occur under a minimum sediment cover of 450 m, sometimes reaching depths of up to 1000 m. The main coal seam in the deposit is ST4. Despite the lack of continuity between the two deposits (Chico Lomã and Santa Terezinha), the layers of the Santa Terezinha deposit can be directly correlated to the Chico Lomã layers, e.g., CL4 = ST4 [54]. Both deposits show similar values for total carbon, ash, sulfur, and calorific values (

Table 1. Comparative parameters between the adjacent Chico Lomã and Santa Terezinha coal deposits [59].

Deposit	Calorific Value (kcal/kg)	Carbon (%)	Ash (%)	Sulfur (%)
Santa Terezinha	3800–4300	28.0–30.0	41.0–52.5	0.5–1.9
Chico Lomã	3700–4500	27.5–30.5	40.0–56.2	0.6–2.0

) [59]. A study considering coalbed methane (CBM) in the Chico Lomã deposit has indicated an average value of 0.374 cm³ gas/g of coal [55].

2.3. CO₂ Adsorption in Coal Seams

Potential reservoirs for CO₂ geological storage vary between saline aquifers, salt caverns, or depleted hydrocarbon reservoirs to unmineable coalbeds and shale gas reservoirs [60,61]. The last two are classified as unconventional reservoirs due to their low permeability. In addition, both reservoirs are organic-rich and have different storage and gas flow properties compared to conventional reservoirs. Therefore, CO₂ geological storage in unconventional organic-rich reservoirs, such as coal seams (≥50% organic content in weight), is driven by its organic microporous structure and inherent CO₂ adsorption properties due to its high internal surface area and, consequently, high sorption capacities [6,62].

Furthermore, due to the high organic matter content present in coal seams, both trap and seal geological parameters do not play a significant role in the storage process, as CO₂ entrapment is achieved through organic matter adsorption behavior [63,64], attributing the permanent and secure storage of CO₂ to coal seams in the medium- to long-term.

The primary mechanism for CO₂ storage in coal seams is adsorption, which differs from the mechanisms in other geologic reservoirs [49]. Thus, the main equations for calculating capacity, such as the CSL methodology [45], the DOE NETL methodology [46], and the in situ coalbed methane resource conversion method [47], are based on adsorption, usually obtained through the Langmuir isotherm. Some equations, as proposed by Van Bergen et al. [43], also consider the enhanced coalbed methane (ECBM) potential, which has been applied in the Santa Terezinha coal deposit [29]. Xu et al. [65] detached the neglected CO₂ storage capacity of free gas and soluble gas with an increasing trend trough depth.

Depth limits for CO₂ injection and storage in coal seams vary in each basin, depending on the geothermal and geo-pressure gradients [46]. However, some technical operational limits are indicated in the literature. For instance, in the case of considering enhanced coalbed methane recovery (ECBM) combined with CO₂ injection, the increase in effective stress induces the cleat closure, reducing permeability to the extent that coalbed methane cannot be produced below 1500 m [60]. At greater depths, supercritical CO₂ can cause swelling of the coal matrix, generating injectivity problems [66,67]. CO₂ storage is also limited by compression costs, which rise significantly below 3300 m [46,68]. In the recent compilation from Wu et al. [69] of 23 pilot experiments and theoretical studies conducted worldwide on coal deposits, the depths range from 273 m to 1260 m.

3. Methodology

3.1. Drillhole Database and Digital Terrain Model

The database used in this study consists of 37 vertical drillholes [55] in a total of approximately 11,584 m, where the average depth per hole is 300 m. Exceptions include two drillholes (CA-85 and CA-86) located further north with depths of ~700 m (Figure 2). All drillholes possess data regarding the top (first) and base (last) of the coal seams, depth (m), cumulative and net coal thicknesses, and information on the underlying geological formation and the basin’s crystalline basement.

A digital terrain model (DTM) was generated from the SRTM (Shuttle Radar Topography Mission), based on the public data available on the EarthExplorer website (https://earthexplorer.usgs.gov, accessed on 27 March 2023) in the QGIS software, version 3.34.11-Prizren. The DTM surface and the drillhole data spreadsheets were loaded in the Leapfrog Geo software, version 2024.1.2 where all subsequent 3D modeling steps, described below, were performed (Figure 4). The 3D solids and volumes were generated in the software (Figure 5) using algorithms based on fast radial basis functions (RBF) [42,70,71]. In general, the software has 4 main interpolation algorithms known as follows: “Intrusion”, “Vein”, “Stratigraphic” and “Deposit” [33,72,73], the latter being the one that best depicted the Chico Lomã coal seams geologically in 3D. A complete explanation of the mathematics behind this method is outside the scope of this paper. However, it is important to note that the methodology relies on the principle that infinite precision is neither necessary nor anticipated when performing fast RBF computations [42,70,71,74].

The 3D model of the coal-bearing zone was generated using the top and base data and the “Deposit” algorithm of the software through implicit modeling [42]. The coal-bearing zone 3D model dips gently one degree northeast N25, as defined by the drillhole intersections (Figure 6).

3.2. Defining the Depth of the Coal-Bearing Zone for CO₂ Storage

To define the favorable CO₂ storage portions that are spatially within the coal-bearing zone model of the Chico Lomã deposit, the CO₂ sorption isotherm data from the sample, identified as “07-166” from Weniger et al. [28], was used. The 07-166 sample consists of a coal sample from the Rio Bonito Formation, collected from a drillhole at a depth of 620.80 m, from the contiguous coal deposit of Santa Terezinha (Figure 1), due to the lack of data available in the literature for Chico Lomã deposit. The TOC value for the 07-166 sample is 43.9%, vitrinite reflectance is 0.92%R_r, moisture content is 0.93%, and ash wield is 41.99% [28]. The 07-166 sample presents a maceral composition with vitrinite content of 53%, liptinite content of 9%, and inertinite content of 38% [28]. The CO₂ sorption isotherms were measured at 45 °C on a raw basis and were determined up to pressures of more than 20 MPa [28]. The temperature level is consistent with a geothermal gradient (28.8 °C/km) determined from the bottom hole temperature of a nearby Petrobras petroleum exploration well [75] and the average annual temperature of the area [55]. The isotherm used in this study exhibits a maximum in the 8.1 MPa and then shows a continuous decline (Figure 7). This is a typical behavior of excess adsorption isotherms [76], where the curve decreases with increasing temperature, because of the exothermic nature of the sorption process [77]. This phenomenon was also observed by Busch et al. [78], Pini et al. [76], and Zhang et al. [79] for CO₂ adsorption on coal, as characterized by a maximum absorption peak. Assuming the interval between 4.2 MPa and 8.1 MPa as the best adsorption range, and a pressure gradient of 10 MPa/km [22], the maximum and minimum depth values of 419 m and 812 m are obtained.

3.3. Clipping the Coal-Bearing Zone Model into the Interest Depth

Initially, a 3D model was implicitly generated based on information from the upper portion of the coal-bearing zone, that is, based on the drillhole intervals overlying the coal-bearing zone. This model represents, in 3D, the vertical distances from the topographic surface to the top of the coal-bearing zone. The maximum and minimum depth limits were defined based on CO₂ sorption isotherm data and then filtered into this model (Figure 8). The intersection between these limits with the DTM generates two polylines (Figure 9A). The two polylines were projected vertically at depth to create a surface, and then, a solid bounded by these polylines was generated to define the volume of interest, where the coal-bearing zone is located between 419 and 812 m depth (Figure 9B). Finally, a cutting operation was carried out between the model of the total coal-bearing zone within the entire study area and the solid block that represents the interval of interest depth (Figure 9C). In the software, surfaces and solid blocks, in the form of vertices and triangles, define the 3D shape of the models and are called meshes. The software operation used is called “clip mesh”. This approach has some similarities with the methodology presented by Kaufmann and Martin [34], although the studies have different goals.

3.4. CO₂ Storage Capacity Calculation

The CO₂ storage capacity calculation in the Chico Lomã coal deposit was performed using the DOE NETL [46] equation as follows:

$$G_{C{O}_2}=Ah{C}_s{\rho }_{C{O}_2}{E}_{coal}$$

where $G_{C{O}_2}$ is the mass estimate of the CO₂ resource of one or more coal beds within the coal-bearing zone. The product of the total area and the thickness ($Ah$) account for the total bulk volume containing the coal to be assessed and was obtained directly from the 3D-generated model of the coal deposit zone. The $C_s$ is the maximum volume of CO₂ at standard conditions that can be adsorbed per volume of coal (e.g., the Langmuir isotherm volume constant), and it is assumed to be on an in situ or “as it is” basis. As mentioned before, in this research we assumed the results of the high-pressure isotherms from Weniger et al. [28], considering the minimum and maximum values of 12.50 and 14.55 m³/t, respectively (Figure 7). ${\rho }_{C{O}_2}$ is the standard density of CO₂. According to Bachu et al. [60], to express the CO₂ storage capacity in mass rather than volume of CO₂, the results have to be multiplied by CO₂ density at standard conditions of 1.873 kg/m³. The storage efficiency factor ($E_{coal}$) represents the portion of the total bulk coal volume that can store the injected CO₂ and ranges between 21 and 48 percent at the 10th to 90th percent probability interval [46]. This methodology can be applied in equations for capacity calculation based on direct CO₂ storage in coal seams like DOE NETL [46], as well as in equations considering ECBM [43,48]. As the innovation presented here basically refers to the definition of the volume in 3D before the calculation, in both types of equations, simply apply the proposed methodology and substitute the 3D volume obtained into the volumes in the ($Ah$) from any of the equations [46,48].

Table 2. Used parameters in the CO₂ storage capacity calculation of the Chico Lomã coal deposit.

Parameter	Values	Source
$Ah$	2320 million m3	This study
$C_s$	12.50 m3/t and 14.55 m3/t	[28]
${\rho }_{C{O}_2}$	1.87 kg/m3	[60]
$E_{coal}$	21% to 48%	[46]
${\rho }_{coal}$	2.05 kg/m3	[55]

shows the values used in this study with their respective sources.

4. Results

The coal-bearing zone volume within the Chico Lomã deposit is 92,044 million m³ within the study area. The coal-bearing zone volume at depths of interest for CO₂ storage, between 419 m and 812 m, is 27,130 million m³. The average thickness of the coal-bearing zone in the drillhole data is 18.7 m [55], and the cumulative coal seam thickness of the two deepest drillholes CA-85 and CA-86 are 24.00 m and 17.76 m, respectively. Therefore, the total coal volume was estimated at 2320 million m³. Previous works that have used 3D modeling have presented different volumes of coal. Araújo et al. [54] present a total coal volume of 1064 million m³ for the Chico Lomã deposit with an average thickness of 0.90 m, while Kalkreuth et al. [55] present a coal volume of 3489 million m³ with an average thickness of 2.41 m. The volumes mainly vary according to the area that has been defined for the deposit. The coal seams show continuity and extension at depth and laterally. Boundaries in this study are in accordance with the availability of drillhole data (Figure 1 and Figure 4).

Applying the coal seam efficiency factors ($E_{coal}$) of 21%, 37%, and 48% (P₁₀, P₅₀, and P₉₀, respectively) from Goodman et al. [46], the total theoretical CO₂ storage capacity of the Chico Lomã deposit would range from 27.2 Gt to 62.2 Gt (

Table 3. Total theoretical CO₂ storage capacity in the Chico Lomã coal deposit, in accordance with different coal seam efficiency factors from Goodman et al. [46].

P10 (21%)	P50 (37%)	P90 (48%)
27.2 Gt	47.8 Gt	62.2 Gt

).

The CO₂ storage efficiency factor from the DOE NETL methodology [46] is a function of geologic parameters, such as area, thickness, and displacement, and efficiency components, such as areal, vertical, gravity, and the degree of CO₂ saturation (with respect to the maximum predicted by the Langmuir isotherm) within the CO₂-accessible deposit, which reflect the portion of a region’s coal bulk volume and primarily the pore internal surfaces of organic matter with which CO₂ is expected to contact. However, this equation cannot be used to obtain information about micropore structures from a single adsorption isotherm, such as the Dubinin–Astakhov equation adsorption model [80].

The efficiency factors for coal seams were determined by using the Log Odds Method when applied with Monte Carlo sampling [81]. The overall storage efficiency factor for coal seams ranges from 21 to 48 percent over the 10 and 90 percent probability range. The P10 and P90 values computed in this manner serve as the nominal lower and upper bounds that demark a plausible range of efficiency factors, defined in a consistent probabilistic manner [46]. Because these limits are based on a combination of data with varying quality and expert judgment, the P10 and P90 limits should be interpreted as general (rather than strictly mathematical) limits [46].

5. Discussion

The CO₂ storage capacity of the Chico Lomã coal deposit has been evaluated for the first time in this study. A similar calculation, although without using the original methodology presented here, was performed in the nearby Santa Terezinha coal deposit (Figure 2), reaching a total CO₂ storage capacity of 13.8 Gt while also considering the coalbed methane (CBM) potential [29].

The total capacity obtained, despite being restricted by the volume of coal within the best adsorption limit, is still higher than that presented for Santa Terezinha [29] and other similar global cases [69]. This is probably because we are working with a projection of the dip of the coal zone at depth and only two deep drillholes (Figure 6 and Figure 8), and due to the specific coal characteristics (maceral composition and vitrinite reflectance) between the different deposits. With the development of the project and new drillholes detailing this information, increasing geological reliability, this number will most likely be lower, as pointed out by Bachu et al. [60].

In the assessment of the Chico Lomã deposit, we chose to use the DOE NETL equation [46] due to its efficiency factors and ease of application, emphasizing the adsorption, which was used in the previous definition of the volume (Figure 7). However, any of the literature equations for coal seams [43,45,46,47,65] and other similar ones can be applied to the workflow we present here since the main difference lies in the determination of the volume before calculation.

Although the Chico Lomã is an unmined coal deposit, the approach introduced here can also be extended to abandoned coal mines by applying the appropriate volume discounts for shafts, pillars, and other underground mine structures, according to Ray and Dey [82].

Following the capacity classification of Bachu et al. [60], which considers the technical and economic aspects in CO₂ storage capacity evaluations, the tonnages discussed and presented in this study are assumed as the Theoretical Capacities (Table 3). To achieve satisfactory performance in the categories of Effective or Practical Capacities [60], it is recommended that future works perform adsorption tests on individual samples of each of the main coal seams (CL1 to CL3, Figure 3) of the Chico Lomã deposit and carry out 3D modeling of each seam individually. As highlighted by [83], experiments in the laboratory need to reproduce the underground conditions as closely as possible in order to be useful for field application. Another important aspect would be to increase the number of samples for adsorption isotherms. Our initial demonstration is based on just one sample. Ideally, in order to achieve a reliable Effective Capacity classification [60], tests should be carried out on several samples from the various coal seams, which can present different heterogeneities in the coal reservoir. For the subsequent classification of Practical Capacity, it is recommended that tests be conducted in a pilot plant [83], with a greater representativeness given by a larger volume of samples or even initial injection tests in field scale [67].

In this work, we assumed that the coal seams of the Santa Terezinha and Chico Lomã deposits would have similar adsorption isotherms because they have some similar characteristics and parameters (Table 1), and because they are contiguous (Figure 1) and almost spatially continuous [54]. This approach can be accepted at this assessment stage because the aim is to validate the methodology and demonstrate a minimum of feasibility for the Chico Lomã deposit as a CO₂ potential reservoir. However, subsequent studies should properly generate adsorption isotherms with samples from the Chico Lomã deposit.

Although the coal from the Chico Lomã deposit has a high average ash content (40.0 to 56.2%), and this content usually results in a blockage of the pores, thus generating a low adsorption capacity [69,84,85], detailed studies, which are beyond the scope of this work, addressing this aspect as well as porosity and permeability should be conducted in the future. For comparison, in the studies conducted by De Silva et al. [47], coal deposits with an ash content of up to 57% were considered for CO₂ storage.

Another important aspect, that should be incorporated into future work, is the structural geology information on faults and fractures, which should be added to the 3D model [86], such as the SE-NW striking fault system presented by Burger et al. [87] for the Santa Terezinha deposit. The lack of structural geology data directly impacts the 3D model and therefore the volume used in the capacity calculation. Incorporating an understanding of structural geology into the geological model is a logical approach that can enhance their accuracy [88]. Studies on natural analog CO₂ reservoirs show that leakage to the surface through faults or other geological structures is globally rare, obtaining very low leak rates, even in places that would not be considered adequately safe for CO₂ storage [89,90,91,92]. The most likely routes for leakage are improperly sealed wellbores [2,93,94].

The methodology proposed in this paper effectively defines the coal volume within the best absorption limit in 3D, providing a scientific basis for determining denser sampling and better defining the location of CO₂ injection points, or even gas production wells (in the case of ECBM), ultimately leading to safer and more cost-effective research programs. The advantage of this method, besides its intrinsic implicit modeling advantages [42], is its ability to provide the smoothest surface of interpolation [40], which is ideally suited for geological modeling [41], and also that it can provide a more realistic volume than the common product of the deposit area multiplied by the average thickness. This methodology is ideal for the initial stages of the Exploration Phase of assessing a deposit or region for CO₂ geological storage, as the Site Selection and Initial Characterization stages [46].

Three-dimensional numerical models that consider regular model grid block formats or simplified tabular coal seams, e.g., Perera et al. [95], Ekneligoda and Marshall [96], Pratama et al. [97], Asif et al. [98], and others, would also benefit from this methodology, adding greater geological reliability with the better representativeness of the contact surfaces generated from the 3D implicit modeling of the geological contacts of the coal seams from the drillhole intervals.

However, drawbacks of the methodology presented here are its high dependence on the representativeness of available CO₂ adsorption isotherms, the assumption of geological homogeneous coal seams, and the number of accessible drillholes with trustworthy geological data that will allow the coal seams to be modeled reliably in terms of their depths. If the adsorption isotherms are not representative or if there are not enough drillholes, the capacity calculations may not be realistic. In terms of pressure gradient, De Silva et al. [47], analyzing a comprehensive set of data for CO₂ storage in coal seams, concluded that better estimation can be obtained by expressing adsorption capacity as a power function of pressure rather than assuming a linear relationship between adsorption capacity and pressure. Future research should focus on refining these aspects, optimizing sample collection and processing, and exploring their applicability in various geological contexts to improve the reliability of the estimative, enabling the CO₂ project to move into the Site Characterization Phase [46].

In a subsequent economic viability assessment, the ECBM potential will also need to be considered along with CO₂ injection in the Chico Lomã deposit, which was not considered in this initial appraisal. Recent studies have shown a high correlation between coal storage capacity, maceral composition, and vitrinite reflectance values [6,85]. For Chico Lomã, the fixed carbon and vitrinite reflectance average values are 13% and 0.63%, respectively [55,99], which also indicate a high potential for CO₂ sequestration.

Another important aspect to take into account is the thermal effects on coals and black shales in the proximity of dikes and sills of the Serra Geral magmatism (Cretaceous) [17,100,101,102,103]. In the present study, this type of information was not included in the 3D geological model. Nevertheless, recent works have modeled the sills and dikes in this region in three dimensions [104,105], so in future assessment studies, these data should be integrated.

The DOE NETL [46] equation is simplified and does not consider CO₂ density variations or fluctuations. CO₂ can reside in nanopores as sorbed and free phases, which may depend on the pressure, temperature, and pore size. However, as mentioned above, the methodology presented here can be incorporated into different flows by considering other more complete volume-based equations.

To assess the potential for CO₂ storage in the region, as the primary factor in the choice of the storage location is the proximity to the capture site [61], data on the location of thermoelectric plants were used from the Agência Nacional de Energia Elétrica [106]. This also provides data on the individual power and fuel for each plant. The CO₂ emission rates in tons per year were estimated for each plant, based on an assumed capacity factor of 80% and a heat rate assumption of 40% efficiency. Emission factors were taken from Eggleston et al. [107] and vary according to the fuel consumed by the power plant. According to Allah et al. [12] and Moore et al. [14], there is an opportunity for increased flexibility in the context of fossil fuel power plants, with CCUS technologies exhibiting lower environmental impacts, being competitive and reliable energy sources compared to renewables.

In the project region, there are 12 power plants within a radius of 55 km [22], and they have a total CO₂ emission per year of approximately 9.2 Mt [106]. Among these plants, 10 are diesel-fueled thermoelectric plants and 2 have natural gas-based sources. Therefore, considering the calculated theoretical average capacity of 47.8 Gt of CO₂ (Table 3) in the Chico Lomã deposit and the emission of 9.2 Mt/year emitted in the region by the 12 power plants, it is possible to achieve a potential for CO₂ storage for at least 520 years. The study area therefore has a high potential for geological storage of CO₂, since the Chico Lomã coal seam shows both lateral and depth continuity, making it possible to install a hub for CO₂ geological storage captured in all the thermoelectric power plants in the region.

6. Concluding Remarks

–

The innovative methodology proposed, which combines the most advanced 3D implicit modeling techniques and coal seam sorption isotherms analytical data, proves to be an efficient alternative for evaluating this type of geological reservoir in CO₂ storage studies, offering a new alternative that could be applied not only to other coal deposits in this southern Brazil province but also to other unmineable coal deposits around the world.

–

The coal seams of the Chico Lomã deposit exhibit thicknesses, volumes, lateral continuity, and depths compatible with the requirements considered worldwide for the geological storage of CO₂ in unmined coal deposits.

–

The theoretical total CO₂ storage capacity of 47.8 Gt achieved in this study is compatible with and could mitigate CO₂ emissions from local thermoelectric power plants, while still providing a surplus for storing other potential stationary industrial sources in the surrounding areas for more than 500 years.

–

Given the great geological heterogeneity within these layers, future studies are recommended to include a greater number of detailed adsorption isotherm analyses for each coal seam, as well as more drillholes reaching the seams at greater depths, to provide more realistic data.