Introduction

The storage of CO2 is one of the alternatives to reduce CO2 emissions to the atmosphere. Geological storage of CO2 is one of the best methods due to the large storage capacity over a long period. Previous study has shown that geological storage can contain huge volumes of CO2 while also adding benefits such as reduced negative impact and enhanced oil recovery (EOR) (Vo Thanh et al. 2019; AlRassas et al. 2021; Khanal and Shahriar 2022; Safaei-Farouji et al. 2022). Geological sequestration in coal seams allows the storing of CO2 since CO2 will bind to the coal, causing the CO2 to be trapped physically in the coal (Wahid et al. 2018). Adsorption is the primary mechanism for CO2 storage in coal seams, accounting for approximately 95–98% of total storage (De Silva et al. 2012). Moreover, CO2 urges CH4 to escape from the coal seam and allows for enhanced coal bed methane (ECBM) which may minimize the cost of sequestration (Day et al. 2010; Anggara et al. 2014, 2016).

Low-rank coal can adsorb CO2 in higher volume than high-rank coal (Kolak and Burruss 2004; Sripada et al. 2018), but low-rank coal has the most significant water adsorption capacity compared to the other coal rank (Liu et al. 2018). Changes in moisture content have become an exciting topic in the field of identifying the possibility of increasing the storage of CO2 in the coal seam (Hao et al. 2018; Gao et al. 2019; Abunowara et al. 2020). The comparisons between dry and wet coal have been analyzed in many studies to identify the moisture effect on CO2 storage capacity (Pan et al. 2010; Švábová et al. 2012; Chen et al. 2018).

Coal powder samples have been mainly used to analyze the influence of moisture on the CO2 adsorption (Gensterblum et al. 2013; Abunowara et al. 2020). Powdering process has affects porous structure, especially in low-rank coal (Mangi et al. 2022). Using coal powder samples for the adsorption experiments requires additional analysis since physical changes cause differences in the amount of gas adsorbed. The varying sizes of coal particles result in different gas adsorption capacities (Anggara et al. 2010). Nevertheless, one of the main problems with using coal powder is that it hardly represents the in situ underground storage conditions that occur in a compact form. To address this issue, one of the ways is to compare gas adsorption on coal powder and compact coal samples. A study has shown different adsorption capacities on coal powder and coal lumps due to different surface areas (Kim et al. 2019).

Various experimental adsorption methods are not only based on coal samples’ conditions but also on pressure and temperature setups. Coal injected with CO2 injected into coal under high pressure was a commonly used setup since it increases the CH4 recovery and CO2 storage rates (Busch et al. 2003b; Zhu et al. 2019). However, injecting CO2 under high pressure resulted in problems such as reduced coal permeability (Wang et al. 2016) and the possibility of over-caprock integrity and leakage (Masum et al. 2022). Temperature is also a critical factor in adsorption since higher temperature will increase the kinetic energy of the gas molecules resulting in lower adsorption capacity (Zhou et al. 2019). Furthermore, the effect of CO2 injection above 4 MPa at 323.15 K on the cumulative desorption of CH4 gas is weakened (Wen et al. 2022).

Isotherm adsorption model is used on experimental data to identify how CO2 adsorbed on coal. Langmuir and Freundlich are two adsorption isotherms parameters that are widely used to analyze adsorption capacity (Kalam et al. 2021). Even though Langmuir and Freundlich are commonly used, the results will vary depending on the adsorbents and gas injected (Guarín Romero et al. 2018; Hao et al. 2021). The adsorption characteristics are correlated with various coal compositions (Laxminarayana and Crosdale 1999; Dutta et al. 2011; Anggara et al. 2014, 2016). Thus far, none of these papers have successfully obtained isotherm parameters using various coal sample conditions.

Coal from South Sumatra Basin has various moisture contents (Sosrowidjojo 2013) and consists of mainly low-rank coal (Amijaya and Littke 2005). South Sumatra coal basin is recognized as one of Indonesia's largest and most important coal mining regions (Amijaya and Littke 2005; Belkin et al. 2009). Coal from this area has CBM potential resources about up to 40% out of total potential resources in Indonesia (Wahid et al. 2018). The evolution of research from CBM to the possibility of ECBM recovery improves ECBM study in the South Sumatera Basin. Coal samples from two coal seams in the South Sumatera Basin were taken and proven to have the potentials of CO2 geological storage and ECBM recovery (Anggara et al. 2010; Anggara 2017). Numerical simulation has also been done on South Sumatera CBM field to discover a possibility of ECBM recovery (Wahid et al. 2018). Furthermore, South Sumatera coal exhibits the heterogeneous coal characteristics, which affect CO2 adsorption on coal (Afikah et al. 2018). Most of these studies have concentrated only on some coal samples representing adsorption characteristics from an area. Meanwhile, different coal seams have different surface areas (Karayiğit et al. 2018) and pore parameters (Wu et al. 2014). The differences would lead to a variety of gas adsorption capabilities. Increasing accuracy is a meaningful value for predictive CO2 adsorption on South Sumatera.

This study aims to comprehensively investigate coal characteristics and sample preparation for CO2 adsorption capacity and analyze isotherm parameters. In contrast with another study, a coal sample was taken from the same coal seam in a different area to highlight the differences in similarity of organic matter for adsorbing CO2. Even though this study used samples with various conditions, CO2 adsorption was done with low-medium pressure at 318.15 K to consider the possibility of CO2 storage and ECBM. The volumetric method was used to support CO2 adsorption in various forms and water conditions (Battistutta et al. 2012; Kim et al. 2019), even though equipment's ability of the volumetric method to detect up to high is not used to the fullest. The progress to identifying CO2 adsorption capacity on coal was detailed in Fig. 1. Firstly, three coal samples were taken from the thickest coal seam in the South Sumatera basin (seam B) from a different area (Fig. 2). Then samples were prepared onto coal powder and coal block in dry and raw coal conditions. CO2 was injected with the volumetric method on 318.15 K and a maximum pressure of 3 MPa. Langmuir and Freundlich's isotherm model were used to obtain the isotherm parameter for the study area. Finally, coal characteristics and organic matter are used to evaluate the CO2 adsorption properties of coal.

Fig. 1
figure 1

General sketch of CO2 adsorption on coal with various conditions

Full size image
Fig. 2
figure 2

The location of the research area in South Sumatra Basin, Indonesia

Full size image

Sample preparation

Coal characteristic measurements

Samples were taken from South Sumatra Basin, Indonesia, and coal seams were selected from the thickest coal seams in three areas, namely West Banko (WB), East Banko (EB), and North Muara Tiga Besar (NMTB). Coal samples slightly differ in coal lithotype where coal samples from WB were banded-dull, coal samples from EB were dull, and coal samples from North Muara Tiga Besar NMTB were banded-dull (Fig. 3). The coal characteristics were performed based on the proximate and petrography analysis. The moisture content of coal powder was determined based on the ASTM D3173-73 guidelines by weighing approximately 1 g of coal, heating it to 105 ℃, and then re-weighing the dry sample at room temperature. Organic petrography was done by crushing the samples into 0.25 mm particles, mounting them in polyester resin, desiccating for 12 h, and polishing one of the flat surfaces. The polished coal sample was analyzed under a reflected light microscope to determine the maceral and huminite reflectance. Maceral analysis was conducted by counting 550 points per sample. Huminite reflectance was determined by analyzing 50 huminite grains as described in the ASTM D2798-06 guidelines. Huminite reflectance can also be defined by the volatile matter value where the equation follows (1) (Rice 1993).

$$R_{0} \left( \% \right) = - 2.172 \times log \left( {VM_{daf} } \right) + 5.092$$
(1)
Fig. 3
figure 3

The lithotype of coal samples from a WB, b EB, and c NMTB

Full size image

Adsorption measurements

The coal samples used for the adsorption analysis were coal powder crushed into 0.25 mm (60 mesh) particles and coal block (1 × 1 cm) created from the irregular-shaped coal, shaped and smoothed with sandpaper (Table 1). EB coal block sample was found to be slightly more brittle than the other coal block samples and have many cleats. The adsorption experiment used coal block and 5 g of coal powder from all coalfields. For dry coals, coal samples were dried for 2–5 h in a vacuum oven at 105 ℃ until the weight of coal remained constant to remove the moisture content. The dried coal samples were transferred abruptly to the sample cell to avoid any exposure to oxidation or moisture after the drying process. The steps of the experimental process to identify CO2 adsorption on coal are figured on the flowchart (Fig. 4).

Table 1 Coal sample from all coal fields in different coal forms
Full size table
Fig. 4
figure 4

Flowchart of experimental process in this study

Full size image

Experimental setups

The volumetric method was selected to determine coal adsorption. This study used different equipment to analyze coal powder and coal block. Conventional equipment analyzes CO2 adsorption on coal powder since this equipment only allows a small volume. Newly manufactured equipment to hold large volumes use for analyzing CO2 adsorption on coal blocks. The coal block adsorption equipment has been curved on the bottom to allow the CO2 adsorption on every side of the cube. For coal powder, the \(V_{{{\text{rc}}}}\) was 40 cm3 and \(V_{{{\text{sc}}}}\) was 75 cm3 (Fig. 5 a). The \(V_{{{\text{rc}}}}\) and \(V_{{{\text{sc}}}}\) were the same for the coal block, 201 cm3 (Fig. 5 b). The entire setup was placed in a water bath at a constant temperature (318.15 K). The \(V_{{{\text{void}}}}\) was determined by subtracting the volume of sample from the volume of empty sample cell (Busch et al. 2003a). The sample volume was determined by dividing the mass of coal over the coal density. Adsorption analysis was conducted under six pressure steps (0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 MPa) at 318.15 K. To reflect the in situ condition, the highest pressure in this experiment was adjusted at 3 MPa as the target was encountered at a shallow depth (Sosrowidjojo 2013). Higher pressure would not be suitable for reservoir conditions. CO2 was subsequently introduced into the reference cell and allowed to equilibrate until no changes in the pressure were observed for 30 min. Finally, the reference cell was opened, and the system was allowed to equilibrate until there was no change in the pressure for 24 h.

Fig. 5
figure 5

Volumetric method for a coal powder and b coal block

Full size image

Gibbs’s excess adsorption equations were used in this study to calculate the amount of adsorbed gas on the samples described as follows Eq. (2) (Goodman et al. 2004).

$$\Delta n^{ex} = \left( \frac{1}{RTm} \right)\left( {V_{rc} \left( {\frac{{P_{rci} }}{{Z_{rci} }} - \frac{{P_{rcf} }}{{Z_{rcf} }}} \right) - V_{void} \left( {\frac{{P_{scf} }}{{Z_{scf} }} - \frac{{P_{sci} }}{{Z_{sci} }}} \right)} \right)$$
(2)

where the compressibility factor was calculated as follows Eq. (3) (Abunowara et al. 2020).

$$Z = 1 + \left( {0.083 - \frac{0.422}{{T_{r}^{1.6} }}} \right)\frac{{P_{r} }}{{T_{r} }} + \omega \left( {0.139 - \frac{0.172}{{T_{r}^{4.2} }}} \right)\frac{{P_{r} }}{{T_{r} }}$$
(3)

where the reduced pressure were described in Eq. (4), and the reduced temperature were described in Eq. (5) (Abunowara et al. 2020).

$$P_{r} = \frac{P}{{P_{c} }}$$
(4)
$$T_{r} = \frac{T}{{T_{c} }}$$
(5)

where the critical pressure of CO2 is 7.39 MPa, the critical temperature of CO2 is 304.2 K and the acentric factor of CO2 is 0.224.

Results and discussion

Coal characteristic

The results of proximate and organic petrography analysis are shown in Table 2. The moisture content of samples from three different area varied, but the fixed carbon and volatile matter values were comparable (Fig. 6 a). Moisture has more substantial changes than fixed carbon and volatile matter compared to the lowest value of moisture content, fixed carbon, and volatile matter (Fig. 6b). A sample from West Banko (WB) showed lower moisture content (16%, a.r) than coal from East Banko (EB) (22%, a.r) and North Muara Tiga Besar (NMTB) (23%, a.r). Organic petrography showed huminite in WB coal sample is 64.43%, EB coal sample is 58.92% and NTMB coal sample is 52%. Liptinite was moderately present (20–25%) but inertinite occur with low concentrate (12–21%). Compared to other samples with the lowest particular maceral content, WB contain the highest huminite, EB contain the highest liptinite and NTMB contain the highest inertinite (Fig. 7). The WB coal sample had low ash yield and mineral due to the coal deposition, and no significant marine influence has been observed in this research area (Amijaya and Littke 2005). Ash yields are generally less than 5% correlating reasonably well with the mineral determined microscopically. Even though the moisture and organic content showed different values, the huminite reflectance was similar (0.49–0.50%). Based on the reflectance huminite classification (Thomas 2013), these coal samples are low-rank coal and classified as high volatile bituminous C.

Table 2 Results of proximate analysis of coal samples
Full size table
Fig. 6
figure 6

a The result of proximate analysis and b the differences coal samples with samples contained the lowest value of moisture, fixed carbon and volatile matter

Full size image
Fig. 7
figure 7

a The result of organic petrography and b the differences coal samples with samples contained the lowest value of huminite, liptinite and inertinite

Full size image

CO2 adsorption measurement

CO2 adsorption was analyzed over two parameters such as coal forms and moisture content. The different parameters showed the different times to reach equilibrium. The raw coal block samples took the longest, whereas dry coal powder took the shortest time to reach equilibrium. On average, the adsorption test on raw coal powder required 12 h to reach equilibrium in every pressure step; meanwhile, the dry condition only required 6–10 h. Experimental data for the raw coal block showed that it required 24–30 h to reach equilibrium in every pressure step. Meanwhile, the dry condition of the coal block needs 14–16 h in every pressure step.

For each parameter, such as coal form and moisture content, the result demonstrated a similar amount of CO2 adsorption (Fig. 8). The highest amount of CO2 adsorption was found on coal in powdered and dry conditions, while the lowest amount in coal samples with block and raw conditions. The difference in adsorption with different parameters gets higher along with higher pressure. The difference in moisture content is more significant than coal form, mainly found on lower pressure (0–1 MPa) when in raw coal, it is tough to find the difference; meanwhile, on dry coal, even in lower pressure, the difference is noticeable. Based on these findings, the method of preparing samples for CO2 adsorption capacity measurement is an essential value. The CO2 adsorption capacity of the coal study area was also compared with similar coal ranks from different coalfields worldwide (Table 3). By the comparison with another sample, this research found a similar trend, such as dry condition has bigger capacity than coal with moisture content, and compact coal sample has a lower capacity than powder coal. The comparison shows this study has a lower value than other research because this research uses lower pressure due to consider not only CO2 storage but also ECBM. There is some uniformity in the outcomes across all condition, but there is still a minor variance. Coal characteristics play a critical role in the results despite crushing and drying. WB and EB coal samples had the highest and lowest CO2 adsorption capability under all condition.

Fig. 8
figure 8

CO2 adsorption on the difference between coal form and moisture: a powder samples under raw conditions, b powder samples under dry condition, c block samples under raw conditions, and d block samples under dry conditions

Full size image
Table 3 Summary of the CO2 adsorption on low rank under various conditions
Full size table

Isotherm analysis

This study chose the Langmuir and Freundlich isotherm model to fit the experimental data. The Langmuir adsorption model defines the relationship between gas concentration and pressure for any coal on monolayer adsorption capacity. Ga detailed calculation as follows Eq. (6) (Tiab and Donaldson 2016).

$$G_{a} = V_{L} \left( {\frac{P}{{P + P_{L} }}} \right)$$
(6)

The Freundlich model is described the heterogeneity of coal surface by occurrence different adsorption energies on multilayers formation (Mahmoud et al. 2019). \(Q_{e}\) calculated with detailed as follows Eq. (7) (Guarín Romero et al. 2018).

$$Q_{e} = K_{f} PCO_{2}^{\frac{1}{n}}$$
(7)

The least-squares method in the Excel solver function was applied to estimate the Langmuir and Freundlich isotherm models from the experimental data and draw the fitted line (Fig. 9). The correlation (R2) from experimental and model show the fitting of the Langmuir as well as Freundlich isotherm model (Table 4), indicating monolayer as well as multilayer adsorption of CO2 was observed.

Fig. 9
figure 9

Comparison of different CO2 adsorption experiment and isotherm fitting curve from WB coal samples with various conditions: a powder samples under raw condition, b powder samples under dry condition, c block samples under raw condition, and d block samples under dry condition

Full size image
Table 4 Constant CO2 adsorption isotherm model
Full size table

Langmuir parameter on CO2 adsorption on coal

The Langmuir volume capacity shows similar results under all conditions, but the difference is discernible. WB coal sample found as the highest, and NTMB was found to have the lowest Langmuir volume capacity of all conditions (Fig. 10). Comparison of other samples on raw and dry conditions with WB coal sample showed moisture content had a significant effect (Fig. 11). The difference on raw condition easier to define than dry condition, expect on raw block coal samples. This condition related to the EB coal sample in the block form has a more visible cleat where gas is more accessible to adsorption into the coal sample. Langmuir pressure showed that the increasing water content and coal forms lead to the increase of Langmuir pressure. Previous studies have reported that PL was reciprocal to \(V_{{\text{L}}}\), especially on the low-rank coal (Gensterblum et al. 2013). \(P_{{\text{L}}}\) in the raw and block condition was higher than the dry and powder. As a result, CO2 is more quickly absorbed in dry and powder conditions than in raw and block ones.

Fig. 10
figure 10

Langmuir parameters for CO2 adsorption of all measured samples: a Langmuir volume capacity and b Langmuir pressure

Full size image
Fig. 11
figure 11

The difference in Langmuir volume capacity on EB and NMTB from WB coal sample

Full size image

Effect of different conditions on CO2 adsorption capacity on coal

According to the preceding discussion, the crushing and drying processes increase CO2 adsorption in coal. Comparing the discrepancy in Langmuir volume is due to various coal condition shows drying process resulted higher differences than crushing process (Fig. 12). Pervious research has shown that crushing coal causes the pore network to become more open (Olajossy 2017), which increases the surface area and produces more available sites to adsorb more gas, so the gas adsorption capacity is more significant in coal powder. However, coal powder and coal block showed similar adsorption capacity in raw conditions. The similarity shows that although changes in the surface area exist, the pore size distribution is still the same (Qi et al. 2017), so the changes are insignificant. A un-grind with a connected pore network provides more adsorption capacity than a grinded sample due to possibly increasing, decreasing, and even blockage pore (Mangi et al. 2022).

Fig. 12
figure 12

The discrepancy in Langmuir volume due to various coal conditions

Full size image

Experimental CO2 adsorption capacity on low-rank coal resulted in the drying process having a significant effect. Low-rank coal contains much water because the absence of water in the pores causes significant changes in the coal (Yu et al. 2013; Olajossy 2017). Eliminating water content leads to losing water film on the pore wall surface and pore space (Zhao et al. 2018) and increasing micropore volume, which expands CO2 adsorption capacity (Yu et al. 2013). The drying process resulted in excessive hydroxyl and carboxyl function groups, which intensified the possibility of CO2 binding with active sites in coal (Abunowara et al. 2020).

Effect of coal characteristics on CO2 adsorption capacity of coal

This study uses correlation analysis (Dutta et al. 2011) to find the significant effect of the coal characteristics on CO2 adsorption on coal. Based on the proximate analysis, in raw condition the moisture effect shows more significant differences than fixed carbon and volatile matter (Fig. 6). The elimination of moisture content by the drying process resulted in fixed carbon, and volatile matter affects more on CO2 adsorption capacity (Fig. 11). Increasing fixed carbon is related to increasing CO2 adsorption capacity (Dutta et al. 2011) and this also happened at WB coal sample with the highest fixed carbon resulted in the highest CO2 adsorption capacity at raw and dry condition. The CO2 adsorption capacity increased with decreasing volatile matter (Ramasamy et al. 2014), and these phenomena resulted in the similarity of CO2 adsorption capacity in dry conditions due to EB and NTMB containing lower volatile matter than WB.

Organic matter content was essential for gas adsorption on coal (Mastalerz et al. 2004). Earlier research discovered huminite has considerable influence on CO2 adsorbs on coal (Crosdale et al. 1998) due to the organic matter has associated micropore connections (Mangi et al. 2022). A similar result was also shown on coal in this experimental result. The experiment on CO2 adsorption on coal samples shows a significant correlation between the role of huminite and CO2 adsorption on coal (Fig. 7). The highest huminite content was found in the WB coal sample with the highest coal adsorption capacity (Fig. 13). Huminite can adsorb CO2 effectivity due to its more micropore presence than other maceral content (Bakshi et al. 2020). WB coal sample contains the lowest liptinite compared with EB and NMTB but does not give significant results due to liptinite only weak positive correlation with pore network (Teng et al. 2017). In contrast to huminite and liptinite, inertinite inhibits adsorption processes (Mangi et al. 2020). The experimental result shows that NTMB contains the highest inertinite related with the lowest CO2 adsorption capacity. Minerals also can reduce gas adsorption capacity due to infilling and blockage of pores, cleats, fracture systems, and lower micropore surface area with inherent and extraneous minerals (Kumar et al. 2019; Mangi et al. 2020). By organic petrography analysis, the coal sample contains a mineral that possibly blocks coal pores and fractures (Fig. 14).

Fig. 13
figure 13

Huminite on coal sample on image under white light

Full size image
Fig. 14
figure 14

Mineral on coal sample under white light

Full size image

Summary and conclusions

The following can be concluded as different coal preparation methods to measure CO2 adsorption of seam B in different areas in the South Sumatera Basin, Indonesia, resulting in CO2 adsorption capacity.

  • Coal samples exhibited monolayer and multilayer adsorption of CO2 at various conditions.

  • CO2 adsorption capacity in all condition shows similar results but have slight differences.CO2 adsorption capacity in raw conditions was hard to find the significant difference between powder and block coal samples due to uncertainty on change pore network. The drying process allowed more accessible sites and increased CO2 adsorption capacity. The drying process increases the effect on coal characteristics, even though the drying process might not depict a realistic situation.

  • Coal sample from WB area more promising for CO2 storage and ECBM due to lower moisture and higher huminite than EB and NMTB areas.

This study is an experimental laboratory approach to comparing adsorption capacity where the preparation such as drying and crushing processes clearly cannot represent the in situ condition. Moreover, it is implausible to be able to dry up low-rank coal to its full potential in natural situations.