Mineralogical and geochemical changes in conglomerate reservoir rocks induced by CO2 influx at Mihályi-Répcelak natural analogue, NW-Hungary

A temporary solution to massive anthropogenic CO2 emissions can be the capture of industrial CO2 from flue gas and sequestering it in geological formations. For safe and effective storage of CO2, interaction processes in the rock-pore fluid–CO2 system should be known. Investigation of natural CO2 accumulations provides valuable examples to what physical and chemical effects could be expected during CO2 influx at future CO2 storage sites. One of the key controlling factors of the processes occurring in natural CO2 reservoirs is the lithology of the storage rocks, which is primarily determined by the formation conditions of these rocks. In this respect, the lithologies of individual CO2 accumulation areas influence the processes between the host rock, the pore fluid, and the CO2 in different ways. In the current study, we focus on a well-studied natural CO2 storage reservoir, namely the Mihályi-Répcelak area, NW Hungary. We provide insight into the so far unstudied conglomerate reservoirs that represent a stratigraphically deeper reservoir unit with significantly different lithology and pore water compositions compared to the sandstone reservoirs. Our results indicate that dawsonite /NaAlCO3(OH)2/ formation also affected the conglomerate reservoirs, which indicates that at least part of the CO2 could be trapped in mineral form. An important role of salinity in reducing the CO2 mineral trapping capacity of the storage system is also demonstrated. Furthermore, H isotope analysis of diagenetic kaolinite was applied to trace the origin of the pore water that was present during the rock formation. Based on the data, dawsonite formation was induced by the flux of meteoric water that infiltrated during a warm and humid period and mixed with ascending CO2.


Introduction
Due to the growth of industrial activity and the continuous expansion of the technological infrastructure since the 19th century, the number of CO 2 emitting point sources and the amount of CO 2 gas releasing into the atmosphere have increased dramatically. The atmospheric proportion of CO 2 has now exceeded 410 ppm under continuous, large-scale emissions (NOAA/ESRL 2019). Carbon dioxide accumulating in the atmosphere is thought to pose a heavy environmental risk to the society today, as it plays a major role in global climate change as a greenhouse gas, thus affecting the ecosystem and water balance of the Earth (IPCC 2007). Therefore, reducing CO 2 emissions has a priority for the future of the Earth, which can only be achieved through extensive international cooperation (e.g.: Kyoto Protocol 1998;2009/31/EC Directive 2009, Paris Agreement 2015. A possible way to reduce emissions is separating the CO 2 from industrial flue gas and storing it in deep geological reservoirs (Carbon Capture and Storage-CCS) (Arts et al. 2008;IPCC 2013;De Silva et al. 2015). One of the most important pre-conditions for using this technology is to understand and predict the geochemical changes occurring during the storage process as accurately as possible. Several methods can help to understand the underground processes, such as laboratory experiments (e.g., Hellevang et al. 2010;De Silva et al. 2015), geochemical model-based approaches (e.g., Xu et al. 2003;Király 2017), and study of natural CO 2 accumulation areas or CO 2 analogues (e.g., Worden 2006;Király et al. 2016). It is important to study characteristic properties of the CO 2 natural analogues to understand what physical and chemical processes in an industrial CO 2 storage project are likely to occur. In other words, what are the necessary conditions for the long-term safety and efficiency of CO 2 trapping in deep geological environments (Watson et al. 2004). Mineral and petrophysical (porosity, permeability) properties of rocks are primarily determined by the conditions of their formation. This is especially true for sedimentary rocks where sedimentation and diagenetic processes primarily influence, among others, porosity, permeability, and reactivity. Therefore, it is essential that we know the environment and formation mechanism of rocks when examining CO 2 natural analogues (Pearce et al. 2005;Holloway et al. 2007).
Former studies have focused mainly on sandstone reservoirs, whereas conglomerate rocks have been neglected in terms of CO 2 storage due to their generally small volume and heterogeneity. However, the main objective of this study is to analyze the lithology and the geochemical character of conglomerate core samples that serve as natural CO 2 reservoirs in the Mihályi-Répcelak area, gaining a more accurate understanding of how lithology and pore water geochemistry influence CO 2 impacts in a natural environment.
This paper focuses on a distinctive isolated conglomerate reservoir at Mihályi-Répcelak area. Previous studies of conglomerate rocks, especially regarding CO 2 storage, are subordinate in the literature despite evidence suggesting natural CO 2 can accumulate in conglomerate reservoirs (Rauzi 1999). Nevertheless, conglomerate reservoirs could have a potential for CO 2 storage as based on our results, mineral trapping of CO 2 might also take place in the conglomerate rocks. Not considering mineral trapping, it was demonstrated by laboratory experiments (Wang et al. 2016) in Janggi basin (South Korea) that, due to favorable reservoir parameters, conglomerate rocks can store almost three times as much CO 2 as sandstones with similar porosity and depth.
Consequently, it also points to the importance of detailed or even targeted investigation of potential conglomerate reservoirs like the ones at the Mihányi-Répcelak area.
The present study investigates a well-known and extensively studied natural CO 2 accumulation at Mihályi-Répcelak area, north-western Hungary, where effects of CO 2 accumulation were examined in details for sandstone reservoirs and their caprocks (Király et al. 2016;Király 2017). A further study from the Mihályi-Répcelak area deals with the question of stable isotope composition of carbonate minerals, mainly dawsonite (Cseresznyés et al. 2017). The results show that dawsonite precipitated from magmatic CO 2 , and the origin of OH is the pore water. One of the main questions of this study is whether the origin of dawsonite is the same in the conglomerate rock as in the sandstone.
Furthermore, accumulated CO 2 appearing in conglomerate rocks is exclusive and has not been studied before at Mihályi-Répcelak area nor worldwide, even though CO 2 trapping could be observed from new perspectives. The results from this study could be used to widen the knowledge about CO 2 -rock-porewater systems and help to make industrial CO 2 geological storage safe and effective. Furthermore, conglomerates in the area represent a deeper stratigraphic unit than the sandstones, with remarkably different lithology and hypersaline pore water (Mészáros et al. 1979).

Sedimentology and stratigraphy of the Little Hungarian Plain
Mihalyi-Répcelak area is located in the Little Hungarian Plain, NW Hungary ( Fig. 1), which is an extensional basin formed by NE-SW-striking faults during the Miocene and filled with Neogene sediments (Horváth 1993;Tari 1994;Kovač et al. 2007).
The pre-Pannonian sedimentary filling of Little Hungarian Plain is similar to other parts of the Pannonian Basin System. From Carpathian to Badenian, marine sediments of the Paratethys and coarse-grained alluvial sediments were formed (Sacchi and Horváth 2002;Kovaĉ et al. 2007;Nagymarosy and Hámor 2012). The marine sediment sequences are interrupted by several tuff interbeddings which are connected to volcanic centers of Pásztori, Szany, and Tét area (Balázs and Nusszer 1987). Sarmatian sediments are represented by restricted brackish deposits of the Pannonian Lake. At the end of the Sarmatian, the Pannonian Lake lost all connection with oceans which caused gradual appearance of freshwater deposits and endemic fauna (Magyar et al. 1999).
Uplifting and erosion of Alpine-Carpathian region caused significant quantities of sediments transported to the Pannonian Basin from the NW and started to fill the basin in the Late Miocene (Pogácsás 1984;Vakarcs et al. 1994;Juhász 1994Juhász , 1998Magyar et al. 1999Magyar et al. , 2013. The first shelf in the Little Hungarian Plain had formed 10 million years ago in less than 1 million years . Starting from the Middle Miocene, the sedimentation was controlled by a prograding delta system arriving from the NW. The sediment sequence is uniform throughout the Pannonian Basin and reaches several km in thickness. This sequence hosts the most significant hydrocarbon, natural CO 2 , and geothermal fluid resources basin wide. The stratigraphically deeper units are characterized by alternating clayey-silty layers with fine-grained sand covered by a silt-dominated sequence. The shallower units are more dominated by medium-to-coarse grain size sand layers that are overlain by clayey layers, the latter representing fluvial sediments (Juhász 1994;Sztanó et al. 2013Sztanó et al. , 2016 Fig. 1C).
At some parts of the basin, the riffles of the basement reached the water level forming islands where abrasion sediments settled (Békés Formation) as assorted, unbedded conglomerates ( Fig. 1C). Base material of these sediments is fine-grained sandstone with rounded gravels. It reaches its maximum thickness at the limb of structural highs, while it pinches out vertically (Juhász 1994). During recent research, Békés Formation was investigated by core samples, which are from 1411.5 to 1446.5 m depth. In every case, this formation is overlain by a dark gray to light gray, laminated to structureless clay marl serving as impermeable clayey cap rock.

Volcanic activity during the Neogene in the Little Hungarian Plain
In the Little Hungarian Plain, Late Miocene magmatic activity took place in two main phases. The Upper Miocene activity is represented by a buried magmatic body characterized by a large positive magnetic anomaly in Pásztori area (Balázs and Nusszer 1987). Several oil exploration drills had reached the magmatic series and discovered a unique series of alkaline trachyte in the Pannonian Basin (Kőrössy 1958). The appearance of sea sediments at the beginning of the volcanic series suggests that volcanism began in a sub-sea environment (Kőrössy 1958;Harangi et al. 1995), but increased above sea level in the Sarmatian (Tari 1994). The second magmatic phase began 6 million years ago and volcanic cones built up by alkaline basalt, basanite, and trachybazalt formed in the Little Hungarian Plain area (Balázs and Nusszer 1987;Harangi et al. 1995).

Forming of CO 2 accumulation
As a result of volcanic activity, several CO 2 reservoirs developed in the area in the sandstone and conglomerate bodies of the Middle Miocene sedimentary sequence (Mészáros et al. 1979) (Fig. 1B). In the study area, 26 of these reservoirs have been exploited and produced in the last decades. The origin of the CO 2 was identified as magmatic (Palcsu et al. 2014;Vető et al. 2014) based on stable isotope composition of the produced gas.

Samples
Two drill cores representing conglomerate facies (RM19-7Rkong and RM19-8Rkong), the only samples available from over 40 boreholes, were studied whose features are shown in Table 1. The drill cores available from the area predominantly consist of layered sandstones, aleurolites, and claystones.
Both studied samples are oligomictic, well cemented, matrix supported conglomerates. In both cases, the matrix is light gray-colored and fine-grained. The appearing clasts can be sorted in three groups, which are the following: (1) kaolinitic clasts that are 0.1-2.5 cm and make at least 50 v/v% of all clasts in the rocks, (2) mm-to-cm scale fossils, and (3) metamorphic and micrite rock fragments.

Petrography
During the sample preparation, two-component araldite resin was used with blue dye to make the porosity of the rock visible. The petrographic examination of the samples was carried out using an NIKON Eclipse LV100 POL type polarization microscope equipped with a NIKON DS-Fi1 digital. To get a better understanding of mineralogical and textural information, scanning electron microscopy (SEM) measurements were carried out. The Amray 1830-type electron microscopy was equipped with EDAX PV 9800 energydispersive X-ray spectrometer. 20 kV accelerating voltage was used in each case; the primary electron current was 1 nA. Backscattered electron (BSE) image and secondary electron image (SEI) were used during the analyses. Petrographic characterization and SEM analysis were carried out at the Eötvös Loránd University, Hungary.
Mineral modes were determined using X-ray Powder Diffraction (XRD) measurements. However, it must be noted, that several mineral phases identified under polarization microscope were not detectable during the XRD measurements. That is why, the mineral composition of the kaolinitic clasts (discussed in detail later) was estimated using image analysis on high-resolution SEM images. Volume of pores pace (in v/v%) was taken by the difference between the total pixel number and the pixel number of all the minerals. After subtracting, the area filled with minerals from the size of the whole image. The error of this porosity calculation method can be up to 5-10 v/v% due to the granules lost during grinding.

Micro-XRD
Check the cleanness of the separated kaolinite samples to the stable isotope analysis, Micro-XRD were carried out at the Institute for Geological and Geochemical Research, Budapest. Micro-XRD analysis was performed following Kovács et al. (2021) on an RIGAKU D/MAX RAPID II diffractometer, which is a combination of a MicroMax-003 third-generation microfocus, sealed-tube X-ray generator, and a curved imaging plate (IP) detector. The diffractometer is operated with CuKα radiation generated at 50 kV and 0.6 mA. The powdered samples for the micro-diffraction measurements The temperature was estimated by average geothermal gradient (Dövényi et al. 1983)  are encapsulated in a borosilicate-glass capillary, with a diameter of 0.3 mm and wall thickness of 0.01 mm, by a vertical manual charging process. Then, the capillary is analyzed by the micro-diffractometer in transmission mode with a beam spot diameter of 100 μm. In each measurement, 0.5-1 mg sample is placed in the funnel end of the capillary and the sample is tapped into the narrow portion. The measurement time takes about 10 min. IP is read by a laser scanning readout system in about 1 min. 2DP RIGAKU software is used to record the diffraction image from the laser readout and the operator can determine the area to integrate for a 2θ versus intensity plot. This plot is read into RIGAKU PDXL 1.8 software for data interpretation. The diffraction patterns were processed using SiroQuant V4 software, and the modal contents were determined by the Rietveld method.

Stable isotope measurements
Water contents and H isotope compositions were determined using a High-Temperature Conversion Elemental Analyser (TC/EA) attached to a stable isotope ratio measuring mass spectrometer (Sharp et al., 2001)  The hydrogen contents were calculated on the base of size of the mass 2 signal peak. The isotope compositions were expressed in the δ-notation (δ 2 H=(R 1 /R 2 −1)·1000, where R 1 and R 2 are the D/H ratios in the sample and the standard, respectively) in permil (‰) relative to V-SMOW. The H isotope compositions of laboratory standards (IAEA CH-7 polyethylene foil and NBS-22 oil) were used for calibration yielded an average reproducibility of <±2 ‰. A δ2H value of − 27.5 ± 0.7 ‰ (n = 4) was obtained for an in-house test amphibole sample (TOB-A1) that had been analyzed before at the University of Lausanne (− 29 ‰, Demény et al., 2005) and at the Université Claude Bernard Lyon 1 (− 28.6 ±6.0 ‰, n = 5; Fourel et al. 2017). Note that the He flushing method without heating resulted in a δ2H value of − 28.8 ± 4.8 ‰ (n = 4), closer to the published values. The NBS-30 biotite standard was measured, but yielded a large scatter (from − 64 to − 56 ‰) in accordance with the observations of Qi et al. (2014).

Calculation of δD values in pore water in equilibrium with kaolinite
The calculations were based on results of δD measurements in kaolinite and published fractionation equations (Faure 1998;Savin and Epstein 1970;Suzuoki and Epstein 1976;Lambert and Epstein 1980). Formation temperature (T = 70 °C) was estimated from the current depth (1411.5-1446 m) of the studied samples, the average geothermal gradient at the Little Hungarian Plain area (40 °C/km) (Dövényi et al. 1983) and the average mean temperature of the area (11 °C) (Hungarian Meteorological Service 2018). Savin and Epstein (1970) and Suzuoki and Epsein (1976) suggested constant kaolinite-water hydrogen-isotope fractionation values of − 30 ‰ and − 23.4 ‰, respectively. On the other hand, Lambert and Epstein (1980) found temperature dependence in the fractionation and gave the following equation: 1000·lnα = − 4.53·106/T2 + 19.4. Assuming 70 °C as formation temperature (see above), the hydrogen-isotope fractionation value is − 19.1 ‰. Since α = (δDkaolinite + 1000)/(δDwater + 1000), the water composition can be calculated using the measured kaolinite compositions. To reduce the uncertainties inherent in the fractionation values given by the studies listed above, all of the calculated water compositions are reported and finally averaged.

Petrography of the conglomerates
The base material consists of moderately abrasive monocrystalline quartz, mica, and plagioclase. They appear as detrital minerals. Accessory apatite, glauconite, Ti-containing minerals (rutile or anatase), pyrite, and zircon were also found. In some cases, diagenetic quartz growth was observed on the detrital quartz particles. Mica appears unoriented and shows no signs of deformation or dissolution. The plagioclase particles of size 300-450 μm occur in minimal quantities (one or two grains) in both samples. The cement material of the rocks is carbonate based on scanning electron microscopic measurements and mainly consists of calcite. In addition, minimal amounts of siderite and ankerite appear.

Fossils in the conglomerates
In the samples, 25-40 v/v% of all clasts are fossils ranging in size from 100 to 2 mm, and in all cases, they are built up by calcite. Fossils in the samples can be classified into two groups. In the matrix Ostracoda, Gastropoda, and various shells and shell fragments are dominant. Additionally, dark brown, rounded micrite fragments Milionida sp., Elphidium sp., and Sprulina sp. foraminifera were found (Galácz and Monostori 1992).

Petrography of kaolinitic clasts in the conglomerates
Kaolinitic clasts in the studied conglomerates are rounded ( Fig. 2A). Results of staining method used in sample preparation implies that porosity in kaolinitic clasts is considerably higher (according to visibility of the blue epoxy) than in the matrix where the used blue dye can be detected only on micrometer scale.
According to the SEM analysis, the main mineral in the kaolinitic clasts is kaolinite (~53-70 v/v %) ( Table 2). Kaolinite appears only in the kaolinitic clasts ( Fig. 2A). The clasts display relic magmatic textures most likely representing intersertal structure with mica, albite, K-feldspar, Ti-minerals (rutile or anatase) apatite, and opaque minerals occurring in the clasts. These phases are considered to be part of the original material. Halite (Fig. 2B), dawsonite, and other carbonate minerals appear as diagenetic minerals. Dawsonite, a new mineral phase in the rocks is the strongest evidence of CO 2 presence (Worden 2006), is limited to the rim of kaolinitic  clasts. The dawsonites in the kaolinitic clasts appear as radial, fibrous minerals. The fibers reach over the edge of the kaolinitic clasts ( Fig. 2A), which indicates that dawsonite is a diagenetic mineral. Between the fibers of dawsonite, relict albite and K-feldspar are observed ( Fig. 2C and D). Modal composition of the kaolinitic clasts in the two samples is similar to each other except in the presence of halite (Fig. 2B) which can be identified only in RM19-7Rkong sample. However, mineral modes strongly differ in the samples. Quantity of albite decreases with the occurrence of dawsonite in the two samples. In the RM19-7Rkong sample, there is 3 v/v% surplus in dawsonite, while it has 6 v/v% less albite than in RM19-8Rkong sample ( Table 2). Pore space in the RM19-8Rkong sample is lower than in RM19-7Rkong (RM19-7Rkong: 20.36 v/v%; RM19-8Rkong: 13.36 v/v%).

Stable isotope measurements in kaolinitic clasts
The relatively high hydroxyl content (OH − ) of kaolinite allowed δD values to be measured in the kaolinitic clasts. Two types of samplings were carried out: (1) homogenized samples from kaolinitic clasts that are dawsonite barren and (2) controlled micro-sampling from preselected areas in the kaolinitic clasts near dawsonite precipitation. The results show that the samples containing dawsonite show slightly different δD values than the ones without dawsonite. The differences may depend on the quantity of dawsonite in the micro-sampling. The presence of dawsonite may cause a decrease in δD values. Results are shown in Table 3.
In dawsonite-bearing kaolinitic clasts, δD values are between − 63.0 ‰ and − 55.5 ‰, whereas in the RM19-8Rkong, the δD values are − 55.4 ‰ and − 55.9 ‰ (Table 3). Table 4 summaries the calculated δD values in pore water considering three different kaolinite water fractionation values, whereas the average δDwater values in dawsonite-bearing and dawsonite-free samples are shown in Table 5.

Depositional environment
Based on the foraminiferas, Miliolina sp., Elphidium sp., Spirulina sp. in the micrite rock fragments, represent euhaline water of Badenian-Sarmatian (15.9-11.6 Ma) (Galácz and Monostori 1992). However, fossils, Ostracoda, Gastropoda, and shell fragments, in the matrix refer to brackish water environment (Galácz and Monostori 1992). This means redeposition that involved the studied conglomerates happened in the Pannonian time and affected the Badenian-Sarmatian sediments. In both studied samples, the appearing clasts are elongated and show a slight orientation. This may refer to episodic events, like storms, mass movements, or changing weather conditions, which caused the redeposition and formation of the conglomerates from Badenian-Sarmatian and Pannonian sediments. The kaolinitic clasts, based on our petrographic observations, represent altered volcanoclastic fragments that have experienced transport resulting in rounded shapes. It is a question whether kaolinitization of the clasts precedes transport and sedimentation or is already the manifestation of diagenetic alteration.
The preservation of original volcanic intersertal texture strongly implies that the alteration process took place after transport and sedimentation. The relics of mica and feldspars, partly kaolinitized and the lack of quartz indicate that the original rocks, from which the kaolinitc clasts formed, were probably neutral pyroclasts. These rocks could be related to the Pásztori volcano, which were active during the same time period as the formation of studied rocks (~ 10 million year ego; Harangi et al. 1995) and marked as a possible source of CO 2 migration ). The dominant textural type, identified in the rocks derived from the volcanism from Pásztori volcano, is also intersertal (Schléder 2001), which also supports the connection between the studied rocks and Pásztori volcano. However, we cannot completely rule out the possibility that the former volcanic clast was transported already as kaolinite.

Dawsonite formation environment
After the deposition and burial of the sediments, diagenetic effects caused remarkable changes on the properties of rocks, especially in mineral composition including the formation of dawsonite. Its appearance as a new mineral is a result of CO 2 -pore water-host rock interaction.
Dawsonite fibers cross the border between the kaolinitic clast and the base material, which indicates that dawsonite Fig. 3 Dawsonite stability at the studied environment. Dawsonite precipitated in the edge of kaolinitic clast (XN, polarization microscope) (A), Albite relict occurred in the environment of dawsonite, which mixed with kaolinite (white circles) (B). CO 2 fugacity-pH diagram (Worden 2006), where the dashed line signed the actual CO 2 fugacity (C). Dawsonite stability depends on Na + concentration, pH, and SiO 2 activity (Hellevang et al. 2013) (D) appears as diagenetic mineral; hence, it was formed after the burial. Feldspar relicts-mostly albite and K-feldsparbetween fibers of dawsonite refer to the breakdown of albite and K-feldspar to form dawsonite (Fig. 3A). This observed process corresponds to models (Worden 2006;Zhou et al. 2014) that state the Al +3 content of dawsonite comes from dissolution of albite and K-feldspars; meanwhile, the source of Na + content of dawsonite could be either the dissolved albite and/or, if K-feldspar was the source of Al 3+ , Na + may have originated from the pore water itself. The presence of considerable concentrations of Na + in pore water is also supported by halite in RM19-7Rkong sample (Table 2), which refers the pore water being (over)saturated in NaCl.
We suggest that the strongly limited nature of dawsonite forming processes in the conglomerates, compared to overlying sandstones (Király et al. 2016), can be related to the high salinity and, therefore, low CO 2 solubility in the system. The appearance of halite confirms the high salinity in the pore water. In a NaCl-H 2 O-CO 2 system, CO 2 activity dramatically decreases as the salt content grows (e.g., Duan et al., 2006). Therefore, the presence of CO 2 is mostly expected as free phase, resulting in much lower reactivity compared to dissolved CO 2 and lower degree of pore water-reservoir rock interaction. However, it is notable that the formation of dawsonite vs. kaolinite depends strongly on the geochemical environment. Furthermore, we should consider the microenvironment of the observed mineral reactions. Kaolinite and dawsonite are intergrown around the rims of kaolinitic clasts in close textural relation with (Fig.3A,B), albite and K-feldspar relicts (Figs. 2D, 3B). This may indicate that a microenvironment favorable for dawsonite formation could have developed only at the rims of the kaolinitic clasts on the kaolinite-dawsonite equilibrium (Fig. 3A, B). In this geochemical system, besides CO 2 fugacity (Worden 2006, Fig . 3C), the formation of kaolinite or dawsonite depends on the pH, Na + concentration, and SiO 2 activity (Hellevang et al. 2013, Fig. 3D). The conditions for dawsonite formation were only appropriate at the rim of the kaolinitic clasts where silicate minerals could dissolve, whereas Na + and Al 3+ were supplied and CO 2 fugacity was adequately high for the stability of dawsonite. The exclusive occurrence of dawsonite on clast rims might be related to sudden increase in porosity from the matrix to kaolinitic clasts and the resulting in abrupt changes in the dynamics of the pore water flow. Thus, the controlled appearance of dawsonite is a combined effect of several phenomena. It is interesting that physical and geochemical conditions for the formation of dawsonite should have been appropriate in the whole kaolinitic clast; however, dawsonite formed only at the clast rims ( Figs. 2A, 3A).

Possible origin of the dawsonite-bearing kaolinite forming pore water
Considering that the rock is not composed of significant amount of hydrogen-bearing minerals (minerals with crystalline water or OHcontent), which would influence the δD values of the percolating water, the hydrogen-isotope composition of kaolinite represents the percolating water during the formation of kaolinite. Based on the rounded shape of kaolinitic clasts ( Fig. 2A), the alteration of sediments into kaolinite clasts may had started during early lithification, which means the present pore water may have come from the infiltrating meteoric water. Since the dawsonite-bearing kaolinite clasts yield almost the same water hydrogen isotope composition (− 35.4 ‰) as the dawsonite-free clasts do (− 33.2 ‰) (see Table 5), the isotopic compositions of the percolating solutions may have been similar during kaolinite and dawsonite formation (related to fluid origin), regardless the chemical differences that might have been determined by water-rock interactions However, the average calculated δDwater value (− 34.2 ‰) does not match the composition of present day meteoric water in Hungary (annual mean of − 62.3 ‰, Czuppon et al. 2018), but has slightly higher values. More likely, the calculated δDwater values reflect climatic condition of Late Miocene/Pliocene, which had largely homogenous, warm, and humid climate (Burch et al. 2006;Kovács et al. 2015), and caused a higher δD value in meteoric water (Fig. 4).
Considering the origin of pore water, the increased local geothermal heat flow, as results of volcanic activity in the area during Upper Miocene, caused emergence of convection flow in groundwater system. Mixing of magmatic water to the infiltrating solutions along with the CO 2 cannot be excluded, but, in this case, the meteoric water should have had even higher H isotope compositions than magmatic water, which has generally lower δD values (around − 80 ‰, Boettcher and O'Neil 1980;Kyser and O'Neil 1984).

Concluding remarks
The present study investigated effects of natural CO 2 flooding on a conglomerate formation at Mihályi-Répcelak area in Hungary. The studied conglomerate rocks formed by episodic redeposition during the Lower Pannonian era (~10 million years ago) and became CO 2 reservoirs in the same era.
The presence of dawsonite is the clearest evidence of CO 2 interaction, which means that at least part of the CO 2 trapped in mineral form. The variability in porosity and mineral composition within the studied conglomerates strongly influences the reaction processes induced by the presence of CO 2 . Beside the rock properties, the geological setting of the studied formation, facies, properties of surrounding rocks, and regional groundwater flow system also influence the CO 2 -related processes.
Our results highlight the prevailing influence of geological setting of reservoirs such as relief of basement, the sedimentary sequence and the hydrogeological properties of the area, thereby the salinity of the system, which dramatically reduces the potential of mineral trapping of CO 2 . Therefore, in case of a possible industrial CO 2 storage project, investigation of the mineral composition and caprocks alone is not sufficient. Significant attention should be given to the complete geological environment of the selected area. Otherwise, efficiency of CO 2 trapping during storage could be compromised, increasing the risk of any industrial CO 2 storage project.
Funding Open access funding provided by ELKH Research Centre for Astronomy and Earth Sciences.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.