The Rio do Rasto Formation is a notable sedimentary record in the Paraná Basin associated with western Gondwana continentalization among the Middle to Upper Permian ([17]; [11]; [25]). This unit reaches thicknesses up to 600 meters and outcrops along the eastern margin of the Paraná Basin in South Brazil. The Rio do Rasto Formation is divided in two lithostratigraphic units: the basal Serrinha Member, and the upper Morro Pelado Member ([9]).

The Serrinha Member consists of greenish-gray and purple claystones and siltstones interbedded with fine-grained gray sandstone ([26]; [11]; [25]). This unit may reach thicknesses between 150 and 250 meters, and the transition to the Morro Pelado Member is concordant and gradual, primarily characterized by an increase in sandy deposits upward.

The Morro Pelado Member corresponds to the thickest interval and may reach thicknesses of 250 to 300 meters ([26]; [11]). This unit is characterized by an increased occurrence of sandy bodies interbedded with claystones and siltstones throughout the succession. These fine-grained rocks become more reddish toward the top, indicating more oxidizing conditions than those in the basal portion ([13]; [25]). Additionally, paleosols and pedogenetic features become more prevalent towards the top of the Morro Pelado Member.

Due to the common intercalation of sandstone and fine-grained rocks with a variety of sedimentary structures, the Rio do Rasto Formation is interpreted as having been deposited by many paleoenvironments. These environments include transitional tidal flats ([26]); a wide deltaic system ([8]); a transition from marine to lacustrine settings ([3]); lacustrine to fluvial settings with associated aeolian deposits ([13]; [23]); a fluvio-deltaic system ([31]); and distributive fluvial systems ([24]; [16]).

Current paleoenvironmental and paleogeographic reconstructions of the Rio do Rasto Formation, as proposed by [16], point to distributive fluvial systems fed by distinctive drainage basins. These systems consisted of basal lacustrine deposits overlain by fluvial, overbank, and aeolian deposits. [30] also details the interaction between fluvial and aeolian processes at different frequencies and magnitudes. Together, these studies demonstrate the architectural complexity of the stratigraphic framework of continental sedimentary successions.

Petrographic thin section preparation was outsourced to the specialized company Petrografia BR, the samples were impregnated with blue epoxy for better visualization of the pore system. Subsequent petrographic analysis was conducted using an Olympus BX 41 optical microscope. This equipment enabled the description of the main mineralogical constituents and textural aspects of the samples, including mineral identification, types and distribution of porosity, grain size, types of intergranular contacts, cement nature, matrix presence, and evidence of diagenetic processes.

The adopted analytical method was point counting, with 300 points registered per thin section. The generated data was subsequently compiled and processed in Microsoft Excel spreadsheets. The samples were compositionally classified according to Folk (1968).

X-ray computed microtomography (Micro-CT) was used for three-dimensional characterization of the samples pore structure. This non-destructive technique is based on the principle of X-ray beam attenuation as it interacts with matter. Variations in material density and composition result in differences in attenuation coefficients, generating contrast that enables distinction between the solid matrix and pore spaces ([19]; [22]).

For the analysis, specimens were prepared as mini-plugs with approximate dimensions of 3 mm in diameter and 9 mm in height at the Porous Media and Thermophysical Properties Laboratory (LMPT/UFSC). Data acquisition was performed using a Zeiss Versa XRM-500 microtomograph.

Tomographic data post-processing involved three-dimensional reconstruction and quantitative analysis of the pore structure. The Avizo 3D software was employed for generating three-dimensional models, and calculating the porosity and permeability values while the C3DE software was used specifically for pore network and throat extraction. A fundamental step in this process was image binarization, where the application of a threshold (T) enabled digital segmentation of the image into two phases: solid matrix and pore space ([28]). To optimize this segmentation, Non-Local Means and unsharp masking filters were applied. This procedure facilitated the quantification of microstructural parameters such as porosity, connectivity, and pore size distribution.

Mercury intrusion porosimetry (MIP) was performed on four samples prepared as cubes with approximately 1 cm edges. The method is based on the principle that mercury, being a non-wetting fluid, requires the application of external pressure to penetrate the sample’s pores ([12]). The relationship between the applied pressure and pore radius is governed by the Washburn equation.

The test enables determination of effective porosity (ϕ) through the ratio between the volume of mercury intruded into the connected pores (Vp) and the total sample volume (Vt), according to the equation:

ϕ = Vp/Vt

In addition to porosity, the method provides pore throat size distribution derived from the curve of intruded volume versus applied pressure.

The petrographic analyses performed on samples from the Morro Pelado Member revealed four types of sandstones (Table 2). Drill log 21 proved to be much richer in quartz, whereas drill log 101 showed a much greater abundance of lithic grains and feldspars (Figure 4).

Figure 4. Framework composition of 15 samples from the Morro Pelado Member. Plotted on the Folk (1968) ternary diagram. In blue, samples from drill log 21; in red, samples from drill log 101.

Table 2. Percentage of porosity, quartz, feldspar, lithics and depth per sample.

4.1.1. Eolian Deposits Subsystem

Sandsheets have a total porosity of 18.3%, as observed in thin section. This porosity is predominantly primary intergranular, accounting for 70.9% of the total. The grains are fine to medium-grained, with shapes ranging from subangular to subrounded. Intergranular contacts are mostly long. Silica cement occurs locally, distributed sparsely within the rock framework without causing significant obstruction of the pore system. The mineral assemblage is dominated by quartz grains, accompanied by feldspars and oxides (Figure 5(A)).

Eolian Dunes exhibit more variable porosity (13% - 26%) in thin section, with notable development of secondary intragranular porosity (10% - 63%), evident in moldic pores and dissolved grains (Figure 5(B)). The grains are typically more rounded and establish elongated and point contacts. Mineralogically, quartz, feldspars, and oxides (24% - 44%) are observed, forming coatings around the grains (Figure 5(C)). Carbonate cement (4% - 8%) partially fills the pore spaces (Figure 5(D)).

4.1.2. Fluvial Channel Deposits

Lateral Accretion Bars exhibit total porosity ranging from 20% to 32%, as observed in thin section, with a predominance of primary intergranular porosity, accounting for 50% to 79% of the total identified pore volume. The grains show textural variation from subrounded to subangular, with intergranular contacts ranging from long to point contacts. The mineral assemblage is primarily composed of quartz, feldspars, and oxides (Figure 5(E)). Carbonate cement occurs subordinately, representing approximately 1% of the total rock volume (Figure 5(F)). Aggradational Sandy Bedforms are characterized by high porosity, ranging from 15% to 33%, as observed in thin section. The predominant porosity in these sandstones corresponds to primary intergranular pores, accounting for 13% to 85% of the total (Figure 5(G) and Figure 5(H)). In turn, secondary intragranular pores (Figure 6(A)) are more developed in facies composed of massive sandstones and sandstones with incipient laminations or low-angle cross-stratification, representing 24% to 52% of the porosity in these lithofacies. Texturally, the sandstones exhibit good grain-size sorting, with the frequent occurrence of elongated to concave-convex intergranular contacts (Figure 6(B)). Cement is observed subordinately and dispersed (Figure 6(C)), insufficient to significantly obstruct the pores. Additionally, the presence of intraclasts dispersed within the sedimentary framework is recorded (Figure 6(D)).

Figure 5. Qz: Quartz, F: Feldspar, Pl: Plagioclase, Ox: Oxides, Lt: Lithics, Carb: Carbonate Cement, P1: Primary intergranular porosity, P2: Secondary intragranular porosity, P3: Moldic porosity. (A) Sample showing granulometric (grain-size) variation. (B) Sample with oxide cutans. (C) Sample with presence of lithic grains, plagioclase, oxides, and carbonate cement (cross-polarized light). (D) Sample with oxide cutans and primary intergranular porosity. (E) Sample exhibiting three types of porosity, along with aggregation of oxides forming cement. (F) Sample with presence of carbonate cement and lithic grains (cross-polarized light). (G) Sample with high primary intergranular porosity, presence of moldic porosity, long and point contacts, and aggregation of oxides. (H) Sample with high primary intergranular porosity and oxide globules.

4.1.3. Overbank Deposits

Terminal Splays composed of sandstones with incipient tangential cross-stratification exhibit the highest primary intergranular porosity in the system (96%) (Figure 6(E)). The predominantly subrounded grains have contacts ranging from long to point contacts. Cement is scarce, and the mineralogical composition is marked by extremely abundant quartz (78%), with small amounts of feldspars (8%) and oxides (9%).

Figure 6. Qz: Quartz, F: Feldspar, Pl: Plagioclase, Ox: Oxides, Lt: Lithics, Carb: Carbonate Cement, C. Silc: Silica Cement, P1: Primary intergranular porosity, P2: Secondary intragranular porosity, P3: Moldic intragranular porosity. (A) Secondary and moldic intragranular porosity, with grains showing concave-convex contacts. (B) Sample with secondary intragranular porosity and concave-convex contacts. (C) Sample with pinpoint silica cement and presence of intragranular dissolution porosity (cross-polarized light). (D) Sample with predominance of primary intergranular porosity and an intraclast. (E) Sample with oxide cutans and moldic intragranular porosity. (F) Sample with an abundance of clay matrix, with grain-size variation up to sand-sized grains. (G) Sample with high content of carbonate cement and oxides (cross-polarized light). (H) Sample with oxide cutans and moldic intragranular porosity.

Horizontally Laminated Deposits are notable for the near absence of porosity due to an abundant clay matrix (70%) (Figure 6(F)). The grains, which range from subrounded to subangular, are enveloped in a fine mass of clay minerals. The mineralogy includes quartz, micas, and oxides (15%).

In Sandstones with Deformed Stratification, the moderate porosity (17%) shows a mixed distribution: 73% primary intergranular, 19% dissolution porosity, and 8% moldic porosity. The subangular to angular grains establish mainly long contacts with some point contacts. Carbonate cement and a significant presence of oxides (41%) are notable (Figure 6(H)).

The obtained petrophysical parameters revealed multi-scale porous systems (Table 3 and Table 4), whose characterization required the integration of complementary techniques. Micro-CT analyses, while essential for evaluating macroporosity (above ~4 µm), had limitations in detecting pores below their resolution, underestimating the total porosity in samples with fine matrix or significant microporosity. Mercury intrusion porosimetry (MIP) demonstrated the presence of a considerable fraction of micropores, increasing the total porosity values compared to micro-CT data. However, even with fair to good porosity (13.84% - 18.07%), the permeability in these systems was low to moderate (0.34 - 23.87 mD), reflecting the critical influence of pore connectivity at reduced scales.

Table 3. Petrophysical parameters obtained from the binarization of micro-CT images.

Table 4. Petrophysical parameters obtained from MIP.

4.2.1. Eolian Deposits

The sandstones associated with eolian sandsheets exhibited a total porosity around 15%, with permeability on the order of 2.7 mD and an average pore diameter of approximately 0.04 µm, as determined by mercury intrusion porosimetry. Data obtained by microtomography revealed a lower effective porosity, of approximately 5%, in addition to evidencing anisotropy in permeability, with values ranging between 8.9 mD and 4.8 mD along two of the analyzed axes (Figure 7(A)).

In the sandstones from eolian dunes, the results indicated porosity below 6%, associated with an absence of detectable permeability within the resolution scale of microtomography (Figures 7(A)-(C)). It is noteworthy that the low-angle laminations and cross-stratifications, typical of the eolian depositional process, could be recognized in the three-dimensional models of pore networks and throats (Figure 7(C)), indicating the influence of the depositional structure on the organization of the pore space.

4.2.2. Fluvial Channel Deposits

The sandstones associated with lateral accretion bars reveal discrepant porosities, ranging from 21.38% to 3.45% (Figure 8(A) and Figure 9(D)). The permeability in the more porous facies proved to be anisotropic, varying from 502.51 mD to 874.35 mD along different axes. In contrast, permeability in the less porous facies was negligible (effectively zero) due to the resolution limit of the equipment used.

The aggradational sandy bedforms exhibit a marked variability in their petrophysical properties, revealing two distinct behavioral patterns. The first group shows favorable characteristics, with porosities ranging from 14.39% to 21.96%, associated with significant permeabilities reaching up to 1299.48 mD (Figures 8(B)-(D)). These values correspond to deposits with low-angle cross-stratification, planar stratification, and massive sandstones.

In contrast, the second group shows more restrictive properties, with porosities between 2.93% and 3.96% and permeabilities below the resolution limit (Figures 9(A)-(C)). However, complementary analyses by Mercury Intrusion Porosimetry (MIP) revealed a microporous structure not previously detected, particularly in massive sandstones with incipient cross-laminations, which showed 18.07% porosity. Despite this considerable value, permeability remained low (0.34 mD), suggesting that the pores, although numerous, have limited connectivity.

Figure 7. Pore network and throat models, 3D models and porosity distribution from the eolian deposits. (A) Eolian Sandsheet sample; (B) (C) (D) Eolian dunes samples.

Figure 8. Pore network and throat models, 3D models and porosity distribution from the fluvial channel deposits of the core sample 21. (A) Lateral accretion bars sample; (B) (C) (D) Aggradational sandy bedforms samples.

Figure 9. Pore network and throat models, 3D models and porosity distribution from the fluvial channel deposits of the core sample 101. (A) (B) (C) Aggradational sandy bedforms; (D) Lateral accretion bars.

4.2.3. Overbank Deposits

The sandstones associated with terminal splays revealed three distinct porosity-permeability behaviors. Deposits with incipient tangential cross-stratification present the most favorable properties, with a porosity of 25.53% and permeabilities ranging between 1787.80 mD and 2428.44 mD along the different axes of analysis (Figure 10(A)).

Figure 10. Pore network and throat models, 3D models and porosity distribution from the overbank deposits. (A) (B) (C) Terminal splays samples.

In contrast, deposits with horizontal laminations show radically different characteristics (Figure 10(B)). Micro-CT scanning registered only 0.33% porosity, while Mercury Intrusion Porosimetry (MIP) revealed higher values (16.2% porosity and 23.87 mD permeability), with an average pore throat diameter of 0.03 µm. This discrepancy indicates the presence of microporosity below the resolution threshold of the micro-CT scan.

Finally, deposits with deformed cross-stratification presented an intermediate behavior. Micro-CT analysis showed 0.905% porosity, contrasting with the 13.84% obtained by MIP, which also revealed a permeability of 3.21 mD and an average pore throat diameter of 0.02 µm (Figure 10(C)).

While eolian deposits commonly exhibit higher porosity and permeability, this generalization requires careful application to systems significantly influenced by water-related processes ([2]). The eolian deposits, comprising eolian sandsheets and eolian dunes, consistently behave as low-permeability reservoirs, a characteristic directly linked to their depositional and post-depositional evolution.

Eolian Sandsheets as described by [16] form as thin, extensive sheets in wet interdune areas. The sample T21AM17 reflects this environment, showing good total porosity (~15% by MIP) dominated by primary intergranular pores (70.9%). The good initial sorting typical of eolian transport created a well-defined pore network. However, the permeability is strikingly low (2.77 mD). The observed discrepancy between good porosity and low permeability is best explained by a combination of early diagenetic processes. While primary intergranular pores are preserved and grain dissolution created significant secondary porosity, carbonate cementation occluded the connecting pore throats. The isolated pores, initiated within feldspars, as proposed by [18] for Permian Rotliegend sandstones, likely commenced prior to deposition and progressed throughout early diagenesis, persisting until the diminishing influence of meteoric waters during burial diagenesis. Preserving porosity but severely reducing connectivity and effective permeability. Furthermore, the permeability anisotropy detected by micro-CT analysis (8.89 mD vs. 4.85 mD) can be attributed to subtle, low-angle lamination (Figure 5(A)) within this facies, which imparts a preferred orientation and heterogeneous flow pathways to the pore system, as observed in Figure 7(A).

Eolian Dunes in contrast to the eolian sandsheets, show a more homogeneous pore system as observed in the porosity distribution graphs in Figures 7(B)-(D). While primary intergranular porosity is present, there is a notable development of secondary intragranular and moldic porosity (up to 63% of the total porosity in some samples). This is a classic signature of dissolution processes, developed during diagenesis through the dissolution of detrital and diagenetic constituents ([5]). Furthermore, the presence of carbonate cement (4% - 8%) partially occludes pores. The model of [16] for wet eolian dunes suggests a high water table, which would favor such early diagenetic like dissolution. The combination of isolated secondary pores and pore-filling cement results in a system with low effective connectivity, explaining why micro-CT shows very low permeability (often below detection limit). One of the 3D models captures the inherited eolian cross-stratification, indicating that the depositional architecture dictates the fundamental skeleton of the pore network. This is consistent with the findings of [15], who notes that high-resolution scans provide a detailed representation of cross-beds.

This subsystem, represented by lateral accretion bars and aggradational sandy bedforms, contains the most prolific reservoirs in the study, aligning with their interpretation as high-energy channel deposits by [16].

Lateral Accretion Bars form by the lateral migration of meandering channels ([16]). Samples like T21AM19 exhibit high porosity (21.39%) and, crucially, very high and anisotropic permeability (e.g., 874 mD vs. 503 mD). One of the factors for high porosity can be explain by the type of rock, several authors have observed that subarkoses sometimes exhibit a greater potential for porosity preservation than quartzarenites. According to [1], the presence of feldspars inhibits the extensive development of quartz cement. This observation is supported by [21], who, in a study of Paleozoic sandstones from Wales, identified subarkoses as the composition with the most favorable porosity preservation potential. The petrography confirms a clean sandstone with well-developed primary intergranular porosity and minimal cement.

Aggradational Sandy Bedforms shows the highest variability, reflecting a range of flow regimes within the channel system. Well-sorted sandstones with low-angle or planar cross-stratification (e.g., T21AM20, T21AM23) exhibit excellent properties (porosity >17%, permeability >1000 mD). Their textural maturity, with point and long contacts, and well-preserved intergranular porosity, leads to a well-connected macroporous system.

In contrast, the massive to incipiently stratified sandstones (Sm/Sli), exemplified by sample T101AM32, present a petrophysical divergence. While micro-CT analysis reveals low porosity and negligible permeability, MIP detects a significant microporous system, with a porosity of 18.07%. This microporosity (0.06 µm average pore diameter) could be attribute to the dissolution of grains as suggested by [27]. However, it is poorly interconnected, resulting in the observed very low permeability of 0.34 mD. The high total porosity is largely reflective of the 52% moldic porosity, which, being isolated, explains the low connectivity and consequent low permeability.

The overbank deposits exhibits the most significant contrasts in reservoir quality, serving as a prime example of how subtle facies variations govern petrophysical behavior. As interpreted by [16], the characteristics of these deposits are governed by the interaction of flood event frequency and intensity, topographic depressions, and drainage configuration.

Terminal Splays—High-Quality Facies: These deposits represent by sample T21AM18 (Sti), present an exceptional porosity (25.53%) and permeability (>1700 mD). This characteristics can be explain by the near-absence of matrix and cement that allowed for the preservation of an unspoiled, highly connected intergranular pore network. This facies shows a more homogeneous porosity distribution (Figure 10(A)) and acts as high-permeability flow path. As noted by [5], the abnormal pore-fluid pressures generated during rapid sedimentation can inhibit compaction and cementation, thereby preserving primary porosity.

Terminal Splays—Low-Quality Facies: Other splay deposits tell a completely different story. The horizontally Laminated Sandstones are dominated by a clay matrix (up to 70%), which completely occludes the primary porosity. Micro-CT rightly detects near-zero porosity and permeability. The significant porosity (16.2%) and moderate permeability (23.87 mD) revealed by MIP are attributed to a network of micropores within the clay matrix itself. This facies functions as a baffle or seal.

Sandstones with Deformed Stratification shows an intermediate character. The deformation processes may have disrupted the original fabric, reducing macroporosity (0.9% in micro-CT). However, the presence of dissolution and moldic porosity (27% of the total porosity in thin section) and a lower clay content than facies Sh creates a system with fair MIP porosity (13.84%) and low but measurable permeability (3.21 mD). The low permeability can be the result of the heterogeneous system within this facies, as observed in Figure 10(C).