Petrography and Fluid Inclusions for Petroleum System Analysis of Pre-Salt Reservoirs in the Santos Basin, Eastern Brazilian Margin

Abstract

The complex interaction of hydrothermal fluids and carbonate rocks is recognized to promote significant impacts on petroleum systems, reservoir porosity, and potential. The objective of this study is to investigate the fluid phases entrapped in the mineral phases of the Barra Velha Formation (Santos Basin), including their petrographic paragenetic relationships, relative timing, temperatures of migration events, and maximum temperature reached by the sedimentary section. The petrographic descriptions (387), Rock-Eval pyrolysis (107), fluid inclusion petrography (14), and microthermometry (428) were performed on core and sidewall samples from two wells from one field of the Santos Basin. Hydrocarbon source intervals were primarily identified in lithologies with high argillaceous content. Chert samples still retain some organic remnants indicative of their original composition prior to extensive silicification. Redeposited intraclastic rocks exhibit the lowest organic content and oil potential. A hydrothermal petroleum system is identified by fluids consisting in gas condensate, light to heavy undersaturated oil, occasionally accompanied by aqueous fluids influenced by juvenile and evaporitic sources, and localized flash vaporization events. These hydrothermal fluids promoted silicification and dolomitization, intense brecciation, and lead to enhanced porosity in different compartments of the reservoir. The relative ordering of paleo-hydrothermal oils and the main oil migration and accumulation events has improved our understanding of the petroleum systems in the basin. This contribution is significant for future regional research on the evolution of fluid systems and their implications for carbonate reservoirs.

1. Introduction

Petroleum reservoirs are complex geological formations that store hydrocarbons and exhibit significant heterogeneity in their properties, such as porosity, permeability, and fluid distribution [1]. These variations occur on different scales and can impact reservoir behavior [2,3], making it challenging to predict fluid flow and optimize production. Understanding and managing this heterogeneity is essential for effective reservoir characterization [2,4], improving recovery rates, and making informed decisions in field development and enhanced oil recovery techniques.

The giant hydrocarbon accumulations found in Aptian stage lacustrine reservoirs in the eastern Brazilian basins and their counterparts in western Africa, referred to as the pre-salt deposits [5,6,7], stand among the most significant discoveries of this century. Recently, evidence on hydrothermal alteration of pre-salt deposits has been presented for the Campos Basin [8,9,10,11,12,13] and the Santos Basin [14,15,16,17,18,19].

Recent terrestrial hydrothermal systems encompass diverse environments where fluids emerge at or near the land surface at temperatures higher than the local ambient air. These systems transport heat and dissolved components through liquid or vapor phases, with fluids originating from varying depths within the Earth and exhibiting a broad range of temperatures and chemical compositions. At the surface, hydrothermal discharge can contribute to water body recharge, as observed in East African Rift lakes, and facilitate the precipitation of travertine and silica sinter deposits [20].

Hydrothermal alteration in carbonate reservoirs, such as those in the pre-salt carbonates of the Macabu Formation in Campos Basin (equivalent to Barra Velha Formation carbonates), involves complex fluid–rock interactions that precipitate new mineral assemblages and modify reservoir properties [11,12,13,15,21,22]. This process occurs when hydrothermal fluids, which can originate from magmatic, metamorphic, or meteoric sources, replace primary minerals by introducing reactants and removing dissolution products [23,24]. Hydrothermal systems require both a heat source and fluid pathways, often controlled by deep fault networks that facilitate high flow rates in transient or episodic regimes [25,26,27]. In this context, it is also common the occurrence of pore-filling solid bitumen and fluids inclusions usually recording higher temperature events than the corresponding burial thermal history of the sedimentary basins [11,13,18,25,28,29,30,31].

In the current study, we aim to explore the fluid phases entrapped as fluid inclusions (FIs), their petrographic paragenetic relationships, relative timing, temperatures of fluid migrations events, and the maximum temperature of the studied section. The obtained results provide new perspectives on the diagenetic evolution of the complex pre-salt deposits. We aim to bring direct evidence of hydrothermal oils, previously suggested by molecular indicators of biodegradation diluted in non-biodegraded oil accumulations [18].

One of the key contributions of this study is the identification and characterization of an early hydrothermal petroleum system for the pre-salt. This system was responsible for generating oil, albeit in small quantities and localized occurrences, before the main oil generation pulse associated with conventional burial processes in typical petroleum systems. Given the complex thermal history of this basin, distinguishing this early hydrothermal system, which affected the sediments prior to salt deposition is a significant finding. This has important implications for accurately modeling the petroleum system and understanding its interactions with the diagenetic processes and products that shaped this important reservoir.

Geologic Setting

The Santos Basin (Figure 1) is one of Brazil’s largest marginal basins, covering an area of approximately 350,000 km² and stretching eastward to depths exceeding 3000 m of water [32]. To the north, the basin is bordered by the Cabo Frio High and the Campos Basin, while to the south it is bounded by the Florianopolis High and the Pelotas Basin [32,33]. The basin’s geological fill is classified into three main stages or supersequences: rift, post-rift, and drift [32,34].

According to the stratigraphic chart for the Santos Basin [32,36] the rift supersequence comprises the basaltic volcanic rocks of the Early Cretaceous Camboriú Formation, which overlay a granitic–gneissic Precambrian basement; the Barremian siliciclastic alluvial conglomerates and lacustrine talc–stevensitic sandstones and mudstones of the Piçarras Formation; and the late Barremian to lower Aptian bioclastic calcirudites and calcarenites and organic shales of the Itapema Formation. The upper boundary of this supersequence is marked by a widespread erosive surface known as the Pre-Alagoas unconformity [32,36].

The post-rift stage corresponds to the thermal subsidence stage, resulting in the formation of a large, relatively flat, and shallow basin [37,38]. The post-rift supersequence, which is the focus of this study, comprises carbonate–Mg–clay rocks of the lower Aptian Barra Velha Formation (BVF) and the upper Aptian Ariri Formation evaporites. Initially, the deposition of the BVF was interpreted as microbial [32,39], but it was later reinterpreted as abiotic precipitation in evaporative lacustrine conditions [40]. The BVF deposits are a combination of in situ fascicular and spherulitic calcite aggregates and an Mg–clay matrix, along with rocks composed of redeposited fragments of these carbonate aggregates. The uppermost part of the Aptian interval is represented by the Ariri Formation, characterized by extensive layered sequences of anhydrite, halite, and other chlorides [32]. These evaporites mark the large-scale entrance of seawater into the system, representing a massive marine transgression [32,41].

The Santos Basin has a complex evolutionary history, marked by multiple magmatic events that influenced its thermal regime [42,43,44]. These events not only led to volcanic deposition and magmatic intrusions but also drove diagenetic alteration of the pre-salt carbonate sequence from the time of its deposition onward. In a conventional basin subjected to a typical burial regime, thermal history follows a relatively linear trajectory. However, in basins with a complex thermal evolution, early magmatic events can significantly elevate the temperature of the sedimentary section well above the expected burial gradient. The temperature anomalies associated with these events could have played a crucial role in the thermal maturation of surrounding rocks, potentially triggering early oil generation in localized areas during periods of hydrothermal activity in the depositional environment. Alternatively, oil generation could have taken place during later hydrothermal circulation. Following these thermal anomalies, the system would gradually return to temperature conditions controlled by burial.

2. Materials and Methods

The studied samples consist of core and sidewall plugs from two wells drilled in one field in the central part of the Santos Basin (Figure 1). The exact location and identification are proprietary information and cannot be disclosed. The presented sample depths are relative to the base of the Ariri Formation for each well. The wells are identified as wells MS-06 and MS-09 from field B. MS-06 has two cores, while MS-09 only has sidewall samples available (

Table 1. Dataset of samples selected from wells MS-06 and MS-09 for this study.

Well	MS-06	MS-09	Total
Core samples	2	-	2
Sidewall samples	-	120	120
Samples interval length (m)	48.08	341.22	389.3
Thin sections	267	120	387
Rock-Eval pyrolysis samples	85	22	107
Doubly polished thick sections	5	8	13
Microthermometry measurements	149	279	428

). Both wells are located at structural highs. MS-06 is located on the west side of the field and MS-09 on the east side (Figure 1 and Figure 2).

2.1. Petrography

For petrographic analysis, the paragenetic evolution of diagenetic mineral phases were examined in 387 thin sections (Table 1) at IPR-PUCRS by counting 300 points [45], using Petroledge^® Workstation software version 3.11.8.1111 [46]. Samples were stained with an alizarine red S and potassium ferricyanide solution to differentiate the carbonate minerals [47]. Petrographic analysis was conducted with Zeiss Axio ImagerA1 (Carl Zeiss AG, Gottingen, Germany) and Leica DM750P (Leica Microsystems GmbH, Heerbrugg, Switzerland) polarized transmitted light microscopes.

The samples were classified according to De Ros, L.F., and Oliveira’s [48] system for the Aptian pre-salt deposits. For paragenetic sequence boundaries, we considered the proposition of Morad et al. [49,50] for eodiagenesis and mesodiagenesis. Eodiagenesis includes processes developed at depths < 2 km under the influence of surface or modified surface waters such as marine, mixed marine–meteoric, or meteoric waters. In contrast, mesodiagenesis includes processes encountered at depths > 2 km and reactions involving chemically evolved formation waters.

2.2. Rock-Eval

Rock-Eval pyrolysis was performed to evaluate the organic and inorganic matter of the deposits [51,52]. The samples’ free hydrocarbons were extracted in a Soxhlet apparatus to remove interference from the present reservoir oil, thus allowing us to study only the original deposited organic matter [53]. The results were used to measure the hydrocarbon potential, the maturity stage, the total organic carbon, and the mineral carbon content of the rock. The maturity stage is relevant to constrain the geothermal history since the organic matter thermal evolution functions as a maximum palaeothermometer [54]. Bulk rock samples (Table 1) were analyzed with a Rock-Eval 7^TM system at IPR-PUCRS. Calibration was conducted using the geological standard IFPEN-160000 (marl; Albian/Aptian) [55].

2.3. Fluid Inclusions

The fluid inclusion petrography of 14 doubly polished thick sections and the microthermometry of 428 fluid inclusions were conducted by H-Expertise Services S.A.S (H-ES, Vandœuvre-lès-Nancy, France) and Fluid Inclusion Technologies Schlumberger (FIT-SLB, Houston, TX, USA). At H-ES, the fluid inclusions (FI) phase transitions were measured at temperatures between −170 and 400 °C using a LINKAM MDS 600 heating–freezing stage (Linkam Scientific Instruments, Surrey, UK) equipped with a Sony Exwave HAD3 (Sony Group Corporation, Tokyo, Japan) color video camera mounted on an Olympus BX 51 microscope (Olympus Corporation, Tokyo, Japan) at GeoRessources, Vandœuvre-lès-Nancy, France. At FIT-SLB, the FI phase transitions were determined using a Fluid Inc. modified USGS heating–freezing stage (Fluid Inc., Oakland, CA, USA). The microthermometric stage in both laboratories was thermally calibrated using synthetic fluid inclusion standards, considering the melting point of CO₂ and ice in pure water (−56.6 °C and 0.0 °C, respectively) and the critical point of pure water (374.1 °C). The accuracy of measurements is estimated at ±0.2 °C for Tm(ice) and ±0.5 °C for Th. All measurements have been reproduced at least 2 times to ensure reproducibility. Fluid inclusion assemblages (FIAs) were defined as the most finely discriminated and petrographically associated group of coeval inclusions. FIAs were described using the approach of [56], being classified according to their genetic type (primary, secondary, and pseudosecondary) and tied in to the paragenetic sequence.

Epifluorescence analysis used a Zeiss HXP 120 V metal halide fluorescence light source (Carl Zeiss AG, Jena, Germany) connected to an Axio Imager A2 microscope (Carl Zeiss AG, Jena, Germany), using a Zeiss filter set 02 (excitation G 365 nm, beamsplitter FT 395 nm, emission LP 420 nm). Quantitative estimation of oil fluid inclusions’ API gravity could not be appraised due the necessity of endmember oil samples representative of the studied field to calibrate the spectrometry device [57]. The qualitative approach of [58] was employed to estimate API density based on fluorescence colors.

The petrographic analysis of paragenetic relationships of the mineral phases hosting the fluid inclusion was subsequently carried out at IPR-PUCRS. This was conducted utilizing the microscope systems mentioned earlier and examining duplicated thin sections obtained from the same samples cut for the microthermometry study.

This study presents the Th of individual FIAs as box and whisker plots. This data representation method is recommended for studies aiming to determine the physical and chemical conditions associated with the formation of a mineral deposits or hydrocarbon reservoirs [59]. Additionally, it aids in developing a comprehensive understanding of the fluid and thermal history of the deposit or reservoir, both temporally and spatially. The homogenization temperature (Th) indicates when a biphasic FI with a liquid phase volume greater than the vapor phase at room temperature homogenizes to a single liquid phase after heating. The vapor homogenization temperature (Th_vapor) highlights cases when a biphasic FI with a vapor phase volume greater than the liquid phase at room temperature homogenizes as a single vapor phase after heating.

For relative ordering, the FIAs for each sample were considered within the paragenetic sequences and the relationship between each host mineral. Whenever possible, petrographic evidence of the relative sequence or diagenetic stage was considered the priority for ordering each FIA. The petrographic criteria we employed consist of the compaction stage, relations with fractures, pore filling, and the engulfing of later minerals. In cases of unknown paragenetic sequence, the homogenization temperatures (Ths) were considered in increasing order, since in the absence of anomalous thermal event evidence, we assumed a simple burial heating history. The final ordering conditions were the similarities in fluid salinity and Th distribution.

For cases of FIAs with oil and aqueous inclusion coexistence, a rough estimate of the pressure was made using the PIT (Petroleum Inclusion Thermodynamic) model [60], considering the original mineral precipitation conditions, assuming the visual estimate of 2D vapor volume percentage and the fluorescence-estimated API degree.

3. Results

3.1. Petrography

The petrographic characterization of 387 representative thin sections (for detailed descriptions, see [61]), revealed that the in situ lithologies are represented by shrubstones, spherulstones, muddy shrubstones, muddy spherulstones, mudstones, and cherts and that the redeposited rocks correspond to intraclastic calcarenites and rudaceous calcarenites [48].

3.2. Rock-Eval

Rock-Eval pyrolysis analyses were performed in 85 samples (58.21–106.29 m, from the base of the evaporites) from well MS-06 and 22 samples (0.29–341.51 m) from well MS-09 (Figure 3, Table S1). In

Table 2. Rock-Eval pyrolysis results grouped by lithological classes, including S2 (oil potential), Tmax, HI (Hydrogen index), OI (oxygen index), and TOC (total organic carbon). Minimum–maximum values; n—number of samples; raw data in Table S1.

Lithological Classes	n	S2 (mg HC/g Rock)	Tmax (°C)	HI (mg HC/g TOC)	OI (mg CO2/g TOC)	TOC (wt%)
Mudstone	5	2.57–54.3	441–447	454–849	19–82	0.57–6.39
Muddy spherulstone	33	0.26–7.82	431–458	123–743	9–316	0.15–1.12
Muddy shrubstone	17	0.26–2.9	437–462	161–427	41–249	0.12–0.74
Shrub-spherulstone	7	0.23–3.03	431–452	60–401	33–226	0.17–0.82
Shrubstone	6	0.13–0.88	441–458	142–337	48–126	0.07–0.26
Calcarenite	26	0.11–3.46	432–448	93–567	23–397	0.11–0.61
Rudaceous calcarenite	7	0.37–1.78	438–443	199–357	53–180	0.17–0.5
Chert	6	0.33–3.88	441–446	221–494	26–201	0.15–0.99

is presented the Rock-Eval pyrolysis results grouped by lithological classes.

Well MS-06 samples presented an hydrocarbon generation potential (S2) of 0.11–54.3 mg HC/g Rock (poor to very good); thermal maturity parameter (Tmax) of 431–462 °C (oil to condensate zone); hydrogen index (HI) of 60–849 mg HC/g TOC; oxygen index (OI) of 9–397 mg CO₂/g TOC; dominant kerogen type I (oil prone) and subordinated II-III (oil and gas prone); total organic carbon (TOC) of 0.11–6.39 wt% (poor to very good); and mineral carbon (MINC) of 5.44–12.06 wt% (predominant presence of carbonates).

Well MS-09 samples presented an S2 of 0.13–3.88 mg HC/g Rock (poor to fair); Tmax of 431–446 °C (oil zone); HI of 194–512 mg HC/g TOC; OI of 10–201 mg CO₂/g TOC; dominant kerogen type II (oil and gas prone) and subordinated III (gas prone); TOC of 0.07–0.99 wt% (poor to fair); and MINC of 0.17–12.31 wt% (predominant absence of carbonates).

3.3. Fluid Inclusions

3.3.1. Well MS-06

Fluid inclusions from 31 FIAs in prismatic quartz at well MS-06 show relatively large homogenization temperature (Th) variation within individual FIAs (Figure 4, Figure 5 and Figure 6; Table S2 and Figure S1). Each assemblage contained between 15 and 1 fluid inclusions. The FIAs of primary light to heavy oil presented a Th of 35–127 °C. Primary aqueous FIAs had a Th range of 45–148 °C and salinity of 13–22 wt% (evaporitic-influenced). In two of those assemblages, biphasic inclusions (L+V) coexist with monophasic (L) inclusions, both exhibiting a salinity of 20–21 wt%. The FIAs of primary gas condensate showed a Th_vapor of 55–135 °C, with one assemblage containing condensate that coexisted with aqueous monophasic and biphasic inclusions, which had a Th range of 50–77 °C and salinity of 21 wt% (evaporitic-influenced). Additionally, the FIAs of primary light to medium oil displayed a Th range of 41–91 °C, with one assemblage coexisting with aqueous inclusions that had a Th of 78–120 °C and salinity of 17–22 wt% (evaporitic-influenced). There was also a heterogeneous FIA of pseudosecondary light oil and gas condensate with a Th of 70–82 °C and Th_vapor of 136 °C, coexisting with aqueous monophasic and biphasic inclusions that had a Th of 46–110 °C and salinity of 15–22 wt% (evaporitic-influenced). Finally, the FIAs of secondary light to medium oil exhibited a Th range of 60–97 °C, with one assemblage coexisting with aqueous inclusions that had a Th of 48–50 °C and salinity of 16 wt% (evaporitic-influenced). The FIA of secondary gas condensate showed a Th_vapor range of 103–165 °C.

Fluid inclusions from six FIAs in blocky dolomite at well MS-06 show relatively narrow Th variation within individual FIAs (Figure 5 and Figure 6; Table S2 and Figure S1). Each assemblage contained between three and one fluid inclusion. The FIAs of secondary aqueous inclusions showed Th values ranging from 48 to 61 °C, with salinity between 17 and 19 wt%. In one of these assemblages, biphasic inclusions coexisted with monophasic inclusions, presenting a salinity of 20 wt%. Additionally, the FIAs of secondary medium to light oil displayed Th values ranging from 70 to 112 °C.

For detailed descriptive results of each fluid inclusion assemblages of well MS-06, please refer to Supplementary Materials File S1.

3.3.2. Well MS-09

Fluid inclusions from four FIAs in fibrous-radiated chalcedony at well MS-09 show relatively narrow Th variation within individual FIAs (Figure 7, Figure 8 and Figure 9; Table S2 and Figure S2). Each assemblage contained between six and two fluid inclusions. One FIA consisted of primary aqueous inclusions with the Th ranging between 70 and 76 °C and a salinity of 22–23 wt% (evaporitic-influenced). Another FIA comprised primary light to medium oil with Th values between 82 and 84 °C. Additionally, the FIAs of secondary light to medium oil exhibited Th values ranging between 54 and 116 °C. In one of these assemblages, the light oil coexisted with aqueous inclusions that had Th values of 85–91 °C and salinities of 18–19 wt% (evaporitic-influenced).

Fluid inclusions from 11 FIAs in flamboyant quartz at well MS-09 show relatively narrow Th variation within individual FIAs (Figure 7, Figure 8 and Figure 9; Table S2 and Figure S2). Each assemblage contained between seven and one fluid inclusions. The FIAs of primary light oil displayed Th values ranging from 45 to 98 °C. The FIA of primary gas condensate had a Th_vapor of 46–48 °C. Additionally, the FIAs consisting of primary aqueous inclusions showed Th values between 60 and 107 °C, with salinity levels in individual assemblages ranging from 22–23 wt% (evaporitic-influenced) to 3–4 wt% (brackish water–seawater equivalent) to 0–2 wt% (fresh water).

Fluid inclusions from 43 FIAs in prismatic quartz at well MS-09 show relatively large Th variation within individual FIAs (Figure 8 and Figure 9; Table S2, Figures S2–S4). Each assemblage contained between 11 and 1 fluid inclusions. The primary aqueous FIAs included Th values ranging from 45 to 88 °C and a salinity of 13–23 wt% (evaporitic-influenced); one FIA containing large vapor bubbles with a Th_vapor of 59–69 °C, for which salinity could not be measured; and another FIA with Th values between 74 and 87 °C and a salinity of 3–4 wt% (brackish water–seawater equivalent). The FIAs of primary light oil exhibited Th values ranging from 52–56 °C, while those of heavy oil ranged from 64 to 95 °C; medium oil showed Th values of 120–122 °C, and the gas condensate had Th_vapor values between 94 and 118 °C. Additionally, the FIAs of primary light to medium oil displayed Th values of 76–88 °C, coexisting with aqueous inclusions that had Th values of 64–86 °C and a salinity of 11–17 wt% (evaporitic-influenced). A heterogeneous FIA consisting of primary light oil and gas condensate had Th values ranging from 109 to 114 °C, with Th_vapor values between 150 and 153 °C. The FIAs of pseudosecondary light to medium oil exhibited Th values from 54 to 94 °C, with one assemblage coexisting with aqueous inclusions that had Th values of 70–103 °C and salinities of 16–23 wt% (evaporitic-influenced). The FIAs of pseudosecondary aqueous inclusions ranged from Th values of 88–97 °C with a salinity of 16–17 wt% (evaporitic-influenced) to 90–107 °C with a salinity of 6–7 wt% (seawater equivalent). The FIAs of secondary gas condensate had Th_vapor values between 80 and 100 °C coexisting with aqueous inclusions that had Th values of 90–102 °C and salinities of 4–6 wt% (seawater equivalent). The FIAs of secondary light–medium oil exhibited Th values ranging from 84 to 106 °C, with one assemblage coexisting with aqueous inclusions that had Th values of 100–115 °C and salinities of 3–5 wt% (brackish water–seawater equivalent). Monophasic aqueous inclusions were observed in the early quartz phases, but attempts to measure Th and salinity were unsuccessful.

Fluid inclusions from one FIA in calcite spherulite at well MS-09 showed relatively narrow Th variation within individual FIAs (Figure 8 and Figure 9; Table S2, Figures S2 and S5). The assemblage contained five fluid inclusions. This FIA consists of primary light oil with Th values ranging from 66–71 °C, coexisting with one aqueous FI that has a Th of 98 °C. The salinity of these inclusions could not be measured due to mineral opacity.

Fluid inclusions from four FIAs in spherulite-replacive blocky dolomite at well MS-09 showed relatively narrow Th variation within individual FIAs (Figure 8 and Figure 9; Table S2, Figures S2 and S5). Each assemblage contained between four and two fluid inclusions. The FIAs of primary aqueous inclusions showed Th values ranging from 70 to 76 °C, with a salinity of 22–23 wt% (evaporitic-influenced), and Th values of 82–86 °C with a salinity of 8–9 wt% (equivalents to evolved seawater). Additionally, there were two FIAs of primary light oil with Th values ranging from 88 to 105 °C.

Fluid inclusions from ten FIAs in blocky dolomite at well MS-09 showed relatively large Th variation within individual FIAs (Figure 8 and Figure 9; Table S2, Figures S2 and S6). Each assemblage contained between six and one fluid inclusions. The FIAs of primary light to heavy oil exhibited Th values ranging from 55 to 118 °C. One assemblage of gas condensate had a Th_vapor of 74–132 °C and two assemblages of light to medium oil with a Th of 80–88 °C, both of which coexisted with aqueous inclusions that had Th values between 74 and 130 °C and salinities of 20–22 wt% (evaporitic-influenced). Furthermore, one FIA of primary aqueous inclusions showed Th values of 66–67 °C and a salinity of 1–2 wt% (fresh water).

For detailed descriptive results of each fluid inclusion assemblage of well MS-09, please refer to Supplementary Materials File S2.

4. Discussion

4.1. Organic Potential and Content

The Rock-Eval data of well MS-06 have higher hydrocarbon potential and total organic carbon than well MS-09. The quantity of carbonates in each well varies from predominant to intermediary at well MS-06 and from predominant to absent in well MS-09, as shown by the measured mineral carbon content, thus marking the silicification intervals when the mineral carbon is low.

When the samples are grouped by lithological classes [48], it is clear that the hydrocarbon-generating intervals are associated with more argillaceous in situ lithologies (mudstones and muddy spherulstones). The samples contain type I and II organic matter, in the oil-generating window. Other in situ lithologies (muddy shrubstone, shrubstone, and shrub–spherulstone) present lower oil potential and organic content, due to differences in clay mineral and organic matter content and preservation, probably related to oxidation conditions, the energy of the depositional sub-environment, and the organic matter accumulation rate. Chert samples still preserve some low oil generation potential and organic content reminiscent of their original composition prior to intense silicification, suggesting that this event occurred before maximum burial of the section. Calcarenites and rudaceous calcarenites presented the lowest oil potential and organic content of the studied wells. This is related to the origin of intraclastic calcarenites and calcirudites through reworking of the in situ deposits by hydrodynamic processes inside the lake [48,64]. Redeposition enhances the diffusion of oxygen, promoting organic matter oxidation, as well as the growth of aerobic microorganisms that support its degradation [65].

Several authors pointed out that the hydrocarbon generation potential of the Barra Velha Formation (BVF) is minor [66,67,68], and our studied wells (MS-06 and MS-09) are no exception. We found only five cases of high organic content and oil generation potential in well MS-06, in which most of the samples (80) were of fair to poor oil potential and organic content. The BVF kerogen in wells MS-06 and MS-09 corresponds to types I-III and II-III, respectively, with an average HI of 322 mg HC/g TOC and average TOC of 0.43 wt%. As discussed by Lenz et al. [67], for well MS-06, it was possible to identify highly fluorescent bacterial amorphous organic matter associated with type I-II kerogen and low-fluorescence amorphous organic matter derived from zooclasts associated with type II-III kerogen.

4.2. Fluid Event Implications

Fluid inclusions are volumes in which pressure and temperature are interdependent variables. Heating biphasic oil and aqueous fluid inclusions to homogenization temperature (Th) allows for the establishment of the minimum trapping temperature of the inclusion. Nevertheless, considering the shallow slope of the isochores of the petroleum fluid systems, oil FI Th can be way below the true temperature of trapping [69,70,71], up to tens of degrees, and this varies according to the composition of the petroleum fluid and the pressure. For aqueous FIAs of low salinity formed at <150 °C, the inclination of the isochores is practically vertical, meaning there is no significant difference between the Th and the real entrapment temperature of the inclusions [69]. Thus, for Figure 6 and Figure 9, the minimal trapping temperature lines only considers the Th of aqueous inclusions.

The box plots in Figure 6 and Figure 9 present some primary and secondary FIAs with wide homogenization temperature value distributions, suggesting the occurrence of post-entrapment processes [59,72]. In Figure 5 and Figure 8, the homogenization temperature and salinity of primary, pseudosecondary, and secondary aqueous fluid inclusions are compared for each well. Based on the distribution pattern, it is possible to indicate stretching processes [58,69] in some FIAs, which re-equilibrated the original entrapment temperature to a gradation of higher values, with maximum values marking peak temperature. For the FIAs with clear occurrence of stretching, only the minimal Th mode of aqueous FIAs was considered as representative of the original entrapment condition. The maximum Th mode of some of the FIAs that presented stretching processes (130–160 °C for well MS-06; and 150–153 °C for well MS-09) are like the estimated maximum Tpeak for burial heating (131–163 °C).

For well MS-06, the aqueous FI salinities indicate the predominance of basinal brines, with fluids at 15–20 wt% (NaCl equivalent). The aqueous inclusions in well MS-09 presented at least two distinct fluids: one of high salinity (20–25 wt%), originally entrapped at a low temperature (45–60 °C), and a low-salinity fluid that varies both above and below seawater salinity (0–8 wt%), originally entrapped at a slightly higher temperature (70–80 °C). Between these two endmembers, there are a few points that can represent two possibilities: that there are four different fluids (20–25; 15–20; 5–10; and 0–5 wt%) with different origins or that there are two different fluids (20–25 and 0–10 wt%), with the composition between these two endmembers corresponding to mixed fluids. High-salinity fluids can be associated with evaporitic-influenced conditions possibly related to connate fluids of pre-salt lacustrine deposits. As low salinity can be found in meteoric or juvenile aqueous fluids [73], the latter possibility is favored by FIA occurrences that had been predominantly entrapped in late, mesodiagenetic phases, which would be already isolated from the depositional fluids. Furthermore, there is no petrographic evidence of surface exposure in the studied samples, which could be related to meteoric water infiltration.

The history of low-temperature fluids preserved in the Santos Basin is absent in the Campos Basin, with only records of high-temperature events between 90 and 150 °C associated with basinal brines [11,13]. The mixing between brine connate and brackish juvenile fluids with contrasting original entrapment temperatures (respectively 45 and 70 °C) is a novelty for the Santos Basin. The only alike occurrence was reported in the carbonate reservoirs of the West African pre-salt Kwanza Basin [74], where mixing between basinal brines with contrasting temperatures, one colder (~55 °C) and the other hotter (~140 °C).

The apparent temperatures of carbonate clumped isotopes for carbonates from the Barra Velha Formation in the Santos Basin range from 36 to 91 °C [75] and from 46 to 73 °C [19]. Temperatures below 40 °C were considered to correspond to depositional condition, while temperatures above 40 °C were related to diagenetic alteration [75]. We presented data of primary heterogeneous aqueous fluid inclusion (L and L+V) entrapped at 45–50 °C in prismatic quartz and blocky dolomite in well MS-06, indicating a low-temperature evaporative context during eodiagenesis [69], which corroborates the precipitation of some of these mineral phases in the depositional context. Some prismatic quartz in MS-09 presented FI homogeneous entrapments with similar temperatures (45 °C), suggesting early mesodiagenetic context.

Lawson et al. [75] considered some potential explanations for the elevated temperatures. A proposal by Farias et al. [19] was that these high temperatures are indicative of thermal springs discharged into pre-salt lakes in the vicinity of carbonate precipitation. Lawson et al. [75] observed no covariance between the temperatures of the clumped isotopes and δ¹⁸O to support this hypothesis. Our data indicate that this is unlikely, due to the absence of records of high-temperature events in original entrapment conditions during eodiagenesis. Another proposed explanation for the elevated temperatures of the clumped isotopes was the result of burial diagenesis during recrystallization of the original carbonate phases or the precipitation of new diagenetic products [75]. As previously presented in our studied section, the maximum Th mode of re-equilibrated fluid inclusion converged with the peak burial heating temperatures; therefore, we can infer that gradual burial heating was responsible for the elevated temperatures of the clumped isotopes.

Occurrences of heterogeneous entrapment of light oil, gas condensate, and aqueous monophasic (L) and biphasic (L+V→L) systems were observed in one FIA of well MS-06, suggesting that hot fluid systems undergo a depressurization event in eodiagenetic contexts (Figure S1; 93.76 m FIA 3). One FIA of well MS-09 also presented heterogeneous entrapment of light oil and gas condensate, indicating depressurization in eodiagenesis (Figure S2; 269.94 m FIA 6). In both cases, their homogenization temperatures were incongruent to define the boiling state, possibly due to re-equilibration processes.

The occurrence of aqueous vapor FI could be associated with the vadose zone, boiling, or flash vaporization [69,76,77]. The pattern of extensive and pervasive silica precipitation in the studied wells and the lack of oxidation and other exposure features suggest a phreatic, saturated condition, ruling out the possibility of a vadose zone. The characteristic of boiling fluids is the coexistence of liquid-rich and vapor-rich aqueous inclusions (heterogenous entrapment) that homogenize at the same temperature [56,78]. In the two occurrences in this study’s wells, the FIAs consist of only vapor aqueous inclusions hosted by prismatic quartz, associated with condensate gas in the case of well MS-06. The remaining possibility is flash vaporization, which occurs when there is a sudden drop in pressure below the saturation point. This rapid pressure decrease causes the liquid to undergo a swift phase transition. Hydrothermal fluids may be present at depths in numerous, poorly connected, and/or highly cemented fractures, resulting in fluid pressures that exceed hydrostatic levels and may approach lithostatic conditions [79]. If the rock fractures due to seismic activity or increased fluid pressure, a pressure decrease to less than the hydrostatic pressure may occur instantaneously, resulting in the conversion of the original liquid into a low-density vapor phase.

Boiling and flashing events are a driving mechanism in the precipitation of epigenetic minerals [30]. Fluid inclusions in Mississippi Valley-type deposits indicate that CO₂ effervescence under rapid pressure drops but in near-isothermal conditions would result in the precipitation of saddle dolomite as the first and predominant mineral phase, while non-isothermal boiling could trigger the precipitation of quartz in a fault-controlled hydrothermal chert reservoir system [30]. The vapor inclusions in well MS-06 samples occur at 76.62 m beneath salt with 75–90 °C. In well MS-09 samples, vapor inclusions occur at 345.74 m with 59–69 °C. This indicates that localized flash vaporization occurred first at the base of MS-09’s studied section and later at the top of MS-06’s studied section. This may suggest that the fluid migration front reached MS-09 before MS-06, which is supported by the salinity distribution of preserved low-salinity fluid found in MS-09.

Five FIAs that presented oil and aqueous inclusion coexistence allowed us to estimate a rough entrapment pressure. For well MS-06, at 76.62 m, the prismatic quartz entrapment pressure of FIAs 1 and 3 was estimated as 22 MPa at 86–88 °C, and the 263.13 m FIA 2 blocky dolomite was 10 MPa at 97 °C. For well MS-09, the prismatic quartz entrapment pressure of the 83.10 m FIA 6 and 269.94 m FIA 1 was estimated as 13 MPa at 81–82 °C. Despite the approximated values, this indicates that homogeneous FIAs were trapped in mesodiagenetic conditions, isolated from the atmosphere, and subjected to significant pressure.

4.3. Fluid Percolation History and Organic Maturation

The available data allowed us to characterize a complex paragenetic sequence and fluid percolation history. For well MS-06 (Figure 10), eodiagenetic blocky dolomite precipitation is associated with evaporitic-influenced aqueous fluids, probably connate due to high salinity, and the absence of oil fluid inclusions suggests that the early dolomite precipitated before the onset of an atypical hydrothermal petroleum system. In the mesodiagenetic stage, it is possible to infer the action of a hydrothermal petroleum system, generating fluids ranging from light to medium oil.

In prismatic quartz, the presence of one FIA (93.76 m #3) of gas condensate coexisting with light oil and with evaporitic-influenced monophasic and biphasic aqueous fluids indicates entrapment during eodiagenesis. During mesodiagenesis, the hydrothermal petroleum events persist in quartz, with fluids composition varying from gas condensate and light-to-heavy oil to some undersaturated light-to-medium oil, sometimes associated with aqueous fluids related to evaporitic-influenced systems. One occurrence (76.62 m FIA 2) of aqueous vapor phases entrapped at 75–77 °C may be associated with flash vaporization.

For well MS-09 (Figure 11), one FIA (269.94 m #6) in prismatic quartz presented evidence of an early hydrothermal petroleum system, marked by the occurrence of gas condensate coexisting with light oil (heterogeneous trapping) during eodiagenesis. During mesodiagenesis, the chalcedony, flamboyant, and prismatic quartz phases were influenced by a hydrothermal petroleum system, with fluids composed of gas condensate, light to heavy oil, and some undersaturated medium oil, in some occurrences associated with aqueous fluids with low to high salinities related to juvenile and evaporitic-influenced fluids. One occurrence (345.74 m FIA 7) of an aqueous vapor phase entrapped at 59 °C in prismatic quartz may be associated with flash vaporization.

The mesodiagenetic replacive and blocky dolomites record the influence of a hydrothermal petroleum system front, composed of fluids ranging from gas condensate to light-to-heavy oil, with some undersaturated medium oil. This system is associated with early mesodiagenetic low-salinity aqueous fluid inclusions related to the percolation of juvenile fluids and later intermediate to high salinities related to evolved and evaporitic-influenced aqueous fluids.

The spherulitic calcite presented only one FIA with light to medium oil associated to associated with aqueous fluid, originally classified as primary. Due absence of vapor trapping, we would expect significant pressure suggesting mesodiagenetic trapping. However, this would be incongruent with the expected paragenetic sequence. Thus, we suggest that the FIA is probably from a secondary event. The mineral opacity compromises the identification and precludes salinity measurement.

The thermal maturation parameter (Tmax from Rock-Eval pyrolysis) from both wells indicated a predominant oil zone, with only 6% of samples (n = 5 from well MS-06) reaching the beginning of the condensate zone. Tmax was used to estimate the re-equilibrated fluid inclusion homogenization temperatures upper mode for burial heating and hydrothermal metamorphism contexts [80,81,82,83]. For wells MS-06 and MS-09, the burial heating Tpeak equivalent from Tmax was 148 ± 9 °C {131–163 °C} ($\stackrel{-}{\mathrm{x}}$ ± 1 σ {min–max}), and the calculated geothermal gradient was ≈25 ± 2 °C/km ($\stackrel{-}{\mathrm{x}}$ ± 1 σ). For a hydrothermal metamorphism scenario, the Tpeak equivalent from Tmax was 172 ± 15 °C {145–195 °C}. The present geothermal gradient calculated for the pre-salt of the Santos Basin [84] ranges from 23 to 27 °C/km at similar depths, based on bottom hole temperatures of 140–160 °C. Thus, for the studied wells, the maximum temperature that the section was subjected to is related to burial heating, and eventual hydrothermal processes that occurred before the burial must have been restrained to lower or equal temperatures than present temperatures.

The subsequent burial heating of the sedimentary section led to the re-equilibration of some FIAs by stretching. The degree of the thermal evolution of the entrapped hydrocarbons is incompatible with the minimum entrapment temperatures of fluid inclusions, suggesting that the thermal event that generated these oils occurred in another location of the basin. The current oil accumulation in the reservoir consists of light to medium API gravity. A comparison with the oil types entrapped in the FI makes clear that this petroleum system, showing a wide range of API gravity, encompassing gas condensate and heavy oil, is significantly distinct. This variation may reflect differences in source rocks and thermal maturity, which may be explained due to the existence of an atypical hydrothermal petroleum system in the area [18,30,85]. Other studies on FIs in cemented sandstones and carbonates from Kwanza Basin pre-salt have also shown that entrapped oils in some fields consist of complex mixtures between two distinct source rocks, involve different maturities of oil from the same source, or consist of blends of degraded and fresh oil [86].

An atypical petroleum system was previously described in the Santos Basin due to the presence of a wide range of FI Th (50–160 °C) hosted in early diagenetic cements, coupled with molecular indicators of biodegradation in pre-salt oils, which are non-biodegraded oil accumulations [18]. This incompatible combination suggests that the generation and biodegradation events of hydrothermal oils under shallow burial conditions were widespread during the early diagenetic history of the pre-salt reservoirs [18]. The source rocks for this system may correspond to in situ generation within the Barra Velha Formation or from the underlying Itapema Formation, which contains the main source rocks for the present oil accumulation in the Santos Basin [66,87,88].

Syn to post rift (~112–~123 Ma) and drift (~38–~110 Ma) magmatic cycles at the Santos Basin [42,43,44] could had provided the heat to the action of hydrothermal petroleum system in the vicinity of the analyzed wells. The hypothesis of in situ generation within the BVF would depend on an anomalous localized temperature event, of which there is no evidence in the studied wells, thus favoring the hypothesis of hot fluid circulation through hydrothermal activity in the area. The observed salinity variation may indicate events of mixing between saline connate fluids and pulses of brackish juvenile fluids. Such juvenile fluids could percolate laterally (Figure 2) in the reservoir section and/or vertically through the reactivation of syn-rift fault systems [42].

Although we have examined the original precipitation and maximum temperatures in the study section, we cannot dismiss the possibility of overlapping sequences between heating events, which may affect intermediate records between fluid–rock interaction events. Previous studies have reported a wide age range and geochronological disturbance of primary carbonate age data of Barra Velha Formation in the Santos Basin [44]. One proposed explanation was the relationship between U-Pb isotopic ages and hydrothermal alteration, suggesting that hydrothermal fluids may trigger U gain or Pb loss during diagenesis, resulting in a younger age. Brito et al. [44] also reported contrasting intensity and pervasiveness of hydrothermal activity between wells, affecting the age of eodiagenetic precipitation in some cases, thus suggesting that overlapping thermal events could have occurred in the BVF.

The studied wells show evidence that calcite spherulites and shrubs, as well as dolomite and silica, were favored in structurally high areas [61], as also shown by Carvalho et al. [21,89] for another field in the area. However, in our study area, we do not see any petrological evidence of either large lake-level fluctuations or of surface exposure (e.g., calcretes with low δ¹⁸O, oxidation, stevensite alteration to sepiolite; [61]) to support the influence of meteoric fluids. As shown by petrographic data [11,12,13,15] and simulations [21], the circulation of hydrothermal fluids can intensely alter the rocks in the vicinity of the faults, generating high rates of dolomitization and silicification. Hydrothermal fluids may also affect areas away from the fault by increasing the porosity of initially porous–permeable layers. MS-06’s core presented some localized minor silicification and dolomitization and a porosity enhanced by the dissolution of the Mg–clay matrix and calcite aggregates [61]. Conversely, well MS-09 presented intense silicification and brecciation at the base of the analyzed section, with enhanced matrix and aggregate dissolution, vugular and channel porosity at the upper interval of the core [61].

The intense brecciation, preservation of low-salinity fluid inclusions, and localized flash vaporization events at the base of MS-09’s studied section indicate that a fluid migration front reached this well before reaching MS-06. This fluid front was responsible for the silicification, dolomitization, and porosity enhancement in different compartments of the reservoir prior to the main pulse of oil charge responsible for the present oil accumulation. We propose that the main oil accumulation began to occur when the reservoir temperatures were at least 100–120 °C, due to the late occurrences of entrapped medium oil during mesodiagenesis, potentially related to the main charge and the end of mineral precipitation due to oil saturation.

The field B fault system (Figure 2) suggests that faulting occurred simultaneously and after the deposition of the BVF. It would be relevant to know which of the faults that cut the field are seals or porous pathways for adjacent fluid, potentially linked to magmatism and the reactivation of rift faults and structures. According to the FI data and the selective pervasive silicification in the wells, we believe that hydrothermal fluid migration reached the east flank of the reservoir structure before reaching the wells on the west flank. It was also considered that the association of the faults with the current structural highs and reservoir sections could play some role as a trigger for carbonate deposition. They could be a pathway for the entry of CO₂-rich fluids that interacted with the oversaturated connate fluid in ions such as Mg, Ca, etc., and would have enhanced the chemical precipitation, thus thickening the sections adjacent to the faults.

5. Conclusions

This study presents an interesting case where an integrated fluid inclusion analysis and reservoir diagenesis investigation, coupled with the evaluation of the maximum thermal burial and hydrocarbon generation of the Barra Velha Formation reservoir, are used to track the process of multiple episodes of petroleum generation–migration.

The hypothesis of in situ oil generation within reservoir facies was disregarded due the absence of anomalous thermal events and the inconsistence between the minimum entrapment temperatures of fluid inclusions and the necessary thermal evolution degree for the generation of entrapped hydrocarbons, thus suggesting that the thermal event that generated the hydrothermal oils occurred in another location in the basin. A limitation of our study is that we cannot rule out the possibility of overlapping sequences between heating events, which may influence the intermediate records of fluid–rock interaction events.

The observed salinity variation in the fluids responsible for mineral precipitation indicate events of mixing between saline connate fluids and pulses of brackish juvenile fluids. This reinforces that these juvenile fluids could percolate laterally and/or vertically in the reservoir section, carrying droplets of hydrothermal oil generated in other sections. This hydrothermal fluid front was responsible for silicification, dolomitization, and intense brecciation and led to enhanced porosity in different compartments of the reservoir, prior to the main pulse of the oil charge at the end of mesodiagenesis, which was responsible for the present oil accumulation. Further studies could clarify the role of the fault systems in the study field as either seals or porous pathways for fluid migration.