Laboratory Modeling of the Bazhenov Formation Organic Matter Transformation in a Semi-Open System: A Comparison of Oil Generation Kinetics in Two Samples with Type II Kerogen

Abstract

In this study, Kerogen conversion and oil production laboratory modeling results in Bazhenov formation source rock samples (Western Siberia, Russia) are presented. Two samples from one well with a similar composition and immature type II kerogen, which were accumulated in the same deep-sea conditions, were used for this investigation. Hydrous pyrolysis was performed under 300 °C, with liquid products and a sample portion collected every 12 h to study kerogen parameters via pyrolysis and the synthetic-oil composition via GC–MS. The transformation of pyrolytic parameters was similar to the natural trend previously determined for Bazhenov source rocks with different maturities. The synthetic oils’ normal alkane composition and biomarker parameters transformed with time. Sedimentary conditions and lithology biomarker parameters presumed to be constant (Pr/Ph, Ph/C18, H29/H30, and DBT/Phen) changed depending on the heating duration. The oil maturation increased slightly. Differences between the samples were detected in hydrocarbon generation endurance (5 and 8 days), n-alkane composition, and C27/C29 and DBT/Phen. A hypothesis about the influence of kerogen variability and mineral matrix on oil production was made. This paper provides the basis for more detailed and accurate investigation of the factors affecting kerogen cracking and hydrocarbon formation.

1. Introduction

Two of the main recent energy resources, oil and gas, are produced primarily from conventional reservoirs. These resources are generated in source rocks and accumulate in the reservoirs through migration processes. It is noteworthy that if there are unconventional reservoirs (more than several-percent porosity) in source formation, hydrocarbons (HCs) can retain there, forming shale oil and shale gas deposits. But enhanced porosity is formed mainly as a result of hydrocarbon generation. This process in sediments is associated with the thermal decomposition of macromolecular organic matter (OM), kerogen, during geological evolution [1]. The amount of formed HCs depends on the concentration and initial structure of kerogen, as well as on the completeness of OM transformation due to geothermal evolution under the heat flows produced by different sources [2].

The study of OM transformation and oil and gas generation processes is a complex goal, and two main approaches are used. The most common is focused on the factual material investigation and creation of general geological concepts based on a single result. It has been determined that there are various kerogen types that differ in composition, characteristics, and, thus, transformation pathways. These differences, in turn, specify different compositions of generated HCs. In this case, a reservoir’s fluid characteristics depend on the accuracy of predictions on oil and gas generation processes in sedimentary basins. For this purpose, the conception of kerogen chemical structure and its individual chemical reaction behaviors during kerogen cracking is necessary.

One of the main problems of kerogen investigation is that even the same type of kerogen is nonuniform. Its structure may differ significantly in the number and position of functional groups and individual elements in bulk volumes [3,4]. As kerogen is macromolecular, it cannot be investigated using standard gas chromatography–mass spectrometry (GC–MS) methods. Other methods, such as FTIR and NMR, assess only individual features of the structure: the amount of aliphatic and aromatic carbon and hydrogen or the presence of individual functional groups or elements (double, triple, C=O, and –OH bonds) [5,6,7,8]. To estimate the kerogen structure, a mathematical modeling approach can be used. Several researchers have published studies on kerogen structure and transformation during oil and gas generation by estimating the positions of functional groups and single, double, and triple bonds [9,10,11,12,13]. However, these estimations are only assumptions. Also, as kerogen structure varies in terms of formation and depends on sedimentary conditions and secondary processes, an OM model should be created for each basin. Correspondingly, oil and gas generation processes should be investigated. Determining rock sample characteristics cannot provide sufficient information. Hence, modeling of oil and gas generation processes should be performed.

Another method of investigating the kerogen structure and its transformation during HC formation is to simulate HC generation. Among the modeling approaches to kerogen cracking processes, two methods are distinguished: computer and laboratory modeling. Basin modeling is the main computer method [14]. However, its primary purpose is to predict reservoir expansion by reconstructing the geological history on a regional level. Laboratory modeling is based on kerogen cracking experiments, where the geological time is equalized by the heating temperature. These studies can be performed in open systems using classical pyrolysis [15] or in closed systems [16,17,18,19]. There are also single studies in which a semi-open system was simulated using specialized complex equipment, including a dropping valve and a combined autoclave with a pressing jacket [20]. Different methods are used as there is a lack of agreement on which system is more similar to in situ process conditions. A semi-open system that allows part of the products to escape the reaction zone after exceeding a certain pressure is presumed to be the most similar one, though experiments using this system need much more effort and cannot be performed in one stage. The need for sophisticated equipment and a several-fold increase in the number of measurements makes semi-open systems uncommon. The compositions of the generated products in semi-open and closed systems are almost the same as those of HCs generated in situ. Closed hydrous pyrolysis is commonly performed in petroleum geochemistry and has been used for various formation investigations [21,22,23,24,25,26]. The aim of these experiments was to assess changes in the qualitative and quantitative composition of the generated HCs, including oil and bitumen biomarker parameters’ behavior under different experimental conditions [27,28,29,30,31]. However, most research has focused solely on the final product. The kinetics of HC composition alteration have been poorly investigated.

In this study, hydrous pyrolysis was applied to the Bazhenov formation (BF) source rocks, which are considered as one of the main oil-source strata in Russia [32]. The experiments based on BF samples heating with water are described in different publications [33,34]. However, they were performed on the exact samples, while BF has nonuniform characteristics through the section and area. The kerogen heterogeneity is expressed both in OM maturity and maceral composition variation [35,36,37,38]. Since various OMs have a different structure, the HC generation processes will be performed in a number of ways [39]. The mineral matrix variability can also affect the kerogen transformation process [40].

A hydrous pyrolysis investigation of two BF rocks samples from the same well, which were formed in similar sedimentary conditions, was performed. To study the kinetics of OM transformation and oil generation, a unique experiment in a semi-open system, which is a series of the closed-system hydrous pyrolysis experiments, was set up. At each stage, a part of the sample and produced synthetic oil were collected for the investigation. The design of the experiment shrinks the role of a generated HC secondary cracking. It also allows for investigating kerogen cracking kinetics. The aim of the study was to investigate kerogen transformation and oil generation and to assess the similarities and differences in the processes for the similar rock samples. Based on these results, the possibility of unifying the section will be verified, and some conclusions on the factors that may affect oil formation in the BF source rocks will be drawn.

2. Geological Setting

The Bazhenov formation (BF) is the largest oil and gas source stratum in the world. It is the main source rock of West Siberia and includes the following analogue formations: Tutleimskaya, Mulyminskaya, Maryanovskaya, Yanovstanovskaya, and Golchikhinskaya (Figure 1) [41].

BF contains OM in several forms: kerogen, hydrocarbon compounds physically bound to kerogen or mineral matrix, and movable HC compounds. If unconventional oil reservoirs were formed in the BF due to the kerogen transformation, movable HC might be recovered using hydraulic fracturing and other special technologies. The following features are characteristic for BF rocks: uneven and poorly predicted lateral and cross-section distribution of lithotypes and OM; its maturity; complex structures of the mineral matrix and the pore space; multicomponent composition of rocks and OM. With a huge distribution area of about 1.2 million km², the thickness of BF sediments mainly does not exceed 50 m. Such a low thickness makes it almost impossible to use the seismic survey methods for the direct assessment of the deposit characteristics. Lateral BF rocks are substituted with the same-aged facies analogues: kerogen-saturated clay–carbonate–siliceous rocks are replaced by gray-colored clays, which are further displaced by sandy–silty glauconite-containing sediments penetrated on the periphery of the Bazhenov epicontinental paleobasin.

One of the main features of BF sediments is the specific rocks composition, which has no analogues in the West Siberian basin sections: an increased content of kerogen formed from aquagenic OM, a set of rock-forming minerals, determined by a different ratio of siliceous minerals (opal, cristobalite, tridymite, chalcedony, and quartz), carbonate and clay minerals. The main mineral compositions are clay–siliceous and siliceous, which were formed mainly by biogenic sedimentation [42] or the bacterial processing of sludge. Secondary processes proceeded at various stages of lithogenesis under the increased heat flow, which can lead to an unconventional reservoir forming. Meanwhile, rocks with relatively high porosity and permeability (up to 25% of pore space in radiolarites) are quite rare, and the entire thickness of BF is characterized by extremely low porosity values (less than 1%) and permeability (0.01–0.1 mD). Therefore, BF rocks have complex structures, and their transformation frequently occurs under elevated heat flow, which might affect the process of HC generation.

3. Materials and Methods

3.1. Sample Collection

To study the oil and gas generation process, two BF samples, Sample A and Sample B, taken from the middle of one section, were selected. The interval between the samples is 1 m. Both samples show similar properties. The rocks accumulated in uniform settings relative to the deep-water sea (water depth is less than 400 m). According to the X-ray diffractometry (XRD) results with a MiniFlex 600 X-ray diffractometer (Rigaku, Tokyo, Japan), samples are composed of silica (59.1 and 54.5%), clay minerals (22.4 and 17.2%), OM (8.4 and 17.1%, according to the pyrolysis investigations), and sodium feldspar (6.5 and 5.2%). There are almost no other minerals in the rocks (the content of pyrite and carbonates is less than 5% in total), and their composition can be considered similar. The reason for the difference might be in the minor difference of terrigene sediment ablation or bioproductivity local activation. Both processes do not affect the OM structure. Kerogen parameters are also almost identical: after removing all HCs with chloroform, hydrogen index (HI) parameters are 626 and 703 mg of HC/g TOC and T_max = 430 °C, which characterizes the OM as an immature marine type.

3.2. Hydrous Pyrolysis

Hydrous pyrolysis was performed for laboratory modeling of oil and gas generation. Pulverized rock samples weighing 50 g were placed in an autoclave (Comvics, Moscow, Russia), and 50 mL of distilled water was added. Autoclaves were heated in the furnace (Oven, Moscow, Russia) under 300 °C, with a water vapor pressure of 90 atm. The experiment duration was 10 (Sample A) and 12.5 (Sample B) days and ended when the amount of released synthetic oil was less than 0.0001 g. During the hydrous pyrolysis, every 12 h, autoclaves were elicited from the furnace, cooled to room temperature, and opened; the released light synthetic oil was collected from the water surface using 10 mL of pentane (chemically pure, Ekos-1, Moscow, Russia). After removing the solvent from the generated HCs via evaporation and achieving a constant mass of fluid, the amount of produced synthetic oil was determined by weighing, with accuracy of 0.00005 g. Synthetic oil and water were separated with a dividing funnel. A total of 1 g of each heated sample was taken from the autoclave for pyrolysis measurements. A volume of 50 mL of water was then added to each autoclave, and the experiment was resumed.

3.3. Pyrolysis

Pyrolysis of the initial samples and powders taken from the autoclaves at each experiment stage was conducted on a HAWK Resource Workstation (Wildcat Technologies, Humble, TX, USA) at a heating rate of 25 °C/min from 300 °C to 650 °C. The sample weight for analysis was 30 ± 1 mg. Before pyrolysis, all samples were dried at a temperature of 87 °C for 2 days. To estimate kerogen properties, HCs were removed from the samples using chloroform (chemically pure, Ekos-1, Moscow, Russia). Extraction was carried out until the concentration in the solution reached 0.000625% or lower. The research methodology and the parameters to be determined are described in more detail previously [15].

The main measured parameters are S2 (mg HC/g rock) and T_max (°C). The S2 parameter after extraction shows the potential of kerogen to generate HCs. T_max is the temperature at which the maximum HC release occurs. The parameter determines the maturity of rock samples [15]. Among the calculated parameters, hydrogen index HI was used. HI (mg HC/g TOC) commonly decreases with an increase in OM maturity. Total organic carbon (wt. %) shows the total content of organic carbon in the sample. To determine OM concentration, the TOC value was multiplied by the proportion of carbon in the immature OM, which is 71% in the BF source rocks of the region. The accuracy of pyrolysis measurements relative to the WT2-HAWK-STD standard was ±10% for S2 and HI measurements, ±2 °C for T_max, and ±5% for TOC. The calculations of other pyrolytic parameters given in the appendices were also described in detail previously [15].

3.4. Synthetic Oil Preparation for GC-MS Analysis

The obtained synthetic oils were separated into saturated and aromatic fractions according to SARA-analysis [43]. The asphaltenes were precipitated by adding a 40-fold excess of n-hexane (chemically pure, Ekos-1, Moscow, Russia) for 24 h and filtering.

Silica gel column chromatography with sequential elution with n-hexane (chemically pure, Ekos-1, Moscow, Russia) and toluene (chemically pure, Ekos-1, Moscow, Russia) was used for the further separation of saturated and aromatic fractions, respectively.

Thus, through hydrous pyrolysis, 20 and 25 synthetic oils were obtained from samples A and B. They were fractionated, and a total collection of 90 samples (45 of each fraction) was analyzed.

3.5. GC-MS Analysis

Gas chromatography (GC) with mass spectrometry (MS) was performed using a GC-MS chromatograph 8890 GC-5977B MSD (Agilent Technologies, Santa Clara, CA, USA) with a HP-5MS capillary column (30 m length × 0.5 mm internal diameter × 0.25 μm film thickness, Agilent Technologies, USA). Data collection and chromatogram processing were performed using MassHunter software (Agilent Technologies, USA, Version 12.1).

Helium was the carrier gas. The flow rate through the column was 1 cm³/min, the volume of injected sample was 1 mm³. The evaporator temperature was 290 °C, and the temperature of the interface was 300 °C. The MS data were acquired in a scan mode (m/z 35 ÷ 600 Da). The relative abundance of compounds was determined from peak areas using selected mass chromatograms for the integration of compounds. The temperature of the ion source was 230 °C, and the energy of ionizing electrons was 70 eV.

For aromatic fraction measurement, temperature conditions were set as follows: initial column temperature was maintained at 60 °C for 3 min, then increased to 180 °C at a rate of 15 °C/min, and then to 300 °C at a rate of 4 °C/min. Isothermal heating exposition at 300 °C was 6 min. The total analysis time was 56 min.

For saturated fraction measurements, temperature conditions were set as follows: the initial temperature was maintained at 60 °C for 3 min, then the temperature was increased to 180 °C at a rate of 27 °C/min, and then to 300 °C at a rate of 6 °C/min. Isothermal heating exposition at 300 °C lasted for 25 min. The total analysis time was 57 min.

The following parameters were used as reference parameters. Differences in sedimentary conditions and lithological composition of the source rocks were determined using following ratios: Pr (i-C19 wist)/Ph (phytan i-C20), H29 (17α(H),21β(H)-norgopan)/H30 (17α(H),21β(H)-gopan), t24 (C₂₄H₄₄ tricyclic terpan)/t23 (C23H42 tricyclic terpan), H35S/H34S (H35S-17α(H),21β(H),22(S)-pentaxisgohomogopan, H34S-17α(H),21β(H),22(S)-tetrakysgohomogopan), DBT/Phen (dibenzothiophene/phenanthrene ratio), and dia/(reg + dia)C27 (ratio of dia- and regular steranes C27). The contribution of OM of different nature is recorded by Pr/C17, Ph/C18 t23/H30, C29/C27 (the ratio of regular steranes C29 (propyl-cholestane)- and C27(cholestane)). To assess the maturity of the parent OM from which specific biomarkers were separated, the parameters Ts/(Ts + Tm) (Ts, 18α(H)-22,29,30-trisnorgopan; Tm, 17α(H)-22,29,30-trisnorgopan), MPI-1 (MPI-1 = 1.5(2MP + 3MP)/(P + 1MP + 9MP) MP is methylphenanthrene) were used. These parameters and their cut-off values are described in previous papers [44,45,46,47,48,49,50].

4. Results

Two kerogen-transformation laboratory modeling experiments used to study the oil and gas generation processes in the BF source rocks were performed. The samples pulverized to a 200-mesh powder had a similar composition with slightly higher amounts of silica and clay minerals in Sample A relative to Sample B (5% for each group of minerals), as well as a higher content of OM in Sample B (17.1 compared to 8.4 in Sample A). The experiments simulated a semi-open system, from which a part of the rock and generated synthetic oil were collected every 12 h.

4.1. The Results of Oil Generation Laboratory Modeling in a Semi-Open System

The pyrolysis results of the origin and heated samples after extraction (Figure 2) demonstrate kerogen properties’ (hydrogen index and T_max) evolution during laboratory modeling. The pattern is equal to the theoretical aspects determined by Kozlova et al. when measuring rock samples with different maturities [36]. During the experiments, T_max significantly increases only after 36–48 h of thermal treatment. Contrariwise, HI lowers from 600–700 mg HC/g TOC to 150–200 mg HC/g TOC during the first 48 h. In Sample B, the value of ca. 150 mg HC/g TOC was achieved in three stages (36 h). In Sample A, the most significant change of HI on ca. 250 mg HC/g TOC was promoted after the first 12 h, and then it took three more stages to achieve a value of 185 mg HC/g TOC. After 2 days of thermal treatment, the kerogen-transformation velocity lowers, and HI changes by less than 10–20 mg HC/g TOC for both samples. The difference in the kerogen-transformation process in the studied samples was also detected by the different resulting T_max values. For Sample A, after 10 days, T_max reached ca. 470 °C, while after 12.5 days of heating of Sample B, the value was ca. 450 °C. The pyrolysis results show the same trend of parameters evolution, but there was a difference between velocity and kerogen transformation efficiency for the samples.

The amount of synthetic oil produced for Sample B is higher than for Sample A (474 mg to 381 mg). The kinetics of product generation also differ significantly (Figure 3). For Sample A, the most intensive hydrocarbon generation is observed in the first 2 days. The maximum yield was obtained after the first 12 h, and then approximately the same amount of HC is released for 1.5 days. After 2 days, the yield decreases exponentially, reaching values close to 0 on the 6–7th day of the heating. For Sample B, two maximum yields of synthetic oil were recorded. The first was achieved after 1.5 days of hydrous pyrolysis, and the amount of generated oil was ca. 100 mg. The second maximum yield was established after 3.5 days of treatment and reached ca. 70 mg. The amount of releasing HC lowered to almost 0 on the 8–9th day of heating.

4.2. Normal Alkane Distribution in Synthetic Oils

Figure 4 shows 71 m/z mass fragments for normal alkane distribution in synthetic oils generated at different stages of hydrous pyrolysis (by the total ion current). N-alkanes were only determined in the first 5–8 days of the laboratory modeling.

In the synthetic oils produced after 5 days of the hydrous pyrolysis of Sample A, most n-alkanes were not identified, and only isoprenoids were detected. For Sample B, normal alkanes were found in oils till the end of day 8 of exposure. The yield of n-alkanes correlates well with the yield of maltenes during heating.

The percentage composition of n-alkanes in synthetic oils at each stage of the experiment was determined using GC–MS (Figure 5 and Figure 6). The synthetic oils generated in Sample A during 0.5–5 days contain C₁₃ to C₃₈ n-alkanes. In the first stages of the experiment (12–48 h), there was no explicit maximum in the composition. After the 3rd day of treatment, the n-alkane C₁₉ became predominant. The secondary maxima for n-alkanes C₁₇ and C₂₁ were also detected. The longer the exposure, the greater the predominance of n-alkane C₁₉ relative to the other ones in the synthetic oils (Figure 5). As mentioned above, only several n-alkanes could be determined in synthetic oils produced after the 5th day of the HC generation laboratory modeling.

The synthetic oils in Sample B contain n-alkanes C₁₁-C₃₆ (Figure 6). Similar to the Sample A synthetic oils, there is no explicit maximum in the n-alkane composition till 4.5 days of the experiment, and the distribution is almost identical. After the 5th day of hydrous pyrolysis, a well-defined double maximum related to the C₁₉ and C₂₁ in n-alkane composition is observed. Also, no compounds with a chain length of less than C₁₆ and more than C₃₀ are detected. After the 8th day, most n-alkanes are not found in these synthetic oils.

4.3. Biomarker Parameters of Synthetic Oils

The kinetics of the synthetic oil generation process were investigated by analyzing biomarker parameters. The GC–MS results show that during laboratory modeling, the amount and ratio of biomarkers vary. Figure 7 shows that the ratio of hopans H29 and H30 and terpans t23 and t24 changes during the hydrous pyrolysis. The amount of terpans (in comparison with hopanes) decreases with the increase in the thermal treatment duration, foremost in synthetic oils from Sample A.

The synthetic oils’ biomarker parameters (listed in methods) star diagrams for samples A and B were used to demonstrate changes in the molecular composition of the generated HC (Figure 8 and Figure 9). Similar to the n-alkane composition of oils generated from Sample A after the 5th day of the experiment, most biomarkers were not detected. The diagram shows that some of the biomarker parameters change in a fairly wide range. Thus, the parameters Pr/Ph, Ph/C18, H29/H30, and DBT/Phen, which are responsible for marine and continental, oxidizing and reducing sedimentary conditions, as well as for the contribution of clay and carbonate materials in a source rock, vary the most. On the other hand, the sedimentary condition parameters t24/t23, H35S/H34S и dia/(reg + dia)C27 do not vary significantly. Maturity biomarker parameters increase with the elongation of thermal treatment. For Sample B, more significant variations in the parameters t23/H30 and MPI-1 were recorded. Biomarkers were detected in synthetic oils through the first 8 days of the hydrous pyrolysis.

The dynamics of parameter variation with time were studied to investigate the differences between Samples A and B (Figure 10). Parameters Pr/Ph, H29/H30, t23/H30, and MPI-1 change with the same pattern. However, if the values for t23/H30 and MPI-1 are equal for the synthetic oils for the same duration of treatment, parameters Pr/Ph, H29/H30 achieve the same changes at different times. The maximum Pr/Ph ratio was specified in generated HC after 3.5 days of Sample A and 5 days of Sample B hydrous pyrolysis (Figure 10a). The most significant differences were recorded for C27/C29 and DBT/Phen parameters.

The results show that at the first stage of the experiment, almost all biomarker parameters in both samples are approximately equal between samples. The behavior of the Pr/Ph and H29/H30 parameters can be considered the same (Figure 10a,c). In both cases, the parameter values systematically grow to the maximum value, then decrease. The difference for the samples investigated is specified by the time the maximum value was achieved. For synthetic oils from Sample A, this time is 3.5–4 days, whereas 5 days are necessary for HCs in Sample B.

Significant differences have been found for parameters C27/C29 and DBT/Phen, in particular, the ratio of C27/C29 steranes, which shows a difference in the source-rock OM nature. In Sample A, synthetic oils systematically decrease from ca. 3.5 to 1. For Sample B, the values between 3 and 5 were recorded until the 4th day of hydrous pyrolysis. Then, C27/C29 slightly decreases to a value of 2. The DBT/Phen ratio in synthetic oils from Sample A decreases sharply during the first day of treatment, then it remains almost constant. In synthetic oils from Sample B, two maxima of ca. 1.3 and ca. 1.9 upon 1.5 and 3 days of the experiment were found. After the second one, a decrease in values up to ca. 0.4 with a substantially exponential pattern was determined.

For maturity parameters, Figure 10f shows the systematic growth of the MPI-1 parameter in the synthetic oils from both samples. The value of the Ts/(Ts + Tm) parameter increases with the same trend. The value of the t23/H30 parameter coincides during the first 5 days of hydrous pyrolysis (Figure 10, e), while, theoretically, it should stay constant. After day 5, t23 terpan was not detected in the synthetic oils from Sample A, while Sample B shows a sharp increase in t23/H30, attaining 7 after 7.5 days of the experiment.

5. Discussion

Several oil-generation processes in the source rocks’ behaviors were established based on the results of the kerogen transformation laboratory modeling experiment in the semi-open system. The parameters that characterize changes in the kerogen structure and the molecular composition of synthetic oils change differently. Theoretically, the process of the geological evolution of the hydrogen index, which expresses the hydrocarbon group number in the OM structure that can transform into HC, decreases with the kerogen maturity, while the T_max parameter increases [51]. An increase in the T_max parameter depends on the kerogen structure strengthening and, thus, on the necessity to apply more energy to release HC. The results of laboratory modeling in the semi-open system agree well with theoretical concepts [36] and other oil-generation experiments performed with a different methodology. In [52], researchers accomplished hydrous pyrolysis in a closed system for 48 h under 300–400 °C on source-rock samples with type II kerogen. Pyrolysis of the heated samples after extraction demonstrated the same changes of HI and T_max in type II kerogen from different formations during maturation (Figure 11).

In the first stages of kerogen transformation, a significant decrease in HI and a slight increase in T_max is observed. With an increase in applied energy (in terms of temperature and time of thermal treatment), HI lowers less intensively, and as the value of about 50–100 mg HC/g TOC is achieved, a significant increase in T_max is found. The results are in good agreement with the pattern of the changes of pyrolytic parameters of BF rocks through maturation measured for samples from various wells [36]. Thus, hydrous pyrolysis experiments in a semi-open and closed systems can simulate kerogen transformation during geological evolution in situ.

The differences in kerogen pyrolysis parameters through the laboratory modeling of Samples A and B establish that the processes are not identical. In Sample A, HI decreases by ca. 1.5 times after 12 h, while in Sample B, the parameter decreases only by 50% after 24 h, and 36 h is enough to reach values less than 200 mg HC/g TOC. On the contrary, in Sample A, the decrease in HI is greater. It can be assumed that kerogen in Samples A and B differs in chemical structure, so different amounts of energy are necessary to break bonds and release HC, and the amount of these HC differs due to the applied energy.

The assumption above is confirmed by the yield of synthetic oils. Figure 3 shows that for Sample A, the maximum yield was produced after 12 h, whereas for Sample B, there are several generation steps that result in the maximum yields after 1.5 and 3.5 days of thermal treatment.

The assumption that kerogen in Samples A and B has different chemical structures and transforms via various pathways can be confirmed by the composition of the generated synthetic oils. Figure 5 and Figure 6 show that the proportion of normal alkanes in Samples A and B oils differs and varies significantly through the experiment. In the initial stages, a fairly broad distribution of compounds with different chain lengths was observed. After 3 days of treatment, n-alkane C₁₉ becomes predominant in synthetic oils from Sample A, and for Sample B oils, the double maximum concentration of C₁₉ and C₂₁ was detected after 4.5 days. The termination of n-alkane generation was found after 5 days for Sample A and after 8 days for Sample B. A decrease in the short-chain n-alkane concentration and the presence of a large number of isoprenoids may indicate that at the final stages of kerogen evolution, its structure contains almost no saturated straight-chain HC groups, compounds with complicated presumably aromatic structure are formed, and the process might require high amount of energy applied with a low HC yield. As the yields at these stages lower significantly, the detected change in the synthetic oil composition would not be observed in experiments in a closed system [25].

A common feature in the n-alkane composition was observed for both samples. With an increase in the experiment time, a secondary maximum with C₂₆ predominance was detected. Its evolution at a certain stage of hydrous pyrolysis indicates changes in kerogen structure. Either new HC radicals become available for cracking after releasing some amounts of synthetic oils, or kerogen structure realignment occurs. Another possible reason is that there are different kerogen macerals in the samples with various chemical structures. Investigations of BF OM maceral composition and their transformation under hydrous pyrolysis show that bioclasts and macerals formed from higher plants or animal organisms remain, entering the oil generation process under higher activation energies [38]. It can be assumed that the studied samples also contain such macerals, which, after 2.5 days for Sample A and 5 days for Sample B, begin to generate hydrocarbons. The difference in the generation initiation time may indicate that the structure of macerals might vary. This confirms the previously mentioned idea that the composition and structure of kerogen in these two samples investigated is different. The variation of the C₂₇/C₂₉ diasterans ratio confirms a difference in kerogen nature. Diasterans C₂₇, C₂₈, and C₂₉ show the contribution of various components in the parent OM composition, such as higher plants, algae, phytoplankton, etc. [53]. In fact, these parameters confirm the initial difference in kerogen chemical structures found by other indirect signs.

The maturity biomarker parameters in synthetic oils show a gradual increase with the experimental duration. Such an increase does not correspond to a sharp and significant increase in the kerogen maturation according to pyrolysis data. Meanwhile, the results are in good agreement with the other kerogen maturation laboratory modeling results [49,54]. On the other hand, the composition of generated fluids is close to the natural ones [55]. Thus, if increased heat flow affects source rocks (within hydrothermal processes, seals, and intrusions), biomarker analysis results of maturity estimation could be inaccurate. Maturity biomarker parameter values might be lower as the isomerization processes proceed slower or are absent under high temperatures and pressures through hydrous pyrolysis [56,57,58,59,60]. Thus, kerogen maturity is better assessed by pyrolysis after extraction, which is in good agreement with in situ processes.

For some of the synthetic-oil biomarker parameters, the patterns with time are different from the trends established on well samples. The most significant difference was determined by the parameters that determine sedimentary conditions and the lithological composition of rocks. Ulmishek showed that the investigated rocks are accumulated in the same marine conditions at a large distance of ca. 200 km from the shoreline, which is confirmed by the trace amount of terrigenous material [61]. The mineral composition of the rocks is similar. Thus, the environmental conditions of the biomarker parameters should be the same and stay constant throughout the kerogen evolution [49]. The sedimentation conditions and rocks’ lithological composition parameters t24/t23, H35S/H34S, and dia/(reg + dia)C27 change insignificantly. However, other parameters (t23/H30, DBT/Phen) stay almost constant in the synthetic oils from Sample A, first increased, and then decreased in Sample B oils. The kerogen structure changes during hydrous pyrolysis caused by either HC release or the embedding into kerogen of both asphaltenes [52] and individual compounds, such as diasterans [62], might be the source of parameter variation. In addition, if diasterans embed the kerogen structure, their ratio changes during hydrous pyrolysis, and the corresponding biomarker parameters shift.

The catalytic effect of minerals on kerogen cracking might be another reason for the biomarker parameter variation. Individual minerals and kerogen are collocated in the rocks not uniformly, and through the kerogen transformation process, some minerals might get in contact with OM and stimulate or inhibit its transformation and different biomarker generation. This effect was illustrated dramatically in [40], where researchers added various minerals to the isolated kerogen and evaluated the released HC composition change through pyro-GC–MS. Their paper demonstrates a significant change in the generated HC composition depending on the added mineral.

Another parameter that varies in the synthetic oils released at different times is Pr/Ph. The parameter characterizes continental oxidative sedimentation conditions at Pr/Ph > 3 and marine reducing conditions at Pr/Ph < 1 values. The onset value of the parameter is ca. 1, which corresponds to relatively deep-sea sedimentation conditions that are consistent with the literature and the lithological description [32]. The maximum values achieved during hydrous pyrolysis are 2.8 in the synthetic oils from Sample A and >5 in Sample B HC. Such variations cannot be defined by the kerogen sedimentary conditions’ metamorphosis and are likely caused by the thermal influence on the composition of normal alkanes and their isomers.

Some variation in the maturity, sedimentation conditions, and lithological composition of rocks biomarker parameters between synthetic oils and natural fluids caused by the differences in the laboratory modelling and natural processes conditions is described in [27,28,49,63]. Meanwhile, other investigations show a similar composition of rock HC and synthetic oils [21,50].

Thus, the described processes might affect the kerogen structure and synthetic oils’ composition variation at different levels. The assumption that different maceral compositions affect the process most significantly might prevail. Meanwhile, the lack of information and methods to determine and construct kerogen structure did not allow us to investigate the process and role of various factors, such as the rock matrix catalytic or inhibitor role and the effect of experimental conditions. Combination of different optical and spectrometry methods might provide more data for the future prediction of in situ HC generation under different conditions and with the application of various technologies.

6. Conclusions

According to the results of the investigation, several conclusions on the process of OM transformation and hydrocarbon generation can be drawn. The hydrous pyrolysis in a semi-open system provides data for kerogen transformation and the dynamic pattern of oil and gas generation. (i) Step-by-step kerogen transformation was performed, and the pyrolysis parameters’ change through the experiment correspond to the trend determined for type II kerogen transformation in situ for the BF and other source rocks. The dynamic pattern draws indirect conclusions about the similarities and differences in the kerogen cracking in the two investigated samples. The time necessary to achieve the same maturity stage and differences in HI and T_max parameter values under the same conditions of thermal treatment suggest that the kerogen structure in the samples varies. (ii) The composition of synthetic oils differs depending on hydrous pyrolysis duration. Diversity in oils from two samples was also found. As the sedimentary conditions are similar for both samples, variations in the n-alkane distribution over the experiment duration and between the samples and changes in the C₂₇/C₂₉ and Pr/Ph parameter suggest differences in kerogen structure. (iii) The variation in oil generation within two similar samples could be specified by the maceral composition of OM. Some researchers also report that the mineral matrix can affect the process of kerogen transformation, as some minerals could function as catalysts or inhibitors. Other authors suggest that kerogen rearrangement and embedding some HC compounds into the kerogen structure could take place through hydrous pyrolysis. The investigation methods for the characteristics affecting organic matter cracking should be improved, and a combination of optical and mass spectrometry methods should be used in different stages of hydrous pyrolysis in a semi-open system to detect all the processes in the system.