Petroleum Geochemistry

Abstract

Petroleum geochemistry has entered its second period of growth. The first period, largely associated with conventional oil and gas, occurred in the 70s and 80s when the classic works on source rock characterization, biomarkers, depositional systems, and petroleum generation, including kinetics and basin modeling were the focus. The second period began slightly after the turn of the century as a consequence of the “unconventional resource” revolution and the interest in distressed resources developed, the focus turned to non-hydrocarbon contaminants, new interest in hydrocarbon expulsion and retention, identification of tight rock pay zones, and the development of organic porosity. This chapter will discuss source rock characterization and formation, petroleum generation, expulsion, and retention, correlation among hydrocarbon accumulations and to their source rock(s), and organic porosity.

1. Introduction

There are five components to a petroleum system - hydrocarbon charge, reservoir, seal, trap, and overburden [1]. When assessing exploratory risk each of these elements is directly assessed except for overburden, which is incorporated into the different risk elements (e.g., overburden is incorporated into charge through thermal maturity, seal and reservoir through porosity and permeability reduction associated with compaction). The absence of any of these elements brings the chance of exploratory success to zero. Hydrocarbon charge is considered the most important component of any petroleum system evaluation [2] because there is no alternative. In frontier regions and play extensions, post-drill assessments have indicated that the absence of hydrocarbon charge is a disproportionate cause of exploratory failure [3, 4]. Significant improvement in exploration efficiency was reported when geochemistry was taken into consideration as compared to simply assessing opportunities by trap size alone [5]. Fundamental to understanding hydrocarbon charge is clarity on its components which include the source rock presence and quality, generation process (maturation), and alteration (e.g., biodegradation, thermal cracking, phase segregation).

The importance of the organic matter to the formation and accumulation of hydrocarbons was fundamentally established by (1) the identification of porphyrins, a chlorophyll derivative, in shales, coals, and crude oils [6], and (2) the observation of threshold level of total organic carbon (TOC), approximately 1.5% as a mean of petroliferous basins, rather than the 0.35% of non-petroliferous basins of the Russian Platform [7].

Since these works, and especially over the past five decades, there has been considerable advancement in the foundational understanding of hydrocarbon charge. There have effectively been two major periods of advancement in petroleum geochemistry. The first growth episode occurred, in part, as a result of advances in analytical methods as well as insights into the controls on source rock development and the processes of hydrocarbon generation, expulsion, migration, and alteration. During this period the application of gas chromatography/mass spectrometry (GC/MS) became routine for the assessment of source rock depositional setting and thermal maturity; and basin models became commonplace, requiring an understanding of the kinetics of hydrocarbon generation. The second growth period came with the increase in the importance of self-sourced petroleum systems and tight rock resources. During this recent phase, the focus has been on the identification of landing zones, hydrocarbon expulsion and retention, hydrocarbon cracking, and the development of organic porosity.

This overview discusses the identification, characterization, and formation of hydrocarbon source rocks, the generation process, the characterization of produced fluids including post-accumulation alteration processes, hydrocarbon migration, and establishing genetic relationships among hydrocarbon accumulations, and to their source rock(s), and organic porosity.

2. Source rock identification and characterization

It is important to establish a consistent definition of source rock. A source rock is a rock that contains sufficient quantities of organic matter that after having achieved the appropriate thermal maturity will generate and expel sufficient quantities of hydrocarbons to result in an accumulation. At this point issues of commerciality are not considered because they are dependent on logistics, the presence of prior infrastructure as well as commodity price.

Petroleum source rocks are atypical and are not uniformly distributed either stratigraphically or spatially [8]. The mean value for organic carbon in fine-grain sedimentary is ~0.7 wt.% with a standard deviation of 0.3 wt.% as established using a statistical approach and more than 15,000 fine-grained rock samples worldwide [9]. It was then noted that source rocks should display above-average TOC levels establishing a threshold TOC of 1.00 wt.% (Figure 1). However, a review of data from a number of world-class source rocks such as the Kimmeridge Clay (North Sea Basin), Green River Formation (western United States), Pematang Formation (Central Sumatra Basin, Indonesia), Bucomazi Formation (Lower Congo Basin, Angola), Hydria-Hanifa Formation (Saudi Arabia), Maykop Formation (South Caspian Basin, Azerbaijan), Shublick Formation (Alaska) and Kazhdumi Formation (Mesopotamian Foreland Basin, Iran), all contained significant stratigraphic intervals where organic carbon contents exceeded several weight percent organic carbon. This indicates that source rocks, in fact, typically contain TOC levels that significantly exceed the 1.0% wt.% TOC threshold.

Figure 1.

Global distribution of total organic carbon within fine-grain sedimentary rocks. Insert represents the organic carbon measured at the type locality of the Kimmeridge clay (United Kingdom), after Bissada 1982 [9].

It has also been suggested that there is an upper limit for TOC that limits a source rock’s effectiveness. It is suggested that at TOC levels of 12 to 15 wt.%, oil is retained within the source rock limiting its effectiveness [10]. This upper limit may also partially explain why most coals do not act as an effective source [11].

It was, however, established early that not all organic matter is the same with respect to hydrocarbon generation and that the assignment of source rock potential based on organic carbon is insufficient. Similar quantities of organic matter can have yields that range over several orders of magnitude depending on the type of organic matter and the thermal maturity (Figure 2). This question of yield was approached using the total generation potential (free hydrocarbons + generatable hydrocarbons: S1 + S2) as determined using Rock-Eval pyrolysis. A threshold of 2.5 mg HC/g rock [9] was considered for a possible oil-prone source rock (Figure 3). This threshold was established as outlined above for organic carbon. A physical reason for this threshold also appears present. This reported threshold is consistent with the previously reported minimum of 825–850 ppm hydrocarbons thought to be required for expulsion to occur [10]. A rock having a total generation potential of ~2.5 mg HC/g rock as it approaches the main stage of hydrocarbon generation approaches a free hydrocarbon content consistent with this threshold. Thresholds for possible gas-prone source rocks are less well-defined, in part, because of their different expulsion mechanisms [12]. Oil expulsion requires that the pore network becomes saturated, and the rock becomes over-pressured. In contrast, gas expulsion can occur through diffusion which simply requires a concentration gradient once the sorption capacity of the source is achieved [13] or in solution within a liquid hydrocarbon phase.

Figure 2.

Comparison of residual generation potential (S2) of samples with similar total organic carbon content. Note that for the same TOC, hydrocarbon yield can vary by order of magnitude.

Figure 3.

Global distribution of total generation potential of fine-grain samples containing a minimum of 0.5 wt.% TOC. Insert represents the total generation potential measured on samples greater than 0.5 wt.% TOC at the type locality of the Kimmeridge clay (United Kingdom), after [9].

The atomic H/C and O/C ratios were used to define three primary kerogen types as an explanation for the observed differences in hydrocarbon yield and product character [14]. This van Krevelen diagram has been modified to provide more specific guidance on product characterization (i.e., oil yield) [15] and visualized here in Figure 4.

Figure 4.

Conventional van Krevelen diagram based on the atomic H/C and O/C ratios. Relative oil and gas yields have been added.

Type I kerogen was defined using the Green River Formation and algal kerogens and has the greatest hydrocarbon yield for a given mass resulting from the abundance of hydrogen. When mature type I kerogen will yield principally oil with a lesser amount of gas. The kerogen structure contains abundant long-chain hydrocarbons [16]. This type of kerogen is principally derived from algal material and often appears associated with marine and lacustrine carbonate depositional systems.

Type II kerogen displays lower atomic H/C and higher atomic O/C ratios than Type I organic matter. It produces both oil and gas upon maturation and was defined using the Schistes Carton Formation (lower Toarcian, Paris Basin, France) and Silurian shales of North Africa. The kerogen structure is much more diverse than Type I kerogen due to the diversity of the organic material that led to its development, which includes algal material, plant cuticle, spores, pollen, and resin, which may be microbially reworked. Although often considered to represent a marine depositional system, such kerogen was found to also dominate in siliciclastic-dominated lacustrine systems, such as the Pematang Formation of Central Sumatra.

As implied, the difference in organic matter type between a clay-rich mudstone and a carbonate source rock rest with one of the foundational differences in the development of these two rock types. Carbonate rocks are generally considered to be autochthonous with both the mineral and organic matter forming at or very near to the depositional site. In contrast, clay-rich mudstones are derived from both inorganic and organic material that is transported to their depositional site, reflecting the provenance of the drainage basin, with the lesser autochthonous contribution.

A subset of Type II kerogen is Type II-S, which contains greater than 6% organic sulfur [17]. This differentiation is important because the C-S bond is weaker than the C-C bond and generation proceeds at lower levels of thermal maturity, producing products with greater amounts of asphaltenes and resins.

Type III kerogen was defined using Cretaceous shales from the Douala and Western Canadian Sedimentary basins. It has lower H/C and more elevated O/C ratios than Type II kerogen. It produces the lowest amounts of hydrocarbons per unit mass and yields principally gas. The kerogen structure is envisioned to be dominated by interconnected aromatic rings, with shorter chain hydrocarbon elements. Although this type of organic matter is often associated with vitrinite (a wood derivative) it may also be derived through the poor preservation (oxidation) of marine organic matter.

As a consequence of thermal maturation and the generation of products including organic acids and hydrocarbons both the atomic H/C and O/C ratios decrease. In the case of Type I kerogen, there is a rapid decrease in the atomic H/C ratio and a modest decrease in O/C ratio with increasing thermal maturation. In contrast, there is a rapid decrease in the atomic O/C ratio and a modest decrease in the atomic H/C ratio for Type III kerogen. These changes result in an inability to differentiate among the different kerogen types using their elemental composition at more advanced levels of thermal maturity and alternative means are required for such kerogens.

Subsequently, a fourth kerogen type has been defined, which represents residual organic matter [18]. It displays very low atomic H/C ratios and highly varied atomic O/C ratios. This material is largely inert and incapable of yielding any significant amount of hydrocarbons. It is dominated by inertinite. This material commonly forms through prolonged transport, very slow sedimentation rates leading to long exposure times, or forest fires.

The aforementioned approach to organic matter characterization requires the isolation of kerogen from the rock matrix. This is a time-consuming process that utilizes hydrochloric and hydrofluoric acids as well as requiring relatively large sample volumes. An alternative was proposed that was rapid and required only grinding as sample preparation and did not require large sample volumes. This method was Rock-Eval pyrolysis, where the sample was heated in an inert atmosphere. Two of the measured parameters are used to calculate the hydrogen index (S2*100/ TOC) and the oxygen index (S3*100/TOC, where S3 represents the CO₂ yield) are substituted for the atomic H/C and O/C ratios, respectively (Figure 5).

Figure 5.

Modified van Krevelen diagram based on the rock-Eval parameters the hydrogen and oxygen indices. Arrows represent changes in parameters as a function of increasing thermal maturity, increasing carbonate (especially siderite) content, and decreasing organic carbon.

Although these indices have become routinely accepted for kerogen characterization there are some limitations that are known to exist and should be considered when interpreting the data. For samples with very high generation potentials, the use of the standard sample size may result in the saturation of the flame ionization detector, which produces an apparent reduction in S2 yield and consequently the hydrogen index making the sample appear more gas-prone than would be implied if elemental analysis on isolated kerogen was used. In addition, several studies have shown that there are mineral matrix effects. These effects are especially notable for samples with lower organic carbon contents. It is suggested that for samples with less than 2 wt.% TOC hydrocarbons are retained by the rock matrix, especially in clay-rich samples. This retention reduces the apparent generation potential and the derived hydrogen index [19]. It was also observed that the oxygen index was sensitive to the presence of carbonate minerals, especially siderite. These effects cause the organic matter to appear more gas-prone than in kerogen isolates. Alternative means of correcting the oxygen index for the presence of siderite-derived CO₂ have been proposed [20, 21], however, these approaches alter the value proposition, which was a rapid and simple means to assess generation potential, organic matter type, and thermal maturity. An alternative approach to organic matter characterization without the possible oxygen index complication relies on the relationship between the hydrogen index and Tmax (Figure 6). This approach is still limited at lower TOC values.

Figure 6.

Alternate means of characterizing organic matter utilizing the hydrogen index and Tmax. Arrows represent changes in parameters as a function of increasing thermal maturity and decreasing organic carbon.

Alternative pyrolysis approaches have been developed that provide additional information. The first adds gas chromatography to the pyrolysis unit and is known as Py-GC. This analytical approach provides a more detailed understanding of the products generated beyond a simple assessment of oil- and gas-proneness [22, 23]. A chromatogram of isolated kerogen through Py-GC with vented free hydrocarbons below 320°C (equivalent to Rock-Eval S1 peak) and then pyrolyzed up to 600°C is produced from what essentially was the Rock-Eval S2 peak (Figure 7). These chromatograms provide information on such geochemical properties as waxiness, relative abundance of naphthenes, and aromatic compounds. The relative abundance of C₁-C₅, C₆-C₁₄, and C₁₅⁺ in the Py-GC was used to assess the oil and gas-proneness of different types of kerogens [23].

Figure 7.

Pyrolysis-gas chromatograms of A - Green River formation (Utah, United States); B - Kimmeridge clay (United Kingdom); C - Talang Akar formation (Indonesia); D - Banquereu shale (Atlantic Canada).

Another thermal extraction-pyrolysis innovation has been developed, which has a more complex temperature ramp and is designed to better characterize the free hydrocarbons present in the rock, where the free hydrocarbons are broken down into four fractions (thermal extraction <350°C), representing C₄-C₅, C₆-C₁₀, C₁₁-C₁₉, and C₂₀-C₃₆ (the four oil fractions in Figure 8). The K-1 peak in Figure 8 represents pyrolysis of kerogen at 350–600°C. The Petroleum Assessment Method (PAM) was developed to better assess the nature of the hydrocarbons present in self-sourced petroleum systems [24].

Figure 8.

Representative PAM pyrolysis of Devonian shale (Western Canadian Basin).

Part of the assessment of the validity of a geochemical assessment is a determination as to whether a sample has been stained (natural processes) or contaminated (anthropogenic processes). This assessment is based on the relationships between the abundance of free hydrocarbons (S1) and the total organic carbon content (Figure 9), and the relationship between Tmax (temperature of maximum hydrocarbon yield) and the transformation ratio (S1/(S1 + S2); Figure 10). These assessments do not differentiate between natural and anthropogenic hydrocarbons additional analyses would be needed for this differentiation.

Figure 9.

The relationship between total organic carbon and S1 yield is used to define the presence of staining or contamination.

Figure 10.

The relationship between Tmax and the production index is used to define the presence of staining or contamination.

The S1*100/TOC (OSI; oil saturation index) ratio was proposed for identifying potentially productive zones, with values greater than 100 mg HC/g TOC being zones of interest [25]. This approach is essentially limited, however, to wells not drilled with an oil-based drilling fluid system.

There has been some recent work that has also led to questions on the validity of source rock assessment and characterization when organic-based drilling fluids are used. Organic-based drilling fluids are becoming more commonly used because of their greater stability at higher temperatures and improved hole stability when shales are water-sensitive [26]. It was reported that the often-used solvent pretreatment of contaminated samples does not permit an assessment of the original in situ characters of the rock [27]. The reported organic carbon, generation potential, hydrogen, and oxygen indices were all impacted by the contamination by the drilling fluid and the solvent extraction of the contaminated samples.

3. Source rock depositional controls

As a consequence of the uniqueness of petroleum source rocks, it has been generally accepted that they form under somewhat distinct sedimentary conditions. It was suggested that nearly half of the known source rock systems lack modern analogs (e.g., anoxic epeiric seaways and anoxic oceans [28]). In general, there have been three principal schools of thought on source rock deposition: 1) enhanced organic preservation, often associated with anoxia; 2) enhanced primary productivity, often associated with oceanic upwelling or riverine transport of nutrients; and 3) sedimentation rate, often associated with either rapid removal of the sediment from the various microbial zones or through the concentration of organic matter through a lack of dilution by sediment (i.e., a condensed section). Arguments have been presented to support each as a stand-alone model.

The enhanced preservation model is largely based on the argument that anoxic environments, where oxygen consumption exceeds supply, favored preservation [29]. Such settings are associated with stratification, reduced circulation, water body isolation, or estuarine flow. The primary argument for this was the presumed relative inefficiency of anaerobic processes, which slows decomposition [30]. However, activity levels of anoxic and oxic microbial communities have been shown to display similarities [31]. It appears that the absence of meibenthos and macrobenthos may be more important than microbial rates because they are more efficient consumers of organic matter compared to microbes [32] and also provide a means to irrigate the sediment through bioturbation [33]. Similarly, the absence of alternative oxidizers such as sulfates also leads to more efficient preservation. This limits the source rock potential of evaporitic settings once gypsum precipitation is initiated, and sulfate reduction may occur. Another argument for enhanced preservation was associated with settling or exposure time within the oxic portion of the water column. It was observed that there were order of magnitude reduction in organic matter preservation efficiency from the shelf to the central ocean basin as a result of exposure time [34]. Further reports suggest that settling time could be reduced through the pelletization process, where the increase in particle size and the incorporation of mineral matter increased the settling rate with added protection coming from the mucilaginous cover that the pellets have after passing through the digestive system [35]. It should be noted, however, that stratification may limit nutrient renewal and lead to oligotrophic conditions, suggesting limited autochthonous input and that under such circumstances terrestrial input may be favored.

The primary productivity model was based on the general concept that elevated amounts of organic matter would be incorporated into the sedimentary record if productivity was high [36]. Higher levels of productivity are associated with regions of nutrient renewal such as coastal upwelling, seasonal water body turnover (which is especially common in lake systems and temperate water bodies), as well as riverine input. Numerous publications attempted to highlight areas of high productivity through time through paleoclimate and paleocirculation modeling (e.g., see [37]). In the modern ocean, there are numerous regions of high productivity, however, that lack significant organic carbon in the sediment. This is clearly documented in the Southern Ocean where an intense upwelling system has been established but is also a region where freshly-oxygenated bottom waters are present. Here the sediment appears dominated by siliceous tests and TOC is minimal, (typically below 1.0 wt.%) as a result of organic carbon’s brief residence time of 15 to 150 years [38]. Attempts to correlate regions of modeled high productivity have had limited success. In part, this is because of factors beyond nutrient availability that influence productivity such as turbidity. For example, the suspended load of the Mississippi River results in limited light penetration at the river’s mouth. The region of elevated productivity is thus shifted further offshore to where the sediment has salted-out.

The discussion on the role of sedimentation rate follows two paths. Early arguments suggested greater potential for organic matter preservation when sedimentation rate was high [39]. It was suggested that rapid sedimentation would reduce the time spent within the various microbial zones ranging from oxidation through sulfate reduction and eventually methanogenesis. This concept appears supported by the positive correlation between sedimentation rate and total organic carbon [40, 41]. The specific relationship appears to differ among lithologies. However, when the sedimentation rate exceeds approximately 20 m/MY, the organic carbon content begins to decrease as a result of dilution by sediment. An increase in carbon content with an elevated sedimentation rate can only occur if the level of primary productivity increases. In contrast, it’s suggested that source rocks are associated with condensed sections, where dilution by sedimentary material has been minimized. An often-cited example of a condensed section source rock is the Shublick Formation in Alaska [42], which also appears to be associated with elevated productivity as suggested by the presence of phosphorites [43]. Not all sediment starved areas develop oil-prone source rocks. It was reported that for a condensed section deposited under oxic conditions such as the Upper Jurassic/Lower Cretaceous of SE France the section is bioturbated and TOC values are less than 0.25% [44]. The influence of sedimentation rate was also noted as part of the preservation model, where more oil-prone material was associated with higher sedimentation rates and inert material was preserved with slow sedimentation rates [29].

In addition to the three working models, it is also important to understand that the reactivity of organic matter is not uniform. It was noted that under oxic conditions planktonic material would degrade more rapidly than the remnants of vascular plants because of chemical differences [38]. Algal amorphous material was easier to decompose than structured organic matter [45]. This was, in part, a result of the greater surface area of amorphous organic material.

It was reported that the three single factor models proposed were insufficient and that a more robust model requires the integration of the three taking into consideration the interplay among them (Figure 11, [46]).

Figure 11.

Workflow to assess the probability of source rock presence and quality based on primary productivity, preservation potential, and sedimentation rate [46].

4. Petroleum generation, retention, and expulsion

Organic matter in source rocks are composed of extractable organic matter (EOM) - bitumen and insoluble organic matter including oil/gas prone kerogen and inert carbon. Under sufficient thermal stress, petroleum is formed incrementally from the decomposition of kerogen and secondary cracking of generated petroleum molecules. This process can be simulated as a series of parallel first-order reactions following the Arrhenius law. A simple reaction of an initial reactant X with mass x generating a product Y with mass y can be represented by:

$$X\stackrel{k}{\to }Y\text{,}$$

$$\frac{\partial y}{\partial t}=-\frac{\partial x}{\partial t}=kx$$E1

$$k=A{e}^{-E/RT}$$E2

where t is the reaction time, k is the reaction rate, A is the frequency factor, E is the activation energy, and R is the universal gas constant 8.314 J∙K⁻¹∙mol⁻¹.

Laboratory anhydrous and hydrous pyrolysis are used to simulate the processes of natural petroleum generation, retention, and expulsion [23, 47, 48, 49, 50, 51, 52, 53]. Burnham systematically documented integration of kinetics and pyrolysis methods to simulate petroleum generation reactions [54]. As shown in Figure 12, it is observed that (1) Type I kerogen generates petroleum over a narrower oil window to decompose a more uniform composition; (2) Type II-S kerogen enters oil-window earlier with lower reaction activation energies to breakdown weaker bonds; in contrast to (3) Type II and Type III kerogens that react with an extended and elevated range of reaction activation energies, respectively, to breakdown mixed kerogens with more complex structures.

Figure 12.

Comparison of activation energy distributions for hydrocarbon generation of four representative different kerogen types (modified after [55]).

In most cases, source rocks contain mixed kerogens. Compositional kinetics was developed to simulate a series of reactions from mixed types of kerogens to form complex petroleum compositions and the secondary cracking of products [56, 57, 58]. Figure 13 shows an example of petroleum primary generation and secondary cracking reactions. Figure 14 shows how these reactions work in a closed system through modeling calibrated with Vaca Muerta Formation data [56]. It shows that (1) asphaltenes and NSO-bearing polar components are formed in the early oil window at 0.5–0.7%R_o, (2) followed by secondary cracking of these components and continuous cracking of kerogens forming saturated and aromatic hydrocarbons in the main oil window at 0.7–1.3%R_o, by then, asphaltenes, NSO-bearing polar components, and large (C₁₅₊) aromatic compounds are fully cracked; (3) Beyond 1.3%R_o, large (C₁₅₊) saturated and small (C₆-C₁₄) aromatic hydrocarbons start cracking, forming light oils (dominant light saturates) and gas hydrocarbons, (4) until 2%R_o where all liquid components are fully cracked to gas and eventually forming dry gas - methane.

Figure 13.

Schematic reaction mechanism of petroleum primary generation and secondary cracking with 17 species (modified after [56]).

Figure 14.

Simulation of petroleum primary generation and secondary cracking in a closed system using calibrated compositional kinetics based on Vaca Muerta Formation data (adapted from [56]).

Natural petroleum systems in the subsurface are semi-closed systems with not only petroleum generation/cracking reactions, but also retention and expulsion. Kinetics and retention models are incorporated into basin modeling together with other necessary geochemical and geological inputs to simulate and quantify petroleum generation, retention, and expulsion in subsurface [56, 59, 60]. Figure 15 shows an example of petroleum generation, retention, and expulsion of Vaca Muerta petroleum system through time and temperature.

Figure 15.

Basin model showing petroleum generation, retention, and expulsion through time and temperature changes, a Vaca Muerta Formation example (adapted from Mei [56]).

Organic matter and clay minerals in source rock have a high sorption capacity for petroleum [61, 62, 63]. As shown in Figure 15 using the Vaca Muerta Formation as an example, the initially high sorption capacity decreases through petroleum generation and sorption with increasing time and temperature. Until the quantity of generated petroleum exceeds source rock sorption capacity, major petroleum expulsion occurs at about 0.85–1%R_o and 120–140°C. This process associates with increasing pore pressure, permeability, and organic porosity. The sorbed components can be further cracked with elevated temperature over time. When thermal maturity is increased to above 1.3%R_o and 160°C, intensive petroleum cracking creates volume expansion and excess pore pressure, which in turn induces rock fracturing and the second stage of major expulsion. Tectonic uplift decreases pressure and temperature, which temporarily stops petroleum generation and expulsion. Continuous burial can result in further cracking and expulsion.

5. Petroleum migration

Within this study, migration is considered the movement of hydrocarbons within a carrier system once they have been expelled from the source rock. This includes the initial movement to the trap as well as any remigration that may occur following the initial accumulation as a result of tectonic movements or the subsequent addition of hydrocarbons.

Hydrocarbon migration is considered the least understood aspect of the petroleum system. This, in part, is a result of our limited ability to observe migration and that we typically see only the results of migration (i.e., the position of the accumulations [64]). Migration is driven by buoyancy, which is controlled by density differences between the migrating hydrocarbons and pore fluids, largely controlled by brine salinity and API gravity [65]. Hydrocarbon migration can occur laterally, vertically, or a combination of the two.

Lateral migration occurs in stratigraphic proximity to the source, but over significant distances potentially exceeding 100 km, with accumulations developing beyond the limits of the generative kitchen. A single stratigraphic unit may contain multiple accumulations. The flow paths or migration patterns are, in general, controlled by structural patterns, where hydrocarbons may be focused or dispersed as shown in Figure 16. Regions of focus are preferred sites for exploration, while dispersive regions are to be avoided [66]. Flow paths are established at the base of low permeability layers. These flow patterns may change through time as a consequence of structural evolution. Carriers may include permeable beds, fracture networks, and certain unconformity surfaces. Depending on the availability of hydrocarbons an examination of structural patterns may also aid in the identification of migration shadows, as well as opportunities for fill and spill establishing up-dip hydrocarbon charge potential. These migration patterns can be altered by strong water movement and the distribution, character (including variability in permeability), and extent of the carrier beds. Sheet sands provide potentially the longest and least controlled migration patterns, whereas isolated reef bodies such as those of the Michigan Basin provide no continuity and are not effective carriers.

Figure 16.

Structural patterns establish general hydrocarbon migration patterns. Regions of focusing and dispersion are identified.

Vertical migration provides a means of transferring fluids across stratigraphic horizons. Accumulations develop above or near the active source. Stacked reservoirs with a common source exist. Surface seepage is common. Although the lateral movement in such systems can be limited, vertical fluid movement can be quite significant, on the order of several kilometers [67, 68].

There are examples where multiple generative kitchens can focus on a common trap. In some cases, these oils may remain distinct, and in others where the oils may mix. Situations exist where sealing faults are present within a structure and no mixing occurs. Such is the case in the Minas field (Central Sumatra, Indonesia) where distinct oils are present on the two sides of the Main Minus Fault Zone.

Migration may be episodic, potentially as a result of fault movement as in the case of deep-water Nigeria, where unaltered oil is introduced into a shallow reservoir where the oil pool has undergone biodegradation [69] or largely continuous and potentially in near real time such as at Eugene 330 Field in Gulf of Mexico [70].

Remigration or dysmigration may result in the loss of hydrocarbons, the repositioning of the remaining hydrocarbons, and changes in oil character (e.g., phase segregation). Remigration may take place as a result of fault movement or structural inversion.

6. Geochemical inversion and correlation

Integrating geochemical inversion, oil to oil and oil to source rock correlations, basin modeling, and regional geology is important to understanding the petroleum system and significantly reducing the risks of petroleum exploration [71, 72]. For clarification, geochemical inversion entails utilizing diagnostic molecular and isotopic characteristics of petroleum collected from seeps, various types of rock samples, and produced fluids to infer (1) the organic-matter type and thermal maturity of the source rock as well as that of the oil or gas at time of generation [72]; (2) the depositional environment (salinity, redox conditions, and lithology) of the source beds [73]; (3) the age of the likely source rocks [74]; (4) accumulation history; and (5) secondary alteration such as biodegradation [75] and migration after expulsion from the source rocks with anomalous or mixing signatures [76, 77, 78]. In addition, petroleum to source correlation entails comparison of the geochemical markers in source-rock candidates with equivalent markers in the petroleum to better understand oil origin and migration history. Furthermore, basin modeling entails analyzing the geological and thermal settings for a stratigraphic sequence in a basin to understand the burial and thermal histories of the source bed, and to deduce the probable occurrence of petroleum generation, expulsion, and migration relative to reservoir deposition and trap formation.

To understand whether an oil accumulation is charged from the direct contact source rock or migration from deeper or downdip kitchens, it is critical to understand source rock maturity based on maturity indicators in source rock and calibrated basin modeling. Maturity indicators such as vitrinite reflectance (R_o) and spore-color thermal alteration index (TAI) are commonly measured using microscopic technologies. Uncertainties include (1) indicators that are based on terrigenous organic matter that are commonly deposited in fluvial deltaic environments or transported to marginal marine settings in post-Silurian age. However, oils generated from aquatic kerogen (amorphous alginate and exinite) in marine or lacustrine environments or older source rocks contain limited or no higher plant materials such as vitrinites to measure R_o, or spores and pollen to measure TAI. When bitumen exists, bitumen reflectance can be used to estimate vitrinite reflectance [79, 80], although discrepancies in derived vitrinite reflectance are common using different algorithms. (2) Recycled vitrinite may not experience the same thermal history as the primary kerogen. (3) Even with the same thermal history, different types of kerogen may achieve different extents of maturity via different kinetics. Therefore, it is important to develop direct measurements of thermal maturity for aquatic kerogens and correlate them to the thermal maturity standard R_o [81]. Transmission light spectroscopy and Raman spectroscopy show promising results [82].

In addition, to infer oil origin, identify oil families (oil to oil correlation), and correlate oil with possible source rocks, it is important to analyze and interpret the chemical compositions of oil and bitumen in source rock(s) when available. Crude oil and bitumen are complex mixtures of organic compounds consisting of four major group types: saturated hydrocarbons, aromatic hydrocarbons, resins, and asphaltenes (SARA). Among these compounds are numerous trace components such as biomarkers that are organic compounds derived from ancient living organisms (algae, bacteria, and plants), that can provide source-diagnostic information and relatively resistant to alteration.

To analyze biomarkers in oil and rock samples, SARA group-type separation is used to prepare saturated and aromatic fractions of oils for GC-MS analysis. This sample preparation is required to avoid coelution interference and enhance the sensitivity and accuracy of the analytical method [83, 84, 85, 86]. Recent advancements using modern analytical technologies such as GC tandem triple quadrupole mass spectrometry (GC-QQQ-MS/MS) and 2D-Gas-chromatography/time of flight mass spectrometry (GC × GC-TOF) with enhanced analytical resolution enabled simultaneous analysis of diverse trace components in whole oil and minimized volatile loss during sample preparation [87, 88, 89]. Sometimes, many of the biomarkers are absent or occur at much-reduced abundance as a result of alteration. New proxies using alteration-resistant compounds such as diamondoids have been investigated [90, 91, 92]. Diamondoids are saturated hydrocarbons with cage-like (bridged cyclohexane) structures. They are derived from the structural rearrangements of saturated hydrocarbons catalyzed by lewis acids (chemical species with an empty orbital that is capable of accepting an electron pair; commonly associated with clay minerals and thermal cracking). Diamondoids are resistant to many alteration processes, particularly stable at higher maturity, and can be used to indicate advanced thermal maturity and cracking.

The analytical results of relative abundances of chemical compositions such as biomarker ratios are commonly used as geochemical indices. The concept, history, and guidelines for geochemical data interpretations with case studies using global samples were systematically documented in the Biomarker Guide [93]. In brief, alteration, source facies, and maturity are interpreted using multiple intact and diagnostic signatures. These interpretations can then be compared with source rock data in a geologic context. Figure 17 shows examples of geochemical interpretations using GC-MS m/z 191 traces of intact terpanes. Figure 17(a) and (b) are two examples of carbonate sourced oil. Figure 17(a) is a low maturity oil from the Triassic carbonate platform in Sicily with characteristics of (1) abundant extended homohopane H35 equal to or more than H34 indicating anoxic water bottom, (2) presence of abundant gammacerane (G) indicating stratified water column which is commonly associated with hypersalinity from carbonate or evaporative settings, and (3) abundant C24 tetracyclic terpane relative to C26 tricyclic terpane indicating a carbonate or evaporite depositional environment. Figure 17(b) is a high maturity oil from a Late Jurassic to Early Cretaceous carbonate source rock from Guatemala with characteristics of (1) abundant norhopane (H29) relative to hopane (H30) and (2) relatively higher amount of Ts than Ts indicating carbonate depositional environment. In contrast, Figure 17(c) shows a Tertiary oil from the Niger Delta (similar examples see [94]). Specifically, the presence of abundant oleanane (O) indicates terrigenous organic matter inputs in post-Jurassic age and is most associated with Tertiary age, and the stair-step pattern of homohopanes with relatively lower abundance of H35 than H34 indicates a suboxic environment commonly associated with clastic facies. The interpretation above is an example of geochemical inversion. In addition, oil-oil and oil-source correlation are to compare the similarities and differences of geochemical characteristics of oil and source rock. To confirm a genetic relationship, multiple available geochemical characteristics (biomarkers such as terpanes, steranes, and isoprenoids, bulk components such as alkanes and aromatics, and others like elements, isotope ratios, and API gravity, etc.) should be interpreted comprehensively in the context of reasonable geological scenarios.

Figure 17.

Examples of GC-MS m/z 191 traces show different oil biomarker (terpane) distribution patterns and indications for different sources (Triassic carbonate oil data is from [89], Jurassic-cretaceous carbonate oil and tertiary deltaic oil are from Geomark database).

7. Organic porosity

With respect to shale, resource plays geochemistry plays a role in establishing source rock potential as well as reservoir potential. Within unconventional reservoirs three porosity types have been characterized: 1) interparticle pores; 2) intraparticle pores; and 3) organic pores. The importance of pore type varies among the different shale plays [95]. Organic porosity is the porosity that has developed or exists with the kerogen, bitumen, and/or pyrobitumen present within the shale play. Organic porosity within plays such as the Barnett Shale (Forth Worth Basin, Texas) and the Longmaxi Formation (Sichuan Basin, China) provides important storage capacity. These pores, because of their small size (often less than 1 μm), are potentially more important for gas systems [96] although organic pores may play a limited role in some oil plays.

Organic pores may be primary, associated with the kerogen’s initial structure, or secondary, where it is hosted in bitumen or pyrobitumen as a function of generation and alteration. The assignment to primary and secondary pores may be complicated because the visual differentiation between the kerogen and bitumen is not simple. The means to differentiate between kerogen and bitumen pores were proposed in the literature [97, 98].

Organic pores display multiple morphologies, bubbly, spongy, or fracture/crack-based pores as shown in Figure 18 [99]. These forms have different formation mechanisms and associations. For example, the bubbly pore type seems to be largely associated with the oil window and maybe artifacts of water droplets [100]. While cracks and fractures may form through devolitization of solid bitumen [101] or volume changes [102]. The distribution of pores further suggests that the nature of the organic matter, as well as the relationship with the mineral matrix, may play a controlling role. Some of the pores may reflect the initial kerogen character.

Figure 18.

SEM photomicrographs of organic matter from the Kimmeridge clay (United Kingdom): A - bubbly organic pores and B - spongy organic pores [99].

Thermal maturity is one of the key controls on organic porosity. The specifics remain poorly understood and are evolving. Porosity has been observed in immature kerogen (e.g., Eagle Ford Shale [100]), with amorphous kerogen being inherently porous, while the cell structure of vitrinite may also provide primary pores. There is some evidence that as the shales enter the oil-window there is a reduction in observed porosity. This is thought to be a result of the generation of bitumen and oil, which fills pre-existing pore space [103], although there are contradictory data that indicates that porosity may begin to develop within the oil-window [100], not at the onset of generation but at a slightly higher thermal maturity (Ro between 0.8 and 0.9% [104, 105]). It was suggested that the pore generation begins with the onset of hydrocarbon generation and increases through the oil-window, with a decreasing rate of organic-pore generation in the gas-window and terminating at 3.5%R_o [106]. These changes in porosity reflect the release of volatiles and the restructuring of the organic matter. It was noted that changes in porosity are not monotonic [107]. It was further suggested that there is an evolution of the porosity type as maturity increases [108]. Pores may coalesce with increasing maturity causing an increase in pore size and complexity. At advanced levels of thermal maturity, there is also some evidence that pores size decreases [109].

In addition to thermal maturity, organic richness has been considered an important controlling factor in the availability of organic pores. In general, a positive correlation appears to exist between organic porosity and carbon content for TOC levels less than 5.5 wt.% [110]. At the higher levels of organic enrichment, the lack of a correlation may be a result of pore collapse facilitated by greater organic matter connectivity and a less-developed mineral framework. It should also be noted that the greater the organic matter network, the greater the potential for interconnectivity within the organic pore network.

The nature of organic matter is also considered a controlling factor in organic porosity. For example, in humic kerogen, there appears to be little organic pore development beyond what was initially present [111]. In contrast, porosity increases can be observed in solid bitumen. It was suggested that the ratio of bitumen to kerogen was a key factor in determining organic porosity [112]. The greater the solid bitumen content the greater the organic porosity.

8. Summary and future work

Since the 1970s, numerous geochemical studies have been conducted around the world, including improvements in analytical methods, the establishment of data interpretation guidelines, analogs for geochemical inversion and correlation, and improvements in fundamental understanding of petroleum generation, retention, expulsion, and migration. Application of geochemical characterization and interpretation plays a significant role in reducing risk in petroleum exploration. Nevertheless, many interpretation ambiguities and uncertainties still exist due to complex and unclear subsurface conditions. As advanced analytical data and greater volumes of data become available, integrating geochemistry, geology, data analytics, and modeling may help to further understand petroleum systems with fewer ambiguities and uncertainties. This integration will establish new concepts, workflows and improve estimates of unknown values in time and space.

Acknowledgments

The authors wish to thank Chevron Corporation for permission to publish this work and thank Jessica Little and Michael Hsieh especially for performing Chevron’s internal review before submission. We also would like to thank GeoMark for permission of using GC-MS traces in their database in this book chapter to show examples of different biomarker patterns in different types of oils.