Available online at www.sciencedirect.

com

Organic Geochemistry
Organic Geochemistry 39 (2008) 11501161 www.elsevier.com/locate/orggeochem

Oilsource rock correlations Limitations and recommendations

J.A. Curiale *
Chevron Energy Technology Company, 14141 Southwest Freeway, Sugar Land, TX 77479, United States Received 21 August 2007; received in revised form 8 January 2008; accepted 3 February 2008 Available online 9 February 2008

Abstract A rigorous, defensible and causal oilsource rock correlation denitively ties an individual source rock sample to an individual crude oil using genetically-based, internally consistent parameter matches. Such a correlation is a relationship established between the two components which is consistent with all known chemical, geochemical and geological information, and meets three criteria: (a) the relationship must be causal the oil must arise (at least in part) from the specied source rock(s); (b) chemical data used in the correlation must be comparable the elemental, molecular and isotopic data derived from the source rock must be of the same type as that derived from the oil; and (c) all available geological data must be supportive clear geological evidence must exist which allows the proposed source rock to have sourced the oil. The three points of this denition are satised, and the correlation is successful, if we solve one analytical and three geological problems. Natural extraction of crude oil from its source rock i.e., expulsion diers mechanically and temporally from laboratory extraction, leading to a correlation problem referred to as extraction dierences, and representing the single greatest analytical uncertainty in correlation eorts. Geologically, three aspects confound oilsource correlations: occurrence of multiple source units and/or source units at diering maturity levels; lateral and vertical depositional variations in source unit(s) organic matter; and lateral and vertical variation in-reservoir unit(s) organic matter. These concerns are overcome by taking ve actions for each oilsource correlation eort. (1) Select representative samples using statistically defensible methods. (2) Establish the inherent compositional variability laterally and temporally due to depositional and maturation processes in each prospective source unit. (3) Assess the extent of migration-induced changes in oil composition, including post-migration changes such as in-reservoir alteration. (4) Support each correlation with migration histories derived from 4d models. (5) Iterate correlation results with new data gathered from ongoing exploration eorts. These ve actions will (a) allow correlation success to be measured more objectively, (b) establish risking parameters for direct use in basin assessment, and (c) provide baseline criteria for use in assessing the reliability of oil source rock correlations. Although these approaches will improve conventional oilsource rock correlation eorts, ultimately they may not be needed if inversion methods become more reliable. The ability to invert compositional data for crude oil i.e., predict the age, lithology, maturity, and other characteristics of a source rock solely through chemical analysis of its expelled oil could eventually result in successful oilsource rock correlations in which source rock analysis for oilsource rock correlation purposes is less critical. 2008 Elsevier Ltd. All rights reserved.

Tel.: +1 281 287 5646; fax: +1 281 276 9351. E-mail address: jcuriale@chevron.com

0146-6380/$ - see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.orggeochem.2008.02.001

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

1151

1. Introduction Correlation of a crude oil to one or more source rocks is a common industrial application of petroleum geochemistry. Conrmation that oil has been generated in the target sedimentary basin is the most critical piece of knowledge a petroleum explorationist can derive; second in importance is the determination of the source(s) of that oil. For this reason, an extensive arsenal of analytical methods is utilized to collect primary data on the organic matter in crude oils and possible source rocks, and various components of these data are used to relate oils causally to their prospective sources. Oilsource rock correlations at various condence levels have been established for the petroleum systems of all major sedimentary basins. Published case studies establishing causal correlations between oil and source rock pairs began to appear in the third quarter of the 20th century (e.g., Hunt et al., 1954; Welte, 1965; Dow, 1974; Williams, 1974) and summaries of oilsource rock correlation studies and methods are now available (Curiale, 1993, 1994a; Waples and Curiale, 1999). Although correlations are fundamentally subjective exercises, the results are rarely challenged either conceptually or on their merits. Numerous published examples correlating one or two oil samples to one or two source rocks purport to establish relationships which are commonly carried basinwide. Other examples utilize a decidedly limited suite of geochemical analytical types and data, often establishing strong chemical correlations but no convincing causality. Still others attempt to enforce a chemical correlation on sample sets whose geologic circumstances make such a correlation quite unlikely. Concerns of this type have been expressed since oilsource correlations were rst attempted, although often only as passing remarks in published literature and industry reports. Interestingly, although these concerns have been duly noted by various authors, most of these authors proceeded with their correlation eorts anyway, without any serious eort to address the noted concerns. This has led to unnecessary uncertainty in published results, and the impression that any correlation is better than no correlation. My objectives here are to gather into a single publication the reasons for the subjectivity of oilsource correlations and to recommend conceptual and practical improvements for the use of these correlations in solving industrial

problems. In addition, I present some criteria for judging the success of practical oilsource rock correlations, in the hope that these criteria will assist in dening the risk associated with the charge component of exploration eorts in both frontier and mature basins. 1.1. Requirements of an oilsource rock correlation Any successful oilsource rock correlation must include three attributes: (a) requirement of causality; (b) comparable chemical data for all samples; and (c) geological support. For the purposes of this paper, I will use the denition of Curiale (1993), derived from previous authors and cited in that publication as: an oilsource rock correlation is a causal relationship, established between a crude oil and an oil-prone petroleum source rock, which is consistent with all known chemical, geochemical and geological information. The three key points of this denition are listed below.  The relationship must be causal. That is, the oil must arise (at least in part) from the specied source rock.  Chemical data used in the correlation must be comparable. That is, the elemental, molecular and isotopic data derived from the source rock must be of the same type as that derived from the oil.  All available geological data must be supportive. That is, clear geological evidence must exist which allows the proposed source rock to have sourced the oil. The absence of any of these three key points necessarily negates the validity of a proposed oilsource rock correlation. That is: the presence of all three points is required, at a minimum, before declaring a correlation to be successful. The importance of both the chemical and geological character of these three denitional points cannot be overemphasized. Establishing chemical similarities between the organic matter in a source rock and that in an oil, even if these similarities involve genetic (i.e., source-derived) molecular and isotopic characteristics, is necessary but insucient. Such a result must also be supplemented by supporting geological data establishing that the source was capable in all spatial and temporal dimensions of having generated a specic oil. These geological data, including the details of depo-

1152

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

sitional history and structural conguration through time, are provided as input to a robust basin model which is used to support the correlation conclusion in the spatial (i.e., uid ow conguration) and temporal (i.e., timing of generation and expulsion) dimensions. Only when this is conrmed can a bona de oilsource rock correlation be concluded with condence. 1.2. Previous approaches Several published, successful oilsource rock correlations have appeared since the rst was presented by John Hunt and colleagues over 50 years ago. Hunt et al. (1954) provided clear chemical and geological evidence in a causal framework for the source of oils and solid bitumens of the Uinta Basin (USA). Two decades passed before publication of the seminal work by Williams (1974) and Dow (1974) on the Williston Basin (USA), in which elemental, isotopic and molecular data were used in successful correlations among three source rocks and oil families. Although later studies became progressively more analytically sophisticated, the concepts established by Hunt, Williams and Dow, from which the denition given above was derived, still dominate oilsource rock correlation approaches. All analytical methods used by petroleum geochemists to characterize oils and the organic matter in source rocks have also been used to correlate oils to source rocks, and the condence we have in oil source correlations parallels the condence we have in petroleum geochemical methods in general. Indeed, the ease and rapidity with which precise analytical data can now be obtained has tended to deemphasize the importance of geological data, because the precision of chemical data tends to be far greater than that of geological data. In part, this is because rapid temporal changes in the depositional conditions of many sedimentary systems make it dicult to sample these systems representatively. Thus, the fundamental non-analytical obstacle to successful oilsource rock correlations is the natural variability of the geological system, and much of this paper focuses on coping with this variability when attempting oilsource rock correlations. 1.3. The problem Our knowledge of variability within depositional systems at all scales has not kept pace with our abil-

ity to generate highly precise analytical data. Indeed, chemical correlations are now easy to establish almost automatic when using some statistical approaches whereas knowledge of natural, nescale variation in source rock deposition (e.g., Curiale, 1994b; Keller and Macquaker, 2001; Barker et al., 2001) is substantially more dicult to acquire. Therefore, although a few aspects of analytical uncertainty remain (mostly based on the type of rock extraction method used see below), the success of an oilsource rock correlation depends largely on our understanding of sedimentary organic variability in the natural setting. This problem making certain that the natural variability of depositional processes is fully considered when establishing oilsource rock correlations requires that lateral and vertical dierences in organic matter distribution in both source and reservoir rocks are accounted for properly. This variability extends from organic dierences in source shales at the laminae scale to the potential for multiple source units to drain toward a single focal point (trap). The remainder of this paper discusses these factors using examples. Individual sections briey address correlation problems arising from the occurrence of extraction technique dierences, multiple source units, lateral/vertical source organic variability and lateral/vertical reservoir organic variability. A closing section proposes recommendations for more accurate and consistent correlations, in an eort to minimize the uncertainty in oilsource correlations. In this manner, more accurate correlations can be used in an industrial context, thus reducing exploration risk. 2. Extraction dierences Most correlation techniques in common use involve molecular and isotopic comparison of components in crude oil with components extracted from a candidate source rock. Natural extraction of crude oil from its source rock usually denoted as expulsion diers chemically, mechanically and temporally from laboratory extraction, leading to a correlation problem referred to here as extraction dierence. Laboratory extraction usually utilizes an organic solvent at its boiling point, occurs over a convenient (i.e., short) time period in a manner that is relatively well controlled and understood (Eglinton and Murphy, 1969), and usually concludes with a time-ecient method of removing most of the unwanted solvent. In contrast, natural

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

1153

extraction (expulsion) of oil from a source rock occurs at greatly dierent timescales, via mechanisms that are much less well understood. Nevertheless, the standard approach of the petroleum geochemist is to relate molecular and isotopic characteristics of an oil with those in organic matter that has been solvent-extracted from a source rock, while making an eort (either implicit or explicit) not to utilize components thought to be dierentially aected by these two modes of extraction (e.g., the lightest components which, although often the bulk of the soluble organic matter in a source rock, are usually lost during sampling or laboratory workup extraction procedures). This is clearly a fundamental analytical constraint in traditional oilsource rock correlation approaches. Although some workers have attempted to circumvent this concern by using extraction uids thought to be more relevant to those present in the subsurface and taking special care to retain native components during workup (e.g., through extraction with supercritical carbon dioxide; Li et al., 1996; Koel et al., 2000), these laboratory approaches still suer from a lack of similarity to natural expulsion. Indeed, most correlations continue to use standard solvent extraction techniques, leading to an unwarranted correlational focus on tetra- and pentacyclic hydrocarbons because of their favorable workup retention characteristics. As a result, the approach represents the single greatest analytical uncertainty in correlation eorts, and will likely remain so. Because the extent and nature of fractionation dierences that occur between solvent extraction in the laboratory and natural extraction during oil expulsion are well known and extensive at the compound class level (e.g., Leythaeuser et al., 1984), a few workers have sought alternatives. A straightforward approach to more closely mimicking natural generation and expulsion is to apply pressure and/ or temperature to prospective source intervals in the laboratory, thereby articially generating an oil for comparison with natural crude oil. Early eorts in this regard involved mechanically compacting organic-rich recent sediments, which led to expulsion of hydrocarbon-rich uids subsequently analyzed and shown to be oil-like (e.g., Shinn et al., 1984). Because most petroleum source units are already lithied, these mechanical compaction eorts involving recent sediment have been replaced by various types of conned and unconned heating in an oxygen-free atmosphere. Of these pyrolytic approaches, conned heating in the presence of

excess water commonly yields the most oil-like expelled uid (Lewan, 1985). Nevertheless, detailed molecular dierences between natural oil and pyrolyzate are still observed, limiting the approach to some extent. Furthermore, the time- and laborintensive nature of the hydrous pyrolysis approach limits its use to non-routine applications, leaving us with continued concerns about extraction dierences when attempting oilsource rock correlations. 3. Mixing from multiple source units and multiple maturity levels Increasing use among petroleum geologists and geochemists of thermal and ow modeling at the basin scale has resulted in an increased awareness of the occurrence and importance of multiple source rock units in a sedimentary basin, and the postsource mixing of oil expelled from these units en route to the trap. In addition, maturity-induced compositional variations in crude oil and in the soluble organic matter remaining in a source rock after explusion can be signicant, and eorts to correlate a crude oil with the residual soluble organic matter in its source rock(s) represent an inherent and fundamental uncertainty in oilsource rock correlation. It is now recognized that few commercial oil and gas accumulations are singly sourced, and even fewer are sourced from a single unit at a single maturity level. Examples of these sorts of mixing range from the classic work of Seifert et al. (1979) on the oils of the Prudhoe Bay Field (Alaska) to a large body of work published over the past two decades for sedimentary basins worldwide (e.g., Peters et al., 1989; Dzou et al., 1999; Chen et al., 2003a,b and references therein). Although the correlation problems associated with maturity dierences can sometimes be alleviated through judicious choice of rock sampling, oils mixed from sources of diering ages usually at dierent maturity as well confound the problem signicantly. Chen et al. (2003a,b), in their extensive study of the triply-sourced oils of the Cainan Field of Chinas Junggar Basin, have established multiple source unit input by using molecular and isotopic mixing evidence as well as by creating articial mixtures of inputs from Jurassic, Triassic and Permian source rocks. Potential variations in contributions from these three sources (Chen et al., 2003b) could result in carbon isotope ratio dierences of greater than 4, pristane/phytane ratios varying by a factor of two, and relative concentrations of b-carotane vary-

1154

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

ing by more than an order of magnitude (Fig. 1). Chen et al. (2003a,b) clearly demonstrate that all three sources are required as contributors to the oil in the Cainan Field, and show the wide range of single-eld oil compositions which can be anticipated when multiple sources are involved. The contribution of expelled products from two or more source units within a single sedimentary fetch area presents a fundamental problem for any oilsource correlation attempt. Accurate correlation of oil contained within a single trap (it is usually assumed that the oil is uniform throughout the trapped volume) and derived from multiple sources requires detailed compositional information for each of the sources, as well as information about lateral and vertical compositional variability in each source due to depositional (the importance of this item is discussed below) and maturity dierences. Furthermore, the extent of mixing (as well as the time and place of mixing) will aect the composition of the resultant oil. Because most of these aspects are either unknown or poorly known, oilsource correlations involving multiple source units and/or maturity levels possess increased risk. This risk can be alleviated to some extent by examining the full molecular weight spectrum of an oil (rather than, say, just the tetracyclic and pentacyclic bio-

marker components). The recent use of multiple molecular fractions has yielded more robust correlations in several instances (Rooney et al., 1998; Waples and Curiale, 1999; Obermayer et al., 2000). Oilsource correlation concerns arising from multiple source inputs are exacerbated even further when widely varying component concentrations are used to support multiple sourcing, as exemplied by the question of source components for the oils of Beatrice Field, oshore Scotland. The Jurassic-reservoired oils of this eld contain b-carotanes, which are used by both Peters et al. (1989) and Bailey et al. (1990) to support a correlation to a Devonian lacustrine source unit. In addition, however, the presence of low concentrations of n-propyl cholestanes are used by Peters et al. (1989) to infer a secondary, Jurassic source input. This co-sourcing conclusion is refuted by Bailey et al. (1990), who state that as a matter of philosophy, we would suggest that it is dangerous to base hypotheses upon components which occur in petroleum only in trace amounts (cf. Curiale, 1994a). Although Peters et al. (1999) cite additional, supporting information for a Jurassic co-source, the multiple-sourcing aspect of oil source correlation eorts will continually raise the question: What molecular concentrations should we require in an oil before we must seek those com-

Fig. 1. Variation in b-carotane/n-C25 hydrocarbon ratio for oils of the Cainan Field, Junggar Basin (NW China; see inset) and articial mixtures of Permian-sourced, Triassic-sourced and Jurassic-sourced oils. The x-axis shows the percentage of Jurassic-sourced oil in the oils of the eld and of the mixtures. The hatchured rectangle shows the range in b-carotane/n-C25 ratio for the Cainan Field oils, and the horizontal line within this rectangle shows the average of the ratio for all Cainan oils. The curves sloping from top left to bottom right show the variability in b-carotane/n-C25 ratio as Permian oil (top curve) or Triassic oil (bottom curve) is added to Jurassic oil. Displays like this are useful in estimating source inputs to multiply sourced oils. This gure is modied after gures in Chen et al. (2003a,b).

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

1155

ponents in a source? For example, are biomarkerbased correlations reliable when a biomarker-rich uid (e.g., a low-maturity oil) mixes with a biomarker-poor uid (e.g., a thermal condensate)? As highly selective and sensitive mass spectrometric detection methods have become more commonplace, the answer to this question becomes more critical, and represents a key concern when attempting oilsource rock correlations in basins containing multiple source units or in source units at multiple maturity levels. 4. Lateral and vertical organic variability in source deposition Inspection of modern depositional settings indicates that organic matter accumulation in negrained sediments can show extensive composialiphatic gas chromatogram n-C27 n-C25 n-C31 n-C29

tional variability, both laterally and temporally (Tyson, 1995). As modern sediments become part of the lithied stratigraphic section, this variability is often preserved as compositional variations at the bulk (organic petrographic) and molecular levels. Lithologically homogenous sections often show molecular dierences over relatively small vertical extents. An example of this is shown in Fig. 2. Two Oligocene rock extracts from northern Alaska, sampled as sidewall cores and separately vertically by less than 100 m (and thus at essentially identical thermal maturity level), are very similar in bulk composition TOC = 1.16% and 1.26%; d13Caliph = 28.4 and 28.6. Yet, these samples show measurable and signicant dierences in acyclic and cyclic biomarker distributions (see chromatograms in Fig. 2, as well as additional details in the gure caption).
m/z 410 mass chromatogram

12, 18, 13(18) -oleanenes

Fig. 2. Molecular distributions in two sidewall core rock extracts from northern Alaska. Both samples are early Oligocene, of identical lithology, and separated by less than 100 m in a lithologically-constant stratigraphic section. A portion of the aliphatic gas chromatogram is shown on the left (shallower sample at the top); prominent odd-carbon number n-alkanes are annotated, with even-carbon number nalkanes eluting in between (not annotated). Note the dierences in the distribution of the n-alkanes, including the greater odd-carbon nalkane dominance in the deeper sample. Note also the obvious dierences in non-n-alkane distribution between the two samples. A portion of the m/z 410 mass chromatogram (GC-MSD data) for each sample is shown at the right; a cluster of olenic angiosperm-derived components, including oleanenes with unsaturation at the 12, 18 and 13(18) positions, is annotated. Note the diering distribution of oleanenes in the two chromatograms, despite constant lithology and maturity. Note also the apparent occurrence of numerous additional, minor (angiospermous?) components in the deeper (lower) sample.

1156

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

Although the example in Fig. 2 shows variability at the 100 m scale, studies at much higher sampling resolution also consistently reveal vertical organic variability, often at the centimeter scale (e.g., Fig. 3), that is as great as that observed at the meter and kilometer scales (Huc et al., 1992; Aplin et al., 1992; Curiale, 1994b; Burwood et al., 1995; Jones et al., 1997; Nijenhuis et al., 1999; Keller and Macquaker, 2001; Barker et al., 2001). Furthermore, the extent of this variability can change signicantly from one depositional setting to another. For example, deltaic source rocks containing Type II/III or III kerogen show much greater variability, both laterally and vertically, than freshwater lacustrine source rocks containing Type I kerogen (cf. Gibling et al., 1985; Lin et al., 2005). An example of this is the Serravallian (mid-Miocene) section of coastal California, where 1 m and 10 m intervals of the Monterey Formation show elemental, molecular and isotopic variability equivalent to that of the entire vertical succession of the Monterey Formation (Curiale, unpublished data). As inferred earlier, the composition of crude oil may be envisioned as a time-averaged product expelled from all candidate sources within a single

fetch area. Just as multiple source units can confound oilsource rock correlations, temporally and laterally rapid internal dierences in the organic facies of each source unit (that is, the lateral and vertical changes in kerogen type within a single source unit) will do the same. The temporal scale of these changes can be decadal or even annual; lateral changes can occur over as little as meters or tens of meters. As a result, the integrated product of expulsion from these various facies i.e., the oil expelled from the source unit is likely to be compositionally unlike that expelled from any individual sample from within the source sequence. An example of this is shown in Fig. 4, where the isotopic distribution of n-alkanes in a Camamu Basin (Brazil) oil is unlike that of any sample of its purported source unit, but nearly identical to that of a weighted average of four prospective samples from the source unit (Curiale and Sperry, 2000). As ner and ner analytical methods are utilized (in this case, compound specic stable isotope analysis for carbon), the observation of mismatches between individual rock samples and crude oils are likely to become more commonplace. 5. Organic variability in carrier bed and reservoir rock Although organic facies variations in oil-prone source units provide a rst-order control on the composition of expelled oil, so-called non-genetic or post-sourcing events can also change an oils composition. In-trap biodegradation, for example, results in several well-recognized compositional changes in crude oil (Larter et al., 2006, and references therein) which, if unrecognized, can result in a failed oilsource correlation eort. Less well-recognized compositional changes that often go completely unnoticed can occur as a result of exchange between oil and carrier/reservoir rock at the molecular level (Philp and Gilbert, 1982; Hughes and Dzou, 1995; Curiale, 2002). Pang et al. (2003) have shown that less than 510% of a molecular contribution from the migration pathway or reservoir can alter biomarker compositions signicantly, a phenomenon which would easily lead to erroneous oilsource rock correlations. This migrationcontamination process is easily interpretable for oils which have migrated into a thermally immature part of the stratigraphic section (e.g., Fig. 5), and is seen at the extreme when recent organic matter is mixed with migrated crude oil at the seawatersediment

Fig. 3. Total organic carbon variability in a 40 cm section covering approximately 40,000 yr of deposition in the Early Albian of southeastern United States (modied after Barker et al., 2001). Note the large (more than an order of magnitude) and rapid variation in organic carbon preservation. Large variations are also observed in hydrogen index (99790 mg/g) and oxygen index (33394 mg/g) in this 40 cm section.

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

-27

1157

dominant lacustrinesourced oil in the basin

-28

-29

Carbon Isotope Ratio (o/oo)

-30

-31

-32

-33

composite of all 4 source rock samples

2050 m 2210 m

-34

2400 m 2610 m

-35

Dashed Lines: 4 source rock samples from an 800m lacustrine section

-36 20 21 22 23 24 25 26 27 28

N-Alkane Carbon Number

Fig. 4. Distribution of carbon isotope ratio values for n-C20 through n-C28 in four lacustrine source rock samples (dashed curves; sample depths are shown in the inset), one composite of data from these four samples (solid curve with circle symbols), and a lacustrine-sourced crude oil (solid curve with diamond symbols). All samples are from the Camamu Basin, oshore Brazil. Additional details are in the accompanying text. The gure is modied extensively after a similar gure in Curiale and Sperry (2000). Note that although no single source rock extract correlates well with the oil, the composite curve of all four source samples is an excellent match to the oil.

interface. In contrast, the phenomenon will often be entirely unrecognizable when oils reside in a mature part of the stratigraphic section. Adding to the problem of deconvoluting the eect of migrationcontamination on a reservoired oil is the likelihood that this molecular exchange will vary in type and extent as a function of lateral and temporal dierences in the syndepositional organic content of the reservoir rock. If these dierences are signicant, and occur at high frequency in the rock, then oils tested from individual portions of the same reservoir in the same eld can appear compositionally dierent from one another (Hughes and Dzou, 1995; Curiale and Bromley, 1996a). The success of oilsource rock correlations in such a setting can deteriorate even further in situations where phase changes during migration tend to fractionate the product compositionally (Gussow, 1954; Thompson, 1988; Curiale and Bromley, 1996b); an example is presented in Fig. 6. These migration-

related eects are additive through time and space, and provide a complicated scenario: indeed, migration-induced eects have the potential to alter an oils composition to such an extent that an accurate oilsource rock correlation might be impossible (Curiale, 2002). 6. Recommended approach The previous sections have provided examples of four situations where oilsource rock correlations can be compromised: where laboratory rock extraction techniques dier dramatically from natural oil expulsion; where multiple source units and maturity levels occur in the basin, resulting in oil mixing within the fetch area; where lateral and temporal organic variability within a source unit (i.e., organic facies dierence) is substantial; and where migration-induced compositional changes in an oil are appreciable and vary at small reservoir scales. These

1158

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

A (oil)

complex to unravel completely. However, specic approaches can be used to minimize associated problems and provide a framework for reducing exploration risk in both frontier and mature basins. These recommendations are summarized below. 1. Select representative samples. a. Analyze at least one oil sample from each fetch volume in the study area, and establish viable oiloil correlations and family groupings. If multiple source units are anticipated in a single fetch volume, the number of oil samples in this volume should at least equal the number of mature source units in the volume. b. Analyze at least one rock sample from each organic facies in each source unit, preferably at a maturity level corresponding to a transformation ratio between 10% and 50% (note: all rock samples must be pre-determined to be source rocks, using standard criteria). 2. Establish the minimum necessary source rock vertical sampling resolution. a. Analyze sucient samples within each source unit to conrm that temporal variability causes vertical source compositional dierences which are less than the compositional dierences observed between any two oil families in the study area. 3. Assess the extent of migration-induced changes. a. If reservoirs occur in thermally immature sections, analyze the reservoired oils to conrm that immature biomarkers (e.g., sterenes, hopenes; ring-unsaturated terrigenous triterpenoids) are absent or in low concentration. b. If long-distance vertical migration pathways are evident (e.g., growth faults), use physical parameters of the uids (e.g., API, GOR) to establish that oil families are, indeed, source-dened, and not inuenced by migrationfractionation. 4. Support all oilsource correlations with 4d modeling. a. Construct a 4d model of the study area to conrm that reasonable migration pathways connect, through both space and time, each proposed source with each of its correlated oil samples. 5. Iterate correlation results with ongoing exploration results.

B (coal extract)

20R

20S

Fig. 5. m/z 217 mass chromatograms of (A) Cook Inlet (Alaska) Sunsh-1 oil, and (B) a solvent-extract of a coal penetrated by the Sunsh-1 well. The prominent peaks in the coal extract are 20S5a,14a,17a-ethylcholestane (early eluting) and 20R-5a,14a,17aethylcholestane, and the low 20S/20R epimer ratio suggests thermal immaturity. This low ratio has been overprinted on the sterane distribution of the Sunsh-1 oil, yielding a 20S/20R epimer ratio which implies a maturity level below that conventionally considered necessary to generated crude oil (gure modied after Hughes and Dzou, 1995). The modied low ratio in this oil is representative of many molecular changes which can result from migrationcontamination, and which make correlation of the oil to its source rock particularly dicult.

Fig. 6. Light-end bias of n-alkanes in oils and condensates of the Vermilion-39 eld, oshore Louisiana, USA, relative to reservoir depth. All samples shown are source-invariant, and were expelled at similar maturity. Shallow oils are progressively enriched in lighter n-alkanes due to migrationfractionation. Post-sourcing compositional changes such as these, if unrecognized, can compromise oilsource rock correlations. Figure modied from (and full details are present in) Curiale and Bromley (1996b); four outliers, shown in the original paper, are not shown here.

diculties are inherent in the complexity of the natural system, and will probably always remain too

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161

1159

a. Maintain the 4d model and the oil/rock databases live, integrating each additional data point and iterating the oilsource correlation conclusions and the model accordingly. These ve recommendations represent optimal conditions for an oilsource rock correlation attempt. Of course, in practice such ideal conditions are rarely achieved. However, setting these recommendations as objectives provides a baseline level for assessing the reliability which decision makers should place in oilsource rock correlations. The lack of oil samples from a specic fetch volume; the lack of source rock samples from within the oil window; the absence of knowledge about vertical facies changes or migration-induced changes all of these situations compromise the success of an oil source rock correlation, and must be factored into the exploration risk assigned to a specic basin or region. In summary, then, the success of an oilsource rock correlation will be proportional to our success in achieving the objectives framed in these ve recommendations. To the extent that we can build robust datasets which contain proper sample types, sucient sampling density, and geological understanding, and to the extent that we support our conclusions with holistic models, oilsource rock correlations will be successful. More importantly, because achieving these objectives will make each correlation defensible, we will successfully reduce the exploration risk in both frontier and mature settings. 7. Parting comment oilsource rock correlations using only oil Traditional oilsource rock correlations, as discussed throughout this paper, establish compositional and geological comparisons between oil samples and rock samples. However, increased understanding of petroleum systems and, in particular, detailed knowledge of the molecular and isotopic composition of crude oils and the organic matter in their source rocks has led to a non-traditional approach to oilsource rock correlation which I will refer to here as the inversion approach. Inversion involves the use of detailed compositional information about a crude oil to infer the characteristics of its source(s). Inversion methods are already in regular use to deduce, for example,

source rock maturity at the time of expulsion, or in-reservoir biodegradation level, and similar techniques for successfully inverting an oils composition to assess its source character will undoubtedly proliferate. Source rock age, environment of deposition and lithology are now routinely predicted from oil composition, based upon pre-established basinal and regional source rock learning sets. The reliability of this oil-only approach currently can be, and should be, subject to serious critical appraisal. Nevertheless, progressively improving datasets and incorporation of geological information into basin models will gradually reduce the error inherent in this predictive approach. In summary, continued understanding of the eect of source rock characteristics on the molecular and isotopic composition of crude oil has the potential to minimize the need for source rock analysis as a correlational approach, and it is likely that the use of inversion methods will increase signicantly in the years to come. Acknowledgements Numerous Unocal and Chevron co-workers have provided ideas and correlation examples which have aected my thinking about oilsource rock correlations, and their help and advice are appreciated. My thanks go to both of these organizations for providing the regional petroleum systems overviews and extensive databases and analytical capabilities necessary to evaluate petroleum source rocks and their associated oils and gases in detail, and for providing approval to present and publish this work. I also appreciate comments from R. Patience, B. Katz, an anonymous reviewer, and especially from M. Fowler, on an early version of the manuscript.
Associate EditorErdem Idiz

References
Aplin, A.C., Bishop, A.N., Clayton, C.J., Kearsley, A.T., Mossman, J.R. Patience, R.L., Rees, A.W.G., Rowland, S.J., 1992. A lamina-scale geochemical and sedimentological study of sediments from the Peru Margin (Site 680, ODP Leg 112). In: Summerhays, C.P., Prell, W.L., Emeis, K.C. (Eds.), Upwelling Systems: Evolution since the Early Miocene. Geological Society Special Publication No. 64, pp. 131149. Bailey, N.J., Burwood, R., Harriman, G.E., 1990. Application of pyrolysate carbon isotope and biomarker technology to organofacies denition and oil correlation problems in North Sea basins. Organic Geochemistry 16, 11571172.

1160

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161 Gibling, M.R., Haraluck, M., Tantisukrit, C., Thanasuthipitak, T., Uttamo, W., 1985. Oil shale sedimentology and geochemistry in Cenozoic Mae Sot Basin, Thailand. American Association of Petroleum Geologists Bulletin 69, 767780. Gussow, W.C., 1954. Dierential entrapment of oil and gas a fundamental principle. American Association of Petroleum Geologists Bulletin 38, 816853. Huc, A.Y., Lallier-Verges, E., Bertrand, P., Carpentier, B., Hollander, D.J., 1992. Organic matter response to change of depositional environment in Kimmeridgian shales, Dorset, UK. In: Organic Matter: Productivity, Accumulation, and Preservation in Recent and Ancient Sediments. Columbia University Press, New York, pp. 469486. Hughes, W.B., Dzou, L.I.P., 1995. Reservoir overprinting of crude oils. Organic Geochemistry 23, 905914. Hunt, J.M., Stewart, F., Dickey, P.A., 1954. Origin of hydrocarbons of Uinta Basin, Utah. American Association of Petroleum Geologists Bulletin 38, 16711698. Jones, T.P., Fortier, S.M., Mosbrugger, V., Roessler, J., Utescher, T., Ashraf, A.R., 1997. 13C/12C ratio double cyclicity in a Miocene browncoal: isotopic signals and orbital forcing. Terra Nova 9, 1923. Keller, M.A., Macquaker, J.H.S., 2001. High resolution analysis of petroleum source potential and lithofacies of Lower Cretaceous mudstone core (Pebble Shale Unit and GRZ of Hue Shale), Mikkelsen Bay State #1 Well, North Slope Alaska. In: Houseknecht, D.W. (Ed.), NPRA Core Workshop: Petroleum Plays and Systems in the National Petroleum Reserve, Alaska. Society of Economic Paleontologists and Mineralogists, SEPM Core Workshop, vol. 21, pp. 3756. Koel, M., Lyovin, S., Bondar, Y., 2000. Supercritical carbon dioxide extraction of Estonian oil shael. Goryuchiye Slantsy [Oil Shale] 17, 225232. Larter, S.R., Huang, H., Adams, J., Bennett, B., Jokanola, O., Oldenburg, T., Jones, M., Head, I., Riediger, C., Fowler, M., 2006. The controls on the composition of biodegraded oils in the deep subsurface; Part II, Geological controls on subsurface biodegradation uxes and constraints on reservoiruid property prediction. American Association of Petroleum Geologists Bulletin 90, 921938. Lewan, M.D., 1985Evaluation of Petroleum Generation by Hydrous Pyrolysis Experimentation, vol. 315. Philosophical Transactions of the Royal Society, London, pp. 123134. Leythaeuser, D., Mackenzie, A., Schaefer, R.G., Bjoroy, M., 1984. A novel approach for recognition and quantication of hydrocarbon migration eects in shale-sandstone sequences. American Association of Petroleum Geologists Bulletin 68, 196219. Li, H., Li, P., Miao, D., Shen, Z., Zhou, G., 1996. Supercritical carbon dioxide extraction of organic matter from petroleum source rocks and its implications. Chinese Journal of Geochemistry 15, 324330. Lin, R., Saller, A., Dunham, J., Teas, P., Kacewicz, M., Curiale, J.A., Decker, J., 2005. Source, generation, migration and critical controls on oil vs. gas in the deepwater Kutei petroleum system. In: Proceedings 30th Annual Convention & Exhibition of the Indonesian Petroleum Association, pp. 447466. Nijenhuis, I.A., Bosch, H.-J., Sinninghe Damste, J.S., Brumsack, H.-J., De Lange, G.J., 1999. Organic matter and trace element rich sapropels and black shales: a geochemical comparison. Earth and Planetary Science Letters 169, 277290.

Barker, C.E., Pawlewicz, M., Cobabe, E.A., 2001. Deposition of sedimentary organic matter in blacks shale facies indicated by the geochemistry and petrography of high-resolution samples, Blake Nose, western North Atlantic. In: Western North Atlantic Palaeogene and Cretaceous Palaeoceanography. Geological Society, London, Special Publication 183, pp. 4972. Burwood, R., De Witte, S.M., Mycke, B., Paulet, J., 1995. Petroleum geochemical characterization of the Lower Congo Coastal Basin Bucomazi formation. In: Petroleum Source Rocks. Springer-Verlag, New York, pp. 235 263. Chen, Jianping, Liang, Digang, Wang, Xulong, Zhong, Ningning, Song, Fuqing, Deng, Chunping, Shi, Xinpu, Tao, Jin, Xiang, Shuzheng, 2003a. Mixed oils derived from multiple source rocks in the Cainan oileld, Junggar Basin, Northwest China; Part I: genetic potential of source rocks, features of biomarkers and oil sources of typical crude oils. Organic Geochemistry 34, 889909. Chen, Jianping, Deng, Chunping, Digang, Liang, Wang, Xulong, Zhong, Ningning, Song, Fuqing, Shi, Xinpu, Tao, Jin, Xiang, Shuzheng, 2003b. Mixed oils derived from multiple source rocks in the Cainan oileld, Junggar Basin, Northwest China; Part II, Articial mixing experiments on typical crude oils and quantitative oilsource correlation. Organic Geochemistry 34, 911930. Curiale, J.A., 1993. Oil to source rock correlation concepts and case studies. In: Engel, M., Macko, S. (Eds.), Organic Geochemistry, Topics in Geobiology, vol. 11. Plenum Press, New York, pp. 473490. Curiale J.A., 1994a. Correlation of oils and source rocks a conceptual and historical perspective. In: Magoon, L.B., Dow, W.G. (Eds.), The Petroleum System From Source to Trap, vol. 60, American Association of Petroleum Geologists Memoir, pp. 251260. Curiale, J.A., 1994b. High-resolution organic record of Bridge Creek deposition, northwest New Mexico. Organic Geochemistry 21, 489507. Curiale, J.A., 2002. A review of the occurrences and causes of migrationcontamination in crude oil. Organic Geochemistry 33, 13891400. Curiale, J.A., Bromley, B.W., 1996a. The Migration of petroleum into Vermilion 14 Field, Gulf Coast, USA molecular evidence. Organic Geochemistry 24, 563579. Curiale, J.A., Bromley, B.W., 1996b. Migration-induced compositional changes in oils and condensates of a single eld. Organic Geochemistry 24, 10971113. Curiale J.A., Sperry, S.W., 2000. An isotope-based oilsource rock correlation in the Camamu Basin, oshore Brazil. In: Proceedings Volume of the 6th Latin American Congress on Organic Geochemistry. Revista Latino Americana de Geoquimica Organica, vol. 4, pp. 5164. Dow, W.G., 1974. Application of oil-correlation and source-rock data to exploration in Williston Basin. American Association of Petroleum Geologists Bulletin 58, 12531262. Dzou, L.I., Holba, A.G., Ramon, J.C., Moldowan, J.M., Zinniker, D., 1999. Application of new diterpane biomarkers to source, biodegradation and mixing eects on Central Llanos Basin oils, Colombia. Organic Geochemistry 30, 515 534. Eglinton, G., Murphy, M.T.J., 1969. Organic Geochemistry: Methods and Results. Springer-Verlag, New York.

J.A. Curiale / Organic Geochemistry 39 (2008) 11501161 Obermayer, M., Osadetz, K.G., Fowler, M.G., Snowdon, L.R., 2000. Light hydrocarbon (gasoline range) parameter renement of biomarker-based oiloil correlation studies: an example from Williston Basin. Organic Geochemistry 31, 959976. Pang, X., Li, M., Li, S., Jin, Z., 2003. Geochemistry of petroleum systems in the Niuzhuang South Slope of Bohai Bay Basin. Part 2: evidence for signicant contribution of mature source rocks to immature oils in the Bamianhe eld. Organic Geochemistry 34, 931950. Peters, K.E., Moldowan, J.M., Driscole, A.R., Demaison, G.J., 1989. Origin of Beatrice oil by co-sourcing from Devonian and middle Jurassic source rocks, Inner Moray Firth, United Kingdom. American Association of Petroleum Geologists Bulletin 73, 454471. Peters, K.E., Clutson, M.J., Robertson, G., 1999. Mixed marine and lacustrine input to an oil-cemented sandstone breccia from Brora, Scotland. Organic Geochemistry 30, 237248. Philp, R.P., Gilbert, T.D., 1982. Unusual distribution of biological markers in Australian crude oil. Nature 299, 245247. Rooney, M.A., Vuletich, A.K., Grith, C.E., 1998. Compoundspecic isotope analysis as a tool for characterizing mixed oils: an example from the West Shetlands area. Organic Geochemistry 29, 241254. Seifert, W.K., Moldowan, J.M., Jones, R.W., 1979. Application of biological marker chemistry to petroleum exploration. In:

1161

Proceedings of the Tenth World Petroleum Congress, Heyden and Son, London, pp. 425440. Shinn, E.A., Robbin, D.M., Claypool, G.E., 1984. Compaction of modern carbonate sediments: implications for generation and expulsion of hydrocarbons. In: Palacas, J.G. (Ed.), Petroleum Geochemistry and Source Rock Potential of Carbonate Rocks American. Association of Petroleum Geologists Studies in Geology #18, Tulsa, Oklahoma, pp. 197 204. Thompson, K.F.M., 1988. Gas-condensate migration and oil fractionation in deltaic systems. Marine and Petroleum Geology 5, 237246. Tyson, R.V., 1995. Sedimentary organic matterOrganic Facies and Palynofacies. Chapman & Hall, London, p. 615. Waples, D.W., Curiale, J.A., 1999. Oiloil and oilsource rock correlations. In: Beaumont, E.A., Foster, N.H. (Eds.), Exploring for Oil and Gas Traps. American Association of Petroleum Geology, Tulsa, pp. 8-18-71 (Chapter 8). Welte, D.H., 1965. Relationship between oil and source rock. American Association of Petroleum Geologists Bulletin 49, 22462268. Williams, J.A., 1974. Characterization of oil types in Williston Basin. American Association of Petroleum Geologists Bulletin 58, 12431252.