Fuel 84 (2005) 159168 www.fuelrst.

com

FTIR and SUVF spectroscopy as an alternative method in reservoir studies. Application to Western Mediterranean oils
A. Permanyera,*, L. Douib, N. Dupuyb, A. Lahcinia, J. Kisterb
a

mica, Petrologia i Prospeccio Geolo ` gica, Mart i Franque ` s, s/n., 08028 Barcelona, Spain Universitat de Barcelona, Facultat de Geologia, Dt. de Geoqu b des Sciences et Techniques de St. Je ro me, Universite dAix-Marseille III, CNRS UMR 6171, Syste ` mes Chimiques Complexes, Faculte ochimie Organique Analytique et Environnement (GOAE), Case 561, 13397 Marseille cedex 20, France Laboratoire de Ge Received 14 April 2004; accepted 23 June 2004 Available online 7 August 2004

Abstract Reservoir geochemistry assessment has traditionally used the gas chromatogram ngerprint method and star diagrams. Recently we tested alternative techniques, such as Fourier transform infra red (FTIR) and synchronous ultra violet uorescence (SUVF) spectroscopy to optimise the evaluation of reservoir continuities, and to characterize the geochemical evolution of oils from individual reservoirs. We used some Western Mediterranean oils to demonstrate that these independent techniques provide results that are in good agreement with each other. GC ngerprints, FTIR and SUVF spectroscopy can describe the oil characteristics and its evolution in the reservoir. We also show that some FTIR parameters can be closely related to the API degree. q 2004 Elsevier Ltd. All rights reserved.
Keywords: Reservoir geochemistry; Fourier transform infra red spectroscopy; Synchronous ultra violet uorescence; Gas chromatography; Mass spectrometry; Western Mediterranean; Offshore Spain

1. Introduction The role of organic geochemistry in the petroleum industry has been overwhelmingly geared towards exploration [1]. It has now become apparent that petroleum geochemistry can play a key role in petroleum exploration appraisal and development [2]. Geochemistry of reservoirs is well documented in the literature with most of the studies focusing on characterization of oils to solve problems, such as vertical and lateral reservoir continuities [37]. Heterogeneities in petroleum populations may be used laterally and vertically for early identication of reservoir compartmentalization. Indeed, structural and functional composition of crude oils may be affected by reservoir

* Corresponding author. Tel.: C34-934021416; fax: C34-934021340. E-mail addresses: albert.permanyer@ub.edu (A. Permanyer), jacky. kister@univ.u-3mrs.fr (J. Kister). 0016-2361/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.fuel.2004.06.027

compartmentalization. Gas chromatography ngerprints of oils are extensively employed to detect changes in composition linked to reservoir connectivity. Star diagrams using selected inter n-alkanes peak ratios identiable on a whole oil chromatogram are currently employed. However, differences in oils cannot always be deduced by employing only chromatographic ngerprints and therefore a new complementary method has been proposed [8,9]. This procedure is based on the Fourier transform infra red (FTIR) and on synchronous ultra violet uorescence (SUVF) spectroscopy. These techniques give reliable information such as total concentration of aromatic compounds or condensation degree of polyaromatic compounds. The main aim of this paper is to demonstrate the capability of the FTIR and SUVF for use as complementary methods to the GC ngerprint method in reservoir geochemistry. The second goal is to develop an alternative method to the API degree technique to determine the maturity degree of oils.

160

A. Permanyer et al. / Fuel 84 (2005) 159168

2. Geological setting and samples Western Mediterranean oil elds are located offshore of the Ebro delta (NE Spain) and were explored and produced mainly between 1970s and 1990s (Fig. 1). Signicant papers

have been published by several authors, who highlighted both geological and geochemical features [1015]. Mesozoic karstied carbonate rocks, as well as Miocene dolomites mainly constitute the reservoirs. The Mesozoic rocks form a horst-graben rift system, where highs make up

n and Rodaballo are the most important elds currently in Fig. 1. Location map of oil elds of the Western Mediterranean. The Casablanca, Boquero production.

A. Permanyer et al. / Fuel 84 (2005) 159168

161

the main reservoir structures sealed by a thick Pliocene to Pleistocene series [1517]. The oils from the Casablanca eld are reservoired in Jurassic karstied carbonates, which n and constitute the major structure in the area. The Boquero Rodaballo elds consist of single Mesozoic highs and are separated from each other and from the Casablanca structure. The Casablanca eld produces at wells 8 and 11, situated in the northern and southern parts of the main structure, n and Rodaballo elds produce at respectively. The Boquero their respective wells. Reservoir top depths increase from Casablanca to Rodaballo elds (25503380 m, respectively). The origin of the studied oils is assumed to be a tertiary source rock, which presumably consists of marine shales deposited in an anoxic environment since the early Miocene [12]. Lateral variations in source rock facies may occur locally and should be considered. 3. Experimental 3.1. Gas chromatography and gas chromatographymass spectrometry Whole oils were analysed by gas chromatography (GC) and chromatograms were compared. The 19 minor

interparafn peak ratios, from C7 to C13 alkanes were selected (Fig. 2) and the peak height ratios were plotted on star diagrams to establish similarities or dissimilarities between the oils. Some classic ratios such as Pr/Ph, Pr/nC17 and Ph/nC18 were also reported and plotted together with the peaks selected before (Fig. 10). Saturated hydrocarbons were also analysed by gas chromatography/mass spectrometry (GC/MS) to identify the signicant oil features and biomarkers. 3.2. FTIR analysis The spectrometer used for the FTIR analysis was a ge -460 coupled to a Nicolet Nic-Plan IR Nicolet Prote microscope. Oil samples were analysed by transmission on a potassium bromide (KBr) thin plate. The lm thickness was about 20 mm. A recording between 4000 and 400 cmK1 was made with the KBr thin plate alone (reference acquisition) and then with the sample. Spectra were recorded with 64 scans and 2 cmK1 resolution. For each sample, preparation and data acquisition were performed three times. All the spectra were normalised for comparison [18]. The main IR bands were determined

Fig. 2. Gas chromatogram ngerprints showing the peaks chosen for constructed star diagrams.

162

A. Permanyer et al. / Fuel 84 (2005) 159168

Fig. 3. FTIR spectrum of a crude oil between 4000 and 400 cmK1. Band areas measured from valley to valley.

thanks to indexes which had been described in earlier studies [19]. These indexes were used to determine and compare the chemical composition of each sample. Assignments of the main IR bands were determined by reference to an earlier work [20]. A deconvolution technique was applied to increase spectral resolution of overlapping infrared bands [18]. The main assignments of the FTIR bands used in this study are described in Fig. 3 and the applied indexes are dened as: P Aliphatic index: (A1460CA1376)/ A P The sum of the area A represents: X A Z A1700 C A1600 C A1460 C A1376 C A1030 C A864 C A814 C A743 C A724 C A2953;2923;2862 ; which is equivalent in Fig. 3 to X A Z A1 C A2 C A3 C A4 C A6 C A8 C A9 C A10 C A11 C A12 C A13 C A14 P Aromaticity index: A1600/ Aaro, (A2/A9CA10CA11 in Fig. 3) Long chains index: A724/(A1460CA1376), (A11/A3CA4 in Fig. 3) Substitution 1 index: A864/A864CA814CA743 Substitution 2 index: A814/A864CA814CA743. The most accurate ratios of these indexes were nally applied for reservoir compartmentalization and oil maturity evaluation. 3.3. Synchronous ultra violet uorescence analyses Fluorescence measurements were carried out using a PerkinElmer LS50B luminescence spectrometer.

The wavelength (Dl) interval between lex and lem was constant and equal to 23 nm. One centimetre thick quartz cells were used. Fluorescence intensity is related to the quantity of aromatic compounds present in the sample. It also depends on various parameters such as temperature, sample concentration and the kind of solvent [21]. The concentration of 10 mg lK1 was chosen to obtain a suitable signal/ noise ratio of the spectrum. The crude oils were dissolved in tetrahydrofurane (THF). The THF solvent and the high dilution used reduced uorescence-quenching effects. The analysis of UV uorescence spectra of standard polycyclic aromatic hydrocarbons, which are present in crude oils, allows us to dene three main regions A1, A2 and A3 (Fig. 4). Each of these spectral regions is characterized by the number of condensed aromatic rings, yielding qualitative information on the nature of the aromatic species present in oils (spectral region from 280 to 580 nm) [22]. The uorescence index (A2/A1 ratio) represents the ratio of the aromatic compounds with three or four rings with respect to the compounds with two rings. The A3 represents aromatic condensation with ve or more aromatic rings.

Fig. 4. Synchronous UV uorescence spectrum of polycyclic aromatic hydrocarbons present in crude oils. Each spectral region is characterized by the number of condensed aromatic rings: A1 (two rings), A2 (three or four rings), A3 (ve rings or more). (THF, tetrahydrofurane solution).

A. Permanyer et al. / Fuel 84 (2005) 159168

163

4. Results and discussion 4.1. Geochemical characteristics Table 1 summarizes the main geochemical parameters along with global data from the analysed oils. All samples have similar characteristics, with some differences (Table 1). Gas chromatograms show a similar distribution of n-alkanes, as well as pristane and phytane isoprenoids, and a minor decrease in nC28C in the Rodaballo oil (Fig. 5). Tricyclic terpanes are similar in the two Casablanca oils n and Rodaballo oils show (Fig. 6). By contrast, the Boquero dissimilarities between the epimers R and S in the C26 and C28 tricyclics, and a clear diminution in the C24 tetracyclic n crude. The C24 tetracyclic is absent in the in the Boquero Rodaballo oil. These particularities could be related to discrete variations in the source rock sedimentary environment, which would be supported by the variation in T s n and Rodaballo (18a (H)-trisnorneohopane) in Boquero with respect to the Casablanca oils, and Ts/Tm inversed ratio n oil (Fig. 6). The C24/C23 tricyclic ratio in the Boquero shows similar values in the Casablanca oils, and higher n and Rodaballo oils, thus suggesting values in the Boquero a predominantly shaly environment (Fig. 7). Pentacyclic terpanes are practically identical in the two n crude shows a Casablanca samples. The Boquero diminution in homohopanes (C31C35) homologues. This diminution is more marked in the Rodaballo oil (quasi absence of homohopanes as well as Ts and Tm)
Table 1 Geochemical data and parameters of the studied oils Sample CASABLANCA-8

(Fig. 6). Tertiary origin for all oils is supported by the presence of oleanane, which is related to angiosperms [23]. The C27, C28, C29 regular sterane distribution (Fig. 8) is very similar for all the samples, and is consistent with the marine origin of source rock. Steranes show a quantitative n and Rodaballo diminution from the Casablanca to Boquero oils (Fig. 9). The isotope signature of the four oils is similar, suggesting that they originated from the same source rock (Table 1). 4.2. GC ngerprints Three groups of oils are identied when peak ratios are plotted in star diagrams (Fig. 10). The two oils from Casablanca compose the rst group, indicating very close ratios, with only one discrete dissimilar ratio (ratio number 11). This result shows that the Casablanca crude presents good homogenization across the eld structure, and behaves as a single reservoir compartment. n and Rodaballo oils form two distinct The Boquero groups. These two oils present star diagrams that differ from each other and from that of the Casablanca oils. This nding clearly shows that the oils belong to different eld structures and must therefore be regarded as independent compartments. The results obtained by GC ngerprints are consistent with the geological structures of the elds in this area. This case constitutes a good example for testing the application of FTIR and UV techniques in reservoir evaluation.

CASABLANCA-11 2555 35.3 20.45 19.30 0.78 0.11 0.07 1.82 0.27 1.85 0.04 0.54 0.88 0.65 0.09 0.50 28.36 26.27 45.38 47.87 49.34 0.92

N BOQUERO 2997 38.7 20.2 18.30 0.80 0.10 0.09 2.18 0.47 2.03 0.10 0.61 0.95 0.73 0.15 1.18 32.44 36.92 30.65 48.81 50.65 0.95

RODABALLO 3388 41.3 19.5 17.75 1.03 0.00 0.00 0.00 1.07 n.d. 0.26 0.84 1.14 0.58 0.23 5.79 27.10 33.19 39.71 58.77 59.41 1.43

Depth of reservoir (m) G2450 8API gravity 32.2 d13C Saturate HC () 20.55 d13C Aromatic HC () 19.65 % Ro equ.Z0.487*(C29aa20S/aa20R)C0.33 0.74 Moretane/hopane (C30) 0.12 Homohopane index 0.07 Gammacerane indexZGamm/C30HOP 1.71 Oleanane indexZOLN/C30 HOP 0.21 C33 22SC22R/C34 22SC22R 1.90 Diahopane/C30 hopane 0.03 C21/C23 Tricyclic terpanes 0.53 C26/C25 Tricyclic terpanes 0.85 C24/C23 Tricyclic terpanes 0.63 2829 Tric/(TricCC2933 Hop) 0.07 P C19C30 Tric/S C29C35 Hop 0.35 % C27 sterane 30.91 % C28 sterane 35.16 % C29 sterane 33.93 C29 20S/(20SC20R) 45.70 C29 bb (20SC20R)/C29 bbCaa (20SC20R) 48.03 C29 aa (20S)/C29 aa (20R) 0.84 20S/20R and bbb/abbCaaaZisomeric ratios from C29 steranes (%) % Ro equ.Zvitrinite reectivity equivalent [24]

164

A. Permanyer et al. / Fuel 84 (2005) 159168

Fig. 5. Gas chromatograms of studied topped oils (C15C).

4.3. FTIR and UV 4.3.1. Aliphatic and aromatic compounds determined from FTIR Fig. 11 shows two infrared spectra obtained from two different oils. Globally the spectra seem to be very close but some small variations can be seen particularly near 1500 cmK1. Fig. 12 shows that Rodaballo is the sample which has the lowest aliphatic rate and the highest number of long chains, n when compared with the three other samples. The Boquero oil has a greater degree of aliphaticity and a lower long chains rate. The two Casablanca oils have the greatest degree of aliphaticity, particularly Casablanca 8. The lowest long chains rate is detected in Casablanca 11. This diagram n does not show a clear differentiation between the Boquero and Casablanca 11 oils.

Fig. 6. GC/MS chromatogram corresponding to terpanes (191 ion).

If we compare the aromaticity vs. FTIR substitution 2 indexes (Fig. 13), the results appear more accurate, with the three groups of oils being better dened. The two n and Casablanca oils form one group, and the Boquero Rodaballo oils each form a group. These groups are

A. Permanyer et al. / Fuel 84 (2005) 159168

165

Fig. 7. Tricyclic terpane diagram showing a discrete variation of source rock deposition environment.

Fig. 8. C27, C28, C29 regular sterane distribution.

Fig. 10. Star diagrams. The Casablanca oils show a closed diagram between n and Rodaballo oils. them, and differ clearly from the Boquero

Fig. 9. Sterane GC/MS chromatograms, ion m/z 217.

Fig. 11. Infrared spectra of two different oils.

166

A. Permanyer et al. / Fuel 84 (2005) 159168

Fig. 12. Aromaticity vs. long chains index diagram.

Fig. 14. Substitution 1 vs. uorescence A2/A1 index diagram. A good discrimination is shown between the three reservoir structures.

consistent with those determined by GC ngerprints (Fig. 10). 4.3.2. Aromatic compounds determined from UV The uorescence indexes are related to a predominant region for each type of condensation. In the studied oils, A3 was very small with the result that the number of compounds with ve or more aromatic rings was considered to be insignicant. In this study, therefore, we present only the results obtained in the A1 and A2 areas. The uorescence index (A2/A1 ratio, Fig. 4) proved to be an accurate parameter to determine reservoir compartments when combined with the FTIR substitution 1 index [9]. In n, Casablanca Fig. 14 the three groups of crude oils, Boquero and Rodaballo can be distinguished very clearly. Casablanca 8 and 11 oils form a very compact group related to the single reservoir structure in accordance with the GC ngerprints (Fig. 10). 4.4. Maturity degree As shown in Table 1, a good correlation exists between API gravity and top reservoir depths.

The results obtained by gas chromatographymass spectrometry suggest that variations in hopanoids and trisnorhopanes may be related to an increase in the maturity degree from the Casablanca to the Rodaballo oils. The increase in maturity was also evidenced by sterane isomerization [24] (Fig. 15A). A good correlation also exists with API gravity (Fig. 15B). The ratio between both aromaticity and aliphaticity indexes shows an evolution path that can be related to the degree of maturity [9]. In Fig. 16 samples are aligned

Fig. 13. Aromaticity vs. substitution 2 index diagram. Three groups of samples can be distinguished.

Fig. 15. (A) Regular sterane isomerization shows increasing maturity from Casablanca to Rodaballo oils. (B) The increasing values of gravity (API8 degree) appear to be related to sterane isomerization.

A. Permanyer et al. / Fuel 84 (2005) 159168

167

following increasing maturation path in accordance with the maturity parameters deduced before. The aromaticity vs. aliphaticity index diagram shows a good correlation between API gravity and sterane isomerization (Fig. 17AC). Consequently, the aromaticity and aliphaticity FTIR indexes constitute good parameters for maturity degree. The most mature sample presents the highest ratios.

5. Conclusion
Fig. 16. Aromaticity vs. aliphaticity index ratio shows a trend related to maturity of oil samples. Direct ratio between maturation degrees expressed as API grades and FTIR parameters.

The analysis of the four Mediterranean oils suggests that the oils originated from a unique marine source rock tertiary in age. Minor lateral changes may be possible in accordance with some small variations in biomarker parameters. Maturity of oils present increasing values from Casablanca-8 to Rodaballo. The FTIR aromaticity and aliphaticity indexes proved to be an excellent maturity parameter, showing a good correlation with API degree and sterane isomerization. In two preliminary studies [8,9] we demonstrate that oil characteristics as well as its evolution in the reservoir can be described in a similar way by three techniques GC, FTIR and SUVF. In this work, the results conrm the potentialities of the FTIR and UV uorescence spectroscopic methods as new tools in the reservoir geochemistry domain. Some relationships between the FTIR indexes appear to be more suitable for determining the compartmentalization of the reservoir. Thus, the relationships between the FTIR substitution 1 index vs. UV uorescence parameters (A2/A1 indexes) proved to be the most accurate in differentiating between oils from different reservoirs. The FTIR aromaticity and aliphaticity indexes can by used as maturity parameters, expressed as a ratio between aromaticity vs. aliphaticity. This shows a very good correlation with API gravity and the degree of sterane isomerization. Thus, the FTIR and UV uorescence techniques, which are both rapid and inexpensive, constitute an interesting alternative for studying the geochemistry of the reservoirs. A good understanding of the reservoirs structure enables us to improve the interpretation of the results.

Acknowledgements The authors acknowledge Repsol-YPF for facilitating the sampling oils. This work was nanced by the DURSI of the Catalonian Government (ACI2002/2; 2001SGR00075 Grup de Geologia Sedimentaria) and Projects BTE20000574-C03-01/02 and BTE2003-06915, from the Spanish Ministry of Science and Technology.

Fig. 17. Relationships between aromaticity vs. aliphaticity FTIR index with (A) API8 gravity, (B and C) sterane isomerization. An important correlation exists in all cases, which conrms the aromaticity/aliphaticity ratio as a good maturity indicator.

168

A. Permanyer et al. / Fuel 84 (2005) 159168 [12] Demaison G, Bourgeois FT. Am Assoc Pet Geologists Stud Geol 1985;18:15161. [13] Seifert WK, Carlson RMK, Moldowan JM. Adv Org Geochem 1981 1983;71024. mpin VF, Casson N. AAPG Treatise Pet Geol, Atlas [14] Seemann U, Pu Oil Gas Fields 1990;A-017:120. [15] Berastegui X, Clavell E. Spec Publ Eur Assoc Pet Geoscientists 1991; 1:35568. [16] Watson HJ. AAPG Memoir 1982;32:23750. [17] Orlopp DE. Proceedings of the Offshore Technology Conference 1988. [18] Doumenq P, Guiliano M, Mille G, Kister J. Anal Chim Acta 1991; 242:13741. [19] Guiliano M, Mille G, Kister J, Muller JF. J Chim Phys 1988;85(10): 96370. [20] Pieri N, Planche JP, Kister J. Analysis 1996;24:11322. [21] Vo-Dinh T, Gammage RB, Martinez PR. Anal Chim Acta 1980; 118(2):31223. [22] Kister J, Pieri N, Alvarez R, Diez MA, Pis JJ. Energy Fuels 1996;10: 94857. [23] Peters KE, Moldowan JM. The biomarker guide, interpreting molecular fossils in petroleum and ancient sediments. Englewood Cliffs, NJ: Prentice Hall; 1993 p. 363. [24] Bein A, Sofer Z. Am Assoc Pet Geol Bull 1987;71:6575.

References
[1] Larter SR, Aplin AC. Geological Soc Spec Publ 1995;86:522. [2] England WA, Cubitt JM. Geological Soc Spec Publ 1995;86:13. [3] Kaufman RL, Ahmed AS, Elsinger RL. GCS-SEPM Foundation Ninth Annual Research Conference Proceeding, New Orleans 1990;26382. [4] Hwang RJ, Ahmed AS, Moldowan JM. Org Geochem 1994;21: 17188. [5] Hwang RJ, Baskin DK. Middle East Pet Geosci Geo 1994;94(2): 52941. [6] Magnier C, Trindade LAF. Rev Latin Am Geoqu Org 1999;5:2537. [7] Magnier C, Trindade LAF, Becker M, Becker W, Cerqueira JR. Proceedings of the International Meeting on Organic Geochemistry, Istambul, Sept. 1999. [8] Permanyer A, Doui L, Lahcini A, Lamontagne J, Kister J. Fuel 2002; 81(7):8616. bufa C, Kister J. Eigth Latin[9] Permanyer A, Doui L, Alberdi M, Re American Congress on Organic Geochemistry 2002;1114. s J, Algaba J, Clavell E, Grimalt J. Org Geochem 1986;10: [10] Albaige 44050. a-Sin eriz B, Querol R, Castillo F, Fernandez JR. Proceedings of [11] Garc the 10th World Petroleum Congress, Bucarest, vol. 4;1980 p. 14.