In¯uence of biodegradation on crude oil acidity and
carboxylic acid composition
W. Meredith
a, S.-J. Kelland
b, D.M. Jones
a,*
aFossil Fuels and Environmental Geochemistry (Postgraduate Institute): NRG, University of Newcastle-upon-Tyne, NE1 7RU, UK bEnterprise Oil plc, Grand Buildings, Trafalgar Square, London, WC2N 5EJ, UK
Received 31 January 2000; accepted 12 September 2000 (returned to author for revision 15 May 2000)
Abstract
Quantitative analysis of separated carboxylic acid fractions of 33 crude oils from the UK, Italy and California, showed that the carboxylic acid fraction is a major factor responsible for the acidity in these oils. It was apparent that biodegradation is the main process that produces high concentrations of carboxylic acids in these crude oils with the extent of biodegradation, as measured from their hydrocarbon compositions, being clearly correlated with their total acid number (TAN). Although probably not important in in¯uencing oil TAN, the distribution of C30±C32hopanoic
acids was also seen to be controlled by biodegradation, increasing in concentration for all but the most biodegraded oils. Hopanoic acids with the 17b(H),21b(H) stereochemistry were found in many of the biodegraded oils, and were thought to be mainly derived from the bacteria that were responsible for the biodegradation of the oil. This may have implications for the timing and mechanisms of the biodegradation involved. The role of C0±C3alkylphenols in
deter-mining oil acidity was investigated and shown not to be a signi®cant factor in the sample set studied. However, a number of undegraded oils, with low carboxylic acid contents were seen to have relatively high acidities, showing that factors other than biodegradation, possibly related to high sulphur content can control oil acidity in certain oil types.
#2000 Elsevier Science Ltd. All rights reserved.
Keywords:Oil acidity; Total acid number; TAN; Carboxylic acids; Naphthenic acids; Hopanoic acids; Alkylphenol; Crude oil biode-gradation
1. Introduction
The acidity of a crude oil is measured by its total acid number (TAN), which is the number of milligrams of KOH required to neutralise the acidity in one gram of oil (Derungs, 1956). High TAN oils (>0.5 mg KOH/g) are less desirable than low TAN oils because of the cor-rosion and re®nery problems they cause, and they there-fore attract a much lower price (Babaian-Kibala et al., 1993; Turnbull et al., 1998). Petroleum acids have been termed naphthenic acids since the earliest identi®ed
acids were saturated cyclic carboxylic acids, although there are potentially many dierent classes of carboxylic acids present in crude oils (Derungs, 1956). The problem of naphthenic acid corrosion was ®rst identi®ed in the 19200s during the re®ning of oils from Romania,
Cali-fornia and Russia (Derungs, 1956), with the ®rst major study on acid structures published by Lochte and Littman (1955). The occurrence of high TAN oils is widespread, e.g. Venezuela, Nigeria, Texas, North Sea (Jayaraman et al., 1986), and their exploitation is becoming more important with time (Babaian-Kibala et al., 1998). Oils can vary in TAN from <0.1, to as high as 8 mg KOH/g (Babaian-Kibala et al., 1998).
There have been reports of the detailed analysis of the carboxylic acid fraction of crude oils, with compounds identi®ed including linear fatty acids, isoprenoid acids,
0146-6380/00/$ - see front matter#2000 Elsevier Science Ltd. All rights reserved. P I I : S 0 1 4 6 - 6 3 8 0 ( 0 0 ) 0 0 1 3 6 - 4
Organic Geochemistry 31 (2000) 1059±1073
www.elsevier.nl/locate/orggeochem
* Corresponding author. Tel.: 191-222-8628; fax: +44-191-222-5431.
monocyclic, polycyclic and aromatic acids (e.g. Lochte and Littman, 1955; Carson and Graham, 1965; Seifert and Teeter, 1970; Seifert, 1975). The cyclic acids have been classi®ed according to their ring structure (e.g. Fan, 1991).
There appears to have been relatively little detailed work published on identifying the compound groups responsible for high TAN values in oils, although clearly the carboxylic, (including naphthenic) acids, which are known to be produced during the in-reservoir biodegradation of petroleum hydrocarbons, could be one such group (Mackenzie et al., 1983). The role of saturated n-acids in the control of oil TAN has been investigated, with both Jae and Gallardo (1993), and Barth et al. (1998) showing that these compounds exert very limited in¯uence. It has been suggested that biode-gradation is one of the processes which may lead to oils having high TAN values, though the relationship is not a simple one (e.g. Behar and Albrecht, 1984; Jae and Gallardo, 1993; Olsen, 1998). It has been suggested that the increase in concentration of acidic compounds observed in many biodegraded oils could be the result of either the neoformation of acids during biodegradation, or the preferential removal of non-acidic compounds, leading to a relative increase in the concentration of the acidic components (Behar and Albrecht, 1984). Organic acids will also be incorporated into the oil during source rock maturation (e.g. Mackenzie et al., 1983), although acetic acid is the major product of organic acid genera-tion, with the concentration of longer chain acids pro-duced rapidly decreasing with chain length (Barth and Bjùrlykke, 1993). These short chained acids are prob-ably not a major contributory factor to oil TAN, since they are very water soluble and will rapidly partition into the aqueous phase (e.g. Reinsel et al., 1994). The relatively high water solubility of lower molecular weight carboxylic acids mean that extensive water wash-ing, often associated with biodegradation, will deplete these components, thus resulting in acid concentrations and distributions which may re¯ect the extent and tim-ings of these two processes (Mackenzie et al., 1983). Thermal degradation may also control the fate of the acids produced during biodegradation, with the linear acids, and especially any unsaturated acids being removed preferentially to the other oil constituents (Mackenzie et al., 1983).
The carboxylic acid fraction of crude oils contains biomarkers which could potentially yield important information on the geological history of petroleums, although these compounds are not routinely analysed in petroleum exploration. However, the concentration and distribution of C30±C32hopanoic acids have been used
as indicators of oil maturation (Jae and Gardinali, 1990), migration (Jae et al., 1988a,b), and biodegrada-tion (Behar and Albrecht, 1984; Nascimento et al., 1999). Recently, hopanoic acids have been shown to be
formed by the oxidation of the corresponding hopanes during the laboratory biodegradation of crude oils (Watson et al., 1999).
Other groups of compounds which could in¯uence the TAN of an oil include low molecular weight alkyl-phenols (C0±C3 alkylphenols), which occur widely in
crude oils (Ioppolo et al., 1992). For example, Sama-dova and Guseinova (1993) noted that in a study of low wax, high TAN crude oils from Azerbaijan, the phenol contents were between two and seven times higher than those of the carboxylic acids.
In this study we investigate the in¯uence of the car-boxylic acid fraction on oil acidity, by comparing the TAN values of 14 undegraded, 13 biodegraded and 6 mixed crude oils from various geological settings and geographical locations world-wide, with the quantitative analysis of their separated carboxylic acid fractions. The relationship between oil TAN, carboxylic acid fraction, sulphur and C0±C3alkylphenol content with the extent
of biodegradation is also investigated in these oils.
2. Materials and methods
2.1. Samples
The sample set comprised 33 oils, 24 from oshore UK (from 13 ®elds), 4 from Italy (4 ®elds), and 5 from a single Californian ®eld.
2.2. TAN and sulphur analysis
The TAN value, and total sulphur content of 19 of the oil samples were obtained by standard methods ASTM D 664 and ASTM D 2622 respectively. The TAN value for each of these oil samples was measured in duplicate. Details of the methodology of the TAN analysis, together with accuracy and precision data are given in the ASTM D 664 method description. Although this is a standard method for the measure-ment of oil acidity there are a number of potential cau-ses of inaccuracy which are described by Piehl (1988). The data for the remaining 14 oils was obtained from oil company assay reports, with analysis by the same stan-dard methods.
2.3. Hydrocarbon analysis
The saturated and aromatic hydrocarbon fractions were isolated from each oil by thin layer chromato-graphy using 0.5 mm thickness Merck Kieselgel 60G1
plates, and light petroleum (b.p. 40±60C) as developer,
in a similar way to that described in Farrimond et al. (1994). The fractions were then analysed by gas chroma-tography (GC), and gas chromachroma-tography±mass spectro-metry (GC±MS), in order to assess the degree of biodegradation, maturity and source characteristics of the oils. Two samples were analysed in triplicate to assess the repeatability of the technique, with procedural blanks analysed to assess contamination.
Gas chromatographic analyses were carried out on a Hewlett Packard 5890A-II instrument ®tted with a Hew-lett Packard 7672 autosampler. A HewHew-lett Packard HP-5 phenylmethylsilicone coated (0.25 mm ®lm thickness) fused silica column (25 m0.25 mm i.d.) was employed, using hydrogen as carrier gas, and a ¯ame ionisation detector. Splitless injection was used, with an oven pro-gramme of 50C (2 min) to 300C at 4C minÿ1, where
it was held for 20 mins. Data were acquired and pro-cessed using a PC based, LabSystems XChrom data system.
GC±MS analyses were carried out on a Hewlett Packard 5890A-II-5972 MSD quadrupole instrument ®tted with a similar HP-5 column, with hydrogen as the carrier gas. Automated splitless (1 min) injection was used and, for the saturated hydrocarbons, the GC oven temperature programme was 40±175C at 10C minÿ1,
175C (1 min) to 225C at 6C minÿ1and 225C (1 min)
to 300C at 4C minÿ1, held for 18 min. Aromatic
hydrocarbon fractions were analysed with an oven pro-gram of 40C (3 min) to 300C (12 min) at 4C minÿ1.
2.4. Carboxylic acid extraction
The carboxylic acid fraction was extracted from each oil sample by a novel non-aqueous ion exchange method (unpublished results). This method is, brie¯y, as follows. Aliquots (0.4 g) of each oil were loaded onto a 4 g SAX quaternary amine solid phase extraction (SPE) column, with the non-acid compounds removed by successive elutions with n-hexane, and dichloromethane (DCM). The crude carboxylic acid isolate was then eluted with diethyl ether containing 2% (v/v) formic acid. This frac-tion was methylated with boron tri¯uoride-methanol (BF3-methanol) (60C for 1 h), and cleaned up by eluting
through a silica SPE column with 60% / 40% (v/v) n -hexane/DCM. 1-adamantane carboxylic acid (Fluka, U.K.) and 5-cholanic acid (Sigma, UK) were used as surrogate (extraction eciency) standards, and the methyl ester of 1-phenyl-1-cyclohexane carboxylic acid as an internal standard for quanti®cation purposes. Analysis was by gas chromatography as for the saturated
hydrocarbon analysis. Quanti®cation was by compar-ison of analyte peak areas with that of the internal standard. The response factors of the analyte com-pounds were assumed to be unity, whilst the response factors of the surrogate standards were calculated and corrected for. Recovery of the surrogate standards throughout the procedure was found to be approxi-mately 90%. Oils spiked with a suite ofn-alkanoic acid standards in the rangen-C10:0ton-C30:0were analysed,
and recoveries of approximately 61% forn-C10:0, 90%
for n-C14:0, and 45% for n-C30:0 were obtained. The
recoveries of aromatic acids were much lower e.g. ben-zoic acid 16% and 2-naphthoic acid 30% (Jones et al., unpublished results). It was thought that the poor recovery of the aromatic acids may be due to the ine-cient methylation of these acids by BF3-methanol (e.g.
Behar and Albrecht, 1984). The re-analysis of acid frac-tions from seven of these oils using diazomethane as a methylating agent showed no signi®cant dierence in appearance in the gas chromatograms of the extracted acids (as methyl esters), and the total concentrations of the extracted acid fractions were within 10% of the results obtained using BF3-methanol. Therefore it was
assumed that the aromatic acids were not a major com-ponent of the total carboxylic acid fraction of these oils. Five samples were analysed in triplicate to assess repeatability of the method used, with procedural blanks analysed to check for contamination. It was found thatn-C16:0andn-C18:0fatty acid contaminants
were introduced during the ion-exchange separations; these were quanti®ed and corrected for during ®nal quantitation.
The concentrations and distributions of the hopanoic acids were investigated by GC±MS, under the same conditions as the saturated hydrocarbon analysis, with the data acquired in both the selective ion monitoring (SIM), and scan modes. The concentrations of the C30±
C32 hopanoic acids were measured by comparison of
their peak areas in their respective m/z 235, 249, 263 mass chromatograms, with that of the added standard 5b-cholanic acid in the m/z 217 mass chromatogram. Response factors of unity were used, so the hopanoic acid data should be considered as semi-quantitative.
2.5. Alkylphenol analysis
The alkylphenols were extracted from each oil sample by the C18 silica solid phase extraction method, described by Bennett et al. (1996). After extraction, the alkylphenol fractions were derivatised with N,O-Bis(trimethylsilyl)tri-¯uoroacetamide (BSTFA) (60C for 1 h) and analysed
by GC±MS, as for the hydrocarbon analysis except for the following oven program: 35C (10 min) to 150C at
2C minÿ1and from 150C to 300C at 8C minÿ1, held
for 20 min. Data were acquired in the SIM mode. Quanti®cation was by comparison with the peak area of
the internal standard, 2-naphthol, with correction for its response factor. Two samples were analysed in triplicate to assess repeatability, with procedural blanks analysed to check for contamination.
3. Results
3.1. TAN and sulphur data
The TAN data from both the direct analysis of the samples, and from oil company assay reports are sum-marised in Table 1. The samples have a range of TAN from 0.1 to 2.67 mg KOH/g, with those oils having a TAN>0.5 mg KOH/g being classi®ed as high TAN.
The sulphur content of these samples range from 0.24 to 4.8% by weight. With the exception of one sample with a sulphur content of 3.1% the oils from the UK are generally low sulphur oils. In contrast, the oils from Italy and California are high sulphur, with a content greater than 1% by weight.
3.2. Hydrocarbon analysis
Gas chromatograms of the saturated hydrocarbon fraction from typical examples of an undegraded oil, a moderately degraded oil, a severely degraded oil and a mixed oil are shown in Fig. 1. The chromatograms of the undegraded oils (e.g. oil no. 5 Ð UK) show a complete envelope ofn-alkanes, which are reduced in a moderately degraded oils (e.g. oil no. 30 Ð California), whilst heavily degraded oils (e.g. oil no. 13 Ð UK) are dominated by resistant biomarker compounds. Some of the oils (e.g. oil no. 22 Ð UK) appeared undegraded, although they contained 25-norhopanes which was attributed to the mixing of a fresh oil with a heavily degraded oil (cf. Volkman et al., 1984; Horstad and Larter, 1997).
The degree of oil biodegradation can be described by both the pristane/nC17ratio (Bailey et al., 1973), and by
assessing the overall distribution of petroleum hydro-carbons. Various classi®cations have been proposed, primarily based on the order of degradation of satu-rated hydrocarbon compound groups (e.g. Connan et al., 1980; Volkman et al., 1984). The oils in this study were ranked according to the 10 point scale of Peters and Moldowan (1993), with level 1 representing an undegraded oil (unaltered suite ofn-alkanes), and level 10 a severely degraded oil (hopanes and diasteranes absent, C26±C29 aromatic steroids attacked). Together
with this classi®cation, the pristane/nC17ratio of each
oil is shown in Table 1.
The ratios of a wide range of source and maturity dependent biomarkers were determined from GC±MS analysis of both the saturated and aromatic hydrocarbon fractions. These ratios could potentially be in¯uenced by the extensive biodegradation apparent in many of the
oils. All of the oils were thought to be Type II (except the Californian oils which were Type IIS) marine sourced oils. The maturity of the oils was variable with the data of one of the maturity parameters, the relative concentration of the C295a(H),14b(H),17b(H) steroidal
hydrocarbon biomarker (abb) isomer to the (aaa+abb) isomers is shown in Table 1.
3.3. Carboxylic acid analysis
The concentration of the total carboxylic acid frac-tion, and the total concentration ofn-acids from each oils are shown in Table 1. Gas chromatograms of the carboxylic acid fraction from an undegraded oil, a moderately degraded oil, a highly degraded oil and a mixed oil are shown in Fig. 1. The gas chromatogram from a typical undegraded oil (e.g. oil no. 5) is domi-nated by a homologous series of saturatedn-acids (C8:0±
C34:0) with a low total concentration of <300mg/g. The
n-acids are also apparent in the gas chromatogram of a moderately degraded oil (e.g. oil no. 30), together with a large number of unidenti®ed peaks and an unresolved complex mixture (UCM) of medium molecular weight branched and cyclic carboxylic acids. Then-acids were not prominent in the gas chromatograms of a typical highly degraded oil (e.g. oil no. 13), which is instead dominated by a higher molecular weight UCM, with a much higher total concentration of acids (up to 9000mg/ g). The indication from the hydrocarbon compositions that six of the samples are mixtures of undegraded and severely degraded oils (e.g. oil no.22) is supported by this carboxylic acid data. The gas chromatograms show the high molecular weight UCM derived from the severely degraded component on which is superimposed a homologous series ofn-acids from the undegraded oil. Also visible on this chromatogram is a series of peaks with a retention time between 65 and 75 min, which were shown by GC±MS (using a dierent temperature programme) to be hopanoic acids.
The C30±C32 hopanoic acids were present in the
majority of the oils analysed, although they were not detected in some of the undegraded oils and in the one most severely degraded oil. As described by Jae and Gallardo (1993), these hopanoic acids formed a pseudo-homologous series, isomeric at positions C-17 and C-21 (ab,baandbb) and at position C-22 (R and S). Exam-ple m/z 235, 249 and 263 mass chromatograms from undegraded, moderately degraded, severely degraded and mixed oils (samples no. 5, 30, 13 and 22) are shown in Fig. 2. The C30,C31 and C32 hopanoic acids were
Table 1
Location of and geochemical data from the 33 oils in the sample set
Sample no.
Field Location Degree of biodegradationb
TAN (mg KOH/g)
Total acids (mg/g)
Calculated TAN (mg KOH/g)c
% Of measured TAN
Total
n-acids (mg/g)
Totalab
hop acids (ng/g)
Totalba
hop acids (ng/g)
Totalbb
hop acids (ng/g)
Sulphur wt.%
Total C0±C3 alkylphenols (mg/g)
Pristane /nC17
C29 sterane
abb/ (aaa+
abb) ratio
1 A UK 1 0.05a 178 0.03 61 11 n.d. n.d. n.d. 0.38a 35 0.60 0.49
2 A UK 1 0.05a 137 0.02 47 10 n.d. n.d. n.d. 0.38a 43 0.61 0.49
3 B UK 1 0.25 140 0.02 10 13 n.d. n.d. n.d. 0.44 38 0.66 0.53
4 B UK 1 0.10 266 0.05 45 14 n.d. n.d. n.d. 0.41a 53 0.66 0.52
5 C UK 1 0.25 222 0.04 15 33 172 229 31 0.23 39 0.56 0.54
6 D UK 1 0.04a 225 0.04 96 5 n.d. n.d. n.d. 0.54a 116 0.57 0.54
7 E UK 1 0.60 245 0.04 7 18 n.d. n.d. n.d. 3.10 31 0.37 0.50
8 F UK 1 0.10a 422 0.07 72 12 183 166 139 0.24a 14 0.63 0.53
9 F UK 1 0.10a 103 0.02 18 9 n.d. n.d. n.d. 0.24a 18 0.60 0.54
10 G UK 6 1.85 4152 0.71 38 11 1066 567 2870 1.20 8 n.d.d 0.53
11 H UK 6 1.40 2439 0.42 30 7 713 860 487 0.91 26 n.d. 0.36
12 I UK 7 2.00 6646 1.13 57 5 202 58 3821 0.60 2 n.d. 0.55
13 J UK 7 2.25 8535 1.46 65 3 700 81 4390 0.62 3 n.d. 0.54
14 J UK 7 2.67a 8697 1.48 56 7 513 99 2056 0.85a 2 n.d. 0.55
15 J UK 7 2.67a 7692 1.31 49 5 456 76 2118 0.85a 2 n.d. 0.55
16 J UK 7 2.67a 7644 1.30 49 2 334 69 1987 0.85a 1 n.d. 0.55
17 J UK 7 2.67a 8479 1.45 54 6 395 102 2042 0.85a 2 n.d. 0.54
18 K UK 9 2.15 7519 1.28 60 8 n.d. n.d. n.d. 0.87 1 n.d. n.d
19 L UK Mixed 0.30 1547 0.26 88 5 1119 962 3964 0.42 9 0.83 0.45
20 M UK Mixed 1.20a 2700 0.46 38 16 1573 1068 328 0.44a 19 0.41 0.44
21 M UK Mixed 1.19a 3239 0.55 46 20 2289 1608 373 0.47a 21 0.41 0.44
22 M UK Mixed 1.60 4558 0.78 49 23 2814 1906 379 0.44a 15 0.41 0.42
23 M UK Mixed 1.20a 3233 0.55 46 15 1686 1131 245 0.44a 19 0.41 0.44
24 M UK Mixed 1.20a 3868 0.66 55 15 1821 1134 275 0.44a 51 0.43 0.44
25 N Italy 1 0.25 33 0.01 2 6 n.d. n.d. n.d. 2.00 9 0.11 0.42
26 O Italy 1 0.72a 210 0.04 5 5 n.d. n.d. n.d. 2.33a 9 0.12 0.42
27 P Italy 1 0.35 225 0.04 11 14 90 45 4 4.80 35 0.15 0.43
28 Q Italy 1 0.15 39 0.01 4 4 n.d. n.d. n.d. 1.60 26 0.13 0.43
29 R California 1 0.20 118 0.02 10 21 n.d. n.d. n.d. 1.40 39 1.04 0.46
30 R California 2 1.00 1511 0.26 26 20 739 463 455 1.70 46 2.43 0.46
31 R California 2 1.00 534 0.09 9 19 1233 610 227 3.90 54 1.91 0.44
32 R California 3 1.25 1877 0.32 26 21 1604 920 614 2.10 34 4.50 0.44
33 R California 4 1.75 3635 0.62 35 29 1211 650 466 2.10 81 n.d. 0.45
a Data from assay reports.
b Scale of Peters and Moldowan (1993). c See text for details.
d n.d., not detected.
W.
Meredith
et
al.
/
Organic
Geochemistry
31
(2000)
1059±1073
3.4. Alkylphenol analysis
The concentration of the C0±C3 alkylphenols were
measured from the relevant mass chromatograms (m/z
151 for phenol,m/z165 for the cresols,m/z179 for the dimethylphenols and m/z 193 for the trimethylphenols), as reported by Bennett et al. (1996). As shown in Table 1 the total concentration of C0±C3alkylphenols in these
samples varies between 1 and 116mg/g.
4. Discussion
4.1. Contribution of isolated carboxylic acids to oil TAN
The strong correlation (r2=0.91) between the
con-centration of the carboxylic acid fraction and TAN (shown in Fig. 3), suggests that this fraction is a major contributor to the TAN of these oils. This correlation can be seen to be valid for oils from a variety of dierent areas and dierent geological settings (UK, Italy, Cali-fornia), with dierent source environments and matura-tion histories.
Below TAN values of 0.5 mg KOH/g the correlation between TAN and total acid concentration is not as clear as that for the higher TAN oils. This indicates that compounds other than carboxylic acids contribute sig-ni®cantly to the TAN in these oils. The poor correlation for the low TAN oils may also be in part due to the analytical errors. Variation in duplicate TAN analysis was around 0.1 mg KOH/g, which is of greater sig-ni®cance in low TAN compared to high TAN oils. In addition relative errors in the acid fraction quantitation are greater in low TAN oils, for example, in replicate analyses of oils with acid concentrations below 200mg/g the relative standard deviation can be up to 17% (at higher acid concentrations of 4000 mg/g this error decreased to around 7%).
However, some individual oils lie away from the trend line and may have unique features which require further investigation that could provide key insights into what other geochemical characteristics control oil TAN. For example, sample nos. 7, 26 and 31 (from the UK, Italy
Fig. 2. Partialm/z235, 249, 263 mass chromatograms showing the C30, C31and C32hopanoic acids (as methyl esters) extracted from: A Ð undegraded oil (no. 5); B Ð moderately degraded oil (no. 30); C Ð extensively degraded oil (no. 13) and D Ð mixed oil (no. 22). Isomers designated as: 1-17a,21b,[22S]; 2-17a,21b,[22R]; 3-17b,21a,[22S]; 4-17b,21a,[22R]; 5-17b,21a,[22R]; 6-17b,21b,[22S]. Mass chromatograms for each oil are normal-ised to the most abundant peak. Total concentrations of C30± C32hopanoic acids for each oil: A Ð 432 (ng/g); B Ð 1657 (ng/ g); CÐ5171 (ng/g) and DÐ5099 (ng/g).
Fig. 3. Concentration of carboxylic acid fraction (mg/g oil) vs. TAN, which shows the strong correlation (r2=0.91) between the acidity and carboxylic acid content of these oils.
and California, respectively), all of which have high sulphur contents, can be seen to have high acidities, (0.6, 0.72 and 1.0 mg KOH/g), despite having low carboxylic acid contents (245, 210 and 534mg/g). The recovery of the surrogate standards for these oils were comparable to that of the remainder of the sample set, and so the low concentration of carboxylic acids isolated from these samples is not an analytical artefact.
If one assumes that the ``average'' acid is a C22
com-pound, with a saturated tricyclic structure, an alkyl side chain and one carboxylic acid group (e.g. Olsen, 1998; Tomczyk et al., 1998), it would have an average mole-cular weight of 334. This is consistent with the retention position of the UCM maxima in the chromatograms of our isolated acid fractions. It is therefore possible to calculate the expected TAN of each oil, if it was con-trolled by its carboxylic acid content alone. The results of these concentrations are shown in Table 1. Excluding the low TAN oils (<0.5 mg KOH/g), the calculated TANs are between 26 and 65% of the measured values. The exceptions are three high sulphur oils (sample nos. 7, 26 and 31) whose carboxylic acid content appears to contribute less than 10% to the total oil TAN, indicat-ing that some other group of compounds is controllindicat-ing the TAN of these oils. The apparently low contribution to the total TAN of the isolated carboxylic acid fraction may be because either the average structure assumed is incorrect (in molecular weight, or with dicarboxylic acids being a major component), or that non-GC amenable, higher molecular weight acids are also contributing to the TAN of the oils. However, it is still apparent that in the majority of these oils, the carboxylic acid fraction is a major contributor to oil TAN.
There is no clear trend in relationship between then -acid concentration of these oils and their TAN (Fig. 4), although there is some evidence of a trend of decreasing
n-acid concentration with increasing TAN. The scatter on the plot is due in part to the eect of the samples with mixed compositions, which have an-acid distribu-tion re¯ecting the undegraded component of the oil, while the acidity is due in most part to the biodegraded component. These ®ndings support earlier work (e.g. Jae and Gallardo, 1993; Barth et al., 1998) which showed that n-acids do not contribute greatly to the TAN of an oil.
4.2. Role of biodegradation in controlling oil TAN
The relationship between the pristane/nC17ratio and
TAN for these oils is shown in Fig. 5. Amongst the undegraded oils from both the UK and Italy there is little variation in this ratio. This suggests that the var-iation in TAN apparent in these oils is due to processes other than biodegradation. Amongst the four Cali-fornian oils a trend of increasing TAN with increasing pristane/nC17 is apparent. Although these oils are
thought to share a similar source, factors other than biodegradation may be in¯uencing this ratio, which in addition to the small sample set may be re¯ected in the relatively poor correlation of this relationship. How-ever, this data is sucient to suggest that for these oils, of up to moderate degradation, the degree of biode-gradation is an important controlling factor on the ulti-mate TAN. This conclusion is supported by the work of Olsen (1998), who demonstrated a strong correlation
Fig. 4. Concentration of totaln-acids (mg/g oil) vs. TAN.
between TAN and biodegradation (assessed by the phytane /nC18ratio) for a suite of Norwegian oils.
The relationship between the degree of biodegrada-tion (as de®ned by the scale of Peters and Moldowan, 1993) and TAN (Fig. 6) shows a trend of increasing TAN with an increase of biodegradation, although there is some degree of scatter (r2=0.74). The relationship
between the level of biodegradation, and the concentra-tion of carboxylic acids in these oils (Fig. 7) shows a
very similar trend to that in Fig. 6, although the corre-lation is stronger (r2=0.86). There is a little scatter in
the concentration of carboxylic acids in the undegraded oils, which indicates that compounds other than car-boxylic acids are responsible for the high acidity of some of the undegraded oils. These data, therefore, strongly suggest that it is the degree of biodegradation of an oil which controls the concentration and distribution of carboxylic acids.
These data support the hypothesis that biodegrada-tion is a major control of oil TAN, which has been pre-viously alluded to (e.g. Behar and Albrecht, 1984; Jae and Gallardo, 1993; Olsen, 1998), but never demon-strated for a large sample set. Although some minor increase in TAN may occur by the preferential removal of other groups, the tenfold increase in the carboxylic acid concentration between the least and the most degraded oils strongly suggests that the neoformation of these acids during biodegradation is the major process leading to high TAN values.
The results of this study suggest that the primary geological and geochemical conditions that appear to give rise to most high TAN, acid crudes are those that favour biodegradation. Those conditions are generally found in reservoirs that are (or were at one time) shallow, with temperatures lower than about 100C (e.g. Bailey
et al., 1973; Connan, 1984). Traditionally, in-reservoir oil biodegradation has been associated with oxygenated meteoric waters (e.g. Palmer, 1993), and carboxylic acids are known to be intermediates during the aerobic biodegradation of petroleum hydrocarbons (for pro-posed pathways, see Singer and Finnerty, 1984). More recently there has been evidence that anaerobic degra-dation is possible, although the mechanisms have tended to be speculative (e.g. Stetter et al., 1993; Rueter et al., 1994; Wilkes et al., 1995; Heider et al., 1999). However, very recently Zengler et al. (1999) reported that long chain alkanes can be metabolised anaerobically by a consortium of bacteria resulting in the formation of methane. It is not clear whether it is aerobic or anaero-bic biodegradation processes which cause the increase in acidity seen in these oils.
4.3. Distribution of hopanoic acids
The relationship between the total C30±C32hopanoic
acid concentration of these oils, and both their degree of biodegradation and TAN (Fig. 8), shows that there is initially a trend of increasing hopanoic acid concentra-tion with increasing TAN, to a maximum TAN value of approximately 2.0 mg KOH/g (r2=0.91). Many of the
undegraded oils contain no detectable hopanoic acids and in those undegraded oils which do contain them, they are present in very low concentrations (<500 ng/g). With an increase in degree of biodegradation through the moderately degraded Californian oils, the concentration
Fig. 7. Degree of biodegradation (scale of Peters and Moldo-wan, 1993) vs. concentration of carboxylic acids (mg/g oil), excluding mixed oils.
Fig. 6. Degree of biodegradation (scale of Peters and Moldo-wan, 1993) vs. TAN, excluding mixed oils.
of hopanoic acids can be seen to increase. The maximum hopanoic acid contents found were present in the exten-sively or severely degraded oils UK oils (5000 ng/g). After this maximum, the concentration of hopanoic acids appears to rapidly decrease, with the most degraded sam-ple (no. 18), containing no detectable hopanoic acids.
The distribution of the C30±C32 hopanoic acids
between the three stereochemical con®gurations are shown in Fig. 9. Both the total 17a,21band 17b,21aacids show an initial increase in concentration with increasing TAN, although there is signi®cant scatter in the data. This trend continues to a maximum TAN value of approxi-mately 1.5 mg KOH/g, (and extensive levels of biode-gradation) before the concentration of hopanoic acids decreased to zero in the most severely degraded oil.
The total 17b,21bacids show a relatively minor increase in concentration with increasing TAN and degree of bio-degradation up to moderately degraded oils, with a TAN of 0.5 mg KOH/g. The extensively degraded oils then showed a dramatic ten fold increase in concentration ofbb
hopanoic acids to a maximum of 4500 ng/g, at a TAN of approximately 2 mg KOH/g. As with the ab and ba
acids, there was then a sharp decrease to zero inbbacid concentration in the most severely degraded oil.
The hopanoic acid concentrations in the mixed oils obscure the above trends, and so are not included on these plots. Sample no. 19 has a very high hopanoic acid content (6045 ng/g), dominated bybbacids (3964 ng/g), despite a relatively low TAN of (0.3 mg KOH/g). The other mixed oils, which are all from the same UK ®eld
(®eld M) show considerable variation in their total hopa-noic acid content (20- 2969 ng/g to 22- 5099 ng/g), which are predominantly of theabandbbcon®gurations. This suggests these acids may be a preserved signal of biode-gradation heterogeneity within the ®eld before the sec-ond charge of oil occurred, or that alternatively some process other than biodegradation are responsible for the distribution of hopanoic acids seen in these oils.
The observed trend of increasing hopanoic acid con-centration with increasing degree of biodegradation (Fig. 8), is in contrast to a study by Behar and Albrecht (1984), which showed an apparent decrease in hopanoic acid concentration with an increasing degree of biode-gradation amongst ®ve unrelated oils. The observed initial increase in hopanoic acid content with increasing biodegradation of the oils in this present study could be due to the preferential removal of other compounds during biodegradation. However, it is more probable that this trend re¯ects either the incorporation of these acids during migration, or the neoformation of hopa-noic acids within the biodegraded oils.
It has been suggested that the incorporation of hopa-noic acids during migration can result in an isomeric distribution of these acids which can be used to describe the maturity of the rocks through which an oil had migrated (Jae et al., 1988a, b). Thebbcon®guration is the least thermodynamically stable of the three isomers, and their presence in an oil has been suggested to indi-cate the incorporation of immature organic matter dur-ing migration (Jae et al., 1988a). The ab and ba
hopanoic acids are formed from the isomerisation ofbb
hopanoids via diagenetic processes (e.g. Seifert, 1975). Therefore, oils containing only theabandbaacids were thought to have migrated through relatively mature rocks (Jae and Gallardo, 1993). It is likely that many of the UK oils would have migrated through Cretaceous and Tertiary mudstones (Lovell, 1990), which due to their depth and consequent temperature would have a maturity which would preclude the presence of bb
hopanoic acids. Consequently, it is unlikely that these oils would have incorporated anybbacids during their secondary migration.
An alternative mechanism for the incorporation into crude oils ofbbacids at relatively high levels of maturity has been proposed by Jae and Gardinali (1990). Those authors indicated that between vitrinite re¯ectance maturity measurements of approximately 0.5±0.8% Ro, there is a secondary generation of hopanoic acids which had been tightly bound to organic geopolymers within the source rock. These acids retained their original immature con®guration and could give rise tobbacids within the produced oil (Jae and Gardinali, 1990; Jae and Gallardo, 1993). However, the absolute concentra-tions ofbbhopanoic acids produced during secondary generation have been shown to be very minor compared to the concentration of hopanoic acids produced during
their primary generation at lower maturities (Jae and Gardinali, 1990; Bennett and Abbott, 1999). In this present study the concentration of bb hopanoic acids can be seen to signi®cantly increase in the biodegraded oils. Therefore, the hypothesis of secondary generation of these acids is unlikely to account for the high con-centrations of bbacids found in many of these highly degraded oils.
The stereochemistry of the hopanoids isolated from bacteria is always 17b(H),21b(H) (Rohmer et al., 1992), and the predominance of bb acids in many of the degraded oils may suggest a biological origin. It can be seen in Fig. 2 that the distribution of hopanoic acids in the degraded oil (sample no. 13) is dominated by the C32
bb[R] isomer. This distribution is also seen in the other severely degraded oils with high hopanoic acid contents.
The dominance of the C32 bb isomer over the other
homologues has been previously reported in recent sediments (e.g. Buchholz et al., 1993; Ries-Kautt and Albrecht, 1989), and is thought to be due to the oxidation and side chain cleavage of the extended C35
tetra-functionalised biohopanoid precursors during diagenesis (Innes et al., 1997). Tetrafunctionalised biohopanoids have been noted to be the most abundant type of functionalised hopanoids in recent sediments (Farrimond et al., 2000).
The presence of bb acids in relatively high propor-tions in some of these oils suggests that they may be derived from the biomass of the micro-organisms responsible for biodegrading the oil. Since this isomeric con®guration is the least thermodynamically stable of the three isomers it also suggests that either biode-gradation is on-going in these reservoirs, or that since it
Fig. 9. Total C30±C32hopanoic acid concentration (ng/g) of: A Ð 17a,21bacids; BÐ17b,21aacids and CÐ17b,21bacids vs. TAN (for key see Fig. 8, excluding mixed oils).
ceased the oils have not been subjected to temperatures high enough to cause the isomerisation of thebbacids to the aborbaforms. Another potential in¯uence on the distribution of hopanoic acids in biodegraded oils is the neoformation ofabandbahopanoic acids from the microbial oxidation of hopane hydrocarbons in the ori-ginal oil (Watson et al., 1999).
The observed variations in the isomeric distribution of the hopanoic acids may also give insights on the mechanisms of the biodegradation processes aecting these oils. The hopanoids which are the precursors of these acids have to date not been detected in obligate anaerobic bacteria, and are present only in aerobes or facultative anaerobes (Ourisson et al., 1987). Therefore, the presence of hopanoic acids suggests that these oils may have been biodegraded at least partly by aerobic processes. Alternatively, it may be possible that types of anaerobic bacteria as yet unidenti®ed, which contain hopanoids were involved.
4.4. Role of sulphur content in controlling oil TAN
In the 33 oils analysed in this study there is no clear relationship between their sulphur content and TAN (Fig. 10). Generally the data is very scattered, but by dividing it into oils from geological areas it can be inferred that the extremes in the sulphur contents are probably dependent on source environment, whereas the TAN of an oil is not. When this relationship is investigated for the UK oils separately, with the excep-tion of the one anomalously high sulphur oil, there does seem to be a slight trend of increasing sulphur content with increasing TAN (Fig. 10). This trend has been
reported previously (e.g. Gransch and Posthuma, 1974), and it is probably a biodegradation eect, where the relatively more resistant sulphur containing compounds are concentrated in the oil, as the less resistant hydro-carbons are removed.
However, as noted previously sample nos. 7 and 26, are undegraded, and no. 31 is slightly degraded. These oils contain low carboxylic acid contents, but have high acid-ities. They also have relatively high sulphur contents, which may play a role in controlling the TAN of these oils since there have been reports that unbiodegraded high TAN oils, which have high sulphur contents occur in the Gulf of Mexico (B.A. Stankiewicz, Shell, pers. comm.).
4.5. In¯uence of maturity on oil TAN
An example of the relationship between the TAN and maturity (as assessed by the C29abb/ C29aaa+abb
ster-ane ratio) of these oils is shown in Fig. 11. Within each geographical region there is little variation in maturity (with the exception of one UK outlier), but great variation in the TAN of the oils. It has been reported that ther-mally immature oils may have high acidities (Seifert, 1975) but this sample set does not contain sucient low maturity oils to test this hypothesis.
4.6. Role of alkylphenols in controlling oil TAN
The relationship between the C0±C3alkylphenol
con-centration and TAN is illustrated in Fig. 12, and shows a considerable scatter with no obvious trends. The absence of an observed increase in alkylphenol concentration with an increase in TAN suggests that alkylphenols do
Fig. 10. Sulphur content (wt.%) vs. TAN.
not exert a dominant in¯uence over the TAN in these oils, although due to their acidic nature it is envisaged that they may play a minor role.
The distribution of alkylphenols within these oils may be due to many interacting factors, including migration biodegradation and water washing (Taylor, 1994). Alkylphenols are an intermediate in a number of degra-dative pathways of petroleum components and it has been proposed that they may be formed during the microbial alteration of aromatic hydrocarbons in crude oil (e.g. Kaphammer et al., 1990; Gibson, 1991). The anaerobic degradation of aromatic hydrocarbons in petroleum also has the potential to produce alkylated phenols (e.g. Vogel and Grbic-Galic, 1986; Grbic-Galic and Vogel, 1987). However, the alkylphenols produced could then be consumed by further biodegradation resulting in the complete oxidation to mineralised pro-ducts which can occur under both aerobic and anaero-bic conditions (Howard et al., 1991). Water washing of the oil, which is often associated with biodegradation would selectively remove many of the alkylphenols, especially those of lower molecular weight which are more water soluble (Taylor et al., 1997). Therefore, although it is clear that C0±C3 alkylphenol
concentra-tions are controlled by a number of factors, in this pre-sent work they show no relationship with oil acidity.
5. Conclusions
This study shows that carboxylic acids are a major group of compounds responsible for the high TAN
values of the oils studied, and that biodegradation is the dominant process that produces high concentrations of carboxylic acids in these oils. However, there are examples of undegraded oils with signi®cant TAN values, which suggests that factors other than biodegradation can result in high TAN oils. Such factors may be related to sulphur content which depends on the source rock depositional environment, and possibly the maturity of an oil.
This study supports previous reports in showing that then-acid concentration of an oil does not signi®cantly in¯uence its TAN, and there is also strong evidence that the C0±C3alkylphenol content of oil does not contribute
greatly to oil TAN.
The distribution and concentration of C30±C32
hopa-noic acids can be seen to be partially controlled by bio-degradation. A trend of increasing concentration of these acids, with increasing biodegradation is apparent for all but the most degraded oil, with the hopanoic acids in the extensively degraded oils being pre-dominantly C32compounds of the 17b(H), 21b(H)
con-®guration. These hopanoic acids with the ``biological''
bb con®guration, are thought to be derived from the bacteria responsible for the biodegradation of the oil, and the presence of this thermodynamically unstable con®guration in crude oils may provide information on the timing of the biodegradation. Since hopanoids have not been detected in obligate anaerobic bacteria, the presence of thebb hopanoic acids in biodegraded oils indicates that either the degradation conditions were not strictly anaerobic, or that the degradation involved a type of previously unreported hopanoid containing anaerobic bacteria.
Further work is being undertaken to investigate the eects of source, migration, maturity and environment of degradation on the few oils that plot away from the carboxylic acid concentration / degree of biodegrada-tion versus TAN trend line.
Acknowledgements
We are grateful to Enterprise Oil plc for funding this work, for supplying most of the samples analysed in this project, and for permission to publish this paper. We are also grateful to Saga Petroleum and Norsk Hydro for their support during the development of the acid extraction method used in this study. This manuscript was greatly improved by the constructive reviewing of Tanja Barth and Winston Robbins, and the editorial comments of Andy Bishop. We acknowledge useful dis-cussions from Steve Larter, Andy Aplin, Ian Head, Jon Watson, Barry Bennett and Artur Stankiewicz, and technical assistance from Paul Donohoe, Kim Noke, Berni Bowler and Ian Harrison.
Associate EditorÐA.N. Bishop
Fig. 12. Total concentration of C0±C3alkylphenols (mg/g oil) vs. TAN.
References
Babaian-Kibala, E., Petersen, P.R., Humphries, M.J. 1998. Corrosion by naphthenic acids in crude oils. Preprints of the American Chemical Society, Division of Petroleum Chem-istry. 3;106±110.
Babaian-Kibala, E., Craig, H.L., Rusk, G.L., Blanchard, K.V., Rose, T.J., Uehlein, B.L., Quinter, R.C., Summers, M.A., 1993. Naphthenic acid corrosion in a re®nery setting. Mate-rials Performance 32, 50±55.
Bailey, N.J.L., Jobson, A.M., Rogers, M.A., 1973. Bacterial degradation of crude oil: Comparison of ®eld and experi-mental data. Chemical Geology 11, 203±211.
Barth, T., Bjùrlykke, K, 1993. Organic acids from source rock maturation: generation potentials, transport mechanisms and relevance for mineral diagenesis. Applied Geochemistry 8, 325±337.
Barth, T., Moen, L.K., Dyrkorn, C., 1998. Comparison of acid numbers and carboxylic acid molecular compositions in biode-graded and normal crude oils. Preprints of the American Che-mical Society, Division of Petroleum Chemistry 3, 134±136. Behar, F.H., Albrecht, P., 1984. Correlations between carboxylic
acids and hydrocarbons in several crude oils. Alteration by biodegradation. Organic Geochemistry 6, 597±604.
Bennett, B., Bowler, B.F.J., Larter, S.R., 1996. Determination of C0±C3 alkylphenols in crude oils and waters. Analytical Chemistry 68, 3697±3702.
Bennett, B., Abbott, G.D., 1999. A natural pyrolysis experi-ment- hopanes from hopanoic acids? Organic Geochemistry 30, 1509±1516.
Buchholz, B., Laczko, E., Pfennig, N., Rohmer, M., Neunlist, S., 1993. Hopanoids of a recent sediment from Lake Con-stance as eutrophication markers. FEMS Microbiology and Ecology 102, 217±223.
Carson, J., Graham, D.W., 1965. Isolation of isoprenoid acids from a California petroleum. Tetrahedron 21, 471±483. Connan, J., 1984. Biodegradation of crude oils in reservoirs. In:
Brooks, J., Welte, D.H. (Eds.), Advances in Petroleum Geo-chemistry 1. Academic Press, London, pp. 300±330. Connan, J., Restle, A., Albrecht, P., 1979. Biodegradation of
crude oil in the Aquitaine Basin. In: Douglas, A.G., Max-well, J.R. (Eds.), Advances in Organic Geochemistry, 1980. Pergamon, Oxford, pp. 1±18.
Derungs, W.A., 1956. Naphthenic acid corrosion Ð an old enemy of the petroleum industry. Corrosion 12, 41±46. Fan, P.T., 1991. Characterisation of naphthenic acids in
petro-leum by fast atom bombardment mass spectrometry. Energy and Fuels 5, 371±375.
Farrimond, P., Head, I.M., Innes, H.E., 2000. Environmental in¯uence on the biohopanoid composition of recent sedi-ments. Geochimica et Cosmochimica Acta 64, 2885±2992. Farrimond, P., Stoddart, D.P., Jenkyns, H.C., 1994. An
organic geochemical pro®le of the Toarcian anoxic event in northern Italy. Chemical Geology 111, 17±33.
Gibson, D.T., 1991. The role of oxygenases in the microbial oxidation of aromatic compounds. Abstracts of the Amer-ican Chemical Society 201, 46.
Gransch, J.A., Posthuma, J., 1974. On the origin of sulphur in crudes. In: In. Tissot, B., Bienner, F. (Eds.), Advances in Organic Geochemistry, 1973. Editions Technip, Paris, pp. 727±739.
Grbic-Galic, D., Vogel, T.M., 1987. Transformation of toluene and benzene by mixed methanogenic cultures. Applied Environmental Microbiology 53, 254±260.
Heider, J., Spormann, A.M., Beller, H.R., Widdel, F., 1999. Anaerobic bacterial metabolism of hydrocarbons. FEMS Microbiology Review 22, 459±473.
Horstad, I., Larter, S.R., 1997. Petroleum migration, alteration and remigration within Troll ®eld, Norwegian North Sea. AAPG Bulletin 81, 222±248.
Howard, P.H., Boethling, R.S., Jarvis, W.F., Meylan, W.M., Michalenko, E.M., 1991. Handbook of Environmental Degradation Rates. Lewis Publishers, Chelsea, MI. Innes, H.E., Bishop, A.N., Head, I.M., Farrimond, P., 1997.
Preservation and diagenesis of hopanoids in Recent lacus-trine sediments of Priest Pot, England. Organic Geochem-istry 26, 565±576.
Ioppolo, M., Alexander, R., Kagi, R.I., 1992. Identi®cation and analysis of C0±C3phenols in some Australian crude oils. Organic Geochemistry 18, 603±609.
JaeÂ, R., Albrecht, P., Oudin, J.L., 1988a. Carboxylic acids as indicators of oil migration: I. Occurrence and geochemical signi®cance of the (17bH,21bH) C30hopanoic acid in geolo-gical samples. Organic Geochemistry 13, 483±488.
JaeÂ, R., Albrecht, P., Oudin, J.L., 1988b. Carboxylic acids as indicators of oil migration: II. Case of the Mahakan Delta, Indonesia. Geochimica et Cosmochimica Acta 52, 2599± 2607.
JaeÂ, R., Gardinali, P.R., 1990. Generation and maturation of carboxylic acids in ancient sediments from the Maracaibo Basin, Venezuela. Organic Geochemistry 16, 211±218. JaeÂ, R., Gallardo, M.T., 1993. Application of carboxylic-acid
biomarkers as indicators of biodegradation and migration of crude oils from the Maracaibo Basin, Western Venezuela. Organic Geochemistry 20, 973±984.
Jayaraman, A., Singh, H., Lefebvre, Y., 1986. Naphthenic acid corrosion in petroleum re®neries: a review. Revue de l'Insti-tut Francais du Petrole 41, 265±274.
Kaphammer, B., Kukor, J.J., Olsen, R.H., 1990. Cloning and characterisation of a novel toluene degradative pathway from Pseudomonas picketti. Abstracts of the 90th Meeting of The American Society of Microbiology. pp. 243.
Lochte, H.L., Littman, E.R., 1955. The Petroleum Acids And Bases. Chemical Publishing Company, New York.
Lovell, J.B., 1990. Cenozoic. In: Glennie, K.W. (Ed.), Intro-duction to the Petroleum Geology of the North Sea. Black-well, Oxford, UK, pp. 273±293.
Mackenzie, A.S., Wol, G.A., Maxwell, J.R., 1983. Fatty acids in biodegraded petroleums. Possible origins and signi®cance. In: Bjorùy, M. et al. (Eds.), Advances in Organic Geochem-istry 1981. John Wiley, Chichester, UK, pp. 637±649. Nascimento, L.R., ReboucËas, L.M.C., Koike, L., Reis, F. de
A.M., Soldan, A.L., Cerqueira, J.R., Marsaioli, A.J., 1999. Acidic biomarkers from Albacora oils, Campos Basin, Bra-zil. Organic Geochemistry 30, 1175±1191.
Olsen, S. D., 1998. The relationship between biodegradation, total acid number (TAN) and metals in oils. Preprints of the American Chemical Society, Division of Petroleum Chem-istry 3, 142±145.
Palmer, S.E., 1993. Eect of biodegradation and water washing on crude oil composition. In: Engel, M.H., Macko, S.A. (Eds.), Organic Geochemistry. Plenum Press, New York, pp. 511±533. Peters, K.E., Moldowan, J.M., 1993. The Biomarker Guide.
Prentice Hall, New Jersey.
Piehl, R.L., 1988. Naphthenic corrosion in crude distillation units. Materials Performance 27, 37±43.
Reinsel, M.A., Borkowski, J.J., Sears, J.T., 1994. Partition coecients for acetic, propionic and butyric acids in a crude oil/water system. Journal of Chemical and Engineering Data 39, 513±516.
Ries-Kautt, M., Albrecht, P., 1989. Hopane derived triterpe-noids in soils. Chemical Geology 76, 143±151.
Rohmer, M., Bisseret, P., Neunlist, S., 1992. The hopanoids, prokaryotic triterpenoids and precursors of ubiquitous molecular fossils. In: Moldowan, J.M., Albrecht, P., Philp, R.P. (Eds.), Biological Markers in Sediments and Petroleum. Prentice Hall, Englewood Clis, New Jersey, pp. 1±17. Rueter, P., Rabus, R., Wilkes, H., Aeckersberg, F., Rainey,
F.A., Jannasch, H.W., Widdel, F., 1994. Anaerobic oxida-tion of hydrocarbons in crude oil by new types of sulphate-reducing bacteria. Nature 372, 455±457.
Samadova, F.I., Guseinova, B.A., 1993. Distribution of acids and phenols in Azerbaidzhan crudes. Chemistry and Tech-nology of Fuels and Oils 28, 515±517.
Seifert, W.K., 1975. Carboxylic acids in petroleum and sedi-ments. In: Herz, W., Grisebach, H., Kirby, G.W. (Eds.), Progress in the Chemistry of Organic Natural Products 32. Springer Verlag, Berlin, pp. 1±49.
Seifert, W.K., Teeter, R.M., 1970. Identi®cation of polycyclic naphthenic, mono-, and diaromatic crude oil acids. Analy-tical Chemistry 42, 180±189.
Singer, M.E., Finnerty, W.R., 1984. Microbial metabolism of straight-chain and branched alkanes. In: Atlas, R.M. (Ed.), Petroleum Microbiology. Macmillan, New York, pp. 1±61. Stetter, K.O., Huber, R., BloÈchl, E., Kurr, M., Eden, R.D.,
Fielder, M., Cash, H., Vance, I., 1993. Hyperthermophilic archaea are thriving in deep North Sea and Alaskan oil reservoirs. Nature 365, 743±745.
Taylor, P., 1994. Controls on the occurrence of phenols in Petroleums and Water. PhD thesis, The University of New-castle upon Tyne, England.
Taylor, P., Larter, S.R., Jones, D.M., Dale, J., Horstad, I., 1997. The eect of oil-water-rock partitioning on the occur-rence of alkylphenols in petroleum systems. Geochimica et Cosmochimica Acta 61, 1899±1910.
Tomczyk, N.A., Winans, R.E., Shinn, J.H., Robinson, R.C. Results of treatment on acid content in a California heavy crude. Preprints of the American Chemical Society, Division of Petroleum Chemistry. 3;1998;123-126.
Turnbull, A., Slavcheva, E., Shone, B., 1998. Factors control-ling naphthenic acid corrosion. Corrosion 54, 922±930. Vogel, T.M., Grbic-Galic, D., 1986. Incorporation of oxygen
from water into toluene and benzene during anaerobic fer-mentation. Applied Environmental Microbiology 52, 200± 202.
Volkman, J.K., Alexander, R., Kagi, R.I., Rowland, J., Shep-pard, P.N., 1984. Biodegradation of aromatic hydrocarbons in crude oil from the Barrow sub-basin of Western Australia. Organic Geochemistry 6, 619±632.
Watson, J.S., Jones, D.M., Swannell, R.P.J., 1999. Formation of carboxylic acids during biodegradation of crude oil. In: Alleman, B.C., Leeson, A. (Eds.), In Situ Bioremediation of Petroleum Hydrocarbon and other Organic Compounds. Battelle, Columbus, Ohio, pp. 251±255.
Wilkes, H., Willsch, H., Rabus, R., Aeckersberg, F., Rueter, P., Widdel, F., 1995. Compositional changes of crude oils upon anaerobic degradation by sulphate reducing bacteria. In: Grimalt, J., Dorronsoro, C. (Eds.), Organic Geochemistry: Developments and Applications to Energy, Climate, Envir-onment and Human History. Selected papers from the 17th International Meeting on Organic Geochemistry, 4±8 Sep-tember 1995, San Sebastian. AIGOA. Donostia-San Sebastian, Spain. pp. 321±323.
Zengler, K., Richnow, H.H., Rossello-Mora, R., Michaelis, W., Widdel, F., 1999. Methane formation from long-chain alkanes by anaerobic micro-organisms. Nature 401, 226± 269.
