International Conference on Gas Hydrates (ICGH) (6th : 2008)


Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


59278-5577.pdf [ 262.93kB ]
JSON: 59278-1.0041093.json
JSON-LD: 59278-1.0041093-ld.json
RDF/XML (Pretty): 59278-1.0041093-rdf.xml
RDF/JSON: 59278-1.0041093-rdf.json
Turtle: 59278-1.0041093-turtle.txt
N-Triples: 59278-1.0041093-rdf-ntriples.txt
Original Record: 59278-1.0041093-source.json
Full Text

Full Text

Proceedings of the 6th International Conference on Gas Hydrates (ICGH 2008), Vancouver, British Columbia, CANADA, July 6-10, 2008, ISOLATION AND MOLECULAR IDENTIFICATION OF HYDRATE SURFACE ACTIVE COMPONENTS IN PETROLEUM ACID FRACTIONS Kristin Erstad ∗1, Sylvi Høiland 2, Tanja Barth 1 and Per Fotland 3 1 University of Bergen, Department of Chemistry, N-5007 Bergen, Norway 2 SINTEF Petroleum Research, N-5008 Bergen, Norway 3 StatoilHydro R&D, N-5254 Bergen, Norway ABSTRACT The anti-agglomerating hydrate behavior observed for some crude oils has previously been related to crude oil composition and to surface adsorption mechanisms. Petroleum acids derived from some crude oils have been found able to convert systems with initially high risk of plugging into easily flowable dispersions. In this work, acid fractions are isolated from three oils with low tendency to form hydrate plugs and from two oils associated with high risk of hydrate plugging by using an ion-exchange resin. The extracts are further separated into four sub-fractions by solid phase extraction (SPE). The chemical composition of the fractions is studied by means of HPLC, GPC, FTIR- and UV/VIS spectroscopy and elemental analysis. The distribution of chemical compound classes in the fractions differs between the non-plugging and plugging oils, and the differences are most distinctive in one of the sub-fractions. The results imply that acid sub-fractions holding a significant proportion of more weakly polar compounds, like ester functionalities, are important for how the hydrate surfaces and the oil phase interact. Keywords: petroleum acids, hydrates, natural inhibiting compounds NOMENCLATURE B-oils Biodegraded oils DBE Double Bond Equivalents DCM Dichloromethane ESI-MS Electrospray ionization-mass spectrometry FTIR Fourier Transform InfraRed GPC Gel Permeation Chromatography HPLC High Performance Liquid Chromatography LDHI Low Dosage Hydrate Inhibitors MeOH Methanol S-oils Sweet, non-biodegraded oils SPE Solid Phase Extraction TAN Total Acid Number TBN Total Base Number UV/VIS Ultraviolet/Visible ∗Corresponding author: Phone: +47 55 58 34 80 Fax: +47 55 58 94 90 Email: INTRODUCTION In petroleum production using multiphase transport systems, conditions leading to stable gas hydrate for- mation can occur. The potential danger of forming hydrate plugs that can effectively plug the pipelines represents a huge challenge in oil production. On the Norwegian continental shelf, the main strategy for preventing hydrate plugging has traditionally been to inject large volumes of methanol or glycol as hy- drate inhibitors, in order to keep the system outside of the stable hydrate formation region. This is a very expensive strategy, and efforts have been made on trying to replace the alcohols with so-called low dosage hydrate inhibitors (LDHI) as a more cost ef- fective way of preventing plugs. These can be either kinetic inhibitors that delay the hydrate crystalliza- tion, or anti-agglomerants where hydrate formation still occurs, but the anti-agglomerants prevent the agglomeration of hydrate crystals into larger clus- ters. Instead the hydrates forms small dispersed par- ticles that are easily transported in the fluid [1–3]. As of today, the use of LDHIs on the Norwegian continental shelf is restricted due to toxicity. Several chemical companies are currently trying to develop environmental friendly alternatives [4]. It is observed that certain crude oils are less associ- ated with risk of hydrate plugging than others, even under thermodynamically stable hydrate formation conditions. This strongly indicates that natural in- hibiting compounds that may be anti-agglomerants are indigenously present in some oils. This feature has been pointed out by several authors [2, 3, 5–8]. The anti-agglomerating behavior of some crude oils is explained by the presence of indigenous surface active compounds that are able to adsorb to the hy- drate surface and create a lyophilic or oil-wet sur- face. This reduces the possibility of hydrogen- bonding between hydrate particles and results in flocculation rather than agglomeration. The process is associated with promotion of water-in-oil emul- sions stabilized by oil-wet hydrate particles [7]. Surface and interfacially active components in crude oils span over a large range of chemical structures and molecular weights, including asphaltenes, resins and naphthenic acids [9,10]. “Naphthenic acids” is a common term for an isomeric mixture of carboxylic acids containing one or several saturated alicyclic rings. In petroleum terminology “naphthenic acids” has become more loosely used to describe the whole range of organic acids found within crude oils [11]. Recently, several authors have emphasized the effect of petroleum acids on hydrate plugging properties. Hemmingsen et al. [3] found that the plugging po- tential of crude oils increased after modifying them by removing the acidic compounds. Bergflødt [6] studied the hydrate inhibiting properties of commer- cial and petroleum naphthenic acids as additives in model systems, and showed that they display an dis- persive effect on the systems. Høiland et al. [12] showed that addition of an acid extract isolated from a crude oil with low tendency to form hydrate plugs into a crude oil with high tendency of forming hy- drate plugs, was able to transform the hydrates from plugging to dispersed. In a recent work, acid ex- tracts from oils with low hydrate plugging potential have been separated into sub-fractions [13]. It is ob- served that even the acid sub-fractions alone are able to impose an alteration in the wettability of freon hy- drates, even at very low concentrations. Although the effect of surface active petroleum acids on hydrate plug prevention is well documented, little is known about the specific structures of active com- pounds in the extracted acid fractions. Whether they exist as certain key compound structures, or whether the inhibiting effect is a result of interplay between several compound types present in the acid fraction, or even whether the bulk crude oil solvent proper- ties impose an influence on the extent of surface ac- tivity, remains to be answered. Borgund et al. [14] observed that the amount of phenolic compounds found in the acid extracts has a strong negative cor- relation with wettability of hydrates, meaning that these compounds are associated with an agglomer- ating/plugging effect. The chemical characterization of the isolated acid fractions represents a considerable challenge due to the high degree of complexity in their composition, in addition to being poorly amenable to commonly used analytical characterization techniques [8, 15]. The ultimate aim and objective of the ongoing work is to identify and determine the chemical structure of the active natural inhibiting components. In this work we isolate and characterize hydrate surface active acid fractions chemically, and investigate if there are differences in the acid fractional composi- tion between oils with low and high plugging poten- tial. The acid extracts are separated into four sub- fractions that are studied separately, providing more resolved matrices. MATERIALS AND METHODS Materials Sample description Acid extracts and sub-fractions are isolated from three biodegraded crude oils with low plugging tendency, and one biodegraded oil with high plugging tendency. In addition, an acid extract from a non-biodegraded oil with high plugging potential is included. The oils are labeled as B-oils (biodegraded) or S-oils (sweet, non-biodegraded), followed by a digit and a letter indicating the field and well number (or different batches from the same well). In addition, the extracts isolated from the oils are assigned a suffix to indicate the type of extract (whole acid extract: acid or sub-fractions: SPEA-SPED. Other characteristics like chemical and physical properties of the crude oils have been published by Høiland et al. [7], Borgund et al. [14] and Barth et al. [16]. Some of the crude oil properties are presented in Table 1. The crude oils are stored in dark containers and in an inert nitrogen atmosphere to prevent oxidation reactions. Prior to sampling the containers are heated to 60◦C and thoroughly shaken to achieve homogeneous samples. All the crude oils originate from the Norwegian continental shelf and are supplied by StatoilHydro. Chemicals All solvents used are of p.a. or HPLC quality. Standards used for GPC are nine polystyrenes pur- chased from Polymer Laboratories (purity 99.5%), n-octadecanoic acid (purity 99+%, Sigma) and 1,12- dodecanedioic acid (purity 99.5+%, Fluka). Experimental methods A general overview of the methods used in the analytical work is given in Figure 1. Crude oil Acid extract SPE- fractions A, B, C and D HPLC Elementalanalysis UV/VIS spectroscopy FTIR spectroscopyGPC Acid extraction • Quantification • Fractionation on SPE-column Quantification Figure 1: Flowsheet of analytical work. Acid extraction The acid extracts are isolated by use of an ion exchange method, described by Mediaas et al. [17], and also recently reported by Borgund et al. [14]. Separation of acid extracts into sub-fractions The procedure has been developed by Borgund et al. [14]. The acid extracts are fractionated into sub-fractions of different composition and polarity using SPE cyano columns (Isolute SPE Cyano (end-capped), International Sorbent Technology, UK, 1000mg sorbent mass, 3 mL reservoir volume). Approximately 30 mg of the acid extract (dissolved in 0.075 mL DCM:MeOH 93:7 (v/v)) is applied onto the column, and the four SPE-fractions are obtained by eluting with the following solvent mixtures: SPEA: hexane:DCM 90:10 (v/v) (20 mL), SPEB: hexane:DCM 90:10 (v/v) (30 mL), SPEC: DCM:MeOH 93:7 (v/v) (20 mL) and SPED: MeOH:DCM 70:30 (v/v) (2 mL) and MeOH:formic acid 95:5 (v/v) (10 mL). The eluent solvents are removed from the fractions by evaporation using a gentle stream of nitrogen gas. The SPE-fractions are then redissolved in DCM:MeOH 93:7 (v/v). Quantification The amount of material in acid extracts and SPE-fractions is determined gravimetrically by de- positing 5 or 10 µL of the samples in DCM:MeOH 93:7 (v/v) solution on a weighing pan of a Cahn electrobalance (range 0.0001-2 mg), letting the solvent evaporate for exactly 20 minutes and noting the weight of the nonvolatile residue. HPLC procedure The HPLC procedure has been developed and de- scribed in detail by Borgund et al. [18]. The method separates organic acid mixtures into four groups of compound classes: weakly polar compounds, saturated carboxylic acids, phenolic compounds and polyfunctional compounds. Samples are dissolved in DCM:MeOH 93:7 (v/v) to concentrations in the range 0.3-2.5 mg/mL, dependent on detector response of each individual sample. A total of 31 blank runs are performed along with the sample runs, revealing a contamination peak eluting at the same place in the chromatogram as the polyfunc- tional compound peak. The contamination peak is very small though, and is calculated to contribute 1±1 mV to the polyfunctional compounds peak areal. When quantifying the peak areas, this value is subtracted from every sample chromatogram. GPC procedure Gel permeation chromatography gives the molecu- lar weight distribution of the samples. A detailed description of the GPC procedure is given by Borgund et al. [14]. A linear relationship between the logarithm of the molecular weights of the Table 1: Crude oil propertiesa. Crude Asphaltene Density TAN TBN Biodeg. Wettabilityc Plugging tendencyd oil (mg/g oil) (g/mL) (mg KOH/g oil) (mg KOH/g oil) level ∆ϕ∗ (plug/dispersion) B4a 3.785b 0.895b 1.02 1.21b 8b 0.35 dispersion B4c 0.240 0.897 1.40 1.10 2 0.38 dispersion B2b 5.810 0.918 2.66 0.89 6 0.26 dispersion B1c 14.136b 0.942b 2.18 2.79b 2b -0.32b plugb S3b 2.800 0.833 0.16 1.16 0 -0.06 plug a The data are taken mostly from Høiland et al. [7], Borgund et al. [14] and Barth et al. [16]. b Measured on a previous sampled batch of the oil. c Positive values of ∆ϕ∗ correspond with low plugging tendency and negative values with high plugging tendency [7]. d From combined field experience and tests performed in high pressure sapphire cells at StatoilHydro R&D centre. standards and their retention times is observed, with a correlation coefficient R2 of 0.98 (data not shown). Three positions on the chromatographic peaks are assessed, giving the molecular weight distribution of the sample. The retention time corresponding to the maximal intensity of the peak gives the medium calculated molecular weights MWmed . In addition, the retention times at the width at half the height on both sides of the peak are assessed, representing high molecular weights MWmax at the leading edge and low molecular weights MWmin at the trailing edge of the peak. FTIR spectroscopy procedure The FTIR spectra are recorded on a Nicolet Protege 460 FTIR spectrometer with a diamond attenuated total reflection (ATR) Dura sampler cell (from SensIR). The ratios between some characteristic absorption bands in the spectra are calculated: the aliphatic, asymmetric stretch C-H stretch band in the region 2920-2930 cm−1, the carbonyl C=O stretch band in the range 1700-1740 cm−1 and the aromatic C=C stretch vibration band in the region between 1580-1610 cm−1. UV/VIS spectroscopy procedure A Hewlett Packard HP 8453 spectrophotometer is used for recording UV/VIS spectra of the collected acid extracts and SPE-fractions. The absorbance is measured in the range 220-600 nm, using a quartz cuvette (1×1 cm). The samples are prepared in concentrations of approximately 0.01 mg/mL in DCM. The ratio between the absorption at the wavelength of maximum absorption on the curves (λmax) and the absorbance at a fixed wavelength (260 nm) is calculated, providing a measure of the aromaticity of the samples. Elemental analysis The elemental composition of the samples (except from SPED-fractions) is measured on a VarioEL III CHNS elemental analyzer. RESULTS AND DISCUSSION Amounts of extractable acids Figure 2 shows the amount of extractable acids of the oils. The acid content in the non-biodegraded oil S3b is very small compared to all the biode- graded oils. This is in accordance with the process of biodegradation being known to increase the acid- ity of crude oils [16, 19]. 0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0 B4a B4c B2b B1c S3b E x t r a c t e d a c i d s  ( m g / g o i l ) Figure 2: Amount of extractable acids from crude oils. The values given are averages of previous extractions [14] and new extractions included in the data set, shown with calculated standard deviations. Relative distributions of material in SPE- fractions The percentage distributions of organic material found in the four SPE-fractions from four of the acid extracts are shown in Figure 3 (the S3b acid ex- tract was not fractionated due to the minute amounts of material extracted from this oil). Overall the SPEA-fraction represents the major part of the or- ganic material. Smaller but significant parts elute in the SPEB- and SPEC-fractions, while the SPED- fraction contains only a very small fraction of the total acid extracts. The B1c extract (high plugging potential) contains a larger relative amount of SPEC and a lower amount of SPEA compared to the three oils with low plugging potential B4a, B4c and B2b. 0,0 10,0 20,0 30,0 40,0 50,0 60,0 70,0 80,0 90,0 100,0 SPEA SPEB SPEC SPED W e i g h t  c o m p o s i t i o n ( % ) B4a B4c B2b B1c SP A B C D Figure 3: Relative amounts of organic material found in the SPE-fractions, determined gravimetrically. The SPE-fractionation is a preparation procedure which has limited capacity to resolve the compo- nents in the samples. HPLC separates the SPE- fractions further into chemical compound classes present within each SPE-fraction. Typical pro- files are shown in Figure 4 (SPE-fractions from B4a). Relative amounts of compound types cal- culated from the chromatographic peak areas are shown in Figure 5 (whole acid extracts and their SPE-fractions). In general, the first sub-fraction, SPEA, is quite similar in composition to the whole acid extract, holding a majority of saturated car- boxylic acids and low amounts of polyfunctional compounds. SPEB is an intermediate fraction with a more equal contribution from these compound classes in addition to weakly polar and phenolic compounds. SPEC is dominated by the presence of polyfunctional compound types, while SPED mainly holds weakly polar- and polyfunctional compounds. The SPEA- and SPEC-fractions of the plugging oil B1c are very similar to the three non-plugging oils B4a, B4c and B2b. However, a distinct difference in composition is observed in the SPEB- and SPED- fractions where B1c contains a considerable higher percentage of polyfunctional compounds. More- 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 47.0 -200 500 1 000 1 400 _ [ y ] g _ _ _ mV min 1 - 6.975 2 - 18.270 3 - 23.395 4 - 32.752 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 47.0 -200 500 1 000 1 400 mV min 1 - 7.019 2 - 18.328 3 - 28.872 4 - 32.897 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 47.0 -200 500 1 000 1 400 mV min 1 - 6.983 2 - 18.119 3 - 20.251 4 - 32.621 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 47.0 -50 100 200 300 mV min 1 - 7.014 2 - 18.205 3 - 27.454 4 - 32.797 Weakly polar compounds Saturated carboxylic acids Phenolic compounds Polyfunctional compounds SPED SPEC SPEB SPEA Figure 4: HPLC chromatographic profiles of SPE- fractions isolated from B4a oil. The chromatograms are divided into chemical compound classes according to Borgund et al. [18] (see upper chromatogram). over, it is interesting that for all samples studied the B1c fractions contains the largest relative amounts of polyfunctional compounds. Like B1c, the S3b acid extract is isolated from an oil with high plugging tendency, but it is the only non-biodegraded oil in the data set (see Table 1). No SPE-fractions are available, which highly restricts the basis of comparison. This is due to the minimal amounts of extractable acids in S3b (Figure 2), mak- ing SPE-fractionation impracticable. Figure 5 (bar chart at the top) shows that the S3b whole acid ex- tract compound class distribution resembles the non- plugging oils. Molecular weight range of acid extracts and SPE-fractions The results from GPC are presented in Table 2. The values given are averages of 2-3 runs of each sam- ple, and they are approximated to the nearest 50 g/mol for the minimum and medium MW-values and to the nearest 100 g/mol for the high MW- values. A very general interpretation of the data Acid extracts 0,0 10,0 20,0 30,0 40,0 50,0 60,0 70,0 80,0 90,0 100,0 weakly polar compounds saturated carboxylic acids phenolic compounds polyfunctional compounds R el at iv e co m po si tio n (% ) SPEA 0 10 20 30 40 50 60 70 80 90 100 weakly polar compounds saturated carboxylic acids phenolic compounds polyfunctional compounds R el at iv e co m po sit io n (% ) SPEB 0 10 20 30 40 50 60 70 80 weakly polar compounds saturated carboxylic acids phenolic compounds polyfunctional compounds R el at iv e co m po si tio n (% ) SPEC 0 10 20 30 40 50 60 70 80 90 100 110 weakly polar compounds saturated carboxylic acids phenolic compounds polyfunctional compounds R el at iv e co m po sit io n (% ) SPED 0 10 20 30 40 50 60 70 80 90 weakly polar compounds saturated carboxylic acids phenolic compounds polyfunctional compounds R el at iv e co m po si tio n (% ) B4a B4c B2b B1c S3b Figure 5: Relative amounts of compound classes present in acid extracts and SPE fractions using HPLC. suggests a spread of molecular weights in the range between approximately 150 and up to 2000 g/mol. The values calculated from the maximum intensity of the peaks (MWmed) represents the majority of the molecules present, giving values between 250- 800 g/mol as a reasonable estimate. This range is quite similar to molecular weights previously re- ported for petroleum naphthenic acids [8,16]. Look- ing at the MWmed-values, all the acid extracts have lower molecular weights than their respective SPE- fractions. One possible explanation may be that the degree of self-association between the acidic molecules present is less in the whole extract than in its SPE-fractions. Aggregation of petroleum as- phaltenes is well-documented [20]. Recently such behavior has also been observed among organic acids in bitumen and bitumen distillates [21]. The MWmed-values of SPEA and SPEB from B1c are somehow lower compared to the other samples, and have the highest value of all in SPED, but also differ- ences between the other three samples are observed. The precision of the MW-estimates is based on the observed standard deviations of the parallel runs. GPC is, however, not an exact method for molecular weight determination of petroleum samples. In prin- ciple the method separates molecules based solely on molecular size. However, other factors such as molecular geometry can affect the separation order [16]. Also it must be recognized that petroleum con- stituents have a range of polarity and aromaticity. This might result in interaction with the gel surface of the column to different degrees, causing adsorp- tion effects. Another problem is the lack of realis- tic standards of chemical nature similar to petroleum constituents [20]. For these reasons the data are in- terpreted with great caution. Still, the data gives valuable information on the molecular weight range of the samples, and a basis for internal comparison. Characterization of acid extracts and SPE- fractions by spectroscopic methods and ele- mental analysis Interpretations of the FTIR spectra are based on the work of Williams and Fleming [22] and Coates [23]. Generally, when comparing the FTIR spectra from acid extracts and SPE-fractions of B4a, B4c and B2b (from the low hydrate plugging potential oils) lit- tle variation is observed. However, the spectra of each individual SPE-fraction support the trend that Table 2: Molecular weight ranges of the samples from GPC. For the S3b sample only data on whole acid extract is available. Sample Fraction MWmin MWmed MWmax (g/mol) (g/mol) (g/mol) B4a acid 200 350 1100 B4a SPEA 250 500 1200 B4a SPEB 200 400 1200 B4a SPEC 200 550 2100 B4a SPED 300 550 1400 B4c acid 150 350 1000 B4c SPEA 250 550 1200 B4c SPEB 200 500 1200 B4c SPEC 300 800 2200 B4c SPED 250 400 1200 B2b acid 150 250 800 B2b SPEA 250 500 900 B2b SPEB 200 350 800 B2b SPEC 200 500 1200 B2b SPED 250 350 1000 B1c acid 150 250 1000 B1c SPEA 100 250 1000 B1c SPEB 150 250 1000 B1c SPEC 200 500 1700 B1c SPED 250 600 1900 S3b acid 150 250 800 is observed in the HPLC analysis (Figure 5); namely that the four SPE-fractions vary much in chemical composition. When comparing these FTIR spectra with the B1c spectra, some major deviations are ob- served, as illustrated in Figure 6, comparing B4a and B1c. Generally the FTIR spectra of the acid extracts have very similar features to SPEA. The characteristic absorption band at approximately 1705-1706 cm−1, represents C=O (carbonyl) stretch from carboxylic acids. This observation is supported by the HPLC results that show that the acid extracts and SPEA- fractions mainly contain saturated carboxylic acids (Figure 5). For the SPEA-spectra, this band has be- come more intense, and for B4a the most intense band in the spectra. In SPEB this band is less in- tense relative to the bands from C-H stretch vibra- tions around 2920-2930 cm−1 for the B4a sample, while in the B1c spectra the intensity of the band is about the same as in the SPEA-spectra. Also, an aromatic signal around 1600 cm−1 becomes visi- ble in the SPEB-spectra. The most obvious features with the spectra from SPEC are the increased absorp- tion above 3000 cm−1, representing O-H stretch. This band seems to be more distinct in B4a than in B1c. Also the C=O stretch band is less distinct and the aromatic signal more prominent in SPEC (in the spectra from B1c the aromatic bond being weaker than the carbonyl). Compounds holding both acid and phenolic functionalities are probably present in this fraction. Knowing from the HPLC results (Fig- ure 5) that the SPEC-fractions hold almost exclu- sively polyfunctional compounds, it is likely that the peaks from FTIR are parts of these more com- plex molecules, rather than from isolated carboxylic acids and phenolic unit structures. Another observa- tion is that in SPEC the C=C aromatic bond at 1583- 1592 cm−1 is located at lower wavenumber than any of the other SPE-fractions (1598-1608 cm−1). This is a strong indication that the aromatic ring is conjugated, meaning that polyaromatic compounds and/or substitution on the aromatic ring most likely are present to a larger extent than in the other frac- tions. In SPED the C=O band is located at higher frequencies, around 1735 cm−1 in the spectrum of B4a, and with strong intensity (also found in B4c and B2b SPED-spectra). This indicates that com- pounds with ester functionalities are more dominant in this fraction. However, it can be seen from Figure 6 that for B1c this peak is much less distinct, and also located at lower frequency (1710 cm−1). This indicates that the composition of the SPED-fraction from B1c differs significantly from B4a, B4c and B2b. Again, relating this observation to the HPLC results it is seen that the SPED-fraction of B1c is very different from B4a, B4c and B2b. The HPLC results show that for these three samples the main part of the SPED-fraction consists of weakly polar compounds. Testing of simple long chained methyl ester standard compounds on HPLC shows that they elute as weakly polar compounds in the HPLC chro- matogram, supporting the findings from FTIR and HPLC. Still the exact structures of the ester com- pounds in the SPED-fractions remain to be deter- mined. Another observation is that in the SPED- fraction of B1c the intensity of the aromatic C=C stretch signal is stronger than the C=O stretch, while in all the three non-plugging oils the C=O stretch signal is more intense. The UV/VIS spectra do not provide detailed in- formation on structures of molecules present in the samples, but give a general measure of the extent of polycondensation and substitution on the aromatic rings present. All the spectra have a sharp absorption maximum peak between 228 and 239  94  96  98 100 % R  90  95  100 % R  2000   4000 Wavenumbers (cm-1)  85  90  95  100 % R  90  95 % R  2000   4000 Wavenumbers (cm-1) SPEA  96  98  100 % R  94  96  98  100 % R  2000   4000 Wavenumbers (cm-1)  95 100 % R  90  95  100 % R  2000   4000 Wavenumbers (cm-1)  98  99  100 % R  90  95 % R  2000   4000 Wavenumbers (cm-1) SPEB SPEC SPED C=O stretch carboxylic acids CH2 asym. stretch C=O stretch carboxylic acids CH2 asym. stretch C=C arom. stretch C=O stretch carboxylic acids CH2 asym. stretch C=C arom. stretch CH2 asym. stretch C=O stretch ester (1735 cm-1) C=C arom. stretch C=O stretch carboxylic acids O-H stretch Phenols? O-H stretch Phenols? Acid extract CH2 asym. stretch C=O stretch carboxylic acids Figure 6: FTIR spectra of SPE-fractions from B4a (non-plugging) and B1c (plugging). All the spectra are stacked for comparison, showing B4a at the top and B1c at the bottom. nm, and with a less sharp shoulder of secondary absorption from approximately 250 nm with a de- scending tail continuing towards the visible part of the spectrum (above 400 nm). It is difficult to make quantitative predictions about substitution patterns based on UV/VIS spectra of complex mixtures. However, it is recognized that the observed patterns are generally typically for e.g. a monosubstituted benzene ring or a two-ring aromatic compound such as naphthalene. Relatively non-polar substituents such as for instance the acetoxyl group in an ester have only minor effects on the shape and position of the absorption peaks of the parent hydrocarbon compound [22]. For all the SPEC-spectra the secondary absorption is less distinguishable from the absorption maximum, which can be related to increased degree of aromaticity and complexity, in agreement with HPLC and FTIR results. The two spectra of SPEB- and SPEC-fractions from B4c illustrate this (shown in Figure 7). The lack of absorption maxima above 350 nm shows that three- or four aromatic ring systems most likely are absent. No differentiation between low and high hydrate plugging potential oils can be detected from the UV/VIS spectra. In Table 3 measured peak ratios from both FTIR and UV/VIS spectra are shown, along with the weight percentages and atomic ratios measured from ele- mental analysis. Also, by combining the GPC data (from Table 2) with elemental analysis data, aver- age formulas and double bond equivalents (DBE) are proposed. The C=O/CH2 asym. ratio represents a rough measure of the carbonyl concentration rela- tive to aliphatic structures. The whole acid extracts and the SPEA- and SPED-fractions of the three sam- Wavelength (nm)250 300 350 400 450 500 A b s o r b a n c e  ( A U ) 0 0.1 0.2 0.3 0.4 Wavelength (nm)250 300 350 400 450 500 A b s o r b a n c e  ( A U ) 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 SPEB SPEC Figure 7: UV/VIS spectra of SPE-fractions isolated from B4c oil. Topp: SPEB. Bottom: SPEC . ples B4a, B4c and B2b have larger relative content of C=O compared with the respective fractions from B1c and S3b. The O/C ratio is not directly compa- rable to the C=O/CH2 asym. ratio as this reflects the total oxygen content relative to all carbon types. No systematically trends are seen for the O/C ratios of the samples, neither between the different oils nor within the fractions from the same oil. Overall the O/C ratios vary from 0.107 to 0.293, which is in the range of many oxygenated organic compounds. One sample (whole acid extract of B1c) deviates much from the others by having a much larger oxygen con- tent (O/C: 0.603). As of today the reason for this deviation is unknown. Generally, the data from C=C/CH2 ratio, the A(λmax)/A(λ260nm) and the H/C correlates poorly. Although these parameters are all measures of de- gree of aromaticity in the samples, they are not directly comparable to each other. This can be explained by the differences in what these ratios represent: the C=C/CH2 ratio does not take into account the amount of terminal aliphatic groups (CH3-groups) present in the molecule. These can be numerous if there is high degree of branch- ing in the aliphatic chains. Secondly, the ratio A(λmax)/A(λ260nm) does not reflect aliphatic groups or compounds present, but is more a measure of the extend of polycondensation in the aromatic ring sys- tems. The C/H ratio is, in the same way as O/C, a measure of the total content of the elements H and C present in any type of compound, but gives a gen- eral measure on degree of aromaticity. For all sam- ples the H/C ratio is significantly lower than 2, indi- cating presence of unsaturation in the acid extracts and SPE-fractions. The SPEB- and SPEC-fractions have lower H/C values than the whole acid extracts and SPEA. The amounts of sulfur and nitrogen are very small. However, it is observed that all SPEC- fractions have considerable larger nitrogen content. The average formulas given are rough estimates of the elemental distribution of all molecules present in each fraction, based on elemental composition and molecular weights. On average the number of oxygen atoms in each molecule is larger than 2, which indicates that the fractions do not solely con- sist of monocarboxylic acids. From these formulas it seems like the B1c SPE-fractions hold fewer oxy- gen atoms per molecule than the same fractions of B4a, B4c and B2b. The parameter giving the double bond equivalents describes the hydrogen deficiency, i.e. the number of double bonds (C=C, C=O or N=O) and/or number of rings present. As can be seen from Table 3 for all samples the value of DBE increases from low to high from whole acid extracts to the most polar fraction SPEC (note that no data is available for the SPED- fraction). This indicates increased complexity of the functionalities of the molecules present. If a ben- zene ring structure is present it represents four DBE, a diaromatic like naphthalene will contribute with seven DBE, and a triaromatic system like phenan- threne has a DBE of ten. From the UV/VIS spec- tra implying a maximum of two aromatic rings, and FTIR analysis showing distinct absorption from car- bonyl functionalites, obviously C=O groups are also contributing significantly to the total DBE. Like for GPC, precautions must be taken interpret- ing and calculating elemental analysis data. Firstly, the uncertainty in the weight percentage measure- ments is estimated to ±0.2%, meaning that most of the values of %N and %S given in Table 3 are es- sentially close to zero. In addition, the elemental analyzer is not calibrated for sulfur measurements, giving a significant uncertainty in the %S measure- ments. In general, deviations as high as 5 weight% are observed for sulfur-containing standard com- pounds. In addition, oxygen content is measured by difference, accumulating the errors in the other elemental data and adding to the overall uncertainty. Still, it is of value to use these data for a very general description of the samples and for internal compari- son. Ta bl e 3: R at io s be tw ee n so m e ce nt ra lb an ds in sp ec tr a m ea su re d fr om FT IR (% R efl ec ta nc e) an d U V /V IS (A bs or ba nc e) an d da ta fr om el em en ta la na ly si s. FT IR U V /V IS E le m en ta la na ly si s Fr ac tio n C =O / C =C / A (λ m ax )/ w t% w t% w t% w t% w t% A to m ic A to m ic A to m ic A to m ic A ve ra ge fo rm ul aa D B E b C H 2a sy m . C H 2 as ym . A (λ 26 0n m ) C H O N S H /C O /C N /C S/ C C xH yO zS m N n B 4a ac id 1. 02 6 nd 2. 30 0 74 .2 10 .6 14 .5 0. 3 0. 4 1. 69 7 0. 14 6 0. 00 4 0. 00 2 C 21 .6 3H 36 .7 0O 3. 16 S 0 .0 5N 0. 08 4 B 4a SP E A 1. 00 0 1. 18 0 4. 07 3 77 .9 10 .4 11 .1 0. 3 0. 4 1. 58 3 0. 10 7 0. 00 3 0. 00 2 C 32 .4 3H 51 .3 4O 3. 46 S 0 .0 6N 0. 10 8 B 4a SP E B 1. 01 8 1. 05 0 2. 65 5 72 .3 7. 6 18 .7 0. 6 0. 8 1. 24 8 0. 19 4 0. 00 7 0. 00 4 C 24 .0 8H 30 .0 5O 4. 68 S 0 .1 0N 0. 17 10 B 4a SP E C 1. 08 2 1. 08 1 1. 46 7 73 .9 8. 6 13 .7 2. 6 1. 3 1. 38 0 0. 14 0 0. 03 0 0. 00 6 C 33 .8 3H 46 .6 8O 4. 72 S 0 .2 2N 1. 01 12 B 4a SP E D 0. 99 7 1. 02 1 2. 67 6 nm nm nm nm nm nm nm nm nm nm nm B 4c ac id 1. 02 7 nd 2. 45 8 64 .9 9. 2 25 .3 0. 2 0. 4 1. 68 7 0. 29 3 0. 00 2 0. 00 2 C 18 .9 1H 31 .9 1O 5. 54 S 0 .0 5N 0. 04 4 B 4c SP E A 0. 95 1 1. 50 7 2. 07 1 68 .6 9. 3 21 .4 0. 3 0. 5 1. 61 9 0. 23 4 0. 00 4 0. 00 2 C 31 .3 9H 50 .8 2O 7. 35 S 0 .0 8N 0. 11 7 B 4c SP E B 1. 02 9 1. 13 1 1. 96 0 75 .4 8. 4 14 .8 0. 5 0. 9 1. 32 8 0. 14 7 0. 00 6 0. 00 4 C 31 .3 8H 41 .6 9O 4. 63 S 0 .1 3N 0. 19 12 B 4c SP E C 1. 10 7 1. 10 6 1. 32 8 74 .1 8. 2 13 .5 2. 9 1. 4 1. 31 9 0. 13 7 0. 03 3 0. 00 7 C 49 .3 3H 65 .0 7O 6. 76 S 0 .3 4N 1. 63 19 B 4c SP E D 1. 01 7 1. 17 1 1. 50 5 nm nm nm nm nm nm nm nm nm nm nm B 2b ac id 1. 01 9 nd 2. 59 2 72 .3 10 .2 16 .9 0. 2 0. 5 1. 67 3 0. 17 5 0. 00 2 0. 00 3 C 15 .0 4H 25 .1 7O 2. 64 S 0 .0 4N 0. 03 3 B 2b SP E A 0. 88 3 1. 47 5 2. 72 8 65 .6 9. 7 24 .0 0. 2 0. 5 1. 76 0 0. 27 5 0. 00 2 0. 00 3 C 27 .3 2H 48 .1 0O 7. 52 S 0 .0 7N 0. 06 4 B 2b SP E B 1. 01 2 1. 07 8 2. 15 3 73 .8 7. 7 16 .9 0. 6 0. 9 1. 24 1 0. 17 2 0. 00 8 0. 00 5 C 21 .5 1H 26 .7 0O 3. 71 S 0 .1 0N 0. 16 9 B 2b SP E C 1. 14 4 1. 14 5 1. 64 3 69 .6 7. 4 17 .8 3. 8 1. 4 1. 25 8 0. 19 2 0. 04 7 0. 00 7 C 28 .9 8H 36 .4 6O 5. 57 S 0 .2 2N 1. 37 12 B 2b SP E D 1. 03 2 1. 09 2 1. 72 5 nm nm nm nm nm nm nm nm nm nm nm B 1c ac id 1. 05 6 nd 1. 83 6 51 .2 6. 9 41 .2 0. 1 0. 6 1. 60 4 0. 60 3 0. 00 2 0. 00 4 C 10 .6 7H 17 .1 0O 6. 43 S 0 .0 4N 0. 03 3 B 1c SP E A 1. 03 9 1. 13 2 2. 56 8 72 .2 9. 8 17 .3 0. 2 0. 7 1. 61 0 0. 18 0 0. 00 2 0. 00 4 C 15 .0 2H 24 .1 8O 2. 70 S 0 .0 5N 0. 03 4 B 1c SP E B 1. 02 0 1. 07 4 1. 70 1 77 .0 7. 6 13 .5 0. 7 1. 1 1. 18 1 0. 13 2 0. 00 8 0. 00 5 C 16 .0 3H 18 .9 3O 2. 11 S 0 .0 9N 0. 13 8 B 1c SP E C 1. 09 9 1. 12 7 1. 47 6 76 .2 8. 5 11 .6 1. 8 1. 9 1. 33 2 0. 11 4 0. 02 1 0. 00 9 C 31 .7 2H 42 .2 3O 3. 62 S 0 .2 9N 0. 66 12 B 1c SP E D 1. 08 0 1. 06 9 1. 64 7 nm nm nm nm nm nm nm nm nm nm nm S3 b ac id c 1. 05 5 1. 08 8 2. 08 3 65 .7 9. 5 24 .1 0. 3 0. 4 1. 73 2 0. 27 5 0. 00 4 0. 00 2 C 13 .6 7H 23 .6 7O 3. 76 S 0 .0 3N 0. 06 3 a T he av er ag e fo rm ul as ar e ob ta in ed fr om th e el em en ta lc on te nt s gi ve n in Ta bl e 3 an d fr om th e M W m ed va lu es in Ta bl e 2. b D B E (d ou bl e- bo nd eq ui va le nt s) ar e ca lc ul at ed as de sc ri be d by W ill ia m s an d Fl em in g [2 2] . c O nl y da ta on th e w ho le ac id ex tr ac ti s av ai la bl e. nd :n ot de te ct ab le (d ue to la ck of C =C ba nd in th e FT IR sp ec tr a) . nm :n ot m ea su re d (d ue to th e m in ut e am ou nt s of m at er ia le xt ra ct ed in th e SP E D fr ac tio ns ,e le m en ta la na ly si s of th es e fr ac tio ns w er e no ta ch ie va bl e) . N ot e th at FT IR pe ak ra tio s ar e m ea su re d in % R efl ec ta nc e. T hi s m ea ns th at a lo w va lu e of th e ra tio m ea ns hi gh er in te ns ity of th e nu m er at or s C =O or C =C ba nd s in th e FT IR sp ec tr a. SUMMARY AND CONCLUSIONS The distribution of acids and chemical compound class compositions differ between the three crude oils with low hydrate plugging potential (B4a, B4c, B2b) and the high plugging potential oil B1c. The main observations are summarized as follows: - FTIR supports the findings from HPLC for the SPED-fraction: the FTIR-spectra of the three non- plugging oils have signals from ester carbonyl func- tionalities that are not found in the plugging B1c. - The plugging oil in this data set differs from the non-plugging oils by having a relatively higher amount of polyfunctional compounds in all the acid fractions, as determined by HPLC. - The differences are especially prominent in the SPED-fraction: HPLC results shows that the non- plugging oils contain a considerably larger propor- tion of weakly polar compounds, while the plugging oil B1c holds a large amount of polyfunctional com- pounds in this sub-fraction. - The chemical composition of the acid extracts seems to be of more importance than the amount of acids present in the oils when considering hydrate plugging potential of biodegraded oils. Chemical characterization of the composition of petroleum acid extracts and their SPE-fractions has been performed, comparing fractions from oils with low and high hydrate plugging potential. We show that separation of the acid extracts into smaller sub- fractions that are studied separately is useful for ana- lyzing such complex mixtures and improves the res- olution. In general, the chemical composition of the acid extracts and SPE-fractions from the oils studied in this work can be summarized as follows: - The SPEA-fraction represents the major part of the whole acid extracts. HPLC results show that this fraction contains almost exclusively carboxylic acids. - The molecular weights range from 150 to 2000 g/mol, but the major part of the components present are found in the range between 250-800 g/mol. - The acid extracts and SPE-fractions contain com- pounds ranging from C10 to C49, with 2-7 oxygen per molecule and a maximum of two aromatic rings present, as a rough approximation. - The SPEC-fractions seem to have the most complex acid composition, including phenols that are most likely included in polyfunctional molecules. Unlike the other samples studied, the SPEC-fractions also contains a significant amount nitrogen. - The complexity increases with the polarity of the fractions, reflected in the DBE-values. Carbonyl functionalities contribute significantly to the DBE- values. Further work is in progress, addressing the challenge of obtaining accurate information on the chemi- cal structures of the natural inhibiting compounds present in the non-plugging oils. ESI-MS is a highly suitable method for determination of polar com- pounds in petroleum, and experiments of structure determination on acid extracts and SPE-fractions us- ing this technique have commenced. ACKNOWLEDGMENTS Inger Johanne Fjellanger is greatfully acknowledged for performing elemental analysis on the samples. The authors want to thank the Research Council of Norway and StatoilHydro for funding the work, Sta- toilHydro also for providing the crude oils and for the permission to publish results. REFERENCES [1] Sloan ED. Clathrate Hydrates of Natural Gases. Second edition, Revised and Expanded. New York: Marcel Dekker Inc., 1998. [2] Leporcher EM, Peytavy JL, Mollier Y, Sjöblom J, Labes-Carrier C. Multiphase transportation: hydrate plugging prevention through crude oil natural surfactants. Society of Petroleum Engineers 1998; paper 49172. [3] Hemmingsen P, Li X, Peytavy JL, Sjöblom J. Hydrate Plugging Potential of Original and Modified Crude Oils. Journal of Dispersion Science and Technology 2007;28(3):371-382. [4] Arjmandi M, Leinweber D, Allan K. Develop- ment of A New Class of Green Kinetic Hydrate Inhibitors. Presented at the 19th International Oil Field Chemistry Symposium, 9-12 March 2008, Geilo, Norway. [5] Fadnes FH. Natural hydrate inhibiting com- ponents in crude oils. Fluid Phase Equilibria 1996;117(1-2):186-192. [6] Bergflødt L. Influence of Crude Oil Based Sur- face Active Components and Synthetic Surfac- tants on Gas Hydrate Behaviour. PhD thesis, University of Bergen, Norway, 2001. [7] Høiland S, Askvik KM, Fotland P, Alagic E, Barth T, Fadnes F.Wettability of freon hydrates in crude oil/brine emulsions. Journal of Col- loid and Interface Science 2005;287(1):217- 225. [8] Hemmingsen P, Kim S, Pettersen HE, Rodgers RP, Sjöblom J, Marshall AG. Structural Char- acterization and Interfacial Behaviour of Acidic Compounds Extracted from a North Sea Oil. Energy & Fuels 2006;20(5):1980-1987. [9] Seifert WK, Howells WG. Interfacially Ac- tive Acids in California Crude Oil. Isolation of Carboxylic Acids and Phenols. Analytical Chemistry 1969;41(4):554-562. [10] Langevin D, Poteau S, Henaut I, Argillier JF. Crude Oil Emulsion Properties and their Ap- plication to Heavy Oil Transportation. Oil and Gas Science and Technology - Revue De L Institut Franscais Du Petrole 2004;59(5):511- 521. [11] Barrow MP, Headley JV, Peru KM, Derrick PJ. Fourier transform ion cyclotron resonance mass spectrometry of principal components in oilsands naphthenic acids. Journal of Chro- matography A 2004;1058(1-2):51-59. [12] Høiland S, Borgund AE, Barth T, Fotland P, Askvik KM. Wettability of freon hydrates in crude oil/brine emulsions: the effect of chemi- cal additives. In: Proceedings of the 5th Inter- national Conference on Gas Hydrates. Trond- heim: 2005. [13] Erstad K, Høiland S, Fotland P, Barth T. The in- fluence of petroleum acids on gas hydrate mor- phology. (in preparation). [14] Borgund AE, Erstad K, Barth T. Fractiona- tion of crude oil acids by HPLC and char- acterization of their properties and effects on gas hydrate surfaces. Energy & Fuels 2007;21(5):2816-2826. [15] Altgelt KH, Boduszynski MM. Composition and Analysis of Heavy Petroleum Fractions. New York: Marcel Dekker Inc., 1994. [16] Barth T, Høiland S, Fotland P, Askvik KM, Pedersen BS, Borgund AE. Acidic compounds in biodegraded petroleum. Organic Geochem- istry 2004;35(11-12):1513-1525. [17] Mediaas H, Grande KV, Hustad BM, Rasch A, Rueslåtten HG, Vindstad JE. The Acid-IER Method - a Method for Selective Isolation of Carboxylic Acids from Crude Oils and Other Organic Solvents. Society of Petroleum Engi- neers 2003; paper 80404. [18] Borgund AE, Erstad K, Barth T. Normal phase high performance liquid chromatography for fractionation of organic acid mixtures ex- tracted from crude oils. Journal of Chromatog- raphy A 2007;1149(2):189-196. [19] Watson JS, Jones DM, Swannell RPJ, van Duin ACT. Formation of carboxylic acids during aerobic biodegradation of crude oil and evi- dence of microbial oxidation of hopanes. Or- ganic Geochemistry 2002;33(10):1153-1169. [20] Speight JG. The chemistry and technology of petroleum. Third edition, Revised and Ex- panded. New York: Marcel Dekker Inc., 1999. [21] Smith DF, Schaub TM, Rahimi P, Teclemariam A, Rodgers RP, Marshall AG. Self-Association of Organic Acids in Petroleum and Canadian Bitumen Characterized by Low- and High- Resolution Mass Spectrometry. Energy & Fu- els 2007;21(3):1309-1316. [22] Williams DH, Fleming I. Spectroscopic Meth- ods in Organic Chemistry. Fifth edition. Lon- don: The McGraw-Hill Companies, 1995. [23] Coates J. Interpretation of Infrared Spectra, A Practical Approach. In: Meyers RA, editor. Encyclopedia of Analytical Chemistry. Chich- ester: JohnWiley & Sons Ltd., 2000. p. 10815- 10837.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items