Oxidation Behavior of Light Crude Oil and Its SARA Fractions

Nov 30, 2017 - Verkoczy and Freitag(9) investigated the oxidation behavior of three Canadian heavy oils and their SARA fractions by TG and autoclave t...
2 downloads 10 Views 2MB Size
Subscriber access provided by Gothenburg University Library

Article

Oxidation Behavior of Light Crude Oil and Its SARA Fractions Characterized by TG and DSC Techniques: Differences and Connections Cheng-dong Yuan, Mikhail A. Varfolomeev, Dmitrii A. Emelianov, Aleksey A. Eskin, Ruslan N. Nagrimanov, Mustafa Versan Kok, Igor S. Afanasiev, Gennadii D. Fedorchenko, and Elena V. Kopylova Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b02377 • Publication Date (Web): 30 Nov 2017 Downloaded from http://pubs.acs.org on November 30, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Oxidation Behavior of Light Crude Oil and Its SARA Fractions Characterized by TG and DSC Techniques: Differences and Connections Chengdong Yuan†, Mikhail A. Varfolomeev †*, Dmitrii A. Emelianov †, Aleksey A. Eskin †, Ruslan N. Nagrimanov†, Mustafa Versan Kok‡, Igor S. Afanasiev§, Gennadii D. Fedorchenko§, Elena V. Kopylova§ †

Department of Physical Chemistry, Kazan Federal University, Kazan, Russia Middle East Technical University, Ankara, Turkey § JSC Zabubezhneft, Moscow, Russia ‡

Abstract: This research is intended to reveal the difference and connection of oxidation behavior between crude oil and its SARA fractions. Thermogravimetry (TG) and differential scanning calorimetry (DSC) techniques were used to characterize oxidation behavior. The results showed that the oxidation behavior of individual SARA components exhibited obvious difference. Saturates showed a weak high-temperature oxidation (HTO) region. Asphaltenes generated more heat in HTO than in low-temperature oxidation (LTO) region. Aromatics showed intense exothermic activity both in LTO and HTO regions. Heat release and mass loss showed a good correspondence in HTO region for all SARA fractions, which means heat release and mass loss were caused by the same reaction mechanism that is believed to be the coke combustion as it is the only significant reaction in HTO region. However, the good correspondence didn’t exist in LTO interval where the reactions are more complicated and multiple-step mechanism should be considered. In addition, it is not quite reasonable to determine the reactivity of SARA fractions only by TG data as little mass loss does not mean reactants are inactive. Kinetic parameters of LTO and HTO reactions were determined by Friedman and Ozawa–Flynn–Wall isoconversional methods. In general, for the crude oil and each fraction, the activation energies of HTO were higher than that of LTO. The additivity of DSC data could be applied quite well in LTO region. However, the predicted curve seriously deviated from the actual situation after 350 0C, which implies the exothermic reaction process of individual components was influenced by the presence of other components. Nevertheless, the total heat release of the measured and predicted values was similar, which makes it possible to predict the heat effect of crude oil from individual SARA components.

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

1. INTRODUCTION Air injection techniques can offer unique benefits and technical opportunities for enhanced oil recovery (EOR) in many candidate reservoirs for both conventional and unconventional oil sources, such as high pressure air injection (HPAI) for water-flooded light oil reservoirs and low permeability reservoirs where water injection is difficult (conventional oil sources), in-situ combustion (ISC) for heavy oil reservoirs, bitumen sands and shale oil (unconventional oil sources).1-3 After air is injected into reservoirs, various oxidation reactions take place between crude oils and oxygen contained in the injected air, which generates heat to reduce oil viscosity (especially for ISC in heavy oil reservoirs) and flue gas to improve swept volume and increase reservoir pressure.4 It is acknowledged that oxidation reactions between crude oil and injected air dictate the overall success of the air injection processes.5 Therefore, to improve the efficiency of air injection process it is necessary to acquire a clear understanding of oxidation behavior of crude oils. The oxidation reactions strongly depend on the reactivity of crude oils that is mostly determined by their composition.6 Generally, crude oils can be divided into four fractions as saturates, aromatics, resins and asphaltenes (SARA fractions).7 In recent years, many different crude oils and their SARA fractions have been studied to characterize oxidation behavior and mechanism. Freitag and Verkoczy 5 studied the LTO behavior of two Canadian heavy oils and their SARA fractions using PDSC and a tubular flow reactor. They found that the LTO rate of saturates is very different from that of other fractions, and the addition of resins and aromatics can repress the early oxidation of saturates. Kok and Gul 8 analyzed the oxidation characteristics and kinetics of Turkish crude oils and their SARA fractions using DSC. Two main reaction stages are observed from DSC curves as LTO and HTO, and saturates shows the minimum heat release. Varfolomeev et al.7 investigated the oxidation behavior of Tatarstan heavy oil and its SARA fractions by DSC and TG. Also, two main reaction stages are detected from DSC curves. However, only one main stage is observed in DTG curves. Verkoczy and Freitag

9

investigated the oxidation behavior of three Canadian heavy oils and their SARA fractions by TG and autoclave tests. Their study indicates that saturates display the slowest reaction rate and asphaltenes are the most reactive for low temperature oxidation under 225 0C in autoclave experiments. The similar results that asphaltenes have higher reactivity than paraffins at 150-250 0C were reported by Moschopedis and Speight.10 Niu et al.11 also indicated that heavy oil and heavier oil compounds are

ACS Paragon Plus Environment

Page 3 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

easier subjected to be oxidized than light oil and light compounds under relatively low temperature. However, there are some different views about the reactivity of different oil fractions. Li et al.12 indicated that paraffin samples with low molecular weight exhibit lower onset temperatures for oxidation reactions than heavier paraffin samples. Kok and Karacan13 reported that for LTO reactions, asphaltenes have very little mass loss and heat release, while saturates show a huge mass loss and the highest amount of heat release. For HTO reaction, saturates display a weak combustion reaction, while asphaltenes release a great amount of heat. Al-Saffar et al.14 concluded that saturates have a high reactivity at low temperature, which can yield hydrocarbon products that lead to another oxidation region. Saturates make a major contribution for both oxygen consumption and fuel deposition for the light oil. The concise review of above studies shows that the oxidation behavior of crude oils and SARA fractions has received considerable attention in the literature. However, the oxidation behavior of crude oils is still not fully understood. Simultaneously, there are still some disagreements among different studies that might be attributed to the difference in oil fractions, methods and evaluation index used in these studies, which requires more investigations to better understand the oxidation behavior of crude oils and SARA fractions. In addition, in these studies, the difference of the oxidation behavior between crude oils and different SARA fractions were mainly described and explained, but the connection between them has been little documented. Therefore, in this study, the oxidation behavior and kinetics of crude oils and SARA fractions will be further analyzed by DSC and TG to help to get a better understanding of oxidation mechanism. Also, another principle objective of this study is to reveal the connection of the oxidation behavior between crude oil and its SARA fractions in terms of heat effect.

2. EXPERIMENTAL SECTION 2.1. Materials The crude oil used in this study was obtained from the Vishanskoe oilfield (Russia). The properties and SARA fractions of the crude oil are shown in Table 1. ASTM procedure was used for analyzing and obtaining the SARA fractions, which is described in details by Varfolomeev et al.7 Silica (ACROS 24037, 0.060-0.200mm, 60Å) was provided by TatchemProduct (Kazan, Russia).

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2.2. Characterization of SARA fractions As is mentioned in Introduction section, the difference in used oils and SARA fractions might result in different conclusions. Therefore, to increase the broader scientific contribution of this work and make it easier to be compared with other possible studies using different crude oils in future, these obtained SARA fractions were further characterized to understand their main components. Gas chromatography (GC) and high-performance liquid chromatography (HPLC) were used for analyzing the main fractions of saturates and aromatics, respectively. Krystall 2000M gas chromatograph (Chromatec) with flame ionization detector was used for measuring the saturate fraction. The temperature of column varied from 100 to 150 0C at a heating rate of 10 0C/min and from 150 to 300 0C at a heating rate of 3 0C/min. Hydrogen was used as a carrier gas. The temperatures of injector and detector were 310 and 250 0C, respectively. High performance liquid chromatograph UltiMate 3000 (Thermo Scientific) was employed for analyzing the aromatic fraction. Experiments were carried out using the following parameters: the temperature of columns was 30 0C (analytic column Restek Pinnacle II Amino 5µm 250 mm * 4.6 mm), the temperature of refractometer cell was 30 0C, flow rate was 1 ml/min, and n-Heptane was used as the eluent. The following analytical standards were used to determine the yield of aromatic hydrocarbon groups and their concentrations: o-xylene (mono-aromatics), fluorene (di-aromatics) and phenanthrene (tri-aromatics). Matrix-assisted laser desorption-ionization time of flight mass spectrometry (MALDI-TOF MS) was employed for analyzing the relative molecular mass of the resins and asphaltenes. Experiments were carried out using Bruker MALDI TOF/TOF Ultraflextreme instrument. Solutions of resins and asphaltenes in toluene were taken for the measurements. The spectra were obtained practically without the participation of the matrix, since it is an aqueous solution that does not mix with toluene. The main fractions of saturates are C11-C16, and the main fractions of aromatics include diaromatics (54 %) and monoaromatics (23 %). The relative molecular mass of resins and asphaltenes are 450 Da and 1300 Da, respectively. 2.3. Thermogravimetry and Differential Scanning Calorimetry Analysis In this study, oxidation behavior was investigated by thermogravimetry (TG) and differential scanning calorimetry (DSC). These thermal analysis techniques have obtained wide acceptance and been widely used in the characterization of the combustion or pyrolysis behavior of hydrocarbons and their derivatives.15-21 In this study, NETZSCH STA 449 F3 Jupiter (Netzsch, Germany) was used to

ACS Paragon Plus Environment

Page 4 of 22

Page 5 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

characterize the oxidation behavior of the crude oil and its SARA fractions. Before experiments, temperature and sensitivity calibrations were conducted in an open alumina crucible filled by the standard samples of seven different metals indium (In), tin (Sn), bismuth (Bi), lead (Pb), zinc (Zn), aluminum (Al), gold (Au) at the different heating rates from 2 to 20 0C/min in dynamic air atmosphere. DSC-TG Experiments were performed under air atmosphere with an air flow rate of 75 ml/min at different heating rates of 4, 6, and 10 0C/min using alumina crucible. The temperature range was 30 – 600 0C. Silica was used as solid matrix for TG and DSC experiments. Before experiments, the crude oil and its SARA fractions were mixed with silica. The baseline correction was conducted for the obtained DSC data for all experiments by the method of “spline”.22

3. RESULTS AND DISCUSSION 3.1. Oxidation Behavior of Crude Oil and its SARA Fractions Characterized by DSC. The oxidation behavior of the crude oil and its SARA fractions was analyzed by DSC experiments. Figures 1-5 show the DSC curves of the saturates, asphaltenes, aromatics, resins and crude oil at different heating rates of 4, 6 and 10 0C/min, respectively. The reaction intervals and peak temperatures obtained from the DSC curves at 4 0C/min are given in Table 2. For the saturates (Figure1), only one obvious exothermic reaction region considered as LTO was observed from the DSC curves. After the LTO, there was an inflection point that indicates the oxidation reaction was transforming into a very weak HTO reaction. It is generally admitted that HTO reaction are the process where deposited solid fuel is completely decomposed.14 With this in mind, it can be concluded that a weak HTO region means that no enough fuel was deposited and the saturates almost completely reacted with oxygen during the LTO reaction, which implies that the saturates used in this study make a small contribution to the HTO reaction of the crude oil. This is consistent with the result reported by Kok et al.13, 23 The asphaltenes behaved quite differently from the saturates. As is shown in Figure 2, there are two exothermic reaction regions in the DSC curves: LTO and HTO regions. Unlike the saturates, the HTO reaction of the asphaltenes was intense and more pronounced than the LTO reaction. The heat release at the HTO region was much higher than that at the LTO region. This means more fuel was formed in the LTO reaction of the asphaltenes and combusted in the HTO region, which contributes to the high heat generation.

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For the aromatics, also two significant exothermic reaction regions (i.e., LTO and HTO) were detected. There, heat released at the LTO region was higher than that at the HTO region. Li et al.6 also observed two obvious exothermic reaction ranges from the DSC curves of pure aromatic components, but in their study the HTO reaction released a much higher amount of heat than the LTO reaction. We assumed that the difference in the molecular structure of used aromatics is the main reason that causes the different results. In this research, the main fractions of the used aromatics include diaromatics (54 %) and monoaromatics (23 %) as mentioned before. However, in their research, the used aromatics are 9,10-dimethylanthracene (C16H14) with three benzene rings and 7,12-dimethylbenzanthracene (C20H16) with four benzene rings that are heavier than the aromatics employed in this study in terms of the number of benzene rings. These heavier aromatics behave similarly to asphaltenes (HTO reaction is stronger than LTO reaction).6 For the resins, the oxidation behavior was more like that of the aromatics. Similarly, two exothermic reaction regions were observed, and the LTO region had a higher heat generation than the HTO region. But the HTO reaction was not as strong as that of the aromatics. However, it can be seen from Table 2 that the resins had the lowest peak temperature in both the LTO and HTO regions, 323.0 and 396.0 0C, respectively. Figure 6 shows a comparison of the heat flow for the crude oil and different SARA fractions at the heating rate of 4 0C/min. These curves were drawn after deducting the mass of silica. Therefore, they reflect the heat flow per unit mass of pure oil or SARA fractions. Heat releases during oxidation for each SARA fraction were calculated according to mass loss. The heat releases are in turn: asphaltenes (34458 J/g) > resins (26525 J/g) > aromatics (20834 J/g) > crude oil (13999 J/g) > saturates (7137 J/g). As shown in Figure 6, the oxidation behavior was very different between individual SARA fractions. The saturates showed a very weak HTO region, released more heat in the LTO region than in the HTO region, and gave the smallest total heat release of 7137 J/g. While the asphaltenes exhibited a weak LTO region, generated more heat in the HTO region than in the LTO region, and had the highest total heat release of 34458 J/g. Among all the SARA fractions, the asphaltenes displayed different oxidation behavior from other fractions where heat release in HTO region was higher than in LTO region, and showed the most intense HTO reaction. As we mentioned before, HTO reaction is a combustion process of deposited fuel. This means that asphaltenes are the main source that produces solid fuel to contribute in the HTO reaction of crude oil. Murugan et al.24 also reported that asphaltenes make a ACS Paragon Plus Environment

Page 6 of 22

Page 7 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

major contribution on coke formation. The aromatics showed intense exothermic activity both in the LTO and HTO region. Therefore, it can be concluded that some solid fuel was formed under the LTO process of the aromatics for the HTO reaction. As can be seen in Figure 6, the DSC curve of the crude oil is close to that of the saturates. This is the expected result as the saturates accounts for 69.3% (by weight) of the crude oil. 3.2. Oxidation Behavior of Crude Oil and its SARA Fractions Characterized by TG-DTG. Under some circumstances, for some complicated reactions, the reaction behavior characterized by TG and DSC is not always in complete agreement. To make a better understanding of the oxidation behavior, TG-DTG plots were drawn together with DSC curves on the same figures for the convenience of comparison. Figures 7-11 show the TG-DTG-DSC curves of the saturates, aromatics, resins, asphaltenes, and crude oil at the heating rate of 4 0C/min. For the saturates (Figure 7), mass loss started from the beginning of heating, and rapid mass loss took place at 64.1 0C, while the temperature point where it started to release heat was about 89.8 0C. It was assumed that some mass loss before exothermic reaction was caused by vaporization. Simultaneously, as can be seen from DTG curve in Figure 7, there is a distinct mass loss region before significant heat release occurred, which can be attributed to the vaporization of the light fractions in the saturates. For the crude oil, the similar scenario was observed that there also exists an obvious mass loss peak caused by vaporization. This means that the vaporization process is a non-negligible stage for the oxidation of saturates and light oils. The similar phenomenon of high mass loss for light oil in low-temperature range was also observed by Li et al. 25 They believe that for light oil, the vaporization of lighter components from the oil is dominant in low-temperature range. For the aromatics (Figure 8), rapid mass loss and heat release occurred almost at the same temperature of approximately 125.0 0C. This means that the evaporation for the used aromatics was almost negligible. However, for the resins and asphaltenes (Figures 9 and 10), an interesting phenomenon was observed that the temperature where rapid mass loss happened was much higher than the temperature where heat release occurred. The heat release started at 139.5 and 129.3 0C for resins and asphaltenes, respectively, but the rapid mass loss took place after 249.1 and 246.7 0C. This indicated that little mass loss doesn’t mean that reactants are inactive or resistant. It is not very reasonable to determine the reactivity of the resins and asphaltenes only by mass loss. In fact, this phenomenon should be

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

attributed to the high reactivity of resins and asphaltenes in low temperature ranges considering that a significant heat release was observed with a small mass loss. Many researchers have indicated that heavy oils and asphaltenes are more reactive than light oils and lighter oil fractions at relative low temperature ranges.2, 9-11 This kind of high reactivity refers to the strong ability to uptake oxygen by oxygen addition reactions. In this scenario, oxygen atoms are chemically bound into the molecular structure of resins and asphaltenes, which releases some heat and generated various oxygenated compounds. Consequently, the mass loss is slow during the oxygen addition reactions. Simultaneously, it is worth noticing that for all the SARA fractions and crude oil there was a good corresponding relation between heat release and mass loss in HTO region, that is, the temperature range of reaction intervals and peak temperature for HTO reactions characterized by DSC and TG-DTG were almost the same. This direct correspondence means that heat generation and mass loss in HTO region were caused by the same reaction mechanism that is believed to be the coke combustion as it is widely accepted to be the only important reaction in HTO reaction. However, in LTO region, this corresponding relation was not so good for all SARA fractions and crude oil, and the thermal behavior and mass loss behavior had a significant difference. For the saturates and aromatics, the fastest mass loss took place ahead of the highest heat release. The peak temperature of the LTO region in DSC curves (332.0 0C for saturates and 341.2 0C for aromatics) was higher than that in the DTG curves (307.0 0C for saturates and 320.1 0C for aromatics). In addition, for the saturates, it can be seen from Figure 7 that there was an extra obvious inflection between 200 and 300 0C in the DSC curve and three regions in the DTG curve of the LTO region. This indicated that LTO reaction should be considered and modeled by multiple-step mechanism. Vaporization step is very important for these fractions. For the resins and asphaltenes, only after approx. 250 0C where the rapid mass loss took place there was a direct correspondence between heat release and mass loss. Before 250 0C, the heat release was already very significant, but the mass loss was very slow. Therefore, the LTO interval of the resins and asphaltenes should be further divided into at least two subintervals. For the crude oil, the scenario was similar with the saturates as it accounts for 69.3% of all components, which is the largest proportion. 3.3. Kinetic Analysis. The oxidation of crude oil is a very complex, and numerous reactions proceed simultaneously. As

ACS Paragon Plus Environment

Page 8 of 22

Page 9 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

we mentioned before, the entire reaction process can be roughly divided into two obvious reaction intervals from DSC curves, which refers to LTO and HTO. In this work, the kinetic parameters of LTO and HTO were determined using software thermokinetics by NETZSCH where the “Model Free” package enables the kinetic parameters activation energy (Ea) and logarithm of pre-exponential factor (log A) to be calculated by Friedman and Ozawa–Flynn–Wall isoconversional models. The calculated kinetic parameters could be considered as apparent data representing complex reactions. The main advantage of these isoconversional methods is to evaluate the activation energy without determining reaction model.26-27 Consequently, it can provide more reliable insights into the intrinsic kinetics of the complicated oxidation process. For differential isoconversional method suggested by Friedman, the following equation is used:27-28 ln ×  ⁄ = ln  + ln −  ⁄

(1)

The logarithm of the conversion rate ln/ is recorded vs. 1/T at different heating rates (β . For OFW integral isoconversional method, for a selected conversion degree , the following equation is used:28 ln = −1.0516 ×  ⁄ + const.

(2)

ln is plotted vs. 1/T. In both two methods, for a constant , the plotted lines should be straight, and the activation energies can be determined from the slope of the line. The conversion dependence of activation energy is presented in the supporting information. Obtained kinetic parameters are presented in Table 3. As we can see that the activation energies calculated by Friedman and Ozawa–Flynn–Wall models were similar, but there were still some differences, which can be explained by a systematic error resulted from improper integration.28 In Friedman method, instantaneous rate values are used, which is very sensitive to experimental noise.28 For OFW method, the systematic error is introduced by the assumption that the value of activation energy is independent of .22 In general, for all SARA fractions and crude oil, the activation energies of HTO reaction were higher than that of LTO reaction. The activation energies of LTO and HTO varied between 102 – 145 kJ/mol and 118 – 155 kJ/mol, respectively. As shown in Table 3, among all SARA components, resins showed the lowest activation energies for both the LTO and HTO. This might be attributed to the chemical instability of resins molecules. When heated, resins are easily oxidized into asphaltenes.29 It was also reported that the reactivity of resins are higher than saturates

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

and aromatics in the process of LTO, and the further oxidation of resins will produce asphaltens-like products.9 With all these in mind, it is easier to understand why resins gave the lowest activation energies. 3.4. Additivity of Crude Oil Oxidation. As mentioned in section 3.1 and 3.2, the general oxidation behavior of the crude oil was closer to that of the saturates. Here, the DSC data were used to determine whether there is some specific connection in oxidation behavior between the crude oil and its SARA fractions, and whether the oxidation behavior of crude oil can be predicted directly from individual SARA components according to additivity rule. The predicted curve was obtained by the additive scheme of the heat flow of individual SARA component according to its proportion in crude oil. Before the additive scheme was conducted, the heat flow data were corrected by mass loss. The predicted curve was compared with the DSC measurement result of the crude oil in Figure 12. As can be seen from Figure 12, a relatively good agreement was shown for the LTO region (before 350 0C) between the predicted and experimental curves. The peak temperatures of the LTO for predicted curve (329.0 0C) and measurement results (329.5 0C) were almost the same, and the temperature ranges of the LTO region were also coincident. This indicated that the additivity of the DSC data could be applied quite well for the LTO region, which partly implies that the exothermic reaction of individual component in the LTO region before 350 0C was nearly independent of the presence of other components and they complied with their own reaction pathway. This allows one to predict the LTO oxidation process of whole oil from individual SARA components. However, the predicted curve gradually deviated from the measurement curves at the later stage of the LTO after 350 0C, and became interlaced in the HTO region. In the measured curve, the HTO reaction was more distinct and occurred earlier than the predicted scenario, and the HTO reaction was finished with higher heat release in a narrower temperature range, which means the rate of fuel formation was faster in actual situation for crude oil. Generally, the later stage of LTO and the transition stage between LTO and HTO are considered as the main region where solid fuel is formed. The disagreement of the predicted and measurement curve at this region at least implies that fuel formation process of one individual SARA component is influenced by the presence of other fractions. They could not comply with their own pathway to generate fuel. Nevertheless, total heat release for the

ACS Paragon Plus Environment

Page 10 of 22

Page 11 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

measured and predicted curves was almost the same. The generated heat of the oxidation for crude oil measured by DSC was 11107 J/g, and the predicted value obtained by additive scheme was 11072 J/g. This makes it possible to predict the heat effect of different crude oils from individual SARA components.

4. CONCLUSION Based on the results of TG and DSC experiments, the following conclusions were drawn about the difference and connection of the oxidation behavior between crude oil and its SARA fractions: (1) Saturates showed a very weak HTO region, and was almost completely combusted in LTO reaction. This fraction made a small contribution to the HTO reaction of the crude oil. While asphaltenes exhibited a weak LTO region, generated more heat in HTO region than in LTO region, and had the highest total heat release and the most intense HTO reaction. Aromatics showed intense exothermic activity both in LTO and HTO region. The heat releases are in turn: asphaltenes (34458 J/g) > resins (26525 J/g) > aromatics (20834 J/g) > saturates (7137 J/g). (2) For saturates and crude oil, rapid mass loss took place before heat release occurred, and there is a distinct mass loss region before significant heat release took place. For aromatics, the rapid mass loss and heat release occurred almost at the same temperature. However, for resins and asphaltenes, the temperature where heat release occurred was much lower than the temperature where rapid mass loss happened due to the strong ability to uptake oxygen by oxygen addition reactions. This indicated that little mass loss doesn’t mean that reactants are inactive or resistant, and it is not quite reasonable to determine the reactivity of resins and asphaltenes only by TG data. (3) For all SARA fractions and crude oil, heat release and mass loss in HTO region showed a good correspondence, which means they were caused by the same reaction mechanism. However, the thermal behavior and mass loss behavior exhibited significant differences in LTO region. Based on the results of DSC and DTG, LTO interval should be considered by multiple-step mechanism and divided into at least two subintervals. (4) Activation energies calculated by Friedman and Ozawa–Flynn–Wall models were similar. The activation energies of LTO and HTO varied between 102–145 kJ/mol and 118–155 kJ/mol, respectively. In general, for the crude oil and each fraction, the activation energies of HTO were higher than that of LTO. Resins showed the lowest activation energies for both LTO and HTO.

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(5) The additivity of the DSC data could be applied quite well for LTO region before 350 0C, which implies the exothermic reaction of individual component in LTO region was nearly independent of the presence of other components and they complied with their own reaction pathway. This allows to predict the LTO process of whole oil from individual SARA component. However, since the later stage of LTO (after 350 0C) where it is considered as the beginning of the deposition of high amount solid fuel, the predicted curve showed serious deviation from the actual situation. It means the reaction process of one individual SARA component is influenced by the presence of other components, at least for the fuel formation process. Nevertheless, total heat release for the measured and predicted curves was almost the same, which makes it possible to predict the heat effect of different crude oils from individual SARA components.

SUPPORT INFORMATION The detailed results of kinetic analysis are provided as support information. This information is available free of charge via the Internet at http://pubs.acs.org/.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected], [email protected] (M. A. Varfolomeev) Notes The authors declare no competing financial interest.

ACKNOWLEDGEMENT This work was supported by the Russian Science Foundation project N17-73-10378.

REFERENCES (1) Kuppe, G.; Mehta, S.; Moore, R.; Ursenbach, M.; Zalewski, E. Heats of combustion of selected crude oils and their SARA fractions. J. Can. Petrol. Technol. 2008, 47, 1−8. (2) Pu, W. F.; Yuan, C. D.; Jin, F. Y.; Wang, L.; Qian, Z.; Li, Y. B.; Chen, Y. F. Low-temperature oxidation and characterization of heavy oil via thermal analysis. Energy Fuels 2015, 29(2), 1151−1159. (3) Gargar, N. K.; Mailybaev, A. A.; Marchesin, D.; Bruining, H. Compositional effects in light/medium oil recovery by air injection: vaporization vs. combustion. J. Porous Media 2014,

ACS Paragon Plus Environment

Page 12 of 22

Page 13 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

17(11), 937−952. (4) Nissen, A.; Zhu, Z.; Kovscek, A.; Castanier, L.; Gerritsen, M. Upscaling kinetics for field-scale in-situ-combustion simulation. SPE Reserv. Eval. Eng. 2015, 18, 158−170. (5) Freitag, N.; Verkoczy, B. Low-temperature oxidation of oils in terms of SARA fractions: why simple reaction models don't work. J. Can. Petrol. Technol. 2005, 44, 54−61. (6) Li, J.; Mehta, S.; Moore, R.; Zalewski, E.; Ursenbach, M.; Van Fraassen, K. Investigation of the oxidation behaviour of pure hydrocarbon components and crude oils utilizing PDSC thermal technique. J. Can. Petrol. Technol. 2006, 45, 48−53. (7) Varfolomeev, M. A.; Galukhin, A.; Nurgaliev, D. K.; Kok, M. V. Thermal decomposition of Tatarstan Ashal’cha heavy crude oil and its SARA fractions. Fuel 2016, 186, 122−127. (8) Kök, M. V.; Gul, K. G. Combustion characteristics and kinetic analysis of Turkish crude oils and their SARA fractions by DSC. J. Therm. Anal. Calorim. 2013, 114, 269−275. (9) Verkoczy, B.; Freitag, N. Oxidation of heavy oils and their sara fractions-its role in modelling in-situ combustion, presented at the seventh petroleum conference of the south saskatchewan section, the Petroleum Society of CIM, held in Regina, october 19-22, 1997. (10) Moschopedis, S. E.; Speight, J. G. Oxidation of Petroleum Fractions. Fuel 1973, 52, 83. (11) Niu, B.; Ren, S.; Liu, Y.; Wang, D.; Tang, L.; Chen, B. Low-temperature oxidation of oil components in an air injection process for improved oil recovery. Energy Fuels 2011, 25, 4299−4304. (12) Li, J.; Mehta, S.; Moore, R.; Ursenbach, M.; Zalewski, E.; Ferguson, H.; Okazawa, N. Oxidation and ignition behaviour of saturated hydrocarbon samples with crude oils using TG/DTG and DTA thermal analysis techniques. J. Can. Petrol. Technol. 2004, 43, 45−51. (13) Kok, M.; Karacan, C. Behavior and effect of SARA fractions of oil during combustion. SPE Reserv. Eval. Eng. 2000, 3, 380−385. (14) Al-Saffar, H.; Hasanin, H.; Price, D.; Hughes, R. Oxidation reactions of a light crude oil and its SARA fractions in consolidated cores, Energy Fuels 2001, 15 182−188. (15) Varfolomeev, M. A.; Nagrimanov, R. N.; Samatov, A. A.; Rakipov, I. T.; Nikanshin, A. D.; Vakhin, A. V.; Nurgaliev, D. K.; Kok, M. V. Chemical evaluation and kinetics of Siberian, north regions of Russia and Republic of Tatarstan crude oils. Energ. Source Part A 2016, 38, 1031−1038. (16) Varfolomeev, M. A.; Nurgaliev, D. K.; Kok, M. V. Calorimetric study approach for crude oil combustion in the presence of clay as catalyst, Petrol. Sci. Technol. 2016, 34, 1624−1630. ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(17) Varfolomeev, M. A.; Nurgaliev, D. K.; Kok, M. V. Thermal, kinetics, and oxidation mechanism studies of light crude oils in limestone and sandstone matrix using TG-DTG-DTA: Effect of heating rate and mesh size. Petrol. Sci. Technol. 2016, 34, 1647−1653. (18) Kök, M. V.; Varfolomeev, M. A.; Nurgaliev, D. K. Thermal characterization of crude oils in the presence of limestone matrix by TGA-DTG-FTIR. J. Petrol. Sci. Eng. 2017. 154, 495−501. (19) Liu, P. G.; Pu, W. F.; Ni, J. H.; Ma, X. P.; Zhang, J.; Liu, M. Thermal investigation on crude oil oxidation kinetics through TG/DTG and DTA tests. Petrol. Sci. Technol. 2016, 34, 685−692. (20) Pu, W. F.; Liu, P. G.; Li, Y. B.; Jin, F. Y.; Liu, Z. Z. Thermal characteristics and combustion kinetics analysis of heavy crude oil catalyzed by metallic additives. Ind. Eng. Chem. Res. 2015, 54, 11525−11533. (21) Vecchio Ciprioti, S.; Paciulli, M.; Chiavaro, E. Application of different thermal analysis techniques to characterize oxidized olive oils. Eur. J. Lipid Sci. Technol. 2017, 119, 1−16. (22) Vyazovkin, S.; Burnham, A. K.; Criado, J. M.; Pérez-Maqueda, L. A.; Popescu, C.; Sbirrazzuoli, N. Ictac kinetics committee recommendations for performing kinetic computations on thermal analysis data. Thermochim. Acta 2011, 520, 1−19. (23) Kok, M. V.; Gul, K. G. Thermal characteristics and kinetics of crude oils and SARA fractions. Thermochim. Acta 2013, 569, 66−70. (24) Murugan, P.; Mani, T.; Mahinpey, N.; Asghari, K. The low temperature oxidation of Fosterton asphaltenes and its combustion kinetics. Fuel Process. Technol. 2011, 92, 1056−1061. (25) Li, J.; Mehta, S.; Moore, R.; Ursenbach, M., New insights into oxidation behaviours of crude oils. J. Can. Petrol. Technol. 2009, 48, 12−15. (26) Mothé, C.; Miranda, I. Study of kinetic parameters of thermal decomposition of bagasse and sugarcane straw using Friedman and Ozawa-Flynn-Wall isoconversional methods. J. Therm. Anal. Calorim. 2013, 113, 497−505. (27) Opfermann, J.; Kaisersberger, E.; Flammersheim, H. Model-free analysis of thermoanalytical data-advantages and limitations. Thermochim. Acta 2002, 391, 119−127. (28) Chrissafis, K. Kinetics of thermal degradation of polymers: complementary use of isoconversional and model-fitting methods. J. Therm. Anal. Calorim. 2008, 95, 273−283. (29) Wang, C.; Zhang, Y. Oil science of storage and transportation, second ed., China petroleum university press, Dongying, China, 2006. ACS Paragon Plus Environment

Page 14 of 22

Page 15 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 1. DSC curves for saturates at different heating rates of 4, 6 and 10 0C/min.

Figure 2. DSC curves for asphaltenes at different heating rates of 4, 6 and 10 0C/min.

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. DSC curves for aromatics at different heating rates of 4, 6 and 10 0C/min.

Figure 4. DSC curves for resins at different heating rates of 4, 6 and 10 0C/min.

ACS Paragon Plus Environment

Page 16 of 22

Page 17 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 5. DSC curves for crude oil at different heating rates of 4, 6 and 10 0C/min.

Figure 6. DSC curves for crude oil and its SARA fractions at the heating rate of 4 0C/min.

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. TG-DTG-DSC curves for saturates at the heating rate of 4 0C/min.

Figure 8. TG-DTG-DSC curves for aromatics at the heating rate of 4 0C/min.

ACS Paragon Plus Environment

Page 18 of 22

Page 19 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 9. TG-DTG-DSC curves for resins at the heating rate of 4 0C/min.

Figure 10. TG-DTG-DSC curves for asphaltenes at the heating rate of 4 0C/min.

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 11. TG-DTG-DSC curves for crude oil at the heating rate of 4 0C/min.

Figure 12. Comparison of the predicted curve from additive scheme and measured results of crude oil.

ACS Paragon Plus Environment

Page 20 of 22

Page 21 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Table 1. Properties of crude oil and its SARA fraction sample

viscosity at 20 0C (mPa·s)

API (°)

density 3 (g/cm )

saturates

Vishanskoe oil

37.2

29.6

0.8713

69.30

SARA fractions (%) aromatics resins 20.1

9.5

asphaltenes 1.1

Table 2. DSC curves parameters of crude oil and its SARA fraction in air atmosphere samples saturates aromatics resins asphaltenes crude oil

low temperature oxidation temperature peak range temperature 0 0 ( C) ( C) 89.8 – 411.5 327.5 124.0 – 401.5 337.0 139.5 – 374.5 323.0 129.0 – 365.5 327.5 109.8 – 416.5 329.5

heating rate 0 ( C /min) 4 4 4 4 4

high temperature oxidation temperature peak range temperature 0 0 ( C) ( C) 411.5 – 566.5 417.5 401.5 – 563.0 467.5 374.5 – 566.5 396.0 365.5 – 561.5 458.5 416.5 – 587.5 428.5

Table 3. Kinetic parameters for the crude oil and its SARA fractions obtained by Friedman and Ozawa–Flynn–Wall models samples saturates aromatics asphaltenes resins crude oil

methods Friedman OFW Friedman OFW Friedman OFW Friedman OFW Friedman OFW

LTO log(A) 7.63 7.96 7.11 7.13 7.26 6.33 6.13 6.69 9.65 9.21

Ea (kJ/mol) 119.04 ± 11.17 121.39 ± 4.05 115.40 ± 5.55 115.11 ± 4.32 139.34 ± 4.54 124.89 ± 4.95 102.20 ± 16.94 106.47 ± 12.03 144.36 ± 16.94 137.02 ± 27.55

ACS Paragon Plus Environment

HTO log(A) 8.11 7.25 8.73 9.10 6.27 6.06 8.02 8.87

Ea (kJ/mol) 152.32 ± 6.92 139.02 ± 15.07 135.92 ± 41.12 136.60 ± 31.08 121.86 ± 12.47 118.66 ± 13.05 146.60 ± 22.01 154.63 ± 31.30

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Graphical Abstract

ACS Paragon Plus Environment

Page 22 of 22