Oxidation Behavior and Kinetics of Light, Medium, and Heavy Crude ...

Article Cite This: Energy Fuels 2018, 32, 5571−5580

pubs.acs.org/EF

Oxidation Behavior and Kinetics of Light, Medium, and Heavy Crude Oils Characterized by Thermogravimetry Coupled with Fourier Transform Infrared Spectroscopy Chengdong Yuan, Dmitrii A. Emelianov, and Mikhail A. Varfolomeev* Department of Physical Chemistry, Kazan Federal University, Kazan 420008, Russia S Supporting Information *

ABSTRACT: The oxidation behavior of three crude oils was characterized by thermogravimetry coupled with Fourier transform infrared spectroscopy (TG−FTIR) to investigate the oxidation mechanism of crude oils. The results indicated that the entire oxidation process can be divided into three main reaction intervals: low-temperature oxidation (LTO) interval (90

5−7 14−16 2−3

13−15 19−21 7−8

the heating of the whole chamber to the temperature of 200 °C and with the heating of the transit line to 250 °C in an absorbance mode. The scans were measured in the spectral range from 4000 to 650 cm−1 with the resolution of 4 cm−1. The background spectra were measured after loading the sample inside the thermogravimetric balance. After starting the experiment, 16 scans were conducted every 20 s. 2.3. Kinetic Methods. In this research, the “model free” package embedded in the software thermokinetics by NETZSCH was employed to determine the kinetic parameters activation energy (Ea) and logarithm of the pre-exponential factor (log A) by Friedman and Ozawa−Flynn−Wall (OFW) isoconversional models using multiple heating rates. These isoconversional methods allow for the evaluation of the activation energy without determining the reaction model.23,24 Consequently, it can provide more reliable insights into the intrinsic kinetics of the complicated oxidation process. The Friedman method is the most common differential isoconversional method. For a linear non-isothermal program, the basic equation of the Friedman method can be written as22 ln(βi (dα /dT )α , i ) = ln A + ln f (α) − Eα /RTα , i

(1)

where β is the heating rate, E is the activation energy, α is the extent of conversion, A is the pre-exponential factor, and R is the gas constant. The value of Eα is estimated from the slope of a plot of ln(dα/dT) against 1/T. The OFW method is an integral isoconversional method. For a given α, the following equation is used:22

ln(β) = −1.0516(E /RT ) + constant

(2)

where the value of Eα is estimated from the slope of the linear plot of ln(β) against 1/T.

3. RESULTS AND DISCUSSION As shown in Table 1, crude oils 1, 2, and 3 have API gravities of 29.6°, 13.9°, and 32.1°, which belong to medium, heavy, and light crude oils, respectively. Figures 1, 4, and 7 show TG/DTG curves of crude oils 1, 2, and 3, respectively. The reaction intervals reflected by TG/DTG were very different among the three oils. Generally, the reaction process of crude oils can be divided into three different intervals based on TG/DTG curves: low-temperature oxidation (LTO), fuel deposition by coking reactions, and high-temperature oxidation (HTO).3,7,10,11,25 There are also some studies where only LTO and HTO were reported.4,20 Before any analysis of reaction intervals and reaction mechanism, it must be mentioned that heating rates had an obvious effect on the temperature range of reaction intervals as reported in other studies, where the different heating rates were used.8,9 The onset temperatures and peak temperatures of all reaction intervals were shifted to higher temperatures with the increase of the heating rate. Therefore, in 5572

DOI: 10.1021/acs.energyfuels.8b00428 Energy Fuels 2018, 32, 5571−5580

Article

Energy & Fuels

Figure 1. TG/DTG curves of crude oil 1 at different heating rates of 5, 10, 15, and 20 K/min.

Figure 4. TG/DTG curves of crude oil 2 at different heating rates of 5, 10, 15, and 20 K/min.

Figure 2. Variation of the absorbance of CO2, CO, H2O, and hydrocarbons in FTIR spectra as a function of the temperature for crude oil 1.

Figure 5. Variation of the absorbance of CO2, CO, H2O, and hydrocarbons in FTIR spectra as a function of the temperature for crude oil 2.

Figure 3. TG/DTG curves of crude oil 1 during the heating process in air and nitrogen atmospheres at the heating rate of 10 K/min.

Figure 6. TG/DTG curves of crude oil 2 during the heating process in air and nitrogen atmospheres at the heating rate of 10 K/min.

5573


Article

Energy & Fuels

Figure 10. TG/DTG curves of crude oil 3 during the heating process in air and nitrogen atmospheres at the heating rate of 10 K/min. Figure 7. TG/DTG curves of crude oil 3 at different heating rates of 5, 10, 15, and 20 K/min.

Figure 11. Dependence of activation energy upon the conversion degree calculated with OFW and Friedman methods for crude oil 1 (LTO). Figure 8. Variation of the absorbance of CO2, CO, H2O, and hydrocarbons in FTIR spectra as a function of the temperature for crude oil 3.

Figure 9. Components of the saturate fraction of crude oils 1 and 3. Figure 12. Dependence of activation energy upon the conversion degree calculated with OFW and Friedman methods for crude oil 1 (coking process).

this study, when it comes to the analysis related to the temperature or temperature range, the data for the heating rate of 15 K/min will be used as a reference. Figures S1, S3, and S4 of the Supporting Information show the FTIR spectra of 5574


Article

Energy & Fuels

Figure 13. Dependence of activation energy upon the conversion degree calculated with OFW and Friedman methods for crude oil 1 (HTO).

Figure 15. Dependence of activation energy upon the conversion degree calculated with OFW and Friedman methods for crude oil 2 (coking process).

Figure 14. Dependence of activation energy upon the conversion degree calculated with OFW and Friedman methods for crude oil 2 (LTO).


gaseous products released in the oxidation processes of crude oils 1, 2, and 3, respectively. It can be seen from these data that main products observed in the gas phase are water, hydrocarbons, carbon monoxide, and carbon dioxide. Stretching vibration bands of the characteristic groups of ketones, aldehydes, alcohols, and carbon acids are not identified. However, it does not mean that they are not formed as intermediate products in oxidation processes. They can have low volatility, and after a certain temperature, they participate in pyrolysis and decomposition reactions with the formation of hydrocarbons, water, and carbon dioxide. Figures 2, 5, and 8 show the variation of the absorbance of water, CO2, CO, and hydrocarbons as a function of the temperature for crude oils 1, 2, and 3, respectively. According to the Lambert−Beer law,26 a linear relationship exists between the absorbance of a typical functional group and the concentration of the gaseous product containing this functional group. Therefore, the relative concentration of one effluent gas during the oxidation process could be represented by the absorbance of the typical functional group contained in this gaseous product.

3.1. TG/DTG−FTIR Analysis of Crude Oil 1 (Medium Oil). For medium oil (Figure 1), there are three distinct reaction intervals during the entire process. The first interval was before 400 °C, which was considered as the LTO interval. The mass loss in the LTO interval was about 78−80%. In the LTO interval, two types of oxidation reactions are believed to be significant reaction pathways: oxygen addition reactions and isomerization and decomposition reactions.27 The oxygen addition reaction can take place even at normal ambient conditions.28 In the lowest temperature range of the LTO interval, the formation of hydroperoxides through the oxidation addition reaction is believed to be the main oxidation reaction path as the following:27 RH + O2 → ROOH

(3)

where RH is any possible aliphatic hydrocarbon or the aliphatic chains attached to naphthenic rings or aromatic rings in the molecule of other components of crude oils and ROOH is a hydroperoxide. Reaction 3 is a simple sum of a series of 5575


Article

Energy & Fuels

Figure 17. Dependence of activation energy upon the conversion degree calculated with OFW and Friedman methods for crude oil 3 (LTO).


were produced before 400 °C had a good corresponding relationship with the reaction intervals shown in DTG curves, and both the maximum of the mass loss rate and the largest absorbance of hydrocarbons occurred at 300 °C, which means that the release of hydrocarbon had the most important effect on the mass loss in this temperature range. These effluent hydrocarbons before 400 °C were assumed to be caused by the evaporation of light fractions in crude oils because the LTO reaction does not produce new light hydrocarbons and the occurrence of cracking reactions to produce light hydrocarbons requires a higher temperature.15 This implies that evaporation played a significant role and controlled the mass loss in the entire LTO region. To further prove this assumption, the pyrolysis experiment of crude oil 1 in a nitrogen atmosphere was conducted to compare to that in an air atmosphere, and the result is shown in Figure 3. In the pyrolysis process in a nitrogen atmosphere, two stages can be observed. Obtained FTIR spectra of evolved gases show (Figure S2 of the Supporting Information) that hydrocarbons are the main products of the processes, which take place in a nitrogen atmosphere. The first stage is the evaporation stage caused by the evaporation of hydrocarbons with a low boiling point and the rupture of weak chemical bonds.15 This stage also finished before about 400 °C, such as the LTO interval at air conditions. Simultaneously, the mass loss trend was also similar to the LTO interval in an air atmosphere, except that the mass loss was slightly slower than in an air atmosphere. The final mass loss in this stage is about 71%, which is close to the total mass loss in the LTO interval in an air atmosphere. All of these signals further confirmed the assumption that, for crude oil 1, evaporation played a significant role and controlled the mass loss in the entire LTO region. As we mentioned before, in the earliest stage of the LTO interval, the oxygen addition reaction to produce hydroperoxides was a dominant pathway. When the temperature was increased to a higher level, the formed hydroperoxides started to participate in a series of decomposition and isomerization reactions. In this study, the start temperature of the isomerization and decomposition reactions is believed to be at 180 °C, at which H2O was released as a typical product from

Figure 18. Dependence of activation energy upon the conversion degree calculated with OFW and Friedman methods for crude oil 3 (coking process).

elementary free-radical reactions, which is believed to start from the following reactions:27 R• + O2 → RO2• •

(4) •

RO2 + RH → ROOH + R

(5)

As shown in DTG curves (Figure 1), the LTO interval seems to be more complicated and should include multiple different processes. As seen from DTG curves, at approximately 50 °C, rapid mass loss took place, the mass loss rate displayed a rapid linear increase, and then the increased trend was slowed down at about 200 °C. Also, in this temperature range, the release curve of hydrocarbon exhibited a small shoulder peak before 200 °C. After this, the mass loss rate showed a further rapid increase, reached a peak at about 300 °C, then quickly decreased, and finally reached a level of approximately 400 °C. Correspondingly, a sharp and large peak was observed in the release curve of hydrocarbons (Figure 2) in the temperature range of 200−400 °C next to the small shoulder peak observed before 200 °C. The temperature intervals where hydrocarbons 5576


Article

Energy & Fuels Table 3. Kinetic Parameters for the Three Crude Oils by Friedman and OFW Methods LTO sample

method

log(A)

crude oil 1

Friedman OFW Friedman OFW Friedman OFW

3.70 3.86 4.07 4.20 3.58 3.47

crude oil 2 crude oil 3

coking process

Ea (kJ/mol)

log(A)

± ± ± ± ± ±

8.02 6.78 11.02 9.61 6.21 4.47

64.77 61.78 70.18 67.15 52.11 49.95

3.68 3.01 4.69 3.79 1.27 1.96

some isomerization and decomposition reactions to produce ketones or alcohols as follows:27 ROOH → RO• + •OH

(6)

RO• + RH → ROH + R•

(7)

•

OH + RH → R• + H 2O

Reactions 6−8 can be simply summarized to reaction 727 •

(9)

•

where RO is an alkoxy free radical, OH is a hydroxyl radical, ROH is an alcohol, and R• is an alkyl free radical. Reaction 7 was considered as the most generally accepted pathway related to the hydrogen abstraction reactions and the formation of alcohols during the oxidation process of crude oils.27,29 Reaction 9 is believed to be one of the main reaction pathways for the decomposition of hydroperoxides to produce most alkyl free radicals. At approximately 300 °C, where the maximum of the mass loss rate and the largest absorbance of hydrocarbons occurred, it was also observed that CO started to be produced, which is followed by the release of CO2 at 328 °C. Simultaneously, the release amount of H2O reached the maximum. For the LTO region, the production of CO2 and CO means that the reaction pathway was transferring in another way. Therefore, it can be concluded that the isomerization and decomposition reactions of hydroperoxides to produce CO2 and CO became more important at 300−400 °C. Reaction 9 is just one type of isomerization and decomposition reaction. Actually, there are many different reaction pathways for the isomerization and decomposition reactions of hydroperoxides, which results in the formation of many oxidation products, including ketones, alcohols, aldehydes, carboxylic acids, etc.27 In addition, instead of the intermolecular reaction, as shown in reaction 5, some intramolecular abstraction of hydrogen by an alkylperoxy radical in the same molecule could take place, which is another type of isomerization reaction that also plays a significant role in the formation of many oxidation products,27 for instance, the decomposition of carboxylic acid to produce CO2.30 In brief, the results of the isomerization and decomposition reactions are to form more highly oxidized products, including some ketones, alcohols, aldehydes, and carboxylic acids, and their condensation compounds, also the same bond-scission products of CO2, CO, and H2O. After the LTO reaction, these oxygen-rich products formed from the conversion of hydroperoxides were left, which is usually called the LTO residue. Therefore, the overall reaction of the isomerization and decomposition processes can be described as the following reaction:27 ROOH → LTO residue + CO2 + CO + H 2O

log(A)

± ± ± ± ± ±

5.42 5.48 4.23 5.17 6.60 5.48

143.10 127.36 185.58 165.45 98.16 79.26

40.68 25.67 32.76 17.63 2.55 2.60

Ea (kJ/mol) 121.64 124.90 103.80 122.68 134.01 118.16

± ± ± ± ± ±

2.33 5.83 16.34 1.18 9.47 5.57

The second reaction interval, as shown in DTG (Figure 1), was at 400−500 °C, which was attributed to the formation of coke through coking reactions. The mass loss in this interval was about 5−7%. Coke is the main fuel for the high-temperature combustion reaction. In this stage, the maximum release amount of both CO2 and CO was obtained at about 430 °C and small shoulder peaks were formed at 300−500 °C. Simultaneously, the release of hydrocarbons could still be observed at this stage. Therefore, it was believed that coke was mainly produced by the oxidative cracking of the LTO residue formed in the LTO reaction, which can be expressed as reaction 11.27

(8)

ROOH + 2RH → 2R• + ROH + H 2O

HTO

Ea (kJ/mol)

LTO residue → coke + CO2 + CO + H 2O + hydrocarbons

(11)

In a real ISC process on the field, coke is also believed to be produced by another reaction pathway, which is the pyrolysis of those non-volatile hydrocarbons, which initially existed in crude oils.27 Interestingly, as shown in Figure 3, the temperature range of the second stage (thermal cracking to produce coke, 400−500 °C) in the pyrolysis process of crude oil at nitrogen conditions was almost the same as that at air conditions. The similar temperature range of the pyrolysis of these non-volatile hydrocarbons in a nitrogen atmosphere is also reported by Hao et al.15 This means that, if some non-volatile hydrocarbons, which initially existed in crude oils, remained after the LTO reaction and were not be oxidized in an air atmosphere, they must be related to the formation of coke by pyrolysis or oxidative cracking. The third reaction interval, as shown in DTG (Figure 1), took place at about 500−650 °C, which was considered as HTO. The mass loss in this interval was about 13−15%. It is widely accepted that the main reaction that occurred in the HTO interval is the combustion of the formed coke.1,3 This can also be partly confirmed by comparing the mass loss behavior of crude oil in air and nitrogen atmospheres (Figure 3). In both air and nitrogen conditions, the second stage (i.e., coke formation stage) was at about 400−500 °C. In a nitrogen atmosphere, after the coke formation, the mass loss almost did not change with the increasing of the temperature and there was no more reaction interval being observed (Figure 3). However, in an air atmosphere, the formed coke can be further combusted and release CO2 and CO, which leads to the third reaction interval (HTO). As shown in Figure 2, a sharp peak was observed for both CO2 and CO at almost the same temperature ranges. The largest absorbance of the two peaks was far higher than the absorbance of CO2and CO in the stages of coke formation and LTO. This means that the combustion of coke released a substantial amount of CO2 and CO. The release amount of CO2 was much higher than that of CO.

(10) 5577


Article

Energy & Fuels

released. Therefore, for the heavy oil, this process can be described as follows:

However, almost no water was observed in the combustion process of coke. Therefore, this process can be expressed as the following reaction: coke + O2 → CO2 + CO

LTO residue → coke + CO2 + H 2O + hydrocarbons (13)

(12)

The HTO interval of the heavy oil was at about 500−650 °C, which is similar to the medium oil. However, the combustion of coke was stronger because more coke was formed in the second stage. The mass loss was about 19−21%. Similar to the HTO interval of the medium oil, a sharp peak was observed for both CO2 and CO, and the release amount of CO2 was far higher than that of CO. However, a small amount of H2O was detected in the coke combustion process of the heavy oil, while no H2O was observed in this stage for the medium oil. This implies that coke obtained from the heavy oil might contain more hydrogen than that of the medium oil. Therefore, the coke combustion process for the heavy oil can be expressed as follows:

3.2. TG/DTG−FTIR Analysis of Crude Oil 2 (Heavy Oil). As seen from Figure 4, also three main reaction intervals were observed, which were LTO, coking reactions, and HTO, in turn, from low temperature to high temperature. However, in the LTO interval, DTG curves showed a successively descending and ascending process, which is different from that of the medium oil, where a process in which the mass loss rate became slow was observed during the descending process. This phenomenon can be attributed to the evaporation of light components. As shown in Figure 5, there was almost no hydrocarbon being observed before 200 °C and only one broad peak with a maximum at 370 °C was observed at 200−400 °C, while a distinct shoulder peak before 200 °C and a sharp peak at 200−400 °C with a maximum at 300 °C were shown in the release curve of hydrocarbons for the medium oil (Figure 3). This implies that the heavy oil (crude oil 2) contained less lighter components that can be vaporized before 200 °C compared to the medium oil. However, the heavy oil itself contained more heavier components that can be vaporized after 300 °C. The total mass in the LTO interval was about 64−66%, which is smaller than that of the medium oil. It was also found that mass loss behavior (total mass loss and mass loss rate) was very similar between air and nitrogen conditions (Figure 6), which once again indicates that evaporation controlled the total mass loss in the LTO interval, as we mentioned in the medium oil part. However, the similarity of the mass loss behavior between air and nitrogen conditions is higher for the heavy oil compared to the medium oil, which implies that for the heavy oil, the oxidation reaction in the LTO interval that can lead to mass loss is weaker than that of the medium oil. For the medium oil, the release of H2O was observed at about 180 °C. However, for the heavy oil, the release of H2O was observed at about 280 °C, where the maximum of the mass loss rate occurred. This means that, for the heavy oil, those isomerization and decomposition reactions of hydroperoxides to produce H2O require a higher temperature and formed oxygen-containing compounds are less volatile and more temperature-stable than in the medium oil. Simultaneously, unlike the medium oil, there were no CO2 and CO being observed in the LTO interval for the heavy oil, which signifies that the isomerization and decomposition reactions to produce CO2 and CO might not occur in this stage. All of these phenomena give a hint that the LTO reaction of the heavy oil might have some differences from the medium oil. Being similar to the medium oil, the coking process also occurred at approximately 400−500 °C for the heavy oil. However, it can be seen that this process was clearly stronger compared to that of the medium oil. The mass loss was 14− 16%, which is 2 times higher than that of the medium oil, which indicates that more LTO residue was obtained in the LTO reaction and participated in the coking process to form more coke. This can be attributed to the nature of the heavy oil that it contained more resins and asphaltenes, especially asphaltenes that are the main components that contribute to the formation of coke.6 Simultaneously, the gaseous products were also different from those of the medium oil. A smaller amount of CO2, no CO, but a bigger amount of hydrocarbon were

coke + O2 → CO2 + CO + H 2O

(14)

3.3. TG/DTG−FTIR Analysis of Crude Oil 3 (Light Oil). As shown in Figure 7, the DTG curves of the light oil were very different from the other two oils. The mass loss rate rapidly increased, even at a very low temperature. The maximum mass loss rate occurred at only 140 °C. From the release curve of hydrocarbons (Figure 8), two large and sharp peaks were observed in the LTO interval. The mass loss was more than 90% in the LTO interval. The big difference in evaporation behavior between the light and medium oils can be attributed to the different contents of VOCs and alkanes. As shown in Table 1, the content of VOCs of the light oil was 24.03%, which is far higher than that for the medium oil (15.03%). VOCs are the lightest hydrocarbon components in crude oils. They are vulnerable to be evaporated at a low temperature range. Simultaneously, Figure 9 shows that the content of the light fractions of C11−C21 hydrocarbons in the light oil was 84.59%, which is also much higher than for the medium oil (70.55%). The high content of VOCs and the light fractions of C11−C21 are considered as the main reasons why the light oil had a much stronger evaporation than that of the medium oil. In comparison of the TG/DTG curves in air conditions with a nitrogen atmosphere (Figure 10), one can see that they were almost coincident before 300 °C, which means that evaporation totally controlled the total mass loss in the LTO interval. Being similar to the medium oil, at about 200 °C, the release of H2O was observed (Figure 8), which means that the isomerization and decomposition reactions of hydroperoxides to produce H2O started to become more obvious. At about 300 °C, where the maximum release amount of hydrocarbon took place, there was an obvious inflection point on the DTG curves, which indicates the descent velocity of the mass loss rate. This implies that the reaction pathway was changing and the new reactions also played an important role in the mass loss. Simultaneously, at approximately 300 °C, it started to release CO2 and CO, and subsequently, the release amount of H2O reached a maximum. On the basis of these signals, it can be concluded that the isomerization and decomposition reactions of hydroperoxides to produce CO2 and CO started to prevail at 300−400 °C. The second stage at 400−500 °C did not show a peak in DTG curves, and only a slight descending was observed, which is very different from the medium and heavy oils. Also, a straight line with a very small inclination was observed in TG 5578


Article

Energy & Fuels

respectively. It can be concluded that the heavier the crude oil, the higher the activation energy in the LTO stage. For the coking stage, activation energies also increased with the decrease of the API gravity of the crude oils. However, for the HTO stage, activation energies were similar for the three oils. As we mentioned before, in this stage, the only significant reaction is the combustion of coke. Therefore, the value of activation energies mainly depends upon the quality of the formed coke by the oxidative cracking reactions of the LTO residue. The similar activation energies implied that the quality of the formed coke from the three crude oils had no significant difference.

curves. The total mass loss in this stage was only 2−3%. Therefore, we can conclude that the coke formation reaction was weak. This can also be verified by the TG/DTG curves (Figure 10) in a nitrogen atmosphere, where only a small inflection was observed in the coke formation stage. However, it can be seen from Figure 8 that the release behavior of CO2, CO, H2O, and hydrocarbon was similar to that of the medium oil. Therefore, it can be inferred that the same coking process occurred as in the medium oil. However, the process was weaker as a result of the nature of the light oil. Also, a weak HTO reaction with a mass loss of 7−8% was observed at 500− 650 °C. In comparison to the medium and heavy oils, the entire oxidation process of the light oil can be described as follows: the lighter oil had a more severe evaporation in the LTO interval, and a smaller amount of coke was formed in the coking process, which consequently leads to a weaker coke combustion process in the HTO interval. This characteristic of the light oil determined that an ISC process is not suitable for light oil reservoirs because not enough fuel can be formed to sustain a stable combustion front. 3.4. Kinetic Analysis. The TG data obtained from four heating rates β = 5, 10, 15, and 20 K/min (as shown in Figures 1, 4, and 7) were used to determine the kinetic parameters using Friedman and OFW methods. As we mentioned before, the entire reaction process can be divided into three stages: LTO, FD, and HTO. In this work, the kinetic parameters for three stages were calculated separately. The first step is to separate the TG dynamic curves into three stages according to the DTG curves. After the separation, the data of each stage were separately exported in the ASCII code and sent to the NETZSCH thermokinetics software for further evaluation of the kinetic parameters. Figures 11−19 show the dependence of Eα upon α calculated by OFW and Friedman methods for different reaction intervals during the non-isothermal oxidation process of the three oils. For all reaction intervals, it was found that Eα slightly varied with α. This is not unexpected as a result of the complicated reaction processes and the complex composition of crude oils. For LTO, it was found that Eα showed a slightly increased trend with α for all three oils, which implies that LTO is not a single-step process. However, the variation of Eα with α was not significant. This means that LTO is more like a multistep process that has one step whose mass loss rate determines the overall mass loss. As analyzed in section 3.1, on the basis of the release curve of hydrocarbons and the good corresponding relationship with the reaction intervals shown in DTG curves, evaporation played a significant role and possibly controlled the mass loss in the entire LTO region. For HTO, Eα was also found to slightly change (either increase or decrease) with α. However, in general, Eα can be considered to be a constant average value. In HTO, the combustion of coke was considered as the only significant reaction. Therefore, Eα is mainly dependent upon the quality of the formed coke. Also, it should be noted that two kinetic approaches Friedman and OZW gave practically the same kinetic parameters, despite the different theoretical basis of these model-free methods. This fact confirms the good quality of experimental data and supports the correct separation of the oxidation behavior of studied crude oils on three stages. For the convenience of the comparison, the values of log(A) and activation energies at α = 0.5 were taken and presented in Table 3. For the LTO stage, the activation energies were between 49 and 53 kJ/mol, between 61 and 65 kJ/mol, and between 67 and 71 kJ/mol for the light, medium, and heavy oil,

4. CONCLUSION On the basis of the results of the TG−FTIR technique, the oxidation process of crude oils was divided into three stages and the possible reaction pathways were described. The following conclusions can be drawn: (1) The oxidation process of light, medium, and heavy crude oils can be divided into three reaction intervals: LTO interval (