Analysis of Characteristic Changes of Blended Very Low Sulfur Fuel Oil on Ultrasonic Frequency for Marine Fuel

Abstract

The demand for very low sulfur fuel oil (VLSFO) with a sulfur content of less than 0.5% has increased since the IMO2020 regulations were published. However, most VLSFOs for marine fuel are produced by blending two fuel oils with different sulfur contents, which causes some problems, such as sludge formation. This study investigates the effect of ultrasonic irradiation frequency (25 and 72 kHz), ultrasonic irradiation time (0, 12, and 24 h), and the blending ratio (marine gas oil (MGO) and bunker-A (B-A) with weight ratios of 25:75, 50:50, and 75:25 on the characteristics of blended VLSFO. After 12 h of irradiation time and a frequency of 25 kHz, the amount of carbon residue decreases with increasing MGO content; it decreases by 33% for 75% MGO. However, at 72 kHz, the carbon residue increases with increasing MGO content, implying that the change in carbon residue depends on the ultrasonic frequency. After 24 h, the carbon residue does not decrease in any scenario; however, it does increase in some cases due to asphaltene reaggregation caused by excessive ultrasonic irritation. The sulfur content decreases by approximately 4% for the 100% B-A condition.

1. Introduction

Since the adaptation of MARPOL (The International Convention for the Prevention of Pollution from Ships) Annex VI in 1997, which placed limitations on the amounts of air pollutants in the exhaust gas emissions from ships, marine emission regulations have become more stringent [1]. Shipping companies and ship owners are required to install new equipment or retrofit their ships to comply with these regulations [2]. The methods to comply with the sulfur oxide (SO_X) regulations include pre-combustion approaches, such as using very low sulfur fuel oil (VLSFO) or operating LNG-fueled ships, and post-treatment approaches, such as exhaust gas cleaning systems (i.e., scrubbers). Exhaust gas cleaning systems have some disadvantages, such as high initial investment, and operating LNG-fueled ships requires building a new vessel [3]. Therefore, to comply with the SO_X regulations, shipping companies prefer to use VLSFO that do not require additional treatment facilities. The use of VLSFO accounted for 78.7% of all compliance approaches, followed by scrubbers at 19.5% and LNG-fueled ships at 1.8%, as of November 2020 [4].

VLSFO for marine fuel can be produced as straight-run, desulfurized, or blended VLSFO. Straight-run VLSFO is produced using sweet crude oil, which has a sulfur content of 0.5% or less. It has the advantage of higher stability compared to blended VLSFO, but it is expensive and only supplied to major ports since sweet crude only accounts for approximately 20% of the total crude oil produced [5]. Desulfurized VLSFO, which is used in conventional desulfurization facilities through the hydrodesulfurization method, requires a significantly high initial investment, making it difficult for small refineries to adopt such facilities [6]. Moreover, the hydrodesulfurization method requires a continuous hydrogen supply. Other technologies, such as oxidative desulfurization (ODS) [7] and ultrasonic-assisted oxidative desulfurization (UAOD) [8], have been studied, but they are challenging to scale-up for large-scale fuel processing. Lastly, blended VLSFO is produced by blending a high sulfur fuel oil (HSFO) with a VLSFO, and most VLSFOs supplied onboard are manufactured using the blending method since it is a simpler process with higher production than other methods [9]. However, the qualities of blended VLSFO, such as sedimentation stability, depend on the manufacturer, duration of the production process, and refinement process of feedstocks, resulting in instability and incompatibility issues [2]. This is because the asphaltenes derived from various hydrocarbon geneses have different compounds, structures, and solubility parameters, which affect sedimentation stability in a constant composition fuel. For instance, an increase in fuel stability is associated with an increase in the average number of aromatic rings in the asphaltene cluster [10], and n-paraffins content is related to the total sediment potentials of asphaltenes [11]. To address the issues concerning instability, several studies have investigated the influence of mixture compositions, such as n-paraffins and residue marine fuels with different genesis on stability [11,12]; moreover, methods have been proposed to determine the stability of fuel mixtures and the solubility of asphaltenes [11,13]. Additionally, although additives have been studied to improve the stability of marine fuels, some additives, such as cutter stocks, still cause quality problems onboard [14,15,16]. For instance, a survey of 192 shipping companies using VLSFO revealed that 55% of them experienced problems in quality [16].

The cavitation phenomenon is explained next. First, fluids with high velocities induce pressure fluctuations within their bulk. When the fluid pressure is below the vapor pressure, cavities (microbubbles) are formed, which shrink or expand due to repeated pressure fluctuations. When the fluid pressure exceeds the saturated vapor pressure, these cavities instantaneously collapse [17]. This phenomenon occurs in high-speed rotators, such as propellers, but it can be artificially generated by exposing the fluid to ultrasound with a frequency of 20 kHz or higher. Ultrasound is commonly used to enhance chemical reactions or break chemical bonds. To elaborate, when cavitation occurs, it forms local hot spots with an instantaneous temperature and pressure of approximately 5000 °C and above 1000 atm, respectively, accelerating oxidative chemical reactions [8]. UAOD is a technology that accelerates the rate of oxidative reactions using ultrasonic cavitation [18,19,20]. Ja’fari et al. [8] reported that ultrasound effectively promotes the desulfurization process, and Margeta et al. [19] reported that UAOD affects up to 87% of the desulfurization of dibenzothiophene in model diesel oil. However, these studies focused on the effect of ultrasonic irradiation to promote the oxidation reaction rate in sulfur with oxidants.

Furthermore, cavitation physically breaks chemical bonds and generates free radicals or micro-emulsions [21]. Consequently, when fuel oil is exposed to ultrasonic irradiation, aromatic hydrocarbons disintegrate into aliphatic hydrocarbons [22], and viscosity is reduced. However, the results of the studies on the changes in viscosity by ultrasonic irradiation are inconsistent [23], and some authors have even reported contradictory results [24,25,26]. In terms of fuel storage, after exposure to ultrasound, the kinematic viscosity of heavy oil becomes non-monotonic based on relaxation time and ultrasonic power density [27]; unfortunately, more detailed studies on this issue are rare. Therefore, the physical change in fuel exposed to ultrasonic irradiation is difficult to predict because of the re-aggregation of compounds cracked by ultrasound [26], causing dependence on frequency, irradiation time, and fuel compounds. Moreover, most of the aforementioned studies have been conducted with probe-type ultrasound units, resulting in the scale-up problems [28]. In our previous study [29], the effect of ultrasound on the content of high molecular weight compounds, such as n-paraffins of over C₁₉, and sulfur compounds was investigated for blended VLSFO. This content verifiably decreased due to ultrasound. However, the aforementioned previous study was limited considering the scale-up to treat large volumes of fuel because it was conducted with a probe-type, and ultrasonic frequency effects were not discussed.

In this study, the effect of ultrasonic frequency (high and low frequencies), ultrasonic irradiation time, and blending ratio on blended VLSFO is investigated to improve its quality. Moreover, flowcell-type ultrasound equipment is utilized to treat large volumes of fuel. Blended VLSFO is prepared by blending MGO (marine gas oil) as a VLSFO with B-A (bunker-A) as an HSFO; moreover, the changes in properties, such as carbon residues, kinematic viscosity, density, pour point, and sulfur, are analyzed considering blending ratios, ultrasonic irradiation times, and frequencies.

2. Experiments

2.1. Materials and Analysis Methods

The feedstocks for blended VLSFO are MGO, which is a straight-run VLSFO, and B-A, which is an HSFO [30]. B-A is an approximately 99:1 mixture of gas oil and residual oil. As listed in

Table 1. Blending ratios of all cases for ultrasonic irradiation experiments.

Case	MGO (wt.%)	B-A (wt.%)	Total (kg)
EBO1	75	25	40
EBO2	50	50	40
EBO3	25	75	40

, MGO and B-A are blended with mass ratios of 75:25, 50:50, and 25:75 (hereafter referred to as EBO1, EBO2, and EBO3, respectively, and EBO is an abbreviation for experimental blending oil). B-A is filled in a tank, heated up to 50 °C, and then MGO is added while circulating B-A using an agitator and pump. After adding MGO, EBOs are circulated for an additional 2 h.

Carbon residue, kinematic viscosity, density, pour point, and sulfur were analyzed by the Korea Petroleum Quality and Distribution Authority, an accredited testing and certification institution. Carbon residue analysis was performed using a micro carbon residue tester (Model: ACR-M3) from TANAKA Scientific Ltd. as per ISO 10370. Kinematic viscosity was analyzed using a viscometer (Maker: Cannon, Model: CAV 2000 Series) and density meter (Maker: Anton Paar, Model: DMA 5000 M), following the ISO 3104 and ISO 12185 standards, respectively. Kinematic viscosity is defined as the ratio of viscosity to density. Pour point was investigated using an MPC-602 from TANAKA Scientific Ltd., following the ISO 3016 standard. Lastly, sulfur was explored using a UV fluorescence spectrophotometer (Maker: Mitsubishi, Model: TS-100V) in compliance with ISO 8754.

2.2. Flowcell-Type Ultrasound Equipment

A typical probe-type ultrasound unit with a cup chamber is used in most sonochemistry lab studies. In this unit, the cavitation effect per unit volume is strong near the horn tip but reduces away from it [31]. Therefore, multiple ultrasonic transducers are needed to maintain the power density constant when the fuel treatment capacity increases, making scale-up difficult.

The flowcell-type ultrasound equipment, as shown in Figure 1, has a lower cavitation effect per unit volume than the probe-type unit in which the horn tip is directly exposed to fuel. However, the capacity can be increased relatively easily by installing as many transducers as needed, increasing the practicality for various industrial sectors [32]. Nevertheless, the quantitative effect between flowcell- and probe-type ultrasound units is difficult to compare because power density depends on the distance from the horn tip and various environmental parameters. Indeed, in terms of structure, the transducers of the probe-type are commonly arranged on a single side, but that of the flowcell-type are arranged around the flow path, inducing the change in power density per unit volume. In addition, compared to the probe-type, which can only be designed for a single frequency, the flowcell-type can be designed for multiple frequencies. The ultrasonic irradiation capacity of the flowcell-type equipment designed in this study is 17 L, and the maximum power is 1500 W. Moreover, it can operate at two different frequencies, 25 and 72 kHz. The fuel circulation tank has a volume of 100 L with a closed-off structure, except for a ventilation cock on the top. An EBO is exposed to ultrasound while passing through the transducer assembly, as indicated by the red box in Figure 1.

2.3. Ultrasonic Irradiation Test

The ultrasonic irradiation test for each blending ratio is conducted at 25 and 72 kHz to compare the ultrasonic effects at low- and high- frequencies. An EBO is exposed to ultrasound for 24 h, and samples are taken before ultrasonic irradiation and after 12 and 24 h based on ultrasonic irradiation time.

3. Results and Discussion

3.1. Carbon Residue

Carbon residue represents the amount of carbon remaining in a sample after evaporating petroleum products by heating to high temperatures above 500 °C. Generally, this carbon residue is composed of high molecular weight compounds such as asphaltenes, and it increases with an increase in heavy fuel oil, leading to ignition delay, after-burning, and deterioration in engine combustibility [33]. B-A has a carbon residue of 0.52%, and MGO has a negligible amount. The carbon residues of EBOs are 0.08, 0.18, and 0.3%, for EBO1, EBO2, and EBO3, respectively.

As ultrasonic frequency increases, the formation, growth, and collapse times of cavities decrease; this reduces the radius of collapsing cavities. Since cavitation energy is proportional to the square of cavity radius, cavity radius rapidly decreases with increasing frequency. Additionally, the reduction in cavity radius forms more cavities per unit volume, increasing cavity density. Consequently, an increase in frequency decreases cavitation energy but increases cavitation density. In contrast, as frequency decreases, cavitation energy increases, but cavity density decreases [34]. Accordingly, as cavity characteristics depend on frequency, the effect of ultrasound also differs with the particle size of sample; the smaller the particle size, the greater the effect of high frequencies, and the larger the particle size, the greater the effect of low frequencies [35]. However, the effect of ultrasound is difficult to explain through only the correlation between particle size and frequency because fuel consists of various hydrocarbon compounds. Indeed, the results in this study are not monotonic with regard to the ultrasonic frequency but are affected by both the blending ratio and frequency simultaneously. Furthermore, the physical and chemical properties of asphaltenes, which are influenced by ultrasound, should be considered to investigate the effect of ultrasound on fuel.

Ultrasound affects fuel such that the chemical bonds of high molecular weight compounds (hereafter referred to as heavy compounds), such as asphaltenes, are broken down into lighter ones, generating free radicals. Nevertheless, when cavitation persists for a specific time, radical formation rate decreases, and radicals branch to other heavy compounds, producing heavier compounds [25]. The point in time when this inflection occurs is referred to as the optimal radiation time. Furthermore, ultrasound changes the solubility of heavy compounds because aromatic compounds are converted into aliphatic compounds [22]. However, aromatic compounds play a significant role in dissolving resins, which absorb heavy compounds [35]; therefore, decreasing the aromatic compounds reduces the solubility of heavy compounds in the fuel [36]. Consequently, the mechanism underlying the effect of cavitation on carbon residue can be summarized as follows: First, heavy compounds crack into lighter compounds, which reduces carbon residue. Second, free radicals are produced during the cracking process but aggregate with other heavy compounds to form even heavier compounds, which are observed as a carbon residue, when the optimal irradiation time is exceeded. Third, the resin solubility in fuel decreases because aromatic compounds are converted into aliphatic compounds, precipitating asphaltenes; consequently, carbon residue increases.

Figure 2 shows the variations in carbon residue with an irradiation time of 12 h, and the value of y-axis is normalized by the initial carbon residue. In Figure 2a, the carbon residue in EBO1 decreases by 25% at a frequency of 25 kHz. The heavy compounds in 25% B-A of EBO1 are cracked because the low frequency scenario has a higher energy, which is effective in cracking heavy compounds, than the high frequency scenario, and the free radicals or resins, formed in cracking process, are re-dissolved in 75% MGO. In contrast, the carbon residue in EBO3 increases by 33% at the same frequency. Although the heavy compounds in 75% B-A are cracked, the free radicals or resins are incompletely dissolved in 25% MGO with lighter compounds because MGO is only 25%; thus, free radicals or resins are combined with other heavy compounds. Simultaneously, the solubility of asphaltenes decreases due to the cracking of aromatic compounds, which accelerates and increases carbon residue. In the previous study [29], which investigates the effect of ultrasound with 19.8 kHz on blended VLSFO, the alkanes of C₁₉ or higher were decreased in a blended fuel oil with a high VLSFO content more than that with a high HSFO content. This is consistent with the results of this study that the heavy compounds decreased by the ultrasound in the similar frequency.

In Figure 2b, the aforementioned trends are reversed at a frequency of 72 kHz. The carbon residue in EBO1 increases by 16%, while that of EBO3 decreases by 9.5%. As frequency increases, the light compounds are effectively cracked because cavity energy decreases, and cavity density increases; thereby, the effect on MGO increases by more than that on B-A. For EBO1, carbon residue increases because the aromatic compounds in MGO are converted to aliphatic compounds, which decrease the solubility of heavy compounds, resulting in increased carbon residue. Although the heavy compounds in B-A break down to reduce carbon residue, their impact is mild because the B-A content is low. For EBO3, which comprises 75% B-A, the cracking effects of aliphatic compounds in B-A are higher than that of aromatic compounds because the bonding energy of aliphatic compounds is lower than that of aromatic compounds.

Figure 3 illustrates the change in carbon residue after 24 h of ultrasonic irradiation time, which shows a monotonic trend with the same blending ratio regardless of frequency. The carbon residue in EBO1, which has a high MGO content, remains constant, and the effect of cracking or branching noticeably saturates before 12 h. Although heavy compounds increase over the optimal irradiation time, they can be dissolved in EBO1, which has a high MGO content. Contrarily, in EBO3 with a high B-A content, carbon residue increases as free radicals or resins re-aggregate over the optimal irradiation time.

In contrast, Figure 4 shows the variation of carbon residue in EBO2 containing the same amounts of MGO and B-A. Any dominant phenomenon after 12 h is difficult to explain since the variation in carbon residue is slight depending on the frequencies because EBO2 has properties of both fuel oils. However, the carbon residue at both frequencies increases after 24 h. Presumably, the optimal irradiation time of EBO2 is exceeded in this case.

3.2. Kinematic Viscosity and Density

The kinematic viscosity of a fluid, which refers to its fluidity, is the ratio of its absolute viscosity to its density.

Table 2. Variation in kinematic viscosity and density.

Condition		Kinematic Viscosity (mm2/s at 40 °C)			Density (kg/m3 at 15 °C)
		0 h	12 h	24 h	0 h	12 h	24 h
EBO1	25 kHz	3.87	3.88	3.89	842.5	842.4	842.5
	72 kHz	3.84	3.92	3.93	842.0	842.5	842.6
EBO2	25 kHz	4.17	4.18	4.19	847.6	847.6	847.7
	72 kHz	4.17	4.18	4.19	847.5	847.6	847.7
EBO3	25 kHz	4.51	4.52	4.52	852.7	852.8	852.8
	72 kHz	4.52	4.53	4.54	852.8	852.9	853.0

lists the kinematic viscosity and density of each scenario. The initial values of kinematic viscosity before ultrasonic treatment are 3.87, 4.13, and 4.51 for EBO1, EBO2, and EBO3, respectively. As the content of B-A increases, kinematic viscosity increases because heavy compounds increase with an increase in B-A, thereby increasing viscosity.

The kinematic viscosity and density are nearly unchanged after ultrasonic irradiation for all scenarios in this study, but carbon residue changes, as discussed in Section 3.1. Interestingly, these results are inconsistent with those in the literature cited earlier, and different studies have reported different results. To elaborate, a decrease in aromatic or heavy compounds reportedly decreases viscosity [37]. Allami et al. [24] reported in their study on biodiesel that density decreases after ultrasonic irradiation; however, Duarte et al. [26] reported, based on a UAOD experiment on feedstock and diesel oil, that density is constant regardless of ultrasonic irradiation. Moreover, Najafi and Amani [25] reported that the viscosity of crude oil temporarily decreases and then increases with time. The aforementioned results are different because the properties of sample oil vary from study to study. The changes in viscosity and density are challenging to estimate and to explain because they depend on various parameters, including hydrocarbon composition, irradiation time, and other parameters.

3.3. Pour Point

The pour point of a fluid refers to the temperature at which it loses its fluidity. This parameter is sensitive to viscous resistance at low temperatures. The possibility of oil temperatures decreasing below their pour points during winter must be considered as it can pose problems during oil transportation. Pour point primarily depends on paraffin content (alkanes of C₁₉ or higher), and this content can change in form to a waxy precipitate at low temperatures. This phenomenon commonly occurs in blended fuel oil and is the most problematic characteristic of blended VLSFO [38].

Figure 5 demonstrates the variations in pour point with ultrasonic irradiation time, and the value of y-axis is normalized by the initial pour point. Note that the pour point increases until 12 h of irradiation time under all conditions. Since this converts aromatic compounds into aliphatic compounds, including paraffin, the pour point increases due to the increase in paraffin content after ultrasonic irradiation. After 24 h, the change in pour point is insignificant except for EBO1 at 25 kHz. The cracking process of aromatic compounds into aliphatic compounds is seemingly in equilibrium before 12 h because the carbon residue is either increasing or constant after 24 h, as shown in Figure 3. However, for EBO1 at 25 kHz, the heavy aromatic compounds in B-A convert into aliphatic compounds continuously, but this does not increase the carbon residue because free radicals or resins can dissolve in 75% MGO, resulting in an increase in the pour point of EBO1 at 24 h.

3.4. Sulfur

This study determines whether water in fuel oil could play the role of an oxidizing agent by analyzing the effect of ultrasound on desulfurization in fuel oil without any additives, such as oxidizers and solvents. The fuel oil used onboard contains some water from supply points, and the water content increases with a decrease in fuel quality; furthermore, water content can increase while a fuel oil is stored in a storage tank due to the mixing of condensate. Therefore, a separate cleaning process is required that incorporates a purifier [39]. Water can be thermally decomposed above 2000 K into H₂, H, H₂O₂, etc., as conveyed in Equation (1) [40], and ultrasonic irradiation can decompose water because cavitation generates hot spots of temperatures over 5000 K with 1000 atm of pressure instantaneously [28]. Moreover, H₂O₂ is chiefly used as an oxidizing agent in the reaction of ODS, as expressed in Equation (2).

$$H_{2}O+Heat \to H_2, H, O, O_2, H_2{O}_2, OH$$

(1)

$$R-S-R^{\prime } \overrightarrow{\stackrel{\left[O\right]}{\to }} R-S{O}_2-R^{\prime }\stackrel{heat}{\to }R-R^{\prime }+S{O}_2\uparrow$$

(2)

Although the sulfur content was decreased by a maximum of 25% in the previous study [29], the results from this study are below 0.2% for all EBOs. This is because the water content in EBOs is noticeably insufficient to remove sulfur, and this point differs from the previous study that used bunker-C as a feedstock, which contains a higher water content. An additional experiment was performed with 100% B-A at a frequency of 25 kHz to investigate the effect of ultrasound on desulfurization. The sulfur in 100% B-A decreases by approximately 4% in 12 h at an ultrasonic frequency of 25 kHz, as shown in Figure 6. B-A is assumed to contain a small amount of water because it contains residual fuel oil. A slight increase in sulfur is observed after 24 h compared with that after 12 h because no SO₂ is extracted during ultrasonic irradiation, resulting in the re-dissolving of SO₂ into fuel oil. SO₂ is known to dissolve easily in fuel oil; thus, a separate extraction process is necessary. Conclusively, the desulfurization effect, by using only ultrasound, in this study is not significant when compared to other UAOD studies that reveal a desulfurization effect of up to 99% with additives. However, this study demonstrates the possibility of desulfurization with water in fuel oil using ultrasound, and efficiency can possibly be increased with a higher water content, which will be investigated in future work.

4. Conclusions

This study investigated the effect of blending ratios, ultrasonic irradiation times, and frequency of ultrasonic irradiation on the properties of blended VLSFO. After 12 h of irradiation, the carbon residue in EBO1 decreased by 25% at a frequency of 25 kHz, while that in EBO3 increased by 33%. However, at 72 kHz, the carbon residue in EBO1 increased by 16%, while that of EBO3 decreased by 10%. The effect of frequency depended on the properties of EBOs. As MGO content increased, a low frequency of 25 kHz was demonstrably more effective than a high frequency of 72 kHz, and vice versa. After 24 h, the variation in carbon residue was independent of frequency; for all frequencies, the carbon residue in EBO1 was constant, but that in EBO3 increased. This is because cavitation persisted beyond the optimal irradiation time, which induced aggregation of asphaltenes. Unfortunately, the optimal frequency and irradiation time could not be analyzed due to a lack of adequate sampling.

The changes in kinematic viscosity and density were insignificant for any of the scenarios, but the pour point increased because ultrasound converted aromatic compounds into aliphatic compounds, including paraffin that precipitated as wax at low temperatures. Although the reduction in sulfur for EBOs was negligible, it decreased by approximately 4.5% under the condition of 100% B-A. Therefore, desulfurization was evidently possible by using water in fuel oil, without the addition of any oxidizing agent; however, its effect was smaller than that achieved by using oxidizing agents, such as UAOD.

The utilized flowcell-type ultrasound equipment has the capability to handle large quantities of fuel oil, and thereby, possibly solve the scale-up problem, a limitation of the prevalent probe-type ultrasound unit. However, the effect of ultrasonic irradiation in this study is difficult to generalize because the optimal frequency and irradiation time differ depending on the properties of fuel oil. In future work, a parameter experiment will be conducted with a shorter sampling period of irradiation time over a wide frequency range and for various fuel types, which is necessary to generalize the effect of ultrasound on fuel oils. Moreover, the changes in the quality of fuels after exposure to ultrasound will be determined to investigate the prospects of long-term fuel storage.