1. Introduction
Marine transportation has been the most important mode of transportation in international trade characterized by good economy and emission per unitized cargo. It is regulated or recommended by most of the coastal countries to switch highway and railway transports to shipping transport for economic and environmental sakes. As a result, shipping will share part of the emission burden of on-road traffic [
1,
2]. This study aims to tackle the challenge of reducing emissions from marine transportation, specifically focusing on particulate matter (PM) and polycyclic aromatic hydrocarbons (PAHs) emitted from emulsified heavy fuel oil (EHFO) used in marine low-speed main engines.
Oxides of nitrogen (NOx), along with oxides of sulfur (SOx) and particulate matter, constitute the primary pollutants in ship exhaust. According to [
3], the sulphur content in fuel oil used on ships operating outside designated emission control areas is limited to 0.50% m/m (mass by mass), a significant reduction from the previous limit of 3.5%. The impact of these substances on the environment and human health cannot be ignored. The MARPOL convention sets requirements for nitrogen oxide (NOx) emissions. The NOx emission limits are to be implemented in three phases. The upcoming ‘Tier III’ emission standards, which are about to take effect, mandate a reduction of approximately 70% in emission limits compared to Tier II [
4]. The IMO’s Tier III NOx emission standards are mandatory for all newly constructed ships operating in designated emission control areas (ECAs) after 1 January 2016. Outside of ECAs, known as non-ECA zones, ships are required to comply with the Tier I or Tier II standards, depending on the engine type and the year of construction. Depending on the catalyst utilized, SCR techniques encompass NH
3-SCR, CO-SCR, HC-SCR, and H
2-SCR and are a focal point of current discussions [
5,
6,
7]. However, SCR systems necessitate precise operational conditions and are prone to challenges stemming from catalysts and reducing agent leakage. Additionally, the practice of slow steaming, aimed at increasing fuel efficiency, is not compatible with many SCR installations, as the lower temperatures in the exhaust gas are not conducive to SCR catalyst performance.
In addition to onboard control, there are multiple approaches to consider, such as replacing it with advanced diesel engines or changing the type of fuel used. There are numerous varieties of fuels available as alternatives to deal with the ship stack emission problems. Fuel blends obtained by adding water, microalgae biodiesel, methanol or butanol are proved to be efficient for emission reductions [
8,
9,
10]. Water-emulsified fuels are practical and economical for NOx reduction compared to expensive large-scale denitrification facilities, such as SCR [
11]. Studies on the onboard commercialization of the emulsified fuel are considered in accordance with the IMO’s (International Maritime Organization) increasing focus on NOx reduction [
12,
13,
14].
On the other hand, modifying fuel characteristics can result in emission variations for various contaminants as well. For instance, compared to the neat fuel, burning an acetone–butanol–ethanol diesel blend can reduce PM emissions by 17.8–50.4% on a four-stroke single-cylinder engine [
15], and burning waste cooking oil biodiesel can reduce CO and PM emissions by, respectively, 7–25% and 18–48% on a 4-cyclinder diesel engine [
16].
Moreover, polycyclic aromatic hydrocarbons (PAHs) and their derivatives are a series of harmful compounds in ship stack exhausts [
17]. They are of great concern due to their mutagenicity and carcinogenicity [
18]. Our previous study has observed that burning the blends of biodiesel and marine gas oil on ship auxiliary engines nevertheless can reduce the bulk emission of PM and PAHs, concentrate the particulate PAHs, and enhance the toxicity of the particle matters [
19]. Similar results are reported in the tests on waste edible oil [
20], animal fat [
21], Fischer–Tropsch [
22] and sugarcane biodiesels [
23] as well. Biomass fuel combustion generates PAH components mainly composed of naphthalene, accounting for almost half of the total, which is also the primary component of polycyclic aromatic hydrocarbons in mineral oil combustion [
24,
25]. In addition, the components of PAHs were also detected in the tail water of the desulfurization process [
15]. Phenanthrene is one of the primary compounds of PAHs generated from the combustion of fossil fuels [
26,
27,
28], it is necessary to make comprehensive assessments on the emissions before the onboard application of emulsified fuel.
Furthermore, the concept of environmentally sustainable and eco-friendly maritime vessels has garnered considerable recognition, prompting a burgeoning body of research in this domain. However, an analysis of the 2020 International Maritime Organization’s (IMO) Data Collection System (DCS) statistics reveals that conventional heavy and light fuel oils continue to dominate marine energy consumption. This underscores the reality that, to date, alternative, cleaner energy sources have yet to supplant traditional fuels within the maritime industry [
29].
The aim of this study is to reveal the effects of water emulsification with heavy fuel oil (HFO) on the emissions of PM and PAH species. Tests will be carried out on a bench with a slow-speed two-stroke marine engine. Both the engine and fuel are the most widely adopted by in-use ship propulsion systems. The PM and nitrated, oxygenated and parent PAHs will be particle size, segregated and monitored to figure out the effect mechanism preliminarily and provide data references for emulsified HFO emissions. Employing this methodology enables the precise monitoring and categorization of particulate matter (PM) and polycyclic aromatic hydrocarbons (PAHs) based on particle size. Such an approach yields initial understandings of the underlying mechanisms and contributes to a comprehensive dataset that is instrumental in shaping forthcoming regulations and driving innovation in the realm of marine emission mitigation.
2. Experimental Setup and Methods
2.1. Experimental Setup
The experiments were carried out on a 6-cylinder two-stroke diesel engine with a low speed of 142 rpm in the Engine Performance Test Lab of Shanghai Maritime University. The main specifications of the test diesel engine are listed in
Table 1. The engine was designed to be representative of a typical low-speed marine engine with burning HFO, and the injection nozzle was not modified when burning EHFO to simulate the fuel changing onboard.
The experiments were conducted with commercial heavy fuel oil (HFO), which is a representative fuel for offshore marine main engines, and prepared emulsified heavy fuel oil (EHFO), which contained water at a rate of 5%vol and was verified to be able to decrease the emission of NO
x in a test carried on a high-speed diesel engine. We believe that it may be possible to produce EHFO onboard by substituting the purification process with clarification. This could be a significant advantage, as it would simplify the fuel preparation process and potentially reduce costs.
Table 2 summarized the full parameters of the test fuels.
The engine loads ranged from idle to 75% and the test operation modes were 25%, 50% and 75% (the most common load point). To obtain reliable samples, the engine was operated for a few minutes trying to ensure that lubricating oil temperature, cooling temperature and exhaust gas temperature were in a steady state when sampling. In addition, the fuel consumption of the test fuels was reported in
Table 3, showing that the engine operated relatively stable during the samplings.
2.2. Sampling Methods and Exhaust Measurements
The initial temperature of exhaust gas was ~350 °C at the sampling point on the stack 20 m downstream the engine exhaust outlet. The sampled exhaust gas was diluted with ambient air and then led the temperature fall below 50 °C for diluted gas. Particulate samples were collected from the diluted gas using an Anderson 8-stage sampler (Thermo-Andersen 20–800, United States) loaded with quartz fiber filters (QFF, 8.1 cm diameter, Munktell MK360, Sweden). The cutoff aerodynamic diameters for each stage were <0.43, 0.43–0.65, 0.65–1.1, 1.1–2.1, 2.1–3.3, 3.3–4.7, 4.7–5.8, 5.8–9.0 and >9.0 μm. Gaseous samples were collected by drawing air (from the outlet of the Anderson 8-stage sampler) through a chamber housing a layer of polyurethane foam plugs (PUF, 6.3 cm diameter and 4.0 cm thick, Tisch Environmental Inc., Maumee, OH, USA). The sampling time was 5 min, and the airflow rate was 28.3 L min−1 for each sampling.
Before sampling, the quartz fiber filter was baked in a muffle furnace for 2 h at 600 °C to remove background pollutants, and we recorded their own weight as initial weight [
30]. The PUFs was pre-processed and cleaned with Soxhlet extraction with dichloromethane. Fuel consumptions were continuously measured during samplings and listed in
Table 3.
2.3. Chemical Analysis Methods
The particulate samples on the quartz filter membranes were wrapped with tinfoil papers and dried in a desiccator for two days until constant weight. The difference of the constant weight and initial weight of membranes was the weight of PM. Then, each quartz filter membrane (containing particulate PAHs) was put in a glass sample bottle, spiked with 5 recovery surrogates of PAHs (naphthalene-d8, acenaphthene-d10, phenanthrene-d10, chrysened12 and perylene-d12) and 5 recovery surrogates of nitrated PAHs (nPAHs) and oxygenated PAHs (oPAHs) (nitronaphthalene-d7, nitroanthracene-d9, nitropyrene-d9, nitrochrysene-d11, anthraquinone-d8), then extracted three times using 20 mL of dichloromethane for 20 min in an ultrasonic bath. PUFs also needed to be wrapped in tinfoil papers and dried for a few days to remove the water. Then, the PUFs were spiked with the same recovery surrogates and extracted with Soxhlet extraction with 80 mL dichloromethane for 24 h. The extracts of PUFs were reduced to 2 mL under a gentle stream of nitrogen and then cleaned up using dispersive solid phase extraction. Finally, the extracted samples were spiked with a known amount of volumetric internal standards (hexamethylbenzene) and sent to instrumental analysis.
PAHs were determined using an Agilent 7890-5975C gas chromatograph coupled with a mass selective detector (GC/MS) equipped with a DB-5 ms capillary column (15 m, 0.25 mm film thickness, Agilent, Santa Clara, CA, USA). Helium was used as the carrier gas at a constant flow of 2.0 mL min−1. The injection volume was 2.0 μL. The temperature of the injector, detector, transfer line and EI source was 280, 270, 280 and 230 °C, respectively. The column oven temperature program was 4 °C min−1 from 60 to 300 °C. Identification of each individual PAH was based on its retention time and the specific primary ion fragment m/z of a corresponding authentic standard.
Negative chemical ionization (NCI) using methane ionization gas (40 mL min
−1) and a source temperature of 200 °C was employed for analysing these nPAHs and oPAHs. The detected compounds have been listed in
Table 4 and
Table 5, along with their retention time (RT) and characteristic ion. The oven temperature program analysis was 40 °C (held 1.7 min), ramped to 150 °C at 20 °C min
−1 and held for 10 min, ramped to 220 °C at 10 °C min
−1 and held for 10 min and finally ramped to 310 °C and held for 15 min. Abundances three times higher than the base peak were detected.
A spiked blank and a reagent blank sample were prepared to assess the recovery of the analytical method. Recoveries of surrogate compounds ranged between 75.2% and 113%. Analyzed reagent blank samples were found to contain undetectable amounts of any interference. The limit of quantification (LOQ) was determined from the standard deviation by analyzing ten blank samples. LOQ was estimated as the mean blank value plus ten times the standard deviation. The LOQs of PAHs in the exhaust and fuels were 0.01 μg/Nm3 and 0.01 μg mL−1 for exhausts and fuels, respectively. The final sample concentrations were surrogate recovery corrected.
Same instrumental analysis and QA/QC processes were used in our previous study [
31]. In this test, recoveries of surrogate compounds ranged between 75.2% and 113%. The final sample concentrations were surrogate recovery corrected.
2.4. Calculation of Emission Factors
The exhaust flow rates for the engine were calculated using the carbon balance method specified in ISO 8178-2 [
32], assuming complete conversion of fuel carbon to CO
2. Based on the fuel consumptions (
Table 3) and pollutant concentrations, the emission factors (EFs) of PM and PAHs were determined using similar methodology in our previous study [
33,
34].
Emission data are reported as power-based emission factors, which can be calculated from the measured pollutant concentrations using the following equations:
where
VE is the emission volume,
VS is the sampling volume and
M is the molecular weight of carbon,
FC is the fuel consumption,
R is the gas constant and Δ
m is the difference of filter mass before and after sampling in the Anderson eight-stage cascade impactor (TE-10-800, TISCH, Maumee, OH, USA). The constant air sampling flow rate is 28.31 L min
−1, pumping for 5 min under each operating mode, and the sampling volume is 142 L.
4. Conclusions
Water emulsification plays a role in altering the emission of PM, PAHs and their derivatives. Burning EHFO could reduce the emissions of PM and LMW-PAHs in the gas phase, with a respective maximum reduction of 13.9% and 40.5% at low operation mode. Nevertheless, burning EHFO could increase the emission of HMW-PAHs in fine particles and pose a consequently higher carcinogenic risk to ecology. Burning EHFO could reduce the nPAH emissions while increasing the oPAH emissions. In addition, the compositions of the nine detected species of PAH derivatives have been altered dramatically, which could be explained by the addition of OH radicals by burning emulsified fuel and the difference in the molecular structure of the derivatives. These findings are valuable for assessing the advantages and disadvantages of the onboard application EHFO.
This study differs from the previous literature on denitrification and fuel-saving methods, providing significant reference values for a comprehensive assessment of emulsified oil and the design of exhaust gas treatment devices, such as the PAH component in the tail water of desulfurization towers. Furthermore, this study presents an exhaustive evaluation of emulsified heavy oil technology, establishing a foundational dataset for comparative analyses in subsequent investigations concerning alternative fuel mixtures and biodiesel applications.