An investigation into the relationship between the formation of thermal ...

1 downloads 0 Views 2MB Size Report
*Department of Mechanical Engineering, National Institute of Technology, ... Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, ...
Bulletin of the JSME

Vol.10, No.2, 2015

Journal of Thermal Science and Technology An investigation into the relationship between the formation of thermal cracked components and PM reduction during diesel combustion using water emulsified fuel Hirofumi NOGE*, Yoshiyuki KIDOGUCHI**, Wira JAZAIR YAHYA ***, Yoko IMAI**** and Kazuo TAJIMA**** *Department of Mechanical Engineering, National Institute of Technology, Maizuru College 234 Aza Shiroya, Maizuru, Kyoto 625-8511, Japan E-mail: [email protected] **Department of Energy System, The University of Tokushima 2-1 Minami-josanjima, Tokushima 770-8506, Japan *** Department of Mechanical Precision Engineering, Malaysia-Japan International Institute of Technology, Universiti Teknologi Malaysia Jalan Semarak 54100 Kuala Lumpur, Malaysia **** Project of Three-Phase Emulsion Technology, Kanagawa University 3-27-1, Rokkakubashi, Kanagawa-ku, Yokohama 221-8686, Japan

Received 17 May 2015 Abstract Water-in-diesel emulsion fuel (W/O) operated in diesel engines, shows a significant reduction of particulate matter (PM). In this paper, PM reduction characteristics by thermal decomposition of W/O10 and W/O20 (10vol.% and 20vol.% of water in W/O respectively) are identified in diesel combustion atmosphere using a plug flow reactor with a co-flow diffusion burner. To analyze initial thermal decomposition at diesel diffusion combustion, the W/O fuels are thermally decomposed in the plug flow reactor first, then the thermally decomposed W/O fuels are introduced into a co-flow diffusion burner as fuel and PM are generated. In high temperature atmosphere without oxygen in the reactor, W/O10 and W/O20 are thermally decomposed and both of them almost produce light hydrocarbons (LHCs) higher than a diesel fuel, which means thermal decomposition before combustion are encouraged by the W/O. Excitation–emission matrix (EEM) method shows that polycyclic aromatic hydrocarbons (PAHs) are produced by both W/O fuels and diesel fuel during the thermal decomposition period but some W/O fuels oxidize a huge amount of PAHs in the later diffusion combustion. CO, CO2 measurements after the combustion of the thermal decomposed substances in the diffusion burner via high temperature reactor reveal that diffusion combustion of W/O fuels contribute to Soluble Organic Fraction (SOF) and Solid reduction which leads to reduction of CO and increase of CO 2 respectively. Key words : Diesel combustion, Water-in-diesel emulsion, PM reduction, Thermal decomposition, Flow reactor, Co-flow diffusion burner

1. Introduction Water-in-diesel emulsion fuel (W/O) is well known for reducing NOx and PM (Particulate Matter) simultaneously from diesel combustion (Henningsen, 1994). Stability of W/O have shunned a practical application in diesel engines. However, to meet a requirement restriction, application of alternative fuels such as W/O, biodiesel fuel, GTL (Gas to liquids), DME (Dimethyl ether) are needed without a major modification of diesel engines. NOx is going to be dropped by 30 to 50% with the lower W/O fuel injection pressure at low and middle load and also maximum PM reduction rate: 94% is obtained at low load compare to diesel fuel (Maiboom and Tauzia, 2011). There are many experimental studies that reveal a NOx and PM reduction rate of diesel combustion using W/O (Bertola and Boulouchos, 2003, Subramanian Paper No.15-00280 [DOI: 10.1299/jtst.2015jtst0024]

0123456789

© 2015 The Japan Society of Mechanical Engineers

1

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

and Ramesh, 2001, Liang et al., 2011). Generally, the percentage of smoke reduction is higher than that of NOx and smoke reduction using W/O is approximately twice as large as diesel fuel (Treea and Svensson, 2007). NOx reduction of W/O is due to a reduction of combustion temperature because the extended Zeldovich mechanism that is the dominant mechanism of NO formation depends on the temperature. Physical and chemical causes on NOx have already been clear, which can take any measurements for NOx reduction definitely. Regarding PM reduction, W/O produces longer ignition delay periods and a high initial premixed burning rate, which enhance oxidation of in-cylinder soot. Micro-explosion is characteristic phenomena of W/O. The phenomena have been experimentally reported in single droplet studies (Tsa and Wang, 1986, Law et al., 1980, Tsue et al., 1996, Watanabe et al., 2009) but micro explosions in diesel sprays under diesel operating conditions have not been reported. Water gas reaction and water gas shift reaction are considered one of causes for PM reduction but these reactions during diesel combustion have not been indicated quantitatively. The change of hydrocarbons into solid soot commonly identified following process: pyrolysis (thermal decomposition), nucleation, coalescence, surface growth and agglomeration. Thermal decomposition is not only primary reaction in combustion but also soot formation (Smith, 1981). All hydrocarbon fuels undergo thermal decomposition and produce polycyclic aromatic hydrocarbons (PAHs) as soot precursors in addition to light weight hydrocarbons. In an actual diesel combustion, soot precursor is formed as a result of both pure fuel thermal decomposition and incomplete combustion. The characteristics of injection duration or its combination and mixing ratio at the early stage of diesel combustion are well known to be important factors for controlling exhaust gas as well as combustion. On that account, oxidation of soot precursor at early combustion stage probably plays an important role in reducing PM. Liu et al. (2014) revealed numerical results that water vapor addition to the oxidizer stream significantly reduces flame temperature and soot loading. The primary pathways for the chemical effect of water vapor are OH + H 2 ↔ H + H2O and OH + OH ↔ O + H2O. The chemical effect of water vapor reduces propargyl, benzene, and pyrene in the early stage of soot formation, which means the chemical effect of water vapor affects soot nucleation more significantly than surface growth. Roberts et al. (2005) also show experimentally and theoretically that primary water, which is an additive in a premixed flame, is soon decomposed into OH, H 2, O and H within the flame and is available for reaction with observed unburned hydrocarbons. Authors studied PM reduction by 10vol.% Methyl decanoate [CH3(CH2)8COOCH3] a mixed fuel. The study revealed that oxygenated fuel can reduce PM in the fuel rich region but detailed PM reduction reaction was uncertain. The report showed that C 2H2 reduction during thermal decomposition makes it possible to reduce PM (Noge et al, 2011). These results are not obtained by using W/O but water is effective for the early stage of soot formation. As thermal decomposition reactions occur at the beginning of diesel combustion, W/O also is to be involved with the thermal decomposition as well as the oxygenated fuel. However seldom studies are reported regarding effect of thermal decomposition of W/O among thermal decomposition, combustion and PM generation. In this study, the effect of W/O on PM concentration will be discussed by analysis of the thermally decomposed products, which are collected at outlet of the flow reactor, and exhaust gas showed in Result and Discussion.

2. Experimental Method The experimental setup shown in Fig.1 consists of a plug flow reactor with infrared reflex furnace and a co-flow diffusion burner. Immediately after fuel was thermally decomposed in the flow reactor at atmospheric pressure, the fuel was fed to the co-flow diffusion burner directly. Thermally decomposed fuels were burned in the co-flow diffusion burner and PM was produced. All experiments were carried out at atmospheric pressure because most of all chemical reactions in diesel combustion depend on temperature. In addition, Cyril et al. (2003) concluded that in-cylinder pressure did not markedly affect the amount of soot produced during combustion by laser-induced incandescence study using rapid compression machine. In this experiment, three kinds of fuel were examined. One was Solvent (normal saturated paraffin, C12H26: 27wt.%, C13H28: 47wt.%, C14H30: 26wt.%) which simulates the main components of diesel fuel, and the others were Solvent-water emulsion fuels constituted by water in Solvent oil about 10 vol.% (W/O10) and 20 vol.% (W/O20) that were produced by three-phase emulsification method (Tajima, 2001, 2002). Plug flow reactor Fuel was first injected by a fuel injector into an evaporator kept at 330℃. The fuel soon vaporized and formed a

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

2

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

Cooled

Exhaust

Exhaust

P GC GC/MS

Filter

Filter

NDIR(CO,CO 2 )

P

C,H,N corder

Exhaust

Fluorescence Spectrophotometer

Fuel injector O 2 /N 2

Carrier gas

T.C.

Heater

Heater

Temp.

T.C. controller

T.C.

Infrared reflex furnace

GC

,

GC/MS Exhaust

P

Filter

Fig.1 Schematic diagram of experimental setup. W/O fuel is introduced into the evaporator with N2 gas and the mixture is thermally decomposed in the high temperature reactor. The thermal cracked fuel mixture is burned by supplying oxidizer at co-flow diffusion burner. PM is collected at burner outlet.

mixture with nitrogen. Then the mixture was introduced into the infrared gold image furnace (ULVAC RHL-E410) where the temperatures were controlled by temperature program controller (ULVAC TPC-1000) with K type thermocouple (OKAZAKI HOSKINS 2300) which measures the mixture temperature directory in the reactor tube. Residence time (tr) in the reactor was a relatively long: tr=40ms to decompose high molecular liquefied fuel in nitrogen atmosphere and controlled furnace temperatures (Tr) were set at 1350K and 1500K to simulate low and high load conditions respectively in a diesel engine. Examined fuel/carrier gas ratio F/G in the flow reactor was set at F/G=0.07, 0.10 and 0.13. Each F/G corresponds to equivalence ratio φb in the combustion vessel atφb =1.0, 1.5, and 2.0 respectively. These experimental conditions are listed in Table 1. A high temperature gas sampler (Yanaco GHS-50) was used to inject thermally decomposed fuels into a gas chromatograph (GC:Yanaco G3800) equipped with FID and TCD. Light hydrocarbons among the decomposed components were analyzed by Squalane 2% packed column (length=3m, 4mm i.d.). Thermally decomposed hydrocarbons were also dissolved in dichloromethane (DCM) at outlet of a plug flow reactor by bubbling. Three-dimensional fluorescence spectra of polycyclic aromatic hydrocarbons (PAHs) in the DCM were scanned by a fluorescence spectrophotometer (Hitachi F-7000) from 200 to 750 nm in the excitation wavelengths and the fluorescence emission was captured at same wavelengths. Each slit of excitation and emission was set at 5 nm, excitation and emission data sampling steps were 50 nm and 10nm respectively. The scanning speed was 1200 nm/min. Co-flow diffusion burner In an atmospheric pressure, axial co-flow diffusion flame was generated using the burner in which fuel and nitrogen mixture were flowed from a centerline tube (22.65mm i.d.). Oxygen and nitrogen were supplied through the outer tube (38mm i.d.) as oxidizer. The mixture velocity and oxidizer velocity were same at outlet of the burner. A cylindrical quartz tube surrounded the burner outlet to prevent excess air inflow. Gravimetric analysis was applied to Table.1 Experimental conditions Fuel Type

Reactor Temp.

F/G (Fuel/Gas) corresponds to Φ b

1350K and 1500K

F/G(Φ b :global equivalence ratio) 0.07(Φ b =1.0), 0.10( Φ b =1.5), 0.13( Φ b =2.0)

Solvent

W/O10 Solvent + Water 10 vol.%

W/O20 Solvent + Water 20 vol.%

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

23

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

measurement of PM that was collected on a ploytetrafluoroethylene (Teflone) filter (MILLIPORE FHLP04700 diameter=47mm, hole=0.45μm) at 4l/min with a diaphragm pump. The Teflone filter was dried at least 12 hours before and after sampling. This study classified PM into Solid and SOF (soluble organic fractions) by measuring filter weight. SOF was extracted from PM on the Teflon filter using the soxhlet extractions for 3 hours. PAHs included in SOF were analyzed by fluorescence spectrophotometer (Hitachi F-7000) mentioned above.

3. Results and Discussion 3.1 PM formation after combustion In this paper, at first PM generated by co-flow diffusion burner was measured. Then to identify the relation among PM, thermally decomposed hydrocarbons produced in the flow reactor and hydrocarbons emitted into the atmosphere through co-flow diffusion burner are compared. Figure 2 shows PM produced by a co-flow diffusion burner combustion by burning the thermally decomposed fuels which are generated in the flow reactor. PM is separated into SOF and Solid to identify main components. As many PM are generated at fuel rich condition : φb=2.0 is selected to be discussed in particular. At the flow reactor temperature : Tr=1350K, W/O10 and W/O20 increase SOF, however they produce less Solid in comparison with Solvent. At low temperature W/O fuels cause Solid reduction but SOF increase. For the reactor temperature : Tr=1500K, PM concentrations are relatively increased at any fuels because many soot precursors such as C 2H2, PAHs and unsaturated hydrocarbons are produced before combustion (Smith, 1981, Noge et al., 2011). However, PM are diminished remarkably by using W/O fuels in comparison with Solvent. PM are reduced proportioning with content’s percentages of water and the percentage reduction of PM are higher than the content’s percentages of water. In a real diesel engine, both 1350K and 1500K will be classified in high temperature. In this experiment, the fuels are thermally decomposed in nitrogen(N2) atmosphere without oxygen(O2) in the flow reactor, which remains original fuel components such as C12, C13, C14 partially and slows the fuel decomposition. The co-flow diffusion flames of Tr=1500K emit yellow color, on the other hand, flames color of Tr=1350K show blue and yellow. The flame colors suggest that the difference of thermal decomposition temperature affect the flame temperature and eventually PM concentration. In a real DI diesel combustion, diesel alternative fuel such as Bio diesel fuel (BDF) and water emulsion fuel increase SOF at low load (Jazair Yahya el at., 2007, Murotani et al., 2007). There is a different result that both Solid and SOF are reduced at all loads in a IDI diesel combustion (Armasa et al., 2005) and the reasons regarding the reduction are explained as effect of micro explosion and OH increase but these phenomena have not been observed directly during diesel combustion. Assuming that temperature of Tr=1350K in this experiment is low considering DI diesel combustion, the ratio of SOF in PM using W/O fuels tend to be higher than that of Solvent.

600

Fuel: Solvent

Fuel: W/O 10

Fuel: W/O 20

Tr=1350K

Tr=1350K

Tr=1350K

Tr=1500K

Tr=1500K

Tr=1500K

400

600

3

(mg/m )

800

PM

200 0

Solid Solid

400

Solid

200

SOF

SOF SOF

0

1.0

1.5

2.0

1.0

1.5 φ b

2.0

1.0

1.5

2.0

Fig.2 PM generated by diffusion combustion of thermal decomposition substances originated from Solvent, W/O10 and W/O20.

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

24

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

Latent heat of evaporation by water may greatly affect decrease of the temperature in combustion, which increases SOF at low temperature condition. Jazair Yahya et al. (2007) observed CO increases at low load and decreases at high load by both waste cooling oil and BDF. There will be a relationship among reaction temperature, SOF and CO in W/O fuel too because an additive of water injection into combustion chamber decreases combustion temperature and increases CO were reported by Moritani et al (2007). Therefore low combustion temperature at low load causes imperfect combustion which will be one of the factors of the SOF increases. The PM reduction by W/O fuel in diesel combustion has many reasons such as increase of ignition delay period and premixed combustion duration (Musculus et al., 2002, Subramanian and Ramesh, 2001, Kee et al., 2003), decrease in the temperature of spray (Liu et al., 2014), the micro-explosion of water which have not been confirmed experimentally in diesel combustion (Sheng et al., 1995, Lif et al., 2007, Huoa et al., 2007) and water gas reaction (Hermann and Huttinger, 1986) , water gas shift reaction (Neeft et al., 1997). PM reduction mechanisms by W/O fuel include both physical and chemical effects in diesel combustion. The essentials of PM reduction mechanism from mentioned above are fuel and oxidizer mixing, combustion temperature control and chemical reaction. The results in Fig.2 are almost consistent with the production tendency of SOF and Solid from a real diesel engine although this experiment doesn’t deal with the main factors for deciding diesel combustion such as injection pressure, air flow in cylinder and ignition delay. Formation of fuel and oxidizer mixture is very important physical method but the outcome of the mixture is decided by temperature and in the mixture, huge chemical reactions are progressing. This experiment examines the relation between chemical reactions controlled by temperature and PM reduction and production without physical effects. The chemical nature of PM reduction and formation is not enough understood. In addition, our previous work concluded that thermal decomposition at first stage of combustion process has a great effect on the combustion at later stage (Noge et al., 2007). There are also reports (D’Anna and Violi, 2005, D’Anna et al., 2001) that water in premixed flame read to soot reduction because water produces hydroxyl radical which may affect unburned hydrocarbons and benzene that are considered as soot precursor. Hydroxyl radical and benzene concentration’s peak are observed in a headstream of a premixed laminar flame where thermal decomposition and initial combustion are probably activated. These intermediates have a property of undergoing chemical reactions easily, so that they interact with another radicals and unburned hydrocarbons at the end of thermal decomposition and initial combustion period. In conventional diesel combustion, diffusion combustion is main combustion period, thermal decomposition before combustion will play an important role and significant to be understood.

3.2 Analysis of Light hydrocarbons at thermal decomposition PM produced through co-low diffusion burner is shown in Figure 2. The co-low diffusion burner consumes thermally decomposed fuel generated into the flow reactor. The properties of PM will depend on the thermally decomposed fuel. Figure 3 shows thermally decomposed fuels : light hydrocarbons (LHCs) produced in the flow reactor over three different fuel carrier gas ratios. LHCs : CH4,C2H6,C2H4,C3H6,C3H8 and C2H2 are produced by each fuel. The kinds and the reaction mass of LHCs show almost same among the fuels. At Tr=1350K, C2H4 is the most dominant species and the other LHCs are produced in following order: CH4>C3H6>C2H2>C2H6>C3H8. Similarly, a lot of C2H4 is produced at Tr=1500K because of β-scission process, which forms free radical and double bond by thermal cracking of hydrocarbons, and remarkable generation of C2H2 and CH4 stands out against Tr=1350K. It is also characteristic that the concentration of CH 4, which is saturated hydrocarbon and stable species, increases while a lot of unsaturated hydrocarbons are generated. Lea-Langton et al. (2013) concluded that the relatively low molecule PAHs product profiles are remarkably similar for BDF, petroleum diesel and GTL fuel from a flow cell reactor study and they pointed out a commonality of mechanism after the initial decomposition of the parent hydrocarbon. On the other hand, a similar experiment using Solvent and BDF mixture fuel carried out in previous study (Noge et al., 2011) revealed that there was a remarkable difference in C2H2 production between BDF and Solvent in thermal decomposition period but the difference of C 2H2 among Solvent, W/O10 and W/O20 cannot be confirmed in this result. As a sequel to thermal decomposition, the calorific values between 1350K and 1500K are different but the thermally decomposed gas is introduced into co-flow diffusion burner through piping controlled at 573K. The temperature at outlet of the burner is approximately constant, therefore, thermally decomposed gas composition will be important factors to decide the flame temperature and PM concentration

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

25

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

8.0

Fuel:Solvent

Fuel:W/O 10 C2H4

1350K

6.0

LHCs vol.%

4.0 2.0

CH4 C3H6 C2H2

C2H6,C3H8

Fuel:W/O 20

1350K

1350K

C2H4

C2H2,C3H6 C2H6,C3H8

C2H4

CH4 C2H6,C3H8

CH4 C2H2,C3H6

0 1500K

1500K

1500K C2H4

C2H4

6.0 4.0

CH C2H2 4

2.0

CH4

0.10

CH4

0.13 0.07

0.10

C2H2

C2H6,C3H8C3H6

C2H6,C3H8 C3H6

C2H6,C3H8 C3H6

0 0.07

C2H2

C2H4

0.13 0.07

0.10

0.13

F/G

Fig.3 Comparison of Solvent with W/O10 and W/O20 on LHCs produced during thermal decomposition in a plug flow reactor.

in this experiment. Since W/O10 and W/O20 contain some percentage of water, hydrocarbon content or Solvent per a volume is decreased. Thus the ratio of total LHCs production per hydrocarbons content in original fuels was calculated. Calculated formula is shown below.



(LHC wt.%  ρ i ) Total LHCs  Hydrocarbo ns Solvent wt.%  ρ sol

(1)

ρi , ρsol : density[kg / m3 ]

Figure 4 shows the production ratio of total LHCs to hydrocarbons in original fuel. At Tr=1350K, F/G=0.07 and 0.10, the LHCs production ratio of W/O10 is higher than Solvent. W/O10 activates thermal decomposition reaction compared with Solvent except at F/G=0.13 in which total amount of LHCs in W/O10 is less than Solvent. On the other hand, W/O20 produces higher LHCs than W/O10 and Solvent at any F/G. At Tr=1500K, W/O10 and W/O20 keep LHCs production ratio larger than Solvent. In addition, the content of water increases the LHCs production ratio in this experiment. PM relatively increases with rising F/G or an equivalent ratio where lots of fuel is used. W/O10 and W/O20 show higher total LHCs/Solvent than Solvent at same F/G, however, W/O fuels can reduce PM in comparison with Solvent. Thus, PM seems to be reduced as the W/O promotes thermal decomposition.

Totala LHCs Each solvent in the test fuels

0.3 Tr=1500K

Tr=1350K W/O20

0.2

W/O10

W/O10 Solvent

W/O20

Solvent 0.1

0 0.07

0.10

F/G

0.13 0.07

0.10

0.13

F/G

Fig.4 Total LHCs production per total Solvent included in the base fuel

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

26

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

3.3 Analysis of Light hydrocarbons at thermal decomposition

Excitation wavelength (nm)

As previous experiments have reported that PAHs are produced in this thermal decomposition atmosphere (Noge et al., 2007, 2011), PAHs were analyzed by fluorescence spectrophotometer (Fig.5(a) and Fig. 5(b)). In general, PAHs formed in combustion have been measured by GC-MS or laser-induced fluorescence (LIF) method but in this experiment, possibility of PAHs measurement by using fluorescence spectrophotometer (FL) is attempted. Analysis of PAHs using FL is impossible to identify PAHs in detail but rough information, which is wavelength of PAHs, is available because almost all PAHs emit light with intrinsic excitation or absorption pattern in the UV and blue spectral regions (D’Anna and Violi, 2005, D’Anna et al., 2001). In addition, fluorescence results are obtained immediately without any pretreatment. Currently, fluorescence excitation–emission matrix (EEM), which involves a collection of sequential fluorescence emission spectra by increasing excitation wavelengths, is used to characterize dissolved PAHs in organic solvent. Although there remains the EEM complex fluorescence signals, we try to apply EEM technique to the investigation of a generation tendency of PAHs. Figure 5 shows fluorescence spectra of PAHs sampled at the flow reactor outlet. A large amount of PAHs are produced atφb=2.0 (F/G=0.13) fuel rich condition in Tr=1350K and Tr=1500K. According to contour graph of PAHs, a high concentration area corresponds to strong fluorescence intensity whose level is scaled by bar plot beside the figure. In the case of Solvent in Fig.5(a), high concentration area is represented by 7. This numeral means that there are 7 excitation plots satisfy the range from 7500 to 10000 fluorescence intensity. At Tr=1350K, there is little difference between W/O10 and Solvent, while W/O20 produces PAHs higher than Solvent. Both W/O10 and Solvent have a strong fluorescence intensity within a emission region from 300nm to 370nm and the excitation wavelength is about 300nm where fluoranthene, naphthalene and phenanthrene exist according to the results of Beltran et al (Dabestani and Ivanov, 1999) and Aizawa et al (2008). W/O20 slightly expands the strong fluorescence intensity until 450 nm by the same excitation wavelength. The expansion of the fluorescence emission area suggests an increase of PAHs. In this case, the addition of chrysene, anthracene and pyrene are predicted (JiJi and Booksh, 2000, Beltrán et al., 1998). W/O20 causes extreme thermal decomposition among these fuels so that more PAHs via LHCs and other thermal decomposition substances might be produced. This is because total LHCs production ratio per hydrocarbons content in original fuels is higher than the other fuels discussed in Fig.4. Active thermal decomposition probably introduces lots of PAHs. The fluorescence intensity of PAHs shown in Fig.5(b) were sampled at Tr=1500K. The excitation wavelength is from 280nm to 450nm and the fluorescence wavelength is emitted from 400nm to 500nm by Solvent, where anthracene, fluoranthene and perylene might be included at relatively high levels (JiJi and Booksh, 2000, Beltrán et al., 1998). On the other hand, the fluorescence area of W/O fuels are shown from 320nm to 500nm.

750

10000

600 5

450

Fuel:W/O20 Tr=1350K F/G:0.13

Fuel:W/O10 Tr=1350K F/G:0.13

Fuel:Solvent Tr=1350K F/G:0.13

2 7

10

5000

14

300 200

7500

18

6 2 6

2500

300

400

500

600

700 300

400

500

600

700 300

400

500

600

700

0

Fluorescence wavelength (nm)

Excitation wavelength (nm)

Fig.5(a) Three-dimensional fluorescence spectra of PAHs produced in thermal decomposition stage at Tr=1350K, F/G=0.13.

750

10000

600

20

450

Fuel:W/O10 Tr=1500K F/G:0.13

20

16

300

400

500

600

Fuel:W/O20 Tr=1500K F/G:0.13

22

30

700 300

400

500

600

23

700 300

400

500

7500

5000

13

12

30

300 200

Fuel:Solvent Tr=1500K F/G:0.13

600

2500

700

0

Fluoresence wavelength (nm)

Fig.5(b) Three-dimensional fluorescence spectra of PAHs produced in thermal decomposition stage at Tr=1500K, F/G=0.13.

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

27

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

The fluorescence emission area of Solvent becomes narrower than those of W/O fuels, which seem to produce a few kinds of high molecule PAHs. The fluorescence spectrums of W/O fuels become wide, but some of the strong fluorescence intensity area move to a short wavelength. Therefore, the W/O generates a greater PAHs than Solvent but the high molecules PAHs are decreased.

3.4 Analysis of PAHs and CO, CO2 in exhaust gas

Excitation wavelength (nm)

PAHs, which is produced from Solvent in thermal decomposition, analyzed by GC/MS in former studies revealed that a little PAHs were detected at Tr=1350K and many PAHs were detected at Tr=1500K (Noge et al., 2007, 2011). Those production tendencies are similar to this study as well, however, as the combustion temperature is higher than the temperature in thermal decomposition, PAHs tend to be produced during the combustion. At this moment, some of PAHs avoiding oxidation will be included in PM or coagulated and aggregated to become PM. PAHs included in PM whose samples were detected by Soxhlet extractor were analyzed by FL and the results are compared with the PAHs produced during thermal decomposition. The measurement conditions are Tr=1350K and Tr=1500K withφb =2.0, which correspond to the conditions of the thermal decomposition: F/G=0.13 experiments. Fig.6(a) shows the results at Tr=1350K. Solvent and W/O10 expand the strong fluorescence intensity area where some kinds of PAHs increase are expected. The thermal decomposition caused by Solvent and W/O10 at 1350K make slow progress as mentioned above, but it is found that strong fluorescence area are expanded because oxygen supplied to the burner probably increase flame temperature where thermal decomposition is more reactive than oxidation. Therefore, the amount of PAHs are likely to be generated if thermal decomposition before combustion is sluggish. W/O20 encourages thermal decomposition before combustion in which somewhat PAHs are generated rather than both Solvent and W/O10, however, it is remarkable the PAHs are easy to be oxidized and hard to be included in PM during combustion because W/O20 shows high total LHCs production ratio per hydrocarbons content in original fuels mentioned in Fig.4, which flammable LHCs may contribute the oxidation of PAHs. PAHs produced in the thermal decomposition at Tr=1500K are oxidized partially during the combustion and the fluorescence intensity and its area are shown in the following order : Solvent > W/O10 > W/O20 in Fig.6(b). The strong intensity area are reduced at any fuel in the changes from the thermal decomposition to the combustion. PM concentration in total are high in comparison with the case of low temperature and large amount of PAHs during thermal decomposition are generated, however, emulsion fuel can oxide most of PAHs in the high temperature flame. The production tendency of the PAHs after combustion shown in Fig.6(a) and Fig.6(b) resembles PM characteristics in Fig.2. In this experiment, as PAHs emitted by the thermal decomposition and the combustion affect PM concentration, this EEM method is effective to evaluate a relation between the PAHs and PM. 750

10000

Fuel:Solvent Tr=1350K φ b:2.0

600 450

19 7 13

300 200

27 9

300

400

500

600

700 300

400

Fuel:W/O20 Tr=1350K φ b:2.0

Fuel:W/O10 Tr=1350K φ b:2.0

7500

5000

15

27

500

600

700 300

6

400

1

2500

500

600

700

0

Fulorescence wavelength (nm)

Excitation wavelength (nm)

Fig.6(a) Three-dimensional fluorescence spectra of PAHs produced in diffusion combustion at Tr=1350K, φb=2.0. 750

10000

Fuel:Solvent Tr=1500K φ b:2.0

600 29 9

450

200

31

300

400

500

15 14

8 21

300

Fuel:W/O20 Tr=1500K φ b:2.0

Fuel:W/O10 Tr=1500K φ b:2.0

600

16 700 300

400

500

600

700 300

400

7500

5000

5

500

2500

600

700

0

Fluorescence wavelength (nm) Fig.6(b) Three-dimensional fluorescence spectra of PAHs produced in diffusion combustion at Tr=1500K, φb=2.0.

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

28

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

Here, the effect of W/O fuel on PM has been reported by analyzing thermally decomposed hydrocarbons such as light hydrocarbons and PAHs. Then the oxidation and growth of PAHs generated in thermal decomposition through diffusion combustion have been evaluated. Finally, to estimate the relation between thermal decomposition and oxidation of W/O about PM reduction, CO and CO2 generated through the diffusion combustion were measured. The measurement condition was fuel rich: φ b=2.0 where much PM and PAHs were detected. CO and CO 2 value per unit Solvent are described in Fig.7. Here the value is calculated as:

CO or CO2 GC area

(2)

Solvent GC area in the each fuel

In thermal decomposition, mentioned above, W/O fuels approximately introduce the increase of LHCs production ratio, which enables the burner combustion easy to promote. As a result, CO 2 / Solvent in Fig.7 is also increased with increasing water. There will be one of relations between thermal decomposition and combustion. Regarding CO / Solvent at Tr=1350K, there is little difference among three fuels. CO 2 / Solvent increases when water in Solvent goes on increasing. W/O promotes thermal decomposition and oxidation of PAHs during the combustion, contributing to particularly solid reduction and thus the value of CO 2 / Solvent at Tr=1350K may have a trade-off relation with solid in PM. At Tr=1500K, reduction of CO / Solvent and increase of CO 2 / Solvent in the exhaust gas generated after the diffusion combustion may be invited by both SOF and Solid reduction. In other words, to reduce PM at certain temperature effectively by using water emulsion fuel, it is favorable to use it in a high temperature atmosphere.

(CO, CO2/Solvent)×10-4

3.0 CO 2.0

φ b=2.0

Tr=1500K

Tr=1350K

1.0 0

10.0

CO2

Tr=1500K

8.0 Tr=1350K

6.0 Solvent

W/O10

W/O20

Fig.7 CO and CO2 generation after diffusion combustion at Tr=1350K and 1500K, φb=2.0.

4. Conclusions To estimate chemical effect on PM reduction by W/O fuel during diesel combustion, this study focuses on analysis of thermally decomposed hydrocarbons: light hydrocarbons, PAHs and exhaust gas: PAHs, CO, CO 2. Flow reactor is used to isolate thermal decomposition before combustion and the thermally decomposed gas was burned by co-flow diffusion burner. In this way, gas analysis is carried out in the condition that one of thermal decompositions is separated from main diesel combustion. 1)W/O fuels can reduce PM. In particular, at low temperature thermal decomposition before combustion, W/O fuels cause a solid reduction, and at high temperature thermal decomposition, W/O fuels contribute to both Solid and SOF

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

29

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

reduction. 2)Promotion of thermal decomposition by W/O fuels leads to PM reduction. 3)In the case of PM is reduced, PAHs produced during thermal decomposition before combustion are easily oxidized by the diffusion combustion with increasing of water in W/O fuels. 4)EEM method is effective for assessing generation and oxidation of PAHs and PM concentration. 5)Diffusion combustion of W/O fuels contributed to SOF and Solid reduction which leads to reduction of CO / Solvent and increase of CO2 / Solvent respectively in the exhaust gas.

Acknowledgment Highest gratitude to the experimental work done by Shinya Imanaka and Masazumi Nishino, financial support by National Institute of Technology, Maizuru College and helpful support of fluorescence spectrophotometer by Kyoto Prefectural Technology Center for Small and Medium Enterprises.

References Aizawa, T., Kosaka, H., Investigation of Early Soot Formation Process in a Diesel Spray Flame via Excitation-Emission Matrix (EEM) using a Multi-Wavelength Laser Source, International Journal of Engine Research 9(1), (2008) , pp.79-97. Armasa O., Ballesterosa R., Martosb F. J., Agudeloc J. R, Characterization of light duty diesel engine pollutant emissions using water-emulsified fuel. Fuel Vol.84, (2005), pp.1011–1018. Beltrán J.L., Ferrer R. and Guiteras J., Multivariate calibration of polycyclic aromatic hydrocarbon mixtures from excitation–emission fluorescence spectra, Analytica Chimica Acta, Vol. 373, Issues 2–3, (1998), pp. 311–319. Bertola A., Li R. and Boulouchos K., Influence of water-diesel fuel emulsions and EGR on combustion and exhaust emissions of heavy-duty DI-diesel engines equipped with common-rail injection system. SAE Paper 2003-01-3146, 2003. Cyril Crua, David A. Kennaird, and Morgan R. Heikal, Laser-induced incandescence study of diesel soot formation in a rapid compression machine at elevated pressures, Combustion and Flame 135 (2003) 475–488. D'Anna A. and Violi A., Detailed Modeling of the Molecular Growth Process in Aromatic and Aliphatic Premixed Flames, Energy & Fuels, Vol.19, Issue 1, (2005), pp.79-86. D’Anna A., D’Alessio A. and Kent J., computational study of hydrocarbon growth and the formation of aromatics in coflowing laminar diffusion flames of ethylene, Combustion and Flame, Vol.125, Issue 3,(2001), pp.1196-1206. Dabestani R. and Ivanov I., A compilation of physical, spectroscopic and photophyisical properties of polycyclic aromatic hydrocarbons, Photochemistry and Photobiology, 70, (1999), pp.10–34. Henningsen, S., Influence of the fuel injection equipment on NOx emissions and particulates on a large heavy-duty two-stroke diesel engine operating on water-in-fuel emulsion. SAE paper 941783, (1994). Hermann G, Huttinger KJ. Mechanisms of non-catalysed and iron-catalised water vapour gasification of carbon. Fuel Vol.65, (1986), pp.1410-1418. Huoa M., Linb S., Liud H. and Lee C. F. F. Study on the spray and combustion characteristics of water–emulsified diesel, Fuel Vol.123, (2014), pp.218-229. Jazair Yahya W., Harada T., Kubo S., Kidoguchi Y. and Miwa K., Improvement of emission in a DI diesel engine fuelled by bio-diesel fuel and waste cooking oil, SAE paper No.2007-01-2029, (2007). JiJi R. D. and Booksh K.S., Mitigation of rayleigh and raman spectral interferences in multiway calibration of excitation−emission matrix fluorescence spectra, Analytical Chemistry, 72, (2000), pp. 718–725. Kee S.S., Mohammadi A., Hirano H., Kidoguchi Y. and Miwa. K., Experimental study on combustion characteristics and emissions reduction of emulsified fuels in diesel combustion using a rapid compression machine. SAE paper 2003-01-1792, (2003). Law C.K., Lee, C.H. and Srinivasan, N., Combustion characteristics of water-in-oil emulsion droplets Combustion and Flame, Vol.37, (1980), pp.125-143. Lea-Langtona A.R., Rossa A.B., Bartlea K.D., Andrewsa G.E., Jonesa J.M., Lia H., Pourkashanianb M. and Williams A., Low temperature PAH formation in diesel combustion, Journal of Analytical and Applied Pyrolysis, 103, (2013), pp.119-125.

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

210

Noge, Kidoguchi, Jazair Yahya, Imai and Tajima, Journal of Thermal Science and Technology, Vol.10, No.2 (2015)

Liang Y., Shu G., Wei H. and Zhang W., Effect of oxygen enriched combustion and water–diesel emulsion on the performance and emissions of turbocharged diesel engine. Energy Conversion and Management. 73, (2013), pp. 69–77. Lif, A., Skoglundh, M., Gjirja, S., and Denbratt, I., "Reduction of Soot Emissions from a Direct Injection Diesel Engine using Water-in-Diesel Emulsion and Microemulsion Fuels", SAE Paper 2007-01-1076, (2007) Liu F., Consalvi J. L. and Fuentes A., Effects of water vapor addition to the air stream on soot formation and flame properties in a laminar coflow ethylene/air diffusion flame, Combustion and Flame, Vol.161, (2014), pp.1724-1734. Maiboom A. and Tauzia. X., NOx and PM emissions reduction on an automotive HSDI Diesel engine with water-in-diesel emulsion and EGR: An experimental study. Fuel 90, (2011), pp.3179-3192. Murotani T., Hattori K., Sato E., Chryssakis1 C., Babajimopoulos A. and Assanis D. N., Simultaneous Reduction of NOX and Soot in a Heavy-Duty Diesel Engine by Instantaneous Mixing of Fuel and Water, SAE paper 2007-01-0125, (2007). Musculus M. P. B, Dec, J. E, Tree D. R, Daly D., Langer D., Ryan T. W. and Matheaus A. C., Effects of water–fuel emulsions on spray and combustion processes in a heavy-duty DI diesel engine. SAE Paper 2002-01-2892, (2002). Neeft J.P.A., Nijhuis T. X., Smakman E, Makkee M and Moulijn J. A., Kinetics of the oxidation of diesel soot. Fuel Vol.76, (1997) pp.1129-1136. Noge H., Yoshihara Y., Kidoguchi Y. and Miwa K., Soot precursor and PM formation in diffusion flame using thermally decomposed aliphatic diesel fuel, Trans. of JSME (in Japanese) Series B, Vol. 73, No. 725(2007), pp. 328-334.. Noge H., Hosomi N. and Kidoguchi Y., A fundamental study of role of thermal decomposition of diesel fuel mixed with bio-diesel fuel on PM reduction, Trans. of JSME (in Japanese) Series B, Vol. 77, No. 774(2011), pp. 360-367. Roberts C. E., Naegeli D. and Chadwell C., The effect of water on soot formation chemistry, SAE Paper 2005-01-3850, (2005). Sheng H. Z., Chen L. and Wu C. K. The droplet group micro-explosions in W/O diesel fuel emulsion sprays, SAE SAE paper 9508552, (1995). Smith OI., Fundamentals of soot formation in flames with application to diesel engine particulate emissions. Progress in Energy and Combustion Science, Vol. 7, Issue 4, (1981) pp.275-291. Subramanian K. A. and Ramesh A., Experimental investigation on the use of water diesel emulsion with oxygenenriched air in a DI diesel engine. SAE paper No.2001-01-0205, (2001). Tajima K., Imai Y., Nakamura A. and Koenuma, M., Three-phase emulsion of hexadecane with dimyristoylphosphatidylglycerol sodium salt in water: Interpretation by new phase transition in bilayer assembly, J. Oleo Sci., 50, (2001), pp.475-484. Tajima K., Imai Y., Nakamura A. and Koshinuma M., Thermal transitions in the bilayer assembly of dimyristoylphosphatidylglycerol sodium salt dispersion in water: a new phase transition through an intermediate, Advances in Colloid and Interface Science, 88, (2002), pp.79-97. Treea D. R. and Svensson K. I., Soot processes in compression ignition engines. Progress in Energy and Combustion Science 33, (2007), pp.272–309. Tsa KC and Wang CL. Puffing and micro-explosion phenomena of water emulsion fuels. SAE Paper 860304, (1986). Tsue M., Segawa D., Kadota T. and Yamasaki H., Observation of sooting behavior in an emulsion droplet flame by planar laser light scattering in microgravity, Twenty-sixth symposium International on Combustion Vol. 26, Issue 1, (1996), pp.1251–1258. Watanabe H., Harada T., Matsushita Y., Aoki H. and Miura T., The characteristics of puffing of the carbonated emulsified fuel. International Journal of Heat and Mass Transfer, Vol.52, Issues 15–16, (2009), pp.3676-3684.

[DOI: 10.1299/jtst.2015jtst0024]

© 2015 The Japan Society of Mechanical Engineers

211