Prediction of the Effects of Ethanol-Diesel Fuel Blends on Diesel

(10) In addition, the effects of using of ethanol-diesel fuel blends at constant equivalence ratios (CER) on the engine performance and exhaust emissi...
4 downloads 0 Views 671KB Size
Energy & Fuels 2009, 23, 1707–1717

1707

Prediction of the Effects of Ethanol-Diesel Fuel Blends on Diesel Engine Performance Characteristics, Combustion, Exhaust Emissions, and Cost Z. S¸ahin*,† and O. Durgun‡ NaVal Architecture and Marine Engineering Department, Faculty of Marine Science, and Mechanical Engineering Department, Faculty of Engineering, Karadeniz Technical UniVersity, Trabzon 61530, Turkey ReceiVed July 24, 2008. ReVised Manuscript ReceiVed December 29, 2008

Ethanol is a promising renewable oxygenated fuel for engines, and many experimental studies on the using of ethanol-diesel fuel blends in diesel engines have been done. But modeling studies on ethanol blends are very scarce. For this reason; the present study intends to investigate numerically the effects of the use of ethanol-diesel fuel blends on the engine performance characteristics such as brake specific fuel consumption (BSFC), brake effective power, brake effective efficiency, exhaust emissions, and cost by using two different turbocharged direct-injection (DI) diesel engines. A computer program has been used for prediction of diesel engine cycles and engine characteristics for the case of neat diesel fuel (NDF) and this program was modified for ethanol-diesel fuel blends. In the diesel engine cycle modeling, a quasi-dimensional phenomenological combustion model previously developed by authors has been used. This model is based on the model originally developed by Shahed and then improved by Ottikkutti, and it has been modified by the authors with new assumptions. By doing some modifications and adaptations in this model it has been converted to the ethanol-diesel fuel blends version. After the engine cycle model for NDF and ethanol-diesel fuel blends was proven to give correct results by comparing with relevant experimental and numerical results, (2-10) % ethanol-diesel fuel blends have been investigated numerically. The results indicate that as ethanol percentage in the mixture increases, BSFC reduces and brake effective efficiency improves significantly and brake effective power increases slightly. On the other hand, equivalence ratio decreases and ignition delay increases for ethanol blends, and combustion duration exhibits generally a decreasing tendency. The concentrations of nitric oxide (NO), the mole fractions of carbon monoxide (CO), and hydrogen (H2) increase at low ethanol ratios because of increase of the temperatures of the cylinder contents. But at high ethanol ratios they decrease because of decreasing temperatures. In the present study, cost analysis has also been performed by using a semi empirical relation given by Durgun. It was determined that ethanol blends are not economical for these engines because the cost of ethanol is higher than that of diesel fuel in Turkey, as well as in many of the other countries, and the decrease in the BSFC is low. In the present study, the effects of the using of ethanol blends at constant equivalence ratios (CER) have also been investigated. In this application, BSFC enhances with increasing ethanol ratios. Also, brake effective power, brake effective efficiency, and combustion duration increase until (4-6) % ethanol ratios at CER and after these ratios they start to decrease. NO concentration and the mole fractions of CO and H2 show generally a decreasing tendency.

1. Introduction Direct injection (DI) diesel engines are widely used as fuelefficient power sources for automotive applications because of their superior efficiency (fuel economy) relative to that of spark ignition and indirect injection diesel engines at equivalent capacity. However, the increase of prices of diesel fuel, stringent emission regulations, and foreseeable future depletion of petroleum reserves make it necessary to research and to develop new technologies such as common-rail system, fuel injection control strategies, exhaust gas recirculation, fuel-related techniques, and so on to meet demands for environment and energy. Studies on alternative fuels, especially renewable fuels, become very common important research areas among fuel-related studies. Nowadays considerable attention has been paid to the development of * To whom correspondence should be addressed. E-mail: zsahin@ ktu.edu.tr. Phone: 0462 752 2805/110. Fax: 0462 7522158. † Naval Architecture and Marine Engineering Department. ‡ Mechanical Engineering Department.

alternative fuel sources in various countries, with particular emphasis on biofuels.1,2 Ethanol is a promising renewable energy source which can be locally produced and used to extend petroleum fuel sources. Use of anhydrous ethanol as a gasoline additive component is readily accepted in many markets around the world. Over the past few decades, researchers have also investigated different techniques of using ethanol in compression ignition engines to extend diesel fuel supplies and gain the benefits of reduced smoke and particulate matter emissions. These techniques can be generally divided into the following three categories:1-3 (1) He, B-Q.; Wang, J-X.; Yan, X.-G.; Tian, X.; Chen, H. Study on Combustion and Emission Characteristics of Diesel Engines Using Ethanol Blended Diesel Fuels. Society of Automotive Engineers Technical Paper No. 2003-01-0762, 2003. (2) Rakopoulos, C. D.; Antonopoulos, K. A.; Rakopoulos, D. C.; Hountalas, D. T. Multi-zone modeling of combustion and emissions formation in DI diesel engine operating on ethanol-diesel fuel blends. Energy ConVers. Manage. 2008, 49 (4), 625–643. (3) Abu-Qudais, M.; Haddad, O.; Qudaisat, M. The effect of alcohol fumigation on diesel engine performance and emissions. Energy ConVers. Manage. 2000, 41 (4), 389–399.

10.1021/ef800587e CCC: $40.75  2009 American Chemical Society Published on Web 02/10/2009

1708 Energy & Fuels, Vol. 23, 2009

(a) Ethanol fumigation to the intake air charge, by using carburetion or manifold injection techniques. It seems that limited the amount of ethanol can be used in this manner because of engine knocking tendency at high loads and prevention of flame quenching and misfire at low loads. (b) Dual injection system that is not considered very practical, as requiring an extra high-pressure injection system for ethanol and, thus, related major design modifications of the cylinder head. (c) Blends (emulsions) of ethanol and diesel fuel by using an emulsifier to mix different fuels to prevent separation, not requiring any modification on the engine side.2 The most attractive and the simplest one of these techniques is using the ethanol blends in diesel engines, and it has been applied in the present study. Many relevant experimental studies on ethanol-diesel fuel blends have been performed in the literature.1,3-9 He et al.1 have investigated the effect of using of ethanol blended diesel fuels on brake specific fuel consumption [g/kWh], brake specific energy consumption [MJ/kWh], smoke and NOx emissions in a DI diesel engine. Their results indicate that with the increase of ethanol ratio in the blends, smoke reduces significantly, brake specific energy consumption improves slightly, and combustion duration decreases. On the other hand, the rate of heat release increases, but ignition delay, brake specific fuel consumption, NOx and unburned hydrocarbon emissions increase. Li et al.4 have investigated experimentally the effects of different ethanol-diesel fuel blends on the performance and emissions of a water-cooled single-cylinder DI diesel engine. In this study it is determined that brake specific fuel consumption and brake thermal efficiency increase, smoke emission decreases, and CO and NOx emissions decrease with the increase of the ethanol content in the blended fuel at the same operating conditions. However, total hydrocarbon emissions increase significantly with increasing ethanol content. Bilgin et al.5 have determined experimentally the variations of performance parameters for a variable compression ignition engine operated with ethanol-diesel fuel blends. The engine has been operated with ethanol-diesel fuel blends having 2, 4, and 6% ethanol on a volume basis, as well as neat diesel fuel (NDF). Their experimental results indicate that the addition of 4% ethanol to diesel fuel increases brake effective power output and brake effective efficiency of the engine, and at the same time BSFC decrease for various compression ratios. Although there are many experimental studies on the ethanoldiesel fuel blends,1,3-9 modeling studies related to this subject are very limited. Rakopoulos et al.2 studied ethanol-diesel fuel blends by applying a two-dimensional multizone thermodynamic model in a DI diesel engine. In their study useful information about combustion has been provided by considering the combustion mechanism by generating maps inside the fuel jet (4) Li, D.; Zhen, H.; Xingcai, L.; Wu-Gao, Z. Physico-chemical properties of ethanol-diesel blend fuel and its effect on performance and emissions of diesel engines. Renewable Energy 2005, 30 (6), 967–976. (5) Bilgin, A.; Durgun, O.; S¸ahin, Z. The effect of diesel-ethanol blends on diesel engine performance. Energy Sources 2002, 24, 431–440. ¨ .; C¸elikten, I.; Usta, N. Effects of ethanol addition on (6) Can, O performance and emissions of turbocharged indirect injection diesel engine running at different injection pressure. Energy ConVers. Manage. 2004, 45 (15-16), 2429–2440. (7) Czerwinski, J. Performance of HD-DI-diesel engine with addition of ethanol and rapeseed oil. Society of Automotive Engineers Technical Paper No. 94054562, 1994. (8) Rakopoulos, D. C.; Rakopoulos, C. D.; Kakaras E. C.; Giakoumis, E. G. Effects of ethanol-diesel fuel blends on the performance and exhaust emissions of heavy duty DI diesel engine. Energy Convers. Manage. 2008, (In press). (9) Rakopoulos, C. D.; Antonopoulos, K. A. Rakopoulos. Experimental heat release analysis and emissions of a HSDI diesel engine fueled with ethanol-diesel fuel blends. Energy 2007, 32 (10), 1791–1808.

S¸ahin and Durgun

showing the distribution of fuel and temperature both before and after wall impingement. They determined that higher fuel to air equivalence ratio areas within the fuel spray exist in a very limited extent when using ethanol-diesel fuel blends instead of diesel fuel. Because the oxygen contained in the fuel is progressively released as the oxygenated fuel is evaporated, this can become available to assist the combustion of the fuel, especially in very rich fuel areas. This can have a favorable effect on lessening the occurrence of pyrolysis, and enhancing soot oxidation rate and NO concentration could also be reduced. It can be seen from the above brief explanations and review that there are many experimental studies on the ethanol-diesel fuel blends.1,3-9 But modeling studies on this subject are very scarce.2 For this reason, the present study aims to investigate theoretically the effects of using of ethanol blends at variable equivalence ratios (VER) on the engine performance and exhaust emissions and to compare with that of NDF in detail in two different DI turbocharged diesel engine. Furthermore, cost analysis has been conducted by using a semiempirical approach given by Durgun.10 In addition, the effects of using of ethanoldiesel fuel blends at constant equivalence ratios (CER) on the engine performance and exhaust emissions have been investigated for low ethanol ratios. 2. Description of the Used Model Developed for NDF In the used cycle model, a multizone thermodynamic based model developed by Shahed11,12 and then Ottikkutti13 has been used and improved with new assumptions to calculate the complete engine cycle and engine characteristics. Detailed information about this model has been given in the authors’ previous studies.14-18 Here, a brief description of the model has been presented. In this model, the spray injected into the combustion chamber is divided into several zones. The boundaries of these zones are determined from lines of CER. Applying the first law of thermodynamics, the ideal gas equation, and other basic relations to the cylinder charge (these zones), a system of ordinary differential equations for cylinder pressure and zone volumes have been obtained. By solving these ordinary differential equations simultaneously during the engine cycle by using the Runge-Kutta 4 method, cylinder pressure and zones volumes can be calculated. Also, the temperatures of the zones can be computed from the ideal gas equation by using the obtained cylinder pressure and zone volumes. (10) DurgunO. AyvazY. The use of diesel fuel-gasoline blends in diesel engines. First Trabzon International Energy and EnV. Sysm.; Karadeniz Technical University: Trabzon, Turkey, 1998; Issue 2, pp 905912. (11) Shahedt, S. M.; Flynn, P. F.; Lyn, W. T. A model for the formation of emissions in a direct-injection diesel engine. In Combustion Modeling in Reciprocating Engines; Mattavi, J. N., Amann, C. A., Eds.; Plenum Press: New York, 1980; Vol. 34, pp 5-368. (12) Shahed, S. M.; Chiu, W. S.; Lyn, W. T. A mathematical model of diesel combustion. Proc. I. Mech. E. 1975, C94/75, 119–128. (13) Ottikkutti, P. Multizone modeling of a fumigated diesel engine. Ph. D. Thesis, Iowa StateUniversity of Science and Technology, Ames, Iowa, 1989. (14) S¸ahin, Z. Effect of using diesel fuel-light fuel blends on combustion and engine performance in diesel engine. Ph. D. Thesis, KTU Graduate School of Natural and Applied Sciences, Trabzon, 2002 (in Turkish). (15) S¸ahin, Z.; Durgun, O. Multi-zone combustion modeling for the prediction of diesel engine cycles and engine performance parameters. Appl. Therm. Eng. 2008, in press, ATE-2007-447R3. (16) S¸ahin, Z.; Durgun, O. High speed direct injection (DI) light-fuel (gasoline) fumigated vehicle diesel engine. Fuel 2007, 86, 388–399. (17) S¸ahin, Z.; Durgun, O. Theoretical investigation of light-fuel fumigation on diesel engine performance and emissions. Energy ConVers. Manage. 2007, 48 (7), 1952–1964. (18) S¸ahin, Z.; Durgun, O. Theoretical investigation of light fuel fumigation and probable developments in diesel engines. 14th International Conference on Thermal Engineering and Thermogrammetry (THERMO); Budapest, Hungary, 2005.

Ethanol-Diesel Fuel Blends

Energy & Fuels, Vol. 23, 2009 1709

In the cycle model, Dent’s19 correlation for spray penetration and Reitz’s correlation given by Heywood20 and Wakuri’s21 correlation for spray angle are used. The effect of swirl on spray penetration and cone angle is incorporated by using Hiroyasu’s22 approach. For determination of the instantaneous total mass and instantaneous mass rates of spray zones, it is required to know the spatial distribution of diesel fuel in the zones. In the present study, the concentration distribution of the fuel along the spray axis is assumed to be hyperbolic, while across the spray it is taken as a normal distribution curve by benefit of information given in literature.11-13 In this model, thermodynamic properties and their partial derivatives have been computed for the unburned mixture and for the burned equilibrium products by using Olikara’s method.23 The instantaneous total heat transfer from the cylinder contents to the combustion chamber and cylinder walls is calculated using Annand’s24 correlation

{[

}

˙ cyl ) A a K (Re)0.7(T - Tw) + [bA(T 4 - Tw4)] Q ω B

]

(1)

where the first and the second terms on the right-hand side show convective heat transfer and radiative heat transfer, respectively. In this equation; B is cylinder bore, A is heat transfer area, K is thermal conductivity of the cylinder gases, Re is Reynolds number based on piston speed and cylinder bore, T is cylinder bulk gas temperature, Tw is cylinder wall temperature, bA ) 3.267 × 10-11 [W/m2 K4], and a is a constant ranging from 0.35 to 0.8 depending on intensity of charge motions. In the used model, a has been selected as 0.35 for the compression and expansion strokes, and its value has been taken as 0.49 for the combustion stroke. But for the test engine used in the experiments this coefficient has been selected as 0.55 during the combustion process by evaluation of the results obtained from a lot of numerical applications. Also, the second term in eq 1 represents the radiant heat flux assuming graybody radiation, and its value is taken as zero for the compression and expansion strokes. Tw has been taken constant, selecting for it an appropriate temperature in the range of 350 to 600 [K] for each engine. Gas transport properties such as dynamic viscosity µ and thermal conductivity K must be determined for solving eq 1. These transport properties have been evaluated for actual mean gas temperature, pressure, and equivalance ratio values. Here, the transport properties have been computed by use of approximate correlations for viscosity and Prandtl number given by Heywood.20 Total heat transfer to the walls is distributed to various zones according to their mass, absolute temperature, and specific heat capacity values.15,25 During the compression and the expansion processes, differential equations for cylinder pressure and temperature given by Heywood have been solved to determine cylinder pressure (19) Dent, J. C. A basis for comparison of various experimental methods for studying spray penetration. Society of Automotive Engineers Technical Paper No. 710571, 1971. (20) Heywood, J. B. Internal combustion engines fundamentals; McGrawHill: New York, 1989. (21) Wakuri, Y.; Fujii, M.; Amitani, T.; Tsuneya, R. Studies on the penetration of fuel spray in a diesel engine. Bull. JSME.3 1960, 9, 123– 130. (22) Hiroyasu, H.; Kadota, T.; Arai, M. Fuel spray characterization in diesel engines. In Combustion modeling in reciprocating engines; Mattavi, J. N., Amann, C. A., Eds.; Plenum Press: New York, 1980; Vol. 36, pp 9-404. (23) Olikara, C.; Borman, G. L. A computer program for calculating properties of equilibrium combustion products with some applications to I.C. engines. Society of Automotive Engineers Technical Paper No. 750468, 1975. (24) Annand, W. J. D. Heat transfer in the cylinders of reciprocating internal combustion engines. Proc. Inst. Mech. Eng. 1963, 177, 973–990.

and temperature values. Also, intake and exhaust processes have been computed approximately by using a semiempirical method given by Durgun.26 In the present study, residual gases which remain in the cylinder from the previous cycle have been taken into account by using the coefficient of residual gases at the intake process calculations. However, the effects of the residual gases were neglected in most of the engine cycle modeling studies. In the used cycle model, residual gas temperature has been chosen approximately at the beginning of the cycle calculations. Then, after completing the cycle calculations, chosen and calculated exhaust temperatures have been compared. If the difference between these values were higher than 2%, the final value has been taken as the exhaust temperature, and the cycle calculations have been repeated again. This calculation procedure has been applied iteratively until the difference between these values became smaller than 2%. Thus, complete cycle control has been performed. In the used cycle model, a correction factor of indicator diagram φi has been used to take into account injection advance, ignition delay, and valve timing effects. The numerical range of the correction factor φi is given as 0.92-0.95 in the literature. But in the used model, injection advance and ignition delay have been considered, and the pressure-volume variation during the combustion process has been determined at a rounded character. Here, only exhaust valve opening advance must be taken into account. For this reason φi was chosen somewhat higher than that of usual values. Thus, φi has been selected as 0.98.26 Finally, indicated work and indicated efficiency obtained from gross cycle simulation have been corrected as follows: ηi )

Wgrφi(1 + Fstφ) Wi(1 + Fstφ) ) Fstφ(1 - f )QLHV Fstφ(1 - f )QLHV

(2)

where Wgr is the work done by per unit mass of the cylinder contents (known as gross work), Wi is indicated useful work, Fst is the stoichiometric fuel-air ratio, φ is the equivalence ratio, QLHV is the lower heating value (LHV) of the fuel, and f is the residual gases mass fraction. In the present study, the LHV has been calculated by using well-known Mendeleyev’s formula.13,24-26 Some empirical values, such as intake pressure and exhaust pressure, and so on, at any engine speed have also been selected by benefit of values given for nominal engine speed. Moreover, a charge-up coefficient φch has been used to take into account the valve timing and superposition effects to minimize cycle simulation errors.15,26 After determining the complete diesel engine cycle, engine performance parameters, such as brake effective power, brake effective efficiency, BSFC, and so on, have been calculated by using relationships given by Heywood20 and Durgun.26 The used cycle model can be used for both naturally aspirated and turbocharged DI diesel engines. In the turbocharged version, a simple calculation procedure given by Durgun26 has been added to the computer code. Detailed information about this model and a flowchart are given in the authors’ previous studies.14-18 The used cycle model can also determine the exhaust emissions. Here mole fractions of CO and H2 have been (25) Kouremenos, D. A.; Rakopoulos, C. D.; Hountalas, D. T. Multizone combustion modeling for the prediction of pollutants emissions and performance of DI diesel engine. Society of Automotive Engineers Technical Paper No. 970635, 1997. (26) Durgun, O. A practical method for calculation engine cycles. Union of Chambers of Turkish Engineers and Architects, Chamber of Mech. Eng. 1991, 383, 18–29. (27) Kolchin, A.; Demidov, V. Design of AutomotiVe Engines; MIR Publishers: Moscow, 1984. (28) Khovakh, M. Motor vehicle engines; MIR Publishers: Moscow, 1979.

1710 Energy & Fuels, Vol. 23, 2009

S¸ahin and Durgun

Table 1. Reaction Rate Constants13

CNmix )

∑ (x CN ) ) x CN i

i

d

d + xethCNeth

computed by applying chemical equilibrium. Also, NO concentration has been calculated by using the three coupled reactions of the extended Zeldovich mechanism. Here the rate of NO concentration at constant volume is expressed as follows.13

(8) 100 100 where xd and xeth are the volumetric percentages of diesel fuel and ethanol, respectively. Fd and Feth are densities of diesel fuel and ethanol, respectively. In the present paper hmin is equal to 1/Fst. As any empirical relation about spray penetration and spray angle for ethanol mixtures has not been given in the literature yet, the empirical formulas developed for NDF have been used to calculate spray penetration and spray angle in this study. To obtain the properties of the combustion products for any blend consisting of two fuels, the following combustion reaction can be written.

d[NO]V1

y13 (xdC13.342H24.74 + xethC2H5O)+

rate constant [cm3/mol/s]

reaction

7.6 1.6 6.4 1.5 4.1 2.0

1-f O+N2fNO+N 1-b NO+NfO+N2 2-f N+O2fNO+O 2-b NO+OfN+O2 3-f N+OHfNO+H 3-b NO+HfN+OH

dt

temperature range [K]

× × × × × ×

103 exp(-38000/T) 1013 109 T exp(–3150/T) 109 T exp(–19500/T) 1013 1014 exp(–23650/T)

2000–5000 300–5000 300–3000 1000–3000 300–2500 2200–4500

) (2kf,1kf,2[O][N2][O2] - 2kb,1kb,2[NO] [O]+ 2

2kf,1kf,3[O][N2][OH] - 2kb,1kb,2[NO]2[H]) ⁄ (kb,1[NO]+kf,2[O2]+kf,3[OH]) (3) where kf and kb are temperature-dependent forward and backward reactions rates for the reactions given in Table 1. The concentrations of H, N, O, N2, O2, and OH in each zone are at their equilibrium conditions in eq 3, and these values have been determined by using Olikara’s method corresponding to the cylinder pressure, zone temperatures, and zone compositions. In an actual engine, the volume of any individual zone changes with crank position and hence the instantaneous NO concentration in each zone depends on both zone volume and reaction rate. The rate of NO concentration in any zone is given as d[NO]i d[NO]Vi [NO]i dVi ) dt dt Vi dt

(4)

where dVi/dt is the variation rate of zone volume which changes with crank position and Vi is the instantaneous cylinder volume. The instantaneous concentration of NO in the cylinder is calculated by combining the instantaneous NO concentrations in the individual zones. 3. Modifications in the Used Model for Ethanol Blends In the present study, the diesel engine cycle model developed originally for NDF is modified for ethanol blends as follows. In the present numerical applications, it is considered that ethanol and diesel fuel is mixed in the fuel tank and injected by using the usual pump-injector system. The main properties such as LHV, stoichiometric air requirement, gas constant, and the cetane number, and so on of any mixture have been calculated by using the following formulas.10,29

∑ (x F Q ) ) x F Q + x F Q x F +x F ∑ (x F ) ∑ (x F h ) ) x F h + x F h ) x F +x F ∑ (x F ) ∑ (x F R ) ) x F R + x F R R ) x F +x F ∑ (x F )

QLHV,mix )

i i

LHV,i

d d

LHV,d

d d

i i

i i min,i

hmin,mix

d d min,d

d d

i i

i i i

d d d

eth eth

eth eth min,eth

i i

(5)

(6)

eth eth

eth eth eth

mix

d d

LHV,eth

eth eth

(7)

eth eth

(29) Durgun, O. Using ethanol-gasoline-isopropanol mixtures in the internal combustion engine. 2nd Combustion Symposium; Istanbul Technical University: Istanbul, Turkey, 1989; pp 325-335 (in Turkish).

[

]

(c + 0.25h - 0.5of) (O2 + φ

3.7274N2 + 0.0444Ar) f [y1H + y2O + y3N + y4H2 + y5OH + y6CO + y7NO+y8O2 + y9H2O + y10CO2 + y11N2 + y12Ar] (9) where xd and xeth are the volumetric percentages of diesel fuel and ethanol, respectively. Here, the mole numbers of all of the products and 12 unknown mole fractions yi have been determined by using Olikara’s method.23 After calculating of these mole numbers and mole fractions, thermodynamic properties and their partial derivatives, depending on temperature, pressure, and equivalence ratio have been computed by using functions given by Ferguson.30 3.1. Calculation of Variation Ratios of the Main Performance Characteristics and Cost Analysis. Variation ratios of brake effective power and other characteristics have been calculated in a similar way, for example, as follows:

(

)

Ne,mix - Ne,d ∆Νe 100[%] ) 100 Ne Ne,d

(10)

where Ne,mix and Ne,d are brake effective powers for ethanoldiesel fuel blends and NDF, respectively. For the cost comparison, the following relationship developed by Durgun10,29 has been used.

[

(

) ]

C2 - C1 ∆be X1 + X2r2 ∆C 100[%] ) 100[%] ) 1+ - 1 100 C1 C1 X1 + X2s2 be (11) where r1 + 1, r2 ) C2/C1 ) 5.90/2.53 ) 2.332, S1 ) 1, S2 ) F2/F1 ) 785/860.3 ) 0.912, C1 is cost of diesel fuel, C2 is cost of ethanol, F1 and F2 are densities of diesel fuel and ethanol, respectively, and ∆be/be is the difference ratio of BSFC. Here, units of (C1, C2), (F1, F2), and be are [TL/lt], [kg/m3], and [kg/ kWh], respectively. The main properties of diesel fuel and ethanol are given in Table 2. 3.2. Accuracy Control of Cycle Model for NDF and Modified Model for Ethanol and Gasoline Blends. Accuracy controls of the cycle models have been performed for NDF, gasoline, and ethanol fumigation, gasoline and ethanol blends, and heat balance. For the cases of NDF and fumigation, detailed comparisons with the previous experimental and other relevant numerical results have also been done in the authors’ previous (30) Ferguson, C. R. Internal combustion engines. Applied Thermosciences; John Wiley & Sons: New York, 1986. (31) Sahin, Z. Experimental and theoretical investigation of the effects of gasoline blends on a single cylinder diesel engine performance and exhaust emissions. Energy Fuels 2008, 22, 3201–3212.

Ethanol-Diesel Fuel Blends

Energy & Fuels, Vol. 23, 2009 1711

Figure 1. (a, b). Comparison of brake effective power and brake effective efficiency values predicted from cycle model and Bilgin et al.’s5 experimental results as functions of ethanol ratio %, respectively.

model have been performed in a single cylinder diesel engine at Karadeniz Technical University, Engineering Faculty Mechanical Engineering Department Internal Combustion Engines Laboratory. For this test engine, at a few instances the maximum difference ratios of the performance parameters reached to the levels of 18%. However, generally a satisfactory conformity between numerical and experimental results has been obtained. These discrepancies could be attributed to the approximately selecting of some of the parameters of the test engine, aging of the exhaust temperature thermocouple, and insufficiency of the spray penetration correlation. For example, when Dent’s spray penetration correlation was used, most of the injected fuel was predicted to burn toward the end of the expansion period. However, when Hiroyasu’s correlation was used, predicted combustion temperatures reached to the levels of 4000 [K]. For this reason, the spray penetration correlation was arranged by averaging Dent’s and Hiroyasu’s correlations after a lot of numerical applications, and the following correlation has been adapted for the test engine. Xt ) 0 . 65Xt,D + 0 . 35Xt,H

Figure 2. Comparison of pressure values predicted from the used cycle model and Rakopoulos et al.’s2 experimental results as functions of crank angle. Table 2. Main Properties of Diesel Fuel and Ethanol5,13,17,20,30 diesel fuel chemical formula molecular mass [kg/kmol] density [kg/m3] lower heating value [kJ/kg] cetane number enthalpy of vaporization (kJ/mol) cost [YTL/lt] (February 2008, Trabzon) 1YTL ) 1.7346 EURO auto ignition temperature [°C] vaporization latent heat [kJ/kg] boiling point [°C] elemental composition, by mass [%]

ethanol

C13.78H24.26 189.966 860.3 42805.17 45 74.08 2.53

C2H5OH 46.07 785 27427 8 42.34 5.9

254 250 180-360 c′ ) 0.871, h′ ) 0.1269

423 840 78 c′ ) 0.521, h′ ) 0.131, o′ ) 0.347

Table 3. Comparison of the Predicted Results from the Cycle Model with Experimental Results for NDF n ) 1300 [rpm]

n ) 1450 [rpm]

compression ratios, ε

∆Ne/Ne [%]

∆be/be [%]

∆ηe/ηe [%]

∆Ne/Ne [%]

∆be/be [%]

∆ηe/ηe [%]

19 20 23

10.671 10.605 8.632

-7.485 -8.459 -7.879

7.087 8.595 7.782

12.889 10.556 10.911

-12.500 -11.144 -10.557

13.278 11.245 11.290

studies.15-17 A satisfactory conformity can be observed between these results. In the present study, some examples of these comparisons results for NDF, gasoline, and ethanol blends have been presented. In Table 3, engine performance characteristics determined by application of the cycle model for NDF compared with authors’ previous experimental results. In Figures 1–3 and Table 4, brake effective power and brake effective efficiency obtained from the used model have been compared with experimental results for ethanol-diesel fuel blends given by Bilgin et al.5 and Rakopoulos et al.2 and experimental results for gasoline-diesel fuel blends given by S¸ahin,31 respectively. Bilgin et al.’s and S¸ahin’s experimental studies which have been used in comparisons with numerical results of the present

(12)

In the test engine, equivalence ratios range between 1.086-1.149. However, equivalence ratios of vehicle diesel engines are generally in the range of 0.650-0.80, and various combustion approximations and parameters which have been developed for vehicle diesel engines are valid for these equivalence ratios. Thus, in the performed computations, more fuel exists in the engine cylinder than that of the actual test engine. Then, as more fuel burns in the last burning zone, calculated combustion and exhaust temperatures become higher than that of experimental results. On the other hand, for determining the mechanical losses and mechanical efficiency, the empirical mean pressure of the mechanical losses formula developed originally for vehicle diesel engines has been used. This relation may not predict satisfactorily the mechanical losses of the test engine. However, as it can be seen from Figure 1a,b and Figure 3a,b predicted values agree reasonably with the measured data generally. Comparisons have been done with Rakopoulos’s experimental results for the mixture containing 15% ethanol. As shown in Figure 2, the calculated cylinder pressure values obtained from the used cycle model are lower than Rakopoulos’s experimental values during the combustion process, but cylinder pressure values are higher than Rakopoulos’s2 values during the expansion process. This difference might have arisen from the spray penetration equation used in the present study. A different spray penetration relation for ethanol-diesel fuel blends is required or a correction factor must be added to the spray penetration relation developed originally for NDF. But a spray penetration formula for ethanol blends has not been found in the literature, as cycle modeling studies for ethanol-diesel fuel blends are very scarce. Nevertheless, the brake effective efficiency value obtained from the used cycle model is about at the level of 0.352 and the brake effective efficiency given by Rakopoulos et al. is 0.329 for 15% ethanol-diesel fuel blend. Thus, the difference in brake effective efficiencies between the used cycle model prediction and Rakopoulos et al.’s experimental result is 6.991%. Also, at the same ethanol ratio, exhaust temperature values obtained from the used cycle model and given by Rakopoulos et al. are 608 and 660 K, respectively. Hence, the difference ratio between exhaust temperatures is 7.88%.

1712 Energy & Fuels, Vol. 23, 2009

S¸ahin and Durgun

Figure 3. (a, b) Comparison of brake effective power and brake effective efficiency values predicted from the cycle model and the experimental results as functions of gasoline ratio %, respectively. Table 4. Comparison of the Predicted Results from the Cycle Model and Bilgin et al.’s Experimental Results for Ethanol Blends ε ) 21, ethanol [%]

∆Ne/Ne [%]

∆ηe/ηe [%]

0 2 4 6

11.691 11.190 9.545 11.767

11.661 14.235 12.892 13.194

Table 5. Main Specifications of the Engines Used for Comparisons of Numerical Results with Experimental Data2,13,14,32 ε D [mm] H [mm] pp [bar] θs [deg] dn [mm] Z nn

Ottikkutti13

Li32

16.8 106.5 127 120-490 -15 0.3 4 4

14.5 138.7 152.4 -14 0.27 6 6

Rakopoulos et al.2 19.8 80.26 88.90 250 0-40 0.25 1 4

test engine14 18-24 90 120 90-250 -22 0.36 1 1

Thus, a satisfactory conformity can be observed in Figures 1–3 and Table 3-4 for NDF and ethanol and gasoline blends. 4. Numerical Results and Discussion In this section, various numerical applications have been performed to determine the effects of using the ethanol-diesel fuel blends in a diesel engine. Variations of the engine performance characteristics and exhaust emissions for ethanol blends have been investigated at two categories by using two different turbocharged DI diesel engines given by Ottikkutti13 and Li.32 The main specifications of these engines has been given in Table 5. Various numerical calculations have been done at three engine speeds. Selected speeds for Ottikkutti’s and Li’s engines are 1500, 1700, and 2100 [rpm], and 1600, 1900, and 2100 [rpm], respectively. Nominal engine speeds of Ottikkutti’s and Li’s engines are 1700 and 1900 [rpm], respectively. In the present study, the selected higher and lower speeds than nominal speeds are denoted as high speed and low speed, respectively. Various numerical calculations have been done at varied equivalence ratios (VER) and CER. Some examples of the obtained results have been presented in the following figures and tables. In Table 5, ε is the compression ratio, D is the cylinder bore, H is the stroke of piston, pp is the injection pressure, θs is the (32) Li, Q. Development of a quasi-dimensional diesel engine simulation for energy and avability analysis. Ph.D. Thesis, University of Illinois, Urbana-Champaign, 1992.

injection advance, dn is the nozzle diameter, Z is the cylinder number, and nn is the nozzle hole number. 4.1. Ethanol Blends at Varied Equivalence Ratios (VER). Figure 4a-c presents the effects of ethanol blends at VER on equivalence ratio, ignition delay, and combustion duration. Equivalence ratio decreases with increasing ethanol percentage in the mixture because the molecular structure of ethanol contains oxygen. This property of ethanol could improve the combustion process. Ignition delay increases as the percentage of ethanol in the mixture increases by the effects of its lower cetane number. Predicted combustion durations increase slightly at low ethanol blend ratios (such as until 2-4%) at high engine speeds. At low speeds and nominal speeds it decreases with increasing ethanol percentage in the mixture. This can be attributed to the oxygen containing ethanol molecular structure and the decreasing equivalence ratio with increasing ethanol ratio. The effects of ethanol at high ratios become dominant than that at low ratios. Thus, the decrement ratios of combustion duration at high ratios are larger.1 As shown in Figure 5a, there is no significant difference between the predicted cylinder pressure values for ethanol blends and NDF. Figure 5b,c indicates the predicted variations of cylinder temperature values for different ethanol blends at 1700 [rpm] and 1900 [rpm]. It can be seen from Figure 5b,c that predicted temperature values increase until approximately 6% ethanol blends. After this ratio, temperature values start to decrease. Because the latent heat of vaporization of ethanol is about 1.5 times greater than that of diesel fuel, this could decrease the temperature values in the cylinder.31 As explained above, the effects of ethanol at high ethanol ratios become dominant. As shown in Figure 6a,b, brake effective power increases slightly until nearly 6-8% ethanol-diesel fuel blends. But after these ratios, brake effective power starts to decrease. The numerical brake effective power increase variations at nominal engine speeds achieved nearly to the levels of 1.8 and 1.2 for Ottikkutti’s and Li’s engines, respectively. The reasons for the attained brake effective power improvement can be explained as follows: Because the molecular structure of ethanol contains oxygen and the equivalence ratio decreases with increasing ethanol ratio, burning of the charge could be improved. Also, most of diesel fuel burns closer to TDC because of the increase of ignition delay, and thus, more energy of diesel fuel could be released near to TDC than NDF. The self-ignition temperature of diesel fuel is lower than that of ethanol (Table 2). Thus, in the injected spray of the blends, diesel fuel could probably initiate the ignition process. Since the boiling point of ethanol is lower than that of diesel fuel,

Ethanol-Diesel Fuel Blends

Energy & Fuels, Vol. 23, 2009 1713

Figure 4. (a, b, c). Predicted variations of ignition delay, equivalence ratio, and combustion duration as functions of ethanol ratio %, respectively. Table 6. Predicted Variation Ratios of Brake Effective Power, BSFC, Brake Effective Efficiency and Cost for Ottikkutti’s and Li’s Engines, Respectively Ottikkutti’s engine n ) 1700 [rpm], ε ) 16.8

Li’s engine n ) 1900 [rpm], ε ) 14.5

ethanol [%]

∆Ne/Ne [%]

∆be/be [%]

∆ηe/ηe [%]

∆C/C1 [%]

∆Ne/Ne [%]

∆be/be [%]

∆ηe/ηe [%]

∆C/C1 [%]

2 4 6 8 10

0.338 0.983 1.729 1.149 0.949

-0.461 -1.382 -2.304 -1.843 -1.382

1.289 2.577 4.124 4.124 4.897

2.371 4.239 6.064 9.387 12.746

0.859 1.092 0.894 0.704 0.118

-0.940 -1.269 -1.175 -1.081 0.564

2.446 3.415 4.052 4.612 4.740

1.878 4.359 7.290 10.235 14.971

Table 7. Predicted Variation Ratios of NO Concentration and the Mole Fraction of CO at 1600 [rpm] and 1900 [rpm] for Li’s and Ottikkutti’s Engines, Respectively

ethanol [%] 2 4 6 8 10

Ottikkutti’s engine n ) 1600 [rpm], ε ) 14.5

Li’s engine n ) 1700 [rpm], ε ) 16.8

∆NO/NO [%]

∆NO/NO [%]

3.268 10.271 3.695 -0.786 -4.755

∆CO/CO [%] 3.965 27.865 0.161 -21.876 -25.376

24.177 6.290 1.569 -2.818 -7.805

∆CO/CO [%] 13.527 -2.759 -9.778 -15.657 -31.854

ethanol would evaporate more effectively before diesel fuel, and after the first autoignition of diesel fuel, fully evaporated ethanol would burn more rapidly than diesel fuel. Thus, it is thought that swirl and additional gas motions would take place in the cylinder as accumulated ethanol burns more rapidly than diesel fuel. These swirl and gas motions would mix rather well the ethanol-diesel fuel and the air mixture through the unburned blend spray. As a result, the diesel engine combustion process could be improved because of faster and more efficient burning of the fuel.5,16,17 At higher ethanol ratios brake effective power starts to decrease because it is well-known that the heating value of the fuel affects the power of any engine.1-4 The heating value of ethanol is lower

than that of diesel fuel. Thus, the lower energy capacity of ethanol-diesel fuel mixtures causes some decrement in the engine power when used in diesel engines without any modifications. At higher ethanol ratios the effects of LHV would be more dominant than that of lower ratios. Thus, as shown in Figure 6a,b, after 6% or 8% ethanol ratios, brake effective power begins to decrease. Brake effective efficiency for NDF is 0.377 whereas it becomes 0.382, 0.385, 0.387, 0.388, and 0.387 for 2, 4, 6, 8, and 10% ethanol ratios, respectively, as at engine speed of 2100 rpm for Li’s engine (see Figure 7a,b). As the percentage of ethanol in the mixture is increased, an increment tendency can be observed in the brake effective efficiency compared to NDF. This could occur because of the cooling effect of the ethanol, as well as more efficient combustion compared to NDF. Since the exhaust gas temperature is lower for the case of ethanol-diesel fuel blend operation, lesser heat loss through exhaust channels occurs, and higher brake effective efficiency could be obtained.34 Also, by reducing of combustion duration with increasing ethanol percentage in the mixture, combustion efficiency could be improved. Similar results have been reported by other researchers.1-3 These also explain (33) Sayın, C.; Uslu, K.; C¸anakc¸, M. Influence of injection timing on the exhaust emissions ofa dual-fuel CI engin. Renewable Energy 2008, 33 (6), 1314–1323.

1714 Energy & Fuels, Vol. 23, 2009

S¸ahin and Durgun

Figure 6. (a, b) Predicted variations of brake effective power as functions of ethanol ratio %.

Figure 5. (a, b, c) Predicted variations of pressure and temperature as functions of crank angle, respectively.

the reasons of being BSFC lower for ethanol blended fuels as shown in Figure 8a,b. Panels a and b of Figure 9 represent variation ratios of brake effective power and brake effective efficiency, respectively. It is well-known that variation of effective efficiency is simply the inverse of BSFC. Also, variation ratios of brake effective power, BSFC, brake effective efficiency, and cost are shown in Table 6 for Ottikkutti’s and Li’s engines. But attained increment ratios in efficiency for ethanol blends are low because it is well-known that the light fuel blending method for diesel engine is not exactly suitable. For this reason the light fuel fumigation method which requires some modifications is widely applied in diesel engines.3,16-18 Whereas light fuel blending at low ratios such as (2, 4, 6, 8, 10) % can be easily applied in diesel engines, and thus some improvements can be achieved. As can be seen from these figures and this table, brake effective power begins to decrease, and increase ratio of brake effective efficiency starts to reduce after approximately 6% ethanol ratio. Experiences gained in the developments and numerical applications of the used cycle model show that the most favorable percentage of ethanol is between 4% and 6% in respect to engine performance characteristics for the used engines.

Figure 7. (a, b) Predicted variations of brake effective efficiency as functions of ethanol ratio %.

As shown in Table 6, ethanol blends are not economical because the cost of ethanol is higher than that of diesel fuel in Turkey, as well as in many of the other countries. It is well-known that ethanol production by actual techniques is generally expensive. However, ethanol may be cheaper in some countries, such as Brazil. If ethanol

Ethanol-Diesel Fuel Blends

Energy & Fuels, Vol. 23, 2009 1715

Figure 8. (a, b) Predicted variations of BSFC as functions of ethanol ratio %.

Figure 9. (a, b) Bar chart diagram of variation ratios of brake effective power and brake effective efficiency at 2100 [rpm] at ε )16.8 for different ethanol ratios %, respectively.

production techniques could be improved and it becomes cheaper in the next years, ethanol blends would probably become economic.

Figure 10. (a, b) Predicted variation of yCO and yH2 as functions of ethanol ratio %.

It is well-known that the most troublesome emission from diesel engines is NO. The mechanism of NO formation is greatly dependent on in-cylinder temperature, oxygen concentration, and residence time for the reaction to take place.33 The predicted variation ratios of the NO concentration at different ethanol ratios are shown in Table 7. The predicted NO concentrations generally increase because combustion temperatures increase slightly at low ethanol ratios, but after 4% ethanol ratios it starts to decrease because of lower temperatures. The mole fraction of CO variation is presented in Figure 10a,b and Table 7 for different ethanol blends. The results show that when the ethanol ratio in the mixture increases, the mole fraction of CO increases until approximately 2-4% ethanol ratios. This trend is because ethanol has less carbon than diesel fuel and ethanol provides lower temperature levels. After these ratios it starts to decrease because of lower flame temperature.3 Higher combustion temperature significantly affects its formation, that is, because higher temperatures increase the dissociation rate of CO2, and thus the formation rate of CO could increase at low ethanol ratios. CO emissions are also primarily effected by the changes in equivalence ratio. So the addition of ethanol which changes the equivalence ratio could change the concentration of CO within the exhaust gases. In the present paper equivalence ratios decrease with ethanol addition. But at low ethanol percentages such as 2% and 4%, decrement ratios of the equivalence ratios are very small, and temperature effects would be dominant than equivalence ratios. At high ethanol ratios, both decrement ratios of temperature and

Table 8. Predicted Variation Ratios of Effective Power, BSFC, Effective Efficiency and Cost at 1600 and 1900 [rpm] for Li’s Engine ε ) 14.5 ethanol [%] 2 4 6

Li’s engine; n ) 1600 [rpm]

Li’s engine; n ) 1900 [rpm]

∆Ne/Ne [%]

∆be/be [%]

∆ηe/ηe [%]

∆C/C1 [%]

∆Ne/Ne [%]

∆be/be [%]

∆ηe/ηe [%]

∆C/C1 [%]

0.775 1.472 1.495

0.387 0.773 1.788

0.345 0.591 0.222

3.243 6.512 10.506

0.939 1.702 2.199

0.233 0.512 1.164

0.434 0.792 0.843

3.085 6.241 9.829

1716 Energy & Fuels, Vol. 23, 2009

S¸ahin and Durgun

Figure 12. Predicted variations of combustion duration as functions of ethanol ratio % at CER.

Figure 13. Predicted variations of temperature as functions of crank angle at CER.

Figure 11. (a, b, c) Predicted variations of BSFC, effective power, and effective efficiency as functions of gasoline ratio % respectively at CER. Table 9. Predicted Variation Ratios of NO Concentration and Mole Fraction of CO at 1600 [rpm] and 1900 [rpm] for Li’s Engine ε ) 14.5

Li’s engine; n ) 1600 [rpm]

Li’s engine; n ) 1900 [rpm]

ethanol [%]

∆NO/NO [%]

∆CO/CO [%]

∆NO/NO [%]

∆CO/CO [%]

2 4 6

37.745 18.973 0.161

-7.547 -10.336 -28.451

28.376 8.261 1.607

-10.800 -13.436 -25.637

equivalence ratios are very high, and thus CO decreases at these ethanol blends. A similar trend has been observed for H2 as shown in Figure 10a,b. As the temperature of the cylinder charge increases, formation of H2 increases up to 2-4% ethanol ratios because the H2O dissociation rate increases at high temperatures.31 The findings about NO and CO emissions are consisted with that of other researchers.2,4

In the present study, the effects of engine speed for various ethanol ratios have also been studied, and it is determined that ethanol blends give better results for engine performance at high engine speeds than that of low engine speeds for used engines. Because the available time for combustion of diesel fuel decreases at high engine speeds, faster mixing rates of diesel fuel with air are required at high speeds. For ethanol blends this could be achieved by faster burning of ethanol and additional gas motions.5,16,17,31 It is well-known that diesel combustion is largely controlled by mixing rates of diesel fuel-air.20 4.2. Ethanol Blends at CER. In this case, injection pressure is adjusted to obtain CER at the selected engine speeds for Li’s engine. Thus, the amount of diesel fuel increases. Also, nozzle cross-section area could be changed to obtained CER. But in the present study, it is preferred to increase injection pressure. In this case, it can be seen from Figure 11a and Table 8 that BSFC increase for this engine, and it is estimates that using ethanol-diesel fuel blends at CER is not economical. In this engine, brake effective efficiency and brake effective power increase slightly with increasing ethanol ratios until 4% and 6%, respectively, at selected engine speeds as shown in Figure 11b,c. Because increasing of the amount of total diesel fuel and ethanol in the cylinder increases, brake effective power and brake effective efficiency. However, at high ethanol ratios they show a decreasing tendency because of the worsened effects of ethanol becoming dominant. Combustion duration increases with increasing ethanol ratios as shown in Figure 12. In this case, larger amount of fuel is burned during the expansion process. Thus, burning a larger amount of the fuel far away TDC results in lower temperatures as shown in Figure 13. Thus, the mole fractions of CO and (34) Ajav, E. A.; Singh, B.; Bhattacharya, T. K. Thermal balance of a single cylinder diesel engine operating on alternative fuels. Energy ConVers. Manage. 2000, 41 (14), 1533–1541.

Ethanol-Diesel Fuel Blends

Energy & Fuels, Vol. 23, 2009 1717

ratios they decrease because of the decreasing cylinder temperature. It was determined that ethanol blends are not economical for these engines because the cost of ethanol is higher than that of diesel fuel in Turkey, as well as in many of the other countries, and the decrease ratio of BSFC is relatively low. Overall Conclusion and Recommendations. (1) Ethanol blends at VER gives better results for engine performance at high engine speeds than that of low engine speeds for these engines because the available time for combustion of diesel fuel decreases at high engine speeds and faster mixing rates of diesel fuel with air are required. This could be achieved by using ethanol blends, thus this forming additional gas motions and faster burning of ethanol. (2) BSFC increases with increasing ethanol ratios at CER. Also, brake effective power, brake effective efficiency, and combustion duration increase until 4-6% ethanol blends at CER, and after these ratios they start to decrease. Mole fractions of CO and H2 exhibit generally decreasing tendencies. However, NO concentration increases at low ethanol ratios, and it starts to decrease at high ratios. Ethanol blends are not economical at CER because of increase of BSFC. (3) It can be said that ethanol-diesel fuel blends at 6% ratio can be used in diesel engines without any modifications; thus, approximately 4% improvement in brake effective power and approximately 2% decrement in brake specific fuel consumption can be achieved. Figure 14. (a, b) Predicted variations of NO concentration and mole fractions of CO and H2 as functions of ethanol ratio % respectively at CER.

H2 decrease (see Figure 14a). As shown in Figure 14b and in Table 9, NO concentration increases at low ethanol ratios although combustion temperature decreases, but it begins to decrease at high ethanol ratios because in this case more amount of diesel fuel exists in the cylinder than that of NDF applications. It is well-known that diesel fuel has a higher tendency to produce NO than ethanol. At the low ethanol ratios the effect of diesel fuel is more dominant than that of high ethanol ratios. On the contrary, NO concentration decreases at high ethanol ratios because of the ethanol effects becoming dominant. 5. Conclusions The main results and recommendations achieved from the present study can be summarized as follows: Main Results. (1) At VER as the ethanol percentage in the mixture increases, BSFC reduces, brake effective efficiency improves significantly, and brake effective power increases slightly. The obtained maximum increase in the variations of brake effective efficiency and brake effective power are at the levels of 5.1% and 2.7%, respectively, for Ottikkutti’s engine. For Li’s engine, the maximum increase in the variations of brake effective efficiency and brake effective power have been calculated as 4.7 and 1.1%, respectively. The obtained maximum decreases in the variations of BSFC are 2.8 and 1.3% for Ottikkutti’s and Li’s engine, respectively. Experiences gained in the developments and applications of the used cycle model show that the most favorable percentage of ethanol is between 4% and 6% in the view of engine performance characteristics for used engines. (2) At VER, equivalence ratio decreases and ignition delay increases for ethanol blends, and combustion duration indicates generally a decreasing tendency for two engines. NO concentration and mole fractions of CO and H2 increase at low ethanol ratios because of the increasing cylinder temperature, but at high ethanol

Nomenclature be ) brake specific fuel consumption [kg/kWh] C ) cost of the fuels [YTL] D ) cylinder bore [m] Fst ) stoichiometric fuel-air ratio [kg-fuel/kg-air] H) stroke of piston [m] QLHV ) lower heating value [kJ/kg] hmin ) stoichiometric air requirement [kg-air/kg-fuel] Ne ) brake effective power [kW] n ) engine speed [rpm] of ) amount of oxygen in the fuel R ) universal gas constant [kJ/kgK] T ) temperature of the contents in the control volume [K] Tw ) wall temperature [K] V ) volume [m3] y ) mole fraction Greek Letters ε ) compression ratio ηe ) brake effective efficiency F ) density [kg/m3] φ ) equivalence ratio x ) volumetric percentage of the fuel ω ) angular speed of the crankshaft [1/s] AbbreViations BSFC ) brake specific fuel consumption CER ) constant equivalence ratio CO ) carbon monoxide DI ) direct injection LHV ) lower heating value NDF ) neat diesel fuel NO ) nitric oxide TDC ) top dead center VER ) varied equivalence ratio Subscripts d ) diesel eth ) ethanol mix ) mixture EF800587E