Fuel 115 (2014) 527–533

Contents lists available at ScienceDirect

Fuel

journal homepage: www.elsevier .com/locate / fuel

Combustion and performance of a diesel engine with preheatedJatropha curcas oil using waste heat from exhaust gas

0016-2361/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.fuel.2013.07.067

⇑ Corresponding author. Tel.: +91 322283161; fax: +91 3222282244E-mail address: hifjur@agfe.iitkgp.ernet.in (H. Raheman).

Priyabrata Pradhan, Hifjur Raheman ⇑, Debasish PadheeAgricultural and Food Engineering Department, Indian Institute of Technology, Kharagpur 721302, India

h i g h l i g h t s

� Improvement in fuel properties by preheating.� Utilization of heat from exhaust gas.� Performance of preheated Jatropha oil vis-à-vis diesel.� Lowered ignition delay for preheated Jatropha oil as compared to diesel.� Lower emissions with preheated Jatropha oil as compared to diesesl.

a r t i c l e i n f o

Article history:Received 11 April 2013Received in revised form 4 June 2013Accepted 18 July 2013Available online 1 August 2013

Keywords:CombustionHeat exchangerPerformancePreheated Jatropha OilReduced emissions

a b s t r a c t

The viscosity and density of CJO (crude Jatropha oil) were reduced by heating it using the heat fromexhaust gas of a diesel engine with an appropriately designed helical coil heat exchanger. Experimentswere conducted to evaluate the combustion characteristics of a DI (direct injection) diesel engine usingPJO (preheated Jatropha oil). It exhibited a marginally higher cylinder gas pressure, rate of pressure riseand heat release rate as compared to HSD (high speed diesel) during the initial stages of combustion forall engine loadings. Ignition delay was shorter for PJO as compared to HSD. The results also indicated thatBSFC (brake specific fuel consumption) and EGT (exhaust gas temperature) increased while BTE (brakethermal efficiency) decreased with PJO as compared to HSD for all engine loadings. The reductions inCO2 (carbon dioxide), HC (hydrocarbon) and NOx (nitrous oxide) emissions were observed for PJO alongwith increased CO (carbon monoxide) emission as compared to those of HSD.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The rapid depletion of conventional fuel and fluctuation of die-sel price in the global market have promoted research for alterna-tive fuels for diesel engine. Among the different alternative fuels,vegetable oil having fuel properties similar to diesel has an accept-able engine performance for short-term operation only [1]. How-ever, long term endurance tests with vegetable oil reported someengine durability issues such as severe engine deposits, piston ringsticking, injector choking, gum formation and lubricating oil thick-ening [2]. These problems are primarily attributed to high viscosityand poor volatility of straight vegetable oils due to large molecularweight and bulky molecular structure. Higher viscosity of vegeta-ble oil (30–200 cSt at 40 �C) as compared to mineral diesel (4 cStat 40 �C) leads to unsuitable pumping and fuel spray characteris-tics. For long running, straight vegetable oils are not suitable asfuels for diesel engines, they have to be modified to bring their

combustion related properties closer to diesel. Undoubtedly,transesterification is well accepted and best suited method of uti-lizing vegetable oils in CI (compression ignition) engine but thisadds extra cost of processing because of the transesterificationreaction involving chemical and process heat inputs. The otheralternative could be use of heated vegetable oils as petroleum fuelsubstitute. Further, heating of oil using exhaust gas from a dieselengine is an attractive proposition.

The viscosity of Jatropha oil was decreased remarkably with in-crease in temperature and it became close to diesel at temperatureabove 75 �C [3]. The density of Jatropha oil was reduced from900.21 kg/m3 to 883.97 kg/m3 by raising the temperature of oilfrom 15 �C to 90 �C [4]. Heating the Jatropha oil between 90 �Cand 100 �C was adequate to bring down the viscosity in close rangeto diesel [5]. Chauhan et al. [6] reduced the viscosity of oil by heat-ing from exhaust gases before feeding to the engine with an appro-priately designed shell and tube heat exchanger with exhaustbypass arrangement. Further, optimal fuel inlet temperature wasfound to be 80 �C considering the BTE and BSEC (brake specificenergy consumption). However, combustion characteristics of

Table 1Fatty acid composition of crude Jatropha curcas oil.

Fattyacid

Systematicname

Formula Structurea wt.%

Oleic cis-9-Octadecenoic

C18H34O2 18:1 34.3–45.8

Linoleic cis-9, cis-12-Octadecadienoic

C18H32O2 18:2 29.0–44.2

Palmitic Hexadecanoic C16H32O2 16:0 14.1–15.3

a xx: y indicates xx carbons in the fatty acid chain with y double bonds; Source:[12].

Table 2Technical specifications of diesel engine.

Particulars Details

Type GF3BMG (TV1)Number of cylinders 1Bore � stroke (mm) 87.5 � 110Cycle 4-strokeMaximum power (kW) 5.5, naturally aspiratedRated speed (rpm) 1500Compression ratio 15.5:1Injection timing (� before TDC) 24Injection type Direct injectionNozzle opening pressure IMEP at 1500 rpm (bar) 5.08

528 P. Pradhan et al. / Fuel 115 (2014) 527–533

Jatropha oil had not been reported. The ignition delay was shorterfor neat rapeseed oil and its blends with diesel as compared to thatof standard diesel. Peak cylinder pressure and maximum heat re-lease rate were decreased with increase in neat rapeseed oil con-tent in blends [7]. Qi et al. [8] reported similar kind of resultsusing Soybean biodiesel and blends with diesel.

In recent years, several attempts have been made to use the es-ters of non-edible oils as substitute for diesel. Hence, it was

1.Load bank 2. Air filter 3. Supply cylinder 4. E

6. Measuring cylinder 7. Engine 8. Fuel filte

11. Fly wheel 12. Data acqui

2

9

11

13

12

10

V

I

1

Fig. 1. Schematic layout of t

decided to choose one such non-edible oil like Jatropha for furtherinvestigation which could provide a suitable substitute for dieselfuel.

Jatropha oil contains higher percentage of oleic acid (34.3–45.8)followed by linoleic acid (29.0–44.2) and palmitic acid (14.1–15.3).The average saturated and unsaturated fatty acids constitute 20.1%and 79.9% of the oil, respectively. The maturity stage of the fruits atthe time of collection is reported to influence the fatty acid compo-sition of the oil [9]. Because of the presence of higher percentage offree fatty acids, it is not desirable to run the diesel engine directlywith Jatropha oil. Therefore an attempt was made to utilize theheat from exhaust gas of diesel engine to reduce the viscosity ofhigh viscous oil to improve its engine performance (see Table 1).

2. Materials and methods

2.1. Experimental setup

A typical engine system widely used in the agricultural sectorwas selected for present experimental investigations and its tech-nical specifications are given in the Table 2. The experimental set-up comprised a constant speed, 5.5 kW, 4-stroke, single cylinder,water cooled, DI diesel engine. The engine was coupled to a threephase, 250 V AC generator. The generator was used for loadingthe engine through an electrical load bank comprising of four heat-ing coils (1500 W, 925 W, 900 W and 875 W) and six electric bulbs(two 200 W, three 100 W and one 40 W). The schematic layout ofthe experimental setup for the present investigation is shown inFig. 1.

The main components of the experimental setup were a dieselfuel tank, measuring cylinders to supply Jatropha oil before heatingand after heating, a helical coil heat exchanger fitted inside the ex-haust gas pipe line for heating Jatropha oil, oil supply line, and per-formance measurement equipments. The exhaust gas flowedthrough the pipe across the helical coil heat exchanger. Helical coil

xhaust gas Analyzer 5. Heat exchanger

r 9. Rotary encoder 10. AC generator

sition system 13. Computer.

3

5

6

8

4

7

he experimental setup.

P. Pradhan et al. / Fuel 115 (2014) 527–533 529

heat exchanger was found to be suitable taking three major factorsinto consideration. Firstly, space is limited. Secondly, under thecondition of laminar flow or low flow rates where a shell and tubeheat exchanger would become uneconomical because of the result-ing low heat transfer coefficients. Thirdly, to increase the exposurearea and residence time for oil.

During the operation of the engine, the crude Jatropha oil wasallowed to flow from the top inlet by gravity and the oil was al-lowed to flow through the coil. The oil traveled the helical pathand thereby gained temperature due to exposure to hot gases withincreased retention time. Finally heated oil came out and was col-lected in a measuring cylinder. The fuel filter was connected to themeasuring cylinder through a valve at the bottom. The engine wasstarted with diesel and once the engine warmed up, it wasswitched over to Jatropha oil. After concluding the tests with Jatro-pha oil, the engine was again switched back to run with diesel be-fore stopping to remove Jatropha oil from the fuel filter and fuelflow line. In all cases, temperature increased with increase in en-gine load. At 80% engine load, exhaust gas temperature beforeand after heat recovery was found to be 310 �C and 226 �C, respec-tively. The oil inlet and oil out let temperature at same load wasfound to be 27 �C and 85 �C, respectively. The temperature of theJatropha oil was maintained within a range of 70 ± 3 �C before itpassed to the fuel filter for all engine loads.

A thermocouple and a temperature indicator were used to mea-sure the exhaust gas temperature. The cylinder gas pressure wasmeasured by a Kistler Model-SN14 piezoelectric pressure trans-ducer mounted on the cylinder head. Crankshaft position and theengine speed were obtained using a rotary Encoder (Model-E50S8) which was connected to the crankshaft.

3. Theory

From the literature, it was revealed that the helical coil is themost suitable heat exchanger to limited space and economic con-straint [10]. Moreover, it increases the residence time and expo-sure area for oil. A copper coil was used as heat exchanger for itshigher thermal conductivity. The number of turns of helical coiland the length of cylinder to accommodate the coil was deter-mined by using theoretical principles for the design of helical coil[11]. The helical coil was designed using Auto CAD and the techni-cal specifications of the heat recovery system are given in Table 3.

3.1. Determination of instantaneous heat release rate

Heat release rate calculations are an attempt to get some infor-mation about the combustion process inside the combustionchamber. Heat release rate (Qn) was calculated from the simplifiedEq. (1) which was derived from 1st law of thermodynamics.

Q n ¼ ½1=ðc� 1Þ�½cPðdV=dhÞ þ VðdP=dhÞ� ð1Þ

where Qn = instantaneous heat release rate (J/�C A), c = ratio ofspecific heats which was taken as 1.35, h = Crank angle (�),

Table 3Technical specifications of the system.

System parameters Dimensions(mm)

Exhaust pipe diameter 55Helix diameter 40Length of helix 210Number of turns 9Pitch of helix 15Coil diameter 6Coil wall thickness 0.9Length of coil 1219

P = instantaneous cylinder gas pressure (Pa), V = instantaneous cyl-inder volume (m3).

3.2. Determination of ignition delay

ID defined as the time interval between start of fuel injection tothe beginning of combustion of injected fuel. Beginning of the fuelcombustion was indicated by the position of the crank in terms ofdegree before TDC where the instantaneous heat release rate be-came positive. The fuel injection timing of the test engine was24� before TDC. Thus from the calculated instantaneous heat re-lease rate value at every crank angle, ignition delay of a test fuelat each load was calculated.

4. Results and discussion

4.1. Fuel properties

The various fuel properties namely density, kinematic viscosity,acid value, flash point of CJO, PJO were determined following theASTM (American society for testing and materials) standards andprocedures and are given in the Table 4. It was also observed thatthe fuel properties of PJO improved by heating.

4.1.1. DensityIt can be seen from Table 2 that densities of CJO and PJO were

found to be about 4.70% and 2.47% higher than that of HSD, respec-tively. The density of CJO was reduced by about 2.13% on heating toa temperature of 70 �C. The higher densities of CJO and PJO as com-pared to HSD may be attributed to the higher molecular weights oftriglyceride molecules present in them.

4.1.2. Kinematic viscosityKinematic viscosity of CJO was found to be 39.97 cSt which was

15.37 times more than that of HSD (Fig. 2). After heating up to70 �C, the kinematic viscosity of CJO was reduced to 20.48 cStwhich was found to be 7.87 times higher than that of HSD. Thiswas due to break down of intermolecular forces and adhesion be-tween molecules. Moreover, the kinematic viscosity of HSD did notvary much with temperature rise. The above results are in confir-mation with the results obtained by Agarwal and Agarwal [5].

4.2. Engine performance

The performance of a DI diesel engine was studied with HSD,CJO and PJO (at 70 ± 3 �C) by varying the engine load by measuringfuel consumption and brake power at governor control range fol-lowing the BIS (Bureau of Indian Standards) procedures. BSFCand BTE were calculated from the measured data.

Table 4Comparison of fuel properties of different fuels.

Fuel type CJO PJO (70 �C) Biodiesel fromJatropha oil

HSD

Density (kg/m3) 890 871 868 850Kinematic viscosity, (cSt) 39.97 20.48 4.80 2.6Acid value (mg KOH/g) 28.0 29.23 0.4 –Flash point (�C) 212 – – 54Calorific value (MJ/kg) 38.65 38.65 39.23 42Carbon (%, w/w) 76.45 76.45 – 85.95Hydrogen (% w/w) 10.45 10.25 – 12.98Nitrogen (% w/w) 3.0 2.95 – 5.15Oxygen (% w/w) 13.12 13.1 – 0.35

Fig. 2. Effect of temperature on kinematic viscosity for Jatropha oil and HSD.

200

300

400

500

600

700

25 50 75 100

BSF

C, g

/kW

h

Engine load, %

HSD

CJO

PJO

Fig. 3. Variations of BSFC for HSD, CJO and PJO at different engine loads.

10

15

20

25

30

25 50 75 100

BT

E, %

Engine load, %

HSD

CJO

PJO

Fig. 4. Variations of BTE for HSD, CJO and PJO at different engine loads.

100

150

200

250

300

350

400

450

0 25 50 75 100

EG

T,

C

Engine load, %

HSD

CJO

PJO

Fig. 5. Variation of EGT for HSD, CJO and PJO at different engine loads.

530 P. Pradhan et al. / Fuel 115 (2014) 527–533

4.2.1. Brake specific fuel consumption (BSFC)BSFC is the ratio between mass flow of tested fuel and effective

power. It can be seen from Fig. 3 that the BSFC reduced with in-crease in engine load for all the fuels tested. At full engine load,the BSFC for HSD, CJO and PJO were found to be 286.88, 319.48and 328.06 g kW�1 h�1, respectively as compared to 569.54,652.15 and 606.74 g kW�1 h�1at 25% engine loading. This wasdue to the higher percentage increase in brake power with increasein engine load as compared to the increase in fuel consumptiondue to relatively less heat losses at higher engine loads.

At 25% engine load, BSFC of CJO and PJO is 14.5% and 6.53%higher than that of HSD. Higher BSFC for CJO might be due to high-er fuel density and viscosity. At full engine load, BSFC of CJO andPJO is 11.36% and 14.35% higher than that of HSD. Higher BSFCfor PJO might be due to lesser ignition delay. The higher fuel con-sumption for CJO and PJO as compared to HSD could be primarilyrelated to the combined effect of higher density and lower energycontent.

4.2.2. Brake thermal efficiency (BTE)Brake thermal efficiency is the ratio of the power output to the

energy supplied through fuel injection. BTE increased with increasein percent load for all the fuels tested as evident from Fig. 4. Themaximum BTE was obtained at full load conditions for all the fuelstested and was 29.88%, 29.15% and 28.33% respectively for HSD,CJO and PJO as compared to 15.05%, 14.28% and 15.32% at 25% en-gine loading. The improved BTE at higher engine load was due tothe reduction in friction loss and increase in brake power with in-crease in percent load. The results obtained are in accordance withChauhan et al. [6] in which the BTE of diesel engine was reported as28.5% and 27.4% when operated with HSD and PJO, respectively.

Though the presence of inbuilt oxygen improved the combus-tion of Jatropha oil, the BTE in general decreased as compared toHSD. This might be due to combine effect of lower calorific valueof fuel and higher fuel consumption. The maximum BTE obtainedfor different fuels as well as HSD was around 29%. At full loadBTE for the CJO and PJO was found to be only 2.44% and 5.18% low-er than that of HSD. This might be due to combined effect of higherBSFC and early combustion.

4.2.3. Exhaust gas temperature (EGT)The temperature of the exhaust gases coming out of the engine

gives an indication of combustion characteristics of the fuel used.The EGT increased with increase in engine load for all the fuelstested (Fig. 5). The maximum EGT was obtained at full load condi-tion for all the fuels tested and was 375, 420 and 418 �C for HSD,CJO and PJO, respectively as compared to 137, 134 and 139 �C atno load condition. The increase in EGT with engine load wasmainly due to increase in the amount of energy released at higherloads because of the burning of increased amount of fuel whichwas injected to meet the extra power requirement to take up theadditional loading; hence more heat rejection to the exhaust gases.Beyond 50% engine load, higher EGT with PJO and CJO compared toHSD was due to lower BTE of the engine. The above results are inagreement with Pramanik [3] in which he has reported higher ex-haust temperature for Jatropha oil.

The difference in performance parameters among HSD, CJO andPJO are significant at 5% level of significance indicating the influ-ence of different fuels on BSFC, BTE and EGT.

4.3. Combustion characteristics

The combustion characteristics of the CJO and PJO were com-pared with that of HSD in terms of CGP (cylinder gas pressure),

(a) No load

(b) Full load

-2

-1

0

1

2

3

4

5

6

-30 -20 -10 0 10 20 30

RO

PR, B

ar/

CA

CA, degree

HSD

CJO

PJO

-4

-2

0

2

4

6

8

10

12

14

-30 -20 -10 0 10 20 30

RO

PR, B

ar/

CA

CA, degree

HSD

CJO

PJO

Fig. 7. Variations of ROPR with respect to crank angle for HSD, CJO and PJO.

P. Pradhan et al. / Fuel 115 (2014) 527–533 531

ROPR (rate of pressure rise), HRR (heat release rate) and ID (igni-tion delay). The variations of CGP, ROPR, HRR and ID with respectto engine loading for the above mentioned fuels are discussed inthe following sections:

4.3.1. Cylinder gas pressure (CGP)The CGP characterizes the ability of the fuel to mix well with air

and burn. After starting of combustion, pressure rose rapidly due tothe expansion of the cylinder contents and reached to a peak fewdegrees before TDC (top dead center). Then it decreased graduallyduring the expansion stroke. The variations of CGP with crank an-gle rotation for HSD, PJO and CJO at no load and full engine load aregiven in Fig. 6a and b, respectively. It can also be seen from thesefigures that the peak CGP occurred earlier for PJO at lower engineloads. At no load condition, the peak CGP occurred at �1�, 2� and0 �C A after TDC, respectively for HSD, CJO and PJO as comparedto �1�, 1� and 1 �C A after TDC, at full engine load. The peak CGPobtained at no load condition was 52.84, 47.86 and 53.25 bar,respectively for HSD, CJO and PJO whereas, it was 73.65, 66.75and 73.57 bar at full engine load. Since, the quantity of fuel burnedincreased with increase in engine load, it caused an increase in theheat energy released which resulted in an increase in peak CGP.

The pressure rise due to combustion started a little earlier forPJO than HSD. This was mainly attributed to the earlier initiationof combustion for PJO. The variations of CGP among HSD, CJOand PJO were found to be insignificant in the later phase of com-bustion. However, the peak CGP reduced for CJO compared toHSD and PJO at all loads. This might be due to poor atomizationof fuel inside the engine cylinder.

4.3.2. Rate of pressure rise (ROPR)ROPR indicates the smoothness of combustion in the engine

combustion chamber. The variations of ROPR with respect to crankangle for HSD, CJO and PJO at no load and full engine load are given

(a) No load

(b) Full load

0

10

20

30

40

50

60

-30 -20 -10 0 10 20 30

CG

P, B

ar

CA, degree

HSD

CJO

PJO

0

10

20

30

40

50

60

70

80

-30 -20 -10 0 10 20 30

CG

P, B

ar

CA, degree

HSD

CJO

PJO

Fig. 6. Variations of CGP with respect to crank angle for HSD, CJO and PJO.

in Fig. 7a and b respectively. It can be observed from these figuresthat for all engine loadings and fuels tested, during the compres-sion stroke, ROPR increased initially due to the expansion forceof the cylinder content as a result of increased temperature. Fewdegrees after fuel injection (24� before TDC), evaporation of the in-jected fuel reduced the cylinder temperature. As a result of this,pressure increased but at a slower rate. After combustion, pressurerose rapidly and ROPR reached a peak value, then ROPR reduced.Finally CGP started reducing during expansion stroke as indicatedby the negative values of ROPR. The peak ROPR for HSD, CJO andPJO occurred at 6 �C A, 5 �C A, and 7 �C A before TDC respectivelyat no engine load as compared to 8 �C A, 9 �C A and 10 �C A beforeTDC, at full engine load. This early occurrence of peak HRR at high-er engine load was due to the decrease in ignition delay with in-crease in engine load. The ID of fuel decreased with increase inengine load resulting in early occurrence of peak HRR which ledto early occurrence of peak ROPR.

The peak ROPR obtained at no load conditions was 4.68, 2.30and 4.84 bar/�C A respectively for HSD, CJO and PJO whereas, itwas 13.09, 8.08 and 12.68 bar/�C A at 100% engine loading. It wasalso observed that the peak ROPR occurred earlier for PJO thanHSD due to early start of combustion for PJO.

4.3.3. Heat release rate (HRR)The HRR curve shows the potential availability of heat energy

which can be converted into useful work. The variations of HRRwith respect to crank angle for HSD, CJO and PJO at no load and fullengine load are presented in Fig. 8a and b respectively. It can beseen from Fig. 8a and b that HRR during the initial stages of com-bustion increased with engine load for all the fuels tested owing tothe increase in the quantity of fuel injected in the combustionchamber. It was found that the peak HRR occurred earlier at higherengine loads for all fuels tested. The peak HRR at no load occurredat 6 �C A, 5 �C A and 7 �C A before TDC, respectively for HSD, CJOand PJO whereas, at full engine load it occurred at 8 �C A, 9 �C A

(a) No load

(b) Full load

-10

0

10

20

30

40

50

-40 -30 -20 -10 0 10 20H

RR

, J

/C

A

CA, degree

HSD

CJO

PJO

-40

-20

0

20

40

60

80

100

120

140

160

-30 -20 -10 0 10

HR

R ,

J/

CA

CA, degree

HSDCJOPJO

Fig. 8. Variations of HRR with respect to crank angle for HSD, CJO and PJO.

Fig. 9. Variations of ID with respect to crank angle for HSD, CJO and PJO.

Table 5Comparison of emission constituents for different fuels at different loads.

Load (%) 0 25 50 75 100

CO (%) HSD 0.106 0.081 0.044 0.037 0.066CJO 0.221 0.132 0.116 0.182 0.226PJO 0.113 0.097 0.075 0.080 0.255

CO2 (%) HSD 3.28 4.26 5.64 7.30 8.68CJO 2.94 4.06 5.54 7.21 8.98PJO 3.17 4.17 5.04 6.72 8.52

HC (ppm) HSD 14 11 17 25 34CJO 31 17 11 19 65PJO 12 17 12 14 43

NOx (ppm) HSD 206 496 997 1670 1827CJO 43 222 887 1273 1302PJO 131 362 711 975 1084

532 P. Pradhan et al. / Fuel 115 (2014) 527–533

and 10 �C A before TDC. As the engine load increased, the ignitiondelay decreased, resulting in early start of combustion. Hence,the HRR rose to the peak more quickly. The peak HRR for PJO oc-curred earlier than that of HSD. Due to longer ignition delay, mostof the injected fuel burned in the later phase of combustion forHSD and CJO at initial loading.

4.3.4. Ignition delay (ID)The variations of ID for different fuels are compared at different

engine loadings in Fig. 9. From this figure it can be seen that thedelay period for all the fuels tested decreased with increase in en-gine load. The ignition delays for HSD, CJO and PJO were calculatedto be 15 �C A, 17 �C A and 14 �C A, respectively at no load condition,whereas, the delays were 13 �C A, 12 �C A and 11 �C A at 100% en-gine loading. This was due to elevated temperature at higher load,which improved the fuel vaporization process and reduced the

chemical delay and hence the overall ignition delay period. Theelevated temperature existing in the combustion chamber at high-er engine loads enhanced the fuel vaporization process and re-duced the chemical delay and hence the overall ignition delayperiod. As the engine load decreased, the residual gas temperatureand wall temperature decreased, which resulted in lower chargetemperature at the time of fuel injection and hence lengtheningthe ignition delay period.

The early start of combustion for PJO could be due to a complexand rapid pre-flame chemical reaction taking place at higher tem-peratures. Therefore, using PJO reduced the ID period as comparedto HSD. But in case of CJO, the ID was higher at initial loads com-pared to HSD. It might be due to higher viscosity of oil and lowerengine temperature.

4.4. Emissions parameters

Constituents of emissions such as carbon monoxide, carbondioxide, unburned hydrocarbon and oxides of nitrogen were re-corded with the help of an online exhaust gas analyzer and are pre-sented in Table 5. The detailed discussions are made in thefollowing sections:

4.4.1. Emissions of carbon monoxide (CO)Carbon monoxide is generated in an engine as a product of

incomplete combustion of the fuel. The values of CO emission de-creased from 0.106% at no load conditions down to 0.037% at 75%load for HSD. The CO emission in general was found to be increasedfor CJO and PJO compared to HSD at any engine load tested. Theseresults are in line with the results obtained by Chauhan et al. [6].Further PJO showed better combustion results compared to CJO ex-cept at 100% load. Among the different fuels tested, CJO gave themaximum CO emission and it decreased with heating. This wasdue to the better atomization of fuel, which led to relatively bettercombustion of the fuel resulting in lower CO emission.

4.4.2. Emissions of carbon dioxide (CO2)The maximum CO2 was obtained at full engine load and was

8.68%, 8.98% and 8.52% respectively for HSD, CJO and PJO whereas,it was 3.28%, 2.94% and 3.17% at no load. At elevated temperature,performance of the engine improved with relatively better burningof the fuel resulting in higher CO2 emission. At full load, CO2 emis-sion for CJO was 3.45% higher than that of HSD and for PJO it was1.84% lower than that of HSD. The above results are in accordancewith results obtained by Agarwal and Agarwal [5].

4.4.3. Emissions of hydrocarbons (HC)Hydrocarbon emissions from CI engine are the direct results of

the non-homogeneity of fuel–air mixture in the combustion cham-

P. Pradhan et al. / Fuel 115 (2014) 527–533 533

ber. The value of HC emissions for HSD increased from 14 ppm atno load to 34 ppm at 100% engine load. The increase in HC is dueto decrease in air–fuel ratio. HC emissions were found to be lowerfor CJO and PJO as compared to HSD at 50 and 75% of engine load.Further, the HC emission for PJO was lower than that of CJO both atlower and higher loading conditions. The reduction in HC emis-sions for PJO might be due to better atomization of fuel molecule,which led to a more complete and cleaner combustion.

4.4.4. Emissions of nitrogen (NOx)Availability of oxygen, higher temperature and combustion

duration are the three main factors which facilitate the productionof NOx. The maximum NOx was obtained at full load conditions forall the fuels tested and was 1827, 1302 and 1084 ppm, respectivelyfor HSD, CJO and PJO as compared to 206, 43 and 131 ppm at noload condition. The NOx emission in general was found to be lowerfor CJO and PJO than that for HSD. This might be due to the supplyof lean mixture in case of HSD. Among all the fuels tested, PJO pro-duced minimum NOx at 50%, 75% and 100% load. The low air–fuelratio and instantaneous chemical reaction in case of PJO promptedlower NOx formation.

5. Conclusion

Based on the experimental results of this study, it can be con-cluded that

1. Both utilization of heat from the exhaust gas and improvementof fuel properties were possible by heating the CJO with exhaustgas using a suitable designed helical tube heat exchanger.

2. The BSFC decreased while BTE increased with increase in engineload for all fuels tested. The maximum BTE obtained for HSD,CJO and PJO was around 29%. At full load, BTE for the CJO andPJO was found to be only 2.44% and 5.18% lower than that ofHSD.

3. PJO exhibited a marginally higher CGP, ROPR and HRR as com-pared to HSD and CJO during the initial stages of combustion forall engine loadings. The peak CGP, peak ROPR and peak HRR forPJO occurred earlier than those with HSD.

4. The ignition delay period for all the fuels tested decreased withincrease in engine load. Using PJO reduces the ID period as com-pared to HSD, but in case of CJO, the ID was higher at initialloads compared to HSD.

5. The exhaust emissions such as CO2, HC and NOx from the enginewhen operated with PJO were reduced on an average by 5.28%,2.67% and 37.2%, respectively as compared to the emissionswhen operated with HSD, whereas CO emissions wereincreased on an average by 85.63%.

On the whole it can be concluded that the designed heat ex-changer could successfully preheated the crude Jatropha oil usingheat from exhaust gas. The fuel properties were improved by pre-heating and it can be used in the diesel engines without any mod-ification as a substitute for diesel.

Acknowledgement

The authors acknowledge the financial support made byDepartment of Science and Technology, Govt. of India, New Delhifor carrying out this research work.

References

[1] Schlick ML, Hanna MA, Schinstock JL. Soybean and sunflower oil performancein a diesel engine. Trans ASAE 1988;31(5):1345–9.

[2] Geottler HJM, Ziejewski M, Knudson A.M. Performance of a diesel engineoperating on blends of diesel fuel and crude sunflower oil at normal andelevated fuel temperature. SAE 1985: Paper No. 852087.

[3] Pramanik K. Properties and use of Jatropha curcas oil and diesel fuel blends incompression ignition engine. Renew Energy 2003;28:239–48.

[4] Veny H, Baroutian S, Aroua MK, Hasan M, Raman AA, Sulaiman NMN. Densityof Jatropha curcas seed oil and its methyl esters: Measurement andestimations. Int J Thermophys 2009;10(1):1007–16.

[5] Agarwal D, Agarwal AK. Performance and emissions characteristics of Jatrophaoil in a direct injection compression ignition engine. Appl Therm Eng2007;27(13):2314–23.

[6] Chauhan BS, Kumar N, Jun YD, Lee KB. Performance and emission study ofpreheated Jatropha oil on medium capacity diesel engine. Energy2010;35:2484–92.

[7] Buyukkaya E. Effects of biodiesel on a DI diesel engine performance, emissionand combustion characteristics. Fuel 2010;89(10):3099–105.

[8] Qi DH, Chen H, Geng LM, Bian YZH. Experimental studies on the combustioncharacteristics and performance of a direct injection engine fueled withbiodiesel/diesel blends. Energ Convers Manage 2010;51:2985–92.

[9] Raina AK, Gaikwad BR. Chemobotany of Jatropha species in India and furthercharacterisation of curcas oil. J Oil Technol India 1987;1987(19):81–5.

[10] Prabhanjan DG, Ragbavan GSV, Kennic TJ. Comparison of heat transfer ratesbetween a straight tube heat exchanger and a helically coiled heat exchanger.Int J Heat Mass Trans 2002;29(2):185–91.

[11] Patil RK, Shendhe BW, Ghose PK. Designing a helical coil heat exchanger. JChem Eng 1982:85–8. December 13.

[12] Gubitz GM, Mittelbach M, Trabi M. Exploitation of tropical seed plant Jatrophacurcas L. Bioresour Technol 1999;67(1):73–82.