9

Click here to load reader

Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

Embed Size (px)

Citation preview

Page 1: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

Applied Energy 100 (2012) 249–257

Contents lists available at SciVerse ScienceDirect

Applied Energy

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

Assessment of liquid fuel (bio-oil) production from waste fish fat and utilizationin diesel engine

Edwin Geo Varuvel a,⇑, Nadia Mrad a, Mohand Tazerout a, Fethi Aloui b

a École des Mines de Nantes, Département Systèmes Energétiques et Environnement (DSEE), GEPEA, CNRS-UMR 6144, 4 rue Alfred Kastler, BP 20722, 44307 Nantes Cedex 03, Franceb Université de Valenciennes, ENSIAME, TEMPO – EA4542 (DF2T), Le Mont Houy, F-59300 Valenciennes Cedex 9, France

h i g h l i g h t s

" To derive the bio-oil from industrial waste fish fat by catalytic cracking." To find the suitability of bio-oil for diesel engines through combustion, emission and performance analysis." To compare the results of neat bio-oil with diesel, bio-oil/diesel blends, fish oil and bio-oil UD.

a r t i c l e i n f o

Article history:Received 31 January 2012Received in revised form 4 May 2012Accepted 7 May 2012Available online 24 July 2012

Keywords:Industrial waste fish fatCatalytic crackingBio-oilDistillation and diesel engine

0306-2619/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.apenergy.2012.05.035

⇑ Corresponding author.E-mail address: [email protected] (E.G. Varuvel)

a b s t r a c t

Increased acceptance of climate change induced by human activities and raising oil demand with unse-cure deliverance compels the searching for alternative fuels. The problems with environmental degrada-tion due to industrial wastes can be reduced by converting some of them into bio-oil. In the present work,the waste from fish processing industry is converted to bio-oil by catalytic cracking. Experiments wereconducted in a direct injection diesel engine of 4.5 kW at 1500 rpm. The different test fuels of diesel, fishoil at 75 �C, bio-oil UD (undistilled bio-oil), B20D80 (20% bio-oil in fossil diesel), B80D20 (80% bio-oil infossil diesel) and neat bio-oil were tested to assess the suitability in diesel engines through combustion,emission and performance characteristics. Experimental results show that the brake thermal efficiency ismarginally higher with neat bio-oil over other test fuels. It is lower with preheated fish oil and it is almostsame for both bio-oil and bio-oil UD. NOx, HC, CO and PM emissions are higher with bio-oil UD comparedto bio-oil. PM, CO and HC emissions are lower with bio-oil over diesel. NOx emissions are lower with bio-oil compared to bio-oil UD but it is still higher than diesel fuel. Addition of diesel with bio-oil reduces theNOx emissions marginally. Intensity of premixed combustion is strong with bio-oil. Ignition delay andcombustion duration are reduced with bio-oil due to high cetane number and oxygen concentration.Bio-oil from waste fish fat by catalytic cracking can be used as a fuel for diesel engines and also the wasteto energy may reduce the environmental and climate change issues due to industrial wastes.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Road transport has become an intrinsic feature of our life style.It is generally recognized that transport activities are a key factorto social, regional and economic cohesion [1] and Europe has lar-gely benefited from affordable and convenient road transportmodes. Today the European Union (EU) road transportation mainlydepends on liquid fuels from fossil fuels. It accounts that more than30% of the total energy consumed for road transport. The EU esti-mates that 98% of the fuel used for transport is based on oil andthat 70% of that oil is imported from non-EU countries. The EU doesnot have the natural resources needed to cover its domestic de-

ll rights reserved.

.

mand for energy. On the other side, road transport’s major impactson human health and the environment are a challenge for Europe’ssustainable development. According to Black [2], sustainable trans-port is defined as transportation that meets the current needswithout compromising the ability of future generations to meettheir own needs. The major obstacles to achieve sustainable trans-port include security of energy supply due to depletion of fossilfuels and emissions of greenhouse gases, combined with long-standing problems of congestion, noise and pollution.

In order to address the described challenges, the EU has decidedto include transport fuels produced from renewable energy sourcewhich is called biofuel as part of its policy strategy towards sus-tainable development [3]. The use of biomass for clean energy gen-eration in the EU could be significantly increased nowadayswithout harming biodiversity, soil and water resources. Also, the

Page 2: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

Table 1Main properties of the waste fish fat.

Properties Waste fish fat

Visual aspect Liquid at 60 �C, brown color, typical smellWater content (%) <0.05%Flash point (�C) 318High heating value (kJ/kg) 39,000Kinematic viscosity (mm2/s) 27Density(kg/m3) 893

Composition of fatty acids (%)Mysteric 1.05Palmitoleic 5.00Palmitic 16.00Stearic 10.50Oleic 45.60Linoleic 20.60

250 E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257

growing interest on biofuel from biomass waste will provide great-er access to clean liquid fuels which would address energy costs,energy security and global warming concerns associated withpetroleum fuels.

An increasing amount of waste has become a second generationenergy resources in Europe, there are several researchers engagedin developing sustainable systems of treating the waste with en-ergy recovery [4–7]. Today all wastes from the municipalities areput on landfills and only a small amount is recycled [8,9]. The con-version of fat as a fuel for diesel engines appears as an attractivesolution to enhance energy reserves of biomass. To make proper-ties of oils and fats similar to diesel, various methods have beenproposed [10–13]. The liquid fuel produced from these methodscan be used in diesel engines or blended with diesel since the prop-erties of bio-oil is very close to diesel [14–18]. The main objectivesof waste to energy are: (i) reduction of the amount of waste that isdeposited on landfill, (ii) generation of energy from a new sourceand (iii) reduction of emissions that contribute to the greenhouseeffect.

More recently, an ambitious target was adopted by the Euro-pean Council in the context of the Energy Package proposed bythe European Commission. A 10% minimum target has been setfor the share of overall petrol and diesel consumption coming frombio-oils in EU transport by 2020 and 25% by 2030 – provided thatsuch biofuel production is sustainable and that second-generationbio-oils become commercially available [19].

2. Present work

In the present investigation, waste fish fat (WFF) has been con-sidered as a potential feedstock for the preparation of bio-oil. Bio-oil was prepared by catalytic cracking and the operating conditionswere optimized with respect to liquid fuel yield and acid value. Thequality of bio-oil was improved through distillation process. Ini-tially experiments were conducted with diesel and fish oil at75 �C (preheated fish oil) in 4.5 kW diesel engine for base line dataat various loading conditions. Then, tests were conducted with bio-oil UD (undistilled bio-oil), B20D80 (20% bio-oil in fossil diesel),B80D20 (80% bio-oil in fossil diesel) and neat bio-oil to analyzethe combustion, emission and performance characteristics. Finally,a comparative study of the combustion, emission and performancecharacteristics was made with all the fuels tested at 80% of full load(i.e. at maximum efficiency point).

3. Bio-oil preparation and analysis

3.1. Preparation of bio-oil

The feedstock used in this work was a fish processing industrywaste and it was obtained from, SIRH group specialized in vegeta-ble, animal and marine oils located in north of France. This waste isthe residue of marine oil treatment which is brown in color andwas used without any special purification treatment. The typicalfatty acid composition has been analyzed by gas chromatographyanalysis (GC/FID). The saturated acids C14:0, C16:0 and C18:0 areidentified. The major fatty acids found are the unsaturated acidsC18:1 and C18:2 responsible for 45.6% and 20.6% respectively ofthe total composition. The important properties of waste fish fatare given in Table 1.

To prepare the bio-oil, the catalytic cracking experiments werecarried out at temperatures ranging from 350 �C to 480 �C withslow heating rate of 2–3 �C/min using a laboratory scale reactor,which is shown in Fig. 1. The fat was introduced in the reactorand then heated by an external electrical resistance. The catalystwas placed just above the fat on a bed with small holes. When

the temperature inside the reactor reaches 350 �C, the reactionstarts. The generated vapor was passed directly over the catalystsurface, before leaving through the top of the reactor. Then it en-ters in a water-cooled, counter flow, heat ex-change which waskept at 15 �C. As a result, three fractions of liquids were collectedin the flask: the first is the pyrolysis water, and the second liquidfraction recovered until the temperature reaches 400 �C. The lastfraction is the bio-oil recovered from 400 �C to 480 �C. After acidityanalysis, it was found that the second fraction was more acidic(acid value equal to 20 mg KOH/g) compared to the third fraction(acid value equal to 0.8 mg KOH/g). Hence, the main interest wasfocused on third fraction of liquid fuel called bio-oil.

3.2. Distillation of bio-oil

The flash point of bio-oil is very low (27 �C) compared to thediesel fuel (56 �C). The volatile compounds reflects low flash pointof bio-oil. The use of bio-oil UD (undistilled bio-oil) in diesel en-gines can cause damage to the engine components like injectionsystem, filters due to corrosion and mechanical wear. A simpleand highly effective distillation to extract the most volatile com-pound was conducted with a flash point of bio-oil relatively closeto that of diesel. The distillation was conducted at temperaturesbetween 130 �C and 140 �C and the mass recovered of the volatilehydrocarbon represent 5% of the total mass of bio-oil. After GC/MS(Gas Chromatography–Mass Spectrometry) analysis, it was foundthat the organic compounds present in the bio-oil were dividedinto eight classes: alkene, alcane, cycloalcane, cetone, aromaticcompound, chlorinated compound, phenol, and ester. The majorcomponents were alkene, alcane, and cetone.

3.3. Analysis of bio-oil

The flash point was measured by NPM 440 model (PENSKY-MARTENS). The acid value was determined by titration withKOH/C2H5OH solution using phenolphthalein as an indicator. Thedensity of the bio-oil was estimated with a pycnometer. The grossheating value was measured using an oxygen bomb calorimeter(model 6200, Parr Instruments Company). To identify the dynamicviscosity, a SV 10 Fibro viscometer was used. The elemental com-positions of the main organic elements (C, O, H, S, and N) weredetermined using an Elemental Analyzer (Flash EA 1112, CE Instru-ments). In order to estimate the chemical compounds of the bio-oil, a GC/MS analysis was carried out. For this purpose a Perkin El-mer TurboMass Gold Mass Spectrometer coupled with a gas chro-matograph CLARUS 500 was used. The column was SBLTM-5 msCapillary type, 30 m in length and 0.25 mm in internal diameter.

Page 3: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

Fig. 1. Schematic diagram of the continuous catalytic cracking unit.

Table 2Properties of different test fuels.

Properties Diesel fuel Bio-oil Undistilled bio-oil B80D20 B20D80 Fish oil

Flash point (�C) 56 57 27 57 56 250Acidity (mgKOH/g) – 0.8 0.2 – – 2HHV (MJ/kg) 45.71 45.10 45.25 45.15 45.22 37.25Dynamic viscosity at 20 �C (Ns/m2) 2.52 2.32 2.11 2.36 2.48 –Density at 20 �C (kg/m3) 830 825 818 826 829 886Kinematic viscosity at 20 �C (mm2/s) 2 1.7 1.72 1.76 1.94 28Auto ignition temperature (�C) 220 230 – 228 223 –Cetane number 52 57 – – – –Pour point (�C) – �5 �5 – – –

E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257 251

The properties of fossil diesel fuel, fish oil, bio-oil UD, bio-oil andits blends are given in Table 2.

Table 3Specifications of test engine.

Make Lister petter

No of cylinders OneType of cooling Air cooledBore � stroke 95.5 � 88.94 mmLength of connecting rod 165.3 mmDisplacement 630 cm3

Fuel injection timing 20� bTDCFuel injection pressure 250 barCompression ratio 18:1Rated power 4.5 kW @1500 rpm

4. Experimental setup

4.1. Performance instrumentation

A single cylinder, four-stroke, air cooled, direct injection, con-stant speed, diesel engine developing power output of 4.5 kW at1500 rpm was used for this work. The important specifications oftest engine are given in Table 3. The fuel injection timing and pres-sure were maintained at 20 �CA (crank angle) bTDC (before topdead centre) and 250 bar respectively. The engine was coupledwith an eddy current dynamometer for the torque measurement.The flow of intake air was measured by a differential pressuretransmitter; type LPX 5481. For temperature measurements, thetest engine was equipped with a series of thermocouples type K.Ambient temperature was measured by an active transmitter for

humidity and temperature, type HD 2012 TC/150. The fuel flowwas measured using a Coriolis mass flow meter. Torque measure-ment was made using a force sensor used in tension and compres-sion of the FN3148 series. It has an accuracy of 0.05% of range ofthe measure.

Page 4: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

Fig. 2. Schematic diagram of the engine setup.

252 E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257

4.2. Emission instrumentation

An exhaust gas analyzer was placed in the exhaust system tomeasure the main pollutants. Emissions of hydrocarbons (HC)were measured by FID flame ionization using a heated hydrocar-bon analyzer (model GRAPHITE 52 M), emissions of nitric oxide(NO) and nitrogen oxides (NOx) were measured using a chemilu-minescence nitrogen oxide analyzer TOPAZE 32 M. Emissions ofcarbon monoxide (CO) and carbon dioxide (CO2) were measuredby absorption of infrared radiation using a 2 M MIR analyzer. Par-ticulate emissions were measured using a dust analyzer in realtime (TEOM model 1105), for measurement and continuous weigh-ing of the mass concentration of particulate exhaust.

4.3. Combustion instrumentation

A control system was installed to measure high-frequency sig-nal, which mainly concern the cylinder pressure, fuel injectionpressure and also the angular position of the crankshaft. The pres-sure in the cylinder was measured at a frequency of 90 kHz using apiezoelectric pressure sensor, water cooled, type AVL QH32D. Theinjection pressure was measured by a piezoelectric pressure trans-ducer, type AVL QH33D, located in between the injection pumpand the fuel injector. The angular position of the crankshaft was

measured by an encoder, type AVL 364C, placed on the flywheel.The schematic of test setup is shown in Fig. 2.

5. Results and discussion

5.1. Analysis of combustion parameters

The variation of cylinder peak pressure for diesel, fish oil at75 �C, bio-oil UD, B20D80, B80D20 and bio-oil at 80% of full loadis shown in Fig. 3. The cylinder peak pressure depends stronglyon the initial combustion rate in diesel engines, which in turn de-pends on the amount of fuel takes part in the premixed combus-tion phase. It can be seen that preheated fish oil has slightlylower peak pressure than diesel due to low cetane number andhigh viscosity. The addition of bio-oil in the blend increases thecylinder pressure. The addition of bio-oil with diesel increasesthe oxygen concentration in the blend due to that increased cylin-der pressure with blends. This enhances combustion rate on ac-count of the flame propagation through the bio-oil addition. Thecylinder peak pressure with neat bio-oil is very high compared toall other test fuels. The fuel inside the combustion chamber reactswith oxygen and releases energy which is used to increase the cyl-inder pressure. The cylinder pressure decreases if diesel concentra-tion in the blend increases. This decrease in cylinder pressure with

Page 5: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

40

50

60

70

80

90

Diesel Fish Oil at 75 °C

Bio-Oil UD

B20D80 B80D20 Bio-Oil

82,3 81,884,7

82,5 83,9 84,4

Cyl

inde

r P

eak

Pre

ssur

e (b

ar)

Fig. 3. Variation of cylinder peak pressure at 80% of full load.

-5

5

15

25

35

45

55

65

-25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50

Crank angle (degree)

Bio-Oil

B80D20

B20D80

Bio-Oil UD

Fish Oil at 75 °C

Diesel

Rat

e of

hea

t re

leas

e (J

/ C

A)

°

Fig. 5. Variation of heat release rate at 80% of full load.

E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257 253

neat diesel and different blends is due to lack of oxygen concentra-tion during combustion which reduces the rate of combustion. Itcan be also seen that the cylinder peak pressure for both bio-oilUD and bio-oil is almost same. The slight increase of peak pressureof bio-oil UD is due to its early ignition as a result of highly volatilecomponents of bio-oil UD with low flash point.

Fig. 4 clearly shows the cylinder pressure – crank angle diagramfor various test fuels at 80% of full load. The injected fuel is beingmixed with air which attains the auto ignition temperature; thecombustion will start and releases energy. This energy is utilizedto increase the in-cylinder pressure. From this, the occurrence ofmaximum pressure is mainly depends on fuel mixing rate. It canbe seen that the maximum peak pressure is shifted by about5.7 �CA (crank angle) with fish oil at 75 �C and 5.3 �CA with dieselaTDC (after top dead centre). The peak pressure of preheated fishoil and diesel is shifted due to longer ignition delay compared tobio-oil, which may delay the start of combustion process. Theoccurrence of maximum pressure for B20D80 and B80D20 is5.1 �CA aTDC and 4.9 �CA aTDC. The shift of occurrence of peakpressure shows the direct reflection on heat release. The maximumpeak pressure occurred very early with bio-oil and bio-oil UD. Itcan be seen that the maximum peak pressure is shifted by about4.7 �CA with bio-oil UD and 4.5 �CA with bio-oil after TDC. Thisshows that the heat energy is released very early with bio-oildue to that the brake thermal efficiency is increased and exhaustgas temperature is reduced.

The rate of heat release with crank angle for diesel, fish oil at75 �C, bio-oil UD, B20D80, B80D20 and bio-oil at 80% of full loadis shown in Fig. 5. It is very clear that premixed combustion phaseof preheated fish oil is significantly lower in comparison with bio-oil and diesel. This is due to the high viscosity of preheated fish oil

0

20

40

60

80

100

-100 -80 -60 -40 -20 0 20 40 60 80 100

Cyl

inde

r P

ress

ure

(bar

)

Crank Angle (degree)

Bio-Oil

B80D20

B20D80

Bio-Oil UD

Fish Oil at 75 °C

Diesel

Fig. 4. Variation of cylinder pressure at 80% of full load.

leading to a reduction in air entrainment and fuel air mixing rates.This results in lesser fuel being prepared for rapid combustion withpreheated fish oil. The diffusion burning, indicated by the secondpeak is higher with fish oil due to the burning of more quantityof fuel in the later part of combustion which lead to lower thebrake thermal efficiency and higher exhaust gas temperature.The premixed combustion with bio-oil is very strong and the diffu-sion combustion is low. Even though the ignition delay of neat bio-oil is less, the premixed combustion rate is quite high. This indi-cates that most of the fuel injected during the delay period is pre-pared for ignition and it is participated during premixedcombustion process. Higher oxygen concentration and cetanenumber of bio-oil increase the rate of combustion. It also showsthat the neat diesel displays lower premixed peak and it increasesby blending with bio-oil. Also, the intensity of diffusion phase ofcombustion showed by the area under the second peak in the heatrelease diagram is higher for neat diesel followed by B20D80 andthen B80D20 compared to bio-oil. At the time of ignition, theamount of fuel–air mixture is prepared with blends and dieselfor combustion is less which results less fuel burned during pre-mixed combustion and more fuel burned at diffusion combustion.The premixed combustion is very high with bio-oil UD but it is notwider compared to neat bio-oil. It is mainly due to the low flashpoint of bio-oil UD which has more volatile components and startsthe ignition early but less fuel is prepared for the premixedcombustion.

Fig. 6 shows the variation of ignition delay for different fuels at80% of full load. Ignition delay provides an indication of the com-bustibility of the fuel injected and mixing of fuel and air. It is longerwith preheated fish oil compared to all other test fuels. This ismainly because of high viscous fuel has larger droplet diameters

0

2

4

6

8

10

12

14

Diesel Fish Oil at 75 C

Bio-Oil UD

B20D80 B80D20 Bio-Oil

11,312,2

10,1

11,510,8 10,7

Igni

tion

Dea

ly (

° CA

)

°

Fig. 6. Variation of ignition delay at 80% of full load.

Page 6: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

33

34

35

36

37

38

39

40

41

42

43

Diesel Fish Oil at 75 C

Bio-Oil UD

B20D80 B80D20 Bio-Oil

39,2

42,8

37,2

39,1

3736,6

Com

bust

ion

Dur

atio

n (

CA

°

Fig. 7. Variation of combustion duration at 80% of full load.

25

125

225

325

425

525

625

725

825

925

1025

Diesel Fish Oil at 75

Bio-Oil UD

B20D80 B80D20 Bio-Oil

852817

1001

866 889 917

Oxi

des

of N

itro

gen

(ppm

)

°C

Fig. 8. Variation of oxides of nitrogen at 80% of full load.

25

75

125

175

225

275

325

375

425

475

525

Diesel Fish Oil Bio-Oil UD

B20D80 B80D20 Bio-Oil

464

523

432457

437 422

Unb

urne

d H

ydro

carb

ons

(ppm

)

at 75 °C

Fig. 9. Variation of unburned hydrocarbons at 80% of full load.

254 E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257

leading to longer droplet evaporation duration. The ignition delayis higher for neat diesel compared to neat bio-oil. It decreases withthe increase of bio-oil mass fraction in the blend. This is due tohigher oxygen present in the blend. This increases the fuel evapo-ration and easily mixes with oxygen, also the combustion startsearlier. Ignition delay is shorter with bio-oil compared to dieseland other blends. The oxygen present in the bio-oil increases theair fuel mixing rate which reduces the physical delay and highercetane number of fuel starts the ignition earlier. Ignition delay islower with bio-oil UD compared to all other test fuels. Bio-oil UDhas highly volatile hydrocarbons and hence it converts quickly tovapor and mixes with air which helps to start the combustionearlier.

Fig. 7 depicts the variation of combustion duration for varioustest fuels at 80% of full load. Combustion duration is marginallyhigher with preheated fish oil compared to bio-oil and diesel. It de-creases as the bio-oil concentration increases in the blend. Theoxygen present in the blend increases the air fuel mixing rate tendsto increase the amount of fuel being prepared during the initialstage of combustion (premixed combustion). This increases thepremixed combustion and reduces the diffusion combustion. It ismarginally higher with diesel compared to bio-oil and bio-oil UD.It is shorter with neat bio-oil compared to all other test fuels. Thisdecrease in combustion duration with bio-oil is due to higherintensity of premixed combustion rate as a result of better mixingof fuel and air, which is clear from the heat release rate curve.

5.2. Analysis of emission parameters

Fig. 8 depicts the variation of oxides of nitrogen emission withdiesel, preheated fish oil, bio-oil UD, B20D80, B80D20 and bio-oilat maximum efficiency point (80% of full load). NOx emissions withbio-oil UD is higher compared to diesel and bio-oil. This is mainlydue to highly volatile hydrocarbons which lead to better premixedcombustion. From heat release diagram, the premixed combustionis very strong with bio-oil UD than diesel and bio-oil. Hence, higherNOx is with bio-oil UD than diesel and bio-oil. NOx emissions areslightly reduced with bio-oil than bio-oil UD but it is still higherthan diesel. The main reason of mixing diesel with bio-oil is to findthe trade-off between NOx and particulate emission with mini-mum use of diesel. NOx emissions decrease as the amount of dieselin the blend goes up. The decrease in NOx is due to the poor mixingof fuel and air which leads to decrease in the premixed combus-tion. The formation of NOx is highly dependent on in-cylinder tem-peratures, oxygen concentration which is 5.34% with bio-oil andresidence time for the reaction to take place. Also, as seen in heatrelease diagram, the initial combustion is very strong with bio-oilwhich increases the combustion temperature. The premixed com-bustion with diesel is lower and the diffusion combustion is higher

compared to bio-oil which shows that burning rate with diesel islower compared to bio-oil.

The variation of unburned hydrocarbon (UHC) emissions withdifferent test fuels at 80% of full load is shown in Fig. 9. It is higherwith preheated fish oil compared to all other test fuels. UHC forma-tion is due to high viscosity and poor volatility of fish oil lead topoor air fuel entrainment. It is lowest with bio-oil, due to sufficientquantity of oxygen present in the fuel itself (5.34%). There is moreoxygen chemically bounded in bio-oil which is an additionalsource of oxygen other than oxygen present in the intake air. Thisoxygen helps the formation of air fuel mixture leading to completecombustion. The ignition is started earlier and good combustion isachieved with bio-oil due to high cetane number. The intensity ofUHC in the exhaust gas increase as the quantity of diesel increasesin the blend. Also, there is a slight increase in UHC emissions withneat diesel. This increase in UHC emissions with neat diesel is dueto the absence of oxygen molecules in the fuel for the oxidationprocess.

Fig. 10 shows the variation of carbon monoxide for various testfuels at maximum efficiency point (80% of full load). The increasedCO emission for preheated fish oil is due to the following reasons;poor atomization, lower cetane number, longer ignition delay andpoor mixture formation. It decreases when the bio-oil is mixedwith diesel. Bio-oil is having higher concentration of oxygen con-tent which leads to greater oxidation than diesel fuel. CO emissionreduces with bio-oil UD and it is further reduced with bio-oil. Dueto the good spray characteristics as a result of low viscosity of bio-oil and bio-oil UD; all the fuel droplets are mixed with air at theend of the compression stroke which results in complete combus-tion. The change in CO emission with bio-oil and bio-oil UD is very

Page 7: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

Diesel Fish Oil Bio-Oil UD

B20D80 B80D20 Bio-Oil

0,070,074

0,057

0,064

0,056 0,054

Car

bon

Mon

oxid

e (%

)

at 75 °C

Fig. 10. Variation of carbon monoxide at 80% of full load.

6.6

6.8

7

7.2

7.4

7.6

7.8

8

Diesel Fish Oil Bio-Oil UD

B20D80 B80D20 Bio-Oil

7,14

7,91

7,327,22

7,59 7,62

Car

bon

Dio

xide

(%

)

at 75 °C

Fig. 11. Variation of carbon monoxide at 80% of full load.

0.001

0.004

0.007

0.010

0.013

0.016

0.019

Diesel Fish Oil Bio-Oil UD

B20D80 B80D20 Bio-Oil

0,014

0,018

0,012

0,014

0,0120,010

Par

ticu

late

Mat

ter

(g/h

)

at 75 °C

Fig. 12. Variation of particulate matter at 80% of full load.

25

26

27

28

29

30

31

32

33

Diesel Fish Oil Bio-Oil UD

B20D80 B80D20 Bio-Oil

29,98

28,64

32,12

30,6

31,832,4

Bra

ke T

herm

al E

ffic

ienc

y (%

)

at 75 °C

Fig. 13. Variation of brake thermal efficiency at 80% of full load.

E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257 255

low. The combustion rate is reduced and hence the carbon monox-ide emission increases with the addition of diesel to bio-oil. It fur-ther increases with neat diesel. The increase in the quantity ofdiesel in the blend decreases the engine performance as well as in-creases the CO emission.

Fig. 11 shows the variation of carbon dioxide emission for dif-ferent fuels at 80% of full load. The CO2 emission is lower with die-sel compared to all other test fuels. The decrease in CO2 emissionwith diesel shows that the combustion is not perfect due to inad-equate supply of oxygen. CO2 emission increases with the additionof bio-oil. It further increases with neat bio-oil. This is due to oxy-gen bounded in the bio-oil contributes better mixing rate and leadsto better oxidation of fuel. It increases to the maximum with pre-heated fish oil. Even though the combustion is not proper with pre-heated fish oil, the CO2 is very high. This is because of more oxygenpresent in the fuel and it is about 10.4%.

The results of particulate matter (PM) emissions of the differenttest fuels at 80% of full load are shown in Fig. 12. The highest par-ticulate emissions are obtained with fish oil at 75 �C. Higher PMemissions with high viscous fish oil are due to poor atomizationof fuel. Higher density and bigger size of fuel molecules containedin the fish oil is considered to result in poor atomization. PM emis-sions are higher with diesel than bio-oil. It is reduced with theaddition of bio-oil with diesel. This clearly indicates that the bio-oil exhibit a clean combustion. PM emissions in a diesel engineare due to carbon and soot particles accumulated and are decidedby intensity of diffusion combustion phase. Preheated fish oil re-quires longer time to mix with oxygen which results in lower pre-mixed combustion phase and longer diffusion combustion phase.PM emissions for bio-oil are very low compared to diesel. The pres-ence of oxygen in the fuel and intake air improve the air fuel mix-

ing rates, help to prepare more quantity of fuel for premixedcombustion. Also, the combustion duration is very short for bio-oil. All the above mentioned reasons lead to reduce the PM in theexhaust for bio-oil and higher blends of bio-oil. There is a slight in-crease in PM emission with bio-oil UD compared to bio-oil. The in-creased combustion duration and longer ignition delay of bio-oilUD increase the PM emissions.

5.3. Analysis of performance parameters

Fig. 13 shows the variation of brake thermal efficiency of thedifferent test fuels at 80% of full load (maximum efficiency point).It is observed that the brake thermal efficiency is lower with pre-heated fish oil. This is due to poor mixture formation as a resultof higher viscosity, density and lower volatility of fish oil. Thereis a marginal increase in brake thermal efficiency with neat dieselfuel. There is an increase in brake thermal efficiency when smallquantity of bio-oil is added with diesel (B20D80) and it is due tothe increase in fuel quality leading to higher premixed combustioncompared to neat diesel. It is observed that the brake thermal effi-ciency of bio-oil is higher compared to all other fuels tested. This ismainly due to high cetane number of bio-oil compared to dieselwhich increases the rate of combustion. High cetane number ofbio-oil starts the combustion early and more energy is releasedduring the beginning of expansion stroke resulting higher mechan-ical output. The brake thermal efficiency decreases with the blendof bio-oil with diesel (B80D20). This slight change in brake thermalefficiency is due to decrease in oxygen content in the blend. Thedecrease in brake thermal efficiency is very less with bio-oil UDcompared to bio-oil.

The variation of exhaust gas temperature with all test fuels at80% of full load is shown in Fig. 14. The exhaust gas temperatureis very high with preheated fish oil. Due to higher viscosity and

Page 8: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

25

75

125

175

225

275

325

375

425

Diesel Fish Oil Bio-Oil UD

B20D80 B80D20 Bio-Oil

371

412

339364 351 347

at 75 °C

Exh

aust

Gas

Tem

pera

ture

( ° C

)

Fig. 14. Variation of exhaust gas temperature at 80% of full load.

256 E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257

poor volatility of preheated fish oil, less amount of fuel is preparedduring the start of combustion. Due to this the premixed combus-tion phase decreases and more amount of fuel is burned in the laterpart of combustion (diffusion combustion phase) which leads to in-crease the exhaust gas temperature. It is lower with bio-oil UD andbio-oil compared to all other fuels. Lower viscosity and good vola-tility of bio-oil and bio-oil UD lead to better mixture formationwhich make the heat release shorter. Also lower ignition delaystarts the combustion earlier and shortens the combustion dura-tion which results in lower exhaust gas temperature compared todiesel. As the concentration of bio-oil increases in the blend, theexhaust temperature decreases compared to diesel. This decreasein exhaust gas temperature is due to shorter ignition delay andcombustion duration as compared to diesel, which shows moreamount of fuel burned during premixed combustion. When theconcentration of bio-oil increases to 80% in the blend, the exhaustgas temperature is almost nearer to neat bio-oil.

6. Analysis of fuel injection system

The experiments were conducted with constant injection tim-ing of 20 �CA bTDC and constant injection pressure of 250 bar forall the test fuels. With optimized injection parameters of fuel injec-tion timing, pressure and nozzle hole size, better performance willbe obtained with bio-oil and its blends rather constant injectionparameters.

From the experimental results, the NOx emissions are higherwith bio-oil and its blends. However, the NOx emissions can belowered further. It is well known that fuel injection timing has a

Table 4Error of instruments used.

Parameter Sensor type

Torque (Tm) Effort sensor (FN 3148Engine speed AVL 364CInjection timing AVL 364CIntake air flow rate Differential pressure trFuel flow rate Coriolis type mass flowIn-cylinder pressure Piezo-electric (AVL QHInjection pressure Piezo-electric (AVL QHIntake air temperature Differential pressure trFuel injection temperature K type thermocoupleExhaust gas temperature K type thermocoupleAmbient air temperature HD 2012 TC/150Relative humidity HD 2012 TC/150Hydrocarbon emissions FID (GRAPHITE 52 M)Nitric oxides emissions TOPAZE 32 MNon-reacted oxygen Infra-red detector (MIRCarbon monoxide emissions Infra-red detector (MIRParticulate matter emissions TEOM 1105Fuel lower heating value Isoperibol calorimeter

strong bearing on the start of combustion and then the attainmentof in-cylinder pressures and temperatures. If the timing is retardedwith respect to TDC position compared to its recommended injec-tion timing, lower in-cylinder pressures and temperatures woulddevelop resulting in less favorable conditions for the formation ofNOx emissions [20].

The design variables of the injection system are related to theshape of the cam profile, to the nozzle geometry, and to the controlparameters influencing the injection pressure and quantity. Thegeometrical properties of the cam profile and the injection param-eters are kept within acceptable limits by the imposed constraints.Once the above-mentioned injection characteristics are optimallymatched to each other for different test fuels, they typically resultin a fuel spray with small droplet size (good atomization), long tippenetration, and narrow spray angle. These spray characteristicsplay an important role to improve engine performance, to reducefuel consumption, and to reduce harmful emission [21]. For thisreason, it seems to be a good idea to involve fuel spray atomizationinto the optimization process of the fuel injection system.

7. Analysis of uncertainty

In measuring any quantity, the results will always differ fromthe true value even with careful experimentation. This error inmeasurement may be either random or systematic. By adding acorrection value, the systematic error can be removed. Random er-ror can only be estimated statistically and cannot be predicted inadvance. Its presence can be detected only when the same quantityis measured again and again under the same conditions and withthe same care.

The uncertainty was estimated based on Gaussian distributionmethod with confidence level of ±2r (95.45% of measured datalie within the limits of ±2r of mean). Thus uncertainty is estimatedusing the following equation.

Uncertainty of any measured parameter ðDXÞ ¼ 2ri=X � 100

ð1Þ

Experiments were conducted to obtain the mean (X) and stan-dard deviation (ri) of any measured parameter (Xi) for 20 readings.From the measured parameters, the uncertainty is computed basedon Kline and McClintock method [22].

In order to get the realistic error limits for any computed quan-tity based on several measured quantities the principle of root-sum-square method is used and the magnitude of the error is givenby

Error

) ±0.1 N m±3 RPM±0.05 �CA

ansmitter (LPX5841) ±1% of measured valuemeter (RHM015) ±0.5% of measured value

32D) ±2 bars33D) ±2.5 barsansmitter (LPX5841) ±1.6 K

±1.6 K±1.6 K±0.2 K±2%±10 ppm±100 ppm

2 M) ±0.25%2 M) 50 ppm

±10 ng/s(PARR 6200CLEF) ±0.25% of measured value

Page 9: Assessment of liquid fuel (bio-oil) production from waste fish fat and utilization in diesel engine

E.G. Varuvel et al. / Applied Energy 100 (2012) 249–257 257

DR ¼ SQRTðð@R=@X1 � DX1Þ2 þ ð@R=@X2 � DX2Þ2 þ . . .

þ ð@R=@Xn � DXnÞ2Þ ð2Þ

Using the Eq. (2) the uncertainty for a given operating conditionwas computed.

The estimated uncertainty values at different operating condi-tions are

Brake power: 0.4–1.7%Specific fuel consumption: 0.5–1.9%Brake thermal efficiency:0.6–1.9%

The error of the various instruments used in this experimentalstudy is given in Table 4.

8. Conclusions

The diesel engine was successfully operated with diesel, pre-heated fish oil, bio-oil UD, B20D80, B80D20 and bio-oil withoutany problems. The following conclusions are made based on thepresent experimental work. The brake thermal efficiency of pre-heated fish oil is lower compared to all other fuels tested. Due toits high viscosity, the combustion is poor which leads to high CO,HC and PM emissions. Addition of bio-oil with diesel improvesthe brake thermal efficiency and reduces the PM, and CO emissionsdue to improvement in the rate of combustion. There is not muchchange in brake thermal efficiency with bio-oil and bio-oil UD.However, NOx, HC, CO and PM emissions are higher with bio-oilUD compared to bio-oil. Eventhough the ignition delay is shortwith bio-oil UD, the rate of combustion is not very good comparedto bio-oil. The brake thermal efficiency of bio-oil is higher com-pared to diesel. Despite of increase in NOx emissions, all otheremissions are lower compared to diesel. Addition of small quantityof diesel with bio-oil decreases the NOx emissions without muchchange in its performance and other emissions. Finally, it is con-cluded that the bio-oil after distillation can be used as a substituteof diesel fuel either partially or fully.

References

[1] Richardson B. Sustainable transport: analysis frameworks. J Transp Geogr2005;13:29–39.

[2] Black WR. Socio-economic barriers to sustainable transport. J Transp Geogr2000;8:141–7.

[3] EC, 2001b. Communication from the Commission to the European Parliament,the Council, the Economic and Social Committee, and the Committee of theRegions on alternative fuels for road transportation and on a set of measures topromote the use of bio-oils. COM 547: Brussels; 2001b.

[4] Eric D. Larson bio-oil production technologies: status, prospects andimplications for trade and development. United Nations Conference on Tradeand Development, New York and Geneva; 2008.

[5] Guru Metin, Koca Atilla, Can Ozer, Cınar Can, Sahin Fatih. Biodiesel productionfrom waste chicken fat based sources and evaluation with Mg based additivein a diesel engine. Renew Energy 2010;35:637–43.

[6] Taufiqurrahmi Niken, Mohamed Abdul Rahman, Bhatia Subhash. Production ofbio-oil from waste cooking palm oil using nanocrystalline zeolite as catalyst:process optimization studies. Bioresour Technol 2011;102:10686–94.

[7] Andersen Otto, Weinbach Jan-Erik. Residual animal fat and fish for biodieselproduction. Potentials in Norway. Biomass Bioenergy 2010;34:1183–8.

[8] Gardner N, Manley BJW, Pearson JM. Gas emissions from landfills and theircontributions to global warming. Appl Energy 1993;44:165–74.

[9] Montejo Cristina, Costa Carlos, Ramos Pedro, Márquez María del Carmen.Analysis and comparison of municipal solid waste and reject fraction as fuelsfor incineration plants. Appl Therm Eng 2011;31:2135–40.

[10] Arpa Orhan, Yumrutas Recep, Demirbas Ayhan. Production of diesel-like fuelfrom waste engine oil by pyrolitic distillation. Appl Energy 2010;87:122–7.

[11] Al-Zuhair Sulaiman, Hussein Ahmedin, Al-Marzouqi Ali H, Hashim Isameldin.Continuous production of biodiesel from fat extracted from lamb meat insupercritical CO2 media. Biochem Eng J 2012;60:106–10.

[12] Lam Su Shiung, Russell Alan D, Chase Howard A. Microwave pyrolysis, a novelprocess for recycling waste automotive engine oil. Energy 2010;35:2985–91.

[13] Helwani, Othman Z, Aziz MR, Fernando WJN, Kim WJN. Technologies forproduction of biodiesel focusing on green catalytic techniques: a review. FuelProcess Technol 2009;90:1502–14.

[14] Senthil Kumar M, Kerihuel A, Bellettre J, Tazerout M. Experimentalinvestigations on the use of preheated animal fat as fuel in a compressionignition engine. Renew Energy 2005;30:1443–56.

[15] Arpa Orhan, Yumrutas� Recep, Argunhan Zeki. Experimental investigation ofthe effects of diesel-like fuel obtained from waste lubrication oil on engineperformance and exhaust emission. Fuel Process Technol 2010;91:1241–9.

[16] Behçet Rasim. Performance and emission study of waste anchovy fish biodieselin a diesel engine. Fuel 2011;92:1187–94.

[17] Murugan S, Ramaswamya MC, Nagarajan G. Performance, emission andcombustion studies of a DI diesel engine using Distilled Tyre pyrolysis oil–diesel blends. Fuel Process Technol 2008;89:152–9.

[18] DeOliveira Edimar, Quirino Rafael L, Suarez Paulo AZ, Alexandre GS. PradoHeats of combustion of bio-oils obtained by pyrolysis and bytransesterification and of bio-oil/diesel blends. Thermochim Acta2006;450:87–90.

[19] Final report of the bio-oils research advisory council. Bio-oils in the EuropeanUnion – a vision for 2030 and beyond. Directorate-General for ResearchSustainable Energy Systems; 2006.

[20] Prasad USV, Madhu Murthy K, Amba Prasad Rao G. Influence of fuel injectionparameters of DI diesel engine fuelled with biodiesel and diesel blends. In:International conference on mechanical, automobile and robotics engineering(ICMAR’2012) Penang, Malaysia; 2012.

[21] Kegl Breda, Kegl Marko, Pehan Stanislav. Optimization of a fuel injectionsystem for diesel and biodiesel usage. Energy Fuels 2008;22:1046–54.

[22] Kline SJ, McClintock FA. Describing uncertainties in single sampleexperiments. Mech Eng 1953:3.