16

Energy Conversion and Management 52 (2011) 849–857

Contents lists available at ScienceDirect

Energy Conversion and Management

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

A comparison of water–diesel emulsion and timed injection of waterinto the intake manifold of a diesel engine for simultaneous controlof NO and smoke emissions

K.A. Subramanian *

Engines and Unconventional Fuels Laboratory, Centre for Energy Studies, Indian Institute of Technology Delhi, New Delhi 110 016, India

a r t i c l e i n f o a b s t r a c t

Article history:Received 23 April 2009Received in revised form 26 December 2009Accepted 8 August 2010

Keywords:Water–diesel emulsionManifold timed water injectionDiesel engine performanceCombustion and emission characteristicsNO and smoke emissions

0196-8904/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.enconman.2010.08.010

* Tel.: +91 011 26591247; fax: +91 011 26581121.E-mail address: [email protected]

Experiments were conducted to compare the effects of water–diesel emulsion and water injection intothe intake manifold on performance, combustion and emission characteristics of a DI diesel engine undersimilar operating conditions. The water to diesel ratio for the emulsion was 0.4:1 by mass. The samewater–diesel ratio was maintained for water injection method in order to assess both potential benefits.All tests were done at the constant speed of 1500 rpm at different outputs. The static injection timing of23� BTDC was kept as constant for all experimental tests.

In the first phase, experiments were carried out to asses the performance, combustion and emissioncharacteristics of the engine using the water–diesel emulsion. The emulsion was prepared using the sur-factant of HLB:7. The emulsion was injected using the conventional injection system during the compres-sion stroke. The second phase of work was that water was injected into the intake manifold of the engineusing an auxiliary injector during the suction stroke. An electronic control unit (ECU) was developed tocontrol the injector operation such as start of injection and water injection duration with respect to thedesired crank angle.

The experimental result indicates the both methods (emulsion and injection) could reduce NO emissiondrastically in diesel engines. At full load, NO emission decreased drastically from 1034 ppm with basediesel to 645 ppm with emulsion and 643 ppm with injection. But, NO emission reduction is lesser withinjection than emulsion at part loads. Smoke emission is lower with the emulsion (2.7 BSU) than withwater injection (3.2 BSU) as compared to base diesel (3.6 BSU). However, CO and HC levels were higherwith emulsion than water injection. As regards NO and smoke reduction, the emulsion was superior toinjection at all loads. Peak pressure, ignition delay and maximum rate of pressure rise were lesser withwater injection as compared to the emulsion. It is well demonstrated through this comparative study thatthe emulsion method has higher potential of simultaneous reduction of NO and smoke emissions at allloads than injection method.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Diesel engines play a major role in power generation, agricul-ture, mass transportation etc. India, one of the fast developingcountries, is known as diesel driven economy as the consumptionof diesel to gasoline is about 5:1. This is mainly due to diesel en-gine operation at higher compression ratio and leaner air fuel ratiothan SI engine resulting in higher thermal efficiency. Even thoughthe levels of HC and CO are very lower in the diesel engines thangasoline engine, however, it emits high levels of NOx and smokeemission. The stringent emission norms pose to a big challengeto the researchers for controlling these emissions.

ll rights reserved.

The main causes of formation of particulate emission from die-sel engines are heterogeneous air–fuel mixture, poor mixing of fuelwith air, high diffusion combustion phase, fuel containing sulfurcontent, high fuel density, etc. Particulate emission could be re-duced by improving mixing rate of fuel with air, enhancing pre-mixed combustion phase by increasing ignition delay, etc.However, it would lead to high in-cylinder temperature resultingin high NOx formation as it is mainly a function of temperature.This conflict nature leads to difficulty in simultaneous control ofNOx and particulate emissions from diesel engines.

Several methods have been tried and reported in literatures tocontrol the emissions. Most in-cylinder control techniques do notsimultaneously reduce NOx and smoke emissions. For example,EGR technique can reduce NOx significantly but it would increaseparticulate emissions where as oxygen enrichment technique

Nomenclature

BTDC before top dead centrebsfc brake specific fuel consumptionBMEP Brake mean effective pressureBSU Bosch smoke unitBP brake powerCO carbon monoxideC1 and C2 constants (C1 = 130 and C2 = 1.4)DI direct injectiondQ/dt heat release rate (W)ECU electronic control unitEGR exhaust gas recirculationemul emulsionHLB hydrophile lipophile balanceHC hydrocarbonH heat transfer coefficientinject injectionNO nitric oxide

NOx oxides of nitrogenppm part per millionsPAH poly aromatic hydrocarbonPC personal computerP cylinder pressure (N/m2)Qw heat transfer to the wall (J)RPM revolution per minuteSI spark ignitionSOx oxides of sulphurt time (s)T average gas temperature (K)TDC top dead centreU internal energy (J)V cylinder volume (m3)Vp mean piston speed (m/s)W/D water to diesel ratioW/O water-in-oil

850 K.A. Subramanian / Energy Conversion and Management 52 (2011) 849–857

could reduce smoke emission drastically but NOx emission wouldshoot up. The combination of many techniques including aftertreatment may be an option to reduce these emissions but theadditional costs including initial investment, maintenance, addi-tional energy consumption by the devices, may be an expensivesolution with system complexity. So, a simple technique needs tobe developed to reduce NOx and Smoke emission simultaneouslywithout fuel penalty.

1.1. Water–diesel emulsion

Modifying the fuel offers a simple way to control these emis-sions as many researchers reported in literatures that the useful-ness of water–diesel emulsion on performance improvement andemission reduction of diesel engines. Water added diesel can re-duce NOx and smoke simultaneously. NO emission decreased dras-tically due to thermal, dilution and chemical effects (enhancementof OH radicals) of water [1]. Smoke reduction may be due toimprovement in mixing rate of fuel with air by micro-explosionphenomenon and increase in premixed combustion phase due tolong ignition delay [2]. The effect of water–diesel ratio up to 0.8by mass was studied and reported a reduction in NOx about 60%and smoke about 50–70% at a given load (BMEP of 5.31 kg/cm2)[2]. However, at lower loads, they reported a slight increase in BSFCcompared to operation on plain diesel as a result of overcoolingand over mixing of the charge. Frederic Barnaud et al. reported oxi-des of nitrogen, smoke and particulate emission at 0.13:1 water–diesel ratio could reduce up to 30%, 80% and 50% respectively [3].In addition, a further reduction in particulate emission of up to90% was obtained with the use of an oxidation catalyst. The authorstudied extensively on water–diesel emulsions with ratio of 0.3:1,0.4:1, 0.5:1 and 0.6:1, and concluded that NOx and smoke reduc-tion with 0.4:1 and 0.5:1 water to diesel ratio at full load was about33.8% and 42%, 25% and 48% respectively [4]. However, there is anincrease in CO and HC levels and drop in brake thermal efficiencyat lower loads. Sheng et al. conducted experiments to study thecombustion mechanism of water in diesel fuel emulsion spray ina combustion bomb and simulated road-load conditions. They re-ported smoke decreased up to 30% and NOx level also decreased,and the water fraction of 6–15% had no significant effect on enginesmoke [5]. Anna lif et al. reported that there is a reduction in NOx

and particulate matter but increase in CO and HC emissions levelwith the increasing the water content in the fuel [6]. In addition,they reviewed also effects of use of water–vegetable oil and

water–biodiesel emulsion in diesel engines. Nadeem et al. reportedthat water emulsification has a potential of significantly reduce theformation of NOx, CO, SOx, particulate matter, soot and hydrocar-bons [7]. They extensively studied the effect of temperature, stir-ring speed and mixing time on emulsion formulation andreported gemini surfactant (trade mark of the surfactant formu-lated by them) has much finer and better distributed water drop-lets as compared to those stabilized by conventional surfactant.This information is important as formulation of water–dieselemulsion needs suitable surfactant. Kadota et al. reviewed recentadvances in the combustion of water–fuel emulsion, phenomeno-logical burning process, ignition process, the flame phenomenaincluding soot concentration profile, etc. [8]. They concluded thatwater emulsion has high potential to increase thermal efficiencyand suppress the emissions such as soot, PAH, carbonaceous resi-due and they also stressed the needs of more experimental teststo identify the dominant mechanism and its full potential. Chernget al. conducted experiments and compared with w/o two phaseemulsion with O/W/O three phase emulsion and they reported thatthree phase emulsion reduces bsfc, CO and NOx emission as com-pared to two phase emulsion (W/O) [9]. Water-in-oil emulsion ismore suitable for diesel engines applications. This means water isenclosed by oil droplet resulting in micro-explosion diesel sur-rounded water particle. Micro-explosion would occur when lowboiling point of liquid (like water) surrounded by a high boilingpoint of liquid such as diesel. As the heat transfer takes place fromdiesel to water during compression stroke in diesel engines, thelow boiling point goes to unstable superheated state leading to mi-cro-explosion resulting in better mixing of fuel with air. The phe-nomenon is discussed in detailed in latter section. PM and NOemission decreased drastically and water–diesel emulsion hasadvantage of no need of huge change in infrastructure [10]. Theuse of water–diesel emulsion in diesel engines not only reducesNOx and particulate emissions simultaneously, it could also im-prove fuel economy at higher loads [2,4]. Abu-Zaid M. studiedthe effect of water–diesel emulsion with different ratio of 0, 5,10, 15 and 20 on performance and exhaust temperature of a dieselengine and concluded the average increase in the brake thermalefficiency for 20% water–diesel emulsion is approximately 3.5%[11]. However, emulsified fuels have the problem of increase inCO, HC and rate of pressure rise.

The water–diesel emulsion along with other techniques such asEGR and different injection timing. could also give high beneficialresults to overcome some problems. It is reported a 55% reduction

Table 1Specifications of the test engine.

Type Four stroke, air cooled, over headvalve, compression ignition engine

Make Kirloskar, TAF1Number of cylinders OneBore 87.5 mmStroke 110 mmDisplacement volume 661.5 ccCompression ratio 17.5:1Static injection timing 23� btdcRated power 6 BHP at 1500 rpm

Table 2Measurement accuracy and uncertainty.

Measurements Accuracy

CO 0.01%NOx ±1 ppmHC ±5 ppmSmoke ±0.1 BSUComputed results Uncertainty (%)Torque ±2.1%Time for air flow ±0.36%Mass flow rate of fuel ±0.82%Brake thermal efficiency ±2.3%Cylinder peak pressure ±1.4%HC ±2%CO ±2%NO ±0.5%

K.A. Subramanian / Energy Conversion and Management 52 (2011) 849–857 851

in NOx and 45% smoke reduction at 20% emulsion with 16.7% hotEGR [12]. A small quantity of hydrogen peroxide (5%) could im-prove the overall performance and emissions of a diesel enginewith water–diesel emulsions [13]. NOx, smoke, CO and HC emis-sions decreased drastically with the water–diesel–hydrogen perox-ide emulsion as compared to plain water–diesel emulsion [13]. Useof 10% diethyl ether along with water–diesel emulsion (0.4:1 byweight) can significantly reduce HC and CO emission without ad-verse effects on NOx and smoke [14]. HC and CO levels drop from75 ppm to 40 ppm and 0.175% to 0.1% respectively at full load ascompared to neat water–diesel emulsion. The performance andemission characteristic of a diesel engine with water–diesel emul-sion can be improved by oxygen enriched air induction. CO, HC andsmoke decreases drastically at 0.4 water–diesel ratio with oxygenconcentration of 24%. However NOx emission shooted up [15].

There is different optimum water–diesel emulsion ratios re-ported in literature. Water–diesel emulsion (15%) was found tobe the best or optimum based on controlling the engine’s emission[7]. Forty percent water–diesel emulsion is the optimum based onthe highest fuel economy [16]. The optimum water–diesel emul-sion ratio was 0.5:1 by mass based on reasonable starting perfor-mance, emulsion stability and viscosity [17]. The optimumwater–diesel emulsion ratio was chosen as 0.4:1 and 0.5:1 basedon brake thermal efficiency, emission reduction, rate of pressurerise and engine rough running and startability [4].

Even though water–diesel emulsion has advantages of emissionreduction and performance improvement of diesel engines, how-ever there was an increase of CO and HC levels and rate of pressurerise. CO and HC emission increased from 35 ppm, 0.12% with basediesel to 60 ppm, 0.15% respectively at 0.4:1 water–diesel ratio [4].In addition, the emulsified fuel increased the ignition delay andrate of pressure rise [2,4].

1.2. Water injection

Apart from the emulsion method, water can also be introducedinto the engine by way of injection into the intake air stream. Theadvantages of injection are versatile of on-line variation of waterquantity, increase of volumetric efficiency due to cooling effect,uniform or homogeneous water distribution in combustion cham-ber, etc. An increase of 1% in the specific heat of the gases in theburned zone results in about 20% reduction in NOx emission andreported a 50% reduction in NOx emission with water injection(0.03 kg per kg of dry air) in the intake manifold [18]. The combi-nation of exhaust gas recirculation (EGR: 17%) and manifold waterinjection gave a 40–50% reduction in NOx [19]. However, CO andHC emissions increased with increase in water content with intakeair [19,20]. Smoke also increased with water induction [20].

It is clearly seen from the literature study that water–dieselemulsion and water injection has advantages of reduction in NOx

and smoke emissions from diesel engines. However, there is noinformation available in literatures on which method (water withemulsion or injection) has higher potential to reduce these emis-sions under similar operating conditions. In this direction, a com-parison between the methods of water injection into themanifold and water–diesel emulsion has been made for assessingpotential benefits of water addition to diesel engines under similaroperating conditions.

2. Experimental details

2.1. Engine and experiments details

A single cylinder, 4-stroke, air-cooled diesel engine with a dis-placement volume of 661.5 cc (87.5 mm bore � 110 mm stroke)

developing 4.4 kW was used for the study. The engine was run atconstant speed of 1500 rpm. The specification of test engine is gi-ven in Table 1. The measurement accuracy and uncertainty is givenin Table 2. Commercially available diesel was used as fuel. A swing-ing field dynamometer was used to load the engine. A detailed lay-out of experimental set-up is shown in Fig. 1. A turbine type flowmeter was used to measure the airflow rate. Both diesel and emul-sion flows were measured on the mass basis. An Inductive typeneedle lift transducer was fabricated to determine the dynamicinjection timing.

2.2. Description of experimental procedure

Experiments were conducted at different loads including therated load (4.4 kW) and over load (4.75 kW) for water–diesel emul-sion and water injection. The water–diesel emulsion ratio of 0.4:1was chosen for the study as it was found to be the optimum basedon the author’s earlier research work [4]. The same water content(water–diesel ratio: 0.4:1 by mass) was maintained for both meth-ods and the results are compared with base diesel. The detailedstudy of optimization of water–diesel emulsion may be referredin literature [4]. Water–diesel emulsion was prepared using a sur-factant (HLB = 7) with help of emulsion preparation apparatuswhich is explained in Section 2.3 and the emulsion was injectedinto in-cylinder during compression stroke using the main injec-tion system. In case of water injection, plain water was injectedinto intake manifold using an auxiliary injector which was con-trolled by an ECU. A static injection timing of 23� BTDC was keptas constant for all experimental tests. The performance and emis-sion characteristics of the engine was measured for both methodsand the results are compared with base diesel. The pressure-crankangle data was measured using piezo-electric transducer and TDCencoder and the data was given as input for calculating ignition de-lay, heat release rate and rate of pressure rise. The dynamic injec-

Fig. 1. Experimental set-up.

Emulsifier (Surfactant): HLB = 7

Fig. 2. Emulsion preparation apparatus.

852 K.A. Subramanian / Energy Conversion and Management 52 (2011) 849–857

tion timing was measured using a needle lift sensor and it wasused for finding out ignition delay.

2.3. Measurement of pressure-crank angle data

A piezo electric pressure transducer was flush mounted on thecylinder head for the measurement of cylinder pressure. An elec-tro-optical sensor was developed to indicate the position of TDCas the crank shaft rotated. A 12 bit analogue to digital converterwas used to store analogue data in digital form on a PC.

2.4. Preparation of emulsion

A mixture of diesel, water, surfactant were circulated severaltimes to form an emulsion using an apparatus comprising of a cen-trifugal pump, glass jar and glass tube with a submergible nozzleportion as shown in Fig. 2. As oil soluble surfactant (HLB < 10)are the best for water-in-oil emulsions, the surfactant with HLBof 7 was used for preparing the emulsion. The surfactant usedwas 1% by weight and stability time for the emulsion was aboutone and half days. After the emulsion prepared in off-line, then itwas immediately used in the engine.

2.5. Development of manifold timed water injection system

An electronically controlled system to inject water into themanifold during the intake stroke for any fixed crank angle dura-tion was developed and installed on the engine. This consisted ofa high pressure water pump which fed a solenoid operated injec-tor. An electro-optical sensor was mounted on the cam shaft to de-tect the position of the piston. This sent out a pulse to trigger thewater injection circuit. The electronic circuit could initiate pulsesof varying width. These pulses were amplified and sent to a sole-

noid operated injector. The amount of water injected per cyclewas controlled by varying the pulse width of the signal fed to theinjector. The injector used was basically a commercially availablegasoline injector. This was mounted on the manifold such thatthe water spray would not impinge on the walls. The injectionpressure was maintained at 2 bar. This pressure was found to leadto a well atomized spray. The complete injection system is indi-cated in Fig. 3. The injector was mounted in the manifold in sucha way that the spray of water will not impinge on the wall.

2.6. Measurement of emissions

Hydrocarbon (HC) and carbon monoxide (CO) emissions weremeasured by means of a Non Dispersive Infrared Analyzer of theHoriba make (measuring range for CO: 0–10 vol.%, HC: 0–10,000 ppm). Nitric oxide (NO) was measured using a Chemilu-minescense analyzer of Rosemount Analytical make (measuringrange: 0–10,000 ppm). It may be noted that NOx emission wasnot measured and only NO emission was measured for the study.

1. Intake manifold 6. Burette

2. Injector 7. Water pump

3. Electronic circuit 8. Heat exchanger

4. Optical encoder signal 9. Flow control valve

5. Water tank 10. Pressure gauge

1 2

7

8 6

10

3 4

5

9

Fig. 3. Water injection system.

10

15

20

25

30

35

20 1 3 4 5Brake Power (kW)

Bra

ke T

herm

al E

ffici

ency

(%)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 4. Comparison of brake thermal efficiency with emulsion and injection.

K.A. Subramanian / Energy Conversion and Management 52 (2011) 849–857 853

Smoke measurement was done using the standard Bosch apparatus(scale: 0–10).

2.7. Calculation of ignition delay period and heat release rate

Ignition delay was calculated based on dynamic injection tim-ing. A computer program developed in the laboratory was usedfor heat release rate diagram using input of experimental pres-sure-crank angle data, and it was calculated using first law of ther-modynamics as given below:

dQ=dt ¼ dU=dt þ P � dV=dt þ dQw=dt ð1Þ

The heat transfer (dQ/dT) was calculated using the Hohenberg’scorrelation as given below [21]

H ¼ C1 � V�0:06c � P0:8 � T0:4 � ðVp þ C2Þ0:8 ð2Þ

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 1 2 3 4 5Brake Power (kW)

CO

(% b

y vo

l)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 5. Comparison of CO emission with emulsion and injection.

3. Results and discussion

The effects of using water–diesel emulsion as the fuel have beencompared with water injection at same water–diesel ratio (0.4:1)on the performance, combustion and emission characteristics ofthe diesel engine.

3.1. Performance characteristics of engine

It is seen from Fig. 4 that the brake thermal efficiency reduces atall outputs below diesel values with injection due to poor combus-tion as a result of reduction in the charge temperature. In the caseof the emulsion, at high outputs, the brake thermal efficiency withthe emulsion is significantly above the base diesel value. This isdue to the enhanced premixed combustion phase and also bettermixture formation with the emulsion. When the emulsion is used,the total injected quantity increases as compared to the dieselmode and this will cause a greater amount of air to get entrainedinto the spray and form a better mixture. In addition, the phenom-enon of micro-explosion could also aid fuel air mixtures prepara-tion and lead to better combustion. The micro-explosion occurswhen the low boiling point of water trapped in the high boilingpoint diesel is heated upon injection into the compressed air. Thisleads to a sudden expansion of water due to vaporization. This pro-cess leads to better dispersion of the diesel which encloses thewater. This phenomenon termed as micro-explosion has been dis-cussed by the researches Murayama [2] and Tsao [22]. Thus the

emulsion is better than water injection as regards brake thermalefficiency. However, the brake thermal efficiency is below dieselvalues only at low loads. It is due to overcooling of charge, whichresults in poor combustion. Such a tendency has been experiencedby earlier researches also [2]. It may be noted that in the case of theemulsion, all the water that is introduced is close to the fuel andthis has a significant effect on combustion. That is the reason forthe reduced efficiency at low loads (the temperature is alreadylow) with the water–diesel emulsion (Fig. 4). With the emulsion,water concentration near the fuel is always the same irrespectiveof the load as long as the water–diesel ratio is held constant. Inthe case of water injection the water is uniformly distributed andhence the water concentration near the fuel is lesser at low loadswhere the amount of fuel injected is low. Thus the emulsion seemsto perform worse than injection at low loads.

3.2. Emission characteristics of engine

CO and HC levels are lower at low loads with the injection ascompared to the emulsion. But it is similar to the water–dieselemulsion at high loads (Figs. 5 and 6). However, it is always higherthan base diesel values due to incomplete combustion and use ofricher mixtures due to lower brake thermal efficiency. It is re-ported in literature that HC level starts to decrease at water to die-sel ratio of 0.5:1 as compared to 0.4:1 and 0.3:1 and the reason forthis is still unclear [3]. Such trends in HC levels have been reportedby other researcher also Bertrand [23]. Matsuo Odaka et al. ob-served the increasing trend of CO and HC emission with water

0 1 2 3 4 5Brake Power (kW)

0

10

20

30

40

50

60

70

80

HC

(ppm

)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 6. Comparison of HC emission with emulsion and injection.

0 1 2 3 4 5Brake Power (kW)

0

200

400

600

800

1000

1200

NO

(ppm

)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 7. Comparison of NO emission with emulsion and injection.

BP: 4.4kW

Die-Inj (643)

Die-emul (645)

Diesel (975)

0

200

400

600

800

1000

1200

NO

(ppm

)

Fig. 8. Comparison of NO emission with emulsion and injection at 100% load.

BP: 1.87kWDiesel (459)

Die-emul (226)

Die-Inj (369)

0

50

100

150

200

250

300

350

400

450

500

NO

(ppm

)

Fig. 9. Comparison of NO emission with emulsion and injection at 40% load.

0

1

2

3

4

5

6

0 1 2 3 4 5Brake Power (kW)

Smok

e (B

SU)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 10. Comparison of smoke emission with emulsion and injection.

854 K.A. Subramanian / Energy Conversion and Management 52 (2011) 849–857

injection [19]. However, the HC levels are higher with emulsion ascompared to base diesel. The HC level is higher with the emulsionas the water is closely in contact with the diesel particles andquenches the combustion process. This is no great difference be-tween the injection and normal diesel modes.

It is seen in Fig. 7 that the emulsion is the more effective inreducing NO levels at given water to diesel ratio. Water injectionalso leads to a significant reduction in the NO levels at high outputswhen the injected water quantity is high. Thus it can find thatwater that is close to the fuel is more effective in controlling theNO level as compared to the condition when it is uniformly distrib-uted in the cylinder. Uniform distribution in the cylinder will leadto a global temperature drop and oxygen concentration drop butlocal presence of water near the fuel can reduce the oxygen con-centration and temperature near the flame. In addition, Miyauchiet al. reported that OH radical concentration increases by wateraddition, which promotes the oxidation of hydrocarbon fragmentsand leads to reduction in NO levels [1]. These factors will beresponsible for the observed trends. Since both the methods com-pared here are mainly to control NO emission at high outputs theyseem to be equally effective on that count as shown in Fig. 8. NOemission decreased from 975 ppm with base diesel to 645 ppmwith emulsion and 643 ppm with injection. At low outputs, emul-sion is better. So, water injection method is not effective on NOemission reduction at part load as shown in Fig. 9. NO level de-creased from 459 ppm with base diesel to 226 ppm with emulsionwhere as it was 369 ppm with water injection at 0.4:1 water todiesel ratio at 1.9 kW power output (40% load).

The smoke emission reduction is most significant with theemulsion as seen in Fig. 10. There is a little change in the smokelevel between the pure diesel and water injection methods. Thiscould be due to the absence of the micro-explosion phenomenoneven though there is an additional benefit from increase in OH con-centration and premixed combustion phase by long ignition delay.Thus the micro-explosion phenomenon may play a major role tocontrol smoke level with water-in-oil emulsion. Murayama [2]

and

Mu

ller-Deth

lefs[24]

reportedth

atboth

micro-explosion

lead-in

gto

betterm

ixing

ofth

eair

and

fuel

and

increase

inO

Hradical

concen

trationcou

ldbe

the

reasons

[2,24].At

ratedload

(4.4kW

),sm

okeden

sitydecreased

from3.6

BSU

with

basedieselto

2.7w

ithem

ulsion

and

to3.2

BSU

with

water

injection

at0.4:1

water

todie-

selratioas

show

nin

Fig.11.How

ever,smoke

emission

increased

atpart

load(Figs.10

and

12).Itm

aybe

noted

fromth

ean

alysis(Figs.

7–12),atallloads,em

ulsion

meth

odh

ash

igher

potentialof

simu

l-tan

eous

reduction

ofN

Oan

dsm

okeem

issionat

allloads

than

injection

meth

od.

3.3.Comparison

between

thepresent

work

with

otherresearchers

workA

briefcom

parisonof

the

present

work

with

other

some

researchers

work

isgiven

inTable

3.Eventh

ough

itis

verydiffi

cult

tocom

parew

ithoth

ersw

orkas

the

difference

indesign

and

oper-atin

gparam

eterssu

chas

engin

etype,

capacityof

engin

e,design

and

operating

parameter,accu

racyof

the

instru

men

t,atmospheric

temperatu

relevel

and

fuel

quality,

the

primary

objectiveof

the

comparison

isto

know

wh

atare

the

researchgaps

availablefor

un

derstandin

gth

epresen

tw

orkas

wellas

tokn

owth

escope

offu-

ture

work.It

may

ben

otedth

atcom

bustion

characteristics

aren

otreported

bym

ostof

the

researchers.M

urayam

adid

an

otablecon

-

BP: 4.4 kW

Diesel (3.6)

Die-em

ul (2.7)

Die-Inj

(3.2)

0

0.5 1

1.5 2

2.5 3

3.5 4

SMOKE (BSU)Fig.11.C

omparison

ofsm

okew

ithem

ulsion

and

injection

at100%

load.

BP: 1.87 kW

Die-Inj

(0.5)

Die-em

ul (0.1)

Diesel (0.3)

0

0.1

0.2

0.3

0.4

0.5

0.6

SMOKE (BSU)Fig.12.C

omparison

ofsm

okew

ithem

ulsion

and

injection

at40%

load.

Table 3Comparison of present work with other researcher’s work reported in literatures.

Researcher’s name Water–diesel emulsion/water injection Brake thermal efficiency Emission characteristics Combustioncharacteristics

Low load High load CO HC NO Smoke

Author (presentwork)

Water–diesel emulsion (0.4) and water injection Decrease Increase Increase Increase Decrease Decrease Available

Murayama et al. [2] Water–diesel emulsion (up to 0.8 by mass) Decrease Increase Increase Increase Decrease Decrease Limited informationBertrand [23] Upto 35 vol% in emulsion No

informationNoinformation

No information Increase Decrease Decrease No information

Park et al. [16] Water–diesel emulsion: 0%, 20% and 40% Noinformation

High No information No information Noinformation

Noinformation

No information

Nazha et al. [12] Water–diesel emulsion: 20% + EGR ; 16.7% Noinformation

Noinformation

Increase Increase Decrease Decrease No information

Coon [25] Water–diesel emulsion: 0–25% Noinformation

Increase Increase Increase Increase No report No information

Afify [26] Water–diesel emulsion: 15%, 30% and 45% byvolume

Noinformation

Noinformation

Increased at highload,decreases at low load

Increased at highload,decreases at low load

Increased Increased No information

Sheng et al. [5] Water–diesel emulsion: 0–20% by volume Noinformation

Noinformation

No information No information Decrease Decrease No information

ease Decrease Decrease No information Decrease Decrease No informationase Increase No information No information No

informationNoinformation

No information

mationNoinformation

Increase Increase Decrease No change No information

to base diesel.

K.A

.Subramanian

/EnergyConversion

andM

anagement

52(2011)

849–857

855

Nadeem [7] Water–diesel emulsion: 5–15% DecrAbu-Zaid [11] Water–diesel emulsion: 0%, 5%, 10%, 15% and 20%) Incre

Odaka et al. [19] EGR:17% + water injection 35 g/kg of air Noinfor

Note: Increase or decrease of performance and emission characteristics as compared

50

55

60

65

70

75

80

0 1 2 3 4 5Brake Power (kW)

Peak

Pre

ssur

e (b

ar)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 14. Comparison of peak pressure with emulsion and injection.

0

2

4

6

8

10

12

0 1 2 3 4 5Brake Power (kW)

Max

.Rat

e of

Pre

ssur

e R

ise

(bar

/ca)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 15. Comparison of maximum rate of pressure rise with emulsion and injection.

80

120

ate

(J/d

eg-c

a)

dieselw/d:0.4-emulw/d:0.4-injec

856 K.A. Subramanian / Energy Conversion and Management 52 (2011) 849–857

tribution in this field of water–diesel emulsion but they did not re-port on water injection. It is well established that water–dieselemulsion could give beneficial results in NOx and smoke/PM reduc-tion but the associated problems of high CO, HC, rate of pressurerise, low BSFC at lower loads, etc. needs to be addressed. Some ofthe problems could be overcome using it along with other tech-niques such as hydrogen peroxide and diethyl ether [13,14]. Theresearchers followed different strategies on preparation ofwater–diesel emulsion using different surfactants, injection meth-od, type of emulsion such as water-in-oil, oil-in-water and threephase emulsion. There is no report available till now includingthe present paper on the effects of surfactants in emulsion on en-gine performance and emission characteristics and it needs to bestudied in future. In addition, emulsion has drawbacks of stabilityproblems, instantaneous control and variation of water quantitywith respect to load, etc. In these aspects, water injection getsmore important to address some problems. In case of water injec-tion, a very few information are available. Masahiro Ishida et al.and Matsuo Odaka et al. reported the benefits of water injectionon NOx emission reduction [18,19]. But CO and HC emission in-creases and no change in smoke emission. Water injection methodmay be an effective technique for NOx emission with penality ofother emissions. If NOx emission is a primary target, EGR may bea viable solution as it does not need any additional system. Thereis no information available on simultaneous reduction of NOx andsmoke emission using water injection. So the lack of informationon both methods in literatures, the research work was carried out.

3.4. Combustion characteristics of engine

The ignition delay is much higher with the emulsion as com-pared to the water injection (Fig. 13). With water injection thetemperature at the time of fuel injection will be lower than dieselvalues as water that is injected during the intake stroke will vapor-ize and cool the air. This will lead to an increase in the ignition de-lay. With the emulsion even though there is no change in the airtemperature at the time of injection, the presence of water alongwith diesel will increase the specific heat of the droplets (sincethe specific heat of water is higher than that of diesel). The dropletsize could also be different for the emulsion as compared to neatdiesel. These factors affect the ignition delay with the emulsion.The ignition delay with the emulsion is 11.7� btdc at 4.7 kW asagainst 9.7� btdc at the same output with water injection. The peakpressure and maximum rate of pressure rise are also higher withthe emulsion due to the high ignition delay (Figs. 14 and 15). Thusengine operation is rough with the emulsion. Premixed combus-tion phase increased as compared to diesel due to long ignition de-

6

8

10

12

14

16

18

0 1 2 3 4 5Brake Power (kW)

Igni

tion

Del

ay (c

a)

dieselw/d:0.4-emulw/d:0.4-injec

Fig. 13. Comparison of ignition delay with emulsion and injection.

320 340 360 380 400Crank Angle (deg)

-40

0

40

Hea

t Rel

ease

R

load : 80%

Fig. 16. Comparison of effect of injection and emulsion on heat release rates at 80%load.

lay and is slightly lesser with water injection than the emulsion. Incase of injection the diffusion combustion phase is higher thanemulsion due to lesser ignition delay as shown in Fig. 16.

K.A. Subramanian / Energy Conversion and Management 52 (2011) 849–857 857

4. Conclusions

The following conclusions are drawn based on experimental re-sults by comparing the two methods at the same water–diesel ra-tio of 0.4:1 as given below.

The brake thermal efficiency is reduced at all outputs belowdiesel values with water injection due to poor combustion. At highoutputs, the brake thermal efficiency with the emulsion is signifi-cantly above the values with water injection. It is even better thanbase diesel operation at full load. In the case of the emulsion thebrake thermal efficiency is below diesel values only at low loads.At full load, brake thermal efficiency is 30.6% with water injectionwhere as 32.6% with water–diesel emulsion at 0.4:1 water to dieselratio.

CO and HC levels are lower at low loads with water injection ascompared to the emulsion. But it is similar to the levels withwater–diesel emulsion at high loads.

Reduction in NO level is less significant with water injection ascompared to the emulsion at low loads. The emulsion is the moreeffective in reducing NO level at a given water to diesel ratio.Water injection also leads to a significant reduction in the NO levelat high outputs when the injected water quantity is high. Sinceboth the methods are studied mainly to control NO emission athigh outputs they seem to be equally effective on that count. NOlevels are 398 ppm, 477 ppm at 60% load and 645 ppm and643 ppm at 100% load with emulsion and injection respectively.

Smoke emission is lower with the emulsion than waterinjection. It was 2.7 BSU with neat water–diesel emulsion as com-pared to 3.2 BSU with water injection at full load.

The ignition delay is much higher with the emulsion as com-pared to water injection. The ignition delay with the emulsion is11.7� btdc at 4.7 kW as against 9.7� btdc at the same output withwater injection. The peak pressure and maximum rate of pressurerise are also higher with the emulsion due to the high ignitiondelay.

The diffusion combustion phase is prominent with water injec-tion than the emulsion.

On the whole water–diesel emulsion is more effective inimproving full load brake thermal efficiency and lowering NOand smoke levels. The method of water injection at the same waterto diesel ratio leads to lesser adverse effects on HC and CO levelsand also to better part load performance. However it is not as effec-tive as the emulsion in reducing smoke and NO levels at a givenwater to diesel ratio. Water–diesel emulsion results in higher igni-tion delays, peak pressures and rates of pressure rise. It can be con-cluded that the emulsion method has higher potential ofsimultaneous reduction of NO and smoke emissions at all loadsthan injection method.

Acknowledgement

The author is very thankful to Prof. A. Ramesh, Department ofMechanical Engineering, I.I.T. Madras for his suggestions duringthis research work.

References

[1] Miyauchi T, Mori Y, Yamaguchi T. Effect of steam addition on NO formation. In:15th Symposium (international) on combustion. The Combustion Institute,USA; 1981.

[2] Murayama Tadashi, Tsukahara Minoru, Morishima Yaushi, Miyamoto Noboru.Experimental reduction in NOx, smoke and BSFC in a diesel engine usinguniquely produced water (0–80%) to fuel emulsion. Society of AutomotiveEngineers. SAE paper no. 780224; 1978.

[3] Barnaud Frederic, Schmelze Pierre, Schulz Philippe. Aquazole™: an originalemulsified water–diesel fuel for heavy-duty applications. Society ofAutomotive Engineers. SAE paper no. 2001-01-1861; 2001.

[4] Subramanian KA, Ramesh A. A study on the use of water–diesel emulsions in aDI diesel engine. In: 2nd International SAE-India mobility conference. IITMadras, SAE paper no. 2001-28-0005; 2001.

[5] Sheng HZ, Chen L, Zhang ZP, Wu CK, An C, Cheng CQ. The droplet group microexplosions in water-in-oil emulsion sprays and their effects on diesel enginecombustion. In: Twenty fifth symposium (international on combustion/thecombustion institute); 1994. p. 175–181.

[6] Anna Lif, Krister Holmberg. Water-in-diesel emulsions and related systems.Adv Colloid Interface Sci 2006;123–126:231–9.

[7] Nadeem M, Rangkuti C, Anuar K, Haq MRU, Tan IB, Shah SS. Diesel engineperformance and emission evaluation using emulsified fuels stabilized byconventional and Gemini surfactant. Fuel 2006;85:2111–9.

[8] Kadota T, Yamasaki H. Recent advances in the combustion of water fuelemulsion. Prog Energy Combust Sci 2002;28:385–404.

[9] Cherng-Yuan Lin, Kuo-Hug Wang. Diesel engine performance and emissioncharacteristics as fuel. Fuel 2004;83:537–45.

[10] Ravikumar TS, Basar Paul D, Michael G.Jensen and Ken Friis Hansen, Emulsifieddiesel – an immediate and effective solution for diesel exhaust emissionreduction. Society of Automotive Engineers. In: 2nd International SAE Indiamobility conference. SAE paper no. 2001-28-0037; 2001.

[11] Abu Zaid M. Performance of single cylinder direct injection diesel engine usingwater fuel emulsions. Energy Convers Manage 2004;45:697–705.

[12] Nazha MAA, Rajakaruna H, Wagstaff SA. The use of emulsion, water inductionand EGR for controlling diesel engine emissions. Society of AutomotiveEngineers. SAE paper no. 2001-01-1941; 2001.

[13] Subramanian KA, Ramesh A. Use of hydrogen peroxide to improve theperformance and reduce emissions of a CI engine fuelled with water dieselemulsion. Society of Automotive Engineers. SAE paper no. 2008-01-0653: SP-2108; 2008.

[14] Subramanian KA, Ramesh A. Use of diethyl ether along with water–dieselemulsion in a DI diesel engine. Fuels and lubricants. SAE transactions. SAEpaper no. 2002-01-2720. 2002;111(4):1361–1367.

[15] Subramanian KA, Ramesh A. Experimental Investigation on the use of water–diesel emulsion with oxygen enriched air in a DI diesel engine. Society ofAutomotive Engineers. SAE paper no. 2001-01-0205: SP-1592; 2001.

[16] Park JW, Huh KY, Park KH. Experimental study on the combustion reduction tocontrol exhaust emissions from heavy-duty diesel powered vehicles. Society ofAutomotive Engineers. SAE paper no. 972961; 1997.

[17] Tsukahara Minoru, Yoshimoo Yasufumi, Murayama Tadashi. W/O emulsionrealizes low smoke and efficient operation of DI engines without high pressureinjection. Society of Automotive Engineers. SAE paper no. 890449; 1989.

[18] Ishida Masahiro, Ueki Hironoku, Sakauguo Daisaku. Prediction of NOx

reduction to control exhaust emissions from heavy-duty diesel poweredvehicles. Society of Automotive Engineers. SAE paper no. 972961; 1997.

[19] Odaka Matsuo, Koike Noriyuki, Tsukamoto Yujiro, Kazuysawa, Yoshida Koichi.Effects of EGR with a supplemental manifold water injection to controlexhaust emissions from heavy-duty diesel powered vehicles. Society ofAutomotive Engineers. SAE paper no. 910739; 1991.

[20] Sarvi Arto, Kilpinen Pia, Zevenhoven Ron. Emissions from large-scale medium-speed diesel engines: influence of direct water injection and common rail. FuelProcess Technol 2009;90:222–31.

[21] Hohenberg GF. Advanced approaches for heat transfer calculation. Society ofAutomotive Engineers. Paper no. 790825.

[22] Tsao KC, Wang CI. Puffing and micro-explosion phenomena of water emulsionfuels. Society of Automotive Engineers. Paper no. 860305.

[23] Bertrand D Hsu. Combustion of water-in-diesel emulsion in an experimentalmedium speed diesel engine. Society of Automotive Engineers. Paper no.860300.

[24] Muller-dethlefs K, Schlader AF. The effect of steam on flame temperature,burning velocity and carbon formation in hydrocarbon flames. Combust Flame27:205–15.

[25] Coon CW. Multi-cylinder diesel engine tests with unstablised water-in-fuelemulsions. Society of Automotive Engineers. SAE paper no. 810250.

[26] Afify EM, Korah NS, Dickey DW. The effect of air charge temperature onperformance, ignition delay and exhaust emissions of diesel engines using W/O emulsion as fuel. Society of Automotive Engineers. SAE paper no. 870555.