Energy Conversion and Management 75 (2013) 141–151

Contents lists available at SciVerse ScienceDirect

Energy Conversion and Management

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

Waste heat recovery from the exhaust of a diesel generator usingRankine Cycle

0196-8904/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.enconman.2013.06.009

⇑ Corresponding author. Tel.: +61 8 8302 5123; fax: +61 8 8302 3380.E-mail address: shekh.rubaiyat@unisa.edu.au (S.N. Hossain).

Shekh Nisar Hossain ⇑, Saiful BariBarbara Hardy Institute, School of Engineering, University of South Australia, Mawson Lakes Campus, SA 5095, Australia

a r t i c l e i n f o

Article history:Received 10 March 2013Accepted 7 June 2013

Keywords:Waste heat recoveryRankine CycleDiesel generationHeat exchanger

a b s t r a c t

Exhaust heat from diesel engines can be an important heat source to provide additional power using aseparate Rankine Cycle (RC). In this research, experiments were conducted to measure the availableexhaust heat from a 40 kW diesel generator using two ‘off-the-shelf’ heat exchangers. The effectivenessof the heat exchangers using water as the working fluid was found to be 0.44 which seems to be lowerthan a standard one. This lower performance of the existing heat exchangers indicates the necessity ofoptimization of the design of the heat exchangers for this particular application. With the available exper-imental data, computer simulations were carried out to optimize the design of the heat exchangers. Twoheat exchangers were used to generate super-heated steam to expand in the turbine using two orienta-tions: series and parallel. The optimized heat exchangers were then used to estimate additional powerconsidering actual turbine isentropic efficiency. The proposed heat exchanger was able to produce 11%additional power using water as the working fluid at a pressure of 15 bar at rated engine load. This addi-tional power resulted into 12% improvement in brake-specific fuel consumption (bsfc). The effects of theworking fluid pressure were also investigated to maximize the additional power production. The pressurewas limited to 15 bar which was constrained by the exhaust gas temperature. However, higher pressureis possible for higher exhaust gas temperatures from higher capacity engines. This would yield moreadditional power with further improvements in bsfc. At 40% part load, the additional power developedwas 3.4% which resulted in 3.3% reduction in bsfc.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Compression Ignition engines, also known as diesel engines, area major type of Internal Combustion (IC) engines. These diesel en-gines have a wide field of applications and frequently used becauseof their higher thermal efficiency. Trucks, buses and earth movingmachineries use high speed diesel engines and output of these en-gines can be as high as 740 kW. Diesel engines are also used insmall power generating units or as standby units for mediumcapacity power generations. Power generation using diesel enginesbecame popular in the last four decades. The main applications ofthese diesel generators are auxiliary or backup power plants inhospitals, airports, hotels and industries those need to ensure reli-able power supply at all times. Engine based power production to-day represents some 10–15% of the total installed capacity in theworld [1].

A brief analysis of heat balance of a diesel engine indicates thatthe input fuel energy is divided into three major parts: energy thatconverts to useful work, energy that loses through the exhaust gasand energy that dissipates to the coolant. In general, diesel engines

have a thermal efficiency of about 35% and thus the rest of the in-put energy is wasted. A considerable amount of energy is expelledto the ambient environment with the exhaust gas despite recentimprovement of diesel engine efficiency. In a water-cooled engineabout 25% and 40% [2] of the input energy are wasted into the cool-ant and exhaust gases, respectively. Johnson [3] found that the to-tal waste heat dissipated can vary from 20 kW to as much as 40 kWfrom a typical 3.0 l engine having a maximum output power of115 kW. It is also suggested that for a typical and representativedriving cycle, the average heating power available from the wasteheat is about 23 kW.

Due to strict regulations on polluting emissions and energy sav-ings, diesel engine is being an object of intensifying research to im-prove its thermal efficiency and to make it more environmentallyfriendly. The thermal efficiency of a diesel engine can be increasedby improving the thermodynamic efficiency of the operating cycleand/or reducing the mechanical losses [4,5]. These techniques re-sult in a reduction in the brake-specific fuel consumption (bsfc),but it appears that the potential for further improvement is limited[6]. An attractive alternative option for further improvement ofbsfc and reductions of specific polluting emissions can be wasteheat recovery (WHR). There are several WHR technologies avail-able and the dominating ones are:

Nomenclature

e heat exchanger effectiveness_m mass flow rate, kg/s

P power, kWgts turbine isentropic efficiencyh specific enthalpy, kJ/kge specific exergy, kJ/kgu specific internal energy at exhaust temperature, kJ/kgu0 specific internal energy at dead state temperature, kJ/kgp0 dead state pressure, barp exhaust pressure at the end of combustionv specific volume at exhaust temperature, m3/kgv0 specific volume at dead state temperature, m3/kgR universal gas constant, kJ/mol KM molecular weight of exhaust gas, kg/mols0(T) absolute entropy of ideal gas at T temperature, kJ/kg Ks0(T0) absolute entropy of ideal gas at T0 temperature, kJ/kg K_Ed exergy destruction rate, kW

cp specific heat at constant pressure, kJ/kg Ku, v, w velocity components in x, y and z direction, respectively,

m/sT temperature, �Ck conductivity coefficient at bulk temperature of fluid, W/

m Kq cold fluid density, kg/m3

s shear stress, N/m2

Subscriptsc cold fluidh hot fluid1 hot fluid inlet2 hot fluid outlet3 cold fluid inlet4 cold fluid outlet

142 S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151

– Mechanical turbo-compounding: an additional power turbine isutilized in the downstream of the turbocharger and it ismechanically coupled to the engine crank shaft by a gear trainto increase the engine power output [4,6]. By using mechanicalturbocompounding, Caterpillar [7] reported an average bsfcreduction of 4.7% for a 50,000 mile extra urban driving test inthe USA. They used an axial-flow power turbine on a 14.6 l die-sel engine. Scania [7] also applied the same technology in a 11 l,six cylinder turbo charged diesel engine and found 5% improve-ment in bsfc at full load. Hountalas et al. [6] performed enginesimulation to study the performance of mechanical turbocom-pounding and reported 4.8% improvement in bsfc at full engineload.

– Electrical turbo-compounding: a system that converts wasteexhaust energy to shaft work using a turbine which is coupledto an electric generator [8,9]. The electricity produced fromthe generator is used to run a motor fitted to the engine crankshaft. Caterpillar has considered this concept in their researchprogramme [9,10] providing an indication of 5% reduction inbsfc whereas Hountalas et al. [6] found 3% bsfc reduction usingthe same principle.

– Thermoelectric system: this technology directly converts a por-tion of the exhaust gas heat to electrical power through thermo-electric phenomenon without the utilization of mechanicalcomponents [6,11]. Due to the low conversion efficiency ofthe system ranging from 6% to 8% [12], the thermoelectric sys-tem can improve the bsfc by less than 1%. Since the cost of thethermoelectric semiconductor materials is relatively high, thethermoelectric system is not yet suitable for practicalapplications.

– Turbocharger: a turbocharger is an exhaust gas driven super-charger. In this system, the available energy in the engine’sexhaust gas is used to run the turbine of the turbocharger whichruns the compressor to increase the inlet air density. The clearobjective of the turbocharger is to increase the specific poweroutput and torque of the engine [4,13,14] with no or slightreductions of bsfc which is less than 1%. Even a turbochargeddiesel engine still rejects 35–40% of input energy through theexhaust gas [15,16]. Because of this fact, applications of WHRcan also be found in turbocharged engines [4,17] as well.

– Rankine Cycle (RC): a steam generator is employed to generatesteam using the exhaust heat which is expanded in a turbineto produce additional power. If an organic fluid is used insteadof water then the system is called Organic Rankine Cycle (ORC).

Hountalas et al. [6] performed engine simulation to study a RCbased WHR system and reported 9% improvement in bsfc at fullengine load. They utilized the heat from both Exhaust Gas Recir-culation (EGR) and Charge Air Cooler (CAC). Recently, Hountalaset al. [18] reported 5.5% bsfc improvement utilizing heat fromthe exhaust gas of a Heavy Duty (HD) truck diesel engine usingRC based WHR system. Katsanos et al. [19] did a theoreticalstudy to investigate the potential improvement of the overallefficiency of a HD truck diesel engine equipped with a bottom-ing RC to recover heat from the exhaust gas. They were able toimprove the bsfc by 7.5% at full load of the engine using wateras the working fluid. Likewise, Dolz et al. [20] was able toachieve 8.5% improvement in bsfc using exhaust heat utilizinga RC based WHR system.

From the aforesaid analyses of different WHR technologies, it isclear that mechanical turbo-compounding and RC based WHR sys-tems are the most promising candidate for improving bsfc of dieselengines. However, increase in engine backpressure and pumpinglosses [21] are the fatal disadvantages of mechanical turbo-com-pounding. Consequently due to these disadvantages, mechanicalturbo-compounding system is not widely used. Between RC andmechanical turbo-compounding systems, Weerasinghe et al. [22]made a numerical simulation comparing their power output andfuel savings. Their results revealed the relative advantages of RCover mechanical turbo-compounding. According to their study,7.8% power was recovered by using RC system whereas only 4.1%power was recovered by using mechanical turbo-compoundingsystem. Vaja and Gambarotta [23] performed thermodynamicanalysis to investigate WHR systems using RC into stationary die-sel engine and reported 6% improvement in bsfc utilizing heat fromboth exhaust and coolant.

With the exception of mechanical turbo-compounding, RC- andORC-based heat recovery systems need to utilize heat exchangersto extract energy from the waste heat. The heat exchanger designis critical as it needs to provide an adequate surface area in order tocope with the thermal duty. The pressure loss across the heatexchangers also needs to be reasonable to avoid back pressure thatwill have a negative impact on the net engine power and efficiency.These are the challenges need to be investigated to design an effec-tive heat exchanger to extract heat efficiently with low pressuredrop from the exhaust of the diesel engines. The pressure of theworking fluid also needs to be investigated to find out the opti-mum pressure for any particular application. However, studies of

S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151 143

literature reviews show that the optimization of heat exchanger interms of design and pressure of the working fluid for RC-basedWHR system are scarce.

In this research, experiments were conducted to measure theexhaust heat available from a diesel engine coupled to a generatorto produce electricity. Two identical shell and tube heat exchang-ers were used to produce super-heated steam. These heat exchang-ers were purchased in the marketplace and installed into theengine exhaust system. Although turbine was not installed in thisexperiment, the available data sufficed to estimate additionalpower using actual turbine efficiency [24–26]. The purchased heatexchangers did not performed optimally during the experiment asthey were purchased from the marketplace and was not designedoptimally for this particular application. Hence, research was con-ducted to design heat exchangers and also find the appropriatepressure of the working fluid to operate optimally for this WHRapplication using commercial CFD software CFX. As mentioned be-fore, most researchers did not investigate these two parameters forWHR applications. Two optimum heat exchangers were designedby varying different important geometrical aspects of the heatexchangers such as shell diameter, length, number and diameterof tubes. To validate the simulation model, the existing heatexchangers were modeled first and then compared with experi-mental results. The steam utilized for this purpose would be re-quired to be super-heated before expanding in the turbine toavoid any condensation in the turbine. So one heat exchanger,explicitly named as ‘vapor generator’ was utilized to produce vaporfrom the liquid and the second heat exchanger named as ‘superheater’ was employed to create super-heated vapor. The authorsbelieve that the use of two heat exchangers to extract heat fromthe exhaust is a new possibility for the improvement of the perfor-mance of the WHR system. These two heat exchangers were usedin two arrangements (series and parallel) and this is a novel idea toinvestigate the performance of the WHR system. These concepts toimprove the performance of WHR system was not found in any lit-erature. Finally, the maximum additional power achievable by theturbine using these heat exchangers was estimated and therebyjustifying the use of WHR to produce additional power and therebyimproving bsfc using RC. Initially, the design and the performanceof the WHR system were optimized at the rated load of the engine.Thereafter, the performance of the WHR system was also investi-gated at part loads as the engine may need to run at part loads.The working pressure at part loads was also optimized in this re-search which is rare in literatures.

2. Experimental setup

The engine used in this study was a 4-stroke, 4-cylinder, watercooled HINO W04D diesel engine which was coupled to a 50 kV Agenerator. The specifications of the engine are presented in Table 1.A 90 kW resistive type load bank was used to load the diesel gen-erator. The schematic of the experimental setup is shown in Fig. 1.

To measure the inlet air-flow rate to the engine a nozzle wasused. An inclined manometer was used to measure the pressuredifference across the nozzle. The accuracy of the manometer was±0.01 kPa. For fuel measurement, a digital weighing scale with anaccuracy of ±1 g and a stop watch were used. Exhaust mass flow

Table 1Engine specifications.

Engine make and model HINO W04DType of engine 4 cylinder water cooled diesel engineBore 104 mmStroke 118 mmCompression ratio 17.9:1Combustion and aspiration Direct injection, naturally aspirated

rate was calculated from the measured air and fuel mass flow rates.The temperatures at different points were measure by K type ther-mo couples. Burdon tube pressure gauges were used to measurethe cold and hot fluids side pressures of the heat exchangers. ADwyer model VFA variable area flow meter was used to measurethe water flow rate into the heat exchangers. The flow meter hadan accuracy of ±5%. Engine speed was measured by a digitaltachometer with an accuracy of ±1 rpm.

At first, baseline experiments were conducted with the dieselgenerator without installing the heat exchangers. The diesel gener-ator was tested at different loads at a constant speed of 1500 rpm.The exhaust temperatures and air flow rate were recorded to calcu-late the available heat energy from the exhaust. Then, the exhaustof the engine was connected to the shell and tube heat exchangersto measure the recoverable energy. Water mass flow rate, water in-let and outlet temperatures, and pressures at different points werealso recorded to calculate the effectiveness of the heat exchangers.Then, these data were utilized to model the existing heat exchang-ers and simulations were carried out and validated with the exper-imental results. The hot fluid (the exhaust from the engine) waspassed through the tubes of the heat exchangers and the cold fluid(water) was flowed through the shell side. Counter flow heat ex-changer orientation was selected for this study.

Error and uncertainty analyses are important issues in experi-mental investigations. The error and uncertainty in the experi-ments are initiated from the selection of instruments, operationalconditions, calibrations, readings, and test planning. The accuraciesof the measuring instruments are already mentioned in the preced-ing paragraphs. Accuracy of the experiment can be ascertained byuncertainty analysis. The uncertainty analysis for the currentexperiment was carried out using the method described by Hol-man [27] and the calculated uncertainties are shown in Table 2.The uncertainty values were within the acceptable limit as foundby other researchers [28–30].

3. Heat exchanger design methodology

As stated earlier, two shell and tube heat exchangers, having90 mm shell diameter with 1 m length and 15 tubes of 15 mmdiameter, were purchased in the marketplace. These heat exchang-ers were installed into the exhaust system and experiments wereperformed to determine the additional power conceivable by thesystem. Being ‘off-the-shelf’, these heat exchangers might not bethe best designed one for this specific application to produce themaximum power. Therefore, various attempts were undertakento optimize the heat exchangers that could potentially generatemaximum additional power. Computational Fluid Dynamics(CFD) techniques were utilized to carry out simulations of the cur-rent heat exchangers using the dimensions of the existing heatexchangers. After obtaining sufficient agreement of the simulationresults with the experimental results, the influences of various de-sign parameters of the heat exchanger on the performance of theWHR system were examined utilizing actual turbine efficiency[24–26]. The steam utilized for this purpose needs to be super-heated before it expands in the turbine. Therefore, one heat ex-changer, explicitly named as the ‘vapor generator’, was utilizedto produce vapor from the liquid and the second heat exchangernamed as the ‘super heater’ was employed to create super-heatedvapor. These two heat exchangers could be positioned in two con-figurations, parallel and series arrangements, as illustrated inFig. 2. In the instance of parallel arrangement, the exhaust fromthe engine was divided into two streams; one stream was traveledinto the vapor generator and the other one into the super heater.Contrariwise, in the series arrangement, the exhaust from the en-gine traveled into the super heater first and then into the vaporgenerator as a single stream.

Air Box

Fuel Tank

Fuel consumption Meter

Generator

Diesel Engine

Cooling System

Inclined Manometer

Flow meter

Thermo couple

Pressure gauge

Legend

Resistive Load bank

Super Heater

Water in

Vapor Generator

Steam Out

Fig. 1. Schematic diagram of the experimental setup.

Table 2Uncertainties of measured values.

Name of the measuredvalues

Uncertainty(%)

Mass flow rate ±0.69Power ±0.20Bsfc ±0.695Thermal efficiency ±0.229Temperature ±0.21Effectiveness ±2.1

144 S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151

4. Theoretical background

4.1. Heat exchanger effectiveness

During optimization of the heat exchanger, effectiveness of theheat exchanger was considered as the performance parameter ofthe heat exchanger. Important geometrical aspects of the heat ex-changer were varied to achieve maximum effectiveness. Effective-ness of the heat exchanger can be defined as the ratio of the actualheat transfer in a given heat exchanger to the maximum possiblerate of heat transfer. Mathematically:

e ¼ Actual heat transferMaximum possible heat transfer

or e ¼ ð_mcpÞhðTc;out � Tc;inÞð _mcpÞminðTh;in � Tc;inÞ

ð1Þ

4.2. Thermodynamic cycle

In the present analysis, Rankine Bottoming cycle is consideredto recover heat from the exhaust of a diesel engine. A waste heatrecovery system using Rankine bottoming cycle consists of a pumpto circulate and increase the pressure of the working fluid, heatexchangers to absorb the heat from the exhaust gas and to gener-ate super-heated vapor, an expander, i.e. turbine to extract powerby bringing the fluid to a lower pressure level, and a condenser toliquefy the vapor before restarting the whole cycle. The configura-tion of this cycle and a T–s diagram is shown in Fig. 3a and brespectively. The power generated by a turbine is calculated fromthe enthalpy drop across the points 3–4 as shown in Fig. 3b consid-ering actual isentropic efficiency of the turbine employing the fol-lowing equation:

P ¼ gts _mcDh ð2Þ

4.3. Exergy analysis

To estimate the energy availability of the exhaust from the die-sel engine, exergy analysis was carried out. During the exergy anal-ysis the following assumptions are considered: (1) the combustionproducts are modeled as air assuming it to be an ideal gas, (2) theeffect of motion and gravity is ignored and (3) the reference envi-ronmental (dead state) temperature and pressure are 50 �C and1.013 bar respectively. The specific exergy of the exhaust was cal-culated by using the following equation [25]:

1st heat exchangerVapour to super heater

2nd heat exchangerLiquid to vapour

1st heat exchangerVapour to super heater

Hot exhaust in

Hot exhaust out

Hot exhaust outHot exhaust in

Series Arrangement

Parallel Arrangement

Water in

2nd heat exchangerLiquid to vapour

Water in

SteamSuper heated steam

Super heated steam Steam

Fig. 2. Heat exchangers arrangements.

S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151 145

e ¼ u� u0 þ p0ðv � v0Þ � T0ðs� s0Þ ð3Þ

where

p0ðv � v0Þ ¼RM

p0Tp� T0

� �ð4Þ

ðs� s0Þ ¼ s0ðTÞ � s0ðT0Þ �RM

lnpp0

ð5Þ

It is also important to calculate the exergy destruction which canidentify the irreversibility in the system. Elimination or reductionsof this will improve the performance of the system. The fluid frictionand temperature difference are the main factors which lead to exer-gy destruction in any system [25]. In this research, the pressuredrops across the heat exchangers is negligible compared to the tem-perature differences. Therefore, in the current study of WHR systemwith heat exchangers, the temperatures difference across the ex-haust gas and water/steam is the main reason of exergy destruction.During exergy calculation the two heat exchangers were consideredas a single control volume as shown in Fig. 4. Now, corresponding toFig. 4, employing exergy balance, the combined destructed exergy iscalculated using the following equation [25]:

_Ed ¼ _mh ðh1 � h2Þ � T0 s01 � s0

2 � R lnp1

p2

� �� �þ _mc ðh3 � h4Þ � T0ðs3 � s4Þ½ � ð6Þ

5. Modeling details

5.1. Computational model

As mentioned previously, the purchased heat exchangers weremodeled and simulations were carried out first to obtain satisfac-tory results. Then, different geometrical aspects of the heat ex-changer were optimized. The existing heat exchanger modeldrawing was created by Computer Aided Design (CAD) softwareSolidWorks 2012. Fig. 5a and b shows the front and cross-sectionalviews of the model respectively. In the model, 30� triangular stag-gered array layout was used for the tube arrangement of the heatexchanger. The geometrical model was then meshed using ANSYSmeshing software. The ANSYS CFX 14.0 was used to solve the equa-tions for the fluid flow and heat transfer. The CFD code ANSYS CFX14.0 is based on the finite volume method. It is a high-perfor-

mance, general purpose fluid dynamics program that has been ap-plied to solve wide-range of fluid flow and heat transfer problems.

5.2. Meshing

In order to make the simulation more accurate, different meshingschemes were used. The solid tubes were meshed using sweep meshwhereas the fluid volumes were meshed using tetragonal-hybrid ele-ments. The final refined mesh was selected by comparing the simu-lation results of model with different mesh density and meshingschemes. The final model has 17,641,872 elements and 7,161,123nodes and grid independent solution was acquired. Fig. 6a and bshows the computational domain of the heat exchangers.

5.3. Governing equations

The numerical simulation was performed with three dimen-sional, steady-states, turbulent flow system. To solve the problem,k–x based Shear–Stress–Transport (SST) turbulent model [31] wasemployed and energy equation was included into the model. Thegoverning equations for the flow and conjugate heat transfer weremodified according to the conditions of simulation setup. As theproblem was assumed to be steady, the time dependent parame-ters are dropped from the equations. The resulting equations are:

Continuity equation : r � ðq~VÞ ¼ 0 ð7Þ

Momentum equations:

x-momentum : r � ðqu~VÞ ¼ � @p@xþ @sxx

@xþ @syx

@yþ @szx

@zð8Þ

y-momentum : r � ðqv~VÞ ¼ � @p@yþ @sxy

@xþ @syy

@yþ @szy

@zþ qg ð9Þ

z-momentum : r � ðqw~VÞ ¼ � @p@zþ @sxz

@xþ @syz

@yþ @szz

@zð10Þ

Energy : qcp u@T@xþ v @T

@yþw

@T@z

� �¼ k

@2T@x2 þ

@2T@y2 þ

@2T@z2

!ð11Þ

5.4. Boundary conditions

For the heat exchanger model, water/steam was referred to asthe cold fluid and exhaust gas from the engine was referred to asthe hot fluid. The cold fluid was considered to be in liquid phase

Q out: to cold source

Q in: from hot source

W inW out,

useful work

PumpCondenser

Heat exchanger Turbine

1

2 3

4

Q out: to cold source

Q in: from hot source

W in

W out, useful work

S

T

1

2

3

4

Fig. 3. Rankine Cycle (a) component (b) T–s diagram.

Hot exhaust inHot exhaust out

Heat exchangers

1 2

3Water in

4Steam out

Control volume for two heat exchangers

Fig. 4. Exergy balance for heat exchangers control volume.

146 S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151

Fig. 5. Heat exchanger (a) front view (b) cross-sectional view.

Fig. 6. Mesh of computational domain (a) side view (b) cross-sectional view.

S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151 147

at 28 �C at selected pressure and it came out as vapor from thevapor generator. It then entered into the super heater as vaporand exited as super-heated vapor at the same selected pressure.The hot fluid was modeled as air with mass flow rate of0.050253 kg s�1 and temperature of 479 �C which were found fromthe experiment at the engine power of 26.57 kW. The operatingpressure of hot fluid is set at 101.325 kPa and both the workingpressure and mass flow rate of the cold fluid were varied.

Fig. 7. Measured exhaust gas temperature and calculated exergy variations withengine power.

6. Results and discussions

6.1. Experiments for initial assessment

To design an effective heat exchanger for heat recovery from theexhaust of a diesel engine, it is required to know how much energy

is available in the exhaust. Therefore, some base line tests wereperformed without installing the heat exchangers. Thereafter, thepurchased heat exchangers were installed into the exhaust of theengine and the performance of the WHR system was investigated.In the following sections these results are analyzed.

6.1.1. Exhaust heat and exergy analysisThe exhaust gas temperature and exergy at various engine pow-

ers is presented in Fig. 7. It is found from the figure that the ex-haust gas temperature and the engine power showed anapproximate exponential relationship. At 26.57 kW the fuel flowrate was 6.48 kg/h which resulted in 479 �C exhaust temperature.Whereas, at a higher power of 33.9 kW, the engine consumed9.0 kg/h of fuel which produced 669 �C exhaust temperature. Morefuel was combusted at higher power which resulted in higher ex-haust gas temperature. Similar trend was also observed by Rama-dhas et al. [32] in their work.

In order to find out the maximum theoretically obtainable en-ergy from the exhaust, exergy analysis can be utilized. Exergy anal-ysis usually aims to determine the maximum performance of athermodynamic system and it has proven to be a powerful toolin thermodynamic analysis of energy systems [33]. Exergy can bedefined as maximum theoretically obtainable work when the sys-tem comes to equilibrium with a reference environment. Defini-tion of exergy for energy in various forms can be found in theworks of Baehr [34] and Ahern [35]. It is found from Fig. 7 thatexergy also increased exponentially with the engine power. Soboth temperature and exergy relationships with power indicatethat the heat recovery is more effective at higher powers of theengine.

It is evident from Fig. 7 that the highest values of both exhaustgas temperature and specific exergy at 33.49 kW were 669 �C and400 kJ/kg, respectively. However, it is found from Fig. 8 that themaximum thermal efficiency and minimum bsfc were found at26.57 kW power. Engine can run continuously at this rated powercondition. Therefore, this point was chosen to model the design ofthe heat exchangers and carry out the simulations. The maximumpower of 33.49 kW was not selected because this is in the overloadregion and usually diesel generator does not operate in this regioncontinuously [36,37].

The total exergy can be calculated from the specific exergyfound in Fig. 7 by multiplying the mass flow rate of the exhaustgas which indicates approximately 30% of the engine’s power iscurrently being wasted, but could be recoverable and convertedto usable form. Similar relationship between exergy and enginepower was reported in the work of Teng et al. [38]. Hung et al.[39] and Larjola [40] were able to recover more than half of theexergy in the high temperature waste heat using an ORC system.When considering heat recovery from this exhaust to produce

Fig. 8. Engine performance curve. Fig. 10. Bsfc improvement with and without existing heat exchangers.

148 S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151

addition power to improve bsfc, there will be some obvious lossesdue to friction, heat loss and irreversibility. In addition, there willbe another drop of power due to heat exchanger effectivenessand turbine efficiency. Considering all these factors, there is stillthe possibility of improving bsfc by more than 10% [22,38,41].

6.1.2. Existing heat exchanger effectivenessAfter having initial estimation of the available energy in the ex-

haust gas, experiments were conducted to extract heat from theexhaust gas with the two existing heat exchangers. The perfor-mance of the heat exchangers was investigated at different powersof the engine which is presented in Fig. 9. The effectiveness wasfound to be 0.44 at 26.43 kW power. It is found from the figure thatthe effectiveness of the heat exchanger increased with power. Asdescribed in the previous section that the exhaust temperaturealso increased with power. The higher exhaust temperature in-creased the heat transfer between the hot and cold fluids resultingin higher effectiveness.

However, it is found from extensive literature review that theeffectiveness of the shell and tube heat exchanger is reported tobe in the range of 0.7 and 0.99 [42,43]. Therefore, the effectivenessof the existing heat exchangers is significantly lower. This findingis one of the motivations of this research to optimize the designof the heat exchanger to recover heat more effectively. Althoughthe turbine was not installed in the experimental setup, the addi-tional power attainable by the existing heat exchangers was esti-mated by using actual turbine efficiency [24,25] taking intoaccount pumping power required by the working fluid. The bsfcand improved bsfc of the engine without and with the existing heatexchangers are presented in Fig. 10. From the figure it is evidentthat the non-optimized heat exchangers managed to improve thebsfc by 3.1% at 26.43 kW. However, Dolz et al. [20] achieved amuch higher value of 8.5% improvement in bsfc by using a RC

Fig. 9. Variation of heat exchanger effectiveness at different power (experimental).

based WHR system. Therefore, it would be interesting to find outthe improvement in bsfc after optimization of the design of theheat exchangers and other parameters like working fluid pressurewere carried out in this research work.

6.2. Modeling of the existing heat exchanger and validation

In the aforementioned sections, it was found that the experi-mental effectiveness of the existing heat exchangers was signifi-cantly lower than the effectiveness reported in differentliteratures. Therefore, based on the available data from the exper-iment, the existing heat exchanger was modeled and simulationswere carried out first to construct the appropriate model whichwas validated by the experimental results. Experimental data ofthe engine running at 1500 rpm and 26.57 kW was selected for thispurpose.

The variation of effectiveness with the power found from CFDsimulation is reported in Fig. 11 along with the experimental val-ues from Fig. 9 for comparison purposes. It is clear from the figurethat the effectiveness found from the experiment was approxi-mately 10% lower than the CFD simulation results. This could bedue to the presence of fouling sources in both hot and cold fluidsin the experiments. The presence of soot and particles in the ex-haust gas increased the resistance of the heat transfer inside thetubes and the untreated water that was used in the experimentcaused fouling effect on the shell side. These effects were not con-sidered in the CFD simulation. Due to these effects, the shell sideand tube side heat transfer coefficients decreased [44] resultinglower effectiveness in the case of experiments. Lei et al. [45] alsoreported an 8% discrepancy between CFD simulations and experi-mental results. Therefore, it can be concluded that the modelwas appropriately constructed and the experimental and simula-tion results were in adequate agreement.

Fig. 11. Comparison of simulation results validation with experimental results.

S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151 149

6.3. Optimization of the heat exchanger

Based on the available data from the experiment, the heat ex-changer design was optimized by computer simulation. Experi-mental data of the engine running at 1500 rpm and 26.57 kWwas selected for this purpose. The important geometrical aspectsof the heat exchanger were investigated to increase the effective-ness of the heat exchangers. These parameters were also examinedat different working pressures of the cold fluid.

As surface area between hot and colds fluid plays an importantrole to transfer heat, in this study the effect of the number of tubeson the effectiveness of the heat exchanger is examined first. Thenumber of tubes was increased inside the initial diameter(90 mm) of the shell by reducing the tube diameter keeping theinitial diameter and length of the shell constant. It was found thatthe effectiveness of the heat exchanger increased as the number oftubes increased. This is due to the increase in surface area of thetubes. The additional heat transfer surfaces resulted in more heattransfer and as a result higher effectiveness of the heat exchangerwas obtained. On the other hand, the shell side fluid velocity alsoincreased with decreasing effective area of the shell side due tothe increase in number of tubes. This higher velocity increasedthe shell side heat transfer coefficient which facilitated heat trans-fer [46]. Consequently, due to higher surface area and higher heattransfer coefficient the effectiveness of the heat exchanger in-creased with number of tubes. Optimum number of tubes wasfound to be 31 for 90 mm shell diameter.

The next important parameter that was optimized was thelength of the heat exchanger. During this optimization, the numberof tubes (31) and shell diameter (90 mm) were kept constant andlength was increased gradually. In the case of longer heat exchan-ger, the hot and cold fluids get more residence time for heat trans-fer. As a result effectiveness increases with the length. Conversely,as the length increases, the heat loss to the environment also in-creases and the heat gain by the cold fluid could be offset by thisloss. Because of this physical phenomenon, the effectiveness ofthe heat exchanger increases with length up to a certain valueand after that it does not increase effectively or even sometime de-creases. Similar behavior was observed in this study. It was foundthat after 2 m length, the effectiveness slightly decreased. There-fore, the 2 m length was selected for the heat exchanger to performfurther analysis.

The effect of shell diameter was investigated next after optimiz-ing the number of tubes and length of the heat exchangers. Thediameter was varied while keeping the number of tubes (31) andlength (2 m) constant. It was found that the effectiveness is higherfor smaller diameter shell. As the diameter of the shell increased,the effective velocity inside the shell decreased for a fixed flowrate. The lower velocity yielded lower heat transfer coefficient[46,47] which impeded the heat transfer between the cold andhot fluids. Thus, the effectiveness was lower for a higher shelldiameter. As the effectiveness decreased with increasing diameterof shell, the initial shell diameter of 90 mm was finally selected forthe proposed heat exchanger.

Table 3Heat exchanger model specifications.

Heat exchanger type Shell and tube counter flow,hot fluid in tubes and cold fluid in the shell

Shell inside diameter 90 mmNo of tube 31Tube arrangement 30� triangular staggered arrayTube pitch 15 mmTube inside diameter 9.44 mmBaffles 50% cut baffles, No. of baffle 7Length of the heat exchanger 2 m

After optimization, the final proposed heat exchanger has ashell diameter of 90 mm, 31 tubes of 9.4 mm diameter, and aneffective length of 2 m. The detailed specifications of the optimizedheat exchanger are presented in Table 3. The effectiveness of thisheat exchanger was found to be 0.76 which is a significantimprovement from the original 0.44.

6.4. Exergy destruction calculation

It is mentioned before that irreversibility in the system causesexergy destruction. Minimizing exergy destruction improves thesystem performance. The exergy destruction in the heat exchang-ers was calculated and presented in Fig. 12. For this calculation,Eq. (6) was used and the two heat exchangers were consideredas a single control volume. Therefore, the exergy destruction delin-eated in Fig. 12 is the combined exergy destruction of the two heatexchangers. Both optimized and non-optimized heat exchangerexergy destruction is compared in this figure. It is found that theexergy destruction rate increased with the increasing exhaust gastemperature. The maximum exergy destruction rates were foundto be 3.5 kW and 4.2 kW at 479 �C for optimized and non-opti-mized heat exchangers, respectively. Similar trend of exergydestruction was also reported by other researchers [48–50]. Highertemperature difference increases the heat transfer rate which in-creases the irreversibility in the system [24,25,46]. Therefore, dueto higher temperature differences between the exhaust, ambientand working fluid the exergy destruction increased with the ex-haust temperature associated with higher power. It is also foundfrom the figure that the exergy destruction was reduced by onaverage 14% for the optimized heat exchanger than the non-opti-mized heat exchanger. At a particular exhaust temperature (i.e.,at a particular power) keeping the inlet temperatures of hot andcold fluids constant, optimized heat exchangers produced lowertemperature differences between the fluids at both end of the heatexchangers due to better heat transfer. These lower temperaturedifferences reduced the irreversibility in the system resulted inlower exergy destruction for the case of optimized heat exchanger.

6.5. Effect of different working pressure

Beside the geometrical aspects of the heat exchangers, the effectof different working pressure of the shell side fluid on the effective-ness of the heat exchangers was also investigated for all geometri-cal aspects of the heat exchanger discussed previously. It wasfound that in all cases the effectiveness was slightly lower at high-er pressure. At higher pressure the shell side fluid molecules weremore compact and molecular diffusion was lower for the samemass flow rate and cross sectional area. This caused a lower heattransfer coefficient which resulted into slightly lower effectivenessat higher working pressure. Thus it can be established that the

Fig. 12. Exergy destruction of non-optimized and optimized heat exchanger.

150 S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151

effect of working pressure on heat exchanger effectiveness for dif-ferent geometrical parameters of the heat exchanger is negligible.Another parameter, the pressure drop across the heat exchangershould also be considered. The maximum pressure drop was foundto be 250 Pa which is negligible as concluded by other researchersas well [19,51–53]. Due to this pressure loss, the engine lostapproximately 0.5% power which was found experimentally at dif-ferent loads. This loss of power will only affect the full load regionof the engine although most of the engine does not operate at fullload [36].

Fig. 14. Bsfc improvement at various loads of the engine.

6.6. Additional power generation and brake specific fuel consumption

6.6.1. At full loadIt was found from the simulation results that both vapor gener-

ator and super heater could have similar designs and therefore, thesame design was used for both heat exchangers. Extra power thatcould be recovered from the exhaust of the diesel engine with theproposed shell and tube heat exchangers is presented in Fig. 13.The additional power generation by extracting heat from the ex-haust gas was calculated at different working pressures. The twoheat exchangers could be used in two different orientations: Paral-lel and Series (Fig. 2). It was found that additional output power in-creased as the working pressure increased for both parallel andseries arrangement of the heat exchangers. The working pressurewas increased keeping the condensing pressure constant. As a re-sult, the enthalpy drop across the turbine increased with increas-ing working pressure resulted in increased power generation.Interestingly, it was found that 10% higher power output wasachieved by parallel arrangement than series arrangement. Thisis due to effective use of available heat from the exhaust gas byparallel arrangement. From the energy balance, it was found thatin series arrangement excess heat was available in the super heaterwhich could increase the mass flow to produce more power. How-ever, for the increased mass flow rate there was not enough heatavailable in the vapor generator to heat the fluid to saturation tem-perature. On the other hand, for parallel arrangement the excessheat was sent to vapor generator to increase the mass flow. As a re-sult the super-heated steam production increased which led to anincrease in power production. The maximum power output of2.9 kW was achieved at 15 bar working pressure. It was found fromthis research that 15 bar working pressure is the maximum possi-ble working pressure attainable for RC based WHR for this 30 kWdiesel generator. This working pressure of 15 bar was limited bythe available exhaust gas temperature for this particular engine.Higher exhaust gas temperatures for higher capacity engines couldresult in higher working pressure which would then produce moreadditional power.

Fig. 13. Additional power output variation with working pressure.

At 15 bar, the proposed shell and tube heat exchanger recov-ered 11% additional power from the exhaust of the diesel engineusing water as the working fluid considering 80% isentropic effi-ciency of the turbine [24–26]. Due to this additional power gener-ation the bsfc of the engine improved 12%. Hountalas andMavropoulos [54] reported 9% improvement in bsfc by computersimulation at 40 bar working pressure. Srinivasan et al. [55] wasable to achieve 7% improvement in bsfc by using ORC system fora dual fuel engine. The higher bsfc improvement of 12% achievedin this research was attributed to the optimization of the heat ex-changer design and working pressure of the fluid.

6.6.2. At partial loadThe heat exchangers designs and working pressure of the fluid

were optimized at full load of the engine which is described inthe previous sections. However, diesel-generator set could run atpart loads as well. Therefore, performance of the WHR systemwas also evaluated at part loads of the engine. These results arepresented in Fig. 14. It was found that at 40% load the WHR systemwas able to develop 3.4% additional power which led to 3.3%improvement in bsfc. The exhaust temperatures at loads lowerthan 40% were not sufficient enough to recover heat to produceadditional power using water as the working fluid. It is also evidentfrom the figure that higher power can be recovered at higher loadfraction of the engine.

7. Conclusions

The exhaust of a diesel engine contains about 40% of the inputenergy and usually this energy is wasted to the environment. Thiswaste energy can be recovered by using WHR system to produceadditional power and thereby improve bsfc and reduce exhaustemissions per kW of power produced. In this research, experimentwas conducted to estimate the available energy in the exhaust gasand performance of two heat exchangers purchased in the market-place were investigated. Then the heat exchangers were optimizedby computer simulation and optimized heat exchangers were usedto estimate additional power generation. The following outcomeswere established from this research:

� The effectiveness of the non-optimized purchased heatexchangers was found to be 0.44 which is much lower than awell-designed heat exchanger. However, after optimization forthis particular application the effectiveness of the heat exchang-ers was improved to 0.76.� Two heat exchangers were used to generate super-heated

steam to expand in the turbine. These two heat exchangerscan have two arrangements: series and parallel. The waste heat

S.N. Hossain, S. Bari / Energy Conversion and Management 75 (2013) 141–151 151

recovery was more effective for parallel arrangement. Parallelarrangement was able to produce 10% more power than the ser-ies arrangement.� The effect of working pressure on additional power generation

was also investigated. Maximum recovered additional powerwas found to be 2.9 kW at 15 bar working pressure at the ratedload of the engine. Thus, an additional 11% power was achievedwith the proposed shell and tube heat exchanger using water asthe working fluid and this indicates 12% improvement in bsfc.This improvement is over 3 unit higher than that found by otherresearchers. This is due to the optimization of design of the heatexchanger and pressure of the working fluid.� More additional power was achieved at higher power regions of

the engine due to higher exhaust temperatures. At 40% part loadthe WHR system managed to produce additional 3.4% powerwhich resulted in 3.3% improvement of bsfc which is lower thanthat achieved at rated due to lower exhaust temperature.

Acknowledgement

The authors gratefully acknowledge the financial and other sup-port received for this project from the Leartek Pty Ltd.

References

[1] Kanoglu M, KazIm IsIk S, Abusoglu A. Performance characteristics of a dieselengine power plant. Energy Convers Manage 2005;46:1692–702.

[2] Hatazawa M, Sugita H, Ogawa T, Seo Y. Performance of a thermoacoustic soundwave generator driven with waste heat of automobile gasoline engine. TransJpn Soc Mech Engineers Part B 2004;70:292–9.

[3] Johnson V. Heat-generated cooling opportunities in vehicles. SAE TechnicalPapers; 2002.

[4] Heywood JB. Automotive engines and fuels: a review of future options. ProgEnergy Combust Sci 1981;7:155–84.

[5] Hountalas DT, Zannis T, Mavropoulos G. Potential benefits in heavy duty dieselengine performance and emissions from the use of variable compression ratio;2006.

[6] Hountalas D, Katsanos C, Kouremenos D, Rogdakis E. Study of available exhaustgas heat recovery technologies for HD diesel engine applications. Int J AlternPropul 2007;1:228–49.

[7] Tennant D, Walsham B, S.o.A. Engineers. The turbocompound diesel engine.Society of Automotive Engineers, 400 Commonwealth Dr, Warrendale, PA,15096, USA; 1989.

[8] Algrain M, Hopmann U. Diesel engine waste heat recovery utilizing electricturbocompound technology; 2003.

[9] Hopmann U, Algrain MC. Diesel engine electric turbo compound technology.Society of Automotive Engineers; 2003.

[10] Hopmann U. Diesel engine waste heat recovery utilizing electricturbocompound technology. In: DEER conference, San Diego, CA, USA; 2004.

[11] Bass JC, Elsner N, Leavitt F. Performance of the 1 kW thermoelectric generatorfor diesel engines. IOP (Institute Of Physics) Publishing Ltd.; 1995. pp. 295–6.

[12] Matsubara K. The performance of a segmented thermoelectric convertor usingYb-based filled skutterudites and Bi2Te3-based materials. Cambridge Univ.Press; 2001.

[13] Hiereth H, Prenninger P, Drexl K. Charging the internal combustionengine. Springer; 2007.

[14] Willard WP. Engineering fundamentals of the internal combustionengine. New Jersey: Editorial Prentice Hall; 2004.

[15] Canakci M. Combustion characteristics of a turbocharged DI compressionignition engine fueled with petroleum diesel fuels and biodiesel. BioresourTechnol 2007;98:1167–75.

[16] Usta N. An experimental study on performance and exhaust emissions of adiesel engine fuelled with tobacco seed oil methyl ester. Energy ConversManage 2005;46:2373–86.

[17] Parlak A, Yasar H, Eldogan O. The effect of thermal barrier coating on a turbo-charged diesel engine performance and exergy potential of the exhaust gas.Energy Convers Manage 2005;46:489–99.

[18] Hountalas DT, Mavropoulos GC, Katsanos C, Knecht W. Improvement ofbottoming cycle efficiency and heat rejection for HD truck applications byutilization of EGR and CAC heat. Energy Convers Manage 2012;53:19–32.

[19] Katsanos CO, Hountalas DT, Pariotis EG. Thermodynamic analysis of a Rankinecycle applied on a diesel truck engine using steam and organic medium.Energy Convers Manage 2012;60:68–76.

[20] Dolz V, Novella R, García A, Sánchez J. HD Diesel engine equipped with abottoming Rankine cycle as a waste heat recovery system. Part 1: study andanalysis of the waste heat energy. Appl Therm Eng 2012;36:269–78.

[21] D.T. Hountalas, C. Katsanos, V. Lamaris. Recovering energy from the dieselengine exhaust using mechanical and electrical turbocompounding. SAETechnical Paper; 2007 [01-1563].

[22] Weerasinghe WMSR, Stobart RK, Hounsham SM. Thermal efficiencyimprovement in high output diesel engines a comparison of a Rankine cyclewith turbo-compounding. Appl Therm Eng 2010;30:2253–6.

[23] Vaja I, Gambarotta A. Internal combustion engine (ICE) bottoming withorganic Rankine cycles (ORCs). Energy 2010;35:1084–93.

[24] Cengel YA, Turner RH, Cimbala JM. Fundamentals of thermal–fluid sciences.3rd ed. Singapore: McGrawHill; 2008.

[25] Moran MJ, Shapiro HN. Fundamentals of engineering thermodynamics. 4thed. USA: John Wiley & Sons; 2000.

[26] Boretti AA. Transient operation of internal combustion engines with Rankinewaste heat recovery systems. Appl Therm Eng 2012;48:18–23.

[27] Holman JP. Experimental methods for engineers. 8th ed. New Work,USA: McGraw-Hill Series in Mechanical Engineering; 2011.

[28] Hosoz M, Direk M. Performance evaluation of an integrated automotive airconditioning and heat pump system. Energy Convers Manage2006;47:545–59.

[29] Huzayyin AS, Bawady AH, Rady MA, Dawood A. Experimental evaluation ofdiesel engine performance and emission using blends of jojoba oil and dieselfuel. Energy Convers Manage 2004;45:2093–112.

[30] Muralidharan K, Vasudevan D, Sheeba KN. Performance, emission andcombustion characteristics of biodiesel fuelled variable compression ratioengine. Energy 2011;36:5385–93.

[31] Ansys CFX-Solver Theory Guide. Release 13.0 ed. Ansys Inc.; 2011.[32] Ramadhas AS, Muraleedharan C, Jayaraj S. Performance and emission

evaluation of a diesel engine fueled with methyl esters of rubber seed oil.Renew Energy 2005;30:1789–800.

[33] Roy J, Misra A. Parametric optimization and performance analysis of aregenerative organic Rankine cycle using R-123 for waste heat recovery.Energy 2012.

[34] Baehr HD. Thermodynamik. Germany: Springer; 2005.[35] Ahern JE. Exergy method of energy systems analysis. New York, United

States: John Wiley and Sons; 1980.[36] Lapuerta M, Armas O, Rodríguez-Fernández J. Effect of biodiesel fuels on diesel

engine emissions. Prog Energy Combust Sci 2008;34:198–223.[37] Bari S. Investigation into the deteriorated performance of diesel engine after

prolonged use of vegetable oil. In: ASME internal combustion engine division2004 fall technical conference. ASME, Long Beach, California, USA; 2004. p. 1–9.

[38] Teng H, Regner G, Cowland C. Waste heat recovery of heavy-duty dieselengines by organic Rankine cycle Part I: hybrid energy system of diesel andrankine engines. SAE Technical Paper; 2007 [01-0537].

[39] Hung TC, Shai TY, Wang SK. A review of organic Rankine cycles (ORCs) for therecovery of low-grade waste heat. Energy 1997;22:661–7.

[40] Larjola J. Electricity from industrial waste heat using high-speed organicRankine cycle (ORC). Int J Prod Econ 1995;41:227–35.

[41] Leising C, Purohit G, DeGrey S, Finegold J. Waste heat recovery in truckengines. Society of Automotive Engineers, 400 Commonwealth Dr,Warrendale, PA, 15096, USA; 1978.

[42] Alireza B. Simple method for estimation of effectiveness in one tube pass andone shell pass counter-flow heat exchangers. Appl Energy 2011;88:4191–6.

[43] Guo J, Xu M, Cheng L. The application of field synergy number in shell-and-tube heat exchanger optimization design. Appl Energy 2009;86:2079–87.

[44] Bell KJ. Heat exchangers: thermal–hydraulic fundamentals and design.Delaware Method Shell Side Des 1981:581–618.

[45] Lei YG, He YL, Chu P, Li R. Design and optimization of heat exchangers withhelical baffles. Chem Eng Sci 2008;63:4386–95.

[46] Kreith F, Manglik RM, Bohn MS. Principles of heat transfer. New York,USA: Thomson Engineering; 2010.

[47] Holman J. Heat transfer. 8th ed. London, England: McGraw-Hill Book Company(UK) Ltd.; 1997.

[48] Qureshi BA, Zubair SM. Second-law-based performance evaluation of coolingtowers and evaporative heat exchangers. Int J Therm Sci 2007;46:188–98.

[49] Wang EH, Zhang HG, Fan BY, Ouyang MG, Zhao Y, Mu QH. Study of workingfluid selection of organic Rankine cycle (ORC) for engine waste heat recovery.Energy 2011;36:3406–18.

[50] Wei D, Lu X, Lu Z, Gu J. Performance analysis and optimization of organicRankine cycle (ORC) for waste heat recovery. Energy Convers Manage2007;48:1113–9.

[51] Hossain S, Bari S. Effect of different working fluids on shell and tube heatexchanger to recover heat from exhaust of an automotive diesel engine. Worldrenewable energy congress 2011 Linköping, Sweden; 2011. p. 764–71.

[52] Rubaiyat SNH, Bari S. Waste heat recovery using shell and tube heat exchangerfrom the exhaust of an automotive engine. 13th Asian congress of fluidmechanics, Dhaka, Bangladesh; 2010. p. 864–7.

[53] Hossain SN, Bari S. Additional power generation from the exhaust gas of dieselengine by bottoming Rankine cycle. SAE 2013 world congress & exhibition.SAE, Detroit, Michigan, United States; 2013.

[54] Hountalas D, Mavropoulos G. Potential for improving HD diesel truck enginefuel consumption using exhaust heat recovery techniques.

[55] Srinivasan KK, Mago PJ, Krishnan SR. Analysis of exhaust waste heat recoveryfrom a dual fuel low temperature combustion engine using an organic Rankinecycle. Energy 2010;35:2387–99.