Analysis of an Increase in the Efficiency of a Spark IgnitionEngine Through the Application of an AutomotiveThermoelectric Generator

JERZY MERKISZ,1 PAWEL FUC,1 PIOTR LIJEWSKI,1

ANDRZEJ ZIOLKOWSKI,1,2 MARTA GALANT,1

and MACIEJ SIEDLECKI1

1.—Faculty of Machines and Transport, Institute of Combustion Engines and Transport,Poznan University of Technology, Piotrowo st. 3, 60-965 Poznan, Poland. 2.—e-mail:andrzej.j.ziolkowski@put.poznan.pl

We have analyzed the increase of the overall efficiency of a spark ignitionengine through energy recovery following the application of an automotivethermoelectric generator (ATEG) of our own design. The design of the generatorwas developed following emission investigations during vehicle driving undercity traffic conditions. The measurement points were defined by actual opera-tion conditions (engine speed and load), subsequently reproduced on an enginedynamometer. Both the vehicle used in the on-road tests and the enginedynamometer were fit with the same, downsized spark ignition engine (withhigh effective power-to-displacement ratio). The thermodynamic parameters ofthe exhaust gases (temperature and exhaust gas mass flow) were measured onthe engine testbed, along with the fuel consumption and electric current gen-erated by the thermoelectric modules. On this basis, the power of the ATEG andits impact on overall engine efficiency were determined.

Key words: Waste heat recovery, engine efficiency, ATEG, RDE

INTRODUCTION

According to the sustainable development policyimplemented within the framework of the Eur-ope 2020 strategy, it is necessary to reduce con-sumption of conventional energy sources (oil, gas,and coal) to build a competitive, low-carbon econ-omy.1,2 This policy is implemented through a num-ber of legal norms that imply limiting emissions ofgreenhouse gases (GHGs), especially CO2. In 2007,the European Commission (EC) proposed a 30%reduction of greenhouse-gas emissions in developedcountries by the year 2020, and suggested that theEuropean Union itself should make a firm indepen-dent commitment to achieve at least a 20% reduc-tion in greenhouse-gas emissions.2 For theautomotive industry, road emissions standards

were introduced for either CO2 emission or fuelconsumption (Table I).

According to Aren et al.,1 CO2 emissions from theroad transport sector constitute 16.4% of globalemissions. Powering vehicles is currently responsi-ble for 38% of annual global demand for oil.1

To meet such CO2 emissions limits, road-vehiclemanufacturers will have to implement new, inno-vative solutions. Currently, the basic directions ofdevelopment of spark ignition (SI) engines aredownsizing, use of gasoline direct injection systems,as well as variable valve timing and valve lift. Avariable geometric compression ratio during engineoperation is also used. There are two types ofdownsizing: static and dynamic.3 Static downsizingtypically refers to reduction of the geometric dimen-sions of the engine—often by reducing the numberof cylinders or displacement volume. Another pos-sibility is to deactivate cylinders during the com-bustion engine work cycle—known as dynamicdownsizing.

(Received September 9, 2015; accepted April 12, 2016;published online May 19, 2016)

Journal of ELECTRONIC MATERIALS, Vol. 45, No. 8, 2016

DOI: 10.1007/s11664-016-4543-0� 2016 The Author(s). This article is published with open access at Springerlink.com

4028

The aforementioned methods aim to increase theoverall efficiency of the engine, thereby reducing itsfuel consumption while maintaining the generatedpower output.4 Most of the chemical energy con-tained in the fuel is drained through the coolingsystem, as well as in the form of heat and pressurein exhaust gases. There is great potential for heatenergy recovery from exhaust gases, as its share inthe energy balance for SI engines can be as high as45%. For CI engines, it is assumed that the share ofthermal energy loss in the exhaust gases is amaximum of 35%. For this reason, systems thatwork in the Clausius–Rankine cycle as well asthermoelectric generators for automotive applica-tions (ATEGs) have been developed andimplemented.5–10

CHOOSING A COMBUSTION ENGINE FORRESEARCH

For analysis of increased spark ignition (SI)engine efficiency through heat recovery fromexhaust gases using an ATEG, Volkswagen’s mod-ern TSI 1.2-dm3 engine was selected (Fig. 1). Thisengine is a consequence of the strategy adopted bythe company, which relies on implementation ofengine downsizing. One of the first solutions of thistype was the TSI 1.4-dm3 engine. The selectedcombustion engine was primarily developed to meetCO2 emissions requirements. According to the man-ufacturer,* the Audi A3 vehicle, equipped with twoversions of the model TFSI 1.2-l engine with powerof 63 kW and 77 kW, reaches CO2 emissions of118 g/km and 127 g/km, respectively, in the NEDCtest. These values are significantly lower than theestablished limit, which in 2015 equals 130 g/km.

The engine block is made of aluminum withinnovative cast-iron cylinder liners. It uses twovalves per cylinder. The inlet valve jacks have aspecial shape, which is responsible for producing aswirl type of turbulence that facilitates formation ofa homogeneous air–fuel mixture. The engine crank-shaft is made of steel with reduced diameter of themain journal and crank bearings with light con-necting rods, which is meant to reduce frictionlosses. The engine has a gasoline direct injection

Table I. CO2 emission and fuel consumption norms around the world1

Countryor region Target year Standard type Unadjusted fleet target Test cycle Penalties

EU 2015 CO2 130 gCO2/km NEDC (WLTPfrom 2017)

Economic fines

2021 90 gCO2/km WLTPJapan 2015 Fuel economy 16.8 km/dm3 JC08 Economic fines

2020 20.3 km/dm3 Public proclamationCanada 2016 CO2 + other GHGs 217 gCO2/mi US combined Economic fines

2025 (proposed) N/A Sales restrictionIndia From 2016 CO2 (130 g/km) NEDC for

low-powered vehicleNot available

2021 113 g/kmChina 2015 Fuel consumption 6.9 dm3/100 km NEDC Economic fines

2020 (proposed) 5 dm3/100 km Public proclamationUSA 2016 Fuel economy/CO2

+ other GHGs36.2 mpg or 225 g/CO2 km US combined Economic fines

2025 56.2 mpg or 143 gCO2/km Sales restrictionSouth Korea 2015 Fuel economy/CO2

+ other GHGs17 km/dm3 or 140 gCO2/km US combined Public proclamation

Mexico 2016 Fuel economy/CO2

+ other GHGs39.3 mpg or 140 gCO2/km US combined Economic fines

Fig. 1. 1.2-dm3 TSI combustion engine chosen for this research.

*Information from the vehicle producer.

Analysis of an Increase in the Efficiency of a Spark Ignition Engine Through the Application ofan Automotive Thermoelectric Generator

4029

system with six-hole injectors that generate injec-tion pressure of 45 bar to 125 bar. To achieve highvolumetric power, the engine is equipped with aturbocharger with an exhaust wastegate valve.

To determine the actual driving profile as afunction V = f(t), tests under actual operating con-ditions were performed using advanced researchtools, including a portable emissions measurementsystem (PEMS). The tests were performed on a testroute with total length of 12.6 km. When selectingthe test route, the authors aimed to reproducetypical urban (involving frequent vehicle stops) andsuburban conditions. The selected test route met thesaid requirements. The ‘‘urban’’ part includedhighly congested streets and involved a variety ofintersections. The ‘‘suburban’’ part was a portion ofthe road number 92—one of the main easternentrance roads to the Poznan agglomeration. Suchvaried road conditions allow analyses to be con-ducted under a wide range of speeds and variedaccelerations. This influences the effective engineparameters, which translate directly into theamount of thermal energy in the exhaust gases.

DETERMINING THE POWER BAND OF THEENGINE BASED ON MEASUREMENTS MADE

WITH THE RDE METHODOLOGY

To determine the engine power band for testbench investigation of the heat recovery from theexhaust gases, tests were carried out under realoperating conditions in accordance with the realdriving emissions (RDE) methodology. This studyselected a vehicle equipped with a 1.2-dm3 TSIengine, which was also located on an engine dynostation. Measurements were made using the latestPEMS instrument—the SEMTECH ECOSTAR.This is another device from the SEMTECH�

series—a successor to the SEMTECH DS analyzer.The main difference between these devices is thatthe SEMTECH ECOSTAR is made up of separatemodules (Fig. 2):11

1. Fuel economy meter (FEM) module for measur-ing the concentration of CO2, CO, and totalhydrocarbons (THC) using a nondispersiveinfrared (NDIR) analyzer,

2. NOx module for measurement of NOx (NO andNO2) concentrations using a nondispersiveultraviolet (NDUV) analyzer,

3. Flame ionization detector (FID) module tomeasure THC,

4. Power distribution module (PDM) to supplypower to each component.

When all the modules (of the gas measurementsection) are used, they must be connected togetherby a controller area network (CAN; green lines inFig. 3). The FEM module is connected via a specialcable to a flow meter for measuring thermodynamicparameters of exhaust gases (mass flow, temperature,

and pressure). This line is supplied a sample ofexhaust gases that passed through the NDIR ana-lyzer (measuring CO2, CO, and THC), directed tothe NOx module with three connectors, whereinmeasurements of NO and NO2 take place. Apartfrom the exhaust gas sample, a special conduittransmits the signal from the vehicle global posi-tioning system (GPS) and a weather sensor used tomeasure atmospheric conditions (temperature,pressure, and humidity), which are connected tothe flow meter. A computer for system control and asignal converter from the vehicle’s diagnostic sys-tem are connected using an AUX2 connector to thePDM, FEM, and NOx modules. Each module isequipped with one AUX1 and one AUX2 connectionpoint. The signal from the power supply module forthe heated wire is sent to the FID module. TheSEMTECH ECOSTAR device can be powered by aPDM module, to which voltage is supplied from thegrid or the car battery. It is also possible to directly

Fig. 2. SEMTECH ECOSTAR instrument, used to measure emis-sions in accordance with the RDE methodology.

Fig. 3. Diagram of SEMTECH ECOSTAR system used for test-ing—section responsible for measurement of gaseous compounds.

Merkisz, Fuc, Lijewski, Ziolkowski, Galant, and Siedlecki4030

connect the modules to the mains. Due to themodular design of the instrument, it can performmeasurements of all gaseous compounds simulta-neously, or separately using single modules.

For the purposes of this study, only measure-ments of CO2 and CO (FEM module), thermody-namic parameters of exhaust gases (temperature,pressure, and mass flow rate), as well as weatherconditions and vehicle speed and position weretaken. These measurements were performed underreal operating conditions on a designated route withlength of 12 km. A detailed description of the testroute is given in Refs. 8 and 12. Measurements werecarried out on a weekday afternoon.

On the basis of the measurements taken inaccordance with the RDE methodology, the charac-teristics of the share of working time intervals inrelation to crankshaft speed and engine load for thetested vehicle were determined, as presented in athree-dimensional (3D) graph in Fig. 4. The powerband of engine operation was defined as the averagetime in the range of 400 rpm and 20 Nm. For morethan half of the total test time, the engine was in therange of 1400 rpm to 3000 rpm with load notexceeding 100 Nm. Single points of work, whoseshare accounted for only 3%, were recorded abovethis range. The share of idling time amounted to31%.

The registered operating points of the combustionengine are presented with the full power character-istic as a two-dimensional (2D) graph in Fig. 5, fromwhich it can be determined that the engine wasmainly operating in the crankshaft rotational speedrange of 1500 rpm to 3500 rpm, where the torque ismaximum. The combustion engine did not reach thevalue of 177 Nm. The maximum torque obtained inthe road test was 144 Nm. No operating points wereregistered in the range of maximum engine crank-shaft speed. This distribution of engine operatingpoints stems directly from the route used in the test.The route has an urban section, where the vehicle

speed limit is 50 km/h, and a suburban section,where the limit is 70 km/h. There are also parts ofthe route which have a 30 km/h or 40 km/h speedlimit. As a result of this, the vehicle cannot obtain ahigh average speed during the test cycle, whichresults in the engine operating at low and mediumloads. During the research, the vehicle reached amaximum speed of 77 km/h and its average speedwas 33.2 km/h.

To best mimic the registered working conditions ofthe test vehicle’s drive unit when testing on theengine dyno bench, it was decided that the researchwould be carried out at three different load charac-teristics with a constant speed of the engine crank-shaft with load changed in 20-Nm intervals (Fig. 5,marked with pink dots): 1800 rpm (six operatingpoints), 2200 rpm (five operating points), and2800 rpm (four operating points). The study alsoincluded the engine idling state (800 rpm). This givesa total of 16 operating points for the engine, at whichtests were carried out under laboratory conditions.

RESEARCH METHODOLOGY USEDFOR ENGINE DYNO TEST BENCH

Measurements of heat recovery from the exhaustgases were carried out on a static engine dyno testbench from Automex company, on which the 1.2-dm3 TSI engine was mounted, generating a maxi-mum power output of 77 kW at 5000 rpm (Table II).The volumetric power indicator of this drive unit is64.2 kW/dm3. The engine meets the Euro 5 emis-sion standards and is equipped with a three-waycatalytic (TWC) converter.

The study used the prototype ATEG developed bythe authors as part of the Applied Research Pro-gramme funded by the National Centre forResearch and Development (Fig. 6; Table III). ThisATEG is modular and consists of three mainelements: a heat exchanger, commercial thermo-electric modules, and cooling elements. The main

Fig. 4. Time share characteristics in the range of engine crankshaftspeeds and loads in a test carried out in accordance with RDEmethods.

Fig. 5. Operating points for the engine of the test vehicle achieved ina test carried out in accordance with RDE methods with its full powercharacteristic.

Analysis of an Increase in the Efficiency of a Spark Ignition Engine Through the Application ofan Automotive Thermoelectric Generator

4031

part of the heat exchanger, onto which the thermo-electric modules are arranged, has a rectangularshape.

Its width is almost twice its height. To ensure auniform temperature distribution along the heatexchanger, ribbing with increasing cross-sectionwas implemented inside it. It was assumed thatfour modules were arranged longitudinally on eachside of the heat exchanger. There are two rows ofmodules on the top and bottom sides of the heatexchanger, and one row on each side, for a total of 24modules. There is one dedicated cooling panel foreach row of modules, attached to the heat exchangerby means of a screw connection. Each radiator isscrewed to a heat exchanger using four screwsplaced at its corners. To ensure uniform pressure ofthe modules on the wall of the exchanger, anadditional locking system was also used in themiddle of the radiators. The screws were tightenedto maximum torque of 20 Nm. This allows use ofthermoelectric modules of different thickness. Aspecial silicated thermal grease was appliedbetween the modules and the walls of the heatexchanger to facilitate heat transfer from the walls

of the heat exchanger to the hot side of the moduleand help reduce possible heat leaks arising from thesurface roughness of the heat exchanger and othermaterial imperfections and faults.

The study used 24 commercial Ferrotec SCTBNORD thermoelectric modules codenamed TMG-241-1.4-1.2,13 connected in series. These modulesare Bi2Te3 based, and each can generate 9.4 W ofpower. The maximum temperature of the hot side ofthe module may not exceed 200�C when usingthermally conductive paste. Without conductivepaste, the temperature should not exceed 175�C.The module generates a maximum voltage of 12.5 Vwith no load, and 6.3 V with load. The modules werechosen based on previous research experience of theauthors with heat recovery from exhaust gasesusing ATEG. Reference8 presents the results of aTEG with a different design, using an earlierversion of the TMG-241-1.4-1.2 modules, whosemaximum power was 7 W. These were selected onthe basis of measurements showing that theyachieved satisfactory operating parameters whilemaintaining resistance to the effects of high tem-peratures. The TMG-241-1.4-1.2 modules are char-acterized by high durability due to the nature of thejoining techniques, junction materials, and diffusionbarriers. For this reason, in the prototype ATEGwhich is the subject of this study, the authors used anewer version of the TMG-241-1.4-1.2 modules fromSCTB NORD A Ferrotec Group Company. Duringthe tests of heat recovery from exhaust gasesperformed on the static engine test bench, thefollowing parameters were measured:

(a) Instantaneous fuel consumption G (g/s),using a mass fuel gauge from Automexcompany,

Table II. Parameters of the tested 1.2-dm3 TSIengine

Parameter Value

Type of engine Spark ignitionDisplacement 1.2 dm3

Number of cylinders 4Stroke 75.6 mmBore 71 mmNumber of valves

per cylinder2

Power output 77 kW at 5000 rpmTorque 175 Nm at 1550 rpm

to 4100 rpmEmissions standard Euro 5Exhaust gasposttreatment

TWC with lambda control

Fig. 6. ATEG installed in the engine testbed.

Table III. ATEG characteristics

Parameter Value

Length 671 mmWidth 200 mmHeight 135 mmInlet diameter 60 mmMaterial Aluminum alloy with

thickness of 2 mmNumber of modules 24Module arrangement Top and bottom

walls—two rowswith four modules each;

side walls—one rowwith four modules

Type of modules used TMG-241-1.4-1.2 byFerrotec NORD

Number of cooling panels 6Generated nominal power 225.6 WGenerated nominal

voltage without load300 V

Generated nominalvoltage with load

151.2 V

Merkisz, Fuc, Lijewski, Ziolkowski, Galant, and Siedlecki4032

(b) Crankshaft rotational speed s (rpm) andengine torque T (Nm), using the stationaryengine dyno test bench made by Automexcompany, equipped with an eddy-currentbrake and inductive engine speed sensor,

(c) Exhaust mass flow rate EFM (kg/h), usingthe SEMTECH ECOSTAR,

(d) Exhaust gas temperature at four measure-ment points t1, t2, t3, and t4 (�C), usingthermocouples,

(e) Temperature of the hot side of the thermo-electric modules t11, t12, t21, t22, t31, t32, t41,t42, t51, t52, t61, and t62 (�C) (first and lastmodule in each row), using thermocouples,

(f) Mass flow mcool (dm3/h) and temperature tcool

(�C) of the coolant,(g) Voltage U (V) and current I (A) generated by

the modules, using a custom-made measure-ment system of the authors’ design.

To register the temperatures and pressures, twoIOTECH Personal Daq/3000 signal converters wereused. The measurement module was equipped witha universal serial bus (USB) interface and high-speed analog-to-digital (A/D) converter (1 MHz/16 bit). The transmitter has 16 single-ended-typeanalog inputs (8 differential inputs), 4 analogoutputs, and 24 digital input/output lines, and it ispossible to program the device in seven ranges from±100 mV to ±10 V. Information from the sensor istransmitted to a computer, which records the dataat a given frequency.

The custom measurement system to measure thevoltage and current generated by the modulesconsists of: the main module (which includes, amongothers, a USB-600 measuring card manufactured byNational Instruments having eight analog inputsand four digital inputs), two converters for voltageand current measurements fitted on a printedcircuit board (PCB), a slide resistor for setting theload of the system, as well as control software. Themain module is supplied with a voltage of 220 V.The software allows for monitoring of the generatedvoltage and current in real time (in numerical andgraphical form) as well as data recording at afrequency of 1 Hz. It uses algorithms to calculatethe generated power in watts and the energyexpressed in kWh. A measuring device records thetotal value of voltage and current of all modules, butalso has the ability to measure these values forindividual modules.

ANALYSIS OF RESULTS

The overall efficiency of an internal combustionengine is a measure of its use of the energycontained in the fuel. It provides insight into theconversion efficiency of chemical energy containedin the fuel into mechanical energy output by themotor to the power receiver. This efficiency is also

described as the inverse of the specific fuel con-sumption and can be expressed by the formula

g0 ¼ Pe

G � GCV¼ 1

BSFC � GCV; ð1Þ

where g0 is the overall efficiency (–), Pe is the enginepower output (kW), G is the instantaneous fuelconsumption (g/s), BSFC is the specific fuel con-sumption (g/kWh), and GCV is the gross calorificvalue (J/g).

To determine the total efficiency of an internalcombustion engine using an ATEG, the powergenerated by the generator PeATEG needs to beincluded in Eq. 1, as follows:

gEþATEG ¼ ðPe þ PATEGÞG � GCV

¼Pe þ ðU � IÞATEG

� �

G � GCV: ð2Þ

Such addition of the power generated by the ATEGPeATEG to the power output of the engine Pe isjustified because the generator produces electricalpower from the thermal energy of the exhaust gasesafter fuel combustion, which in an internal combus-tion engine is treated as waste as the heat createdby burning the air–fuel mixture is not used. Thisincreases the overall efficiency of the engine,because a greater amount of energy is obtainedfrom the combustion produced by the same amountof fuel. In this way, the heat loss in the exhaustgases is also limited.

The determination of the overall engine efficiencyon the engine dyno bench used the measuredparameters of crankshaft speed and engine load(used to calculate the engine power output) and thefuel consumption per second Gs. This consumptioncan also be calculated from the emissions per secondof CO2, CO, and THC in accordance with the carbonbalance method as modified by the authors.14 Aspart of the tests carried out, the relationshipbetween the fuel consumption (as measured byAutomex instrumentation) and the consumptioncalculated on the basis of the parameters measuredby the PEMS equipment was determined; morethan 30 trials were performed for each operatingpoint of the engine, which statistically can betreated as a large test sample. Figure 7 shows therelationship between the average values of instan-teneous fuel consumption for the three enginespeeds.

The datasets were linearly interpolated, obtain-ing a coefficient of determination R2 close to 1. Thisreveals very good agreement between the resultsobtained using the two methods. Thus, the carbonbalance method to calculate the fuel consumptionper second is shown to be reliable and can be usedfor results obtained both on the engine dyno benchand under real conditions (RDE methodology).

The specific fuel consumption in the studiesconducted was in the range of 238.5 g/kWh to384.4 g/kWh (Table IV). The lowest value indicatesthe highest overall efficiency of the engine, which

Analysis of an Increase in the Efficiency of a Spark Ignition Engine Through the Application ofan Automotive Thermoelectric Generator

4033

was recorded for S = 2200 rpm and loads of 60 Nmand 80 Nm, with efficiency of 34.5% and 35.9%,respectively. For S = 1800 rpm and T = 100 Nm, theoverall efficiency was 32.6%, whereas forS = 2600 rpm and T = 69 Nm, it amounted to 33.7%.For idling (the lowest possible speed of the crankshaftwithout load) and neutral (increased speed of thecrankshaft without load), the BSFC and efficiencywere not calculated, because the engine does notgenerate any power output for transmission to thereceiver. It only generates power as needed to over-come the resistance of its operation. Due to theexistence of these points in the real road drivingconditions, it was decided to also include these pointsin the analysis. In these cases, only the value of thepower generated by ATEG is presented, withoutdetermining its effect on overall engine efficiency.

The ATEG power was determined based on themeasured voltage and current generated by the

thermoelectric modules at each measurement point.The tests began with the engine stabilized ther-mally and with generator wall temperature notexceeding 50�C. The duration of the measurementat each point was 60 s to 90 s. Before recording theparameters after setting the operating point of theengine, the system was given time until tempera-tures on the walls of the ATEG stabilized. After thisperiod, recording of working parameters was per-formed. This scheme was adopted in each loadcharacteristic. When changing the range of crank-shaft speed, the ATEG conditioning time was sig-nificantly prolonged to lower the temperature of itswalls to fixed values. This measure was used toreduce the impact of the thermal capacity of thealuminum on the heat flux from the exhaust gasesconducted to the thermoelectric modules. Due to thenumber of measurements and the measured param-eters, only exemplary temperature distributions forthe exhaust system and on the walls of the heatexchanger (hot side of the thermoelectric modules)are presented herein. Measurements were carriedout using water from the sanitary system as coolantwith temperature of 10�C to 12�C at volumetric flowrate of �0.045 m3/h.

For neutral gear at S = 1800 rpm, the exhaustgas temperature between the turbocharger and thecatalytic converter averaged 421.5�C throughoutthe duration of the measurement (Fig. 8). TheATEG was placed at a distance of �1.5 m from theend of the catalytic reactor’s confusor. The inlettemperature to the generator averaged 91.4�C. Thetemperature drop in the generator was 30.3�C. Atmaximum load (100 Nm) for the same rotationalspeed of the engine crankshaft, this drop was140.8�C, and the average temperature on the inputwas t4 = 491.5�C.

For the above measuring point, the highesttemperature was recorded on the bottom wall ofthe heat exchanger (a substantial part of theATEG), being equal to t51 = 45.5�C (Fig. 9). Theminimum temperature value recorded occurred atthe point t11. The temperature difference betweenthese two points was 7�C. This temperature differ-ence is not very significant, meaning that thethermoelectric modules generated similar voltageand current values. Considering the temperaturedistribution on the walls of the ATEG along eachrow of modules, it was found that the greatesttemperature drop occurred on the left side wall ofthe generator: Dt3 = t31�t32 = 6.4�C. On other walls,the recorded temperature differences varied withinthe range of 0.3�C to 4�C. In the case of maximumload (with S = 1800 rpm), the highest temperaturealso occurred at the point t51 and was equal to175�C.

To determine the effect of the ATEG on the totalefficiency of the tested engine based on the data atthese measuring points, it was first necessary todetermine the power generated by the thermoelec-tric modules PeATEG using the values of voltage and

Fig. 7. Linear interpolation of average values of instantenous fuelconsumption recorded on an engine test bench and calculated on thebasis of parameters measured by the PEMS.

Table IV. Comparison of specific fuel consumptionand overall efficiency of tested engine

Enginespeed (rpm)

Torque(Nm)

BSFC(g/kWh)

Efficiency(%)

900 0 – –1800 0 – –

20 295.3 28.640 285.6 28.760 266.4 30.780 258.5 31.7

100 250.8 32.62200 0 – –

20 368.4 22.440 275.5 29.860 238.5 34.580 228.3 35.9

2600 0 – –20 384.4 21.340 283.6 20.160 243.8 33.780 237.3 34.5

Merkisz, Fuc, Lijewski, Ziolkowski, Galant, and Siedlecki4034

current obtained using the developed measure-ment system. At all the load characteristics, atrend of increased ATEG power generation withincreasing engine torque (Fig. 10) could beobserved. This is logical because, at higher loads,the engine produces more exhaust gases at highertemperature. The highest values of PeATEG were asfollows:

(a) 90.1 W at 1800 rpm(b) 98.8 W at 2200 rpm(c) 189.3 W at 2600 rpm

For engine idling (S = 900 rpm and T = 0), PeATEG

did not exceed 1 W. The obtained test results did notreach the ATEG nominal power, which is equal to

225.6 W. The system generated at most 84% of itsmaximum nominal power.

The overall engine efficiency, including theenergy generated by the ATEG, was determinedusing Eq. 2. Then, the obtained values were com-pared with the determined total engine efficiency(Table IV). According to this assessment, the great-est efficiency improvement was obtained forS = 2600 rpm (Fig. 11). The energy generated bythe ATEG over the tested range of engine operationdoes not constitute a significant share of the totalenergy. This is mainly a result of the failure toachieve the nominal power of the generator. Theanalysis above does not take into account the powerloss occurring in the ATEG associated with resis-tance to the flow of exhaust gases.

Fig. 8. Temperature distribution in exhaust system of tested engine with ATEG for neutral gear at S = 1800 rpm.

Fig. 9. Temperature distribution on a substantial part of the ATEG heat exchanger–hot side of the thermoelectric modules for neutral gear atS = 1800 rpm.

Analysis of an Increase in the Efficiency of a Spark Ignition Engine Through the Application ofan Automotive Thermoelectric Generator

4035

To determine the efficiency of the ATEG in termsof conversion of exhaust gas thermal energy intoelectrical energy, one needs to determine theamount of thermal energy from the relation

QEXHAUST ¼ m�

EX�cpEX � DTEX; ð3Þ

where QEXHAUST is the exhaust gas energy losses

(enthalpy) (kW), m�

EXis the exhaust gas mass flow (kg/

s), cpEX is the exhaust gas specific heat at constantpressure [kJ/(kg K)], and DTEX is the exhaust gastemperature (K). The ATEG efficiency gTEG wasdetermined from the relation

gTEG ¼ PATEG

QEXHAUST; ð4Þ

where gTEG is the efficiency of the ATEG (%), PTEG isthe ATEG power (W), and QEXHAUST is the exhaustgas energy losses (enthalpy) (W).

In determining the efficiency gATEG, the powerlosses resulting from the exhaust gas flow resistanceinside the heat exchanger were not taken intoaccount. For the load characteristics at S = 1800rpm, the maximum efficiency gATEG did not exceed0.8% (Fig. 12). For the other load characteristics, thisefficiency was between 0.5% and 1.3%.

CONCLUSIONS

We present a study of the impact of a prototypeATEG on the overall efficiency of a modern SI enginedesigned in accordance with downsizing principles.The walls of the ATEG exhibited an almost uniformtemperature distribution, both along and betweenthe walls. Temperature differences amounted to amaximum of 7�C for neutral gear, being larger forhigher loads. The measuring points did not reach themaximum power output of the ATEG. Only forS = 2600 rpm and T = 80 Nm was more than 80%of the nominal power output achieved. Energy lossesarising as a result of the flow resistance of exhaustgases in the heat exchanger of the generator were notaccounted for in this research. The authors plan toconduct further research including:

(a) Performance measurements of heat energyrecovery using the designed prototype ATEGsystem at operating points reflecting ruraland highway driving conditions,

(b) Determination of the energy losses due to theflow resistance of exhaust gases through theheat exchanger,

(c) Adjustment of the design and construction ofthe ATEG to further increase the efficiency ofheat recovery from exhaust gases,

(d) Integration of the ATEG cooling system withthat of the vehicle.

ACKNOWLEDGEMENTS

The research was funded by the National Centre forResearch and Development (Narodowe Centrum Ba-dan i Rozwoju) research project within the AppliedResearch Programme (Contract No. PBS1/A6/7/2012).

OPEN ACCESS

This article is distributed under the terms of theCreative Commons Attribution 4.0 InternationalLicense (http://creativecommons.org/licenses/by/4.0/),which permits unrestricted use, distribution, andreproduction in any medium, provided you give

Fig. 10. Calculated power generated by the thermoelectric genera-tor modules (ATEG) during tests.

Fig. 11. Increase in overall efficiency of test engine through appli-cation of ATEG in exhaust system.

Fig. 12. Power and efficiency of ATEG depending on engine speed.

Merkisz, Fuc, Lijewski, Ziolkowski, Galant, and Siedlecki4036

appropriate credit to the original author(s) and thesource, provide a link to the Creative Commonslicense, and indicate if changes were made.

REFERENCES

1. F. Aren, L. Mezzan, A. Doyon, H. Suzuki, K. Lee, and T.Becker, The Automotive CO2 Emissions Challenge. 2020Regulatory Scenario for Passenger Cars (Arthur D. Little,Rome, 2014). http://www.adlittle.com/downloads/tx_adlreports/ADL_AMG_2014_Automotive_CO2_Emissions_Challenge.pdf.

2. European Commission, Communication from the Commissionto the European Parliament, the Council, the European Eco-nomic and Social Committee and the Committee of the Regions.An Integrated Industrial Policy for the Globalisation Era Put-ting Competitiveness and Sustainability at Centre Stage.(European Commission, Brussels, 2010) http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CCO:2010:0614:FIN:EN:PDF.

3. M. Brze _zanski and K. Sliwinski, Combust. Engines. 119(2004).

4. M. Bajerlein and L. Rymaniak, Appl. Mech. Mater. 518, 96–101 (2014).

5. P. Bombard, C.M. Invernizzi, and C. Pietra, J. Electron.Mater. (2010). doi:10.1016/j.applthermaleng.2009.08.006.

6. J. Dong, Y.J. Wang, R. Zhang, and B. Wang, J. Electron.Mater. (2014). doi:10.1007/978-3-662-45043-7_15.

7. N. Espinosa, M. Lazard, L. Aixala, and H. Scherrer, J.Electron. Mater. (2010). doi:10.1007/s11664-010-1305-2.

8. J. Merkisz, P. Fuc, P. Lijewski, A. Ziolkowski, and K.T.Wojciechowski, J. Electron. Mater. (2015). doi:10.1007/s11664-014-3522-6.

9. K. Shiho, P. Soonseo, K. SunKook, and R. Seok-Ho, J.Electron. Mater. (2011). doi:10.1007/s11664-011-1580-6.

10. D. Tatarinov, M. Koppers, G. Bastian, and D. Schramm, J.Electron. Mater. (2013). doi:10.1007/s11664-013-2642-8.

11. SEMTECH� ECOSTAR (Gaseous). Getting Started.Quick Reference Manual. Document: 9510-159 (SensorsInc., USA, 2014).

12. J. Merkisz, P. Fuc, P. Lijewski, and A. Ziolkowski, Appl.Mech. Mater. 390, 343–349 (2013).

13. Thermoelectric module TMG-241-1.4-1.2 SCTB NORD AFerrotec Group Company.

14. Fuc, J. Merkisz, P. Lijewski, A. Merkisz-Guranowska, andA. Ziolkowski, in Energy Production and Management inthe 21st Century: The Quest for Sustainable Energy (Vol-ume 1), ed. C.A. Brebbia, E.R. Magaril, and M.Y. Khodor-ovsky (WIT Transactions on Ecology and the Environment,New Forest, 2014), pp. 369–378.

Analysis of an Increase in the Efficiency of a Spark Ignition Engine Through the Application ofan Automotive Thermoelectric Generator

4037