TOPICAL COLLECTION: INTERNATIONAL CONFERENCE ON THERMOELECTRICS 2019

The High Potential for Waste Heat Recovery in Hybrid Vehicles:A Comparison Between the Potential in Conventionaland Hybrid Powertrains

MARTIN KOBER 1,2

1.—German Aerospace Center (DLR), Institute of Vehicle Concepts, Pfaffenwaldring 38-40,70569 Stuttgart, Germany. 2.—e-mail: martin.kober@dlr.de

In future mobility, the mix of different drive trains will probably be muchmore diverse than it is today. According to a large number of scenario anal-yses, a predominant number of vehicles will continue to be based on theinternal combustion engine (ICE), while an increasing number of hybridvehicles are expected. To achieve the required reductions in CO2 emissions itis necessary to investigate all potential technologies for efficiency improve-ment. Therefore in this work the potential of waste heat recovery is examinedfor conventional and hybrid vehicles. Due to the fact that in an internalcombustion engine approximately 2/3 of the fuel’s chemical energy dissipatesas waste heat, the potential for the recovery of this energy in all ICE drivenpowertrains is, in principle, high. The results of this work show that in hybridvehicles the highest share of the energy supplied by the fuel is lost in theexhaust gas. In order to further elaborate this result, we conduct an exem-plary examination of two comparable vehicles of the compact class within theworldwide harmonized light duty test cycle. Measurement data from the twovehicles at the roller dynamometer is used. The result shows that the aver-aged exhaust gas heat flow of the conventional vehicle is 5.0 kW. For thehybrid vehicle, driving in the charge sustaining mode, the averaged exhaustgas heat flow results in 8.1 kW. The comparison shows that the temperaturelevel of this exhaust gas is even higher than that of the conventional vehicle.In addition, this work shows that through the higher temperatures, the exergyin the exhaust gas is higher in hybrid vehicles even if the combustion engineworks with a higher efficiency. In the exemplary comparison the averagedexergy of the exhaust gas is 3.2 kW for the conventional and 5.7 kW for thehybrid vehicle. As a result of this work, the high potential for waste heatrecovery in hybrid vehicles could be demonstrated.

Key words: Waste heat recovery, conventional and hybrid powertrains,thermoelectric generator, power generation, exergetic potential

List of Symbolscp Specific heat capacity (J/kg K)_E Exergy flow (W)h Heat of combustion (J/kg)m Mass (kg)_m Mass flow (g/s)p Pressure (Pa)

_Q Heat flow (W)T Temperature (K)U Internal energy (J)V Volume (dm3)gC Carnot efficiency (�)u Crank angle (�)

AbbreviationsCO2 Carbon dioxideDLR Deutsches Zentrum fur Luft- und

Raumfahrt/German Aerospace Center(Received August 21, 2019; accepted January 31, 2020;published online February 13, 2020)

Journal of ELECTRONIC MATERIALS, Vol. 49, No. 5, 2020

https://doi.org/10.1007/s11664-020-07991-5� 2020 The Author(s)

2928

DSG Dual-clutch gearboxICE Internal combustion engineORC Organic Rankine cyclePHEV Plug-in hybridTEG Thermoelectric generatorTEM Thermoelectric moduleTSI Turbocharged stratified injectionWLTC Worldwide harmonized light duty driving

test cycleWLTP Worldwide harmonized light duty vehicles

test procedure

IndexesA AmbientCO CoolantF FuelHG Hot gasI InLeak Leakagemax Maximummi AverageO OutW Wall

INTRODUCTION

One of the major challenges for future automotivedevelopment is to achieve the required reductions ofthe CO2 emissions. Therefore, it is necessary toinvestigate all potential technologies for efficiencyimprovement. In an internal combustion engine(ICE) approximately 2/3 of the fuel’s chemicalenergy dissipates as waste heat. In particular, theexhaust gas offers the highest potential for wasteheat recovery due to its high temperature level. Oneof the promising technologies for automotive wasteheat recovery is the use of thermoelectric generators(TEG). Therefore this technology has been underinvestigation for several years at the DLR—Insti-tute of Vehicle Concepts in Stuttgart.1–5

According to the majority of current scenarioanalyses, it is to be expected that the majority ofnew registrations in the future will continue to bebased on an internal combustion engine. It isassumed that in 2030 this share will account for

more than 80% and in 2040 more than 70% of newregistrations. In addition, an increasing number ofthese vehicles with ICEs will be hybrid vehicles.6,7

According to Ref. 7, which gives a summary of theliterature, global car sales in 2030 are expected to beabout 150% of the sales in 2015. This leads to theprediction that, in absolute terms, more vehicleswith combustion engines will be sold in 2030 thanare sold today.7

The aim of this paper is to consider futureinfluences of development in the vehicle sector ofpotential waste heat recovery. In particular, thedifference between conventional and hybrid vehicleswas investigated from this standpoint.

REFERENCE VEHICLES, DYNAMICDRIVING CONDITIONS

AND MEASUREMENT DATA

Reference Vehicles

In order to compare the available waste heat flowsof the different vehicle concepts, two petrol-poweredreference vehicles of the compact class wereselected. Figure 1 shows the main characteristicsof the chosen reference vehicles.

To represent a common conventional vehicle, theVolkswagen Golf VII with a turbocharged enginewith 1.2 L displacement was selected. As a hybridvehicle, the Opel Ampera (Chevrolet Volt) with1.4 L displacement was chosen. Both vehicles areshown in Fig. 2, while they are being measured onthe roller dynamometer test bench of the DLR—In-stitute of vehicle concepts in Stuttgart.

The Opel Ampera with its power train as a power-split hybrid represents the majority of current plug-in hybrid electrical vehicles (PHEV). Other compa-rable examples of power-split hybrids are MercedesC 350e, BMW 330e, Volkswagen Golf GTE andToyota Prius. The propulsion system of the OpelAmpera is shown in Fig. 3.8

The kinematic architecture of the power-splitPHEV has a planetary gear set and three clutches.The clutches connect and disconnect the powersources in order to realize different operatingmodes. To compare the hybrid vehicle with theconventional vehicle, the hybrid vehicle was oper-ated in the charge sustaining mode. In this mode,

Fig. 1. Main characteristics of the reference vehicles.

The High Potential for Waste Heat Recovery in Hybrid Vehicles: A Comparison Between thePotential in Conventional and Hybrid Powertrains

2929

Fig. 2. Reference vehicles at the roller dynamometer test bench.

Fig. 3. Propulsion system of the Opel Ampera—Kinematic architecture of the power-split PHEV. (Reprinted with permission of Ref. 8).

Fig. 4. Internal combustion engine of the Opel Ampera and its operating characteristics, shown in a torque-speed diagram. (Reprinted withpermission of Ref. 8)

Kober2930

the cycle is started with an empty traction battery,so that the required energy must be supplied by thecombustion engine.

In hybrid vehicles, the combustion engine isoperated at higher load points than in conventionalvehicles. At these higher load points, the combus-tion engine has higher efficiencies. As shown inFig. 4 (orange curve), the combustion engine of theOpel Ampera is operated at various high-loadpoints. Figure 4 also shows that the specific fuelconsumption is lower at the high operating points.

Dynamic Driving Conditions

Dynamic driving conditions were chosen in orderto show a meaningful comparison of the two propul-sion systems. In particular, the worldwide harmo-nized light duty driving test cycle (WLTC) wasselected as part of the related test procedureWLTP.9

As shown in Fig. 5, the used WLTC is 1800 s longand has an average speed of 46.5 km/h and a topspeed of 131 km/h. For the following comparison,the vehicles were measured at 20�C ambient tem-perature. The dynamic velocity profile in Fig. 6shows that a very low load level is required in thefirst third. A high speed is only driven in the last323 s of the cycle, in the so-called extra high part.

The velocity profile and the vehicle mass mainlydetermine the energy required by the propulsionsystem of the vehicle. As shown in Fig. 1, the massof the hybrid vehicle is 400 kg higher.

Measurement Data from the RollerDynamometer

For the following comparison, measurement datafrom the roller dynamometer test bench were used.The measured fuel consumption for both vehicleconcepts is shown in Fig. 1. The value of the hybridvehicle marked with * refers to the charge sustain-ing mode.

The exhaust gas temperatures were measured atthe point after the catalytic converter. The mea-surement position used for the Opel Ampera isshown in Fig. 7.

Figure 8 shows the measured exhaust gas tem-perature and mass flow for the Opel Ampera. It canalso be seen from the exhaust measurement datathat the combustion engine operates in a hybridmode. For example, in the first third of the cycle, theexhaust gas mass flow shows that the combustionengine does not cover the low load points as inconventional drive trains. When the combustionengine is used, it varies between different high loadpoints. In these phases, more electrical energy isconverted from the fuel and the phases with lowload are driven purely electrically. Finally, with thehybrid vehicle in charge sustaining mode, theenergy required to move the vehicle within thespecified driving cycle is converted from fuel. Thedifference is that the combustion engine works onlyat high load points as described above.

POTENTIAL ANALYSIS FOR WASTE HEATRECOVERY IN VEHICLE APPLICATIONS

In this section, the measured data of the exhaustgas temperatures and mass flows will be comparedfor the conventional and hybrid vehicles, with theaim of analyzing the potential for waste heatrecovery, e.g. with the aid of thermoelectricgenerators.

Figure 9 shows a comparison of the exhaust gastemperatures and mass flows within the WLTC. Itcan be seen that the hybrid vehicle has significantlyhigher exhaust gas temperatures. On the otherhand, the maximal reached exhaust gas mass flowsover the cycle are slightly lower.Fig. 5. Dynamic driving conditions of the WLTP driving cycle.

Fig. 6. Velocity profile of the WLTP driving cycle10.

The High Potential for Waste Heat Recovery in Hybrid Vehicles: A Comparison Between thePotential in Conventional and Hybrid Powertrains

2931

Analysis of the Heat Flow

The calculation of the exhaust gas heat flow ( _QHG)includes the exhaust gas mass flow ( _mHG) andtemperature (THG) and is carried out according toEq. 1.

_QHG ¼ cpHG

��THG

TA� _mHG � THG � TAð Þ: ð1Þ

However, in addition to the heat flow, the tem-perature level is important for an efficient use of thewaste heat. Therefore, Fig. 10 shows the exhaustgas heat flow over the exhaust gas temperature.

It can be seen that the hybrid vehicle has higherheat flows in the exhaust gas than the conventionalvehicle. In addition, these higher heat flows are stillat a higher temperature level.

Fig. 7. Measurement position of the exhaust gas temperature after the catalytic converter of the Opel Ampera.

Fig. 8. Measurement data of the exhaust gas mass flow and temperature of the Opel Ampera.

Fig. 9. Comparison of the exhaust gas mass flow and temperaturebetween the two reference vehicles within WLTC. Fig. 10. Heat flow over the exhaust gas temperature for both

reference vehicles within the WLTC.

Kober2932

Analysis of the Exergetic Potential

In order to include the higher temperature level ofthe exhaust heat flows in the evaluation, the exergy

flows ( _EHG) are calculated according to Eq. 2 below.

Thereby the exhaust heat ( _QHG) flow is multipliedwith the Carnot efficiency (gC).

_EHG ¼ _QHG � gC: ð2Þ

The Carnot efficiency is calculated according toEq. 3 from the temperature level of the exhaust gas(THG) to ambient (TA).

gC ¼ 1 � TA

THG: ð3Þ

Thus the exergy flow represents the entire poten-tial for a waste heat recovery system. For example,the conversion efficiency of a thermoelectric gener-ator (TEG) or an organic Rancine cycle (ORC) alsodepends on the Carnot efficiency.

Figure 11 shows the exergy flows of the twovehicle concepts over time in the WLTC. Thisclearly reflects the fact that the combustion enginein the conventional vehicle operates at many partialload points, resulting in low exergy flows in the firstpart of the cycle. In contrast, the hybrid vehicleshows significantly higher exergy flows, especiallyin the first part of the cycle. These higher exergyflows occur in the hybrid vehicle over a shorterperiod of time.

An analysis of the absolute frequency of thedifferent exergy flows is shown in Fig. 12. Theabsolute frequency was calculated over the time ofthe WLTC by a summation of the amount for eachexergy flow level. Figure 12 shows that in conven-tional vehicles the low exergy flows occur with avery high absolute frequency. The higher exergyflows occur only rarely. On a closer look at Fig. 11, itcan also be seen that in the conventional vehicle thehigh exergy flows occur only in the last part of thecycle (extra high part).

With the hybrid vehicle, on the other hand,medium to large exergy flows occur with a highabsolute frequency. This can be explained by thefact that the combustion engine is operated athigher load points.

The maximum and average values within theWLTC are shown as summary of the analysis inFig. 13. The average heat flow of the hybrid vehi-cle’s exhaust gas within the WLTC is about 62%higher. Figure 1 shows that the measured fuelconsumption of the hybrid vehicle is about 20%higher. This value must be subtracted when com-paring the potential. Nevertheless, the average heatflow in the exhaust gas of the hybrid vehicle is stillsignificantly higher than in the conventional vehi-cle. An explanation for this fact will be given in thefollowing section. In addition, the analysis hasshown that the exhaust gas temperatures are

Fig. 11. Exergy flow in the exhaust gas of both reference vehicles within the WLTC.

Fig. 12. Distribution of the absolute frequency of exergy flows—bothreference vehicles within the WLTC.

Fig. 13. Results of the exergetic potential analysis within the WLTC.

The High Potential for Waste Heat Recovery in Hybrid Vehicles: A Comparison Between thePotential in Conventional and Hybrid Powertrains

2933

higher in the hybrid vehicle. As a result, themaximum Carnot efficiency is also higher than inthe conventional vehicle. However, the averageCarnot efficiency in the WLTC is not significantlyhigher, since the longer standstill phases of thecombustion engine reduce it.

Nevertheless, the average exergy flow calculatedfrom the exhaust gas heat flow and the Carnotefficiency is a remarkable 78% higher in the hybridvehicle. This is due to the fact that in the transientcourse of the cycle, the high Carnot efficienciesoccur together with high exhaust gas heat flows.

ENERGY FLOWS IN CONVENTIONALAND HYBRID VEHICLES

Here, an explanation will be given for the factthat, in the comparison of the measurement datafrom the roller dynamometer, the exhaust heatflows are higher in the hybrid vehicle. Even thoughin this vehicle concept the combustion engine isoperated with a higher efficiency, more heat is lostin the exhaust gas. The explanation shall be givengenerally concerning the thermodynamic relationswithin an internal combustion engine. Thereforethe first law of thermodynamics according to Ref. 11is mentioned in Eq. 4. In this context, the mechan-ical work, the wall heat losses and the exhaust gasenthalpy will be discussed.

�pdV

du|fflfflffl{zfflfflffl}

Mechanical work

þdQF

du�dQW

du|fflfflffl{zfflfflffl}

Wall heat losses

þhIdmI

du

þ hOdmO

du|fflfflfflffl{zfflfflfflffl}

Exhaust gas enthalpy

�hOdmLeak

du

¼ dU

duð4Þ

Focusing on the mechanical work, we see that thisis higher in the hybrid vehicle because the internalcombustion engine is operated at a higher loadpoint. Figure 4 shows that at higher load points theICE has a lower specific fuel consumption and thusa higher efficiency. On the other hand, the

measured values in the exhaust tract show, asalready described, higher exhaust gas enthalpyflows.

The first law of thermodynamics is balanced bythe fact that the wall heat losses are significantlylower.

Figure 14 shows the wall heat losses exemplaryfor an internal combustion engine as a function ofthe engine speed and load according to Ref. 11 It canbe seen that in idling and at a medium speed around26% of the energy supplied by the fuel (QF) is lost inthe form of wall heat (QW). In contrast to this, at fullload only 12% of the energy supplied by the fuel arelost as wall heat. Thus in this example the differ-ence in wall heat losses is 14%.

Beyond this general example, the measurementdata of the two vehicles in the WLTC will be usedfor the next step. The measured exhaust gastemperatures, mass flows, speeds and torques ofthe combustion engine are used. With the help ofthis data, the energy flows in the conventional andhybrid vehicles were analysed. The mechanicalwork of the combustion engine was calculated fromthe torque and the speed. The calculation of theexhaust gas enthalpy is based on Eq. 3 with themeasured exhaust gas temperatures behind thecatalytic converter. In order to consider the decreasein temperature until the measuring point, theexhaust gas temperature was increased by a max-imum of 100 K with a time-dependent compensationto calculate the exhaust gas temperature directlyafter the engine. For the calculation of the wall heat

Fig. 14. Share of wall heat losses in relation to the supplied fuelenergy and as a function of the load and engine speed. (Data usedfrom Ref. 11)

Fig. 15. Energy flow diagram of the conventional vehicle in WLTCaverage.

Fig. 16. Energy flow diagram of the hybrid vehicle in WLTC averageand operated in the charge sustaining mode.

Kober2934

losses, a function was formed in dependence on loadand speed according to Fig. 14, and thus the wallheat losses were calculated for both vehicles.

Figures 15 and 16 summarize the average valuesin the WLTC. The converted fuel energy, which iscalculated from the measured fuel consumption, isused as 100%. The miscellaneous arrow representsthe losses neglected in this consideration. Theseadditional losses are based on convection and radi-ation to the environment, as well as blowby andother heat flows into the coolant. The adjustmentwas made at this point in order to be able to presenta general comparison of the two vehicle conceptsbased on their three main energy flows.

As a result, Figs. 15 and 16 show that hybridvehicles lose their highest percentage of waste heatflows via the exhaust system. In relation to thesupplied fuel energy, the wall heat losses in thehybrid powertrain are lower, and thus the wasteheat in the exhaust tract is higher by 7 percentagepoints. Also, the efficiency of the combustion engineis 4 percentage points higher in average over thecycle. The higher efficiency of the combustionengine and the higher waste heat flows in theexhaust gas are related. Both result from the factthat the combustion engine is operated at a higherload point. Furthermore, as described in the previ-ous chapter, the exhaust gas temperatures in hybridvehicles are higher, which means that the exergeticpotential is higher in addition to the higher wasteheat flows.

The higher potential for waste heat recovery inhybrid vehicles described here results from the factthat the combustion engine is operated at higherload points. This is the case not only in hybridpowertrains. Many other efficiency measures alsoresult in an increase in the load point of thecombustion engine, such as down-sizing or cylinderdeactivation. This means that, with a wide range ofefficiency measures on combustion engines, there isalso an accompanying higher potential for wasteheat recovery.

CONCLUSIONS

A high potential for waste heat recovery in futurevehicles has been shown in this work. In manyefficiency measures, the load point of combustionengines is increased in order to improve the effi-ciency. This is used, for example, in hybrid drivetrains as well as for down-sizing and cylinderdeactivation. This work demonstrates that at ahigher load point the combustion engine has ahigher efficiency and, at the same time, the heatlosses in the exhaust gas are higher. This has beenexplained thermodynamically by the fact that thewall heat losses are lower at higher operatingpoints.

In a comparison between two vehicles of thecompact class, this relationship was proven on thebasis of measurement data from the roller

dynamometer test bench. A conventional vehicle(Volkswagen Golf) and a hybrid vehicle (OpelAmpera) were compared. The hybrid vehicle withits as a power-split hybrid power train representsthe majority of current plug-in hybrid electricalvehicles (PHEV). The comparison shows that themeasured waste heat flow of the hybrid vehicleduring operation in the charge sustaining mode is62% higher. In an additional analysis it was shownthat not only the exhaust heat losses themselves arehigher in hybrid vehicles, also the temperature levelis higher. This results in higher exergy flows in thehybrid vehicle. The analysis showed that the aver-aged exergy flow within the WLTC is 78% higher forthe hybrid vehicle. Accordingly, the potential forwaste heat recovery in hybrid vehicles is about 1.8times higher than in conventional vehicles. Fur-thermore, an additional heat input from the exhaustsystem can lead to a high increase in efficiency andcomfort due to better thermal management. Theseeffects can be even higher in hybrid vehicles than inconventional vehicles. In order to limit climatechange, all promising technologies must be used.As shown, the recovery of waste heat, for examplethrough a thermoelectric generator, can make adecisive contribution in the future.

ACKNOWLEDGMENTS

Open Access funding provided by Projekt DEAL.

OPEN ACCESS

This article is licensed under a Creative CommonsAttribution 4.0 International License, which per-mits use, sharing, adaptation, distribution and re-production in any medium or format, as long as yougive appropriate credit to the original author(s) andthe source, provide a link to the Creative Commonslicence, and indicate if changes were made. Theimages or other third party material in this articleare included in the article’s Creative Commons li-cence, unless indicated otherwise in a credit line tothe material. If material is not included in the ar-ticle’s Creative Commons licence and your intendeduse is not permitted by statutory regulation or ex-ceeds the permitted use, you will need to obtainpermission directly from the copyright holder. Toview a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

REFERENCES

1. M. Kober, Holistic Optimization of Thermoelectric Genera-tors for Automotive Applications: Reaching a Cost BenefitRatio of 81 €/(g/km). (elib DLR, 2016), https://elib.dlr.de/125922/. Accessed 10 Jan 2019.

2. M. Kober, Thermoelectric Generators for Automotive Appli-cations: A New Approach to Reach the Cost-Benefit Target.(elib DLR, 2016), https://elib.dlr.de/115947/. Accessed 10 Jan2019.

3. M. Kober and H. Friedrich, Thermoelectric Generators(TEG) with high power density for application in hybridcars. (elib DLR, 2016), https://elib.dlr.de/107902/. Accessed10 Jan 2019.

The High Potential for Waste Heat Recovery in Hybrid Vehicles: A Comparison Between thePotential in Conventional and Hybrid Powertrains

2935

4. C. Hafele, Entwicklung fahrzeuggerechter Thermoelek-trischer Generatoren zur Wandlung von Abgaswarme inNutzenergie, Forschungsbericht 2016-08 (Koln: DeutschesZentrum fur Luft- und Raumfahrt, 2016).

5. M. Kober, C. Hafele, and H. Friedrich, Methodical ConceptDevelopment of Automotive Thermoelectric Generators. (elibDLR, 2016), https://elib.dlr.de/79163/. Accessed 10 Jan 2019.

6. U. Kugler, J. Brokate, C. Schimeczek, and S. Schmid, Pow-ertrain scenarios for cars in european markets to the year2040. (elib DLR, 2017), https://elib.dlr.de/114744/ Accessed10 Jan 2019.

7. G. Berckmans, M. Messagie, J. Smekens, N. Omar, L.Vanhaverbeke, and J.V. Mierlo, Energies, 10(9). (2017). https://doi.org/10.3390/en10091314.

8. D. Grebe and L. Nitz, Voltec—the propulsion system forChevrolet Volt and Opel Ampera. AutoTechnology (ATZ) 11,28–35 (2011).

9. WLTP DHC, Development of a World-wide Worldwide har-monized Light duty driving Test Cycle (WLTC). (UN/ECE,2014), https://www.unece.org/fileadmin/DAM/trans/doc/2014/wp29grpe/GRPE-68-03e.pdf. Accessed 10 Jan 2019.

10. M. Kober, J. Electron. Mater. (2020). https://doi.org/10.1007/s11664-020-07966-6.

11. R. Pischinger, M. Klell, and T. Sams, Thermodynamik derVerbrennungskraftmaschine (Wien: Springer, 2009).

Publisher’s Note Springer Nature remains neutral withregard to jurisdictional claims in published maps and institu-tional affiliations.

Kober2936