Fuel 97 (2012) 119–124

Contents lists available at SciVerse ScienceDirect

Fuel

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

Effect of low ambient temperature on fuel consumption and pollutant and CO2

emissions of hybrid electric vehicles in real-world conditions

Robert Alvarez, Martin Weilenmann ⇑Empa, Swiss Federal Laboratories for Materials Science and Technology, Laboratory for Internal Combustion Engines, Ueberlandstrasse 129, CH-8600 Duebendorf, Switzerland

a r t i c l e i n f o

Article history:Received 17 November 2010Received in revised form 10 January 2012Accepted 11 January 2012Available online 25 January 2012

Keywords:Hybrid electric vehicleLow ambient temperatureFuel consumptionPollutant emissionsReal-world

0016-2361/$ - see front matter � 2012 Elsevier Ltd. Adoi:10.1016/j.fuel.2012.01.022

⇑ Corresponding author. Tel.: +41 44 823 48 69; faxE-mail address: robert.alvarez@empa.ch (R. Alvare

a b s t r a c t

Hybrid electric vehicles (HEVs) can potentially reduce vehicle fuel consumption and CO2 emissions byusing recuperated kinetic vehicle energy stored as electric energy in a hybrid system battery (HSB).Low ambient temperatures can affect the overall HEV powertrain operation under warm-up and hot driv-ing conditions and, consequently, affect fuel consumption and emission performance. The present studyinvestigates the influence of low ambient temperatures on HEV fuel consumption and pollutant and CO2

emissions for five in-use HEV models. Chassis dynamometer measurements have been conducted at dif-ferent set ambient temperatures using a real-world driving cycle suitable for investigating vehicle cold-start emissions. The main observation is that the amount of HEV cold-start extra emissions (CSEEs) of reg-ulated pollutants are reduced by 30% to 85% on average in comparison to sample CSEEs of conventionalgasoline vehicles. The results for HEV CSEEs of CO2 and fuel consumption are mainly similar than those ofconventional gasoline vehicles except for CSEEs of some HEVs at the ambient temperature of 23 �C. There,increased CSEEs are observed that exceed maximum sample CSEEs of conventional gasoline vehicles,reaching values for CO2 between 155 [g/start] and 300 [g/start] even though the test runs were initiatedwith maximum initial state of charge (SOC) of the HSB. Because SOC of the HSB considerably influencesthe fuel consumption of HEVs, this aspect should be further investigated in regard to the effect of lowambient temperatures on HEV fuel consumption and CO2 emissions. Moreover, no particular influenceof low ambient temperatures on HSB performance was observed during hot-phase operation.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Hybrid electric vehicles (HEVs) represent a promising approachto reduce vehicle fuel consumption and exhaust emissions of CO2.An additional electric powertrain including an energy storage de-vice, typically a rechargeable battery or supercapacitors, is com-bined with an internal combustion engine to provide the desiredoverall vehicle power output. This configuration makes it possibleto design and employ the internal combustion engine in its mostefficient operating conditions and to recuperate kinetic vehicle en-ergy during deceleration for further use, which leads to reducedoverall CO2 vehicle emissions [1–3]. HEVs are typically categorizedaccording to their capability for full electric driving (full HEV) ornot (mild HEV). HEV sales are increasing [4] and expected to attaina considerable market share in the near future [5] because of spe-cific CO2 vehicle emission reduction policies [6] and legislation[7,8]. Furthermore, HEVs are assumed to pave the way forelectricity-based powertrain solutions, such as plug-in hybrid elec-tric vehicles (PHEVs), electric vehicles (EVs) or fuel cell vehicles

ll rights reserved.

: +41 44 823 40 44.z).

(FCVs) [9]. Initial studies have already been conducted to deter-mine the real-world pollutant emission performance of HEVs [10].

An important characteristic of HEVs is the influence of lowambient temperatures on fuel consumption and pollutant andCO2 emissions. In general, emissions during cold start become par-ticularly relevant in comparison to hot emissions for modern vehi-cles with state-of-the art aftertreatment systems, especially forlow ambient temperatures [11]. The time elapsed until hot opera-tion of both the aftertreatment system and the powertrain isachieved determines the cold-start extra emissions (CSEEs) of thevehicle. The specific attribute of HEV powertrains of providingelectrical assistance to the combustion engine can therefore influ-ence this heating-up and the resulting HEV CSEEs, particularly inthe case of full HEVs because of the possibly intermittent operationof its combustion engine. Furthermore, the operational perfor-mance of the hybrid system battery (HSB) employed in HEVpowertrains may be affected by low ambient temperatures, whichcan also lead to increased fuel consumption and CO2 emissions. Inthis regard, field tests to investigate the real-world fuel economy ofHEVs and PHEVs in low ambient temperature conditions have al-ready been conducted [12], and the results show that importantfuel savings can be achieved with such vehicles at low ambienttemperatures depending on their layout and operation.

120 R. Alvarez, M. Weilenmann / Fuel 97 (2012) 119–124

To study the effect of low ambient temperatures on fuel con-sumption and pollutant and CO2 emissions of HEVs, an experimen-tal investigation with five in-use full and mild HEV models hasbeen conducted on a chassis dynamometer. Test runs with the In-rets Urbain Fluid Court (IUFC15) driving cycle have been per-formed, which consists of 15 repetitions of an urban real-worlddriving pattern and is suitable to investigate the effect of cold starton vehicle emission performance. The test results are discussed indetail to highlight the relevance of low ambient temperatures onHEV operation regarding fuel consumption and pollutant and CO2

emissions. The implications of the findings are outlined and possi-ble alternatives are proposed.

2. Methodology

2.1. Vehicle sample

The main characteristics of the in-use HEV models selected forthe test series are summarized in Table 1. HEVs 1–3 representthree identical examples of a single HEV model [13] with differentmileages. All of the HEVs used in this study are categorized as fullhybrids except HEV 5, which is a mild hybrid. As a mild hybrid,HEV 5 is capable of providing electrical assistance to the combus-tion engine and recuperate kinetic vehicle energy during decelera-tion but is not capable of full electric driving in low-speed urbandriving conditions. Note that no particular servicing was carriedout before the test runs except a general vehicle function check.

2.2. Experimental program

The repetitive cold-start real-world driving cycle Inrets UrbainFluid Court (IUFC15) was considered to investigate the effect oflow ambient temperatures on the fuel consumption and the pollu-tant and CO2 emissions of the selected HEVs. This cycle was devel-oped within the ARTEMIS [14] research program and consists of 15repetitions of a short representative European real-world urbandriving pattern [15] that are equally divided into three sectionsand hereafter referred to as subcycles, see Fig. S1. The characteris-tics of the IUFC15 cycle allows to obtain urban real-world vehiclepollutant emissions for each subcycle in different vehicle warm-up phases from cold start to hot operation and additionally ensuresthat enough hot-phase subcycle repetitions are carried out to de-rive representative average hot stabilized emissions. The impactof vehicle warm-up on pollutant emissions is therefore reflectedin the evolution of the emission performance within the singleidentical subcycles, making the IUFC15 cycle suitable for investi-

Table 1Main characteristics of the considered vehicle sample. cert.: certification; cat.: category; I

Characteristic HEV 1

Vehicle Make (–)ModelInertia settinga (kg)Gearbox (–)Cert. cat. (–)1st cert. (–) February 06Mileage (km) 32,768

IC engine Displacement (cm3)Rated power (kW)

Electric motor Rated power (kW)HSB Type (–)

Nom. voltage (V)No. of cells (–)

Level of hybridization (–)

a Empty mass plus 100 kg.

gating various aspects of the effect of real-world cold start on vehi-cle pollutant and CO2 emissions [11,16].

Single test runs with this driving cycle have been executed atambient temperatures of 23 �C, 8 �C and �7 �C for each HEV. Thevehicle soak time has been followed according to the provisionsof Council Directive 70/220/EEC [17]. The HEVs were precondi-tioned to the maximum initial state of charge (SOC) of their HSBs,defined by having no additional charge leading to the HSB in con-stant-speed vehicle traction mode, which simulates coasting condi-tions. This start-up SOC configuration of the HSB avoids any HSBrecharging process by the combustion engine during vehicle coldstart and, consequently, considerable additional CO2 emissions thatwould distort the envisaged CSEE estimation for HEVs [18]. In suchvehicle cold start conditions, the operation of the combustion en-gine of HEVs is only adjusted to additionally heat up the powertrainand aftertreatment system, making its operational demand mostcomparable to conventional vehicles. Furthermore, the air condi-tioning systems of the HEVs were turned off during the test runs be-cause its operation considerably affects fuel consumption and CO2

emissions of both conventional vehicles and HEVs [10,19].

2.3. Experimental setup

Fig. 1 shows the overall experimental setup employed for thetest series. The exhaust was sampled with a constant volume sam-pling (CVS) system. Exhaust emissions of regulated pollutants andCO2 were measured according to the statutory procedure specifiedin Council Directive 70/220/EEC [17] of storing a sample of dilutedexhaust for each of the three cycle sections in a Tedlar gas samplingbag and analyzing its content offline after completion of the testrun. Time-resolved measurements of raw exhaust pollutant andCO2 emissions were also performed, correcting the resulting signaltraces with respect to time delay due to the length of the samplelines by applying a static time shift. In both cases, adequate exhaustgas analyzers were employed as specified by Council Directive70/220/EEC [17]. The fuel consumption was then calculated fromthe measured emissions of regulated pollutants. The time-resolvedHSB wire current was measured with a clamp-on ammeter (LeCroyCP500) to meet the criteria specified in Regulation ECE R-101 ofCouncil Directive 70/220/EEC [17]. In addition, the terminal voltageof the HSB was measured using differential probe analyzers (LeCroyADP305). Both measurements were recorded with a digitalsampling oscilloscope (LeCroy WaveRunner 44Xi).

The chassis dynamometer and its settings were applied accord-ing to the provisions of Council Directive 70/220/EEC [17]. Thedriving resistance of the vehicle was simulated using the coast-down data provided by the manufacturer, and the inertia settings

C: internal combustion; HSB: hybrid system battery; nom.: nominal; No.: number.

HEV 2 HEV 3 HEV 4 HEV 5

Toyota Toyota HondaPrius II GS450 h Insight1425 2030 1301CVT CVT CVTEuro-4 Euro-4 Euro-5August 06 June 05 February 07 May 0960,761 104,266 22,335 27561497 3456 133957 218 6550 147 10NiMH NiMH NiMH201.6 288 100168 240 84Full Full Mild

Dilution TunnelBlower

ebuTdetae

H

Dilution Air Inlet

ChassisDynamometer

Heated Conduit

HFM

Online sampling

CVS-System

I V

BagOnline

NOx

CO

CO2

CH4

T.HC

(hot)

(dry & cold)

SampleBags

Fig. 1. Schematic diagram of the test setup. HFM: hot-film air-mass flow meter; CVS: constant volume sampling; V: measurement of the HSB terminal voltage; I:measurement of the HSB wire current.

R. Alvarez, M. Weilenmann / Fuel 97 (2012) 119–124 121

were set at empty vehicle mass plus a payload of 100 kg (Table 1).The ambient temperature of the test cell was controlled to 23 �C,8 �C and �7 �C for the respective test runs, and the relative airhumidity was set to 50% where possible, i.e., for 23 �C. All HEVswere operated with the same standard fuel with low sulfurcontent.

2.4. CSEE estimation

The effect of low ambient temperatures on the fuel consump-tion and pollutant and CO2 emissions of the considered HEVswas assessed by determining the respective cold-start extra emis-sions (CSEEs) for the test runs conducted with the IUFC15 cycle atthe different ambient temperatures. Out of the different methodsto determine CSEEs, the so-called subcycle method is selected be-cause it is best suited for vehicles with unsteady hot emissions [16]like HEVs. This method is established and has already been appliedin several studies investigating vehicle CSEEs [11,16]. In thesestudies, the emission results from the cycle IUFC15 are separatedinto a warm-up phase and a hot phase, depending on the CO2 emis-sion performance in the single subcycles, which are calculated outof the measured time-resolved CO2 emission measurement. Theaverage hot stabilized pollutant and CO2 emissions of the hot phaseare then subtracted from the respective subcycle emissions of thewarm-up phase to obtain absolute CSEEs per vehicle start.

3. Results

3.1. CSEEs of regulated pollutants

The CSEEs of regulated pollutants obtained for the single HEVsat the different ambient temperatures are summarized in Fig. 2and compared to sample CSEEs of gasoline vehicles of the certifica-tion category Euro-4 (G4), which were obtained within an earlierstudy applying the same determination methodology [11]. Theaverage pollutant CSEEs of the selected HEVs are reduced by 30%to 85% with respect to the average CSEEs of the G4 sample. Thisperformance is particularly remarkable for HC emissions, where

most pollutant CSEEs recorded for the HEVs are even below theminimum detected CSEEs of the G4 sample.

These observations highlight the good emission performancefor pollutant CSEEs of HEVs. A specific advantage of HEV power-trains is assumed to be responsible here: the ability of the electricdrivetrain to assist the combustion engine when providing driveenergy allows adjusting the operation of the latter during cold startto heat up the three-way catalyst as quickly as possible. This heat-ing-up occurs without the excessive formation of pollutants thatshould further be abated in the three-way catalyst because peakload demands to the combustion engine are avoided thanks tothe electric assistance, especially for full HEVs. The observed pollu-tant CSEEs of HEVs are in line with the overall good pollutant emis-sion performance of HEVs in real-world driving conditionscompared to gasoline vehicles, as reported in [10]. In that study,little influence of the initial SOC of the HSB on regulated pollutantemissions is also reported for cold-started driving cycles, whichindicates that their CSEEs may not vary considerably on accountof initial SOC of the HSB.

3.2. CSEEs of CO2 and fuel consumption

The results obtained for the CSEE analysis of CO2 emissions andfuel consumption of the selected HEVs for the different ambienttemperatures are given in Fig. 3. In contrast to the CSEEs of regu-lated pollutants, the HEV CSEEs of CO2 and fuel consumption aremainly similar to the respective sample CSEEs of the gasoline vehi-cles G4 [11]. The CSEEs of some HEVs exceed the average sampleemissions of the G4 sample and even their maximum sample emis-sions at the ambient temperature of 23 �C. A specific attribute ofHEV powertrains can again explain this finding: the measuredhot stabilized emissions of CO2 and fuel consumption of HEVsare reduced in comparison to the respective average sample emis-sions of G4 due to the electric assistance, but the particular fuelused to heat up the three-way catalyst of HEVs remains approxi-mately the same. This circumstance influences the respectiveCSEEs more than it does for gasoline vehicles, which leads to in-creased CSEEs for HEVs.

0102030405060708090

CO

[g/s

tart]

ambient temperature [

G4HEV 1HEV 2HEV 3HEV 4HEV 5

0

5

10

15

20

25

HC

[g/s

tart]

ambient temperature [

G4HEV 1HEV 2HEV 3HEV 4HEV 5

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

-30 -20 -10 0 10 20 30 -30 -20 -10 0 10 20 30

-30 -20 -10 0 10 20 30

NO

x[g

/sta

rt]

ambient temperature [

G4HEV 1HEV 2HEV 3HEV 4HEV 5

161.2 39.2

°°C] °C]

°C]

Fig. 2. Measured CSEEs of regulated pollutants of the selected HEVs for the different ambient temperatures compared to the average CSEEs of a sample of Euro-4 gasolinevehicles [11]. The error bars represent maximum and minimum sample emissions.

0.000.010.020.030.040.050.060.070.080.090.10

Rec

up. e

l. dr

ive

0.000.010.020.030.040.050.060.070.080.090.10

Prov

ided

el.

driv

e

ener

gy [k

Wh]

ene

rgy

[kW

h]

negligiblefor mild HEV 5

A23 C 8 C -7 C23 C 8 C -7 C

B

HEV 1 HEV 2 HEV 3 HEV 4 HEV 5HEV 1 HEV 2 HEV 3 HEV 4 HEV 5

° ° ° ° ° °

Fig. 4. Average and standard deviation of electric drive energy provided from (A) and recuperated to (B) the HSB in full electric vehicle operation determined in the single hot-phase subcycles of the IUFC15 cycle for the selected HEVs at the different ambient temperatures; el.: electric; recup.: recuperated.

050

100150200250300350400450500

CO

2[g

/sta

rt]

ambient temperature [

G4HEV 1HEV 2HEV 3HEV 4HEV 5

0.00

0.05

0.10

0.15

0.20

0.25

0.30

-30 -20 -10 0 10 20 30 -30 -20 -10 0 10 20 30

FC [

l/sta

rt]

ambient temperature [

G4HEV 1HEV 2HEV 3HEV 4HEV 5

°°C] °C]

Fig. 3. Measured CSEEs of CO2 and fuel consumption of the selected HEVs for the different ambient temperatures compared to average CSEE of a sample of Euro-4 gasolinevehicles [11]. The error bars represent the maximum and minimum sample emissions.

122 R. Alvarez, M. Weilenmann / Fuel 97 (2012) 119–124

Furthermore, a significant drop in the CSEEs is also detected forHEV 2 at �7 �C and 23 �C and for HEV 4 at 8 �C ambient tempera-ture. In those instances, a pronounced short-term use of electricenergy out of the HSB for full electric driving is observed in one

of the subcycles during the warm-up phase, see Fig. S2, whichresults in lower overall CO2 emissions and fuel consumption forthis period and, consequently, lower CSEEs. This occurrence maybe rather exceptional and mainly due to the uncommon initial

0.98

1.00

1.02

1.04

1.06

1.08

1.10

1.12

-30 -20 -10 0 10 20 30

rel.

hot C

O2

emis

sion

[-]

ambient temperature [ °°C]

G4

HEV 1

HEV 2

HEV 3

HEV 4

HEV 5

Fig. 5. Relative average hot-phase CO2 emissions of the selected HEVs and of the G4sample in the IUFC15 cycle for the different ambient temperatures.

R. Alvarez, M. Weilenmann / Fuel 97 (2012) 119–124 123

maximum SOC of the HSB. However, the resulting CO2 emissionsand fuel consumption are also considered to be representativefor the respective HEV operation. In contrast, the mild hybridizedHEV 5, which has the most similar operational behavior of thepowertrain compared to conventional gasoline vehicles, featuresalmost identical CSEEs compared to sample CSEEs of the G4vehicles.

The CSEEs of CO2 and fuel consumption presented in Fig. 3 areto be judged optimistically because the test runs were started withfully charged HSB. Otherwise, the selected HEVs would either notprofit from the availability of extra electric drive energy or wouldimmediately need to recharge the HSB with the combustion engineto a certain SOC level as defined by their powertrain control sys-tem, either way leading to increased emissions of CO2 and fuel con-sumption during the warm-up phase. The latter can be pronouncedwhen considering the sensitivity of the initial SOC of the HSB onHEV CO2 emissions in urban driving patterns [18], such as theIUFC15 cycle, and therefore reflected in considerably increasedCSEEs for HEVs.

3.3. Hot-phase HEV operation

Another interesting aspect is the hot-phase operation of theHEV powertrains at different ambient conditions because varyingambient temperatures may affect the underlying electrochemicalprocesses in the HSB depending on its type and, therefore, influ-ence its performance. Consequently, the resulting CO2 emissionsand fuel consumption can also be affected [10,18]. Therefore, theaverage electric drive energy provided from and recuperated tothe HSB in full electric vehicle operation was determined for theselected HEVs in the single measured hot-phase subcycles of theIUFC15 cycle at the different ambient temperatures, see Fig. 4.No particular differences in the HSB performance are observed dur-ing these driving conditions for the different ambient tempera-tures. A slight trend towards reduced performance with lowerambient temperatures can only be observed for HEV 1, but is notstatistically significant because of the standard deviation of thedata series. This slight trend is reflected in the clearly steeper in-crease of the average hot-phase CO2 emissions of HEV 1 for lowerambient temperatures compared to those of the other HEVs andthe G4 sample, see Fig. 5. In general, however, a significant influ-ence of low ambient temperatures on the HSB performance ofthe selected HEVs resulting in more pronounced CO2 emissionsand fuel consumption can be excluded given the test results.

4. Summary and conclusions

The present experimental investigation with five in-use HEVsoffers insight into the effect of low ambient temperatures on theirfuel consumption and pollutant and CO2 emissions in real-world

driving conditions. The test results show that the CSEEs of regu-lated pollutants of HEVs are reduced by 30% to 85% on average incomparison to sample CSEEs of conventional gasoline vehicles ofthe same certification category. In contrast, the CSEEs of CO2 andfuel consumption are mainly similar to those of conventional gas-oline vehicles except for the test results at the ambient tempera-ture of 23 �C, where some HEVs feature further increased CSEEs.

Both observations are explained by the specific attributesof HEV powertrains, in which the electric assistance to thecombustion engine allows an adapted and more selective opera-tion of the latter. Apart from saving fuel and reducing CO2 emis-sions while driving, this circumstance permits a quick heating-upof the three-way catalyst without emitting pronounced amountsof pollutants during that period, thus keeping the resulting pollu-tant CSEEs low. But the fuel used for this heating-up becomes morerelevant in comparison to the fuel savings while driving due to thehybrid powertrain, which leads to somewhat increased CSEEs ofCO2 and increased fuel consumption with respect to conventionalgasoline vehicles, especially for full HEVs.

It can be concluded that the influence of low ambient tempera-tures on the emission performance of HEVs is mainly similar to theinfluence on gasoline vehicles. Namely, lower CSEEs are observedfor regulated pollutants and similar CSEEs for CO2 and fuel con-sumption with selective exceedance at the ambient temperatureof 23 �C. The CSEEs for CO2 and fuel consumption are assumed tobe particularly sensitive to the initial SOC of the HSB, becausethe latter considerably influences the fuel consumption of HEVs[10,18]. This aspect should at least be taken into account or furtherinvestigated to evaluate in detail the influence of low ambienttemperatures on fuel consumption of HEVs. Therefore, cold startrepetition measurements with HEVs with different initial SOC oftheir HSBs appear most suitable. Alternatively, preconditioningthe HEVs before such test runs to the SOC of the HSB in regularvehicle operation, as defined by its powertrain control system,should be considered. Moreover, no particular influence of lowambient temperatures on the HSB performance of HEVs and, con-sequently, on their fuel consumption and CO2 emissions are ob-served in hot-phase driving conditions.

Acknowledgements

The authors thank the Swiss Federal Office for the Environment(FOEN) for principally funding the study and Toyota Motor Europeand Honda R&D Europe for providing valuable technical informa-tion and support.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.fuel.2012.01.022.

References

[1] Sundström O. Optimal Control and Design of Hybrid Electric Vehicles[dissertation]. Zurich (CH): Swiss Federal Institute of Technology; 2009.

[2] Sundström O, Guzzella L, Soltic P. Torque-assist hybrid electric powertrainsizing: from optimal control towards a sizing law. IEEE Trans Contr Syst T2010;18(4):837–49.

[3] Sundström O, Guzzella L, Soltic P. Optimal hybridization in two parallel hybridelectric vehicles using dynamic programming. In: Proceedings of the 17th IFACworld congress, Seoul, Korea; 2008 [July 6–11].

[4] Alternative Fuels & Advanced Vehicles Data Center (AFCD). US Hybrid-ElectricVehicle Sales; 2010. <http://www.afdc.energy.gov/afdc/data/vehicles.html>.

[5] International Energy Agency – Hybrid and Electric Vehicle ImplementingAgreement (IEA-HEV). Outlook for hybrid and electric vehicles; 2008. <http://www.ieahev.org/pdfs/ia-hev_outlook_2008.pdf>.

[6] An F. International comparison of policies to reduce greenhouse gas emissionsfrom passenger vehicles. In: Sperling D, Cannon JS, editors. Driving climate

124 R. Alvarez, M. Weilenmann / Fuel 97 (2012) 119–124

change – cutting carbon from transportation. Amsterdam: Academic Press. p.143–64.

[7] Regulation EC 443/2009 of the European Parliament and of the Council of 23April 2009 setting emission performance standards for new passenger cars aspart of the Community’s integrated approach to reduce CO2 emissions fromlight-duty vehicles. Official Register of the European Union; 2010. <http://eur-lex.europa.eu/Notice.do?val=496639:cs&lang=en&list=496639:cs,&pos=1&page=1&nbl=1&pgs=10&hwords=443/2009~>.

[8] Fontaras G, Samaras Z. On the way to 130 g CO2/km – estimating the futurecharacteristics of the average European passenger car. Energy Policy2010;38:1826–33.

[9] Van Mierlo J, Maggetto G, Lataire Ph. Which energy source for road transport inthe future? A comparison of battery, hybrid and fuel cell vehicles. EnergyConvers Manage 2006;47:2748–60.

[10] Fontaras G, Pistikopoulos P, Samaras Z. Experimental evaluation of hybridvehicle fuel economy and pollutant emissions over real-world simulationdriving cycles. Atmos Environ 2008;42:4023–35.

[11] Weilenmann M, Favez JY, Alvarez R. Cold-start emissions of modern passengercars at different low ambient temperatures and their evolution over vehiclelegislation categories. Atmos Environ 2009;43:2419–29.

[12] Laurikko J, Pellikka AP. Fuel economy and energy consumption of hybrid andplug-in hybrid cars in Nordic climate conditions. FISITA 2010: In: Proceedings

of the FISITA world automotive congress, Budapest, Hungary; 2010 [May 30–June 4].

[13] Danisch R, Goppelt G. Der neue Toyota Prius. Auto Technol 2004;3:186–9.[14] ARTEMIS: Assessment and Reliability of Transport Emission Models and

Inventory Systems; 2010. <http://www.trl.co.uk/ARTEMIS>.[15] Joumard R, Andre JM, Rapone M, Zallinger M, Kljun N, André M, et al. Emission

factor modelling and database for light vehicles – Artemis deliverable 3. Bron(F): Inrets; 2007 [Report No.: LTE 0523].

[16] Favez JY, Weilenmann M, Stilli J. Cold start extra emissions as a function ofengine stop time: evolution over the last 10 years. Atmos Environ2009;43:996–1007.

[17] Council Directive 70/220/EEC of 20 March 1970 on the approximation of thelaws of the Member States relating to measures to be taken against airpollution by gases from positive-ignition engines of motor vehicles; 2010.<http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:1970L0220:20070101:en:PDF>.

[18] Alvarez R, Schlienger P, Weilenmann M. Effect of hybrid system batteryperformance on determining CO2 emissions of hybrid electric vehicles in real-world conditions. Energy Policy 2010;38:6919–25.

[19] Weilenmann M, Alvarez R, Keller M. Fuel consumption and CO2/pollutantemissions of mobile air conditioning at fleet level – new data and modelcomparison. Environ Sci Technol 2010;44:5277–82.