Automotive test drive cycles for emission ...

24 downloads 70513 Views 216KB Size Report
about the representative drive cycles used by the various vehicle certification and regulatory authorit- ... 1 ECE R49 13-mode cycle used for automotive heavy-.
Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering http://pid.sagepub.com/

Automotive test drive cycles for emission measurement and real-world emission levels-a review S Samuel, L Austin and D Morrey Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering 2002 216: 555 DOI: 10.1243/095440702760178587 The online version of this article can be found at: http://pid.sagepub.com/content/216/7/555

Published by: http://www.sagepublications.com

On behalf of:

Institution of Mechanical Engineers

Additional services and information for Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering can be found at: Email Alerts: http://pid.sagepub.com/cgi/alerts Subscriptions: http://pid.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://pid.sagepub.com/content/216/7/555.refs.html

>> Version of Record - Jul 1, 2002 What is This?

Downloaded from pid.sagepub.com by guest on September 6, 2012

555

Review Paper

Automotive test drive cycles for emission measurement and real-world emission levels—a review S Samuel, L Austin and D Morrey* School of Engineering, Oxford Brookes University, Oxford, UK Abstract: The emission levels produced by any vehicle are dependent on the mode of operation of the vehicle and technology behind the vehicle design. The test drive cycles employed to measure the emissions produced by vehicles should adequately represent the real-world driving pattern of the vehicle to provide the most realistic estimation of these levels. However, there is increasing concern about the representative drive cycles used by the various vehicle certication and regulatory authorities. This paper reviews the various drive cycles used for gasoline engine vehicles in Europe and the United States, and the impact of various factors and their inuence on real-world emission levels. The proposed new drive cycles of the United States and Europe are considered. From the work reviewed, it can be concluded that the amount of pollutant levels from automotive vehicles are underestimated because of the characteristics of the existing drive cycles. While much work remains to be done with the development of new drive cycles to represent real-world driving patterns, some useful conclusions can be drawn regarding the impacts of the factors reviewed here. The impacts of the factors reviewed in this paper can be characterized to improve estimations and simulations of the real-world emission levels of the vehicle. Keywords: drive cycles, real-world emission, emission performance, driving pattern, driving behaviour, automotive, internal combustion engine NOTATION CARB ECE EPA EUDC FTP INRETS MEET NEDC NREL ORNL SFTP ¨V TU

1 California Air Research Board UN Economic Commission for Europe Environmental Protection Agency Extra-Urban Drive Cycle federal test procedure Institut National des Recherches sur les Transports et leur Securite´ methodologies for estimating air pollution from transport new European driving cycle National Renewable Energy Laboratory Oak Ridge National Laboratory supplemental federal test procedure ¨ V Rheinland Sicherheit und TU Umweitschutz GmbH

This review work concerns the characteristics of the driving cycles employed for emission measurement in Europe and the United States. In Europe, concerns about atmospheric pollution have been steadily rising since the 1980s, and since January 1993 all new petrol passenger vehicles sold within the European countries have been equipped with three-way catalysts. Also there is concern regarding the ability of the current drive cycles to represent the realworld driving patterns [1–5]. Some events such as sharp acceleration at high speeds are not common in actual driving but they contribute a disproportionate share of total carbon monoxide (CO) emission. Such events are not adequately represented in the test drive cycles used in European countries [6 ]. This paper presents a detailed literature review to identify the real-world driving patterns available and proposed new driving cycles from various agencies for gasoline powered vehicles. 2

The MS was received on 30 October 2001 and was accepted after revision for publication on 27 March 2002. * Corresponding author: Oxford Brookes University, School of Engineering, Gypsy Lane Campus, Headington, Oxford OX3 0BP, UK.

INTRODUCTION

TEST DRIVE CYCLES

The various vehicle certication and regulatory authorities use two basic philosophies. According to the rst, the

Downloaded from pid.sagepub.com by guest on September 6, 2012

D12401 © IMechE 2002

Proc Instn Mech Engrs Vol 216 Part D: J Automobile Engineering

556

S SAMUEL, L AUSTIN AND D MORREY

ances of these cars are tested using ECE drive cycles. The ECE cycle performed on a chassis dynamometer is used for emission certication of light-duty vehicles in Europe and is shown in Fig. 2. The entire cycle includes four modes of driving as shown in Fig. 2. The extraurban driving cycle ( EUDC ) segment was added after the fourth ECE cycle to account for more aggressive high-speed driving modes. The maximum speed of the EUDC cycle is 120 km/ h [8–10]. 2.2 FTP test drive cycles

Fig. 1

ECE R49 13-mode cycle used for automotive heavyduty applications [5, 7] (BMEP, brake mean eVective pressure)

driving cycle is made up of a series of repetitions of a composite of various vehicle-operating conditions representative of typical driving models as shown in Fig. 1. The Economic Commission for Europe ( ECE) and Japanese cycles reect this philosophy. The methodology of testing of an engine for the legislative emission norms shown in Fig. 1 is applicable to automotive engines for heavy-duty applications. However, Fig. 1 shows the general philosophy of testing an engine followed in European Union ( EU ) countries and Japan [5]. According to the second, the composite of driving modes is an actual simulation of a road route. The United States, Canada, Australia, Sweden and Switzerland all use a version of a federal test procedure ( FTP). Individual drive cycles and their characteristics will be now considered in turn.

2.1 European emission legislation and ECE test drive cycles Strict new emissions requirements were introduced in European Community ( EC ) countries in the early 1990s eVectively requiring new petrol-engine passenger cars to be tted with catalytic converters. Emission perform-

Fig. 2

The FTP cycle is shown in Fig. 3. The FTP cycle [11] consists of 23 cycle tests in order to represent diVerent modes of driving. The last 5 cycles of the 23 cycle test are simply a repeat of cycles 1 to 5 after 10 min of vehicle soak. Cycle 11 is a demanding test of an engine controller’s air–fuel ratio performance. Most of the cruise portion of cycle 11 consists of low-engine-load throttle transients that typically bring out the worst in an engine controller performance in the real world. Cycle 2 and the equivalent of the twentieth cycle of the FTP require the greatest vehicle speeds and acceleration of the Environmental Protection Agency ( EPA) 23 cycle test. The highest loads are seen within the rst 60 s of cycle 2. In general an entire FTP cycle could be divided into a ‘transient’ portion and a ‘stabilized’ portion with a total cycle time of 1874 s, a driving distance of 17.8 km and an average speed of 34.1 km/ h. Two such cycles are used: one with the vehicle at an ambient temperature of 16–30 °C before start (cold cycle), and one with the engine control system hot (hot cycle) after a 10 min shutdown after running the cold cycle as shown in Fig. 3. 2.3 Comparison of ECE and FTPP drive cycles Vehicle emissions come principally from three sources: the exhaust, evaporative emission from fuel systems and crankcase ventilation gases. To set the standard to the maximum allowable level of emission in grams per mile, two major aspects must be dened, namely the driving cycles to be used and nally the sampling method. A comparison study carried out by Hadded et al. [12] considers various FTP and ECE emission limits. The

ECE test drive cycle (BS, beginning of cycle; ES, end of cycle) Downloaded from pid.sagepub.com by guest on September 6, 2012

Proc Instn Mech Engrs Vol 216 Part D: J Automobile Engineering

D12401 © IMechE 2002

AUTOMOTIVE TEST DRIVE CYCLES FOR EMISSION MEASUREMENTS

Fig. 3

FTP-75

emission limits are similar, although the vehicles are being tested using diVerent types of drive cycle. The work presented here reviews the drive cycles used for emission measurement, which has a strong inuence on emissions values [1–4]. The existing test drive cycles and the characteristics of ECE and FTP cycles are given in Table 1.

3

FACTORS AFFECTING WORLD EMISSION LEVELS

Actual emission levels from vehicle sources are a function of several variables, namely driver behaviour, travelrelated highway network characteristics and vehicle characteristics. The individual factors are dealt with in the following sections.

3.1 Driver-related factors Recent researchers [4, 13] have found that emissions are also aVected by smoothness and consistency of vehicle speed, which are heavily aVected by driving behaviour and traYc conditions. Sharp acceleration from overtaking or changing lanes, merging on to a freeway from a Table 1

Test cycle

Purpose

Transient FTP 75

Cars and light-duty trucks Cars and light-duty trucks Cars and light-duty trucks

ECE EUDC (2000)

557

slip road or leaving a signalized intersection impose heavy loads on the engine, which result in high emission levels. Vehicle accelerations produce emissions due to an engine operating strategy called power enrichment. When heavy loads are placed on the engine during acceleration, vehicles are designed to operate with a richer fuel–air mixture to prevent engine knock and damage to the catalytic converter [14–16]. This provides good driving performance but overloads the catalytic converter, thereby producing high levels of emissions due to the substantial reduction in catalytic converter eYciency. In this case CO levels are most aVected, followed by volatile organic compounds ( VOCs) but there is little eVect on the NO emission. x Sierra Research [4] has reported that aggressive driving with many accelerations results in CO emission levels 15 times higher, and VOC levels 14 times higher, than those resulting from average driving. They obtained these results by comparing time–speed–emission traces for the same 11 km trip, as shown in Fig. 4. Using instrumented vehicles in Atlanta, researchers [4] found that most vehicles spend less than 2 per cent of the total driving time in this mode, but that this small fraction of the trip can account for up to 40 per cent of the total emissions produced.

Comparison of ECE and FTP cycles

Distance simulated (km)

Average speed (km/h)

Maximum speed (km/h)

Maximum acceleration (m/s2)

17.86

34.1

91.2

1.6

Chassis dynamometer, include a cold start segment and frequent stops

4.052

18.8

62.6

0.487

Chassis dynamometer, based on Paris driving conditions. Emissions measured after 40 s idling

6.955

50

120

0.395

Addition to ECE. Freeway driving

Comment

Downloaded from pid.sagepub.com by guest on September 6, 2012

D12401 © IMechE 2002

Proc Instn Mech Engrs Vol 216 Part D: J Automobile Engineering

558

S SAMUEL, L AUSTIN AND D MORREY

Fig. 4

Driver-related emission: time–speed graphs for an ‘average’ driver and an ‘aggressive’ driver (data from Sierra Research [4])

3.2 Acceleration and speed 3.2.1 Acceleration For most passenger vehicles, the maximum acceleration rate ranges from about 3.6 to 2.1 m/s2 as vehicles start from rest, depending upon their weight–power ratio. Normal acceleration rates for passenger cars are signicantly lower than the maximum acceleration rates and are typically estimated at about 1.1 m/s2 [13]. The newer medium-size vehicles which are available now in the market have a maximum acceleration capacity of 5 m/s2 [17]. Data collected by the Oak Ridge National Laboratory (ORNL) for 1300–1600 individual vehicle data points and the predicted emission by Ahn et al. [2] are shown in Fig. 5. This illustrates that both instantaneous speed and acceleration signicantly aVect the emission. Vehicle acceleration becomes the most inuential factor for CO and hydrocarbon (HC ) emission levels, especially at high speeds. The minor variations in speed and throttle position, which have been found to result in signicant emission increases relative to true steady state operation, have also been reported in the recent work of the EPA [18]. 3.2.2 Average speed Predicted emission levels [1] as functions of average vehicle speed are shown in Fig. 6. Based on the predictions presented by Ahn et al. [2], emission levels are generally high under low-speed congested driving conditions and fall at intermediate speeds in low-density traYc conditions. All emission levels increase with increasing vehicle speed [9, 18]. 3.3 Contribution of auxiliary loads The increasing impact of auxiliary loads on vehicle fuel economy is a critical parameter for the designers of these

items. However, the impact of heating, ventilation and air conditioning has escaped the FTP and ECE emission test drive cycles. Very little research [19–21] has been conducted into the impact of climatic control loads and auxiliaries on vehicle emission levels. Research conducted to date includes the investigation of the impact of (a) air-conditioning loads, (b) ventilation control equipment and (c) cabin warm-up auxiliaries. The results from the modelling work of Farrington et al. [20] show that air-conditioning systems can increase tailpipe emissions signicantly, more than doubling the CO and nitrogen oxides (NO ) depending on x the engine modelled. These results are given in Table 2. This modelled work demonstrates the engine–engine variation and also the eVect of the coeYcient of performance (COP) of the air-conditioning system on vehicle emission levels. The COP of the air-conditioning system is the measure of the cooling eYciency of the system at a specied constant temperature. The net COP is dened as the product of the air-conditioning system COP and the compressor COP. In the past, engine manufacturers have simulated the air-conditioning load by adding load to the engine during testing. The EPA also simulated the impacts of air-conditioning during emission testing by increasing the load on the vehicle. Testing has shown [18] that these simulations do not adequately represent the true impact of air-conditioning use on emission levels. When vehicles are tested on a chassis dynamometer with the air conditioning actually ‘on’ in an environmentally controlled test cell, the emission impacts observed have been considerably greater than those seen by simulating the impact of air conditioning on emission tests by the added load of the engine. While much work remains to be done, the broad outlines of the impacts of

Downloaded from pid.sagepub.com by guest on September 6, 2012

Proc Instn Mech Engrs Vol 216 Part D: J Automobile Engineering

D12401 © IMechE 2002

AUTOMOTIVE TEST DRIVE CYCLES FOR EMISSION MEASUREMENTS

Fig. 5

559

Vehicle emission as a function of vehicle speed and acceleration (data taken from Ahn et al. [2])

Downloaded from pid.sagepub.com by guest on September 6, 2012

D12401 © IMechE 2002

Proc Instn Mech Engrs Vol 216 Part D: J Automobile Engineering

560

S SAMUEL, L AUSTIN AND D MORREY

these changes can be characterized at this stage for real-world emission.

3.4 Road gradient The review work of Ericsson [9] showed that hill ascents produce consistently high NO emission levels, while x descents result in low levels. The extra emissions produced while travelling uphill are not necessarily fully compensated by a corresponding reduction in emissions when travelling downhill. Emission ratios for diVerent gradient classes reecting the change in emission rate compared with that on a at road were developed. The reported work shows that the road gradient could be an important factor governing emissions produced by a car with a catalytic converter if the required engine performance is outside the range for which the engine management system is optimized. A detailed work on the eVect of road gradient on vehicle emission and fuel consumption has been carried ¨ V [22] under the methodologies for estimating out by TU air pollution from transport (MEET ) project. The emission measurements for the gradient classes from 6 to+6 per cent were carried out using chassis dynamometer tests. These gradients represented the gradients encountered in the Swiss and German road networks. The ECE drive cycle, US FTP 75, US highway driving cycle and a special autobahn cycle were used for the study of road gradients For steeper gradients the driving pattern of these cycles could not be achieved and, hence, a special driving pattern was developed and used by ¨ V beyond+2 and 2 per cent gradients. A greater TU inuence of road gradient on vehicle emission was ident¨ V. However, all legislative test drive cycles ied by TU assume roads with zero gradient and also straight roads without any horizontal curvature. The eVect of average speed on vehicle emission

Fig. 6

3.5 Horizontal curvature

Table 2

Predicted increase in tailpipe emissions resulting from the use of two diVerent air-conditioning systems during the SCO3 drive cycle Engine net COP=2.25

Net COP=1.25

Base cycle FTP

HC (%)

CO (%)

NO x (%)

HC (%)

CO (%)

NO x (%)

1.5 l 1.9 l 3.0 l 3.0 l

31 4 24 18

22 51 26 11

52 39 29 31

50 13 46 29

50 125 68 20

113 58 56 54

Geo Saturn Dodge Toyota

An experimental and computer program developed by the Federal Highway Administration (FHWA) [23] on highway eVects on vehicle performance estimated the eVect of horizontal road curvature on vehicle performance and emission levels. Preliminary tests carried out by the FHWA using light-duty vehicles showed that the increase in engine torque due to horizontal curvature is signicant when compared with the engine torque required to drive the vehicle on straight road. The eVect of horizontal road curvature can have a signicant contribution to the real-world emission in a country such as the UK where roundabouts and islands in the road networks are widely used.

Downloaded from pid.sagepub.com by guest on September 6, 2012

Proc Instn Mech Engrs Vol 216 Part D: J Automobile Engineering

D12401 © IMechE 2002

AUTOMOTIVE TEST DRIVE CYCLES FOR EMISSION MEASUREMENTS

4

UNCERTAINTIES OF ECE AND FTP CYCLES

The current ECE and FTP drive cycles estimate emission solely based on changes in average speed. This onedimensional approach cannot adequately describe the underlying distribution of speeds and acceleration which vary with vehicle, driving behaviour, road conditions and the level of congestion, each of which produce potentially large variations in emission levels [21, 25–27]. The FTP cycle does not distinguish between driving on arterial highways and driving on freeways, although each has a quite distinct frequency distribution of speed and acceleration likely to result in diVerent emission levels. The EUDC drive cycle attempts to reect the driving pattern in urban areas. However, the sharp acceleration and aggressive behaviour of drivers are not clearly reected in the EUDC drive cycle. In addition, the results of some research [9, 13] shows that real-world emission levels are more sensitive to the level of acceleration than speed. However, these drive cycles rely on average speed as the sole describer of traYc ow. The variability of speed, road grade and variation in acceleration are not reected in these drive cycles. Another factor is the limitation of drive cycles, virtually all emission testing has been based on a limited set of test cycles with questionable representation of specic traYc ow conditions. There is a pressing need for the development of test drive cycles sensitive to a wider variety of driving patterns than current cycles. These need to be truly representative of the ranges of vehicles on the road today to enable collection of vehicle activity data to use as inputs to traYc ow simulation and emission estimation models. A comparison of three drive cycles [4] with approximately the same average speed is given in Fig. 7, a base-

561

line FTP cycle, a real-world drive cycle on an arterial road and a freeway drive cycle in Los Angeles. The average speeds of all these three drive cycles were similar. However, the acceleration behaviours of the drive cycles are entirely diVerent. The work of the Sierra Research Centre demonstrated that the FTP underestimates driving at higher speeds and acceleration [4], both of which are sources of high emission. The correlation between the fuel economy of the vehicle and the emission produced by the vehicles are available in references [7] and [24, 28]. The fact that fuel-eYcient cars produce fewer pollutants is evident from the new vehicle technology. The factors that aVect the fuel economy of the vehicle are given in Table 3. The drive cycles, which should represent real-world driving patterns, should reect the inuence of these factors in order to accurately predict the real-world emission levels.

5

NEW DRIVE CYCLES

Research to improve the understanding of the emission characteristics of motor vehicles under real-world driving conditions has begun around the world [1, 3, 4, 6, 8, 9]. The research to date has focused on the eVects of driving patterns on emissions at sharp accelerations and high speeds. Neither is well represented by current drive cycles, and it is suspected that they are major reasons for current underestimation of emission levels. The EPA [18] has initiated research to review the current vehicle testing procedures in order to ensure that they reect actual driving conditions by conducting surveys of driving behaviour in selected cities. The California Air Research Board (CARB) has also sponsored similar research in the Los Angeles area [4]. The research sponsored by the EPA has conrmed that sharp accelerations and high speeds are not well represented in the baseline FTP drive cycles. The current maximum acceleration rate of 1.47 m/s2 is frequently exceeded in on-the-road driving. Similarly, the current maximum FTP speed of 90.7 km/ h is routinely exceeded [4, 5, 20]. CARB tested 125 light-duty vehicles on newly developed drive cycles [4]. There are seven freeway cycles and three arterial cycles that have been constructed from actual driving. A new test drive cycle was recently added to the ECE cycle in order to represent the high-speed behaviour of the engine. However, the real-world acceleration characteristics are not included in the EUDC cycle.

5.1 Proposed FTP cycles

Fig. 7

Envelope of speeds and accelerations for three cycles with similar mean speeds [4]

A new US emission test, the supplemental federal test procedure (SFTP) has recently been introduced in order to measure tailpipe emissions with the air-conditioning

Downloaded from pid.sagepub.com by guest on September 6, 2012

D12401 © IMechE 2002

Proc Instn Mech Engrs Vol 216 Part D: J Automobile Engineering

562

S SAMUEL, L AUSTIN AND D MORREY

Table 3

Factors aVecting the fuel economy reductions caused by various factors [7, 24, 28] Reduction in fuel economy (%)

Factor

Conditions

Average

Maximum

Temperature Idling /warm-up Defroster Head wind Uphill driving Poor road conditions Congested traYc

20 °F (7 °C ) versus 77 °F (25 °C) Winter versus summer Extreme use 20 mile/h (32 km/h) 7% grade Gravel, curves, slush, snow, etc. 20 mile/h (32 km/h) versus 27 mile/h (43 km/h) average speed 70 mile/h versus 55 km/h Hard versus easy 0.5 in (12.7 mm) Extreme use

5.3 Variable with driver Analogous to air conditioning on some vehicles 2.3 1.9 4.3 10.6

13 20 6 25 50 15

N/A* 11.8