STUDY ON
EMISSION CONTROL TECHNOLOGY FOR HEAVY-DUTY VEHICLES FINAL REPORT VOLUME 5 IN-USE CONFORMITY TESTING OF EMISSION CONTROL DEVICES
CONTRACT N° ETD/00/503430 Study prepared for the European Commission – DG ENTR (Enterprise) Joint effort by MIRA Ltd, United Kingdom PBA, United Kingdom LAT/AUTh, Greece TU Graz, Austria TNO Automotive, Netherlands Vito, Belgium
May 2002
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
This part of the project was carried out by
Vito Centre of expertise
Energy Technology Boeretang 200 B-2400 Mol Belgium Contact
In collaboration with
Guido Lenaers Tel. 32 - 14 33 58 14 Fax 32 - 14 32 11 85
[email protected]
TNO Automotive Johan Verlaak
Volume 5
Iddo Riemersma Tel. 32 - 14 33 58 62
Tel. 31 - 15 269 67 45
Fax 32 - 14 32 11 85
Fax 31 - 15 269 68 74
[email protected]
[email protected]
Development of proposals for in-use conformity testing of emission control devices
Page 2
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
0
CONTENTS
0
Contents.......................................................................................................................................3
1
Executive summary ....................................................................................................................5
2
Introduction ................................................................................................................................7
3
Approach .....................................................................................................................................8
4
3.1
Translation of system for cars to heavy-duty vehicles ..........................................................8
3.2
Currently used/investigated heavy-duty in-use conformity procedures ................................8
3.3
On-road emission testing for in-use conformity ....................................................................9
3.4
Applicability of in-use conformity for future technology .....................................................9
In-use conformity testing of emission control devices ..........................................................10 4.1
Light Duty in-use conformity ..............................................................................................10
4.1.1
Directive CAP2000 on light-duty in-use conformity in the USA ..........................................10
4.1.2
Directive 98/69/EC on light-duty in-use conformity in Europe .............................................10
4.1.3
Test programmes of the Air Resources Board (ARB) of California on in-use conformity....11
4.2
Currently used/investigated heavy-duty in-use conformity procedures ..............................12
4.2.1
Evaluation of heavy-duty in-use conformity in the USA .......................................................12
4.2.2
Evaluation of heavy-duty in-use conformity in Europe .........................................................21
4.2.3
Comparison of status of heavy-duty in-use conformity in EU and USA ...............................33
4.2.4
Differences between EU and USA heavy-duty vehicle use via WHHD engine cycle ...........33
4.2.5
Applicability of USA in-use conformity approach in EU ......................................................34
4.3
On-road emission testing for in-use conformity ..................................................................38
4.3.1
From the USA study tour .......................................................................................................38
4.3.2
Manufacturers’ vision on on-road measurement systems ......................................................38
4.3.3
Review of known on-road systems.........................................................................................39
4.3.4
Proposed equipment EPA.......................................................................................................52
4.3.5
Review of possible on-road test procedures ...........................................................................54
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 3
EC-DG ENTR
ETD/00/503430
4.3.6
Review of heavy-duty in-use conformity systems .................................................................57
4.3.7
Financial implications ............................................................................................................58
4.3.8
Potential of on-road system use for in-use conformity testing ...............................................59
4.4 5
Emission control technology for heavy-duty vehicles
Applicability of in-use conformity for future technology ...................................................61
Conclusions................................................................................................................................62 5.1
Momentarily no regulations.................................................................................................62
5.2
Differences between USA and EU ......................................................................................62
5.3
NTE approach looks promising ...........................................................................................63
5.4
On-road measuring needs further investigation...................................................................64
6
Recommendations.....................................................................................................................66
7
References..................................................................................................................................67
8
Acronyms and abbreviations ...................................................................................................71
9
Annexes......................................................................................................................................77 9.1
Annex 1: Results questionnaire on in-use compliance ........................................................78
9.2
Annex 2: Delivery goals of CRADA...................................................................................80
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 4
EC-DG ENTR
1
Emission control technology for heavy-duty vehicles
ETD/00/503430
EXECUTIVE SUMMARY At present there are no heavy duty vehicle in-use compliance (IUC) regulations in either the US or the
EU. However, the not to exceed (NTE) limits incorporated into the US engine certification procedure could open the way for future IUC testing executed on a dynamometer or on-road. For on-road tests the NTE limits are well suited, as this means that the vehicle can be operated normally in traffic (instead of following a prescribed speed cycle). Dynamometer tests offer compatibility with certification legislation, and good accuracy as well as reproducibility. On-road testing yields better cycle bypassing resistance at equal to possibly better costeffectiveness (to be confirmed). OBD is not an adequate replacement for IUC testing. It is currently unclear as to whether on board monitoring (OBM) could replace IUC testing and it will need further evaluation as this technique develops. In the US the vehicle has to comply with the NTE limits within a defined area underneath the torquerpm curve of the engine. However, remaining for the required 30 seconds within the NTE zone can be difficult and the question arises as to whether real life emissions are sufficiently well reflected in the NTE test results. Nevertheless, the NTE approach maximises the certainty that in-use emissions are under control. More field experience with the NTE approach will show how well the concept works and where improvements and further research might be necessary. The definition of the NTE zone and the carved out areas, the engine classification and the test methodology could reflect conditions in the EU. However, in the interest of harmonisation, the methodology should be the same in as many aspects as feasible. In the long term it might be advisable to include NTE testing on the world harmonised heavy duty (WHHD) engine test cycle. This would provide a link between the certification cycle and in-use conformity testing as NTE testing would be common to both. If future IUC regulations require on-board measurements specific test procedures will need to be developed. They should cover aspects such as test conditions, cycles, emission limits and should provide a link to the certification and durability regulations. On-road (or on-board) measurement systems to verify IUC could use the NTE concept during real world driving conditions. However, at present, no on-road systems are available that match the demanding specifiVolume 5
Development of proposals for in-use conformity testing of emission control devices
Page 5
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
cations set out by the US EPA. As the NTE limits are only 1.5 times the certification limits, on-road measurement systems need to offer the same accuracy as laboratory grade systems in a suitcase package. Also, the demand for brake specific emission figures necessitates accurate information on the torque. It may take several years before an adequate system can be developed. Given the amount of R&D still necessary a phase-in of on-road measurement systems may occur. It is likely that they will be firstly used to determine real life emission data, and for the evaluation of different technologies under real life driving conditions etc. Subsequently these systems will be developed sufficiently to enable their use as an IUC screening tool. Finally, on-road systems may serve as enforcement tools, possibly without the need for additional dynamometer measurements. Until then, the on-road systems may serve as screening tools i.e. using dynamometer measurements as a follow-up test should the vehicle fail to pass the on-road test (see above). As an alternative dynamometer measurements can be used for IUC tests, meanwhile evaluating the on-road measurement systems (in pilot projects) until these are ready to serve as enforcement tools.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 6
EC-DG ENTR
2
Emission control technology for heavy-duty vehicles
ETD/00/503430
INTRODUCTION In this volume in-use emission conformity (IUC) of heavy-duty (HD) vehicles is investigated. IUC may
be seen as a third step in the assessment of compliance with legislation where type approval is the first step and conformity of production the second. For light-duty vehicles IUC is already regulated. For heavy-duty vehicles this is not the case yet. We will investigate particular aspects that address to HD IUC: •
The measures being adopted in the United States as a consequence of the ‘Consent Decree’ to evaluate their potential for in-use conformity checking in Europe.
•
The potential of using on-road emission measuring equipment within the scope of a European in-use conformity testing scheme as a means of ensuring that vehicle emissions remain within conformity over a period of durability consistent with the requirements developed in Volume 4.
The work on IUC was split into tasks, the most important of which are listed below. Currently used/investigated HD IUC procedures First the IUC situation in the US is evaluated, particularly the NTE limit approach and the work carried out under the Consent Decrees. Then the existing IUC programmes in the EU are discussed. Finally the applicability of the US IUC approach in the EC is evaluated. On-road emissions testing for in-use conformity The potential of using on-road emission testing for IUC is evaluated. Because this is a completely different approach compared to existing procedures the evaluation of the topic will be rather extensive, involving the following aspects: •
Manufacturers’ view of on-board measurement systems
•
Review of known on-board measurement systems
•
Proposed equipment by the EPA
•
Review of possible on-road test procedures
•
Financial implications
•
Potential of on-board systems for IUC testing
Finally, overall conclusions and recommendations are presented.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 7
EC-DG ENTR
3
Emission control technology for heavy-duty vehicles
ETD/00/503430
APPROACH At this moment there are no European requirements for manufacturers of HD vehicles and engines to
ensure emission durability over a certain period and over all appropriate operating conditions. In other vehicle classes and in other parts of the world however, some basic ideas about in-use conformity testing can be found. In order to be able to evaluate these ideas properly, each one is identified together with a discussion on advantages and disadvantages of several aspects (accuracy, costs, practicability, etc.).
3.1 TRANSLATION OF SYSTEM FOR CARS TO HEAVY-DUTY VEHICLES For light-duty (LD) and small commercial vehicles, in-use conformity (IUC) is regulated through Directive 98/69/EC. This system deals with possibilities for the manufacturer to prove the in-use conformity, with test procedures, with the selection of test vehicles, as well as with the measures to be taken if the requirements for the in-use conformity are not fulfilled. The mechanism for enforcement of in-use conformity testing to heavy-duty vehicles should follow that laid down for light-duty and small commercial vehicles and referenced in the type-approval framework Directive (Directive 70/156/EEC, as amended).
3.2 CURRENTLY
USED/INVESTIGATED
HEAVY-DUTY
IN-USE
CONFORMITY
PROCEDURES Up to now IUC testing has been based on the homologation test procedure, i.e. the vehicles tested must comply with the homologation vehicle using the same test cycle. Recent developments in the United States focus on the real world emissions, i.e. are the emissions during actual vehicle use such as can be expected from homologation testing? In the United States, new measures are being adopted as a consequence of the ‘Consent Decrees’. These decrees between the EPA, the Department of Justice (DOJ) and seven engine manufacturers require meeting the US2004 standard on NOx and NMHC emissions 15 months ahead of time. The new measures will be evaluated. At this moment several European countries, like Germany and The Netherlands, have inuse conformity programmes running as well. Comparing these programmes with the US situation will give additional information. In this case, an analysis of differences between the US and EU situation will be made with the help of experience gained by TNO Automotive in the course of a project for the development of a World Harmonised Heavy Duty engine test Cycle (WHDC). It will provide good insight into the potential of applying this in the EU. Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 8
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
3.3 ON-ROAD EMISSION TESTING FOR IN-USE CONFORMITY One of the possibilities with high potential for in-use conformity testing is on-road emission measurements. In this part several aspects of on-road emission testing will be discussed. Opinions on in-use conformity The information from the US visit will be analysed and the vision of the manufacturers on in-use compliance will be presented. Existing systems and proposed equipment Minimum specifications will be set for the measuring equipment. Available information about existing products will be checked for the products’ applicability. Regarding existing and future engine and emission after-treatment technologies, test procedures will also be checked for their reproducibility and representativity. Practical / organisational issues Besides the technical feasibility, the practical feasibility will be assessed. This includes such topics as time necessary for testing, necessary test infrastructure apart from the measuring equipment (tests on public road, private proving ground, chassis-dynamometer, ambient conditions during testing, etc.). Financial implications Available cost estimations will be given for in-use conformity testing. Potential of on-road system for in-use compliance Based on the information gathered, the potential for using on-road emission measurement for in-use compliance will be evaluated.
3.4 APPLICABILITY OF IN-USE CONFORMITY FOR FUTURE TECHNOLOGY As this project is focussed on the emission control technology, the future after-treatment devices and their control strategies should receive special attention as well. Most of these devices are very sensitive to transient behaviour of the engine, and in some cases (e.g. a particulate filter) a shift in time is introduced between the moment emissions are produced in the engine and when they are released from the exhaust. The influence these effects have on a test result cannot be neglected, and certainly need appropriate attention.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 9
EC-DG ENTR
4
Emission control technology for heavy-duty vehicles
ETD/00/503430
IN-USE CONFORMITY TESTING OF EMISSION CONTROL DEVICES
4.1 LIGHT DUTY IN-USE CONFORMITY For light-duty, in-use conformity systems are being applied both in the USA and Europe. In the following an overview of some systems is given.
4.1.1 Directive CAP2000 on light-duty in-use conformity in the USA The Environmental Protection Agency (EPA) in the United States of America has promulgated, via the Federal Rule 40CFR86, regulations for engines and vehicles that limit exhaust emissions. To arrive at this legislation very extensive research was carried out. All stakeholders were involved in the decision process that resulted in the final rule. In this legislation in-use conformity (IUC) is only required for light-duty (LD) gasoline vehicles. For this the Compliance Assurance Program 2000 (CAP2000) is applicable. This part of the federal rule 40CFR86 prescribes how light-duty vehicles have to be checked for their in-use conformity. There are no requirements for heavy-duty engines/vehicles as is mentioned in § II.C p. 59910 of the mentioned final rule [1].
4.1.2 Directive 98/69/EC on light-duty in-use conformity in Europe In-use conformity has just been introduced in the EU legislation for passenger cars (stage Euro 3). Some details are still under discussion, and experience with the legislative approach has still to be gathered. Directive 98/69/EC requires manufacturers of passenger cars and light commercial vehicles to ensure that the emissions performance of their vehicles is maintained while being used by their customers up to a period of 80,000 km or five years of use, whichever is the sooner. From 1 January 2005 the obligation will increase to 100,000 km or five years. This is checked on the basis of comprehensive information supplied by the manufacturer to enable the authority to audit the information. If the information is not sufficient the authority may request additional information. If still not sufficient, the authority may request a sample of in-use vehicles to be taken for full emissions testing to assess conformity. If conformity is not confirmed, a recall plan may be initiated to require the correction of emission system faults on the vehicles found to be at fault.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 10
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
In a number of countries investigations were conducted into IUC testing for light-duty vehicles [2, 3, 4, 5]. The information that can be gathered from these projects can be useful in defining the IUC testing for heavy-duty vehicles. An in-use conformity programme was carried out by the Umweltbundesamt (UBA) in collaboration with similar programmes in The Netherlands and Sweden for light-duty vehicles. The UBA states that the IUC test shall be executed by an independent party [2]. Since 1991 an in-use conformity programme for light-duty vehicles is running in Sweden. No information was available about this programme when writing the report.
4.1.3 Test programmes of the Air Resources Board (ARB) of California on in-use conformity The ARB is currently conducting two separate ongoing in-use vehicle test programmes [6], the ‘InUse Conformity Programme’ and the ‘In-Use Vehicle Surveillance Programme’. Each of these programmes is similar as to which vehicles are borrowed from the public and tested for exhaust and evaporative emissions at the ARB's Haagen-Smit laboratory in El Monte. However, the data generated from these programmes are utilised for different purposes. The ARB's ‘In-Use Conformity Programme’ is a key strategy to aid California in meeting ambient air quality standards. The goal of this programme is to ensure that manufacturers' vehicles meet emissions standards throughout their useful lives. To accomplish this task, the ARB seeks a limited sample of vehicles from a given engine family and duplicates the manufacturers' vehicle emissions certification tests. The vehicles are procured, restored to the manufacturers' specifications, and tested in accordance with the Code of Federal Regulations. ARB and the manufacturers' representatives are present to oversee all aspects of the test programme. Should a non-conformity situation occur within a given engine family, the ARB will work with the manufacturer to correct the problem on all affected vehicles. The corrective action is usually in the form of a state-wide recall in which the manufacturer will notify all affected vehicle owners and state when and where to seek the recall repair. The cost of the repair and service is free to the vehicle owner. The primary objective of the ARB's ‘In-Use Vehicle Surveillance Programme’ is to determine a fleet "snapshot" of baseline emissions for the mobile source emissions inventory. Secondary objectives include the evaluation of present and future emission control programs. ARB's current programs include: 1. Evaluating the cost effectiveness of the Smog Check Program. 2. Gathering information on deterioration rates of emission control equipment for in-use vehicles. 3. Evaluating evaporative emissions. 4. Gathering chemical by chemical (speciated) exhaust and evaporative profiles for in-use vehicles. Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 11
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
5. Evaluating experimental vehicle test cycles. The above mentioned programs are voluntary and mainly on cars. In 2000 a limited number of cars (4 as opposed to 40 in 1999) were tested. Those vehicles had been driven less than 90,000 miles (145,000 km).
4.2 CURRENTLY
USED/INVESTIGATED
HEAVY-DUTY
IN-USE
CONFORMITY
PROCEDURES
An in-use conformity system for heavy-duty vehicles differs from a light-duty system. For light-duty vehicles, emission data are given in g/km, as required by the vehicle homologation procedure. For heavyduty vehicles, emissions have to be measured in g/kWh, as a heavy-duty engine can be found in many different heavy-duty vehicles each aimed at a particular use. Thus the emissions are expressed as brake specific. Consequently, IUC for light-duty cannot just be copied to heavy-duty but needs adaptation. It is necessary for heavy-duty vehicles to know the engine load, for example by acquiring the engine torque. Measuring the engine torque in a vehicle however is quite challenging. So, investigations on heavy-duty in-use conformity systems are ongoing.
4.2.1 Evaluation of heavy-duty in-use conformity in the USA
Up to now light-duty in-use conformity (IUC) testing is based on the homologation test procedure, i.e. the vehicles tested must comply with the homologation vehicle using the same test cycle. Recent developments in the United States focus on the real world emissions, i.e. the emissions during actual vehicle use. These may differ from homologation testing as it is defined now. In this paragraph we will first discuss the not-to-exceed (NTE) limit approach. This may be a useful tool in defining test procedures for IUC testing.
4.2.1.1
The not-to-exceed (NTE) limit approach
Up until now, test cycles with a well defined torque and engine speed profile were used to control whether the engines manufactured fulfilled the requirements on emissions. This enabled the engine manufacturers to optimise their engines on emissions for these test cycles. For the engine loads that were not defined
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 12
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
in the test, no emission limits were applicable. As a result, real world emissions could be higher than the emissions that were intended to be achieved by introducing the limits. To cope with this problem, the not-toexceed approach was introduced. The NTE procedures apply under engine operating conditions (within the range specified in the NTE control area) that could reasonably be expected to be seen in normal vehicle operation and use [7]. The NTE procedure defines limited and specific engine operating regions (i.e. speed and torque conditions, indicated as the NTE control area, see figures 1 and 2) and ambient operating conditions (i.e. altitude, temperature, and humidity conditions) which are subject to the NTE emission standards. If the temperature is outside the range 13°C and 51°C, or the humidity is outside the range from 7.14 to 10.71 g H2O/kg dry air, than correction factors are allowed (below 1675 m altitude). The engine speeds A, B and C that are mentioned in the figures comply with the European Steady state Cycle (Euro 3). Emission results from this test procedure, integrated over a time window from 30 seconds, must be less than or equal to 1.5 times the Federal Test Procedure (FTP) standards for NOx, NMHC, and PM. The new NTE requirements are phased in starting with the 2007 model year and are consistent with the new FTP standards.
Figure 1 NTE Zone for Heavy-Duty Diesel Engines -- C Speed < 2400 rpm speed C = nlo + 0.75*(nhi-nlo), nlo at 50% Pmax, nhi at 70% Pmax (PM carve-out not applicable in Final Rule)
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 13
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
Figure 2 NTE Zone for Heavy-Duty Diesel Engines -- C Speed > 2400 rpm speed C = nlo + 0.75*(nhi-nlo), nlo at 50% Pmax, nhi at 70% Pmax (PM carve-out not applicable in Final Rule)
This approach must ensure that no excess emissions occur in points of the map that are not covered by the standard type test procedure and/or under circumstances differing from those of the type approval test. It is also meant as a safeguard against so-called ‘defeat devices’ and/or ‘irrational control strategies’. As the current approach in the US is to make use of the ‘control area’ as defined in the current European legislation for heavy-duty engines this would link the European and US legislation. Comments on NTE from parties involved When the NTE approach was proposed by the EPA, engine manufacturers argued that the NTE limits were not appropriate due to several aspects. EPA has adapted the proposal based on the comments that were made [8]. The aspects that were stated to oppose the NTE-approach were the following: •
The impact of new technologies have not been considered in this regulation
•
due to the inherent variability of emission maps of HDDE’s, the limit that results from 1.5 x FTP limit (1.25 for 2002 under the Consent Decrees, cfr. 4.2.1.3) cannot be achieved for all operational conditions (e.g. low engine speed, high load)
•
it was not demonstrated that the limits are technologically feasible
•
high brake specific emissions at low engine speeds are not taken into account
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 14
EC-DG ENTR
•
Emission control technology for heavy-duty vehicles
ETD/00/503430
to avoid measurement of emissions in a critical operating condition for the engine, one should consider the overall emissions
•
condensation and corrosion problems in the intake system (durability)
•
problems in achieving PM limits using after-treatment and formation of sulphate A number of aspects that were raised by the manufacturers against the NTE approach can already be
met today (January 2002). A number of engine manufacturers have demonstrated that they can comply with 90% of the NTE area applying cooled EGR, advanced turbo-charging systems and high pressure electronically controlled fuel injection. Some aspects cannot be met today, but can be overcome within the timeframe: •
emissions under high load situations
•
high temperatures of blower
•
controllability of the emission control systems (actuators)
•
availability of appropriate sensors
•
condensation and corrosion issues
•
emission of particulates (PM) when using after-treatment and forming of sulphate
To overcome these problems, the legislative requirements will become compulsory phased in over time and deficiency provisions are included in the NTE standard. EPA stated that all the NTE standards will have to be met by model year 2007. In 2010 all vehicles have to comply without exception.
4.2.1.2
EPA’s research on in-use conformity
At the moment the EPA is carrying out a program on in-use conformity for heavy-duty vehicles [9]. One of the purposes of this program is to demonstrate the feasibility of on-road testing for heavy-duty trucks and buses. Also it is investigated whether the NTE approach can be used as a conformity tool. The measurements that were executed in this project focussed on NOx emissions. More information is given below. On-road measuring On-road measuring is carried out with mobile measuring equipment. The purpose of this is to evaluate whether it will be possible in the near future to measure exhaust emissions in vehicles on-road, with a portable sampling system. This could make in-use conformity measurements more cost-effective. At this moment, it looks to be a good approach, although improvements in the measurement equipment have to be made.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 15
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
Not-to-exceed limits The not-to-exceed (NTE) limit approach is seen as the most appropriate way of testing IUC. This approach ensures the control of emissions over a broad range of real-world conditions, reduces the need to scrutinise emissions control strategies, and enables cost-effective in-use conformity monitoring. The not-to-exceed limits are designed to apply under any engine operating conditions, that could reasonably be expected to occur during normal vehicle use, under a wide range of ambient conditions. This type of limit is therefore ideally suited for in-use conformity testing, although currently there is no requirement to measure NTE emissions in-use. Measurements Based on the experience gathered during the tests, some initial conclusions could be made. Most of the current-technology engines that were tested, complied with NOx NTE limits. Some required follow-up. Testing according to the on-road NTE procedure apparently captured worst-case NOx operation. It was also concluded that obscure operation was not an issue. Further it was experienced that the used test method complemented Euro/FTP to ensure NOx conformity over a broad range of operation. Further investigations Further work is needed on improving the procurement of vehicles and the testing efficiency. For doing the testing, it should be possible to use commercially available equipment with enhanced capabilities of user friendliness, minimised testing time and cost, PM measurement, automated data analysis and improved overall accuracy. Also investigations have to be conducted into improving the understanding of uncertainty in results. Finally quality control procedures should be standardised.
4.2.1.3
Consent Decree by Department of Justice (DOJ) and Environmental Protection Agency (EPA)
In the United States new measures are being adopted as a consequence of the ‘Consent Decree’, announced on October 22, 1998 by DOJ and EPA against 7 heavy-duty diesel engine manufacturers (Caterpillar, Cummins Engine Company, Detroit Diesel Corporation, Volvo Truck Corporation, Mack Trucks, Renault Véhicules Industriels and Navistar International Transportation Corporation) [10].
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 16
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
A court settlement was reached between the EPA, Department of Justice, California ARB and the engine manufacturers over the issue of high NOx emissions from heavy-duty diesel engines during certain driving modes. Since the early 1990’s, the manufacturers used engine control software that caused engines to switch to a more fuel-efficient (but higher NOx) driving mode during steady highway cruising. The EPA considered this engine control strategy an illegal “emission defeat device”. Provisions of the Consent Decree included the following: •
Civil penalties for engine manufacturers and requirements to allocate funds for pollution research
•
Upgrading existing engines to lower NOx emissions
•
Supplemental Emission Test (steady-state) with a limit equal to the FTP standard and NTE limits of 1.25* × FTP (with the exception of Navistar) (this test cycle is very similar to the European test cycle ESC)
•
Meeting the 2004 emission standards by October 2002, 15 months ahead of time (*) For vehicles >2007, the NTE limit 1.5xFTP applies
The Consent Decree also forces each manufacturer to spend $ 2,000,000,- (approx. 2,000,000,- EUR, whereof no more than 20% in phase I and II) on an in-use testing programme. The actions are split into several phases: Table 1
Phasing of actions due to ‘consent decree’ Phase
Action
Time Limit
The manufacturer shall conduct engineering studies to determine the correlation, accuracy, I
precision, and repeatability of existing mobile monitoring technologies. Further engineering studies have to be included to determine the highest degree of accuracy and precision of re-
1 Sep 1999
ported engine output torque achievable. The manufacturer shall develop in-use testing procedures (variety of on-road missions, variety II
of seasonal conditions, variety of engine life time) to be used in connection with phases III and IV of the in-use testing program. The test procedures shall also include identification of driv-
1 Nov 1999
ing routes. III IV
The manufacturer shall conduct emissions testing on a variety of its in-service diesel engines to characterise real world emissions from such diesel engines.
1 Oct 2000
The manufacturer shall conduct on-road conformity monitoring on its heavy-duty diesel engines. Vehicle selection procedures and data reporting requirements are set out.
West Virginia University (WVU) was contracted by the seven manufacturers to carry out the studies mentioned in phase I and II in the table above. An overview of the results is given in the following paragraphs.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 17
EC-DG ENTR
4.2.1.3.1
Emission control technology for heavy-duty vehicles
ETD/00/503430
Phase I
The object of this phase [11] is to evaluate the current available technologies to measure on-road emissions from heavy-duty diesel engines. A survey of the available systems (15) is given. Some of the systems that are mentioned later in this report (see 6.2.1) were not taken into account in this survey. As a result of this study, the specifications of an On-Road Emission Measurement System (OREMS) were given. Table 2
Specifications of an on-road emission measurement system (OREMS) Measuring
Mass emissions of NOx, UHC (unburned hydrocarbons), CO, CO2 Exhaust flow rate Engine speed and torque
Data acquisition
Account for time lags and response functions
Design
Portable Accommodate a broad range of exhaust designs Function accurately over a wide range of ambient conditions and varying altitudes Additional power source (e.g. portable generator)
None of the available systems fully complied with the desired specifications. WVU developed the Mobile Emission Measurement System (MEMS). This system cannot measure CO and UHC. Engine torque is derived from the motor management ECU. The system was compared to the EPA measuring system ROVER (Remote On-board Vehicle Emissions Recorder). This comparison proved to be very difficult as the Rover unit under test was not able to deliver brake specific emissions nor to produce NTE test results. WVU added these features and thus the brake specific emissions are not unique Rover results. However, the time based mass emissions of both systems were comparable. 4.2.1.3.2
Phase II
West Virginia University also carried out a study to develop testing procedures [12]. Several routes (real world situations) were evaluated with a heavy-duty truck. It was concluded that during only a small part of the route (from 20 to 50%) the load conditions of the engine were for 30 seconds within the NTE area. A survey of vehicles from different manufacturers that can be tested on-road is given (have to be equipped with engine ECU with appropriate protocol). The accuracy of the engine torque measurement is limited. For the loads within the NTE area, the error is in the order of 10% for a 30 second window. A larger window can reduce the error.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 18
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
Conclusions WVU report on phase II: • Remaining in NTE zone may fail • Measurement results from laboratory equipment and MEMS show good agreement (maximum difference 5%), measurement results from MEMS on-road and on test bench differ less than 5% for NOx. • Quality control/quality assurance plan is proposed 4.2.1.3.3
Phase III
During the US visit (October 2001), it became clear that this phase had not yet been started. The manufacturers stated that they were waiting for the EPA to provide information on the measurement system to be used. 4.2.1.3.4
Phase IV
No actions have taken place so far (January 2002).
4.2.1.4
Manufactures’ view of in-use conformity
Caterpillar stated that they develop engines to comply with NTE. The introduction of NTE limits has required rewriting development procedures and software, and extensive testing. According to Caterpillar the NTE will result in a 2-3% increase in fuel consumption (yet they claim to be able to meet the 2002/04 limits with no fuel consumption penalty). In-use conformity is not yet a part of the final rule. DDC believe that the taxpayer should pay for the cost of in-use conformity, not the engine manufacturers. DDC would not commit itself when asked whether the NTE limits had changed its philosophy towards engine development. The company stated that it is its philosophy to meet all emission standards competitively.
4.2.1.5
Averaging, banking and trading
One of the difficulties of in-use conformity checking is that the Federal regulations apply to engine families. However, the manufacturers can use averaging, banking and trading provisions for their total production volume. Thus if a tested engine/vehicle type exceeds the emission limits specified for that engine, the manufacturer may compensate for this with another engine/vehicle type which produces emissions below the limits. As long as the emissions produced by all the manufacturers’ engines (the total production volume) are below a well defined level, the manufacturer is allowed to produce and sell these engines. As a consequence the evaluation of the IUC testing can become very complicated, due to the fact that the total production volume of an engine and vehicle manufacturer has to be considered. Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 19
EC-DG ENTR
4.2.1.6
Emission control technology for heavy-duty vehicles
ETD/00/503430
Conclusion about the heavy-duty in-use conformity situation in the USA
To prepare for the actual legislation, extensive discussions have been held between the EPA, engine manufacturers and other stakeholders. As a result of these discussions, the original regulation proposed by the EPA was adapted. By doing so, it is possible to force the engine and vehicle manufacturers to take actions on emission control. However, in-use conformity for heavy-duty vehicles is not yet in legislation. Consent decree Manufacturers are awaiting the measurement system to carry out phase III of the consent decree. At this moment it is not clear what equipment has to be used. Therefore, the original time schedule from within the consent decree cannot be maintained. Non-conformity penalties (NCP) EPA introduced non-conformity penalties (NCP) as a possibility for manufacturers to cope with the consent decree emission limits. According to this a manufacturer can produce a new vehicle or engine without complying with the emission limits, but by paying a penalty. A public hearing on this will be held on the 15th February 2002. At the time of finalising the report no information on this could be found or was given by the EPA upon our request. In-use testing programs The EPA is running a heavy-duty in-use conformity program. Experience shows that it will be possible to do cost effective IUC testing by using portable emission measuring system, but further investigations and testing need to be done. When writing this report, we were not aware of any actions by the vehicle or engine manufacturers. Miscellaneous Heavy-duty in-use conformity will possibly not be compulsory before 2011, as was stated during the US visit by an EPA spokesman. It was also stated that EPA wants to be in compliance with Europe. Averaging, banking and trading will complicate the rule making for in-use conformity.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 20
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
4.2.2 Evaluation of heavy-duty in-use conformity in Europe 4.2.2.1
In-Use conformity programmes for heavy-duty
For heavy-duty vehicles no legislation exists yet, but again some or even extensive experience exists on a national level in The Netherlands, that runs a relatively large programme (TNO Automotive, Delft), and Germany, that runs a smaller programme (Umweltbundesamt). The testing is executed on a vehicle basis (whereas the type approval test is on the engine) and on the basis of steady state testing (as is the current European Stationary Cycle (ESC) test).
4.2.2.2
The Netherlands
In 1994 TNO Automotive started its HD vehicle In-Use Compliance programme at the request of the Dutch Ministry of Spatial planning, Housing and the Environment. Until now, it has been possible to simulate the homologation test procedure with acceptable accuracy on the vehicle (using a chassis dynamometer), rather than testing the engine on an engine dynamometer, as prescribed by the type approval procedure. Advantages of this methodology developed by TNO, are the cost-effectiveness and the possibility of using the homologation test results as a reference. With the Euro 3 legislation coming into effect, new test cycles with transient elements are added to the homologation procedure, where previously it was only a stationary test cycle. As a result of this, the test methodology used so far no longer applies to the whole procedure. Because it is anticipated that in the near future emissions from HD vehicles will be increasingly related to transient engine behaviour, a transient test cycle is also important for IUC purposes. As the current testing methodology used by TNO is not applicable for a transient engine test cycle, alternative options have been identified and assessed that may serve the aims and demands of the Dutch IUC programme. Also the possible role that OBD can play for IUC purposes was evaluated. This has resulted in the following conclusions and recommendations: •
With current and future engine technology, emissions from transient engine behaviour are becoming increasingly important for the total of emissions produced. This calls for the need of transient elements as part of an IUC test procedure, and, for the purpose of reliable emission factors, the use of test cycles that correspond to real-life driving conditions.
•
The ETC test, being an engine based test cycle, is not suitable for IUC purposes as it is impossible to perform a transient engine test on a HD vehicle.
•
In the future, the aims of the Dutch IUC programme are best fulfilled by a stationary engine test cycle on the chassis dynamometer, extended with transient elements such as an ELR for the evaluation
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 21
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
of compliance. For the production of emission factors, a combination with transient vehicle-based test cycles is required. •
OBD will not fulfil all of the primary aims set for the Dutch IUC programme, so it cannot be seen as a substitute. However, for the evaluation of compliance it can deliver a useful contribution. The role of OBD may change in the future, if it is extended with an emission measuring functionality (OBM).
4.2.2.3
Germany
In 1998 an HD IUC test programme was set up by the Umweltbundesamt (UBA) [4]. Measurements were carried out on a chassis dynamometer. All the tested engines complied with the regulations. It was concluded that the selection and the preparation of the vehicles to be tested are important. Also the problem of an appropriate IUC testing procedure has to be solved. A number of alternative test methods on in-use conformity testing, were evaluated. The main difficulty is the fact that the test has to be transient. The methods mentioned are: •
Removing the engine from the vehicle to do a test on an engine dynamometer is the best way to compare with the original approval test. In practice this is rather difficult, even more so when emission reduction technologies are used (see also section 4.2.2.7).
•
Putting the vehicle on a chassis dynamometer is practically very simple. However, chassis dynamometers with the required power rating are very rare. Another difficulty is that transient testing will require dynamic corrections for drive train and auxiliary losses (see also section 4.2.2.7).
•
One could also perform steady state testing and correct the emission values using a dynamic map to calculate emission factors. The question remains whether the aging of the engine or the after treatment system can be simulated in this map.
•
A promising solution seems to be to disconnect the drive shaft and couple it to a dynamometer. In this situation drive train losses have still to be taken into account. Otherwise the effort is comparably small and correlation to type approval is good.
•
Another possibility is the introduction of on-board measuring (OBM) systems for emissions. When this is the case, IUC can be seen as checking the OBM system on a temporary basis.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 22
EC-DG ENTR
4.2.2.4
Emission control technology for heavy-duty vehicles
ETD/00/503430
Sweden
For heavy-duty vehicles there is no programme known. The Swedish Environmental Protection Agency was contacted, but no answer was received at the time of writing this report.
4.2.2.5
Manufacturers’ view
From the questionnaire sent out (see volume 1 of this final report) the following results (see annex 1 for tables) can be given on the subject of in-use conformity testing (‘manufacturers’ refers to all the respondents to the questionnaire, that means vehicle, engine and component manufacturers): •
The manufacturers are not convinced of the fact that in-use conformity testing will reduce the real world emissions.
•
They state that OBD and OBM will make IUC superfluous.
•
The in-use conformity testing shall be executed by the vehicle manufacturer or a type approval authority. The cost issue i.e. what are the costs and who will carry them needs addressing.
•
On whether the test shall be done on the road, on a chassis test bench or on an engine test bench, most of the manufacturers disagree.
•
An IUC test should occur only once in the lifetime of a vehicle, at a random (mile)age (minimum 24 months or 200,000 km).
•
The parameters that have to be tested are not yet defined.
These results give a rough idea on the manufacturers’ opinion on IUC. It will surely be useful to discuss this matter in further detail with each of the manufacturers when proposals are laid down. In a discussion the following was stated by Mr. Signer (European Engine Alliance) on the matter of in-use conformity. There is a major problem of ‘chip tuning’ with in-use vehicles. Apparently Volvo (one of the first manufacturers to introduce electronics on their engines, and hence having the most experience) has found that 10% of their trucks have the electronics manipulated to increase power output, with no regard for the emissions. This can be done by replacing chips, reprogramming the on-board computer, inserting a device that alters the signals, etc. Thus the in-use emissions/fuel consumption performance is likely to be very different to the theoretical performance. This raises interesting issues regarding who takes the responsibility for in-use emissions if the engine has been tampered with. Presumably the engines are returned to the manufacturer’s settings if the vehicle is returned to the manufacturer under guarantee or for I/M tests.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 23
EC-DG ENTR
4.2.2.6
Emission control technology for heavy-duty vehicles
ETD/00/503430
Conclusions about the heavy-duty in-use conformity situation in Europe
Several national programmes on heavy-duty in-use conformity are running in the EU. They are based on steady state testing executed on a chassis dynamometer. As for Euro 3 ETC and ELR transient testing has entered the regulation. It is investigated how the IUC steady state test can be replaced/upgraded with transient elements.
4.2.2.7
Assessment of possible test procedures for IUC testing
This section as well as section 4.2.2.8 is based to a very large extent on a paper by TNO Automotive presented at the Intertech Conference held on 15-17 October 2000 in Berlin [5]. As mentioned in section 4.2.2.2 TNO Automotive conducts a large LD and HD IUC programme for the Dutch government. Considerable experience has been gained with HD chassis dynamometer steady state tests. Vito has complemented the text from its experience in on-road testing. In October 2000 the Euro 3 legislation came into effect for new engines entering the market. The ECE R49 13-mode stationary test was replaced by a European Stationary Cycle (ESC). This new stationary cycle is rather similar to the ‘old’ 13-mode test, except for the location (and weight) of the mode points. The methodology for measuring a stationary engine cycle on a chassis dynamometer (applied by TNO in the Dutch IUC programme) can therefore still be applied. Furthermore, a European Load Response test (ELR) was added to the homologation procedure to restrict the smoke emission during transient engine loads. If the chassis dynamometer is designed to keep the speed of the rolls constant at varying loads, this test can also be simulated on the vehicle. Alternatively, the test could even be done on the road, as the measurement equipment can easily be made portable (only a smoke opacity analyser is needed for the ELR). For gas engines and diesel engines equipped with exhaust after-treatment systems the Euro 3 legislation prescribes the European Transient Cycle (ETC). This is a dynamic engine cycle, so engine torque and speed are defined as a function of time. Simulating this engine cycle on a (transient) chassis dynamometer poses the following problems: •
As the ETC is an engine cycle, also gear shifting points are defined in it. The gear ratios of a vehicle will not exactly match those of the test cycle, so the engine speed will deviate from the prescribed one after a shift has taken place. The correction needed to compensate this speed difference will introduce an error in the test result. Alternatively, the ETC can be tested in one gear. As a consequence, the whole driveline then has to be accelerated or decelerated quickly in a short period. Because of the rotational inertia within the driveline, this may cause the wheels to slip.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 24
EC-DG ENTR
•
Emission control technology for heavy-duty vehicles
ETD/00/503430
As a result of inertia effects, vibration, torsion and hysteresis within the driveline it is extremely difficult to predict the engine torque accurately from the torque on the rolls. Measuring the engine torque directly from the engine crankshaft may be an alternative, but is quite complex at this moment. Note that the moments of inertia of components within the driveline need to be known, either from the manufacturer or via empirical evaluation (e.g. coast-down tests).
•
The same dynamic effects also make it very hard to control the engine torque by adjusting the torque on the rolls. The torque needed to compensate for these effects may even be of an order of magnitude higher than the actual engine torque prescribed by the test cycle. These problems will make a simulated ETC on a chassis dynamometer very difficult (with correspond-
ingly high inaccuracy), if not impossible. The only alternative is to dismount the engine from the vehicle, and test it on a transient engine dynamometer. This is an undesirable method for IUC testing from the point of cost effectiveness and practicability. The conclusion therefore can only be that the ETC, or any other transient engine test cycle for that matter, is not an appropriate cycle to be used for IUC testing. As a consequence, the emission values determined for the ETC during homologation cannot act as a reference for IUC testing. The market share of HD vehicles with gas engines or diesel engines with exhaust after-treatment systems under Euro 3 legislation is expected to be quite low. Therefore, the problem of the ETC being unfit for IUC testing purposes will be only a real issue when Euro 4 legislation comes into effect (2005). With current and future engine technology, emissions from transient engine behaviour are becoming increasingly important as to the total of emissions produced. So, the information that an in-use vehicle complies during a transient test cycle is more valuable than compliance with only a stationary test cycle. For the determination of emission factors a transient vehicle test cycle is more preferable, especially if this represents the real-life operating conditions for the vehicle type observed. In the following section a number of possible options for IUC testing will be identified.
4.2.2.8
Options for practical IUC testing of HD vehicles
The benefit of a transient IUC test procedure has been indicated in the previous paragraph. However, the ETC, or any other engine based transient test cycle, can obviously not be used for this purpose, so an alternative procedure has to be thought of. Any alternative IUC test procedure also brings the need for a reference, in order to evaluate measured emissions during the vehicle life. If the IUC test were to be legislated, the homologation test procedure could be expanded with this test in order to provide the reference. In this paragraph the assumption is made that, where necessary and feasible, the corresponding legislative arrangements for expanding the homologation test procedures have been made for each option.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 25
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
The options that will be discussed can be classified in two groups: •
Laboratory tests: a vehicle is driven to a test laboratory, where instrumentation and testing take place in a conditioned environment
•
On-road tests: a vehicle is instrumented in a laboratory, and tested on the road or a circuit
Generally, a test in a laboratory will have a better reproducibility and accuracy because of the conditioned environment and the higher standard of the measurement equipment. On-road tests, on the other hand, have the advantage of lower investment costs. Now the options will be identified and briefly described. This list may not be complete, but the authors have only considered the most promising options. Simple test methods that for example only observe the emitted smoke have therefore not been taken into account. Other methods can be seen as a variant of the options mentioned here. Laboratory tests 1. Transient engine test cycle The engine is dismounted from the vehicle, and tested over a speed/torque-time test cycle on an engine dynamometer. This option is already identified as being costly and impracticable for setting up an IUC programme of substantial size. 2. Stationary vehicle engine test cycle This is the method used for the current IUC programme in the Netherlands, in which a stationary engine test cycle (e.g. the ESC) is tested on a vehicle on a chassis dynamometer. Some transient elements need to be added (where the dynamometer is suitable for this), in order to evaluate basic transient performance. An ELR test is a good example of such a test. 3. Transient vehicle test cycle This is a speed-time test cycle that is driven on a HD transient chassis dynamometer. Such a facility is not widely available; at this moment there are only three in Europe. On-road tests 4. Transient vehicle test cycle A speed-time test cycle is driven on the road. If other traffic is driving on this road as well, it will probably conflict with the test cycle. So an empty road (e.g. test track or circuit) is required for this test. The emission measurement equipment is carried on board the vehicle. 5. Vehicle cycle with target speeds This is also a speed-time test cycle, but with periods in which a constant target speed is defined, instead of an exact prescribed speed pattern. The acceleration is left to the ability of the vehicle. Such a cycle is Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 26
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
easier to carry out on a road with other traffic present, but a test track is desirable. The emission measurement equipment is carried on board the vehicle. 6. Real-life arbitrary cycle with Not-To-Exceed limits The idea of this method is to have a set of limits in the legislation that may never be exceeded during any driving condition. Therefore, the test driver can drive normally in traffic, without having the restrictions of a test cycle. 7. On-Board Diagnostics system In principle, OBD is not a method for IUC testing, but it can be used as an instrument to prevent engines from malfunctioning. Assuming the emissions from a properly functioning engine correspond with those of the homologation engine, it can be seen as a method for obtaining compliance. The reason to include OBD in this list is therefore to illustrate the contribution that it may deliver towards IUC testing. If in the future OBD would also be capable of measuring emissions (On-Board Monitoring; OBM), its role towards IUC testing would have to be re-assessed. For the last two options there is no need to extend the homologation procedure with an additional test to serve as a reference cycle; in case of option 5 only a set of NTE limits has to be legislated. Legislating a vehicle-based test cycle to serve as a reference cycle (options 3, 4, and 5), will prove to be very difficult. HD engines appear in a variety of different vehicle types and drivelines. A number of vehicle related aspects that have an influence on the test results, are therefore hard or even impossible to prescribe in a test procedure (weight, driving resistances, driveline efficiencies, gear ratios). In order to make a comparison between each of these options, their ability to serve the aims and demands of an IUC programme is assessed. In the table below the results are shown. Indicated per option is the extent to which it serves the aim or demand, going from ‘very well suited’ (+ +) to ‘unsuitable’ (- -).
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 27
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
Table 3 Table with assessment of possible testing
OBD system
limits
Real-life with NTE
Target speed cycle
cycle
Transient vehicle test
namometer
cycle on chassis dy-
On-road Transient vehicle test
namometer
cycle on chassis dy-
Stationary vehicle test
namometer
cycle on engine dy-
Transient engine test
Lab Tests
Legislation compliance
++
+
-
-
-
-
+
Emission factors
+
0
+
0
0
0
--
Maintenance effects
+
++
++
+
+
-
0
Cycle bypassing
+
-
+
+
+
++
-
Cost effectiveness
--
+
0
+
+
+
++
Practicability
--
0
+
0
0
+
++
++
+
+
-
-
--
0
Reproducibility, accuracy, comparative
The following will motivate the assessment for each of the aims and demands separately. Legislation compliance A first condition for evaluating compliance is that the test itself is suitable to be legislated. As it was concluded that HD engines are applied in a variety of vehicle types and drivelines, thereby practically preventing a vehicle-based test cycle from being legislated, none of the options that make use of a speed-time pattern are appropriate to evaluate compliance on. Obviously a transient engine test cycle fulfils the requirements of legislation compliance best. In this respect, a stationary engine test cycle can only be considered as an option if transient elements are added to the test, derived from the transient engine test cycle. Another important aspect is that the test needs to be accurate and reproducible. Because of the conditioned environment and the more accurate emission measurement equipment, the laboratory tests are more appropriate than on-road tests. For the arbitrary cycle with NTE limits and OBD, reproducibility is in fact no issue as the limits set may not be exceeded under any driving condition. Still, the laboratory tests have a higher ranking, as they not only show if an engine complies, but also provide better insight into the ratio of
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 28
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
test results with respect to the homologation results. Moreover, OBD can only check a number of engine parameters, but is no absolute guarantee for compliance of emissions. Emission factors For the determination of representative emission factors (defined as a statistically correct average emission value for a certain group of vehicles) the tests used should be representative for real-life driving conditions. Also, the level of reproducibility and accuracy of these tests should allow the investigation of the influence of different parameters on the emission performance of the vehicle type under study for the purpose of bottom-up emission modelling. Such an emission model can provide representative emission factors, and serves as a tool to calculate the effect of traffic measures, introduction of new emission reduction techniques etc. These purposes can be served best by either a dedicated vehicle test cycle, driven on a HD chassis dynamometer, or a dedicated transient engine test cycle on the engine dynamometer. In both cases the laboratory environment allows the variation of one parameter while keeping the others constant. The influence of this parameter can be determined in this way. It should be noted that the number of influencing parameters for future more complex Euro 4 and 5 vehicles will probably increase. The resulting test matrix to investigate the effect on emissions of each of these parameters will be accordingly large and costly to execute. There is however no alternative. Also, the more dedicated the engine or chassis dynamometer cycle the better the resulting emission factor will be. The large variety in HD vehicles and in HD engine use does not facilitate the elaboration of one dedicated cycle which is valid for all. In practice the cycle will hardly ever be fully dedicated to the vehicle thereby reducing the accuracy of the emission factor to the extent of imperfection of the cycle. As a result the highest mark (++) can not be given in Table 3. The vehicle test cycle can also be driven on the road but will produce less accurate test results due to the unconditioned testing method. This is also valid for the target speed cycle. Emission factors measured on-road from a test cycle with NTE limits give a good impression of real-life emissions, but these are not necessarily representative for real-life vehicle use. Due to the fact that these results are not reproducible, they will not allow for bottom-up emission modelling. However, a database covering all of the NTE test results will probably yield an average result which can be seen as a representative in-use emission factor (assuming that all relevant influences on emissions are equalised by the high number of test results). If only a selection of the database can be used for a detailed investigation of a particular problem, the number of test data decreases (as well as the equalising effect), resulting in less accuracy. Examining the influence of a particular parameter by keeping the others constant is not possible. Note also that for emission factor purposes all test data need to be used and not just the ones from within the NTE area. Otherwise the result is not representa-
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 29
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
tive as the engine operates considerable time out of the NTE area (see sections 4.2.5.2 and 4.2.5.3). So the three on-road ‘cycle’ options got a ‘0’ as mark in Table 3. Finally, it would be an asset to emission factor work if laboratory and on-road test results would complement each other. Assuming that OBD at this moment can only inform about the status of the engine (On-Board Monitoring is not yet technically feasible), it delivers no contribution to the determination of emission factors. This situation may change if in the future emissions can be measured accurately by OBM. As the emissions due to transient vehicle use are becoming more important, the relevance of a stationary engine for the purpose of emission factors is only limited. Maintenance effects To research the effects of maintenance on emissions, laboratory tests have the advantage of being accurate and reproducible. Therefore, the resulting effects on emissions can be attributed to re-adjustments made to the engine. The lower reproducibility of on-road tests is preventing the correlation of the resulting effects towards the state of maintenance (especially for the target speed option). OBD does not provide any direct relationship between maintenance and emissions, but can give valuable information about the defect history of the vehicle. Cycle bypassing The detection of cycle bypassing strategies conflicts with the demand for reproducibility, as this makes it easier for the engine management system to recognise a test cycle. The fact that on-road testing methods are never exactly identically, prevents the use of cycle bypassing strategies. OBD is not able to detect cycle bypassing, assuming that it is only checking the engine behaviour. Besides, the manufacturer develops both the OBD system as well as a potential bypassing strategy. The role of OBD in detecting cycle bypassing may change if in the future emissions can be checked by an OBM system. Cost-effectiveness The laboratory tests will demand for high investment costs in facilities and equipment, especially for transient test facilities. Once this investment is made, the costs per vehicle test are dependent on the number of tests i.e. the more tests the lower the costs per single test. This does not account for the transient engine test, which asks for the engine to be dismounted from the vehicle, instrumented, and installed on the engine dynamometer. For the on-road options only portable emission measurement equipment is required, which involves relatively low investment costs as long as the life time and service costs are comparable to laboratory equipVolume 5
Development of proposals for in-use conformity testing of emission control devices
Page 30
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
ment. This remains to be proven. If needed a test track to perform the on-road test cycle can be hired. The costs per individual vehicle test will be largely dependent on the time elapsed to instrument the vehicle. An analysis executed by the EPA [9] on a dedicated on-road system gives an average instrumentation time of two hours followed by a maximum of two hours of measurement. This presents a cost effective approach. The ultimate goal -as defined in EPA’s call for a Cooperative Research And Development Agreement (CRADA) (see section 4.3.4)- is to reduce the instrumentation time to half an hour. Under the NTE option the measurement time can be extended as the vehicle will be monitored during its normal exploitation. When this goal is achieved on-road measurements will become very cost effective. Before and after the testing the vehicle stays in exploitation as the measurement system is transported to the vehicle. This is opposite to laboratory measurements. However, there is also a cost associated with transporting the measurement system. Furthermore, cost-effectiveness will be influenced by the level of use that can be made from the test results of the different options. If for instance -next to IUC- the result from one option can also be used for e.g. emission factor work this yields a considerable positive effect on cost-effectiveness. Overall, in the future on-road will most likely be more cost effective, if the CRADA goals can be met. Given this analysis the options ‘Stationary engine test cycle on chassis dynamometer’ (possibly with limited transient parts) and the ‘three on-road cycle options’ have been given the same marks. When in future more cost data are available a new analysis will show the most cost effective option. OBD is very cost-effective from a government point of view, as it is paid for by the customer. It only needs to be checked for proper operation and a possible indication of malfunctioning. Practicability Performing a transient engine test by dismounting the engine from the vehicle is already considered to be very impracticable. The laboratory transient vehicle test cycle, on the other hand, is relatively easy to perform, because little or no instrumentation is needed on the vehicle. The on-road measurements require more instrumentation, and the installation of the on-board emission measurement equipment. Depending on the test method the vehicle is driven to the test circuit or is just driven on public roads. This is the case for the real-life arbitrary cycle with NTE limits. For the vehicle test cycles it has to be mentioned that the development of the dedicated test cycles will have to be carried out before introduction of such a testing method. Accuracy, reproducibility, and comparison with homologation test results Generally, the laboratory tests have an advantage over the on-road tests because of the conditioned environment.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 31
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
For OBD and the real-life arbitrary cycle with NTE limits these demands are of less importance, as for those options there is no reference to homologation test results. The assessment table and the accompanying motivations, lead to the following conclusions: •
Generally, the best choice of test method depends on the particular aims the legislator has set for the IUC programme. The choice can be derived from Table 3. Overall, laboratory tests offer better compliance with certification legislation, better research of maintenance effects on emissions and better accuracy as well as reproducibility. On-road testing yields better cycle bypassing resistance at equal to possibly better cost-effectiveness (to be confirmed).
•
Vehicle based test cycles are not suitable for legislation purposes, as HD engines appear in a variety of different vehicle types and drivelines.
•
Evaluating the legislation compliance can only be actively monitored by laboratory tests that use an engine based test cycle. Alternatively, OBD systems can be applied, but they can only check a number of engine parameters, which is no guarantee for compliance.
•
From the viewpoint of reliable emission factors the transient laboratory tests are more suitable for the elaboration of bottom-up emission models that enable detailed emission studies, whilst on-road NTE emission tests have the advantage of providing real-life in use emissions. A possible benefit can be gained if a combination of transient laboratory tests with complementary on-road test results (for validation purposes) is applied.
•
If the objective of the IUC programme is to obtain detailed information concerning maintenance effects, the best alternative is to use laboratory tests, for reasons of accuracy and repeatability.
•
The role of OBD is one that can deliver a useful contribution, but cannot replace IUC testing. This role may change if OBD is extended with an emission measuring functionality (OBM).
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 32
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
4.2.3 Comparison of status of heavy-duty in-use conformity in EU and USA
In the USA, engine and vehicle manufacturers are forced by the ‘Consent Decrees’ to undertake actions to prepare for the coming directives on in-use compliance for heavy-duty vehicles. In Europe national programmes on IUC are running in The Netherlands and Germany. No actions are enforced on the manufacturers yet.
4.2.4 Differences between EU and USA heavy-duty vehicle use via WHHD engine cycle It was planned that an analysis of differences between the USA and EU situations would be made with the help of experience gained by TNO Automotive in the course of a project for the development of a world harmonised Heavy Duty (WHHD) engine test cycle. This would provide good insight into the potential of applying this in the EU. Due to the fact that the project on the WHDC is not yet finished, limited data are available [13]. Detailed information could not be incorporated in this report. Some main results relevant to this study are: •
The average engine power over the combined idle and power delivery modes is 30% for the EU, 26% for the US and 20% for Japan (World: 26%).
•
Motoring and idle modes amount up to 40% of total time
•
the region in between idle and 30% of maximum power amounts up to 25% of total time for the EU and 40% for the US (World: 30%)
•
for the remaining 35% in the EU and 20% in the US (World: 30%) the engine delivers more the 30% of its maximum power. So the NTE carve-out below 30% of maximum power (see section 4.2.1.1) eliminates 70% (World) of all engine modes.
•
A World Harmonised Steady state Cycle (WHSC) is under development as a basis for IUC. The test should be incorporated in the certification testing to serve as reference.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 33
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
4.2.5 Applicability of USA in-use conformity approach in EU 4.2.5.1
In-use conformity testing based on the NTE approach
As explained in sections 4.2.1 and 4.2.2 there is no HD IUC regulation in place either in the USA or in the EU. Looking at the upcoming Euro 4 regulation, in the EU a slightly adapted LD Euro 3 IUC system is proposed. In the USA no IUC regulation is incorporated in the 2004 and 2007 sections of the Code of Federal Regulation (40CFR86). However, the NTE limits are incorporated as part of the engine certification procedure. This could possibly open the way for a future IUC based on these NTE limits. The actual IUC NTE testing could be executed on a dynamometer or on-road. For on-road tests the NTE limits are well suited, as this means that the vehicle can be operated normally in traffic (instead of following a prescribed speed cycle). This USA situation is here referred to as USA IUC “approach”. In the following both the applicability of this approach as such and as within the EU is discussed.
4.2.5.2
Not-to-exceed limits in general
The NTE limit definition enforced on a HD vehicle looks to be an ideal tool for IUC. Within relatively wide intervals for environmental parameters such as temperature, relative humidity, etc. the vehicle has to comply with the NTE limits for the defined NTE control area under the torque-rpm curve of the engine (see section 4.2.1.1). As the NTE limits are only 1.5 times the FTP certification limits some areas under the torque-rpm curve are carved out i.e. those areas with high brake specific emissions. Otherwise the limits are not feasible. If the engine had to fulfil NTE limits in each 30 second window within the NTE control area including the carve-outs a few seconds in these regions could cause the engine to fail passing. As the engine was believed to stay only a brief time in the carved out areas during its normal operation, emissions out of these areas would contribute little to its real-life emission behaviour. This would justify the carve-outs. Comments on the test procedure In phase II of the work carried out by West Virginia University under the Consent Decrees on-road testing revealed that remaining for 30 seconds within the NTE zone can be quite difficult (see section 4.2.1.3.2). The resulting low NTE availability poses a problem in itself, in that many measurements from within the NTE zone have to be rejected. The question arises if in this way all real-life emissions are sufficiently well ‘weight reflected’ in the NTE test results. Clearly this cannot be the case. Let us take the example of the emissions carve-out under the 30% Maximum Power curve in the torque-speed diagram. If an engine frequents this region regularly, not only will the NTE availability be low, but also many measurements just above the 30% curve will be rejected and thus underweighted. If this concerns regions under the NTE Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 34
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
zone with relatively high brake specific emissions, then NTE testing will not reflect what it has been intended for. Moreover, if an engine manufacturer ‘adapts’ his engine so as to frequent more the carve-outs more, with high brake specific emissions, then he can influence the final results in his favour. This effect can be enhanced by the kind of cycle that is driven. Another approach to the NTE limits could be to minimise the carve-outs in combination with higher limits, i.e. more then 1.5 times the FTP limits. Alternatively, the carve-outs could become regions with smaller weight factors assigned, thus countering for the higher brake specific emissions. This would give a much higher NTE availability and a better weighted result, thereby leaving less room for engine “adaptation”. Also the NTE 30 second window must be evaluated. A smaller window might be undesirable, because a high but short emission event can cause the engine to fail the test. A too large window on the other hand might attenuate transient emissions too much. More field experience with the NTE approach will show how well the concept works and where improvements and further research might be necessary. Still it is believed that the NTE limits offer a good solution for in-use conformity checking. Given a good approach the engine is tested under a wide variety of in-use conditions thus maximising the certainty that the in-use emissions are under control.
4.2.5.3
NTE in Europe
The USA IUC approach when applied in the EU might need adaptations related to the differences in vehicle use between the USA and the EU (see also further ‘Road load compared to NTE area’). The NTE concept itself is believed to be universally applicable. However, the definition of the NTE zone and the carved out areas, the engine classification and the (test) methodology could be specific for the EU situation. NTE methodology The methodology should be as alike as possible in as many aspects as feasible. Given the fact, that in the USA the NTE limits are FTP limit based, this would mean that in the EU the base could be the ETC cycle. If the ESC could be the base in both the USA and EU this would lead to further harmonisation of emission regulation. The Euro 3 ESC is incorporated in the CFR as Supplemental Emission Test (SET) for future regulation. The ESC however is a stationary test cycle, thus leading to a situation where limits for dynamic NTE testing would be based on limits of a steady state test. A way out of this is presented by the World Harmonised Heavy Duty (WHHD) engine test cycle. This cycle is dynamic and intended for universal use. Moreover, it is designed to reflect real-life engine use. In this way in-use testing limits are based on limits of a dynamic and realistic certification cycle which seems Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 35
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
a logical approach. Furthermore, it might be advisable to extend the certification testing to include NTE testing along the WHHD cycle. This yields a firm link between certification and in-use conformity testing as NTE testing is a factor common to both. It is important to note that NTE testing within the certification procedure excludes the use of stationary tests such as the ESC to link the NTE limits to. The NTE testing is intrinsically dynamic and of course not designed for use with a steady-state cycle. The path along the WHHD cycle presents a long term approach. It might well take until 2012-2015 before a concept like the WHHD cycle is incorporated in the USA, Japanese and EU emission legislation. This 10 years plus timeframe may look like a long period, however it must be taken into account that the US 2007 and Ero 5 (2008) regulations are already in place. The WHHD cycle then has to be fitted into the different new regulations. A lead time has to be given to the engine manufacturers – a four years minimum period by legislation in the USA – to let them prepare for the new situation. Thus a 10 years plus timeframe seems to be realistic. Chassis and engine dynamometer versus on-road measurements In section 4.2.2.8 an extensive assessment covering different aspects of possible test methods for IUC is presented. In the light of world wide harmonisation a common set of specifications for on-road systems should be used across the USA and the EU. Also, the test protocol should be universal. Engine classification The engine classification within the USA NTE approach (see section 4.2.1.1) is based on the C speed with 2400 rpm as spill. Whether this is applicable for the EU needs further investigation. A comparison needs to be made between the USA and EU HD engines. However, as the USA upcoming regulation includes the EU ESC cycle as a Supplemental Emission Test (SET) and as the definition of the C speed is identical to ESC and NTE, not many problems seem to arise as to differences in engines. Therefore, engine classification should pose little problem.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 36
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
Road load compared to NTE area The definition of the NTE zone and carved out areas under the torque-rpm curve of the engine is given some general thought at the beginning of this section. The applicability in the EU is questionable. In particular the carve-out below the 30% maximum power curve needs attention. In the EU -and even more in the US- heavy-duty engines mounted in on-road vehicles operate for a considerable time underneath this curve [13] (see section 4.2.4). This is especially true for MHDDE and LHDDE. The data gathered in the study on the WHHD cycle should be examined to help propose a new carve-out at a lower than 30% maximum power curve. Unfortunately these data are not public at the time of writing this report. It might be necessary to have separate definitions for this carve-out for engines with C speeds below 2400 rpm and for those above. Further actions The items of this section should be the subject of discussions with the engine manufacturers and the EU member states, for example within the Motor Vehicle Emission Group (MVEG) that is chaired by the EC. Thus a common European approach can be developed. Moreover, in the light of harmonisation of emission regulations, the EC and the US EPA should confer on their intended approaches to IUC. It should be stressed that the ideas, advice and comments in this section of the report (on USA IUC applicability in general and in the EU) are developed by Vito from experience in on-road in-use emission measurements. Of course this does not mean that no other, maybe better approaches are possible. The discussion surrounding IUC has only just started.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 37
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
4.3 ON-ROAD EMISSION TESTING FOR IN-USE CONFORMITY 4.3.1 From the USA study tour On-road or roadside measurement systems could be used as a screening tool for in-use conformity. That is, these systems could be used to identify possible problem engines (or engine families). These engines could then be tested, using a chassis dynamometer and finally on an engine dynamometer for full in-use conformity checking. The latter testing would be rather expensive and a system would need to be devised that did not inconvenience vehicle operators. One manufacturer suggested that it might be more appropriate to simply to check that the after-treatment device is in place and is working. The Consent Decrees require the manufacturers to contribute towards the development of an on-road measurement system. This system is known as MEMS (Mobile Emissions Measurement System) and is being developed at West Virginia University. The EPA is also developing two alternative systems, called Rover (Real time On-road Vehicle Emission Reporter) and Spot. The latter is also known as PEMS (Portable Emissions Measurement System), and has been developed primarily for developing emission factors for offroad vehicles. One of the main difficulties has been measuring the exhaust flow rate. Ford Motor Company has also developed a system for gasoline vehicles know as Preview (see 4.3.3.1).
4.3.2 Manufacturers’ vision on on-road measurement systems Cummins believes that MEMS is better than Rover, but are waiting for EPA’s official response. None of the three manufacturers have yet tested MEMS. Most of the testing undertaken by the University of West Virginia has used a Mack engine. The HD engine manufactures believe that the EPA will use on-road measurement as a screening tool, with further chassis and engine dynamometer tests required to confirm non-conformity, at least initially. Only when on-road measurement has been shown to correlate well with engine dynamometer tests can it be used as a conformity enforcement tool. The EPA confirmed that this would require additional rulemaking (and a four year lead time). Therefore, it could not come into force until 2011 at the earliest. As the emission levels decline from 2007, it will become increasingly difficult to undertake on-road measurements. The ‘noise’ is likely to be of the same order of magnitude as the measurements. Comparing on-road emission measurements with the ‘Not to Exceed’ limits will also require a robust method of converting gram/second into g/brake horsepower hour for measurements averaged over a minimum of 30 seconds. The Consent Decrees require on-road measurement of torque with increasing accuracy.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 38
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
In Caterpillar’s view the development of a sufficiently accurate on-road measurement device is a long way off. Cummins believe that it is not really on the political agenda yet and that an on-road test of the aftertreatment device will probably be sufficient.
4.3.3 Review of known on-road systems A list has been prepared of the mobile emission measurement systems (measuring and acquiring the mass of HC, CO, CO2 and NOx, PM optional), that are known. These systems were developed by automotive manufacturers, manufacturers of measurement equipment, universities or research institutes, and will be reviewed. It is not obvious to compare the different on-board emission measurement systems on common specifications, as this is not a standard technique. Every author describes his system in his own way together with ‘in-house’ specifications. Furthermore, definitions can vary between authors, for example accuracy can be given as a typical accuracy or as an average figure, etc. Nevertheless, in the description of the systems an effort is made to discuss them under eight parameters: Name
Acronym and/or full name of the system
Use
Typical use of the system, e.g. for diesel fuelled vehicles, especially for IUC testing, etc.
Principle
Measurement principle, e.g. on raw or diluted exhaust gas, bag collection or continuous measurement, dynamic measurements or not
Methodology
How the mass emissions results are calculated from the measurements, e.g. emission concentration measurements coupled with exhaust gas flow rate from direct determination
Emissions covered
The measured exhaust emission components are detailed
Apparatus
Analyser make and type, etc.
Inaccuracy
Measurement results on-road are referred to measurement results on a chassis dynamometer.
Weight/size
Portability of the system
If no information is available on a parameter the corresponding line is left blank.
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 39
EC-DG ENTR
4.3.3.1
Emission control technology for heavy-duty vehicles
ETD/00/503430
Systems developed by automotive manufacturers:
Caterpillar A portable bag collection system was developed by Caterpillar to quantify fuel specific NOx emission levels for in-use diesel engines in 1982 [11]. Name Use
In-use emission measurements on HD diesel engines
Principle
Raw exhaust in bag, with water removal before the bag. No dynamic measurements
Methodology
Concentration measurement, fuel specific emissions
Emissions covered
NOx
Apparatus
Bag collection
Inaccuracy
10% on a concentration base
Weight/size
‘suitcase’ size
Ford Ford started to develop an on-road emission measurement system in the early nineties. This resulted in a system for measuring gasoline vehicles [11]. Name
OBE On-Board Emissions
Use
emission measurements on gasoline LD vehicle for simultaneous comparison to remote sensing of exhaust
Principle
continuous diluted exhaust measurement
Methodology Emissions covered
CO2, CO, HC, NOx
Apparatus
FTIR plus dilution tunnel
Inaccuracy
< 10%
Weight/size Further developments of this system resulted in Preview [14]. The Ford Preview system (for measuring gasoline vehicles) appears to be well developed. It can measure emissions down to the SULEV levels with a good correlation with chassis dynamometer measurements. This system has largely been developed to reduce engine optimisation costs, as it allows operability and emissions to be calibrated together, rather than
Volume 5
Development of proposals for in-use conformity testing of emission control devices
Page 40
EC-DG ENTR
Emission control technology for heavy-duty vehicles
ETD/00/503430
calibrating for operability and then going into the laboratory for an emissions test. It is likely that Sensors Inc, which has been involved in the development of Preview, will market the product, which might also include a diesel version (see 4.3.3.2). Name
Preview Portable Real-time Emission Vehicular Integrated Engineering Workstation
Use
In-use emission measurements on LD gasoline vehicles
Principle
Continuous raw exhaust measurement
Methodology
Concentration measurements coupled with exhaust gas flow rate derived from engine parameters
Emissions covered
CO
CO2
Apparatus Inaccuracy Weight/size
NO
HC
Co-development with Sensors Inc.
