Regulations Requiring Onboard Diagnostic
Systems on 2010 and Later Heavy-Duty
Engines Used in Highway Vehicles Over
14,000 Pounds; Revisions to Onboard
Diagnostic Requirements for Diesel
Highway Vehicles Under 14,000 Pounds
Summary and Analysis of Comments
United States
Environmental Protection
Agency
-------
Regulations Requiring Onboard Diagnostic
Systems on 2010 and Later Heavy-Duty
Engines Used in Highway Vehicles Over
14,000 Pounds; Revisions to Onboard
Diagnostic Requirements for Diesel
Highway Vehicles Under 14,000 Pounds
Summary and Analysis of Comments
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
v>EPA
United States EPA-420-R-08-018
Environmental Protection _. , „„„
Agency December 2008
-------
Table of Contents
Index of Commenters - HDOBD 1
Index of Commenters - Service Information Availability (SIA) 1
Introduction to this Document 2
General statements in support of the proposed rule 2
General statements in opposition to the proposed rule 2
General statements in support of other comments 2
I. Overview 3
I.A Background 3
I.B What is EPA Proposing? 3
I.B.1 OBD Requirements for Engines Used in Highway Vehicles Over 14,000 Pounds GVWR 3
I.B.2 Requirements that Service Information be Made Available 4
I.B.3 OBD Requirements for Diesel Heavy-duty Vehicles and Engines Used in Vehicles Under
14,000 Pounds 4
I.C Why is EPA Making this Proposal? 5
I.C.1 Highway Engines and Vehicles Contribute to Serious Air Pollution Problems 5
I.C.2 Emissions Control of Highway Engines and Vehicles Depends on Properly Operating
Emissions Control Systems 5
I.C.3 Basis For Action Under the Clean Air Act 5
I.D How has EPA Chosen the Level of the Proposed Emissions Thresholds? 6
I.E World Wide Harmonized OBD (WWH-OBD) 7
I.F Onboard Diagnostics for Diesel Engines used in Nonroad Land-based Equipment 9
I.F.1 What is the baseline nonroad OBD system? 10
I.F.2 What is the appropriate level of OBD monitoring for nonroad diesel engines? 10
I.F.3 What should the OBD standardization features be? 11
I.F.4 What are the prospects and/or desires for international harmonization of nonroad OBD?12
II. What are the Proposed OBD Requirements and When Would They be Implemented? 12
II.A General OBD System Requirements 12
II.A.1 The OBD System 13
II.A.2 Malfunction Indicator Light (MIL) and Diagnostic Trouble Codes (DTC) 13
II.A.3 Monitoring Conditions 17
II.A.4 Determining the Proper OBD Malfunction Criteria 18
II.B Monitoring Requirements and Timelines for Diesel-Fueled/Compression-lgnition Engines.. 20
II.B.1 Fuel System Monitoring 22
I I.B.2 Engine Misfire Monitoring 23
II.B.3 Exhaust Gas Recirculation (EGR) System Monitoring 24
II.B.4 Turbo Boost Control System Monitoring 24
II.B.5 Non-Methane Hydrocarbon (NMHC) Converting Catalyst Monitoring 25
II.B.6 Selective Catalytic Reduction (SCR) and Lean NOx Catalyst Monitoring 26
II.B.7 NOx Adsorber System Monitoring 27
II.B.8 Diesel Particulate Filter (DPF) System Monitoring 28
II.B.9 Exhaust Gas Sensor Monitoring 29
II.C Monitoring Requirements and Timelines for Gasoline/Spark-lgnition Engines 30
II.C.1 Fuel System Monitoring 30
II.C.2 Engine Misfire Monitoring 30
II.C.3 Exhaust Gas Recirculation (EGR) Monitoring 30
II.C.4 Cold Start Emission Reduction Strategy Monitoring 30
I I.C.5 Secondary Air System Monitoring 30
II.C.6 Catalytic Converter Monitoring 30
II.C.7 Evaporative Emission Control System Monitoring 30
-------
II.C.8 Exhaust Gas Sensor Monitoring 31
II.D Monitoring Requirements and Timelines for Other Diesel and Gasoline Systems 31
II.D.1 Variable Valve Timing and/or Control (WT) System Monitoring 31
II.D.2 Engine Cooling System Monitoring 31
II.D.3 Crankcase Ventilation System Monitoring 33
II.D.4 Comprehensive Component Monitors 33
II.D.5 Other Emissions Control System Monitoring 35
II.D.6 Exceptions to Monitoring Requirements 35
II.E A Standardized Method to Measure Real World Monitoring Performance 36
II.E.1 Description of Software Counters to Track Real World Performance 36
II.E.2 Proposed Performance Tracking Requirements 37
II.F Standardization Requirements 37
II.F.1 Reference Documents 37
II.F.2 Diagnostic Connector Requirements 38
II.F.3 Communications to a Scan Tool 39
II.F.4 Required Emissions Related Functions 40
II.F.5 In-use Performance Ratio Tracking Requirements 43
II.F.6 Exceptions to Standardization Requirements 45
II.G Implementation Schedule, In-use Liability, and In-use Enforcement 45
II.G.1 Implementation Schedule and In-use Liability Provisions 45
II.G.2 In-use Enforcement 46
II.H Proposed Changes to the Existing 8,500 to 14,000 Pound Diesel OBD Requirements 47
II.H.1 Selective Catalytic Reduction and Lean NOx Catalyst Monitoring 50
II.H.2 NOx Adsorber System Monitoring 50
II.H.3 Diesel Particulate Filter System Monitoring 50
II.H.4 NMHC Converting Catalyst Monitoring 50
II.H.5 Other Monitors 50
II.H.6 CARB OBDII Compliance Option and Deficiencies 50
11.1 How do the Proposed Requirements Compare to California's? 50
Are the Proposed Monitoring Requirements Feasible? 52
III.A Feasibility of the Monitoring Requirements for Diesel/Compression-lgnition Engines 52
III.A.1 Fuel System Monitoring 52
III.A.2 Engine Misfire Monitoring 54
III.A.3 Exhaust Gas Recirculation (EGR) Monitoring 55
III.A.4 Turbo Boost Control System Monitoring 57
III.A.5 Non-Methane Hydrocarbon (NMHC) Converting Catalyst Monitoring 59
III.A.6 Selective Catalytic Reduction (SCR) and NOx Conversion Catalyst Monitoring 60
III.A.7 NOx Adsorber Monitoring 64
III.A.8 Diesel Particulate Filter (DPF) Monitoring 64
III.A.9 Exhaust Gas Sensor Monitoring 68
III.B Feasibility of the Monitoring Requirements for Gasoline/Spark-lgnition Engines 68
III.B.1 Fuel System Monitoring 68
III.B.2 Engine Misfire Monitoring 69
III.B.3 Exhaust Gas Recirculation (EGR) Monitoring 69
III.B.4 Cold Start Emission Reduction Strategy Monitoring 69
III.B.5 Secondary Air System Monitoring 69
III.B.6 Catalytic Converter Monitoring 70
III.B.7 Evaporative System Monitoring 70
III.B.8 Exhaust Gas Sensor Monitoring 70
III.C Feasibility of the Monitoring Requirements for Other Diesel and Gasoline Systems 70
III.C.1 Variable Valve Timing and/or Control (WT) System Monitoring 70
III.C.2 Engine Cooling System Monitoring 70
III.C.3 Crankcase Ventilation System Monitoring 72
-------
III.C.4 Comprehensive Component Monitoring 72
IV. What are the Service Information Availability Requirements? 73
IV.A Comparison of the Aftermarket Service Industry Below 14,000 Pounds and Above 14,000
Pounds 73
IV.B Information to be Made Available by OEMs 75
IV.B.1 Definition of Emissions-related Information 75
IV.B.2 Other Information 76
IV.B.3 Regulating the Cost of Service Information 77
IV.C Requirements for Web-based Delivery of the Required Information 78
IV.C.1 General Comments on Web Site Availability 78
IV.C.2 Short-Term, Medium-Term, and Long-Term Access 80
IV.C.3 Pricing Approval 82
IV.C.4 Limitations on Makes and Models 82
IV.C.5 Length of Information Availability 83
IV.C.6 Correcting Broken or Deleted Web Links on a Weekly Basis 83
IV.C.7 Printing of Information from Web Sites 84
IV. D Availability of Training Information 84
IV.D.1 Delivery of Training Information 84
IV.D.2 Availability of Classroom Training 85
IV.D.3 Duplication of Training Information 86
IV.E Service Information for Third Party Information Providers 86
IV.F Recalibration Information 87
IV.G Pass-through Reprogramming Capabilities 88
IV.H Availability of Generic and Enhanced Information for Scan Tools for Equipment and Tool
Companies 89
IV.H.1 Generic Information 89
IV.H.2 Enhanced Information 89
IV.I Availability of OEM-Specific Diagnostic Scan Tools 90
IV. 1.1 General Provisions for Tool Availability 90
IV.I.2 Requiring Training as a Condition of OEM-specific Scan Tool Purchases 94
IV.J Reference Materials Being Proposed for Incorporated by Reference 95
IV. K Other Comments Received 95
IV.K.1 Costs Associated with the Rule 95
IV.K.2 Alignment with Other Agency and State Requirements 97
IV.K.3 Scope of the Rule 97
IV.K.4 Timing 99
IV.K.5 Persons Entitled to Access 100
IV.K.6 Liability Concerns 101
IV.K.7 Compliance Flexibility 102
IV.K.8 The Service Information Requirements Must Comply with the CAA and Federal Law 102
IV.K.9Tier1 Suppliers 104
IV.K.10 Adding References to "Authorized Service Network" 105
IV.K.11 Errors in Section References 105
V. What are the Emissions Reductions Associated with the Proposed OBD Requirements? 106
V.A Emissions Reductions Associated with the 2007HD Highway Rule 106
VI. What are the Costs Associated with the Proposed OBD Requirements? 106
VI.A Variable Costs for Engines Used in Vehicles Over 14,000 Pounds 108
VLB Fixed Costs for Engines Used in Vehicles Over 14,000 Pounds 109
VI.C Total Costs for Engines Used in Vehicles Over 14,000 Pounds 112
VI.D Costs for Diesel Heavy-duty Vehicles and Engines Used in Heavy-duty Vehicles Under
14,000 Pounds 113
VII. What are the Updated Annual Costs and Costs per Ton Associated with the 2007/2010 Heavy-
duty Highway Program? 113
-------
VILA Updated 2007 Heavy-Duty Highway Rule Costs Including OBD 113
VII.B Updated 2007 Heavy-Duty Highway Rule Costs per Ton Including OBD 113
VIII. What are the Requirements for Engine Manufacturers? 114
VI11.A Documentation Requirements 114
VI11.B Catalyst Aging Procedures 114
VIII.C Demonstration Testing 114
I.C.1 Selection of Test Engines 114
I.C.2 Required Testing 115
I.C.3 Testing Protocol 116
VIII.C.4 Evaluation Protocol 116
VIII.C.5 Confirmatory Testing 116
VIII.D Deficiencies 117
I.E Production Evaluation Testing 117
I.E.1 Verification of Standardization Requirements 118
I.E.2 Verification of Monitoring Requirements 118
I.E.3 Verification of In-use Monitoring Performance Ratios 118
IX. What are the Issues Concerning Inspection and Maintenance Programs? 119
IX.A Current Heavy-duty I/M Programs 119
IX. B Challenges for Heavy-duty I/M 119
IX.C Heavy-duty OBD and I/M 119
X. Statutory and Executive Order Reviews 122
X.A Executive Order 12866: Regulatory Planning and Review 122
X.B Paperwork Reduction Act 123
X.C Regulatory Flexibility Act (RFA), as amended by the Small Business Regulatory
Enforcement Fairness Act of 1996 (SBREFA), 5 USC 601 et.seq 123
X.D Unfunded Mandates Reform Act 123
X.E Executive Order 13132: Federalism 123
X.F Executive Order 13175: Consultation and Coordination with Indian Tribal Governments .. 123
X.G Executive Order 13045: Protection of Children from Environmental Health and Safety Risks123
X.H Executive Order 13211: Actions that Significantly Affect Energy Supply, Distribution, or Use123
X.I National Technology Transfer Advancement Act 123
XI. Statutory Provisions and Legal Authority 123
XII. Other comments 123
IV
-------
Index of Commenters - HDOBD
"^ Commenter
#
1 Community Board #1-M
2 Texas Commission on Environmental Quality
3 Volvo Powertrain
4 Maryland Department of the Environment
5 European Association of Internal Combustion Engine
Manufacturers (Euromot)
6 Northeast States for Coordinated Air Use Management
(NESCAUM)
7 Freightliner
8 Engine Manufacturers Association (EMA)
9 New Jersey Department of Environmental Protection
10 Truck Manufacturers Association
1 1 American Trucking Associations (ATA)
12 Caterpillar Inc.
13 Cummins Inc.
14 Truck Renting and Leasing Association (TRALA)
1 5 New York State Department of Environmental
Conservation (NY State DEC)
16 National Automobile Dealers Association (NADA)
17 California Air Resources Board (CARB)
18 Mitsui Mining & Smelting Co., Ltd. *
* Denotes a comment received after the close of the comment period.
Document ID #
2005-0047-0015
2005-0047-0020
2005-0047-0021
2005-0047-0022
2005-0047-0023
2005-0047-0024
2005-0047-0025
2005-0047-0026
2005-0047-0027
2005-0047-0028
2005-0047-0029
2005-0047-0030
2005-0047-0031
2005-0047-0032
2005-0047-0033
2005-0047-0034
2005-0047-0035
2005-0047-0040
Index of Commenters - Service Information Availability (SIA)
Entry # Commenter
1 Hughes Telematics, Inc.
2 Caterpillar Inc.
3 Volvo Powertrain
4 Automotive Aftermarket Industry Alliance; Automotive
Engine Rebuilders Association; Automotive Parts
Remanufacturers Association; Heavy Vehicle Maintenance
Group (referred to throughout as "Aftermarket
Associations")
5 Engine Manufacturers Association (EMA)
6 National Automobile Dealers Association (NADA)
7 Volvo Powertrain
Document ID #
2005-0047-0037
2005-0047-0038
2005-0047-0039
2005-0047-0041
2005-0047-0042
2005-0047-0043
2005-0047-0044
-------
Introduction to this Document
The outline for this document is identical to the outline used in the preamble to our proposed
rule (72 FR 3200). We have summarized comments and placed those comments under the heading
which most accurately characterized the nature of the comment. As such, issues can be identified
by section number within which those issues have been placed. The reader should keep in mind
that we have attempted to place a comment into only one issue area or section number within this
document. For example, we received comments pertinent to threshold levels and/or feasibility of our
proposal with respect to heavy-duty OBD on applications under 14,000 pounds (see section II.H).
However, in section I.B.3 of this document, we note that we received no comments pertinent to
section I.B.3 which was the preamble section that provided an overview of our proposed
requirements for OBD on applications under 14,000 pounds. We make such a note since no one
commented on that overview section and, instead, we have summarized the comments and made
our responses under section II.H.
General statements in support of the proposed rule
The following commenters expressed general support for the HDOBD portions of the
proposed rule.
Community Board #1 (2005-0047-0015)
Texas Commission on Environmental Quality (2005-0047-0020)
NESCAUM (2005-0047-0024)
General statements in opposition to the proposed rule
The following commenters expressed general opposition to the proposed rule.
None.
General statements in support of other comments
The following commenters expressed support for the comments submitted by the Engine
Manufacturers Association, 2005-0047-0026:
Volvo Powertrain, 2005-0047-0021, p. 1
Cummins Inc., 2005-0047-0031, p. 3
Caterpillar Inc., 2005-0047-0030, p. 2
Truck Manufacturers Association, 2005-0047-0028, p. 1
National Automobile Dealers Association, 2005-0047-0034, p. 1
The following commenters expressed support for the comments submitted by the Engine
Manufacturers Association, 2005-0047-0026, where those comments pertain to nonroad OBD:
Euromot, 2005-0047-0023, p. 3
-------
I. Overview
We received no comments pertaining to this section that require analysis.
I.A Background
We received no comments pertaining to this section that require analysis.
I.B What is EPA Proposing?
I.B.1 OBD Requirements for Engines Used in Highway Vehicles Over 14,000 Pounds GVWR
Comments:
Engine manufacturers face many challenges in providing engines to meet the needs of
customers, EPA, the California Air Resources Board ("ARB"), other government agencies, and their
own businesses. During the past five years, engine manufacturers have poured huge resources into
meeting stringent new federal and California emissions standards that began in 2007 and that will be
fully realized by 2010. The new emission standards will reduce engine emissions by more than 90%
and those reductions will come through using improved engine design, advanced aftertreatment
systems and low-sulfur fuel. The 2007/2010 heavy-duty engine emission standards will result in
diesel technology - long known for being the most durable and energy-efficient - having the right to
also be called clean. Engine manufacturers continue to invest significant resources to develop and
produce engine and aftertreatment technology meeting the stringent new emission standards. They
have already devoted and will continue to devote thousands of hours of engineering time and
expertise and thousands of hours of time in the emissions test cell to achieve those standards.
Manufacturers have begun producing compliant engines for 2007. But, while they expect to meet the
fully-realized new emission standards by 2010, future success is by no means assured.
At the same time, manufacturers are addressing the challenges of the new nationwide
manufacturers' run heavy-duty in-use test program and new California engine manufacturer
diagnostic ("EMD") standards which began in 2007. Those EMD standards are the first step toward
comprehensive heavy-duty OBD requirements as embodied in the Proposed Rule, which would
require engine manufacturers to develop and produce heavy-duty OBD technology to monitor all
engine systems to stringent new requirements beginning in 2010. Engine manufacturers do not
know how they will meet the new heavy-duty OBD requirements or even if they will be capable of
meeting those challenging requirements. What they do know is that the Proposed Rule will require a
major investment of manufacturer resources to invent the monitoring technology and develop it to a
point that it can be used with confidence on 2010 and later engines. Manufacturers will undertake a
workload, invent technology, and invest costs to make changes that must be implemented and
engineered on significantly more engine models and ratings, and recouped on far fewer units of sale,
than ever has been required in any other OBD program. This challenge is made all the more difficult
by the fact that engine manufacturers are still developing the technology to meet the underlying
emission standards for 2010.
OBD is technically complex, and means sophisticated new systems placed on engines and
vehicles. Regulating how manufacturers use OBD and monitor their engine emission control adds
more complexities and new challenges to produce engines that are compliant with 2010 and later
standards. But heavy-duty engine manufacturers have very little experience with regulated OBD
systems. Although some companies have experience with engines and vehicles under 14,000 Ibs.
(light- and medium-duty), most have none. All engine manufacturers will need to devote substantial
time and effort to meeting the new rule, but those without experience will have special challenges to
-------
overcome - from something as basic as understanding OBD terminology to something as complex
as creating the algorithms and writing the software code used for monitoring.
Moreover, many of the monitoring strategies that would be required to meet EPA's proposed
OBD standards depend on the development and use of accurate and durable sensor technology.
Engine manufacturers do not produce sensors and do not control their development or availability in
the market. Nor do engine manufacturers control the accuracy and reliability of those sensors.
Technology-forcing standards are appropriate if the entities held responsible for meeting those
standards have control of the development of the technology. In this case, engine manufacturers are
required to meet the technology-forcing standards, but have no such control.
Finally, what has worked for light- and medium-duty OBD will not necessarily work for heavy-
duty engines. The two industries are very different. The heavy-duty industry is generally a non-
integrated industry, meaning that engine manufacturers sell their products - engines - to customers
who take those engines and incorporate them into many different types of vehicles, with many
different types of transmissions, customer specifications and performance requirements. Engine
manufacturers simply cannot predict all the possible variations in which their engines will be used
and they do not have control over vehicles. EPA has recognized that fact, in part, in the Proposed
Rule, and has proposed to limit the requirements as much as possible to engines. But in the non-
integrated heavy-duty engine and vehicle industry, there is an extreme burden associated with
calibrating OBD monitors for use in a myriad of different vehicle configurations. Further changes
must be made to the Proposed Rule to limit engine manufacturers' responsibility for vehicle matters
outside their control.
EMA, 2005-0047-0026, p. 1-3
National Automobile Dealers Association, 2005-0047-0034, p. 3
Agency Response:
We have carefully considered the points made by commenters in developing our final rule.
While sensor technology has progressed rapidly, there remain some limitations in accuracy and
durability. As such, we have relaxed some 2010-2012 NOx thresholds to levels at which current
technology can detect, and we have provided some PM-related monitoring requirements that our
testing and industry testing show can be met. We have also addressed industry comments as
regards the myriad of different vehicle configurations by providing greater flexibility in some
monitoring requirements (e.g., cooling system monitoring) without compromising the expected
effectiveness of the OBD systems. Most importantly, the sensors needed to comply with the final
requirements do indeed exist, and they exist today, at least for the 2010-2012 requirements. Some
of the 2013 requirements—notably the DPF monitoring requirements—would require, we believe, a
soot sensor with greater sensitivity and durability than what exists today. In that case, we cannot
state that the necessary sensors exist today. However, we believe that they will exist in time for
2013 compliance, and we will keep abreast of technological advances in the coming years in case
our requirements have to be modified. We address these issues in more detail in the following
sections.
I.B.2 Requirements that Service Information be Made Available
Please refer to Section IV of this document for a summary and analysis of all service
information availability comments and issues.
I.B.3 OBD Requirements for Diesel Heavy-duty Vehicles and Engines Used in Vehicles Under
14,000 Pounds
-------
We received no comments pertaining to this section that require analysis.
I.C Why is EPA Making this Proposal?
I.C.1 Highway Engines and Vehicles Contribute to Serious Air Pollution Problems
We received no comments pertaining to this section that require analysis.
I.C.2 Emissions Control of Highway Engines and Vehicles Depends on Properly Operating
Emissions Control Systems
We received no comments pertaining to this section that require analysis.
I.C.3 Basis For Action Under the Clean Air Act
Comments:
EPA is obligated under CAA Section 202(a)(3)(A) to propose standards that are
technologically feasible and cost effective. EPA has failed to meet it s obligations with respect to the
proposed HDOBD program. EPA has failed fully to analyze and consider the technological feasibility
of the proposed OBD thresholds and requirements, the practically of meeting many of the
requirements, and the workload and cost burden proposed to be placed on manufacturers. EPA has
not provided an adequate analysis of technological feasibility and the cost effectiveness of its
proposal.
EPA is obligated by Section 202(a)(3)(C) of the CAA to provide four years leadtime for the
promulgation of new standards and at least a three year period of stability for those new standards.
However, if EPA adequately addresses the technological feasibility, practicality and cost
effectiveness issues discussed throughout EMA's comments, while maintaining the proposed
implementation phase-in, engine manufacturers believe they would be able to comply with such a
program in 2010. Given a limited amount of time in which to address the challenges of HDOBD,
manufacturers can succeed only if the changes recommended by EMA are made. If EPA makes
EMA's recommended changes, engine manufacturers may not be pressed to rely on their legal
rights to leadtime and stability guaranteed by the CAA.
EPA has not met its burden to show that the proposed requirements are cost effective. EPA
has both underestimated costs and has not fully analyzed cost effectiveness.
EMA, 2005-0047-0026, p. 47-49
While Freightliner understands that EPA strove to align its OBD requirements to those of
CARB where feasible, we ask that the EPA adhere in future rulemakings to lead time and stability
limitations. The specific authority to require OBD in HD highway engines clearly stems from CAA
section 202(m)(1) which allows the Administrator to promulgate OBD regulations under CAA section
202(a). In turn, regulations promulgated under CAA section 202(a) are bound by the lead time and
stability requirements of that subsection at subparagraph (3)(C). EPA has already promulgated
regulations applicable to the year 2010 (40 CFR §86.007.11). Following Congress' stability mandate
-------
then, EPA must maintain its HD highway engine regulations without change until 2013. At the very
least, with EPA publishing its proposal after the commencement of model year 2007, Congress' lead
time mandate prohibits EPA from regulating OBD until model year 2012.
Freightliner, 2005-0047-0025, p. 1
Agency Response:
These comments can be summarized as: EPA has failed to show feasibility; and, EPA has
failed to provide proper lead time. EMA and Freightliner are incorrect in claiming that EPA is bound
by section 202(a)(3) of the Act for these OBD regulations. That provision applies only to standards
applicable to emissions of particular pollutants from heavy duty engines, not to OBD requirements
for such engines. This rule is governed by section 202(m) of the Act, which explicitly discusses
promulgation of OBD regulations, including regulations for heavy duty engines. The provisions of
section 202(a)(3), including the lead time and stability provisions, are therefore not applicable to this
rule. In addition, we believe that this issue is moot as a practical matter. On a national level, we
fully expect that manufacturers will sell the same OBD system that is being developed for use in the
State of California to comply with California's analogous OBD requirements. California made final
those requirements in 2005 which provided four years lead time to industry. The requirements we
are finalizing for 2010 compliance are less stringent in terms of OBD thresholds than those finalized
by California back in 2005. Therefore, lead time appears not to be an issue in practical terms. We
believe that the changes made to the final requirements should satisfy commenters as they have
themselves suggested.1
EMA mentions cost effectiveness and, while cost effectiveness is not mentioned in CAA
Section 202(m), we believe we have properly taken costs into consideration in promulgating this rule.
See the final technical support document contained in the docket where we show our estimated
costs as roughly $60 to $70 per engine for engines that are placed in vehicles selling for $50,000 to
more than $100,000 a piece. As for technological feasibility, we have revised some of our
thresholds for the 2010-2012 model years to address feasibility concerns raised by industry during
development of this final rule. We believe that these changes result in a feasible set of requirements
as discussed in our responses to comments in Sections II and III and by commenters themselves.2
As for the thresholds for model years 2013 and later, we have not made changes relative to our
proposal because we still believe that monitoring and sensing technology will advance such that the
thresholds will be feasible. We discuss this in more detail in our responses to comments in Sections
III.A.6 through III.A.9. If, in the 2010/2011 timeframe we are made aware that such advances have
not occurred, we will need to address the issue via possible changes to the 2013 and later
thresholds.
I.D How has EPA Chosen the Level of the Proposed Emissions Thresholds?
Comments:
We support requiring stringent OBD thresholds (i.e., OBD detection at lower emissions
levels) that will, among other things, induce manufacturers to produce more durable emission
controls. Accordingly, we support using the emissions thresholds listed in Tables II.B-1 and II.C.-1
as trigger points for requiring malfunction indicator light (MIL) illumination and storing diagnostic
1 See memorandum to Docket ID No. EPA-HQ-OAR-2005-0047 from Todd Sherwood, "Meetings with the
Engine Manufacturers Association," document ID # EPA-HQ-OAR-2005-0047-0053.
2 See memorandum to Docket ID No. EPA-HQ-OAR-2005-0047 from Todd Sherwood, "Meetings with the
Engine Manufacturers Association," document ID # EPA-HQ-OAR-2005-0047-0053.
-------
trouble codes (DTC). These thresholds are likely to achieve the balance sought by EPA between
environmental protection, system capabilities, and avoidance of repairs where costs are high
compared to emissions benefits.
NESCAUM, 2005-0047-0024, p. 4
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 7-8
EPA's proposal does not include the step-down in increased stringency for malfunctions of
aftertreatment devices as ARB's regulation does for 2013 and 2016 model years. While EPA has
indicated that it will continue to monitor industry's progress towards meeting the ARB thresholds and
could potentially revisit the appropriateness of more stringent thresholds in the future, ARB believes
it is appropriate for EPA to adopt the more stringent thresholds now to provide a clear goal for
industry to design for. EPA could still monitor industry's progress, as ARB does, and as noted above,
could potentially revisit the appropriateness at a later date if changes need to be made. But,
adoption now of thresholds that align with those developed jointly with EPA and adopted by ARB
would provide a consistent target to industry to design to for a single system that meets both
requirements. To date, engine manufacturers have not provided any new technical data indicating
that these thresholds may not be feasible or adjustments may need to be made. In fact, as the first
2007 model year applications are now being certified with various elements of aftertreatment, ARB
and manufacturers are starting to see actual data indicating what thresholds can be achieved now
and in the immediate future and ARB believes the data shows promising results that the
manufacturers are on track.
California Air Resources Board, 2005-0047-0035, p. 3
Agency Response:
As we stated in the preamble to the proposed rule (see 72 FR 3205), we believed that the
proposed thresholds would strike the proper balance between environmental protection, OBD and
various sensor capabilities, and avoidance of repairs whose costs could be high compared to their
emission control results. Since that time, we have learned that certain sensor capabilities have not
advanced as we expected and have, therefore, revised upward some OBD thresholds for the 2010-
2012 model years. One must keep in mind that increasingly stringent OBD thresholds (i.e., OBD
detection at lower emissions levels) may lead to more durable emission controls due to a
manufacturer's desire to avoid the negative impression given their product upon OBD detection.
Such an outcome would result in lower fleetwide emissions while increasing costs to manufacturers.
However, increasingly stringent OBD thresholds may also lead to more OBD detections and more
OBD induced repairs and, perhaps, many OBD induced repairs for malfunctions having little impact
on emissions. Such an outcome would result in lower fleetwide emissions while increasing costs to
both manufacturers and truck owners. Furthermore, increasingly stringent OBD thresholds may
increase the likelihood of false malfunction detections (false positives) which has no impact on
emissions while adversely impacting the perception of OBD. All of these factors must be carefully
balanced and we believe we have done that properly with our final requirements. The California Air
Resources Board is willing to go forward with some uncertainty as regards the ability of
manufacturers to meet the complete set of "stepped down" thresholds in 2013 and 2016. EPA is not
so comfortable moving forward with that level of uncertainty.
I.E World Wide Harmonized OBD (WWH-OBD)
Comments:
-------
Ultimately, EMA's goal is alignment of the EPA and VWVH-OBD requirements such that
engine manufacturers could certify one engine for sale that could meet OBD requirements
internationally. Given the existence of the national and international processes which have not yet
led to a fully aligned approach, EMA recommends that EPA complete this rulemaking and, in a later
rulemaking, evaluate and propose an approach to allow EPA to specify VWVH-OBD as an alternative
OBD solution.
EMA, 2005-0047-0026, p. 45-47
Cummins Inc., 2005-0047-0031, p. 5-6
The allowance of WWH-OBD would be beneficial provided it did not result in a second or
third type of OBD for highway HD engines. A second or third type of OBD would add unnecessary
burdens on engine manufacturers and serve to confuse the service industry for little if any added
value. The commenter also provides a list of OBD elements that EPA should change in the final rule
to align fully with WWH-OBD requirements, listed below, to help support manufacturers' ability to
compete in the international marketplace.
Fewer emission threshold monitors and fewer system monitors (e.g., cooling
system, cold start aids
Less rigidity regarding emission threshold monitor calibrations/lower fidelity
emission threshold monitoring
Fewer malfunctions requiring continuous MIL illumination (i.e., align with the WWH-
OBD "discriminatory" MIL display logic)
Fewer certification steps and certification requirements
No permanent diagnostic trouble codes
No performance monitor ratios
No per-trip readiness reporting
No commanded tests and results
Fewer data stream and freeze frame parameters required
No enforcement provisions
No production evaluation requirements
Fewer emission demonstration tests
EMA, 2005-0047-0026, p. 45-47
WWH OBD as a solution could be beneficial to commercial vehicle market by not having as
many continuous MIL malfunctions. It also could be beneficial to vehicle maintenance industry
because WWH OBD provides a malfunction classification system that could direct them to the
malfunction with the worst emission impact so that they could be fixed first. Additionally it could be
beneficial to manufacturers and maintenance industry if, in addition to the US 50 states, other
regions around the world were following one type of HD OBD.
Cummins Inc., 2005-0047-0026, p. 5-6
Agency Response:
We appreciate the comments received on this topic. As made clear in our proposal, we were
not able to propose the WWH-OBD regulation since it was not yet a final document. It is important
to note that California already had HD OBD requirements in place and we must be cognizant of
harmonization both with California and at the international level. The California OBD requirements
were analogous to the WWH-OBD requirements, but were not identical. At industry's request, we
patterned our proposal after the California regulation and, in the 2010-2012 timeframe, our
requirements were identical to California's. We continue to like certain aspects of the WWH-OBD
regulation (e.g., the malfunction classification system mentioned by Cummins) but we do have
-------
concerns with the lack of specificity regarding some of the VWVH-OBD requirements relative to the
specificity we have in our final requirements. We have not changed anything in our final rule with
respect to this issue but, should this topic come up for further consideration in a future rule, we will
consider the comments summarized here in making any possible future proposal.
I.F Onboard Diagnostics for Diesel Engines used in Nonroad Land-based Equipment
Comments:
TCEQ also supports the consideration of future rulemaking that would require OBD systems
on non-road heavy-duty diesel engines. Non-road heavy-duty vehicles are a source of NOx and PM
and, similar to heavy-duty vehicles, have an extended useful life lasting many years. The
implementation of OBD on non-road heavy-duty engines would assist in the maintenance and repair
of these vehicles and provide a means to ensure these engines continue to emit at low emissions
levels.
Texas Commission on Environmental Quality, 2005-0047-0020, p. 1
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 2
NY State DEC, 2005-0047-0033, p. 1
Considering OBD for nonroad must take into account nonroad market characteristics, such
as:
Low volume applications in highly diversified markets
Broad power ranges with a variety of different test cycles
Severe space, weight and handling constraints, especially in lower power categories
Technically challenging and complex transfers of on-highway technology to nonroad
applications
Consequently, adopting HDOBD requirements for nonroad applications is not regarded as
appropriate and any such considerations by EPA must consider making available flexibility
provisions.
Euromot, 2005-0047-0023, p. 3
EMA appreciates EPA's recognition of the extreme diversity of the nonroad land-based
market. MEA has previously enumerated the issues associated with technology transfer from on-
highway to nonroad applications (see EMA comments on the NRT4 proposal, dated August 20, 2003,
contained in docket # A-2001-28). Moreover, technology transfer to nonroad engines outside the
typical highway HD horsepower range of 200-600 horsepower compounds the level of complexity
which affects manufacturers' ability to transfer OBD from highway to nonroad engines. EMA
believes that proposing the same OBD requirements for nonroad as for highway HD is not
appropriate; the "natural progression" from highway to nonroad, as suggested by EPA, is not
appropriate.
EMA, 2005-0047-0026, p. 43
Caterpillar Inc., 2005-0047-0030, p. 4
Agency Response:
We appreciate the time, effort, and thought put into preparing these comments and will
consider them when we begin development of any possible future nonroad OBD proposals.
-------
I.F.1 What is the baseline nonroad OBD system?
Comments:
Monitoring and diagnostic systems are commonly used today on electronically-controlled
engine platforms. These are typically the higher horsepower engines that currently meet more
stringent emissions standards. These diagnostic systems are used to ensure customer satisfaction,
product performance, and in-use emission compliance. By contrast, lower horsepower engines can
meet emission standards using mechanically-controlled engine platforms and, as such, have few
diagnostic systems today. The first step for EPA is to engage in dialogue and review with industry
over the use case(s) for OBD. Industry and regulators should have a common understanding of
goals and expected environmental results of implementing OBD prior to moving forward.
EMA, 2005-0047-0026, p. 44
Caterpillar Inc., 2005-0047-0030, p. 6
Cummins Inc., 2005-0047-0031, p. 6
Nonroad OBD, if implemented, should not extend beyond the engine sizes and horsepower
range associated with highway engines.
Caterpillar Inc., 2005-0047-0030, p. 6
Agency Response:
We appreciate the time, effort, and thought put into preparing these comments and will
consider them when we begin development of any possible future nonroad OBD proposals.
I.F.2 What is the appropriate level of OBD monitoring for nonroad diesel engines?
Comments:
The diverse range of operating characteristics for nonroad engines and the differences in
typical duty cycles compared to highway engines strongly suggest taking somewhat different
approaches. Therefore, we support the concept of developing nonroad OBD requirements that rely
more heavily on monitoring component performance (e.g., after-treatment devices, sensors, and fuel
systems), compared to monitoring emissions thresholds. However, we support including emissions
threshold approaches for nonroad OBD systems where practical. For example, if certain engine
families are commonly used to operate nonroad equipment under prolonged steady-state conditions,
an emissions threshold approach may be quite practical. In addition, analogous to the drive cycle
options to be made available for OBD monitoring of highway vehicles, it may be possible to identify
common nonroad duty cycles for which an emissions threshold monitoring approach is practical.
NESCAUM, 2005-0047-0024, p. 3-4
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 6
If regulatory authorities proceed with OBD requirements for nonroad diesel engines, those
requirements should be focused on aftertreatment devices only and should be limited to functional
monitoring - without emission threshold monitoring - with no tracking of in-use performance ratios.
In the context of the nonroad market, designing a single OBD approach that would ensure frequent
monitoring events on all possible applications would be almost impossible. Furthermore, EPA or any
other regulatory body should not adopt monitoring requirements for equipment. There is extreme
diversity in nonroad equipment and little, if any, emissions benefit from equipment and drivetrain
diagnostics. Moreover, engine manufacturers do not manufacture equipment and cannot, and
10
-------
should not, be held responsible for diagnostics beyond aftertreatment devices. Nonroad OBD
requirements should be implemented no earlier than 2020 to allow manufacturers to focus
development activity on requirements of other future regulations including highway HDOBD, nonroad
Tier 4, and Tier 3 and 4 locomotive and marine.
EMA, 2005-0047-0026, p. 44-45
Caterpillar Inc., 2005-0047-0030, p. 4-6
Cummins Inc., 2005-0047-0031, p. 6
Unit costs and in-use maintenance costs will be significantly reduced if the OBD monitoring
requirements for nonroad engines essentially parallel those for on-highway engines. TMA
recommends that nonroad OBD requirements minimize potential mechanical differences between
on-highway and nonroad engines.
Truck Manufacturers Association, 2005-0047-0028, p. 2
Agency Response:
We appreciate the time, effort, and thought put into preparing these comments and will
consider them when we begin development of any possible future nonroad OBD proposals.
I.F.3 What should the OBD standardization features be?
Comments:
Current practices in the nonroad service network are adequate to support needed scan tool
interface and product serviceability. The extremely high cost of nonroad equipment downtime has
already driven the market to robust diagnostics, accurate troubleshooting and a service
infrastructure focused on minimal customer downtime. Nevertheless, the global nature of the
nonroad industry would merit (in a perfect world) from a single communication protocol. However,
there are more than one protocol, data link and connector used today in the nonroad market with
electronically controlled engine platforms. Whatever the ultimate design choice is, it is desirable that
a link does exist conforming to a recognized standard and that the connector be accessible to a
service technician. It also is desirable that the use of a dedicated MIL be optional. EMA anticipates
that many of these standardization issues will be worked out through the ongoing SAE and ISO
standardization and/or the WWH-OBD processes.
EMA, 2005-0047-0026, p. 45
Cummins believes that a standardized interface would be a benefit (connector, protocol,
critical subset of services) for the nonroad industry. Allowing SAE J1939 and ISO 15765-4 is
appropriate.
Cummins Inc., 2005-0047-0031, p. 6
On-highway engine data link parameters number in the hundreds. However, nonroad
parameters number in the thousands. While Caterpillar does support J1939 on many products, there
is a need to continue use of proprietary protocols. Some nonroad equipment is manufactured in
extremely small volumes, and some have annual sales volumes of 10 or less. These products are
very complex, have numerous control modules, and require specialized communication needs. From
a cost and development time perspective, it is impractical to obtain standardized or public approval
for communication parameters required for these products because much of the information is so
specific to a product and unlikely to be used by other manufacturers. These factors essentially drive
11
-------
the industry to require the use of proprietary protocols. Caterpillar is requesting the EPA consider the
need for proprietary protocols and their co-existence with standardized protocols.
Caterpillar Inc., 2005-0047-0030, p. 7
The Department also supports requirements for wireless communication protocols for diesel
non-road land based equipment so that onboard information can be read by a universal scan tool or
other offboard device. In order for an engine emissions problem to be effectively diagnosed and
repaired, there must be the ability to download stored onboard information. Common communication
protocols that are readable by universal scan tools are extremely important. Universal scan protocols
will enable equipment owners and service providers to diagnose engine and emission control system
problems for a wide variety of equipment without the requirement of purchasing multiple specialized
scan tools. This capability especially becomes important as equipment ages, becomes more prone
to malfunction, and manufacturer support diminishes as newer products are introduced.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 6
Agency Response:
We appreciate the time, effort, and thought put into preparing these comments and will
consider them when we begin development of any possible future nonroad OBD proposals.
I.F.4 What are the prospects and/or desires for international harmonization of nonroad OBD?
Comments:
EMA believes that international harmonization of nonroad OBD is essential. EMA will actively
participate in the development of a nonroad global technical regulation under the direction of UNECE
WP-29 World Forum of Global Harmonization of Vehicle Regulations. The government/industry
experience gained during the development and ultimate approval of the WWH-OBD (on-road) global
technical regulation will prove invaluable to the nonroad OBD development process. EPA is
encouraged to participate in the development of a nonroad world harmonized OBD global technical
requirement rather than promulgate a separate U.S. regulation. However, if a WWH effort leads to
allowing different solutions for any region that adopts a newly developed WWH Nonroad OBD
regulation then that would not be desired because it would mean there are possibly more
certification combinations to satisfy than there are today.
EMA, 2005-0047-0026, p. 45
Caterpillar Inc., 2005-0047-0030, p. 7
Cummins Inc., 2005-0047-0031, p. 6-7
Agency Response:
We appreciate the time, effort, and thought put into preparing these comments and will
consider them when we begin development of any possible future nonroad OBD proposals.
II. What are the Proposed OBD Requirements and When Would They be Implemented?
II.A General OBD System Requirements
12
-------
What We Proposed:
This section contained the general proposal that engines be equipped with an OBD system,
that the system contain a malfunction indicator light (MIL) to alert the driver of a problem, and also
store diagnostic trouble codes to assist repair technicians when making repairs. We also presented
the concept of monitoring conditions and how they differed for different OBD monitors. Lastly, this
section presented our proposal for addressing infrequent regeneration adjustment factors when
determining OBD thresholds.
II.A.1 The OBD System
Comments:
Since heavy-duty vehicles, especially diesel vehicles, have an extended useful life often
lasting hundreds of thousands of miles, the need to detect emissions related problems throughout
the operational period is important in reducing nitrogen oxides (NOx) and particulate matter (PM)
emissions.
Texas Commission on Environmental Quality, 2005-0047-0020, p. 1
NESCAUM, 2005-0047-0024, p. 3
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 8
National Automobile Dealers Association, 2005-0047-0034, p. 3
Agency Response:
We agree with these comments and require that the OBD system remain functional
according to its original design throughout the lifetime of the engine.
II.A.2 Malfunction Indicator Light (MIL) and Diagnostic Trouble Codes (DTC)
Comments:
Commenters stating an endorsement of the ISO engine symbol (F01) as the OBD MIL:
NESCAUM, 2005-0047-0024, p. 7
American Trucking Association, 2005-0047-0029, p. 2
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 8
Cummins Inc., 2005-0047-0031, p. 10-11
NY State DEC, 2005-0047-0033, p. 1
California Air Resources Board, 2005-0047-0035, p. 1-2
Commenters stating an endorsement of the ISO emissions symbol (F22) as the OBD MIL:
Caterpillar Inc., 2005-0047-0030, p. 4
Commenters stating an endorsement for having one MIL for all emission related
malfunctions:
American Trucking Association, 2005-0047-0029, p. 2
Truck Manufacturers Association, 2005-0047-0028, p. 2
13
-------
The F01 symbol has been used by HD highway vehicles for many years and is understood to
have a meaning different than what the OBD MIL intends to convey. The F22 symbol would better
communicate the need for emissions related repair and would reduce confusion for operators of HD
highway vehicles. Since EPA prohibits the use of the OBD MIL for non-OBD purposes, two amber
warning lamps will be expected on HD vehicles - one OBD lamp and one non-OBD lamp. Therefore,
it is best to continue to use the F01 symbol for non-OBD events (as today) and the F22 for the OBD
MIL.
Caterpillar Inc., 2005-0047-0030, p. 4
ISO warning light symbols should be configured to be easily understood by the equipment
operator. In this regard, we support using the engine symbol as proposed by EPA. The symbol
preferred by the Department of Transportation is confusing and therefore would be less likely to
properly inform the operator of an engine or emissions control system-related problem.
NESCAUM, 2005-0047-0024, p. 7
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 8
The symbol selected for the OBD MIL must be acceptable for use across all North American
markets. There are advantages and disadvantages to either F22 or F01 ISO symbols - TMA does
not have a specific recommendation on this selection.
Truck Manufacturers Association, 2005-0047-0028, p. 2
The proposed MIL symbol, the ISO F01 engine symbol, is already widely used as a check
engine light (with the word "check" written across it). Insofar as the two symbols are similar, their
meanings blur. Freightliner appreciates that EPA has tailored its proposal to match CARB where
possible and appreciates the use of the same symbol. However, EPA should clarify the meaning of
the MIL symbol and, at the same time, create uniformity across the HD highway market by either
choosing a symbol not already in use or by more precisely specifying the symbol (e.g., requiring that
the ISO engine symbol be accompanied by the letters "OBD"). If EPA chooses to deviate in its
symbol from that described in the CARB regulation, EPA should require that CARB change their
requirement to maintain uniformity (via the CARB waiver process currently being considered, see 72
FR 8726).
Freightliner, 2005-0047-0025, p. 2
Freightliner supports the proposed requirement for a key-on/engine-off MIL bulb check as
short as five seconds. Currently, all of Freightliner's other bulb checks are three to five seconds.
CARB requires a 15-20 second bulb check (see 13 CCR 1971.1(d)(2.1.2)). This long bulb check is
likely to generate confusion amongst drivers. EPA should not grant a CARB waiver until CARB
eliminates their potentially confusing bulb check requirement.
Freightliner, 2005-0047-0025, p. 2
Cummins Inc., 2005-0047-0031, p. 11
EPA's proposed rule deviates from the CARB adopted heavy-duty, light-duty, and medium-
duty OBD rules in that it requires the MIL to only illuminate fora minimum of five seconds in the key-
on, engine off position as a bulb check feature. CARB regulations mandate a minimum of 10
seconds and this was a result of discussions primarily with inspection and maintenance program
managers and inspectors that were having difficulty discerning the MIL from the other warning lights
that all illuminate during a bulb check. Ten seconds was chosen as a reasonable value to ensure
inspectors have sufficient opportunity to look for and locate the MIL during the bulb check with
14
-------
minimal risk of falsely failing the vehicle because it extinguished too quickly. Five seconds was
considered and rejected as insufficient given the number of warning lights on today's vehicles.
California Air Resources Board, 2005-0047-0035, p. 2
Requirements to monitor datalink telltale systems and light emitting diode (LED) MIL lamps -
both of which have a high degree of reliability - for proper circuit function and lamp illumination is
difficult and costly and may prompt manufacturers to opt for older-style, inherently less reliable
lamps and driver interface systems than LEDs and datalink systems. Freightliner requests that EPA
omit the requirement for MIL circuit monitoring and that, during CARB's waiver process, EPA require
that CARB do the same.
Freightliner, 2005-0047-0025, p. 2
Cummins Inc., 2005-0047-0031, p. 9
Integrated dash panels should be allowed to provide the MIL function, and the MIL
functionality and wait-to-start lamp functionality requirements should be written such that the
applicable suppliers be burdened with satisfying them - i.e., the engine manufacturer should not be
the responsible party for an item under the control of the vehicle manufacturer or other member of
the heavy truck industry.
Cummins Inc., 2005-0047-0031, p. 9-10, p. 12
LED technology MIL and wait-to-start lamps should be exempted from comprehensive
component monitoring requirements.
Truck Manufacturers Association, 2005-0047-0028, p. 3
EPA should allow the existing engine Amber Warning Lamp and engine Red Stop Lamp
along with the newly proposed OBD MIL. Implementation of the OBD MIL should be aligned with
CARB (i.e., mandatory use of an OBD MIL should begin in 2013).
Cummins Inc., 2005-0047-0031, p. 10
We support the concept of defining a driving cycle according to a specified period of
continuous engine-on operation. This will help to ensure that OBD monitors that run only once per
driving cycle will operate frequently enough to detect system malfunctions and that sustained engine
operation does not effectively turn off these monitors. We support the concept of requiring certain
monitors to run continuously throughout the driving cycle, including certain threshold monitors (e.g.,
fuel system monitor) and most circuit continuity monitors.
NESCAUM, 2005-0047-0024, p. 4
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 8
EPA should consider shortening the driving cycle to one hour rather than the proposed four
hour maximum. More frequent monitoring is preferable to less frequent longer intervals that may
miss triggering engine malfunction codes in extended drive cycles.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 8
The definition of the term "pending DTC" is inconsistent with the definitions used in the
CARB light and medium-duty OBDII regulation, the CARB HDOBD regulation and the WWH-OBD
regulation. The EPA definition states that a pending DTC is a "diagnostic trouble code stored upon
15
-------
the detection of a potential malfunction." A potential malfunction is then defined as meaning that
"conditions have been detected that meet the OBD malfunction criteria but for which more drive
cycles are allowed to provide further evaluation prior to confirming that a malfunction exists." The
commenter recommends a definition that clarifies that a pending DTC be stored only for conditions
detected on the current or most recent drive cycle. The commenter also recommends removing the
term "potential DTC" from the regulation and replacing it with the recommended "pending" DTC term.
Cummins Inc., 2005-0047-0031, p. 4-5
Caterpillar strongly supports EPA's definition of "pending" DTCs. Caterpillar believes that the
original definition of "pending" which would allow DTCs to be simultaneously "pending" and
"confirmed" is not logical and would cause confusion for the HD service industry.
Caterpillar Inc., 2005-0047-0030, p. 4
Agency Response:
Regarding the MIL symbol (F01 versus F22), only Caterpillar supported F22 so we have
decided to remain consistent with our proposal and require the F01 symbol. Other commenters
argued that the F01 symbol has been used in the past and, as such, the OBD MIL should not use
the F01 symbol and should use something different. However, given the lack of support for the F22
symbol and knowing of no real viable alternatives to the F01 or F22 symbol, we believe that the F01
symbol is the best choice and have faith that engine manufacturers will be able to educate their
customers on the new significance of that symbol should drivers see it illuminated on their dash.
Regarding the illumination time upon key-on, CARB argues that a minimum of 10 seconds
should be stipulated while others argue for the 5 second given current practice of 3-5 seconds. A
system designed to meet the 10 second CARB minimum would clearly meet our 5 second minumum.
We believe that inspectors, should HDOBD one day become part of inspection and maintenance
programs, can be trained if necessary to find the illuminating MIL given a 5 second window and,
therefore, we consider our proposed 5 second interval acceptable
Regarding LED based wait-to-start lamps and MIL lamps, we agree with commenters and
have changed our final rule such that these lamps need not conduct circuit checks. We will monitor
the situation in-use to ensure that the reliability claimed by industry is reality. Note that a message
must be present via the datastream indicating any situation where the MIL is being commanded 'on'
by the OBD system. Therefore, anyone can easily determine that a lack of MIL illumination at the
dashboard when the MIL is being commanded 'on' suggests that there is a problem with the MIL
circuit.
Regarding timing of the mandatory MIL as Cummins commented, we agree with the
comment and require that existing lamps be used in the 2010-2012 timeframe and the mandatory
lamp be used beginning in 2013. This remains as we proposed.
Regarding the drive cycle related comments and expressed desires for more frequent
monitoring, we understand these comments but have some reservations which we believe outweigh
the concerns expressed by commenters. Please refer to our response in section II.A.3 for more
detail.
Regarding the comment from Cummins, and others, on the definition of pending DTC, the
intent behind the definition put forth in the proposal was not to generate an entirely new definition of
an existing term. As we understand, there are slightly different methods to storing and deleting
pending DTCs depending on which communication protocol is used (i.e., SAE versus ISO protocols).
16
-------
Our intent was to define the term in a generic enough way so as to convey its general meaning, but
leave it to industry to determine how to best inform their repair technicians about the status of
malfunctions and potential malfunctions (i.e., leave it to the standards setting bodies to standardize
the implementation). As a result of comments received, we believed that our intent, while good, had
served to generate confusion and leave open the possibility that someone would use, say, the SAE
standard but store/erase pending DTCs according to the ISO standard. In subsequent discussions
with EMA, it was decided to leave things generic and up to the manufacturer how best to deal with
pending DTCs. We consider our primary concern to be the MIL-on DTC so are willing to structure
the requirement for pending DTCs as industry prefers.3
As for the requirement to define a pending DTC as a DTC stored only for events detected on
the current or most recent drive cycle, we do not believe this is necessary. The CARB definition is a
"diagnostic trouble code stored upon the initial detection of a malfunction (e.g., typically on a single
driving cycle) prior to illumination of the MIL..." We believe that the erasure clarification mentioned
above along with the proposed definition of pending DTC makes the EPA regulation consistent in
every way with the CARB HDOBD regulation.
II.A.3 Monitoring Conditions
Comments:
We support the general monitoring conditions as proposed. Particularly, we support the
concepts that:
monitors should run during conditions that are technically necessary to ensure robust
detection of malfunctions, avoiding false passes and false indications of malfunction;
enabling criteria should ensure monitoring will occur during normal vehicle operation;
monitoring should occur during at least one FTP transient cycle or SET; and
monitors will run at least once per driving cycle in which the applicable monitoring
conditions are met.
In regard to the 4th general monitoring condition above, we have taken note that throughout
the proposal for various monitors, "monitoring must occur every time the monitoring conditions are
met during the driving cycle in lieu of once per driving cycle as required for most monitors." Among
those for which monitoring is required only once per drive cycle are so-called "major monitors (e.g.,
catalyst, EGR, CDPF, other diesel aftertreatment devices)". The proposal is unclear as to why some
components are monitored only once per drive cycle, whereas others apparently will be monitored
whenever the applicable conditions are met. We urge EPA generally to require monitors to operate
whenever the applicable conditions are met unless there is some compelling reason to monitor only
once per driving cycle.
NESCAUM, 2005-0047-0024, p. 5
In addition to the general monitoring conditions, NESCAUM supports requirements for in-use
performance tracking for the 11 listed system components (§ 86.010-18(d)(1), FR page 3292). We
take note of the fact that initially EPA is proposing a minimum in-use performance ratio of 0.100 for
all monitors specifically required to track in-use performance (i.e., monitors must make valid
diagnostic decisions during 10 percent of the vehicles trips) and that this ratio may be revised
downward, following initial years of implementation as EPA works with industry to gather data on in-
use performance ratios. We further take note of the fact that 10 percent is a minimum, subject to first
meeting the general monitoring conditions. For example, if a particular monitor is capable of
ensuring robust detection of malfunctions during 50 percent of vehicle trips, then the higher
3 See memorandum to Docket ID No. EPA-HQ-OAR-2005-0047 from Todd Sherwood, "Meetings with the
Engine Manufacturers Association," document ID # EPA-HQ-OAR-2005-0047-0053.
17
-------
percentage requirement would prevail. Further, we assume that any decision to revise the in-use
performance ratio for any particular monitor will require further revision of this regulation, so will be
subject to a new public comment process.
NESCAUM, 2005-0047-0024, p. 5
EPA's preamble indicates that EPA is committed to working with industry to collect and
reassess in-use data but it appears limited to the extent that it would only consider revising the
minimum ratios "lower as appropriate". CARB's stated intent for the minimum ratios of 0.100 is to
set an interim ratio that can then be raised or lowered, as appropriate. Based on the statistical
analysis done when developing the in-use ratios for light- and medium-duty vehicles, a ratio of
0 .100 reflects fairly infrequent in-use monitoring for a substantial portion of the fleet and it would not
be considered acceptable by ARB for long term use. For the particulate matter filter alone, this ratio
could translate to as little as three monitoring events per year for a typical medium-duty application.
California Air Resources Board, 2005-0047-0035, p. 2
Agency Response:
We understand NESCAUM's concern that major monitors are only required to run once per
drive cycle. Major monitors usually override the emission control system to in order to detect a
failure of the component they are monitoring. Increased emissions are created during monitoring
override, so it is not desirable to run these monitors from both the stand point of the manufacturer
who is trying to meet the emissions standard and the EPA who wishes to have vehicles create the
fewest emissions possible while still determining the robustness of the emissions control system.
We understand both NESCAUM's and CARB's concerns with having the initial in-use
performance ratio set to 0.100. This introductory ratio does not take effect until 2013. Data will be
gathered by the manufacturer and submitted to EPA as specified in the final regulation at §86.010-
18(j)(3). The minimum ratio may be increased or decreased based on real world data.
Manufacturers should not be setup to fail based on this new requirement. It is EPA's opinion that
starting with an in-use performance ratio of 0.100 is fair and reasonable, and may be increased
when the data available to justify it.
II.A.4 Determining the Proper OBD Malfunction Criteria
Comments:
EMA does not support the inclusion of infrequent regeneration adjustment factors (IRAFs),
as proposed in §86.010-18(f)(2), for numerous reasons, including feasibility and stringency concerns,
the workload burden, and because further analysis is necessary before it can be determined whether
and how IRAFs should be applied to OBD monitors.
Applying IRAFs to OBD thresholds increases the stringency of the OBD standards and
makes them infeasible. When designing engines to meet emissions standards, manufacturers must
leave "headroom" or margin to account for variability and other factors that may increase engine or
OBD emissions in a given situation. Maintaining that headroom is essential to manufacturers' ability
to comply with the thresholds. Adding IRAFs - whether they are emission certification adjustment
factors or uniquely-calculated adjustment factors - erodes or eliminates that margin, assuring that
the OBD threshold standards would not be technologically feasible. Adding IRAFs is unnecessary
since regeneration emissions are already accounted for in the emissions standards.
Requiring IRAFs for OBD creates an unreasonable workload. EPA has no basis for adopting
a requirement over which EPA has expressed substantial concern at 72 FR 3211 where EPA
expresses concern that "manufacturers may find themselves in a difficult iterative process calibrating
18
-------
such monitors that, in the end, will not be correspondingly more effective." The commenterthen
explains that, assuming there are 13 OBD threshold monitors and two regeneration devices (e.g.,
DPF and NOx adsorber), a manufacturer would have to determine unique upward adjustment factors
(UAF) and downward adjustment factors (DAF) for 26 unique (i.e., OBD specific) IRAFs. The
process for conducting this effort and the required testing is then described, including descriptions of
the difficulty of generating "perfect threshold parts" and the time and testing burden needed to do so.
The commenterthen states that all the testing on all the emission threshold monitors for all
applicable test cycles and each infrequent regeneration device and engines/aftertreatment translates
to an enormous amount of engineering resources, expense, test cell time, and leatime required to
obtain the data necessary to develop unique IRAFs for each OBD threshold monitor. In fact, engine
manufacturers estimate that the proposed requirements would increase their OBD threshold
development work far more than double that which manufacturers currently predict for achieving
threshold compliance without including IRAFs. Having had no experience with determining the
impacts of regeneration events on OBD emissions and developing appropriate adjustment factors,
engine manufacturers do not believe that engineering analysis is sufficient for fulfilling these
requirements.
EMA, 2005-0047-0026, p. 25-29
Cummins Inc., 2005-0047-0031, p. 4
Uncertainty exists around the process for determining IRAFs for emissions certification for
2010, let alone for OBD. For 2007 emissions IRAFs, there were multiple, ever-changing guidance
documents issued from EPA, as late as the second half of 2006 when manufacturers were in the
middle of certifying their products. The changing guidance resulted in re-development work and
wasted testing. EPA and the manufacturers will revisit the process for determining IRAFs for 2010
emissions certification. The lack of solid guidance in this area makes the inclusion of OBD IRAFs,
which will require even more complex process development, infeasible for 2010.
Cummins Inc., 2005-0047-0031, p. 4
The costs of adding IRAFs far outweigh the benefits. Given the high cost and the minimal
anticipated benefits from requiring that IRAFs be considered, EPA should not adopt IRAF
requirements for HDOBD.
EMA, 2005-0047-0026, p. 29-30
Using emissions certification IRAFs for OBD certification, as suggested by EPA as perhaps
being the best approach, is technically incorrect. Those IRAFs are based on baseline engine
emissions not on emissions using OBD threshold parts. EPA should not adopt a compromise
position that is not technically correct - it is bad regulatory policy and wrong from an engineering
perspective. Doing so would not advance air quality. EPA should not adopt requirements to adjust
malfunction thresholds for regeneration emissions.
EMA, 2005-0047-0026, p. 30
Under the proposed rule, it is possible that the MIL may be triggered by a component that is
subject to periodic maintenance checks and operating within its expected range of function. In such
cases, premature replacement or repair of engine components may be required to prevent the MIL
from illuminating. Also, trucks may be forced out of service for maintenance for malfunctions not
affecting emissions, which is beyond the scope and intent of this proposed rule. TRALA also
encourages the consideration of an OBD system that helps the driver understand whether or not a
malfunction is emission-related and requires the truck to be pulled out of productive service for repair.
19
-------
Truck Renting and Leasing Association, 2005-0047-0032, p. 2
EPA's preamble opens the discussion for consideration of alternatives to including the
regeneration emissions in determining malfunction criteria. As EPA certainly knows from certification
test results submitted by the manufacturer, the inclusion of regeneration emissions and the
frequency with which they occur is critical in determining the average emissions from the engine and,
as a result, compliance with the tailpipe emission standards. The same principle holds true for
regeneration emissions and frequency when determining malfunction criteria. To the extent that
deterioration and malfunction of a component increases the frequency of a regeneration event or
increases the emissions of the event itself, that increase in emissions should be considered in
establishing the threshold for the monitored component. In fact, for some components such as the
oxidation catalyst where the primary purpose is for the regeneration event itself, failure to consider
the regeneration events would essentially exclude the component from monitoring and let the
increase in emissions go unchecked in-use.
California Air Resources Board, 2005-0047-0035, p. 2-3
Agency Response:
This issue remains a difficult issue to resolve. We agree with the CARB comment that
regeneration emissions are clearly part of the emission standard. In fact, we consider those
emissions to be a very important part of our emission standards hence our inclusion of them in the
standards. What is less clear is whether those emissions should be part of the OBD threshold.
Further, we agree with industry comments that the burden imposed by including regeneration
emissions in the OBD threshold may well be very large (as we noted in our proposal at 32 FR 3211).
In the end, we simply do not believe that the burden imposed is properly aligned with the benefit of
including the regeneration emissions in the OBD threshold. Therefore, we have changed our final
provisions in a small way by stating, in §86.010-18(f)(2), that the manufacturer need not adjust
emissions to reflect the regeneration emissions when determining OBD thresholds. Note also that
we believe our new DOC monitoring requirements (i.e., detecting the inability to achieve a 100
degree C temperature change or achieve the regeneration temperature, etc.) will serve the purpose
of detecting malfunctions associated with the DPF regeneration system which is the primary purpose
of including regeneration emissions in the OBD threshold, at least where the NMHC catalyst is
concerned. Should CARB continue with their inclusion of the regeneration emissions, we will closely
monitor things as they develop to ensure that we are not missing the identification of malfunctions
causing significant emission impacts and would consider including the regeneration emissions in a
possible future OBD proposal.
II.B Monitoring Requirements and Timelines for Diesel-Fueled/Compression-lgnition Engines
What We Proposed:
The proposal contained the following table showing our proposed thresholds.
20
-------
Thresholds for>14,000 pound Certification (g/bhp-hr) - EPA Proposed
Component/Monitor
NMHC catalyst system
NOx catalyst system
DPF system
Air-fuel ratio sensors upstream
Air-fuel ratio sensors downstream
NOx sensors
"Other monitors" with emissions
thresholds (see section II.B)
MY
2010-2012
2013+
2010+
2010-2012
2013+
2010-2012
2013+
2010-2012
2013+
2010+
2010-2012
2013+
NMHC
2.5x
2x
~
2.5x
2x
2.5x
2x
2.5x
2x
~
2.5x
2x
CO
~
~
~
2.5x
2x
~
~
2.5x
2x
NOx
~
+0.3
~
+0.3
+0.3
+0.3
+0.3
+0.3
+0.3
+0.3
PM
~
~
0.05/+0.04
0.05/+0.04
0.03/+0.02
0.03/+0.02
0.05/+0.04
0.05/+0.04
0.05/+0.04
0.03/+0.02
0.03/+0.02
Comments:
EMA provided the following table of suggested OBD malfunction thresholds.
EMA, 2005-0047-0026, Appendix B, p. 11
EMA Proposed Thresholds for >14,000 pound Certification (g/bhp-hr)
Component/Monitor
NMHC catalyst system
NOx catalyst system
DPF system
Air-fuel ratio sensors upstream
Air-fuel ratio sensors downstream
NOx sensors
"Other monitors" with emissions
thresholds (see section II.B)
Misfire
Note: Boldfaced text and entries of "--"
MY
2010-2012
2013+
2010+
2010-2012
2013+
2010-2012
2013+
2010-2012
2013+
2010+
2010-2012
2013+
2010+
NMHC
Functional Check
Only
Functional Check
Only
—
—
2.5x
2x
2.5x
2x
2.5x
2x
No
CO
2.5x
2x
2.5x
2x
NOx
+0.6
+0.3
+0.3
+0.3
+0.3
+0.6
+0.3
+0.3
PM
0.10/+0.09
0.10/+0.09
0.03/+0.02
0.03/+0.02
0.05/+0.04
0.05/+0.04
—
0.03/+0.02
0.03/+0.02
Requirement
denote differences from the EPA proposal.
Agency Response:
We have changed some thresholds for the 2010-2012 model years, many of which align our
final thresholds with the EMA suggestions. Our final thresholds are shown below.
21
-------
Thresholds for>14,000 pound Certification (g/bhp-hr) - EPA Final
Component/Monitor
MY
NMHC CO NOx PM
NOx catalyst system
DPF system
Air-fuel ratio sensors
Air-fuel ratio sensors
NOx sensors
"Other monitors" with
upstream
downstream
emissions
thresholds (see section II. B)
2010-2012
2013+
2010-2012
2013+
2010-2012
2013+
2010-2012
2013+
2010-2012
2013+
2010-2012
2013+
2.5x
2x
2.5x
2x
2.5x
2x
2.5x
2x
+0.6
+0.3
2.5x +0.3
2x +0.3
+0.3
+0.3
+0.6
+0.3
2.5x +0.3
2x +0.3
0.05/+0.04
0.05/+0.04
0.03/+0.02
0.03/+0.02
0.05/+0.04
0.05/+0.04
0.05/+0.04
0.05/+0.04
0.03/+0.02
0.03/+0.02
Note: See preamble Table II.B-1 and §86.010-18(g), Table 1.
With respect to the NMHC catalyst monitoring, we have eliminated the thresholds and are
requiring functional checks only. Please refer to our responses under sections II.B.5 and III.A.5
below for more detail. We have aligned our NOx catalyst thresholds with EMA in the 2010
timeframe. Please refer to our responses under sections II.B.7 and III.A.7 below for more detail. We
have not revised our PM thresholds for DPF monitoring relative to our proposal, but we have added
an optional monitoring requirement for the 2010-2012 model years. This optional requirement
provides an effective monitoring approach using available technology and thereby satisfies the
feasibility concerns raised by EMA. Consistent with EMA, we have eliminated our NMHC thresholds
associated with the NMHC converting function of the DPF. We have retained NMHC thresholds
associated with the DPF in the context of the regeneration frequency monitoring requirement.
Please refer to our responses under sections II.B.8 and III.A.8 below, which pertain to DPF
monitoring, for more detail. Regarding NOx sensors, we have aligned our NOx thresholds with EMA
in the 2010 timeframe but have retained our PM thresholds unlike suggested by EMA. Please refer
to our responses under sections II.B.9 and III.A.9 below for more detail. As for misfire, we have not
followed the EMA suggestion to eliminate the requirement. Please refer to our responses under
sections II.B.2 and III.A.2 below for more detail.
II.B.1 Fuel System Monitoring
What We Proposed:
We proposed that fuel system malfunctions related to injection pressure, injection timing,
injection quantity, and feedback control be individually detected prior to emissions exceeding the
thresholds for "other monitors." Further, we proposed that pressure and feedback related
malfunctions be monitored continuously and that quantity and timing related malfunctions be
monitored once per trip.
Comments:
For fuel injection quantity, EPA should eliminate thresholds and require only functionality or
circuit continuity testing. For fuel injection timing, EPA should eliminate thresholds and require only
functionality or circuit continuity testing and should limit that requirement to common-rail systems.
22
-------
In lieu of eliminating thresholds for injection timing and quantity, EPA should limit the scope
of threshold monitoring to those failure modes that are detectable during idle and deceleration fuel
cutoff.
Threshold monitoring for fuel injection pressure should be eliminated for fuel systems that do
not use common rail fuel injection because only common rail systems directly measure fuel injection
pressure. Threshold monitoring of fuel injection timing should be required only on non-common rail
fuel systems.
EMA, 2005-0047-0026, p. 14-15
Caterpillar Inc., 2005-0047-0030, p. 3
Agency Response:
We have made considerable changes to the fuel system monitoring requirements for diesels.
Please refer to our response in section III.A.1 for details of those changes.
II.B.2 Engine Misfire Monitoring
What We Proposed:
We proposed that, for 2010-2012, a continuous engine misfire be detected during engine idle.
For 2013 and later, we proposed that engines equipped with combustion sensors monitor
continuously for misfire during the full operating range and detect a malfunction prior to emissions
exceeding the thresholds for "other monitors."
Comments:
EPA's proposal for diesel misfire detection is without justification and imposes significant
costs without any benefits. Diesel misfire detection requirements should be eliminated from the final
rule.
EMA, 2005-0047-0026, p. 16
National Automobile Dealers Association, 2005-0047-0034, p. 3
If EPA's primary concern with respect to misfire is with monitoring of homogeneous charge
compression ignition (HCCI) engine technology, then misfire monitoring requirements should be
limited only to the use of such technology and should not be required of other technologies where
misfire monitoring is redundant, costly and unnecessary.
EMA, 2005-0047-0026, p. 17-18
Agency Response:
We disagree with EMA's concerns that diesel misfire detection adds cost with no benefit.
We believe that costs for detecting misfire on engines without combustion sensors are small and
have been done by manufacturers for several years to comply with California's OBDII regulation on
CARB's medium-duty applications. As for engines equipped with combustion sensors, again we
believe that costs are minimal since the combustion sensing technology can easily detect lack of
combustion (i.e., misfire). We are concerned with misfire on all engines not just engines with HCCI
technology as mischaracterized by EMA. The mention of HCCI in our regulation is simply an
example of systems expected to employ combustion sensors. Misfire on a diesel (just like misfire on
a gasoline engine) is a significant emissions concern, and may damage the after treatment system.
23
-------
Elaboration on the condition and frequency of misfire is not necessary, since different amounts of
misfire at different speeds and loads will either cause an emissions problem and/or damage the after
treatment system.
We have more discussion of misfire monitoring in our response to issue III.A.2.
II.B.3 Exhaust Gas Recirculation (EGR) System Monitoring
What We Proposed:
We proposed that malfunctions of the EGR system related to low flow, high flow, slow
response, feedback control, and cooler performance be detected prior to emissions exceeding the
thresholds for "other monitors." Further, we proposed that flow and feedback related malfunctions
be monitored continuously, response related malfunctions be monitored whenever conditions were
met, and that cooler malfunctions be monitored once per trip.
Comments:
Continuous monitoring of EGR flow (low and high) is neither necessary nor feasible. It is
reasonable to limit operation of the monitors to those areas that yield the best separation between
normal and malfunctioning systems as long as a minimum performance ratio is met. The EGR
monitoring requirement should be defined as a requirement to run "whenever the entry conditions
are met" rather than being defined as "continuous monitoring" as proposed. This would result in
monitoring strategies designed such that they detect the failure modes that exhibit an effect on
emissions throughout the engine operating range, which extends beyond the entry conditions of the
monitor.
EMA, 2005-0047-0026, p. 20
Agency Response:
We disagree with the commenter that continuous monitoring of the EGR system is not
necessary. EPA believes that EGR system continuous monitoring is necessary for specific monitors
on complex systems and it can be accomplished by a combination of both system and
comprehensive component monitoring. By reducing the monitoring requirements to less than
continuous, (i.e. once per trip), malfunctions under some operating conditions may not be detected if
those operating conditions happen to differ from the monitoring conditions. Because EGR operates
under almost all operating conditions, and because the possibility exists for malfunctions to manifest
under only specific operating conditions, all operating conditions should be monitored. However, we
believe that there may have been some confusion as regards our proposed "continuous" monitoring
requirements. We have revised slightly our proposed EGR monitoring conditions to provide greater
clarity to the "continuous" monitoring requirement and believe that this change addresses the
primary concern raised by the comment. Please refer to our response under section III.A.3 below for
a more detailed discussion on this issue and the revision we have made for the final rule.
II.B.4 Turbo Boost Control System Monitoring
What We Proposed:
We proposed that malfunctions of the boost control system related to underboost, overboost,
variable geometry slow response, feedback control, and undercooling be detected prior to emissions
exceeding the thresholds for "other monitors." Further, we proposed that underboost, overboost,
24
-------
and feedback related malfunctions be monitored continuously, that slow response related
malfunctions be monitored whenever conditions were met, and that undercooling related
malfunctions be monitored once per trip.
Comments:
Boost monitoring should not be done continuously, but should be done whenever entry
conditions are met.
EMA, 2005-0047-0026, p. 20
Agency Response:
As noted above under II.B.3 for EGR monitoring, we disagree with the commenterthat
continuous monitoring of the turbo boost system is not necessary. EPA believes that turbo boost
system continuous monitoring is necessary for specific monitors on complex systems and it can be
accomplished by a combination of both system and comprehensive component monitoring. By
reducing the monitoring requirements to less than continuous, (i.e. once per trip), malfunctions under
some operating conditions may not be detected if those operating conditions happen to differ from
the monitoring conditions. Because EGR operates under almost all operating conditions, and
because the possibility exists for malfunctions to manifest under only specific operating conditions,
all operating conditions should be monitored. However, we believe that there may have been some
confusion as regards our proposed "continuous" monitoring requirements. We have added a new
provision to the turbo boost monitoring conditions that is analogous to the provision noted above for
EGR monitoring conditions. This new provision, as with EGR, provides greater clarity to the
"continuous" monitoring requirement, and we believe it addresses the primary concern raised by the
comment. Please refer to our response under section III.A.4 below for a more detailed discussion
on this issue and the new provision we have added for the final rule.
II.B.5 Non-Methane Hydrocarbon (NMHC) Converting Catalyst Monitoring
What We Proposed:
We proposed that malfunctions related to NMHC conversion efficiency be detected prior to
emissions exceeding the thresholds for "NMHC catalyst." We also proposed that, should the NMHC
converting catalyst be used to assist other aftertreatment devices, that malfunctions be detected if
that assistance is no longer occurring. Further, we proposed that conversion efficiency and
aftertreatment assistance be monitoring once per trip.
Comments:
EPA should not adopt an emissions threshold monitoring requirement for DOCs and should,
instead, require only functional monitoring.
EMA, 2005-0047-0026, p. 14
Agency Response:
We agree with this comment and have eliminated our emission threshold for NMHC catalyst
monitoring. As discussed in more detail in section III.A.5, we know of no good means to monitor the
NMHC conversion function. The exotherm generated during normal NMHC conversion is too low to
be accurately detected. To generate an exotherm that can be detected, such as that which occurs
25
-------
during a DPF regeneration event, excess fuel must be used which not only increases that rate of
deterioration of the device but also generates emissions. As such, the exotherm monitoring
approach can verify that sufficient temperatures are being generated to assist in DPF regeneration,
but we see no good reason to increase deterioration and emissions on a per-trip basis to generate
such an exotherm for the purpose of evaluating NMHC conversion efficiency. In summary, we still
require a functional check of the NMHC catalyst to ensure that it is providing the necessary
exotherm to assist in any regeneration functions it may have, and we still require a functional check
to ensure that some level of NMHC conversion is occurring, but we have eliminated the NMHC
threshold associated with that monitoring. Please refer to our response under section III.A.5 for
more detail.
II.B.6 Selective Catalytic Reduction (SCR) and Lean NOx Catalyst Monitoring
What We Proposed:
We proposed that malfunctions related to conversion efficiency, active/intrusive reductant
delivery, active/intrusive reductant quantity, active/intrusive reductant quality, and feedback control
be detected prior to emissions exceeding the thresholds for "NOx catalyst system." Further, we
proposed that conversion efficiency and reductant quality be monitored once per trip and that
reductant delivery, quantity, and feedback control be monitored continuously.
Comments:
An adequate supply and proper type of reductant will be critical to the functioning of SCR
systems for NOx control. Therefore, we believe that in all circumstances, there should be an
alternative indicator capable of readily notifying the operator of a problem with the reductant level
and reductant type. The Driver Warning System, as described in EPA's Novembers, 2006 Draft
Guidance Document for Certification Procedure for Light-Duty and Heavy-Duty Diesel Vehicles
Using Selective Catalyst Reduction (SCR) Technologies includes the necessary elements for such
an alternative indicator (i.e., visual warning, escalating in intensity, distinguishable from general OBD
monitors). In addition to the alternative indicator, if the reductant tank becomes empty or is filled with
an ineffective reductant (e.g., water), a MIL should be illuminated and DTC registered.
NESCAUM, 2005-0047-0024, p. 6-7
The Department disagrees with providing manufacturers the option of delaying illumination of
the MIL; storage of a MIL-on DTC; and using an alternative indicator for notifying the vehicle
operator when the reductant tank is empty, the reductant tank contains improper reductant, a
malfunction exists in the reductant system, or deterioration of the reductant delivery system occurs.
If the MIL is not activated when a malfunction of this type is detected and reconfirmed a MIL-on DTC
will not be stored. An activated MIL and a stored MIL-on DTC for this malfunction should be included
if HD OBD I&M is implemented. Malfunctions of these types in the reductant system would likely
cause the engine to exceed the proposed emissions thresholds and should trigger illumination of the
MIL and storage of a MIL-on DTC. This type of malfunction should be treated no differently than any
other malfunction that would result in emissions from an engine exceeding the proposed thresholds.
The Department would support the activation of an alternative indicator if the MIL is also activated
and a MIL-on DTC stored.
The Department supports the allowance for immediate MIL deactivation and DTC(s) erasure
once the OBD system has verified the reductant tank has been properly refilled and the MIL has not
been activated for any other malfunction. The Department recommends addition of an indicator light
that activates when the reductant tank reaches a low level, similar to the low fuel level indicator in
light-duty vehicles, which will allow the vehicle operator sufficient time to replenish the tank before it
is empty. While requiring the low level indicator may be beyond the scope of HD OBD, it would be a
26
-------
useful tool for the vehicle operator and would avoid a significant number of unnecessary failures in
an I/M program. The Department would support the use of a low level warning light for all areas that
may require vehicle operator attention, such as low reductant tank level, low fuel tank level, and low
or high battery or system voltages.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 9
The current proposed NOx emissions threshold of NOx standard/FEL plus 0.3 g/bhp is not
achievable and should be revised in the final rule to the NOx standard/FEL plus 0.6 g/bhp. EMA
supports EPA's proposal not to further reduce the NOx aftertreatment emissions thresholds for 2013
and later.
EMA, 2005-0047-0026, p. 12
A late comment suggested that EPA require that the OBD system be capable of verifying
that the liquid in the urea tank is indeed urea. The commenter believes that such a requirement
should be placed on the OBD system rather than allowing the system to rely on NOx sensors or
other sensors to monitor the SCR catalyst's NOx conversion which would only alert the driver to a
low conversion efficiency but would not identify the problem as being an improper liquid in the urea
tank.
Mitsui Mining & Smelting Co, Ltd., 2005-0047-0040, p. 1
Agency Response:
In general, the comments concerning the warnings and/or driver inducements associated
with the urea fill level and quality are being addressed by our manufacturer guidance on that subject
(see Manufacturer Guidance letter CISD-07-07, March 27, 2007). That guidance specifies a need to
have a separate driver warning system to indicate malfunctions associated with the reductant fill
level and the reductant quality. The guidance does not specify that a urea quality sensor be used.
We do not believe that a quality sensor is required provided there are other sensors - presumably
NOx sensors or oxygen sensors - capable of monitoring NOx conversion over the SCR catalyst.
Those sensors would be capable of detecting a loss of NOx conversion which would illuminate the
OBD MIL, although they would not necessarily be capable of pinpointing the problem as being
associated with urea quality. Provided the SCR catalyst is being monitored for proper NOx
conversion, a urea quality sensor could be considered redundant and representing unnecessary cost.
There may be other reasons beside emissions control - such as safety and/or durability of the
reductant system - for which manufacturers may choose to employ a urea quality sensor, but those
reasons are beyond the scope of an emissions-related OBD system.
We have revised our NOx threshold forthe 2010-2012 timeframe from the NOx FEL+0.3 to
FEL+0.6 based on our understanding of NOx sensor capabilities. We are keeping our proposed
threshold of the NOx FEL+0.3 for model years 2013 and later given our continued confidence that
sensor technology will continue to improve. Please refer to our response under section III.A.6 below
for more detail.
II.B.7 NOx Adsorber System Monitoring
What We Proposed:
We proposed that malfunctions related to adsorber system capability, active/intrusive
reductant delivery, and feedback control be detected prior to emissions exceeding the thresholds for
27
-------
"NOx catalyst system." Further, we proposed that adsorber capability be monitored once per trip
and that reductant delivery and feedback control be monitored continuously.
Comments:
The current proposed NOx emissions threshold of NOx standard/FEL plus 0.3 g/bhp is not
achievable and should be revised in the final rule to the NOx standard/FEL plus 0.6 g/bhp. EMA
supports EPA's proposal not to further reduce the NOx aftertreatment emissions thresholds for 2013
and later.
EMA, 2005-0047-0026, p. 12
Agency Response:
We have revised our NOx threshold forthe 2010-2012 timeframe from the NOx FEL+0.3 to
FEL+0.6 based on our understanding of NOx sensor capabilities. We are keeping our proposed
threshold of the NOx FEL+0.3 for model years 2013 and later given our continued confidence that
sensor technology will continue to improve. Please refer to our response under section III.A.7 below
for more detail.
II.B.8 Diesel Particulate Filter (DPF) System Monitoring
What We Proposed:
We proposed that malfunctions related to the DPF filtering performance, regeneration
frequency, regeneration completion, NMHC conversion, active/intrusive reductant injection, and
feedback control be detected prior to emissions exceeding the thresholds for "DPF system." We also
proposed that a missing DPF substrate be detected. Further, we proposed that all of these functions
be monitored whenever conditions were met.
Comments:
EPA should revise the proposed DPF threshold monitoring requirement to a threshold of the
PM standard orFEL+0.09 g/bhp-hr, or an absolute level of 0.10 g/bhp-hr, whichever is higher, and
maintain that threshold through 2013 and beyond, until advancements in sensor or detection
technology justify further changes. Further, EMA supports EPA's language that allows an exception
to the DPF monitoring requirements to exclude detection of specific failure modes such as partially
melted substrates, if the most reliable monitoring method developed requires it. Current monitoring
technology is very limited in terms of detecting "non-homogenous failures," which is what such
language is intended to address. EMA also supports maintaining the 2010 threshold through 2013
and beyond, which differs from the ARB approach to step the threshold down in 2013. EMA agrees
with EPA that there is no technical data supporting such a step in 2013 (72 Fed. Reg. 3255).
EMA, 2005-0047-0026, p. 9
EMA knows of no practical method to determine the NMHC conversion capability of the DPF.
EPA should eliminate the proposed requirement to monitor NMHC conversion efficiency of the DPF.
EMA, 2005-0047-0026, p. 10
Agency Response:
28
-------
The PM thresholds are not being changed for the final rule. We understand that there are
manufacturers that intend to certify systems to that threshold level. However, for the final rule, we
have added a new alternative monitoring requirement for DPF filtering performance. This alternative
requirement will be available during model years 2010-2012 only and would allow the system to
detect a malfunction when a detectable decrease in the in the expected pressure drop - relative to a
clean, nominal filter - occurs for a period of 5 seconds or more. The monitoring area for this
alternative approach is determined using the test cycles and procedures for the supplemental
emissions test (SET) under §86.1360-2007. The monitored area shall include all engine speed and
load points greater than a region bounded by a line connecting mode numbers 2, 6, 3, and 13 (i.e., a
line connecting A100, A75, B50, and C50). At engine speeds greater than "speed C", the monitor
shall run whenever engine load is greater than 50%. The detectable change in pressure drop is
determined by operating the engine at the "B50" speed and load point (as described in the SET test
procedures), observing the pressure drop on a clean, nominal DPF, and multiplying the observed
pressure drop by 0.5 or other factor supported by data and approved by the Administrator (see
§86.010-18(g)(8)(ii)(A)). We believe that this alternative addresses the feasibility concerns raised by
EMA. With this alternative approach, it is the pressure sensing capability of existing delta pressure
sensors which define a minimum detectable change in the pressure drop across the DPF. In
addition, this optional approach will only monitor during higher load conditions, where the delta
pressure across the DPF is significant and within the working range of the sensor. For 2013 and
later model years, when tailpipe soot sensors are expected to be available and capable of detecting
the quantity of PM passing through the DPF, the thresholds will remain as we proposed. Regarding
the NMHC conversion monitoring in our proposal, which required the OBD system to detect loss of
NMHC conversion for DPFs that converted NMHC emissions (e.g., catalyzed DPFs), we have
eliminated this requirement in the final rule because we know of no good way to monitor the NMHC
conversion efficiency with the accuracy required to detect very minor losses in efficiency. Please
refer to our response under sections III.A.5 and III.A.8 below for more detail.
II.B.9 Exhaust Gas Sensor Monitoring
What We Proposed:
We proposed that malfunctions related to sensor performance be detected prior to emissions
exceeding the applicable thresholds. We also proposed that malfunctions related to circuit integrity,
feedback functions, monitoring functions, and heater performance and circuit integrity be detected
prior to those functions being lost. Further, we proposed that sensor and heater performance be
monitored once per trip, that monitoring functionality be monitored whenever conditions were met,
and that circuit integrity and feedback functionality be monitored continuously.
Comments:
EMA recommends that EPA revise the proposed NOx emissions threshold in the final rule to
the NOx standard/FEL plus 0.6 g/bhp based on the capability of NOx sensor technology. In addition,
the reference to a PM threshold requirement for NOx sensors should be eliminated from Table 1, as
it is not appropriate to require monitoring of PM emissions in relation to a NOx sensor based on the
premise that the NOx sensor is meant to measure and control NOx not PM. EMA generally
supports the Agency's proposal regarding air/fuel ratio sensor monitoring.
EMA, 2005-0047-0026, p. 13
Agency Response:
29
-------
We agree with EMA's concern and have changed the 2010 through 2012 NOx threshold to
the NOx FEL+0.6g/bhp based upon our understanding of NOx sensor capability. We discuss this in
more detail in our response to issue III.A.9, below. We disagree with EMA's comments that there is
a valid reason to remove the PM threshold. PM increases as NOx decreases. If a NOx Sensor
shows an artificially high NOx value, the control system will decrease NOx and increase PM.
Therefore, the PM threshold will remain.
II.C Monitoring Requirements and Timelines for Gasoline/Spark-lgnition Engines
II.C.1 Fuel System Monitoring
We received no comments pertaining to this section that require analysis.
II.C.2 Engine Misfire Monitoring
We received no comments pertaining to this section that require analysis.
II.C.3 Exhaust Gas Recirculation (EGR) Monitoring
We received no comments pertaining to this section that require analysis.
II.C.4 Cold Start Emission Reduction Strategy Monitoring
We received no comments pertaining to this section that require analysis.
II.C.5 Secondary Air System Monitoring
We received no comments pertaining to this section that require analysis.
II.C.6 Catalytic Converter Monitoring
We received no comments pertaining to this section that require analysis.
II.C.7 Evaporative Emission Control System Monitoring
What We Proposed:
We proposed that gasoline evaporative emission control systems be monitored and
malfunctions detected. We proposed no monitoring for diesel evaporative emission controls. Note
that diesel engines and vehicles have no evaporative emission control system.
Comments:
30
-------
To the extent that heavy-duty gasoline and diesel vehicles contribute evaporative emissions,
how will they be addressed?
Maryland Department of the Environment, 2005-0047-0022, p. 1
Agency Response:
Gasoline evaporative emissions are addressed via §86.010-18(h)(7). Diesel engines do not
contribute significantly to evaporative emissions since diesel fuel has very low volatility which results
in negligible evaporative emissions.
II.C.8 Exhaust Gas Sensor Monitoring
We received no comments pertaining to this section that require analysis.
II.D Monitoring Requirements and Timelines for Other Diesel and Gasoline Systems
II.D.1 Variable Valve Timing and/or Control (WT) System Monitoring
What We Proposed:
We proposed that WT system malfunctions related to achieving the commanded valve
timing and/or control within a crank angle and/or lift tolerance and slow system response be detected
prior to emissions exceeding the thresholds for "other monitors." Further, we proposed that these
malfunctions be monitored whenever conditions were met rather than once per trip.
Comments:
EMA generally supports the Agency's proposal regarding WT system monitoring.
EMA, 2005-0047-0026, p. 24-25
Agency Response:
We agree with this comment.
II.D.2 Engine Cooling System Monitoring
What We Proposed:
We proposed that cooling system malfunctions related to proper thermostat function and
engine coolant temperature (ECT) sensor readings be detected. Further, we proposed that
malfunctions tied to the thermostat be monitored once per trip and that most ECT malfunctions be
monitored once per trip except that circuit malfunctions must be monitored continuously.
Comments:
31
-------
EPA must eliminate the cooling system monitoring requirement from the proposed HDOBD
requirements, and reduce the engine coolant temperature sensor requirement to comprehensive
component monitoring with rationality.
EMA, 2005-0047-0026, p. 22
We ask the EPA not to impose OBD requirements that are impractical for non-vertically
integrated engine manufacturers to implement, specifically, the cooling system monitoring
requirement to determine if an engine has reached a warmed-up temperature and thermostat
monitoring. Because the cooling system is essentially both a vehicle-installed system and has high
variability from vehicle-to-vehicle, it is practically impossible for an engine manufacturer to diagnose
accurately. Such diagnosis would require unique OBD calibrations for each engine vehicle
combination.
Caterpillar Inc., 2005-0047-0030, p. 3
National Automobile Dealers Association, 2005-0047-0034, p. 3
OBD failure detection on the performance of the engine cooling system (and the charge air
cooling system) must recognize that a prolonged period of severe use, in addition to lack of ordinary
care by vehicle owners and operators, may inappropriately indicate that components have failed
when they may only need to be cleaned. Such indications of "emission control system failures" may
reduce the confidence vehicle owners have in the HDOBD system and may lead to adverse
behavior, such as ignoring an illuminated MIL, on the premise that it is only a dirty radiator.
Stringent diagnostic requirements on vehicle manufacturer-provided heat exchangers must be
avoided to minimize this risk.
Truck Manufacturers Association, 2005-0047-0028, p. 2
Agency Response:
The direct emission impact of a malfunctioning thermostat or cooling system is only a
secondary reason for the requirement. The primary reason the cooling system is monitored for
proper operation is because the engine manufacturer itself elects to use engine coolant temperature
as a primary enabling criteria for monitoring of nearly every emission critical component. The
requirement for cooling system monitoring is simply stated as a requirement to verify that the engine
properly warms up to the highest temperature required by the engine manufacturer for monitoring of
other components. The relative stringency of this monitor is a direct result of how high the
manufacturer requires engine coolant temperature to get before monitoring other components and
engine manufacturers can effectively desensitize algorithms to vehicle factors by enabling other
monitors at lower temperatures. While we understand engine manufacturers' concerns that actions
by truck builders and users can impact their monitor design, the intent of OBD systems is to have
monitoring of the emission components during real world operation of heavy-duty vehicles. Failure to
achieve the necessary warmed-up temperatures required for monitoring would effectively mean
monitoring is indefinitely disabled on real world vehicles, thus negating nearly the entire OBD system.
Verifying the cooling system is operating properly is a crucial and necessary element to ensure OBD
systems continue to operate on real world vehicles throughout their life.
We suspect that the OBD requirements will result in the limitations and specifications that the
engine manufacturer will have to place on the vehicle builders to ensure the engine remains in a
legally certified configuration. This cooling system monitoring requirement may result in additional
calibration work or classification of the exact extent to which the vehicle builders can modify the
32
-------
cooling system to maintain a compliant system. Failure to do so could result in MIL illumination.
However, while eliminating the cooling system monitoring requirements would avoid this potential
MIL illumination, it would also re-open the possibility that such a system would be put into service
and all of the OBD monitors would be disabled for the entire life of the vehicle.
Subsequent to the closing of the comment period, we met with EMA to discuss, among other
things, cooling system monitoring provisions.4 Based on those discussions, we have changed the
requirement to allow the manufacturer to have a temperature lower than the "normal 20 degrees
below the nominal thermostat regulating temperature" at which a malfunction must be detected
provided the ambient temperature is between 20 F and 50 F. To do so, the manufacturer must
present data justifying the new temperature to be reached at the lower ambient temperatures. EMA
also commented on the number of trips for engine cooling system monitor when we met with them.
We don't feel that increasing the number of trips to more than two for any OBD monitor is acceptable,
since it will lead to decreased trust in the reliability of faults. Monitoring conditions for diagnostics
must be setup such that diagnostics only run in regions where they are reliable.
II.D.3 Crankcase Ventilation System Monitoring
Comments:
EMA generally supports the Agency's proposal regarding crankcase ventilation system
monitoring.
EMA, 2005-0047-0026, p. 24-25
Agency Response:
We agree with this comment.
II.D.4 Comprehensive Component Monitors
What We Proposed:
We proposed that, in general, the OBD system must detect a malfunction of any electronic
engine component or system that either provides input to or receives commands from the onboard
computer(s). Further, we proposed that malfunctions related to circuit continuity and/or out-of-range
values be monitored continuously and that malfunctions related to input data rationality and/or output
component functional response be monitored whenever conditions were met.
Comments:
Regarding glow plug and intake air heater system monitoring, EPA should eliminate the glow
plug and intake air heater system monitors as a requirement for 2010. EPA should conduct further
analysis of HD vehicle confounding issues related to these cold start aids to establish what is
needed fora future rulemaking, if any. Should EPA determine that such monitoring is appropriate
for 2013 and beyond, the monitoring should be for functional response and circuit continuity only
with no monitoring for low to moderate degrees of degradation, comprehensive component
monitoring should be required for only the operating controls for power relays or independent
4 See memorandum to Docket ID No. EPA-HQ-OAR-2005-0047 from Todd Sherwood, "Meetings with the
Engine Manufacturers Association," document ID # EPA-HQ-OAR-2005-0047-0053.
33
-------
controller feedback, and detection only of a single glow plug failure should be required. Regarding
other comprehensive component monitors, EMA is generally supportive of the Agency's proposal.
EMA, 2005-0047-0026, p. 19
National Automobile Dealers Association, 2005-0047-0034, p. 3
TMA notes that the examples cited for comprehensive component monitoring do not include
operator controls, and submits that operator controls are ill-suited for comprehensive component
monitoring. Examples of operator controls include accelerator pedals, cruise control switches, brake
switches, and clutch switches. Defining rational use of individual controls and combinations of
controls by the operator would be fraught with exceptions and subsequent errors of omission for the
rationality monitors. Most commercially available switches, purchased by vehicle manufacturers for
operator controls, are not available in forms that can be readily diagnosed for circuit failures. Adding
comprehensive component monitoring to these circuits would be accomplished at significant
expense to vehicle manufacturers with no additional emissions reduction benefit. Vehicle
manufacturers also strongly recommend that remotely-mounted fuel filters be excluded from
comprehensive component monitoring requirements. The term "fuel preparation systems" is overly
broad and could be construed to include fuel filtration systems that are not mounted on the engine.
These systems sometimes indicate excess water in diesel fuel and work to separate it out, but they
do not measure fuel qualities such as cetane level or sulfur level. Many filters heat the fuel before it
is filtered and water is separated to minimize the impact of wax crystals on vehicle performance.
Indeed, if a water separator becomes clogged with wax, the engine stalls because it becomes fuel
starved—clearly not an emissions problem. Addressing full water bowls and gelled fuels in fuel filters
are considered routine maintenance and should not be MIL-on failures. Heater elements in fuel
filters are self-regulating and are not powered by the engine control system.
Truck Manufacturers Association, 2005-0047-0028, p. 3
Agency Response:
For the final rule, we have changed the regulation consistent with EMA's comments to allow
circuit checks only on glow plugs for 2010-2012. For 2013 and beyond, the regulation requires a
functional check of the sensors. The functional requirement can be met by determining the current
used by the sensor is rational. Intake air heaters will require both a circuit and a functional check in
2010, since all sensors needed to do the testing are available.
We disagree with TMA on monitoring of some operator controls. The primary reason some
operator controls are monitored for proper operation is because the engine manufacturer itself elects
to use them as parameters in monitoring strategies of emission critical components. All inputs to
monitors must be monitored to prevent both false MIL illuminations and disabling of monitors. If an
input such as the cruise switches are not used in any diagnostic or used to switch between
emissions strategies, then they do not need to be diagnosed. Inputs such as accelerator pedal,
brake switch, and clutch switch will probably be used as inputs to diagnostics or emissions controls,
and will need to be diagnosed. We understand the manufacturers do not have experience with
diagnosing operator controls, but we have seen from both the medium duty and light duty industries
that such diagnosis is capable of being robust.
We agree with TMA that fuel filters that are self regulating and do not prepare the fuel for
combustion (such as heating the fuel to a specific temperature) do not need to be diagnosed. We do
believe under some conditions, a fuel filter that is not operating properly may cause a malfunction of
a monitor for fuel pressure, fuel quantity, or fuel timing to occur. We do not feel this is an issue since
there is something to repair. In the future, if fuel filters are capable of heating fuel for proper
combustion, we would expect that feature to be diagnosed.
34
-------
II.D.5 Other Emissions Control System Monitoring
What We Proposed:
We proposed that, for other emission control systems that are not otherwise specifically
addressed be monitored, and that the manufacturer submit a plan for Administrator approval of the
monitoring strategy, malfunction criteria, and monitoring conditions prior to introduction on a
production engine.
Comments:
We received no comments pertaining to this section that require analysis.
II.D.6 Exceptions to Monitoring Requirements
What We Proposed:
We proposed that certain monitors could be disabled under specific conditions related
generally to ambient conditions. Further, we proposed that most such disablements be approved by
the Administrator.
Comments:
We expect that monitors will be automatically re-enabled whenever an extreme condition is
no longer in effect. NESCAUM does not support the use of systems that need to be manually re-
enabled. As experience is gained with OBD systems, manufacturers will have opportunities to
improve the reliability of OBD systems. Therefore, we urge EPA not to grant open-ended
authorizations to disable monitors, but rather require manufacturers to investigate improvements to
the reliability of OBD systems and sunset the exceptions to monitoring requirements in subsequent
model years. In addition, whenever a monitor is disabled, a subsequent OBD scan should reveal the
disablement. We have questions regarding disablement for low temperature and low fuel levels.
Regarding low temperature, we assume the disabled monitors would be those affected by cold start
conditions. However, even under extreme cold conditions, the engine eventually will reach normal
operating temperature, allowing monitors affected by cold start conditions to operate properly. We
assume these monitors can be re-enabled at this point, regardless of ambient temperatures, but the
proposal appears to allow for continued disablement until ambient temperatures rise above 20
degrees. If this is in fact EPA's intent, we request an explanation. Regarding low fuel level, 15
percent of nominal tank capacity may represent a large volume of fuel, particularly in a large vehicle
such as a heavy-duty truck. We therefore request an explanation as to how EPA determined that the
15 percent threshold is appropriate across the entire fleet of affected engines.
NESCAUM, 2005-0047-0024, p. 6
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 9-10
While supporting the concept of "permanent" DTCs, there is concern over the ready status
allowances of §86.010-18(k)(4)(i)(A) which would allow certain monitors to be set to "ready" despite
not having run in-use by satisfying the disablement allowances of §86.010-18(i)(5)(ii). Many areas
will routinely experience temperatures below 20 degrees Fahrenheit for several weeks. In
consideration of permanent DTCs as an I/M failure criterion, a vehicle with a permanent DTC could
take several weeks until the temperature-based enable criterion was achieved to perform the
needed drive cycles to erase the permanent DTC. Section 86.010-18 should require the HD engine
manufacturers to report to EPA those monitors affected by the allowable cold weather disablement;
35
-------
whether the status of these monitors can/will be reported as "ready" when the temperature enable
criterion is not met; and to list permanent DTCs that could be affected by prolonged cold
temperature.
NY State DEC, 2005-0047-0033, p. 2-3
Agency Response:
Regarding disabling monitors when certain conditions are present- ambient temperatures
below 20 degrees F and/or altitudes above 8,000 feet - these are long standing provisions in the
<14,000 pound OBD requirements and, we believe, are important elements to our HDOBD program
given the difficulties associated with monitoring under such extreme conditions. Note that the
provision requires the manufacturer to request such disablements and to support those requests
with data or engineering analyses. Therefore, we believe that we have provided protection against
manufacturers using these provisions as a means of disabling monitors during any significant portion
of real world driving. While it is true that OBD monitors may be disabled for extended periods in
areas where such extreme conditions are common, that outcome is far better than keeping monitors
active and triggering false malfunctions. As for manual re-enablement of monitors, this would
certainly not be the case. Any such disabled monitors - and it is important to note that while specific
monitors may be disabled the entire OBD system is not disabled -would be re-enabled either when
the necessary operating conditions are met or at the next key-on provided the extreme conditions
were no longer present. There is no specific requirement to keep track of such disablement or to
communicate this disablement via an OBD scan. We remind the commenter of the rate based
monitoring requirements whereby we can keep track of how often monitors run. We consider this a
more important metric than how often it is disabled. As for re-enabling monitors once the engine has
warmed-up even though ambients may still be low, we do not disagree with this. However, some
monitors affected by cold ambients at start-up - and again, we remind that it would be specific
monitors being disabled and only upon request with supporting data - may only run at start-up (e.g.,
some evaporative system monitors run only at start-up). Such a monitor could not be re-enabled
during that trip once start-up conditions have passed, although the monitor would be re-enabled at
the next start-up provided the extreme conditions were no longer present. As for the fuel level
disablement, this would be allowed only for monitors impacted by low fuel levels. If the manufacturer
cannot provide data or analysis to substantiate their request, they would not be allowed to disable
any monitors based on fuel level.
Regarding the comment from the NY State DEC, we believe that the regulatory provision
requires this information to be provided to EPA prior to being allowed to employ any such
disablement(s).
II.E A Standardized Method to Measure Real World Monitoring Performance
II.E.1 Description of Software Counters to Track Real World Performance
Comments:
EPA has proposed that numerators and denominators cease incrementing if a malfunction is
detected on a sensor or signal that is used to determine the conditions required for incrementing.
Further, if the condition no longer exists, the OBD system must once again begin incrementing these
counters. However, EPA further requires that, if a malfunction exists that prevents these counters
from incrementing, and the system receives a "clear codes" command, these counters should start
incrementing again. In some cases, such a requirement could cause the OBD system to increment
36
-------
these counters erroneously. This requirement also adds unnecessary complexity. Instead, EPA
should allow manufacturers the option of whether to restart the incrementing of these counters
immediately after a scan tool "clear codes" command or after a key-off cycle following a "clear
codes" command.
EMA, 2005-0047-0026, p. 39
Agency Response:
We disagree with EMA's concern about re-enabling incrementing of numerators and
denominator after a clear codes event or when a malfunction no longer exists. We will clarify the
meaning of a malfunction here, since many EMA members do not have experience with OBD. Once
a malfunction has been determined during a drive cycle, it cannot be considered to be gone until the
end of the drive cycle, since it may reoccur at a later time during the drive cycle. A malfunction is
only considered to exist if a pending DTC (or MIL-on DTC) has been logged. A malfunction
continues to exist until one of two things happens. 1) The pending DTC is cleared. 2) The MIL fora
malfunction has been extinguished.
Manufacturers must put in protect to prevent counters from incrementing more than once per
drive cycle. Once the protection is in place, the "clear codes" command will not erroneously allow
incrementing of of counters. Most manufacturers use bits to keep track of what needs to be
increment then increment the counters at the end of a driving cycle. We do feel this protection is
necessary though we expect the number of "clear codes" commands that occur during driving to be
statistically insignificant.
II.E.2 Proposed Performance Tracking Requirements
We received no comments pertaining to this section that require analysis.
II.F Standardization Requirements
Comments:
We support standardization of various features, including diagnostic connectors, computer
and wireless communication protocols, hardware and software specifications for service technician
tools, information communicated by the onboard computer, methods for accessing onboard
information, numeric designations of DTCs, and service manual terminology. Effective
standardization facilitates diagnosing and repairing malfunctions and potential use of OBD checks in
heavy-duty I/M programs.
NESCAUM, 2005-0047-0024, p. 6
Agency Response:
We agree with these comments.
II.F.1 Reference Documents
Comments:
37
-------
Volvo Powertrain supports the CARB and EMA position that OBD communication
standardization is not to be required before 2013.
Volvo Powertrain, 2005-0047-0021, p. 1
In §86.010-18(k), EPA should add language to the introductory paragraph to clarify that the
standardization requirements, which do not become effective until 2013, must be met "when
specified." As written, the implication is that datalink standardization could be required in 2010. In
addition, the dates for SAE J1939 and J1979 should be updated to more recent versions. Also, SAE
J2534, a recommended practice for pass through reprogramming, is not referenced elsewhere in
§86.010-18(k) or §86.013-18(k) and so should not be referenced at all in this section but should only
be referenced in §86.010-38 which pertains to service information availability.
EMA, 2005-0047-0026, p. 36
Agency Response:
As in the CARB rule, our intent was to propose that standardization begin in the 2013 model
year. The proposed regulatory text, at §86.010-18(k)(1), is confusing where it reads "The OBD
system must conform with the following Society of Automotive Engineers (SAE) standards and/or the
following International Standards Organization (ISO) standards." That text has been removed so
that §86.010-18(k)(1) now reads "Reference materials. The following documents are incorporated by
reference, see §86.1". The standardization requirements are then contained elsewhere in individual
paragraphs of §86.010-18 and superseding sections.
Reference to SAE J2534 has been removed from §86.010-18 and superseding sections
since, as noted by EMA, that recommended practice is relevant only in §86.010-38 pertaining to
service information availability.
II.F.2 Diagnostic Connector Requirements
Comments:
The requirement to locate the connector in the driver's foot-well is inappropriate for some
vehicle types, especially those without a driver's side door (e.g., a bus). Since this connector is
installed by vehicle manufacturers and these vehicles may be designed after the certification data for
the engine is provided, alternate locations will need to be requested for specialty vehicles on an on-
going basis.
EMA, 2005-0047-0026, p. 37-38
ATA supports the use of a standard data link connector conforming to SAE J1962 orJ1939-
13 as well as use of a generic scan tool and communication protocol.
American Trucking Association, 2005-0047-0029, p. 5
TMA members supported their customers' efforts in the American Trucking Association's
Technology and Maintenance Council to develop TMC RP1202. TMC RP1202A recommends use of
either a connector specifically for SAE J1708/J1587orthe J1939-13 connector. The J1939-13
connector includes provisions for J1708/J1587. Use of the J1939-13 connector in the future for HD
OBD access will not deprive vehicle owners of access to J1708/J1587. SAE J1968 (2002) does not
provide accommodations for J1708/J1587. Vehicles that select the SAE J1978/J1979 approach will
38
-------
need a separate connector for J1708/J 1587, if this data stream is continued to be supported by
engine manufacturers in 2013.
Truck Manufacturers Association, 2005-0047-0028, p. 4
Agency Response:
Regarding the EMA comment that alternate locations will need to be requested for specialty
vehicles, we agree and have provided for their allowance upon Administrator approval (see §86.010-
Regarding the TMA comment about the standard diagnostic connectors, we do not allow the
mixing of protocols on standard connectors (on the defined communication pins). Manufacturers
have not been prohibited from using multiple styles of diagnostic connectors on the same vehicle. In
addition, there are discretionary pins on the standard diagnostic connectors that are available for
their use.
II.F.3 Communications to a Scan Tool
Comments:
Volvo Powertrain is currently planning to use the World Wide Harmonized OBD
communication standard (ISO 27145) at the earliest possible time, and requests that ISO 27145 be
an optional communication standard within §86.013-18 and within §86.010-18 in the event that EPA
requires standardization in the 2010-2012 timeframe.
Volvo Powertrain, 2005-0047-0026, p. 1
In the main, Hughes supports EPA's proposal but believes that EPA should provide the
option to vehicle owners and operators to utilize wireless communication protocols for data
transmission. Hughes believes that EPA should allow heavy truck owners and operators the
benefits derived from wireless data transfer.
Hughes Telematics, Inc., 2005-0047-0037, p. 1-2.
Agency Response:
The ISO 27145 standard is being developed as part of the Worldwide Harmonized Heavy-
duty OBD global technical regulation (WWH-OBD).5 We will consider allowing that standard for
model year 2013 and later implementation, and may issue a technical amendment, direct final rule,
or proposed rule to address it.
Regarding the comment from Hughes, we do not believe that our rule prohibits wireless data
transfer so the option suggested by Hughes is available. So, while we do not intend to require
wireless data transfer at this time, it is allowed as long as the required wired data transfer is also
provided.
Global Technical Regulation Number 5: Technical Requirements for On-board Diagnostic Systems for
Road Vehicles; ECE/TRANS/180/Add.5; 23 January 2007, see
www. unece.org/trans/main/wp29/wp29wgs/wp29gen/wp29glob_registry. html
39
-------
II.F.4 Required Emissions Related Functions
Comments:
The proposed NTE-related requirements of §86.010-18(k)(4)(ii)(B)(4J have nothing to do with
the purpose and function of OBD requirements - detecting and correcting malfunctions in key
emission control system components. There is no adequate justification for including NTE-related
"in-use testing program" requirements in an OBD rulemaking. Nor is there justification for requiring
such data in the HD in-use testing (HDIUT) rulemaking. The data EPA seeks are otherwise readily
available (in the relatively few instances in which such information will be needed) to the Agency.
The HDIUT rule stemmed from the settlement of multiple NTE-related lawsuits filed against the
Agency, and resulted from many months of detailed negotiations regarding all aspects of the HDIUT
program, including the manner in which in-use data would be recorded and reported. EPA should
not now unilaterally include under the cover of the HDOBD rule requirements that were considered
and rejected as part of a negotiated, good faith resolution to litigation. As part of the HDIUT program,
very specific second-by-second data - including all of the NTE-related data at issue in the HDOBD
rule - will be recorded and reported to EPA pursuant to an expansive electronic data submission
template. In addition, the HDIUT rule requires that the manufacturer make available to EPA, within
60 days of an EPA request, written descriptions of all NTE deficiencies and parameters defining all
NTE limited testing regions with sufficient detail for EPA to determine if a particular deficiency or
limited testing region will be encountered in the emission test data from the portable emission-
sampling equipment and field testing procedures. The NTE requirements that EPA has proposed
are inconsistent with the negotiated settlement between industry and the Agency and impose
unnecessary and unjustified costs without benefit. These requirements should be eliminated from
the HDOBD rule.
EMA, 2005-0047-0026, p. 22-24
EPA should revise the calibration ID (CAL ID) and calibration vehicle number (CVN)
language to allow manufacturers to have one or multiple CAL IDs/CVNs per engine control module.
A single, unique CAL ID and its CVN is ill suited for the HD industry where engine control software is
often managed as classes of software objects. Variation in engine systems and accessories have
made the exercise of creating unique top-level part numbers for engine assemblies inefficient as the
number of top level part numbers exceeded the number of unique component parts. Moreover, light
duty engine manufacturers intend to leverage existing software architectures that utilize multiple CAL
IDs and prefer to sustain the existing OBD II (13 CCR 1968.2) software architectures. Furthermore,
for engine control systems that use component-driven software architectures, each component class
has one or more members which tailor the performance of engine control system to match the
differing mechanical components used across engines. Members from separate classes are
combined to create a full set of operating instructions and data constants for the engine control
system. Software implementing the scan tool interface proposed by §86.013-18(k) and evaluated
under §86.013-18(j)(1) is one example of a such a class. Specific identification of this class as its
own CAL-ID is desirable for the purposes of aiding the administration of production evaluation
standardization testing for OBD Groups. Other examples where separate identification of classes is
desirable are to separately identify operating instructions (or algorithms) from data constants
(calibration). The commenter suggests the following text, in lieu of that proposed, to simultaneously
provide regulators with the desired information while giving manufacturers additional flexibility to
document software that is standardized among product lines:
(k)(4)(vi) Software calibration identification (CAL ID). Each
engine shall contain at least one CAL-ID, which can be directly
traced to the certified engine family. Engine control systems that
simultaneously support more than one certified engine family shall
provide a CAL-ID that displays a unique value for each engine
40
-------
family supported. The value displayed shall match the engine's
operating engine family performance specification. Additional
CAL-IDs may be provided at the manufacturer's option. When
additional CAL-IDs are provided the 1st CAL-ID shall be directly
traceable to the certified engine family. Each CAL-ID provided
shall represent a unique stream of data, any changes to the data,
including a single bit, shall be represented by a separate CAL-ID.
All CAL-IDs shall be reported through the standardized data link
connector in accordance with the SAE J1979/J1939 specifications.
EMA, 2005-0047-0026, p. 36-37
HD vehicles have a history of being built with multiple responses to a request for the vehicle
identification number (VIM) via the datalink. While engine manufacturers will do their best to ensure
that devices ordered by the customer, including the engine control system, will provide only one
response to a request for a VIN via the datalink, they cannot assure EPA that the customer (e.g., the
vehicle manufacturer and/or truck owner) cannot and will not install aftermarket devices that also
reply to a request for the VIN via the datalink.
EMA, 2005-0047-0026, p. 37
TMA members support the ability to provide a VIN by extending the existing collaborative
agreements with their engine suppliers, and not requiring the VIN to be provided exclusively by the
engine control system.
Truck Manufacturers Association, 2005-0047-0028, p. 4
How will HD engine manufacturers allow for the modification of the electronic VIN (in the
case of an engine swap) but still maintain adequate security measures to ensure the validity of this
data element?
NY State DEC, 2005-0047-0033, p. 3
The datastream requirements of §86.010-18(k)(4)(ii)(A) through (C) may need to be adjusted
to better suit the emissions control and diagnostic methods achieved from 2010 to 2016.
Manufacturers should be allowed to petition to omit data in section (B) that is not relevant to their
system designs. EPA should also institute some mechanism to update the list to the relevant
content.
EMA, 2005-0047-0026, p. 38
Section 86.010-18(k)(4)(ii)(B)(7) requires that the OBD system report the most accurate
calculated load and torque parameter values that are calculated within the applicable electronic
control unit. Most accurate is stated as being of sufficient accuracy, resolution, and filtering to be
used for the purposes of in-use emission testing (e.g., with a PEMS device). However, the torque
information that was created in the standardized protocol was created for improved power train
integration, not for PEMS for which the torque precision may not be adequate. Proprietary protocols
(versus the standardized protocols) use different scaling which can lead to more precise load or
torque information than that required through the public protocols. Users of the raw proprietary data
have to know the sampling techniques, noise factors, and other characteristics to properly filter the
data for use. For these reasons, this paragraph of should be removed from the regulation.
Cummins Inc., 2005-0047-0026, p. 9
41
-------
EPA should clarify the the examples given in §86.010-18(k)(4)(iv)(B) - pertaining to use of
separate diagnostic trouble codes for circuit checks and out-of-range checks of a given monitor - are
not prescriptive and it is not intended to require the OBD system to always discriminate malfunction
conditions. That is, these examples should not modify the full requirements that are given in
§86.010-18(i)(3)(ii)(A) and (B) -which pertain to the comprehensive component malfunction criteria
for input components.
EMA, 2005-0047-0026, p. 38
While the option to report readiness status using the MIL appears to be consistent with
CARB's approach, it is not sufficiently clear. In 2010, the manufacturer may provide readiness
status in a manufacturer-defined format, if the manufacturer does not voluntarily use J1979 or
J1939-73 definitions to provide this data. EPA should clarify that there is no intent to require two
trips to confirm ready status, but should ensure that the language is consistent with that adopted by
CARB.
EMA, 2005-0047-0026, p. 39
EPA should revise the requirement that permanent diagnostic trouble codes be erased upon
reprogramming of the engine control computer. When engine controllers are reprogrammed, the
erasure of permanent DTCs is not always required. If erroneous DTCs are resolved with a
reprogramming event, the proposed OBD regulation already permits the engine controller to
automatically clear any permanent DTCs immediately after the OBD system has tested them once
and determined a malfunction no longer exists. EPA should not make it mandatory that permanent
DTCs be made erasable after a reprogramming event but, rather, it should be left up to the
manufacturer whether these codes are required to be erased upon reprogramming, made erasable,
or remain un-erasable upon reprogramming.
EMA, 2005-0047-0026, p. 39
There should be a method to determine whether or not a particular engine family or
equipment chassis is OBD equipped. This information is especially important to have during the
early OBD phase-in period. It would also be useful for the scan tool to be able to determine what
type of OBD system is present (e.g. OBD1 vs. OBD2). The states need some method to determine if
the vehicle can be OBD inspected during an annual/roadside inspection.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 7
Agency Response:
Regarding the New Jersey comment that the type of OBD system be communicated via the
data link connector, we have put forth such a requirement (see §86.010-18(k)(4)(ii)(A)(3) and
§86.010-18(k)(4)(ii)(B)(3)).
Regarding the EMA comment that NTE type information is not pertinent to an OBD rule, we
ask where is such information pertinent? Within the context of an NTE based rule or a HDIUT
program rule, we suspect that EMA might make the argument that this requested information was
really OBD-related and should not be handled via a NTE or HDIUT program rule. We believe the
OBD regulation is the proper place to require this information be made available since the OBD rule
sets the communication requirements. As a general principle though, the primary topic of a rule
does not prohibit inclusion in that rule of items that do not seem pertinent to that rule. We proposed
that this information be available because we believe it will help us in our in-use testing program.
42
-------
Regarding the EMA comment that reporting the readiness status using the MIL is not
sufficiently clear, we believe that the current text contained in section (k)(4)(i)(C) is clear, sufficient,
and consistent with language adapted by CARB. The text clearly states that the reporting of the
readiness status using the MIL is for "all monitored components or systems", and does not imply that
it is for a single monitor or that multiple trips may be required.
Regarding the EMA comment concerning erasing diagnostic trouble codes upon
reprogramming of the engine control computer, we believe that all DTCs should be cleared. We
have given great consideration to this issue and believe that when the controller has been
reprogrammed, all diagnostic information related to the previous program is no longer valid and
should be cleared. Not clearing them could easily lead to incorrect diagnosis of issues, and
association of faults with incorrect software and/or calibrations.
Regarding the EMA comment that the EPA shoud clarify the examples given in §86.010-
18(k)(4)(iv)(B), we agree that the text is confusing and could be interpreted to appear contradictory
to section §86.01 0-1 8(i)(3)(ii)(B). The text of section §86.01 0-1 8(k)(4)(iv)(B) has been modified (by
adding the phrase "to the extent possible") to be more clear and consistent with section §86.010-
Regarding the EMA comment that the EPA should allow multiple CAL IDs and CVNs per
engine control module, the EPA has agreed to allow this upon request. The text of sections
§86.01 0-1 8(k)(4)(vi) and §86.01 0-1 8(k)(4)(vii)(A) will be updated to allow multiple CAL IDs and
CVNs per emissions critical powertrain control unit. This change would make the requirement
consistent with CARB's most recent OBDII update (November 2007), and we believe that CARB
may consider this change in their next HDOBD biennial review.
Regarding the Cummins comment that the OBD system must report the most accurate load
and torque parameter values, we are confident that the parameters specified in section §86.010-
18(k)(4)(ii)(B) can be calculated in such a way that they meet the specifications of J 1939 and the
requirements of sections 18(k)(4)(ii)(B). If the parameters in question are of higher resolution than is
specified in J1939, then they would need to be scaled in a manner to satisfy the J 1939 specifications.
Regarding the EMA comment concerning the datastream requirements of section §86.010-
18(k)(4)(ii)(A) through (C), manufacturers would not need to report values of parameters not relevant
to their system design. If the list of parameters in this section needs to change in the future, we will
address them appropriately at that time.
Regarding the comment by the EMA concerning multiple responses to a request for the
vehicle identification number via the datalink, we have not required that the VIN be located on a
specific electronic module. Which module responds to a VIN request has not been specified, only
that there is a single response. This is a system design issue that will have to be coordinated
between the vehicle manufacturer and the manufacturer of all modules on the datalink.
Regarding the comment by the NY State DEC and the comment concerning modification of
the electronic VIN (in the case of an engine swap), we have not specified that the electronic VIN be
located in a module that would necessarily need to be moved with the engine. The data validity,
how to reprogram the VIN, and which module contains the electronic VIN are all system design
issues that will have to be handled by the vehicle manufacturer.
I.F.5 In-use Performance Ratio Tracking Requirements
43
-------
Comments:
The definition provided in §86.010-18(k)(6)(i)(B) for determining the engine idle time is too
prescriptive. Methods for assessing engine idle time based on engine load can be equally effective
and should not be eliminated by a prescriptive definition. Instead, EPA should define a minimum
performance standard which would not prohibit alternate methods. The commenter suggests a
revised version of (k)(6)(i)(B) to read as follows:
(B) Total idle run time shall include all time where the accelerator pedal is
released by the driver, the vehicle is not moving, engine speed is greater than or
equal to 50 to 150 rpm below the engine's normal, warmed-up idle speed (as
determined in the drive position for vehicles equipped with an automatic
transmission), and power take-off not active. Total idle run time may include
additional time where the engine is being operated at low loads, as determined by
the manufacturer.
The rationale for this suggestion includes the nature of existing variable reluctance sensor
(VRS) technology use for vehicle speed measurement in vehicles, and potential implications of the
existing text when compared against the anti-idle provisions of 13 CCR 1956.8. Given that the
sensitivity of the vehicle speed system is subject to factors beyond the direct control of the engine
manufacturer, it is impractical for engine manufacturers to assure that a numeric standard can
always be met. Other regulations (e.g., those by the National Highway Traffic Safety Administration
(NHTSA)) govern vehicle speedometers and odometers. Less prescriptive regulation on EPA's part
would minimize potential conflicts with those other regulations. Moreover, EPA's definition must not
conflict with the anti-idling provisions of 13 CCR 1956.8 which require a tamperproof idle shutdown
system. EPA's proposed text, when interpreted literally, suggests that all time with the accelerator
depressed is not idle time, even if the accelerator is depressed by some mechanical means (e.g., a
broomstick).
EMA, 2005-0047-0026, p. 38-39
Agency Response:
We will change some of requirement (k)(6)(i)(B) as requested by EMA. As requested by
EMA in August of 2008,6 we will change "vehicle speed less than or equal to one mile per hour" to
"vehicle speed less than or equal to one mile per hour or engine speed less than or equal to 200 rpm
above normal warmed-up idle (as determined in the drive position for vehicles equipped with an
automatic transmission)". We will not include "Total idle run time may include additional time where
the engine is being operated at low loads, as determined by the manufacturer." We want a
consistent statistic, and allowing each manufacturer to determine which low load conditions to be
included will not give us a comparable measure.
Other commentors have requested that the minimum load for incrementing the denominator
be changed from 15% to 50%, and that we add that either load or vehicle speed be used for
incrementing the denominator. The purpose of the denominator is to have a standard drive cycle for
the ratio static. We cannot allow some manufacturers to use load while others use vehicle speed,
and still have a consistent statistic to measure by. The purpose of the "300 seconds of engine load
above 15%" was to have criteria that would represent that an engine had been doing work for 300
seconds. After consideration, we have determined that "above 1150 engine RPM for 300 seconds"
is a better measure of engine work, and will replace the 15% load in the regulation. We will also add
language to allow "vehicle speed greater than 25 mph for 300 seconds" for 2010 to 2012. After 2012
only "above 1150 engine RPM for 300 seconds" will be allowed.
6 See memorandum to Docket ID No. EPA-HQ-OAR-2005-0047 from Todd Sherwood, "Meetings with the
Engine Manufacturers Association," document ID # EPA-HQ-OAR-2005-0047-0053.
44
-------
II.F.6 Exceptions to Standardization Requirements
We received no comments pertaining to this section that require analysis.
II.G Implementation Schedule, In-use Liability, and In-use Enforcement
II.G.1 Implementation Schedule and In-use Liability Provisions
Comments:
The regulations should be implemented by an earlier date than proposed.
Community Board #1, 2005-0047-0015, p. 2
NESCAUM is concerned about the inconsistency regarding the date when EPA proposes
that all engine families and ratings become liable to certification thresholds (2019), compared to the
effective date for California engines (2016).
NESCAUM, 2005-0047-0024, p. 4
EMA supports EPA's proposed implementation schedule and we urge EPA to finalize the
implementation schedule as proposed. EMA also recommends that EPA become actively engaged
with ARB in the upcoming biennial review of ARB's heavy-duty OBD diagnostic requirements in
order to achieve alignment of ARB's implementation requirements with those of EPA. Allowing
adequate, if not generous, time for implementation will serve to help guarantee OBD system stability
and in-use performance and thus any hoped for environmental improvements.
EMA, 2005-0047-0026, p. 31
National Automobile Dealers Association, 2005-0047-0034, p. 4
EMA supports EPA's proposal to allow interim in-use compliance standards set at two times
the emission threshold level for each monitor for a period of three years after the time an applicable
engine rating is required to meet full OBD requirements. However, additional compliance flexibility is
needed for the DPF through 2019. Given manufacturer feasibility concerns at the threshold level
proposed, and to be consistent with the CARB HDOBD requirements (see 13 CCR 1971.1(m)(3.1)),
EPA should allow for a two times in-use threshold through 2019. EPA should also urge CARB to
align with EPA on other in-use compliance flexibility provisions.
EMA, 2005-0047-0026, p. 31-32
The proposed OBD implementation schedule, requiring manufacturers to install OBD in all
engines by 2019, is too long. A 2016 deadline is sufficient to implement full OBD.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 3
Agency Response:
Regarding the comment from Community Board #1, unfortunately there simply is not
sufficient time to implement HDOBD any earlier than the 2010 model year.
45
-------
Regarding comments on inconsistency with CARB on our implementation schedule, which
has not changed for the final rule, we believe it is consistent with that put forward by CARB in their
July 2005 version of 13CCR 1971.1. In proposed §86.016-18(o)(2), which has been moved to
§86.010-18(o)(3)(ii) in the final rule, the manufacturer is required to "implement an OBD system
meeting the requirements of this section on all engine ratings in all engine families." The commenter
may be confusing ourin-use compliance provisions of proposed §86.016-18(p)(2), which has been
moved to §86.010-18(p)(3)(ii) in the final rule, which states that certain OBD systems must meet
thresholds in-use that are double those required for certification. This in-use provision is consistent
with CARB's 1971.1(m)(3.2).
Regarding EMA's request for a 2x in-use threshold through 2019, as provided by CARB, we
do not agree. CARB made available CCR section 1971.1(m)(3.1) specifically to address those
engines meeting their more stringent 2013 DPF PM thresholds and to afford manufacturers a full six
model years of the 2x in-use threshold. We have not included that more stringent 2013 DPF PM
threshold. We, similarly, will have afforded manufacturers a full six years of the 2x in-use threshold
for the DPF (from 2010 through 2015). If we decide at some point to include the tighter 2013 DPF
thresholds, we would consider allowing for a 2x in-use threshold through 2018 for engines subject to
it. Excluding the issue of the more stringent DPF PM threshold, we believe we are entirely
consistent with the CARB requirements with respect to implementation dates and intermediate in-
use compliance standards.
Regarding the New Jersey comment that 2019 is too long for OBD installation, we believe
there may be some misunderstanding. In fact, we are requiring that all engines be equipped with
OBD systems that monitor the complete emission control system for model years 2013 and later.
We note that an extrapolated OBD system, as discussed in our regulation, is still a fully functioning
OBD system that monitors the entire emission control system. There are in-use flexibilities that
extend through 2018, consistent with flexibilities provided by ARB in their HDOBD program, but
these flexibilities affect the in-use threshold level (i.e., not the certification threshold level) and not
whether an engine is or is not equipped with an OBD system.
II.G.2 In-use Enforcement
Comments:
The USEPA needs to include in this rule adoption enforcement provisions specific to HDDV
OBD that include emission control warranty, recall and other in-use enforcement provisions
applicable to the certifying party. Specifically, the USEPA must propose that the party certifying the
engine and OBD system (typically, the engine manufacturer) also be the responsible party for in-use
compliance and enforcement actions. Outside of neglect and tampering, the certifying party would
be the culpable party for noncompliance identified during in-use or enforcement testing. In cases
where remedial action would be required (e.g., recall), the certifying party would take on the
responsibility for arranging to bring the vehicles back into compliance and providing state specific
repair and recall data.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 5
Agency Response:
We believe that we have made it clear in section II.G.2 of our proposed (72 FR 3253) and
final preambles that the certifying party (presumably the engine manufacturer) will be the responsible
party when and if we need to pursue in-use compliance and/or enforcement actions.
46
-------
II.H Proposed Changes to the Existing 8,500 to 14,000 Pound Diesel OBD Requirements
Comments:
We support EPA's proposal for harmonized malfunction thresholds between HD engines
certified over and under 14,000 pounds.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 8
EMA, 2005-0047-0026, p. 41
Having already certified engines in the under 14,000 pound category for the 2007 model year
does not really allow for meaningful notice and comment regarding the 2007 threshold relaxations
proposed by EPA. If, as EMA recommends, EPA makes modifications to the thresholds for engines
in >14,000 pound vehicles, the same modifications should be made to thresholds for engines in
<14,000 pound vehicles such that harmonization is maintained.
EMA, 2005-0047-0026, p. 41
Regarding the chassis certified OBD requirements (subpart S), EMA supports EPA's
proposed approach to specify vehicle-based thresholds (i.e., g/mile thresholds). This is a better
approach than that adopted by CARB which requires manufacturers to initiate a time and workload
intensive initiative to determine unique vehicle-based thresholds based on detection capability and
then to pursue approval on an individual basis. However, EMA has concerns. In some cases, the
g/mile threshold levels do not align with the corresponding g/bhp-hr thresholds (i.e., those for engine
certified systems) when compared on a multiplicative basis. This is especially true for the proposed
additive NOx thresholds, which do not consider the difference in the NOx standards between the
8,500-10,000 pound and 10,000-14,000 pound categories. For example, the additive NOx threshold
for the NOx catalyst (and other monitors) for 2010 and later is specified as +0.3 g/mile. Considering
the NOx standard is 0.2 for 8,500-10,000 pounds and 0.4 for 10,000-14,000 pounds, EPA's proposal
results in an equivalent threshold multiplier of 2.5x and 1.75x, respectively. EPA proposes more
stringent NOx and PM thresholds to most monitors in the 2013 model year, which is inconsistent
with the engine-based requirements. Of particular concern is the 2013 model year DPF threshold of
+0.04 g/mile, which results in a multiplier of 3x the PM standard of 0.02 g/mile. This compares
unfavorably with the 4x multiplier for 2010 through 2012. It also compares unfavorably with the 5x
PM multiplier for g/bhp-hr thresholds in 2013. EPA should not adopt more stringent malfunction
thresholds for any monitor effective in 2013 for NOx and PM.
EPA must finalize vehicle based g/mile thresholds that are equivalent on a multiplicative
basis to those for engines. The same feasibility issues that apply to the engine based OBD
requirements also apply to vehicle based OBDII issues. The commenter provides the following
tables of suggested thresholds.
EMA, 2005-0047-0026, p. 41-42 and Appendix B
47
-------
EMA Proposed Thresholds for 8,500-14,OOP pound Chassis-dyno Certification (g/mile)
Component/Monitor
NMHC catalyst system
NOx catalyst system
DPF system
Air-fuel ratio sensors upstream
Air-fuel ratio sensors downstream
NOx sensors
"Other monitors" with emissions
thresholds
Misfire
MY
2010-2012
2013+
2007-2009
2010+
2010-2012
2013+
2007-2009
2010-2012
2013+
2007-2009
2010-2012
2013+
2007-2009
2010-2012
2013+
2007-2009
2010-2012
2013+
2007+
NMHC
Functional
Check Only
Functional
Check Only
—
—
2.5x
2.5x
2x
2.5x
2.5x
2x
2.5x
2.5x
2x
CO
2.5x
2.5x
2x
2.5x
2.5x
2x
NOx
4x
4x
3.5x
2.5x
2.5x
3.5x
2.5x
2.5x
4x
4x
4x
3.5x
2.5x
2.5x
PM
10x
10x
5x
3x
3x
5x
5x
5x
5x
4x
4x
No requirement
Note: Boldfaced text and entries of"--" denote differences from the EPA proposal.
48
-------
EMA Proposed Thresholds for 8,500-14,OOP pound Engine-dyno Certification (g/bhp-hr)
Component/Monitor
NMHC catalyst
system
NOx catalyst system
DPF system
Air-fuel ratio sensors
upstream
Air-fuel ratio sensors
downstream
NOx sensors
"Other monitors" with
emissions thresholds
Misfire
MY
2010-2012
2013+
2007-2009
2007-2009
2010+
2010-2012
2013+
2007-2009
2007-2009
2010-2012
2013+
2007-2009
2007-2009
2010-2012
2013+
2007-2009
2007-2009
2010+
2007-2009
2007-2009
2010-2012
2013+
2010+
Std/FEL
All
All
>0.5 NOx
<=0.5 NOx
All
All
All
>0.5 NOx
<=0.5 NOx
All
All
>0.5 NOx
<=0.5 NOx
All
All
>0.5 NOx
<=0.5 NOx
All
>0.5 NOx
<=0.5 NOx
All
All
NMHC
Functional
Check Only
Functional
Check Only
—
—
2.5x
2.5x
2.5x
2x
2.5x
2.5x
2.5x
2x
2.5x
2.5x
2.5x
2x
CO
2.5x
2.5x
2.5x
2x
2.5x
2.5x
2.5x
2x
NOx
1.75x
+0.6
+0.6
1.75x
+0.5
+0.3
+0.3
1.75x
+0.5
+0.3
+0.3
1.75x
+0.6
+0.6
1.75x
+0.5
+0.3
+0.3
PM
0.10/+0.09
0.10/+0.09
0.05/+0.04
0.05/+0.04
0.03/+0.02
0.03/+0.02
0.05/+0.04
0.05/+0.04
0.05/+0.04
0.05/+0.04
—
—
—
0.03
0.03
0.03
0.03
No requirement
Note: Boldfaced text and entries of"--" denote differences from the EPA proposal.
Agency Response:
We agree that changes to the above 14,000 pound OBD thresholds, based on technological
feasibility, must also be made to the under 14,000 pound OBD thresholds. This is, in fact, what we
have done for the final rule where the engine certified thresholds are identical above and below
14,000 pounds.
Regarding the chassis based NOx thresholds being different for 10,000 to 14,000 pound
applications when considered on a multiplicative basis versus and additive basis, we disagree with
this comment. We have stated the OBD threshold as an additive threshold of the NOx FEL (or
standard) +0.6 g/bhp-hr (for engines) and +0.6 g/mi (for vehicles). Stating these thresholds on
multiplicative terms, they would be 4x the NOx FEL (for engines having a NOx FEL of 0.2 g/bhp-hr)
and 1.75x (for vehicles having a NOx standard of 0.4 g/mi). But, were we to state both NOx
thresholds on multiplicative terms, striving for consistency, we would state the vehicle-based
threshold as 4x the NOx standard, or 1.6 g/mi (for a vehicle having a NOx standard of 0.4 g/mi).
This could then be argued as being the NOx standard+1.2 which would clearly be far less stringent
than the engine-based threshold. The logic becomes circular and no right answer exists except the
level that can be done given the technology expected to be used. In the end, we believe that the
+0.6 threshold is feasible and is the appropriate threshold for vehicles in the 10,000 to 14,000 pound
range. The end result is thresholds as shown in the table below which we believe our thresholds
provide greater consistency (column labeled "EPA Result") than do EMA's (column labeled "EMA
Result"). This same argument can be made for the PM thresholds as commented by EMA.
49
-------
Weight
8.5-1 OK
10-14K
>14K
EMA versus EP/
NOx std
0.2 g/mi
0.4 g/mi
0.2
-------
thresholds are appropriate. Consistent with the 2004 EPA-CARB memorandum of agreement, we
urge EPA to strive to harmonize the federal heavy-duty OBD program with California's.
NESCAUM, 2005-0047-0024, p. 4
It is critical to establish a uniform, nationwide HDOBD program and is imperative that no
conflicts exist between the federal and CARB regulations. The trucking industry must be provided
common engines in all fifty states to control costs and maintain the variety of engine and vehicle
choices the industry needs. Without EPA leadership, disharmonized OBD requirements could
adversely lead to the California standard becoming a de facto national standard should California
requirements be more stringent than federal standards.
American Trucking Association, 2005-0047-0029, p. 2
A state-by-state patchwork of varying standards for OBD systems would be impractical and
make compliance excessively costly. Commercial trucks regularly travel interstate in the normal
course of commerce. Both scheduled and unscheduled repair and maintenance on a specific
vehicle can frequently take place in different jurisdictions. Fleets should not have to invest in
different maintenance equipment with separate diagnostic tools for trucks that may travel interstate.
Truck Renting and Leasing Association, 2005-0047-0032, p. 1
Any new EPA OBD standard should do no harm to fleet turnover, thereby helping to achieve
expected environmental benefits on a timely basis. One way to avoid undue OBD standard
complexity, and thus performance concerns and excessive costs, is to ensure harmonization. EPA
should do everything in its power to see that any final OBD standard it issues is harmonized with
California's standard. In some instances, this means EPA should strive to follow California's lead, in
other instances it means EPA should exert its authority and influence to see that California's OBD
standard is revised appropriately to reflect what is practical, achievable and feasible. California's
biennial review process and EPA's waiver approval process are but two of several venues to
address this important issue.
National Automobile Dealers Association, 2005-0047-0034, p. 2
Agency Response:
In general, we agree with commenters that a nationwide HDOBD program is critical and
most beneficial to all stakeholders. As noted in the preamble to our proposal, we cannot, at this time,
argue the appropriateness of the 2013 NOx threshold for NOx aftertreatment and the 2013PM
threshold for DPFs that are unique to the CARB regulation. For that reason, at this time, we are not
finalizing those thresholds.
The concern expressed by TRALA will not come to pass. No one will be required to invest in
different maintenance equipment or separate diagnostic tools. Both the CARB and EPA regulations
contain the same requirements for service tools and communication protocols.
Regarding the NADA comment that EPA do everything in its power to harmonize with
California by either following California's lead or exerting authority to see that California follow EPA's
lead, EPA's intent is to harmonize with California where possible and to engage California in
discussion where we cannot harmonize. That said, regarding exerting our authority, we can only do
what the Clean Air Act authorizes us to do.
51
-------
I. Are the Proposed Monitoring Requirements Feasible?
III.A Feasibility of the Monitoring Requirements for Diesel/Compression-lgnition Engines
III.A.1 Fuel System Monitoring
What We Proposed:
We proposed that fuel system malfunctions related to injection pressure, injection timing,
injection quantity, and feedback control be individually detected prior to emissions exceeding the
thresholds for "other monitors." Further, we proposed that pressure and feedback related
malfunctions be monitored continuously and that quantity and timing related malfunctions be
monitored once per trip.
Comments:
Regarding fuel injection quantity, one of EPA's suggested means of monitoring—by
delivering fuel to one cylinder at a time during a deceleration fuel cutoff condition and measuring the
effect this has on crankshaft speed fluctuation—is limited to small injected quantities of fuel and,
thus, it would not necessarily detect problems occurring at higher commanded fuel rates. EPA's
suggestion of an extension of cylinder balancing tests that manufacturers use to improve idle quality
suffers the same problem. In summary, combustion sensors (also suggested by EPA) are an
expensive and unproven technology and are infeasible due to cost a reliability, air-fuel ratio sensors
(also suggested by EPA) are unproven, and crankshaft speed fluctuation approaches are limited in
their range of applicability, if feasible at all.
Regarding fuel injection timing, EPA again suggests a crankshaft speed fluctuation approach.
Using crankshaft speed fluctuation suffers from the problems described for fuel injection quantity and
also would require the fluctuations ot be accurately correlated to crankshaft position which is difficult
for engine manufacturers to do. Heavy-duty engines typically have at least six cylinders, making
resolution of the specific contribution of individual cylinders more difficult. Also, since timing is
electronically controlled and can vary across the engine operating map, validation of timing at idle
will not necessarily confirm that it is functioning correctly elsewhere. Thus, EPA's suggested
monitoring method does not cover the entire range of engine operation. Looking for an electrical
feedback signal from the injector that would indicate injector opening/closing (another method
suggested by EPA) has not been proven by EPA and, thus, cannot be used as a basis for
demonstrating feasibility.
Regarding fuel injection pressure, non-common rail fuel systems are not able to monitor fuel
injection pressure, although they are able to monitor injection timing.
EMA, 2005-0047-0026, p. 14-15
Agency Response:
The comments raised issues on the techniques of monitoring of the fuel system with
limitations that do not cover the entire engine operating range and the ability to monitor fuel injection
pressure on non-common rail fuel systems. Additional comments include the request to change
threshold monitoring to functional monitoring for non-common rail fuel systems and limiting the
52
-------
scope of threshold monitoring to those failure modes that are detectable during idle and deceleration
fuel cutoff.
The current proposal requires "continuous" threshold monitoring for fuel system pressure and
threshold monitor "once per drive cycle" for fuel system quantity and timing. The requirements for
fuel system monitoring does not differ for "Common Rail" (CR) or "Electronic Unit Injector" (EUI) type
systems.
The differences between CR and EUI fuel system hardware cause distinct system
advantages and disadvantages along with individual system operating monitoring challenges. By
design, CR systems are able to accurately monitor fuel pressure directly and measurement of
quantity and timing during most engine operation conditions. Some EUI systems are not able to
independently monitor pressure, quantity or timing but can monitor the overall fuel system
functionality by crankshaft speed fluctuation during steady state events (e.g. idle or decel conditions)
or other methods. EUI systems by design are able to achieve higher fuel injection pressures and are
critical to lowering engine out emissions in future engine designs.
For some of the major emission control systems and components, the proposed heavy-duty
OBD regulation requires malfunctions to be identified before any problem becomes serious enough
to cause vehicle emissions to exceed the standards by a certain amount above the threshold. EPA is
proposing more stringent emission thresholds for major components and systems (e.g., EGR and
fuel system) located upstream of the aftertreatment as the aftertreatment is expected to compensate
for some of the emission increase caused by a deteriorated emission control component, thereby
reducing the actual impact on tailpipe emissions even though a failed component exists. Specifically
for fuel system, reduced fuel pressure causes increased particulate matter emissions which may be
captured by the downstream particulate trap.
In lieu of these system differences, the regulation has been modified to reflect the differences
of CR and EUI systems based upon the capabilities deemed feasible for current and future fuel
system and component designs.
For common rail type fuel systems, the 2010 regulation will remain at "continuous" threshold
monitoring for fuel system pressure and at least "once per drive cycle" for both quantity and timing
monitors. The "once per drive cycle" monitoring is achievable during conditions such as deceleration
operation. EPA agrees with the comment that a non-failure detection at one operating point does not
necessarily indicate a lack of failure at a different operating point. Therefore, in 2013, the regulation
will remain at "continuous" threshold monitoring for fuel system pressure but be changed to "when
conditions are met" for fuel system quantity and timing. This will incorporate more monitoring points,
where feasible, which will be more effective at capturing failures at various operation points but does
not require monitoring in non-feasible operating areas. The manufacture has the option to combine
the quantity and timing monitors into a single monitor for both 2010 and 2013 regulations as both
indicate a failure of the fuel injector. The allowance to combine malfunctions was part of our
proposal and remains part of our final rule (see proposed and final §86.010-18(a)). That said, we
have added new text in §86.010-18(g)(1) to clarify this fact and highlight where the Administrator is
open to such combining by manufacturers. We would still expect the manufacturer to demonstrate
that their approach is robust and effective at detecting malfunctions.
For EUI systems, the 2010 regulation has been changed to require functional monitoring
"once per drive cycle" for fuel system pressure, quantity and timing. We are making this change
based on current OBD system capability. As noted above, the manufacturer has the option to
combine the pressure, quantity and timing monitors into a single monitor as both indicate a failure of
the fuel injector (see §86.010-18(a)). This addresses the issue of discrete pressure monitoring on
EUI systems while ensuring system integrity. In 2013, to be consistent with CR type systems and
53
-------
because we believe manufacturers may add additional sensors and will certainly gain experience
with monitoring strategies, the regulation will be remain consistent with our proposal by requiring
threshold monitoring. Further, we have changed the monitoring conditions to require monitoring
"when conditions are met" for fuel system pressure, quantity and timing. The continuation of the
option to combine the monitoring of pressure, quantity and timing into one system monitor remains in
place (see §86.010-18(a)). This will incorporate more monitoring points, where feasible, to capture
additional failures at various operation points but does not require monitoring in non-feasible
operating areas.
In summary, EPA believes that robust fuel system monitoring is necessary and can be
accomplished by a combination of both system parameter and comprehensive component
monitoring during conditions that can be monitored. It is anticipated that manufacturers will design
fuel system monitoring strategies that accurately control and monitor fuel flow, both directly and
indirectly, under both transient and steady state load conditions to meet the heavy-duty OBD
regulation for both 2010 and 2013.
III.A.2 Engine Misfire Monitoring
What We Proposed:
We proposed that, for 2010-2012, a continuous engine misfire be detected during engine idle.
For 2013 and later, we proposed that engines equipped with combustion sensors monitor
continuously for misfire during the full operating range and detect a malfunction prior to emissions
exceeding the thresholds for "other monitors."
Comments:
Diesel engine misfire occurs in two different ways:
Improper fuel injection which can result from an insufficient quantity of fuel injected,
inadequate fuel atomization, or mistimed fuel injection; and/or,
Failure of the fuel to auto ignite upon proper fuel injection which can result from
inadequate compression of the air-fuel mixture within the cylinder.
The commenter notes that the latter of these two causes of misfire—failure of the fuel to
auto-ignite—is likely due to poor fuel quality rather than any failed component and argues that poor
fuel quality is not an engine failure and, thus, should not result in any MIL illuminations. The
commenter further argues that misfire under non-idle conditions is rare and presents noticeable
performance problems that would cause operators to promptly seek corrective action.
Under most operating conditions, diesel misfire does not result in a significant increase in
emissions. If misfire results from a failure to inject any fuel, the emission impact is negligible.
Misfire from poor atomization due to low injection pressure, mistimed injection or inability to auto-
ignite results in increased engine-out HC emissions, but these are oxidized by the catalytic exhaust
aftertreatment systems.
As a practical matter, the tailpipe HC emissions are only significantly increased by misfire
during cold conditions before the engine warms up to normal operating temperature. Misfire is most
prevalent and is most likely to have a measurable effect on tailpipe emissions only when the engine
is warming up under idle conditions. OBD monitoring of misfire is not justified by these insignificant
emissions.
Robust detection of misfire is very difficult. Under light-load or idle conditions where misfire
is most prevalent, the torque pulses are weak. Crankshaft speed fluctuations associated with torque
pulses depend on the rotating inertia of the vehicle driveline which may be influenced by
54
-------
engagement/disengagement of engine-driven devices such as fans, air compressors, air
conditioning systems, etc. Thus, detection of misfire would likely require unique calibrations for each
driveline resulting in substantial costs and possibly false failures.
EPA erroneously presumes that there is a method that a manufacturer can use to induce a
regular pattern of misfires during emission testing (by requiring that thresholds be determined via
defining the percentage of misfires that would result in emissions that exceed the OBD threshold).
Emissions threshold monitoring for misfire is problematic because there is no reasonable approach
for inducing a regular pattern of misfires and, hence, no means to correlate a misfire percentage with
an exceedance of a threshold. EPA has failed to demonstrate feasibility for diesel engines.
EPA's proposed misfire monitoring requirements will have an impact on vehicles and vehicle
manufacturers requiring changes to the way vehicle manufacturers do business to ensure that OBD
systems work properly. The monitoring requirements will force engine manufacturers to make
system changes that will force vehicle manufacturers to make vehicle changes thereby raising the
costs of OBD-equipped vehicles. Meanwhile, non-OBD engines will be available at lower cost
leaving OBD engines at a competitive disadvantage. As a result, EPA's stated goal of gaining
experience with 2010-2012 engines may be substantially defeated. EPA must avoid that result by
eliminating all misfire monitoring requirements for diesel engines.
EMA, 2005-0047-0026, p. 15-18
Agency Response:
We do not agree with EMA where it pertains to misfire both from an emission impact
perspective and a detection perspective. The emissions impact of misfire on cold starts can be
significant. Misfire on a warmed engine can damage the after treatment system if they occur in a
large enough quantity.
We also do not agree with the implication that misfire due to poor fuel quality should not be
detected. If a vehicle is misfiring a driver needs to be notified. Assuming the poor quality fuel
misfiring is at a rate great enough to set a MIL, the driver should be taking the vehicle in for repair.
The repair will be to drain the fuel. It is appropriate to have MIL illuminate provided there is
something to fix.
EMA is concerned that robust detection is very difficult. The requirement for 2010-2012 is to
be able to find a completely dead cylinder in 1000 crank revolutions. We believe this requirement is
fair and can be accomplished for 2010.
EMA is concerned that it will not be able to induce a random pattern of misfire. Misfire
generation boxes have been created for many years. The concept of randomly shutting off fuel to a
cylinder to induce misfire has been around for many years. We do not believe it is beyond
manufacturers capabilities to create a misfire generator.
Regarding OBD equipped engines being at a competitive disadvantage, we disagree. In fact,
we believe that an OBD system on an engine can be competitive advantage. OBD can require less
costly repairs because catching failures early can prevent damage to the aftertreatment systems,
and other expensive components. Also, mandatory emission compliance can lead to fines in some
states if vehicles are not maintained. A vehicle with a well designed OBD system can be a real cost
saver for the purchaser.
III.A.3 Exhaust Gas Recirculation (EGR) Monitoring
What We Proposed:
55
-------
We proposed that malfunctions of the EGR system related to low flow, high flow, slow
response, feedback control, and cooler performance be detected prior to emissions exceeding the
thresholds for "other monitors." Further, we proposed that flow and feedback related malfunctions
be monitored continuously, response related malfunctions be monitored whenever conditions were
met, and that cooler malfunctions be monitored once per trip.
Comments:
EPA has not sufficiently justified the need for continuous monitoring for EGR high and low
flow. Continuous monitoring for a system - as opposed to a component under comprehensive
component monitoring - has not even been defined. For components, continuous monitoring is an
absolute requirement only for out-of-range and circuit failures which are problems that can be
detected at any engine operating condition. Continuous monitoring is not feasible for EGR high and
low flow monitors and is meaningless at certain engine operating conditions because a low or zero
EGR rate is commanded thereby providing slim or no opportunity for assessment. During transient
operation, the errors of estimated EGR - used for diagnostics - become high because of the
dynamics of the system and of the sensor response time making such monitoring prone to errors
and false Mils.
EMA, 2005-0047-0026, p. 20
Agency Response:
The commenter raised issues on continuous monitoring of the exhaust gas recirculation
(EGR) flow and recommended the monitoring requirements for EGR high and low flow be changed
to "whenever the entry conditions are met" rather "continuous monitoring" as proposed.
The current proposal requires continuous threshold monitoring for EGR low and high flow
conditions, "when conditions are met" threshold monitoring for EGR slow response, and "once per
drive cycle" threshold monitoring for EGR cooler slow response.
For some of the major emission control systems and components, the proposed heavy-duty
OBD regulation requires malfunctions to be identified before any problem becomes serious enough
to cause vehicle emissions to exceed the standards by a certain amount exceeding the threshold.
EPA is proposing more stringent emission thresholds, where feasible, for major components and
systems (e.g., EGR and fuel system) located upstream of the aftertreatment as improved
aftertreatment systems are expected to compensate for some of the emission increase caused by a
deteriorated emission control component. These thresholds will allow earlier detection of the effected
component to reduce excessive emissions during deteriorated conditions and possibly reduce
damage to the aftertreatment system. Specifically for EGR system design, excessive EGR flow
causes increased PM emissions, and insufficient EGR flow causes increased NOx emissions.
To determine the necessary EGR flow rates and control EGR flow, EGR systems normally
use the following components: an EGR valve, valve position sensor, boost pressure sensor, intake
temperature sensor, intake (fresh) airflow sensor, and tubing or piping to connect the various
components of the system. EGR temperature sensors and exhaust backpressure sensors are also
commonly used. EGR is not a stand alone emission control device; it is carefully integrated with the
air handling system.
Understanding the limitations of "continuous monitoring" for various systems affecting
emissions, including the EGR system, EPA has clarified the continuous monitoring requirements
within the EGR monitors to allow disabling of system monitoring during operating conditions that are
56
-------
determined to be technically necessary to ensure robust detection of malfunctions. System
monitoring is expected to occur under conditions which may be reasonably expected to be
encountered in normal vehicle operation and use. When continuous system monitoring is suspended,
the system must continuously detect malfunctions of input components as specified in the regulation
under the comprehensive component monitoring requirements. This will ensure the system is
continuously monitored by either direct system monitoring or indirect component monitoring during
all operating conditions. Importantly, such monitoring requirements can be and are already being
met by the industry.
The EGR monitoring conditions, as proposed, contained a provision for disabling of
continuous EGR monitoring under certain conditions (see proposed §86.010-18(g)(3)(iii)(D)). For
the final rule, we have made slight changes to those provisions meant to provide greater clarity
without changing the content or intent of the provision. The language now allows the manufacturer
to request Administrator approval to disable temporarily the EGR system monitor(s) under specific
ambient conditions (e.g., when freezing may affect performance of the system) or during specific
operating conditions (e.g. during transient, extreme low or high flow conditions). The manufacturer
must be able to demonstrate via data or engineering analysis that a reliable system monitor cannot
be run when these conditions exist because it cannot distinguish robustly between a malfunctioning
system and a properly operating system. The manufacturer is still required to maintain
comprehensive component monitoring requirements as required.
In summary, we believe that EGR system continuous monitoring is necessary and can be
accomplished by a combination of both system and comprehensive component monitoring. It is
anticipated that manufacturers will design EGR systems that accurately, continuously control and
monitor EGR flow, both directly and indirectly, under both transient and steady state load conditions
to meet the heavy-duty OBD regulation. With continuous system monitoring clarification, the
proposal requiring continuous threshold monitoring for EGR low and high flow conditions, "when
conditions are met" for EGR slow response and "once per drive cycle" for EGR cooler slow response
will remain unchanged.
III.A.4 Turbo Boost Control System Monitoring
What We Proposed:
We proposed that malfunctions of the boost control system related to underboost, overboost,
variable geometry slow response, feedback control, and undercooling be detected prior to emissions
exceeding the thresholds for "other monitors." Further, we proposed that underboost, overboost,
and feedback related malfunctions be monitored continuously, that slow response related
malfunctions be monitored whenever conditions were met, and that undercooling related
malfunctions be monitored once per trip.
Comments:
EMA believes that continuous monitoring of boost is neither necessary nor feasible.
Continuous boost monitoring would suffer from errors due to the dynamics of the system and sensor
response time during transients. Boost monitoring should not be done continuously, but should be
done whenever entry conditions are met.
EMA, 2005-0047-0026, p. 20
Agency Response:
57
-------
The commenter proposes Turbo Boost monitoring should be preformed "whenever entry
conditions are met" instead of "continuously" and states that continuous monitoring is neither
necessary nor feasible. Also, EMA comments that some boost systems do not control to a desired
boost.
The current proposal requires continuous threshold monitoring for turbo boost- under and
over boost conditions, "when conditions are met" threshold monitoring for VGT slow response, and
"once per drive cycle" threshold monitoring for charge air cooler slow response.
Proper boost control is essential to optimize emission levels. Even short periods of over- or
under-boost can result in undesired air-fuel ratio excursions and corresponding emission increases.
Additionally, the boost control system directly affects exhaust and intake manifold pressures.
Another critical emission control system, EGR, is very dependent on these two pressures and
generally uses the differential between them to force exhaust gas into the intake manifold. If the
boost control system is not operating correctly, the exhaust or intake pressures may not be as
expected and EGR system may not function as designed.
Understanding the limitations of "continuous monitoring" for various systems affecting
emissions, including the fuel system, EPA has clarified the continuous monitoring requirements for
specific monitors to allow disabling of system monitoring during operating conditions that are
determined to be technically necessary to ensure robust detection of malfunctions. System
monitoring is expected to occur under conditions which may reasonably be expected to be
encountered in normal vehicle operation and use. When continuous system monitoring is suspended,
the system must continuously detect malfunctions of input components as specified in the regulation.
This will ensure the system is continuously monitored by either direct system monitoring or indirect
component monitoring during all operating conditions and can be met by the industry.
This clarification is made, in the final rule, by adding a new provision for the turbo boost
monitoring conditions that provides greater clarity without changing the intent of the proposal. The
language now allows the manufacturer to request Administrator approval to disable temporarily the
turbo boost system monitor(s) under during specific operating conditions (e.g. during transient,
extreme low or high flow conditions). The manufacturer must be able to demonstrate via data or
engineering analysis that a reliable system monitor cannot be run when these conditions exist
because it cannot distinguish robustly between a malfunctioning system and a properly operating
system. The manufacturer is still required to maintain comprehensive component monitoring
requirements as required.
In addition, addressing the concern on systems that do not have a boost pressure system,
the regulation has been changed to reflect that the OBD system must detect a malfunction of the
boost pressure control system prior to a decrease from the manufacturer's commanded boost
pressure, or expected boost pressure instead of commanded boost pressure on engines not
equipped with a boost pressure control system.
In summary, we believe that turbo boost system continuous monitoring is necessary and can
be accomplished by a combination of both system and component monitoring. It is anticipated that
manufacturers will design turbo boost systems that accurately, continuously control and monitor
turbo boost flow, both directly and indirectly, under both transient and steady state load conditions to
meet the heavy-duty OBD regulation. With continuous system monitoring clarification, and the
statement to include engines with no boost systems, the proposal of continuous threshold
monitoring for turbo boost for both under and over boost conditions, "when conditions are met" for
VGT slow response and "once per drive cycle" for charge air cooler slow response will remain
unchanged.
58
-------
III.A.5 Non-Methane Hydrocarbon (NMHC) Converting Catalyst Monitoring
What We Proposed:
We proposed that malfunctions related to NMHC conversion efficiency be detected prior to
emissions exceeding the thresholds for "NMHC catalyst." We also proposed that, should the NMHC
converting catalyst be used to assist other aftertreatment devices, that malfunctions be detected if
that assistance is no longer occurring. Further, we proposed that conversion efficiency and
aftertreatment assistance be monitoring once per trip.
Comments:
There is not monitoring technology available to meet the proposed DOC threshold monitoring
requirement. A DOC cannot be monitored for oxygen storage because they have no ability to store
oxygen. DOCs do not need to store oxygen because there is always excess oxygen in diesel
exhaust.
The only monitoring technology available to monitor the DOC is temperature sensing of
catalyst exotherms. Such monitoring cannot differentiate between a good DOC and one
deteriorated to 2.5x the NMHC standard. The commenter references SAE paper 2005-01-3602
which highlights the limited feasibility of the catalyst temperature monitoring approach and
demonstrates that the separation between a good catalyst and a threshold catalyst using the
approach is very poor and would result in both false Mils and undetectable failures. The
commenter then argues that, given the likely increase in engine-out NMHC emissions from 2010
engines due to the lower NOx standard, and the resultant higher-efficiency DOC needed to meet the
0.14 g/bhp-hr NMHC standard, the ability to meet the proposed requirements via a functional
monitor is unlikely (since threshold monitoring would be required).
EMA, 2005-0047-0026, p. 13-14
Agency Response:
We agree with the EMA's position and the conclusions of SAE 2005-01-3602 (Diagnostics
for Diesel Oxidation Catalysts) in that temperature- and exotherm-based monitoring strategies do not
provide the accuracy necessary to detect a DOC which fails to convert NMHCs at a 2.5x threshold
level. We also believe that a strategy of intentional fuel dosing to create an exotherm in the DOC -
for the purpose of monitoring DOC performance, - is not an accurate method for determining a failed
part, and execution of such a strategy would needlessly increase tailpipe emissions.
Since the DOC is used to create an exotherm which enables DPF regeneration (exotherm
created by conversion of excess HCs in exhaust) and to provide an optimum ratio of NO-to-NO2 in
the exhaust feedgas (through oxididation of NO to NO2 ) for proper SCR function, failure of this
device will be manifested in the reduced performance of the DPF and SCR catalyst (as these
devices which rely on robust HC and NO oxidation to meet emission standards). For example, if the
DOC fails to create a sufficient exotherm for DPF regeneration, the trap will soon plug with soot, and
the DPF monitor for incomplete regeneration will set a fault. In a similar manner, if the DOC cannot
provide the proper NO-to-NO2 ratio in the exhaust for proper SCR function, the SCR catalyst
monitor NOx will set a fault.
Given that current monitoring technologies cannot detect a threshold failure of a DOC, and
that monitors for components downstream of the DOC will mature faults should the expected
performance of the DOC deteriorate, we will not require threshold monitoring of NMHC conversion
59
-------
over the NMHC conversion catalyst. However, note also that we believe our new DOC monitoring
requirement - which requires detecting the ability of the DOC to achieve a 100 degree C exotherm
(temperature increase) within 60 seconds, achieve the necessary DPF regeneration temperature,
and maintain this temperature throughout the regeneration event - will serve the purpose of
detecting malfunctions of the DOC associated with the DPF regeneration function.
III.A.6 Selective Catalytic Reduction (SCR) and NOx Conversion Catalyst Monitoring
What We Proposed:
We proposed that malfunctions related to conversion efficiency, active/intrusive reductant
delivery, active/intrusive reductant quantity, active/intrusive reductant quality, and feedback control
be detected prior to emissions exceeding the thresholds for "NOx catalyst system." Further, we
proposed that conversion efficiency and reductant quality be monitored once per trip and that
reductant delivery, quantity, and feedback control be monitored continuously.
Comments:
EPA assumes that NOx sensors will be used by all engine manufacturers to control NOx
engine and aftertreatment systems in 2010. EPA further assumes that a viable NOx sensor will be
available for 2010 production engine systems. Those assumptions are not accurate. It is not
necessarily the case that all engine manufacturers will use NOx sensors for underlying NOx
emissions control for 2010. If engine manufacturers can control engine aftertreatment system
emissions without NOx sensors, manufacturers will opt not to use them. If compliance with the
underlying emission standards does require a NOx sensor, manufacturers likely would apply the
sensor in a position where accuracy limitations can be overcome - upstream of the NOx catalyst
where NOx levels are higher but the operating environment is more aggressive. Control of NOx
aftertreatment might be based on an estimate of NOx produced by the engine. In that case,
manufacturers would avoid adding a costly NOx sensor if performance of the system is adequate.
For example, for urea SCR systems, urea solution is metered in proportion to the NOx flowing into
the SCR catalyst. An estimate of NOx produced by the engine would be used for this parameter. For
lean NOx trap systems, the regeneration of the NOx trap is done when storage of NOx in the
catalyst nears a certain level relative to trap capacity. An estimate of NOx produced by the engine
would be used to calculate how much NOx is stored in the catalyst. (While NOx is being stored,
positive and negative errors in NOx estimate will tend to cancel as they are summed together.)
If NOx sensors are used for underlying engine-aftertreatment system control, engine
manufacturers would use a NOx sensor of a different range and resolution/accuracy than that
required for OBD monitoring. Current NOx sensors do not have the much narrower range and far
greater accuracy that would be required for OBD monitoring purposes, nor have they been shown to
have the necessary durability for OBD monitoring.
Engine manufacturers have not seen the 2006 NOx sensor accuracy and durability
improvements upon which EPA is relying for its proposed NOx threshold. Indeed, what engine
manufacturers know to date is that there is one NOx sensor supplier that is advertising a target
accuracy of 10%. That "target" is only a target, and has not been proven. And it is based on the
supplier's experience with Euro 4 SCR engines, with only limited experience on EGR-equipped
engines. Engine manufacturers have not completed long-term EGR evaluations on NOx sensors.
Engine manufacturers also believe that the application and placement of NOx sensors will have an
impact on accuracy. The data available to engine manufacturers shows poor accuracy on a 1500
ppm-range sensor, which is a far broader range than would be required for OBD-sensing technology,
and is limited to use on a light-duty engine at 120,000 miles, not the full 435,000- mile useful life
required for heavy-duty engines, much less the full operating life of the engine.
With regard to durability, the picture is similar. Engine manufacturers' best data shows a
target- an unproven target- of 185,000 miles durability. That number falls far short of the 435,000-
60
-------
mile-plus durability needed for heavy-duty engines. And mileage durability is not the only factor.
Even sensors that are not in use, but are sitting on a shelf waiting to be installed, lose durability,
leading sensor manufacturers to consider very limited time warranties on sensors. Added to that are
the problems with use in the real world, including temperature limitations for the sensor control
module and packaging and wiring requirements that do not reflect actual operating conditions, as
well as problems with potential electromechanical interference from the remainder of the engine
system. Research has been conducted at Southwest Research Institute on the durability of NOx
sensors. Again, the limited data - limited to 6,000 hours of use - show insufficient accuracy and
decreases in sensor response rate at low hours.
Finally, the currently available NOx sensors also are very sensitive to transients and do not
provide reliable results during transient operation. Therefore, the FTP cycle is a poor choice for
monitoring the NOx aftertreatment system as it contains almost no steady-state points. During most
of the FTP cycle, the NOx sensor returns no values. Steady-state operation is much more common
for heavy-duty applications than for light-duty, so monitoring strategies based on the steady-state
test cycle should be acceptable for NOx aftertreatment and NOx sensor certification testing.
NOx sensors under development with the accuracy necessary to meet the stringent OBD
requirements will be very expensive, and will require validation to demonstrate capability to maintain
the required long-term accuracy with good reliability. Much more work must be done, and significant
improvements in NOx sensor technology must be made, before such sensors can be used for
reliable monitoring of NOx emissions for OBD purposes in time for 2010. Indeed, recent research
reveals that the accuracy of current NOx sensor technology is not capable of achieving the proposed
EPA NOx emissions thresholds requirements (see, "Threshold monitoring of urea SCR Systems,"
SAE Paper # 2006-01-3548). At best, current sensor technology would allow measurement to a
threshold at just below 4x the NOx standard of 0.20. And EPA has failed to provide a reasonable
basis upon which to make a determination that an accurate and durable NOx sensor is projected to
be available in time for 2010. Nor has EPA adequately addressed the material and development
costs for developing an accurate and reliable NOx sensor that can meet the thresholds proposed in
the NPRM.
In sum, the NOx aftertreatment threshold that EPA has proposed is not technologically
feasible. Manufacturers' development and validation activities for providing commercially available
products in 2010 require that NOx sensor technology be available within the next few months.
Manufacturers cannot wait for potential performance improvements in NOx sensor accuracy.
EMA, 2005-0047-0026, p. 10-12
American Trucking Association, 2005-0047-0029, p. 3
The trucking industry cannot afford to be burdened with faulty or over-sensitive sensor
monitoring due to EPA's technology-forcing mandates. Regarding the OBD thresholds for NOx
aftertreatment, NOx sensors must be accurate and durable enough to provide predictable results for
monitoring throughout an engine's actual life. Sensor technologies must be sound and thoroughly
tested by the regulatory deadlines and, if not, EPA must adjust the implementation dates under the
rule.
American Trucking Association, 2005-0047-0029, p. 3
Truck Renting and Leasing Association, 2005-0047-0032, p. 2-3
Agency Response:
We agree with commenters that detection of NOx aftertreatment malfunctions at the
proposed threshold levels will be very difficult. The comments focus primarily on the existence of an
acceptable NOx sensor for monitoring NOx aftertreatment devices and the need to raise the
threshold from the proposed level due to the inaccuracy of the NOx sensors expected to be available.
We are somewhat concerned about the possibility of engines with NOx aftertreatment devices and
61
-------
perhaps no sensors capable of managing those devices, let alone detecting malfunctions in those
devices.
The feasibility of meeting the proposed OBD requirements for monitoring of NOx
aftertreatment is not predicated on the successful development of NOx sensors. In our proposal, we
discussed monitoring methods for NOx adsorbers that use wide-range air-fuel sensors, not NOx
sensors, and these sensors are already being used on diesel applications less than 14,000 pounds.
These sensors can be used to quantify the performance of a NOx adsorber and determine if it is
above or below the proposed emission thresholds. This technique is very similar to a monitor of a
NOx adsorber currently being used on a light-duty gasoline vehicle today and we are not aware of
technical limitations that would prevent such a technique from also being used successfully on a
heavy-duty diesel application.
For SCR systems, we believe that a robust NOx sensor will be required not only to meet
proposed OBD requirements, but to meet the 2010 emissions standards. We believe that a NOx
sensor will be needed to measure accurately the real-time NOx emission levels and determine the
proper amount of reductant (urea) dosing. Manufacturers have argued that a NOx sensor may be
available for use as an SCR system control input. Such a sensor would be placed upstream of the
SCR catalyst to measure NOx levels to determine the needed urea injection, but would not be
placed downstream of the catalyst to provide any form of feedback to the control system.
Manufacturers have further argued that an upstream sensor, where NOx levels are higher, being
used for control purposed would not require the accuracy as would be required of a downstream
sensor used as feedback and/or for OBD purposes. While this may be true, we question the
effectiveness of, as well as manufacturers' willingness to implement, a NOx control strategy that
meters urea based on a NOx sensor upstream of the SCR catalyst with no regard for urea and/or
NOx levels emitted downstream of the device. At minimum, it seems that a NOx sensor would be
placed downstream of the catalyst which, when compared to the upstream sensor, could indicate a
lack of NOx conversion activity. This could trigger a dashboard light and cease urea injection which
would save the driver money and eliminate urea emissions into the environment. In the end, the
ability of a downstream sensor to detect at various NOx levels - whether those proposed by EPA or
those suggested by EMA - is of secondary concern to the primary concern of excess NOx and/or
urea emitted from the tailpipe. We believe that there must be some assurance that the SCR catalyst
is indeed performing at least some measure of its intended purpose.
Regarding NOx sensor accuracy, the EMA comment states that NOx sensors may become
durable and accurate enough for SCR control when located upstream of the SCR catalyst, but they
still will not be accurate enough for OBD purposes in the lower concentration environment
downstream of the catalyst. However, as suggested above, we believe that a system without a
feedback mechanism is risky and unlikely. First, we do not believe a manufacturer will be able to
robustly and reliably meet the NOx emission standards for the full useful life with only an upstream
NOx sensor and no ability to adapt, adjust, or compensate for degradation in the SCR catalyst
during the useful life. Catalyst degradation could be substantial over a period of 435,000 miles and
failure to compensate for it will likely result in over or under-dosing of reductant and, consequently,
higher emissions. Second, a sensor that is accurate enough to measure upstream NOx levels during
a variety of engine speeds and loads is also likely to have enough accuracy to measure downstream
NOx levels during a portion of engine speeds or loads where engine out NOx emissions are higher.
Monitoring during this subset of operation could be sufficient to properly detect malfunctioning SCR
systems. Third, other monitoring techniques could be used including intrusive strategies that
temporarily interrupt (or hold constant) reductant injection to assess the performance level while
exposing the downstream sensor to higher concentrations.
Ultimately, this is more than an OBD issue - it is an emissions compliance and emissions
standard feasibility issue. For SCR to be successful, it must first be capable of meeting the emission
62
-------
standards robustly through the useful life. As such, a NOx sensor seems necessary to properly
correct and compensate for catalyst deterioration over the useful life period. Techniques that use
SCR without a NOx sensor do not appear to be viable or robust enough to reliably meet the stringent
NOx emission standards for the useful life of 435,000 miles. Accordingly, if such a sensor becomes
available to make the SCR technology viable, then such a sensor is also available to perform some
level of the necessary OBD monitoring.
The question then becomes, what is the appropriate level of OBD monitoring or, in other
words, what is the appropriate threshold level? In the context of our heavy-duty highway in-use test
program, we have conducted extensive testing using portable emissions measurement systems
(PEMS) and have learned that, for model years 2010 through 2012, the PEMS devices are capable
of detecting emissions levels of roughly 0.44 g/bhp-hr. If we were to add that to the NOx standard of
0.2 g/bhp-hr, we would arrive at a level of 0.64 g/bhp-hr. On one hand, it might seem inappropriate
to require OBD detection at 0.5 g/bhp-hr- knowing that the OBD monitor is not actually measuring
emissions levels but is instead measuring NOx and oxygen concentrations in the exhaust and
correlating them to a g/bhp-hr level - when an actual emissions measurement device can only
detect at 0.64 g/bhp-hr. However, an important distinction exists in these numbers that must be
considered. With the PEMS device, we are measuring actual tailpipe emissions in the field. It is
hard enough to measure emissions in a test cell with laboratory based computers and constant
volume sampling and other sophisticated laboratory equipment. It is another level of difficulty to take
that into the field and measure emissions on an actual truck as it drives down the road. OBD
monitoring can be considered to be more analogous to the laboratory situation. Engineers are
monitoring engine and exhaust system performance characteristics in a laboratory setting and
correlating those to an emission level. Once those correlations are determined, calibrations are
developed that look for those engine and exhaust system performance characteristics during real
world operation. If those performance characteristics are detected, a malfunction is presumed to
exist and a diagnostic trouble code of some sort is stored. For that reason, we believe that OBD
thresholds should in fact be set at stringent levels since their design is based on laboratory settings.
Nonetheless, emissions levels may in fact be somewhat higher in the field than would be expected
upon detecting the correlated levels programmed into the OBD calibration. This is particularly true
during the initial years of implementation. For that reason, we have included in-use compliance
levels that are double those required of certification demonstration engines. Further, we have
included in-use compliance provisions for most OBD engines (those having "extrapolated" OBD
systems) that are not even evaluated against actual emissions levels (see §§86.010-18(p), 86.013-
18(p),and86.016-18(p)).
Manufacturers have also raised concerns over NOx sensor durability. We note that the
emission standards do not mandate that the sensor be able to last for the full useful life of 435,000
miles. The regulations allow engine manufacturers to have maintenance or service/replacement
intervals and require periodic replacement of the sensor. If a NOx sensor cannot reliably last for
435,000 miles, a manufacturer could pursue this option and rely on periodic replacement of the
sensor. Such replacement would not be for OBD purposes, but for continued compliance with the
emission standards. This is not unlike conventional oxygen sensors in the early years of light-duty
gasoline OBD implementation. Early sensors had service intervals of just 30,000 miles and,
gradually, sensor performance and durability has improved to the point that today's sensors
generally last more than 100,000 miles without any required replacement intervals. NOx sensor
technology will likely follow the same path, with durability and performance improving over time and
providing longer and longer time intervals between replacements.
We believe that it is not currently possible to know what level of threshold is the most
appropriate level. Manufacturers appear to focus their efforts on the other aspects of the emission
control system - the engine design, fuel strategy, aftertreatment devices, regeneration strategies -
and worry about the OBD system only after those design elements have been decided upon. In
63
-------
fairness, we cannot necessarily argue that the limited resources be used primarily for controlling the
emissions from the tailpipe. But, that suggests that no one today can know what future OBD
systems can or cannot do given that the emission control systems they will be monitoring have not
yet been determined. After years of experience with OBD and tracking its development and
capabilities, it has become clear that, despite our sense that OBD should be a critical element to the
design of an emission control system, it is often the last element considered by manufacturers.
Given the wide variety of approaches manufacturers are expected to take in meeting
emissions standards, it is likely that different OBD systems and different OBD monitoring strategies
will be capable of detecting malfunctions at different emissions levels. While it appears that current
technology is capable only of detection at 0.8 g/bhp-hr or 4x the standard - the level suggested by
EMA - it seems unreasonable to take a present day detection level and make that the threshold
level for all future engines. In the end, we believe that the NOx aftertreatment threshold level
suggested by EMA is the most appropriate level for 2010 through 2012, but that the threshold level
we proposed is most appropriate for 2013 and later. This strikes the proper balance between forcing
manufacturers to design effective OBD monitoring strategies while also providing the flexibility to
certify systems when good faith efforts fail to comply fully with the requirements.
III.A.7 NOx Adsorber Monitoring
What We Proposed:
We proposed that malfunctions related to adsorber system capability, active/intrusive
reductant delivery, and feedback control be detected prior to emissions exceeding the thresholds for
"NOx catalyst system." Further, we proposed that adsorber capability be monitored once per trip
and that reductant delivery and feedback control be monitored continuously.
Comments:
See entries under III.A.6.
Agency Response:
See our response to comments under section III.A.6.
III.A.8 Diesel Particulate Filter (DPF) Monitoring
What We Proposed:
We proposed that malfunctions related to the DPF filtering performance, regeneration
frequency, regeneration completion, NMHC conversion, active/intrusive reductant injection, and
feedback control be detected prior to emissions exceeding the thresholds for "DPF system." We also
proposed that a missing DPF substrate be detected. Further, we proposed that all of these functions
be monitored whenever conditions were met.
Comments:
The technology that EPA has identified for DPF monitoring is not workable, and the
proposed thresholds are not feasible. EPA must relax the emission thresholds that have been
proposed and allow more time to assess the feasibility of such a stringent emission threshold in the
future.
64
-------
Substantial uncertainties are associated with the EPA's proposed DPF monitoring method.
Those uncertainties make it impossible to use such a method to reliably detect whether and at what
level a threshold has been exceeded. Those uncertainties exist in three primary areas: the soot
accumulation model; deducing accumulated soot from a delta pressure measurement; and,
correlation of PM emissions with disparity between modeled soot loading and soot loading deduced
from DPF pressure drop measurements. The commenter provides many sources of uncertainty
within each of these three primary areas, including variation in fuel characteristics (Biodiesel, Cetane,
sulfur, aromatics, additives, impurities, etc.), assessing completeness of active regenerations (active
regenerations may be interrupted before going to completion), a lack of correlation between
pressure drop and filter soot loading due to variation in permeability and uniformity of soot layer,
imprecision or uncertainty in assessing exhaust flow rate (influenced by vehicle design), low
sensitivity of pressure drop to soot load (see Appendix A, Figure 2), imprecision of pressure drop
measurement (influenced by vehicle exhaust system design), the confounding effects of ash loading,
and correlating PM emissions with the disparity between modeled soot loading and soot load
deduced from DPF pressure drop measurements.
The commenter acknowledges that some of the variabilities noted above are factors that
engine manufacturers can unilaterally control and will control to the best of their ability such as drift
and engine-to-engine differences in fuel injection rates, timing and pressure; engine-to-engine
differences in EGR rates; variation in engine-out O2, NO and NO2 emissions; variation in oil
consumption rate, and variability in FTP and ESC PM emission measurement. But engine
manufacturers cannot precisely and absolutely control even those factors. As a result, these
uncertainties will contribute to the proposed method's unreliability.
Moreover, many factors are outside of engine manufacturers' control. Vehicle OEMs control
or partially control many variabilities such as variation in engine airflow, vehicle-to-vehicle
differences in charge cooling effectiveness. Only if vehicle OEMs restrict vehicle designs or engine
manufacturers develop OBD threshold calibrations that are unique for each vehicle type can these
sources of variability be diminished. But such approaches are unrealistic. Vehicle OEMs should not
be restricted in their vehicle designs. And engine manufacturers cannot predict all the unique
calibrations that may be required for each and every vehicle in which their engines may be used,
and they cannot control where the engines are ultimately used.
The commenter also points out that many of the variabilities noted above are the result of
limitations of scientific knowledge (e.g., variation in engine airflow; uncertainty/imprecision in
assessing exhaust flow rate; imprecision of pressure drop measurement; lack of correlation between
pressure drop and filter soot loading due to variation in permeability and uniformity of soot layer;
uncertainties associated with effectiveness of periodic ash cleaning processes; variation in ash
morphology, sintering, and non-uniform deposition; lack of correlation between filter pressure drop
changes and PM increases for various types of DPF cracks/leaks). Although engine manufacturers
and other experts have expended and continue to expend substantial effort to develop the
underlying science to allow these processes to be more fully understood and to be accounted for in
the modeling, the extent of progress over the next few years is not clear.
Finally, several of the variabilities noted above relate to in-use operating factors that are
beyond the scope of what engine manufacturers can know and account for (e.g., variation in fuel
characteristics, variation/uncertainty in temperature of soot layer, variability in soot composition,
uncertainty in assessing completeness of active regenerations, inability to incorporate full range of
operating sequences, variation in oil ash content). As a practical matter, it will not be possible to
reduce these sources of uncertainty without imposing impractical limitations on the usage of the
engines and vehicles.
65
-------
Lastly, the commenter states that repeatability of test results, which forms the basis for
model development, also is an issue. Even under controlled laboratory conditions, there is significant
uncertainty in the repeatability of loading and regeneration.
Given the uncertainties of soot load/oxidation model-based approaches for detection of DPF
leakage, industry has also focused on the availability of soot sensing devices for detection of soot
leakage past the DPF. The commenter states that, to their knowledge, sensor suppliers are still at
"concept" stage for such sensors and are not ready to propose possible future availability or
performance specifications. Even if prototype sensors were available today, there is insufficient
leadtime to develop and assure the required accuracy of such sensor-dependent diagnostics for use
on engine products available for commercial sale in 2010. When EPA has determined that a soot
sensor is available and feasible for DPF threshold monitoring, an appropriate rulemaking can be
initiated.
Diverting resources from the challenges associated with implementation of reliable DPF
systems to develop an overly -aggressive OBD monitor that will not be capable of meeting the
proposed thresholds could jeopardize the success of DPF systems. DPF failures will be rare and
when failures do occur the delta pressure monitor will detect them.
Current medium-duty (8,500-14,000 pounds) DPF monitors are capable of detection only at
the 9-1 Ox the PM standard level. Not everyone is even able to meet that level. Notably, medium-
duty vehicles certified to CARB's OBDII (13 CCR 1968.2) represent a limited range of engines in a
defined set of applications, where the engine-vehicle configurations were known and for which
monitoring calibrations could, in principle, be most readily accomplished. By contrast, HD has a far
greater range of engines and potential applications.
Regarding the NMHC threshold associated with the DPF monitor, a commenter states that
the DPF NMHC conversion capability is typically lower than that of the DOC, and the normal
temperature rise is much lower than that which would be detectable with the temperature sensors.
Even if excess NMHC were intrusively "fed" to the DPF, the NMHC conversion capability of a typical
DPF would not create a measurable exotherm. Without a measurable exotherm, even functional
monitoring of the NMHC conversion capability of the DPF cannot be done.
EMA, 2005-0047-0026, p. 5-10 and Appendix A
American Trucking Association, 2005-0047-0029, p. 4
Agency Response:
We agree with many of the arguments made by commenters in that uncertainties exist in the
technology expected for OBD use. Similarly, uncertainties exist in the technology expected for
control of DPF systems. We believe that is true despite the fact that DPF systems are being certified
and are being implemented on 2007 model year engines and vehicles are being sold into the
marketplace. However, we believe that the uncertainties are not uncertainties over the ability of the
technology to work, but rather uncertainties over how much improvement is possible and how much
improvement will take place during the initial implementation years that we have entered only
recently. A perfect analogy is the situation that existed in the mid-1970s and the early days of
gasoline catalytic converters. Those systems improved dramatically over the initial years of
implementation and continue to improve even today after more that 30 years since their introduction.
The comments concerning the proposed DPF monitoring requirements, and the proposed
NOx aftertreatment monitoring requirements, mirror those submitted in response to EPA's and
CARB's initial gasoline-based light-duty OBD proposals. Industry claimed, in general, that the
requirements were too stringent, that they could not be met, that they would be too costly, and that
66
-------
they would not deliver sufficient benefits. Despite those claims, the OBD systems in place today
detect malfunctions at greater stringency than originally proposed, they have not been so costly as
to be detrimental to automobile sales, and they have provided benefits to consumers and, we
believe, to manufacturers themselves.
We believe that it is not currently possible to know what level of threshold is the most
appropriate level. Manufacturers appear to focus their efforts on the other aspects of the emission
control system - the engine design, fuel strategy, aftertreatment devices, regeneration strategies -
and worry about the OBD system only after those design elements have been decided upon. In
fairness, we cannot argue that each manufacturer's limited resources should not be used primarily
for controlling the emissions from the tailpipe. But, that suggests that no one today can know what
future OBD systems can or cannot do given that the emission control systems they will be
monitoring have not yet been determined completely. After years of experience with OBD, tracking
its development and capabilities, it has become clear that, despite our sense that OBD should be a
critical element to the design of an emission control system, it is often the last element considered by
manufacturers.
Given the wide variety of approaches manufacturers are expected to take in meeting
emissions standards, it is likely that different OBD systems and different OBD monitoring strategies
will be capable of detecting malfunctions at different emissions levels. While some strategies may
be capable only of detection at 0.10 g/bhp-hr PM or9xthe standard-the level suggested by EMA-
it seems unreasonable to take a present day worst case detection level and make that the threshold
level for all engines. This reasoning is borne out in the current state of certification to the OBD
requirements for vehicles less than 14,000 pounds. Some of those systems have been certified to
detect at 12x to 16x the applicable PM standard (note that they have been certified with a deficiency)
while others have been certified to detect at 5x to 6x the applicable standard. Does this suggest that
the threshold should be 16x the standard or 5x the standard? We believe that current systems
should be expected to detect at 5x the standard and, where the system is unable to do so, an
alternative monitoring approach can be used.
Instead of monitoring to a PM emissions threshold, manufacturers will have the option of
defining a "detectable" change the observed delta pressure across the DPF. It is this change in
delta pressure which indicates that exhaust flow is bypassing the filter (i.e. some portion of the filter
area is missing). This detectable pressure drop can be determined by running the engine at a single
speed/load point, determing the delta pressure at that condition, and multiplying the observed value
by 0.5. The engine speed & load conditions for running this detectable pressure drop test, based on
the test cycles and procedures for the supplemental emissions test (SET) under §86.1360-2007, will
include all engine speed and load points greater than a region bounded by a line connecting mode
numbers 2, 6, 3, and 13 (i.e. A100, A75, B50, and C50)). At engine speeds greater than "speed C",
the monitor shall run whenever engine load is greater than 50%. The detectable change in pressure
drop is determined by operating the "B50" speed/load point, observing the pressure drop on a clean,
nominal DPF, and multiplying the observed pressure drop by 0.5 or other factor supported by data
and apporved by the Administor. Since the alternative monitor will run under engine speed, load,
and exhaust flow conditions which result in a meaningful pressure drop across the DPF - and the
detectable change in pressure drop is determined at an operating point where exhaust flow and the
resulting pressure drop, are significant, we believe that this is a reasonable approach to establishing
a "minumum detectable" pressure drop and monitoring area. For example, if the pressure drop
observed on a clean, nominal filter at the "B50" speed/load point was 4 kPa, the "minumum
detectable" delta pressure would be 2 kPa (4 x 0.5). Whenever the engine is operating within the
boundaries of the monitoring area, the output of the delta pressure sensor will be compared to the
expected pressure drop under similar conditions on a clean, nominal filter minus the detectable
pressure drop. A permanent fault is set if the actual pressure drop across the DPF is less than the
67
-------
calculation of expected pressure drop minus the detectable pressure drop for a period of period of
five seconds or more, indicating a failed DPF.
We should note that, while some systems are being certified today at 5x the standard, none
are certified to detect any and all possible failure modes of the DPF. Hence our provision that allows
certification even though all failure modes cannot be detected provided the manufacturer can make
such a demonstration (see §86.010-18(i)(5)(i)). We also note that our in-use compliance thresholds
are double those for certification engines through 2015, and that extrapolated OBD engines are not
even evaluated against emission thresholds until 2016 (see §86.010-18(p)(2) for 2013-2015 model
years and compare to §86.010-18(p)(3) for 2016-2018 model years).
In the end, we believe that the threshold levels we have proposed strike the proper balance
between forcing manufacturers to design effective OBD monitoring strategies while also providing
the flexibility to certify systems when good faith efforts fail to comply fully with the requirements.
III.A.9 Exhaust Gas Sensor Monitoring
What We Proposed:
We proposed that malfunctions related to sensor performance be detected prior to emissions
exceeding the applicable thresholds. We also proposed that malfunctions related to circuit integrity,
feedback functions, monitoring functions, and heater performance and circuit integrity be detected
prior to those functions being lost. Further, we proposed that sensor and heater performance be
monitored once per trip, that monitoring functionality be monitored whenever conditions were met,
and that circuit integrity and feedback functionality be monitored continuously.
Comments:
As discussed with respect to NOx aftertreatment monitoring, the accuracy of current NOx
sensor technology is not capable of achieving the EPA NOx emissions thresholds requirements in
time for 2010. To be able to create a diagnosis based on the NOx sensor, the sensor supplier would
have to lower the tolerances in time for the engine manufacturers to use these sensors, no later than
mid-2007. Manufacturers must make design decisions with respect to 2010 technology by mid- to
late-2007, and, therefore, must rely on NOx sensor technology as it exists today. Based on current
NOx sensor technology, manufacturers expect to be able to monitor to thresholds at the NOx
standard/FEL plus 0.60, not plus 0.30.
EMA, 2005-0047-0026, p. 12-13
Agency Response:
We understand EMA's concern that NOx sensor technology will not support the 2010 NOx
threshold and have changed the NOx threshold to +0.6g/bhp-hr.
III.B Feasibility of the Monitoring Requirements for Gasoline/Spark-lgnition Engines
III.B.1 Fuel System Monitoring
We received no comments pertaining to this section that require analysis.
68
-------
III.B.2 Engine Misfire Monitoring
Comments:
EPA's proposed misfire monitoring requirements for HD gasoline engines generally align
with CARB's, with the exception that EPA did not include a provision from CARB's regulation for
engines with more than eight cylinders. That provision provides for relief on a case-by-base basis if
an engine is incapable of monitoring for misfire under positive torque conditions over the full
operating range of the engine up to redline (see 13 CCR 1971.1(f)(2.3.5)7). It is absolutely essential
that EPA include this provision in the final rule for HD gasoline engines. Ten cylinder spark ignition
(i.e., gasoline) engines are used in trucks over 14,000 pounds and cannot detect misfire at the
higher speed/loads. This is because crankshaft acceleration levels are less pronounced when a
misfire occurs as the number of cylinders increases. Current misfire detection technology is not
capable of meeting full-range misfire requirements on those engines. Ten cylinder engines are also
used in vehicles under 14,000 pounds, and manufacturers make use of an OBDII provision (see 13
CCR 1968.2(e)(3.3.5)) that is similar to 1971.1(f)(2.3.5). Manufacturers have attempted to
implement other monitoring approaches, such as neural network, but have found them to be
incapable of meeting the full-range monitoring requirement. CARB has evaluated the issue and
concluded that the provision is needed. EPA should include this provision in the federal HDBOD
final rule.
EMA, 2005-0047-0026, p. 42-43
Agency Response:
We have no argument with the legitimacy of the CARB provision with respect to misfire
detection on engines with more than eight cylinders. Our intention would have been to address this
industry concern in a manner identical to that used by CARB, and to call any system needing such
relief a deficient system. Such a determination would carry with it no monetary penalty since we do
not charge fees association with deficiencies. However, we do have limitations on carry-over of
deficiencies which, we believe, manufacturers would need to violate in order to sell engines with
more than eight cylinders since, presumably, those engines would need this relief indefinitely. For
that reason, we see a need to provide for the relief in the regulatory language rather than addressing
the need via our deficiency provisions. This change has been made in the final rule (see §86.010-
.B.3 Exhaust Gas Recirculation (EGR) Monitoring
We received no comments pertaining to this section that require analysis.
.B.4 Cold Start Emission Reduction Strategy Monitoring
We received no comments pertaining to this section that require analysis.
.B.5 Secondary Air System Monitoring
We received no comments pertaining to this section that require analysis.
7 Note that this is a corrected reference. The EMA comments erroneously referred to the analogous
sections of 13 CCR 1968.2.
69
-------
III.B.6 Catalytic Converter Monitoring
We received no comments pertaining to this section that require analysis.
III.B.7 Evaporative System Monitoring
We received no comments pertaining to this section that require analysis.
III.B.8 Exhaust Gas Sensor Monitoring
We received no comments pertaining to this section that require analysis.
III.C Feasibility of the Monitoring Requirements for Other Diesel and Gasoline Systems
III.C.1 Variable Valve Timing and/or Control (VVT) System Monitoring
We received no comments pertaining to this section that require analysis.
III.C.2 Engine Cooling System Monitoring
What We Proposed:
We proposed that malfunctions related to the cooling system not warming up as expected be
detected since cooling system temperature status is an enable criterion for so many OBD monitors.
Further, we proposed that the thermostat be monitored for proper operation and that the engine
coolant temperature sensor be monitored for circuit continuity, out-of-range values, and to ensure
that it provides rational temperature readings.
Comments:
EPA's proposed cooling system requirements are impractical, overly burdensome and costly,
and do not adequately account for the unique nature of the heavy-duty industry. Thus, they should
be eliminated from the final rule.
The light-duty method of creating a model of heat flow to/from the engine and comparing that
to actual coolant temperature to ensure that the cooling system is behaving normally is impractical, if
not impossible, to robustly achieve in the horizontally integrated HD engine market. HD engine
manufacturers often sell engines without knowing the chassis, transmission, and cooling system with
which that engine will be installed. The connections to the cooling system - engine radiator, cab
heater, defroster, sleeper heater, but heater, air conditioner, transmission cooler, exhaust throttles,
fuel fired heaters, air coolers, block heaters, etc. - are all components that are not supplied by, nor
controlled by, the engine manufacturer. However, all those devices affect cooling system behavior
and, thus, engine manufacturers cannot practically and cost-effectively achieve a robust diagnostic.
The commenter suggests that the proposed requirement would result in each vehicle into which an
engine manufacturer's engines were place would need to be calibrated separately since one
calibration would not work for all of that manufacturer's engines. The commenter states that such an
70
-------
outcome is technically possible but so cost ineffective as to be realistically impossible because the
calibration burden would be enormous and, in the end, would be the task of the vehicle manufacturer.
EMA, 2005-0047-0026, p. 20-22
If EPA requires OBD to register cooling system malfunctions, one of the following may
happen: bus OBD systems will frequently register faults during otherwise regular operation, or truck
tractor OBD systems will almost never register faults. The only way to get around such problems
would be for EPA to require that engine manufacturers calibrate each engine (and each engine
rating, insofar as engine power strongly affects engine heat rejection rates) with each cooling system
in each vehicle. The task would be untenably burdensome. In all likelihood, manufacturers like
Freightliner would have to extrapolate cooling system settings across a wide range of vehicle
configurations (not merely bus versus tractor, but including such configuration options as the type
and power draw from alternators and other accessories). Such extrapolation would engender the
possibility of false Mils. Moreover, when an OBD-equipped vehicle is repurposed (e.g., converted
from a highway tractor to an intracity tractor or dump truck, which is a common practice in HD
highway fleets), the warming rate could change in such a way as to lead to false Mils. Freightliner
requests that EPA eliminate thermostat monitoring requirements in HD highway applications.
Freightliner, 2005-0047-0025, p. 3
Agency Response:
The comments raised by industry are of great concern because the nature of the comments
extends beyond OBD and to emissions compliance itself. The aftertreatment devices expected for
2010 compliance - DPF systems and either SCR or NOx adsorber systems - are very much
dependent on the level of heat available in the exhaust. The comments suggest that engine
manufacturers will have no control over the amount of heat being removed from the exhaust for use
as cabin heating, etc. If true, how do manufacturers intend to comply with emissions standards
during real world operation? The comments suggest that the engines and emissions control
systems will be designed with the emissions test in mind - where no heat is removed from the
exhaust for cabin heating, etc. - and with little regard for real world operation. These comments
suggest that the OBD requirements to monitor the engine cooling system are more important than
we believed initially.
That said, the direct emission impact of a malfunctioning thermostat or cooling system is only
a secondary reason for the requirement. The primary reason the cooling system is monitored for
proper operation is because the engine manufacturer itself elects to use engine coolant temperature
as a primary enabling criteria for monitoring of nearly every emission critical component. The
requirement for cooling system monitoring is simply stated as a requirement to verify that the engine
properly warms up to the highest temperature required by the engine manufacturer for monitoring of
other components. The relative stringency of this monitor is a direct result of how high the
manufacturer requires engine coolant temperature to get before monitoring other components and
engine manufacturers can effectively desensitize algorithms to vehicle factors by enabling other
monitors at lower temperatures. While we understand engine manufacturers' concerns that actions
by truck builders and users can impact their monitor design, the intent of OBD systems is to have
monitoring of the emission components during real world operation of heavy-duty vehicles. Failure to
achieve the necessary warmed-up temperatures required for monitoring would effectively mean
monitoring is indefinitely disabled on real world vehicles, thus negating nearly the entire OBD system.
Verifying the cooling system is operating properly is a crucial and necessary element to ensure OBD
systems continue to operate on real world vehicles throughout their life.
71
-------
We suspect that the OBD requirements will result in the limitations and specifications that the
engine manufacturer will have to place on the vehicle builders to ensure the engine remains in a
legally certified configuration. This cooling system monitoring requirement may result in additional
calibration work or classification of the exact extent to which the vehicle builders can modify the
cooling system to maintain a compliant system. Failure to do so could result in MIL illumination.
However, while eliminating the cooling system monitoring requirements would avoid this potential
MIL illumination, it would also re-open the possibility that such a system would be put into service
and all of the OBD monitors would be disabled for the entire life of the vehicle.
III.C.3 Crankcase Ventilation System Monitoring
We received no comments pertaining to this section that require analysis.
III.C.4 Comprehensive Component Monitoring
What We Proposed:
We proposed that malfunctions of any electronic engine component or system not otherwise
covered by our other proposed requirements be detected. Further, we proposed that such
components or systems be monitored for circuit continuity and out-of-range values and to ensure
they provide rational information (input components) and proper functional responses (output
components).
Comments:
Regarding the proposed requirements to monitor glow plugs and intake air heater systems,
EPA has failed to establish a reasonable justification for the requirements or an analysis of the
technological feasibility of detecting low to moderate degrees of glow plug degradation. Directly
monitoring these devices would require adding hardware to today's systems, which EPA has
indicated it intends to avoid. Because of the high current and inrush current flows, monitoring
hardware will likely remain separate from the engine control system. Moreover, intake air heaters
and glow plugs are unlikely to experience gradual deterioration and will, instead, fail to function
altogether. Such total failures are detectable via continuity monitoring making threshold monitoring
unnecessary. Degradation aside, intake air heaters can be assessed for providing heat using the
engine's intake air temperature sensor which would not likely find low to moderate levels of
degradation nor isolate specific failures; however, service bay diagnostic approaches do so today.
EMA, 2005-0047-0026, p. 18-19
Regarding requirements to monitor cold start aids, such requirements will impact vehicles
and vehicle manufacturers by requiring significant changes to the way vehicle manufacturers do
business to ensure OBD systems work properly. EPA must carefully consider such impacts and the
potential impacts on suppressing sales of OBD-equipped engines/vehicles.
EMA, 2005-0047-0026, p. 19
EPA should consider the possibility that OBD requirements to monitor cold start aids may
result in them being added in the aftermarket rather than by engine manufacturers. Cold start aids
are customer-selected items and are not required for emissions compliance. Engine manufacturers
cannot be responsible for monitoring aftermarket parts over which they have no control.
EMA, 2005-0047-0026, p. 19
72
-------
Agency Response:
We have changed the regulation to allow, for 2010 through 2012 model years, circuit checks
only for glow plug monitoring. For 2013 and beyond, we will continue to require both a circuit check
and a functional check, as proposed. The functional requirement can be met by determining if the
current used by the sensor is rational. Intake air heaters will require both a circuit and a functional
check in 2010, since all sensors needed to do the testing are available. We do not believe that the
monitoring of cold start aids will drastically change the way vehicle manufacturers do business. Cold
start monitoring requirements are already in place for California Medium Duty vehicles without issue.
We agree that manufacturers are not required to monitor aftermarket parts.
IV. What are the Service Information Availability Requirements?
IV.A Comparison of the Aftermarket Service Industry Below 14,000 Pounds and Above 14,000
Pounds
Comments:
The Engine Manufacturers Association (EMA) commented that the heavy-duty service
industry is significantly and substantially different from the light-duty vehicle service market. The
light-duty service industry operates on a wide scale, with tens of thousands of service facilities
across the nation - whether franchised or independent - to meet the needs of millions of vehicles. In
marked contrast, the heavy-duty service industry operates on a much smaller and more
individualized scale, with far fewer service facilities and, although a limited number of products, an
enormous number of variations on how those products may be configured and calibrated. Indeed,
the volume of heavy-duty vehicles is approximately 1/40th that of light-duty vehicles.
EMA also commented that aftermarket service providers have testified at a California public
workshop that there was no need to extend service information availability requirements to the
heavy-duty industry. The current service information infrastructure is already established and
adequate to meet the needs of the heavy-duty engine and vehicle service industry, and assures that
emission-related repairs are carried out.
EMA further commented that, in the Preamble, EPA cites data which shows a slight increase
in general maintenance and overhaul by "the independent sector" (72 Fed. Reg. 3266) as evidence
of the need for service information availability. As described above, heavy-duty service information
already is widely available. Moreover, the vehicle census data upon which EPA relies is now a
decade old, and the increase in third party general maintenance service is not attributed to a specific
cause or for a particular purpose. Such increases do not establish a "need" for service information.
In fact, increases in third party maintenance could easily be the result of compliance with OSHA
regulations, increasing costs of workers' compensation insurance, storage and disposal costs for
hazardous materials like used engine oil and anti-freeze, or increases in direct labor costs, which
may have led vehicle owners to decide that it was no longer practical to perform routine preventive
maintenance or general maintenance at their own locations during the period from 1992 to 1997.
The industry now has over 10 years of experience with a vehicle fleet that is predominantly
equipped with electronically-controlled engines. Without follow-up data from 2002 and 2007 that
would confirm or deny such a trend, and that would provide a better comparison of the impact of
engine electronics on service rates, EPA's data does little to support the Proposed Rule.
73
-------
EMA concludes that the heavy-duty service and repair industry works so well that users and
manufacturers cooperate on the design and development of service equipment, service procedures,
and vehicle design. The Technology and Maintenance Council ("TMC") of the American Trucking
Association has collaborated with industry on maintenance and design practices for decades.
Lastly, EMA commented that imposing regulatory requirements for service information is like
trying to fix something that, in the heavy-duty industry, is not broken. The need is not there. Even
EPA must have some sense of the lack of need, because it has requested comment on the need for
service information provisions in the heavy-duty industry, as well as the applicability of service
information requirements to the heavy-duty service industry. If EPA is going to impose regulatory
requirements, then those should be limited to supporting what engine manufacturers already do.
Instead, much of EPA's proposal would require significant changes in how engine manufacturers
provide information. In fact, the Proposed Rule would require complex, substantial, and time-
consuming changes in the current heavy-duty service information infrastructure - changes that will
result in increased costs for manufacturers, and in increased costs for providing service information
and tools.
Caterpillar commented that, based on the EPA's own comments regarding the US Census
data, on a percentage basis the ratio of vehicles being serviced by the aftermarket for light-duty is
approximately 2.5x higher than that of heavy-duty. When a 1/40th sales volume difference between
heavy-duty and light-duty is factored in, the effective aftermarket volume of for heavy-duty repair is
less than 1% that of light-duty on a "number of vehicles" basis. Therefore, the effective heavy-duty
aftermarket volume is extremely small in comparison. Also, the EPA should consider that heavy-duty
vehicles are serviced through different channels than the light-duty industry. A greater percentage of
heavy-duty vehicles are owned in large numbers and thus serviced in larger numbers in company
fleets.
Caterpillar also commented that many of the proposed Service Information requirements are
already available to the aftermarket. For example, ordering of service literature or offline training can
currently be purchased from Caterpillar dealers, and online service information accounts have been
purchased by many Caterpillar customers. Regardless of the number of vehicles a business owns,
each company is equally able to purchase service information from their local dealer for servicing of
Caterpillar products.
The Aftermarket Industry Alliance (AAIA), the Automotive Engine Rebuilders Association
(AERA), the Automotive Parts Remanufacturer Association (APRA), and the Heavy Vehicle
Maintenance Group (HVMG) (hereafter referred to as AAIA, ET.AL), collectively commented that
the heavy duty independent aftermarket is composed of thousands of shops throughout the nation
which service and repair heavy duty vehicles but which are not dealerships or authorized service
providers of the heavy duty vehicle or component manufacturers. Truck and other heavy duty vehicle
owners benefit by having access to the affordable and convenient repair provided by independent
shops. Absent this competitive alternative to manufacturer authorized repair facilities, owners would
be forced to pay higher repair prices and endure longer wait times in order to obtain repairs to their
vehicles. Thus the independent aftermarket helps ensure that vehicle owners get their vehicles
repaired in a timely and economic fashion and that clean air goals are achieved.
AAIA, ET. AL, further commented that it is estimated that there are over 25,000 independent
heavy duty engine and transmission service shops and heavy duty machine shops in the United
States. These shops perform approximately 25% of the general maintenance on heavy duty vehicles
and 30% of the major overhauls on them. This works out to about 1 million general repairs per year
and the same number of major overhauls.
74
-------
Agency Response:
EPA agrees with commenters that there are very clear and distinct differences between light-
duty and heavy-duty aftermarket service providers and that our regulations should reflect these
differences as appropriate. However, EPA does not agree that the fact that service information is
currently available in some form to the heavy-duty aftermarket necessarily precludes the need for
regulations for the availability of heavy-duty service information. While heavy-duty manufacturers do
currently have avenues in place for making this information available, the purpose of the regulations
for heavy-duty service information is to ensure that there is some level of uniformity and consistency
across manufacturers to ensure that aftermarket service providers have reasonable access to
service information. With these regulations, EPA has taken into account the existing infrastructure
of information availability and wherever possible and appropriate, and have structured these
regulations to the extent possible to ensure uniform and consistent information availability to
aftermarket service providers.
IV.B Information to be Made Available by OEMs
IV.B.1 Definition of Emissions-related Information
Comments:
EMA commented that the purpose of service information requirements is to assure that
information is available for the purpose of undertaking emissions-related service repairs. It follows,
then, that EPA's proposed definition of "emissions-related information" must be revised to assure
that engine manufacturers are responsible only for providing information that is truly emissions-
related.
EMA commented that EPA's proposed definition would include "information regarding any
system, component or part of an engine that controls emissions and any system, component and/or
part associated with the engine, including, but not limited to: the engine, the fuel system and ignition
system;..."(emphasis added). EMA objects to the italicized language as proposed in that it is far too
broad. Instead, EMA commented that EPA should adopt the following revised language in place of
the italicized language: "and any system, component and/or part that is part of the diagnostic
strategy for an OBD monitor."
AAIA, et. al. commented that they have a concern with the proposed rule is that the scope of
the information to be provided by the rule is much narrower than under the light duty rule. Under the
light duty rule, emissions-related information is defined to include any part or component of a vehicle
associated with the powertrain system and any information for any system, like the transmission,
which is likely to impact emissions. Under this proposed heavy duty rule, emissions-related
information appears to be limited to that same type of information but only that related to the engine
rather than the entire vehicle. Moreover, the references to the powertrain system and the
transmission have been noticeably deleted. We can see no justification for this limitation. If the EPA
is requiring OBD systems for heavy duty engines and those systems monitor and trigger a default
because of some system failure or problem outside the engine system, then this information also
must be made available to the aftermarket so that its service providers can provide the same level of
service as the engine manufacturers authorized dealers and service networks. We encountered this
same problem in California when the Air Resources Board was finalizing its heavy duty regulation.
The engine manufacturers argued that in many instances the parts and components of other
manufacturers were combined with their engines to assemble a suitable vehicle for the customer.
They argued that because the information on these other systems was not proprietary to them and
75
-------
might not be available to them, they should not be required to provide it. Our position was that if
another vehicle system could adversely affect emissions and if the engine manufacturer elected to
have its OBD system monitor that system, then it had the responsibility to make the emissions
related information available for whatever was being monitored to anyone who needed to fix it,
whether that be a dealer or an aftermarket service provider. California agreed and added language
in Section 1969(d)(9)(B) of its regulations to require that if manufacturer elects to have its OBD
system monitor inputs from the transmission then information related to what was being monitored
and how to repair it had to be made available to the aftermarket. We believe that this language is still
too narrow and that the requirement should apply to any system monitored by the OBD system,
including the transmission. But at a minimum, EPA should at least adopt the same language as
California and allow access to transmission system information.
Agency Response:
EPA agrees that it is reasonable to clarify the definition of "emissions-related" information as
put forth by EMA and the regulatory language will amended accordingly.
EPA does not agree with AAIA et al. It is not necessary to adopt the language from the
CARB regulations to clarify that manufacturers have to make available information for any systems,
including transmission systems, that could impact emissions or for systems that otherwise are
monitored by the OBD system. EPA has a detailed description of how we define emissions-related
information §86.1808-1 OQ(3)(ii)(E) of the regulatory language, which includes a reference to
transmission systems and we believe that this definition is sufficient. Additionally, EPA does not
agree that the scope of information to be provided under this rule is any narrower in the light-duty
rule. The only explicit information limitation discussed in this rule is information pertaining to
transmission systems, Other than that limitation, EPA believes that this rule is otherwise very clear
that any and all information need to diagnose and service emissions-related repairs must be made
available to after market service providers.
IV.B.2 Other Information
Comments:
Cummins commented that that there is some service information that we should not be
required to provide to aftermarket service providers. In particular, Cummins commented that
information related to warranty repairs should not be required to be made available to the
aftermarket. This information should only be made available to service locations that are authorized
to perform warranty repairs. Cummins recommended adding the following language to regulations:
"Manufacturers may take steps to restrict warranty and customer assurance plan information
used only for the purpose of providing such manufacturer covered repairs to only those
repair locations authorized by the manufacturer."
Cummins also commented that EPA should add the following regulatory language to
Sections 86.010-38 (j)(6)(i)
Informal recall service information such as engineering notes and/or sketches are not
required to be made available as long as this information is not made available to
manufacturer franchised dealerships in the form of manuals.
Cummins comments that recall service information such as personal notes, documents,
spreadsheets, charts, and diagrams that is not formally published should not be required to be made
76
-------
available on our web site. To require control of this informal documentation would be overly
burdensome to the manufacturer.
Agency Response:
EPA agrees that it is not necessary for informal information that is not made available to
dealerships to be made available on manufacturer web sites and will change the regulatory
language for this section accordingly with one clarification regarding warranty information. While we
believe that there is some warranty information that does not need to be made available to
aftermarket service providers, such as OEM-specific warranty labor rates, other general warranty
information relevant to making emissions-related repairs must be made available to aftermarket
service providers.
IV.B.3 Regulating the Cost of Service Information
Comments:
EMA commented that EPA has not cited any authority for its proposal to regulate the pricing
of service information by manufacturers. EPA cannot cite any authority, for neither the CAA nor other
federal law provide any such authority to EPA. According to EPA, the legislative history for the
service information provisions in the CAA included Congress's intent that manufacturers should be
able to recover "reasonable costs" for providing service information. But Congress did not take the
further step to establish EPA's authority over costs and pricing issues in the language of the statute.
In particular, requiring consideration of "the ability of the average aftermarket technician or shop to
afford the information" (§86.010-38Q(80(i)(E)) in pricing decisions is completely outside the scope of
the CAA, as it does not even touch on Congress's intent to ensure that manufacturers "recover
reasonable costs for making information available" (72. Fed. Reg. 3269).
The American Truck Dealers Division of the National Automobile Dealers Association (ATD)
commented that there is absolutely no justification, statutory or otherwise, for EPA to regulate the
cost of OBD-related emissions information and any attempt to do so would exceed EPA's authority.
ATD further commented that any heavy-duty OBD rule designed to make essential OBD-
related service information generally available must defer to the marketplace to provide such
information at market prices. Had it been an issue of concern, the statute would have mentioned
regulating cost or price. Moreover, just as there has been no demonstration of any present or
potential future unavailability of heavy-duty OBD-related emissions information, there has been no
demonstration of any present or potential future OBD-related service information cost concerns. In
short, the proposal appears to be addressing problems that don't exist.
Assuming adequate the authority and justification, any attempt by EPA to regulate the cost of
necessary OBD-related emissions information must reflect the cost to dealerships for the same or
similar information. With its comments, ATD provided the results of survey it's conducted to get
some idea of what dealers pay for tools, information, and training.
Agency Response:
EPA disagrees with the EMA and ATD comments regarding regulating the cost of service
information. There is significant precedent established throughout the history of EPA's role in
regulating service information that the cost of service information is critical to its availability. EPA
has long held the position, supported by the legislative history of the service information provisions in
the Clean Air Act, that we do have the authority to regulate the cost of service information. The
legislative history of the Clean Air Act clearly intends that manufacturers cannot price information
77
-------
such that it effectively renders it unavailable, or in such a way that it becomes a profit center for the
manufacturer.
The provisions being finalized today place no particular restrictions on what manufacturers may
charge for access to their information or their scan tools, only that manufacturers must make
information and tooling available at a fair and reasonable price. Generally, EPA has interpreted this
to mean that information and tool should be priced along the lines of what dealers are charged,
taking into consideration that manufacturer's can recoup at least some of their costs for making them
available to the aftermarket.
IV.C Requirements for Web-based Delivery of the Required Information
IV.C.1 General Comments on Web Site Availability
Comments:
EMA commented that they generally support making certain heavy-duty information available
via Web sites in 2010. However, EPA must revise the Proposed Rule to clarify that publicly-available
service information Web sites are not required of heavy-duty engine manufacturers until at least one
year after the start of commercial production of the engine. This is not a retroactive rule as was the
case when the under-14,000-lb. requirements were implemented. In fact, with the extremely longer
useful lives of heavy heavy-duty engines - 435,000 miles - there is no need for public information
availability any earlier than one year after the start of commercial production. Given the complexities
manufacturers will be facing in meeting the underlying emission standards and OBD, and the fact
that commercial customers will seek repair and service from manufacturer facilities during the
warranty period, manufacturer-specific Web sites should not be required until at least one year after
the start of commercial production of an engine product.
.AAIA, et al. commented that section (j)(12)(i) which contains the time frame when the engine
manufacturers have to make the emissions-related information available to the aftermarket has also
been modified from it light duty predecessor in a way that could adversely affect the aftermarket.
Both sections require that the information be made available within six months of model introduction,
or at the same time it is made available to the manufacturer's franchised dealers. However, the light
duty rule says that this information must be made available at the time the first of these events
occurs.
AAIA, et al commented that the heavy duty rule omits this language, and sees no reason for
it. If the engine manufacturer provides the information to its dealers prior to six months after
introduction of the model, it should make the information available to the aftermarket at the same
time. We request that the "whichever is earlier" language from the light duty rule be reinserted in the
heavy duty one.
Caterpillar commented that, even before any heavy-duty service information regulation was
in place, Caterpillar was already providing service information to its dealers and customers through
online, digital offline, and paper publications. This information could be purchased through
Caterpillar dealers. Caterpillar anticipates that the EPA's proposed Service Information Regulation, if
adopted, will neither promote any beneficial change to the level of information already available
online, nor cause a notable increase in service information sales because businesses that need
service information already have access to it.
78
-------
AID commented that they support a requirement that the delivery of emissions-related
service information occur through individual engine manufacturer (OEM) web sites. On the light-duty
side, OEM web sites have proven to be effective for service facilities seeking information necessary
to make OBD-related emissions diagnoses and repairs. At the same time, EPA should not micro-
manage these web sites. The only essential issue of importance is whether necessary OBD-related
emissions information is being made available in a generally user-friendly and timely manner. If a
final rule is issued, EPA's heavy-duty OBD web page should link to the various OEM web sites.
AID also commented that any OEM website mandate should include adequate lead-time.
Given the long emissions warranty period for heavy-duty vehicles/engines, AID suggests OEMs be
allowed up to one year after information is made available to dealerships to post that information on
their web sites.
Volvo commented that EPA should revise section 86.010-380(11) to allow manufacturers to
provide access to text-based service information to the aftermarket in the same manner that it is
currently provided to their dealerships rather than search-engine assisted, on-line viewing of
emissions-related repair information.
Agency Response:
Regarding the EMA and AID comment on when the websites need to publicly-available,
EPA believes what EMA is requesting is that EPA finalize a provision that service information itself
should be not required to made available on the manufacturer websites until at least one year after
the start of commercial production of the engine. EPA is somewhat confused by this
recommendation. It our understanding that most manufacturers currently have web sites available
for their dealer networks and will be using these existing web sites to satisfy these regulations.
Further, EPA is providing manufacturers until July 1, 2010, which is several months after the
introduction of the 2010 model year and over a year following promulgation of this rule, before they
are required to initiate their web sites. It is not clear to EPA why manufacturers would want to invest
resources to implement restrictions to aftermarket service providers for after a certain time period
when they don't feel the need to do so now.
While EPA agrees that service is generally performed by dealers while the engine is still
under warranty, EPA proposed, and will finalize a provision that any emissions-related service
information defined in this regulation must be made available to aftermarket service providers at the
same time it is made available to the dealer network.
Regarding the AAIA, et al. comment regarding the omission of the language, "whichever is
earlier" from section Q(12)(i), EPA agrees that this language was mistakenly omitted from this
section and will amend the final regulations accordingly.
In response to Caterpillar's comments, EPA appreciates that Caterpillar has several avenues
available, including a full-text information web site, for aftermarket service providers to obtain service
information. EPA believes that the parameters for web site availability being finalized in this action
will not require Caterpillar to make significant changes to existing infrastructure.
In response to the Volvo comments, EPA understands that Volvo currently provides all
service information to their dealers and the aftermarket in the form of PDF documents rather than
HTML or other web based formats that are more conducive to a traditional search engine. It is also
EPA's understanding that Volvo currently makes this information available for free to the aftermarket
and intends to keep doing so because they do not believe it is cost effective to develop a traditional
e-commerce site. First, there is nothing in these regulations that would preclude a manufacturer
from presenting information in PDF format so EPA believes that Volvo would not have to make any
79
-------
changes to the way they present this information. In addition, because Volvo makes this information
available for free, the need for searchability is not as essential because an aftermarket service
provider has free and unlimited access to all the material on the site and do not have to make any
monetary investments in order to view documents to ensure that it is the information they need. To
the extent that Volvo changes this existing model in the future and would charge for access, Volvo
would be required to allow aftermarket service providers either a free preview or implement search
capabilities so that aftermarket shops can reasonably assess that the information they are about to
purchase is in fact the information they are seeking.
IV.C.2 Short-Term, Medium-Term, and Long-Term Access
Comments:
EMA commented that EPA has proposed a specific structure of "tiered access" for service
information subscriptions over manufacturers' Web sites. EMA does not support the tiered structure
that would require manufacturers to provide subscriptions for certain specified time periods.
For heavy-duty manufacturers, requiring specified periods of time for access would require a
change in how manufacturers provide information today. Generally, heavy-duty engine
manufacturers' Web sites have been very reasonably priced, such that a yearly subscription is
typical. EPA has again applied a light-duty paradigm to this aspect of the Proposed Rule, and it is
not appropriate.
EMA further commented that short periods of access make it more difficult for manufacturers
to recoup the costs of providing a service information site and that the inability to recoup operating
and administrative costs for a service information rule site creates a fixed cost that must be spread
across a far smaller number of engines compared to light-duty industry volumes. The inclusion of
short-term access periods suggests that persons may perform the repair of heavy-duty diesel
engines for only short, infrequent periods of time. Occasional use suggests that the persons involved
do not seek to repair heavy-duty diesel engines as a principal line of business, and in fact may not
meet industry standard qualifications. On the other hand, persons who have invested in a business
to provide heavy-duty service likely are in it for the long term and need and want access to service
information for the long term.
Caterpillar commented that a single, yearly tier subscription is what Caterpillar currently
provides to its dealers and customers. Requesting manufacturers to create service information or
distribution methods that are different than that currently offered is not cost effective for an industry
with significantly lower product volumes and service information subscription rates. A single tier of a
yearly subscription is non-discriminatory for the aftermarket. Also the current cost of Caterpillar's
service information is not significant when compared to the fixed and variable costs of running a
legitimate repair business or compared with the owning and operating costs of managing a vehicle
fleet, and therefore Caterpillar believes the pricing for a yearly service information subscription
should not be a hindrance for those engaged in heavy-duty engine repair.
Caterpillar's further commented that its service information ordering website, that was
established because of ARB regulation, has been online for over 4 months in 2007, and to date, has
generated only a single sale of service information subscription which was from outside the state of
California. While this does not represent the volume of aftermarket service information subscriptions,
it should indicate there is not a significant need for service information to be made available for
ordering online. The EPA should allow manufacturers to direct service information customers to
dealers for purchasing of this information. Any company engaged in the repair of heavy-duty
vehicles will need to have some business relationship with a Caterpillar dealer so that parts can be
ordered for engine repair.
80
-------
Caterpillar recommends that the EPA allow manufacturers to provide ordering of service
information by phone thus utilizing a manufacturer's dealer network in lieu of having online
purchasing of service information. The cost of providing an online website for e-commerce is not
justified based on the expected sales volumes. As of May 2007, Caterpillar has comparatively priced
service information with the light-duty industry when in fact, due to the significantly lower volume of
service information subscriptions in the heavy-duty market, it would be reasonable for Caterpillar's
service information to be much higher in cost. Our recommended changes to the EPA's proposed
regulation will help keep service information at an affordable cost to the aftermarket.
AAIA, et al commented that a tiered approach to availability and pricing, similar to that
provided in the light duty aftermarket, is necessary to provide the flexibility to allow the diversity of
repair work performed by the aftermarket service shops to continue to be available at a reasonable
price. While many shops specialize in one engine manufacturer or another and would most likely
subscribe to that manufacturer's web site on an annual basis, many others service a variety of
customers with different makes and models of engines and would need access to the information for
only a limited period of time. And even those facilities that specialize in the products of one
manufacturer or another will often be required to service another brand to satisfy the needs of their
customers. To require these facilities to bear the cost of an annual subscription for the information
would be to de facto deny it to them. Therefore, unless a tiered approach to availability and pricing is
adopted as part of the regulation (or the manufacturer makes the information available at one low
price for all aftermarket service providers) a large segment of the heavy duty aftermarket will be
economically barred from effectively using the information on the web site.
Agency Response:
While EPA believes that there may be advantage for some form of shorter term access for
the web sites, EPA believes that it is not necessary to finalize a provision that would require short-,
mid- and long term access to the manufacturer's web sites that are otherwise required by this
regulation in exactly the same fashion as we finalized for the light-duty rule. EPA finalized the tiered
approach for the light-duty arena because of the nature of the service and repair work done by the
light-duty aftermarket. First, there are approximately 22 light-duty vehicle manufactures. That fact,
combined with the fact that the light-duty aftermarket will generally service numerous makes and
models, made it sensible to adopt the tiered approach to accommodate the nature of the how
aftermarket service is performed for light-duty vehicles. On the heavy-duty side, EPA believes that
there is less need for truly short term subscriptions given that there are relatively few HD
manufacturers (approximately 6-10) and currently nearly all heavy-duty manufacturers have web
sites available with yearly subscriptions that are reasonably priced (generally $1500 or less per year).
In addition, EPA believes that it would be unduly burdensome to require manufacturers to
develop the tiered approach to web site access given that their web sites are already developed and
available for use by the heavy-duty aftermarket. When we finalized the tiered approach for the light-
duty service information rule, the vast majority of manufacturers had not yet developed their web
sites and, therefore, was able to develop the infrastructure for tiered access much more efficiently.
In addition, based on our experience in implementing the light-duty service information rule, we are
aware that it is costly for light-duty manufactures to maintain the tiered approach on their web sites.
However, light-duty manufacturers have an increased opportunity to recover at least some of those
costs given the relatively larger volumes of aftermarket service that purchase the information. To
require heavy-duty manufacturer to retroactively adopt the tiered approach to their existing web sites
would be resource intensive and not cost effective given the relatively smaller number of heavy-duty
aftermarket shops that would subscribe to the web sites and could ultimately drive up the cost that
heavy-duty manufacturers would need to charge aftermarket service providers.
81
-------
While EPA would encourage manufacturers to make available a reasonable range of web
site subscriptions options, EPA will be finalizing the same provision for web site access that CARB
uses. It will not require that manufacturers make a range of subscription options available. However,
any manufacturer intending to only offer a yearly subscription must demonstrate to EPA that a single
subscription option would be comparable to the rates that a competitor charges for monthly or daily
subscriptions.
In response to Caterpillar's comments, EPA understands this comment to be recommending
that EPA not require heavy-duty manufacturers to develop and maintain an e-commerce site where
an aftermarket shop could directly purchase their subscriptions to the manufacturer's service
information web sites. EPA believes that Caterpillar is recommending that, while they would still
have a web site, Caterpillar would like the option to allow a manufacture to establish a process
where an aftermarket service provider would call a 1-800 number or an authorized dealer in order to
process the transaction, including the payment and the issuing of passwords and user ids.
Section 86.010-300(4) of the regulations state that '....a manufacturer shall provide or
cause to be provided... .a manufacturer specific World Wide Web site...." EPA believes that this
regulatory language provides the type of flexibility that Caterpillar is looking for to implement this
requirement. To the extent that Caterpillar or any other manufacturer wants to implement something
other than a traditional e-commerce web site, EPA will work with that manufacturer to ensure that
the process they put in place does not put undue burden on the aftermarket to gain access to the
web sites and otherwise meets on the requirements for web based access to information as finalized
in this regulation.
IV.C.3 Pricing Approval
Comments:
EMA commented that EPA has proposed that manufacturers obtain approval from EPA of
the pricing structure of their Web sites 6 months in advance of launching their Web sites. Such a
requirement is overly burdensome and unnecessary. Pricing approval should not be needed if EPA
eliminates the tiered access approach. At a minimum, EPA should revise the requirement to no more
than 60 days' advance notice.
Agency Response:
EPA believes it is reasonable to finalize a requirement that manufacturers must obtain
approval from EPA of the pricing structure of their web sites within 60 days of launching their web
sites with any changes needed to comply with this regulation.
IV.C.4 Limitations on Makes and Models
Comments:
EMA commented that EPA has proposed that manufacturers may not limit subscriptions to
just one make or one model. EMA commented that this is different than what heavy-duty engine
manufacturers do today, and must be revised to reflect current access in the heavy-duty market. In
many cases, heavy-duty engine service providers only want or need information on one model.
Along with elimination of the tiered pricing structure, EPA should eliminate any requirement to offer
service information on all makes and models within a single subscription.
Agency Response:
82
-------
EPA does not agree with EMA that we should eliminate requirements to provide service
information for all makes and models within a single subscription. Since EPA is not finalizing
provisions that would require short-, mid-, and long-term access to the manufacturers web sites, it is
even more important that aftermarket service providers have full access to all of the required
information within a single subscription rate in order to keep the price of access reasonable. In
addition, manufacturers who currently have web sites available to their dealer networks do not limit
access by make or model. To the extent that these manufacturers will open their web sites to the
aftermarket, we do not believe that these manufacturers would not spend the resources to limit
aftermarket access in anyway. Therefore, in order for consistency across manufacturer web sites,
EPA will finalize a provision that manufacturers may not limit web site access to one make or model.
IV.C.5 Length of Information Availability
Comments:
EMA commented that EPA has proposed that engine manufacturers maintain the required
full-text information on their Web sites for a minimum of 15 years after model introduction. EPA also
has requested comment on whether it should require information to be retained for a longer period of
time due to the longer service life of heavy-duty engines and vehicles. EPA should not lengthen the
time required to retain information in full-text format on Web sites for longer than 15 years. The
provisions to provide access to archived CDs for data older than that to the extent that the same
information is provided to a manufacturer's own dealers will be sufficient.
Caterpillar also supports a requirement to keep online service information for 15 years.
AAIA, et al commented that the fifteen year limitation on the availability of information on a
web site, which they believe is too short for light duty vehicles, is way too short for heavy duty
vehicles. In its 2006 Transportation Energy Data Book, the US Department of Energy estimated the
median lifetime for heavy duty trucks manufactured in 1990 at 28 years. EPA itself in the fleet
characterization for its Mobile 6 emissions model estimated that in 1996 over 2 .6 million heavy duty
vehicles over 15 years old were still on the road. And it is these older vehicles which are more likely
to be taken to aftermarket service providers for repair. If EPA allows information on vehicles over 15
years old to be deleted from the engine manufacturer's web site, the aftermarket will be severely
disadvantaged. We believe that the information should remain available on the web site fora
minimum of 25 years.
Agency Response:
While EPA agrees with the comments of AAIA et al. that heavy-duty engines have a
significantly longer useful life than light-duty engines, we do not believe it is necessary to require
manufacturers to maintain full-text information on their web sites for longer than 15 years.
Manufacturers will be required to post indexes of information and how to order that information once
the 15 year window has expired. We believe that this will be more than sufficient to meet the needs
of information availability to the aftermarket.
IV.C.6 Correcting Broken or Deleted Web Links on a Weekly Basis
Comments:
Cummins commented that requiring that every link must be checked weekly is unreasonable
and that EPA should change the language in Sections 86.010-38 (j)(4)(ix) to read:
83
-------
"Correct or delete any reported broken Web links on a weekly basis."
Agency Response:
EPA did not intend this language to imply that every link must be checked weekly, only that
know broken or deleted web links be addressed in a timely manner. In order to be clear, EPA will
change the regulatory language of this section as recommended by Cummins.
IV.C.7 Printing of Information from Web Sites
Comments:
Cummins commented that EPA is requiring the ability to print out any and all of the materials
required to be made available on the manufacturers Web site, including the ability to print it at the
user's location. However, some of the materials available on their web site, cannot be printed on
8.5x11 or 11x17. For example, wiring diagrams may be as large as 36"x72". In addition, some
material is copyrighted. Cummins comments that EPA should remove this requirement.
Agency Response:
While EPA agrees that there may be some material that is not easily printable in the average
shop, we do not believe it is necessary to remove this requirement altogether. We will clarify the
regulatory language such that only material that the Web sites allow for printing of documents that
can reasonably be printed.
IV.D Availability of Training Information
IV.D.1 Delivery of Training Information
Comments:
EMA and Cummins commented that EPA has proposed that training information be shipped
within 24 hours of request, with no consideration for weekends and holidays. This requirement
appears to be based on the light-duty paradigm where all manufacturers use third party providers for
the provision of service information (which is not true for heavy-duty). Shipment within 24 hours is
overly burdensome and would impose unnecessary costs on manufacturers. Orders for training
materials are not requested or shipped daily, and engine manufacturers should not be required to
expend the resources to dedicate one person to this job for which the expected sales rates will be
extremely low. Manufacturers should not need to staff fulfillment centers on weekends and holidays
to meet a 24-hour deadline. Shipment of training information within 3 business days is a more
appropriate turnaround time. In its service information rule, California recognized that fact and has
provided additional time, including not requiring shipment of information on Saturdays, Sundays, and
federal or California holidays.
In addition, Cummins commented that EPA should remove the requirement to fax the
information to the requestor if less than 20 pages as fax technology is outdated.
ATD commented that the statute makes no mention of training or training information, neither
should any final rule. Truck dealerships clearly invest thousands of dollars annually in off-site,
satellite, and web-based product specific technician training. However, ATD understands that since
84
-------
the light-duty rule took effect, there virtually has been no demand for "dealer" training information
from the service aftermarket. In addition, third-party training providers, while not covered by the
statute, are readily able to obtain the information they need to conduct service information training.
Volvo commented that they do not support the use of outside training because of the danger
of improper use and that EPA should remove §§86.010-380(11)(ii) and (j)(11)(iii) from the
regulations entirely.
Agency Response:
In response to the EMA and Cummins comments on the shipping requirement, EPA agrees
that it is reasonable to allow a 3 business day turnaround for the distribution of training information
and will change the regulatory language accordingly. EPA also agrees with the Cummins comment
that is reasonable to remove the requirement that information less than 20 pages be faxed to the
requestor and will change the regulatory language accordingly.
In response to the AID comment regarding the general provision to require that training
information be made available, EPA believes that training materials are covered within the universe
of "information including instructions for making emission related diagnosis and repairs." EPA does
not dispute that dealerships make significant investments in training technicians and that there are
third party training resources available to the heavy-duty aftermarket. We are simply requiring that
manufacturers make available these same training materials to aftermarket service providers.
In response to the Volvo comments, EPA does not agree that we should remove §86.010-
38(j)(11)(ii) from the regulations. This section of the regulations require that manufacturers make
available directly to aftermarket service providers an index of all of their available emissions related
training materials and how those materials can be ordered. This provision is not intended to provide
an avenue for third party training providers to use that information to conduct their own training.
Rather it is intended to be a direct way for an interested aftermarket service provider to have access
to manufacturer specific training materials. EPA will include changes to the regulatory language to
make sure this is clear and will retain §86.010-38(j)(11)(ii) in the regulations.
With regards to §86.010-380(11)(iii), this provision would require manufacturers to allow to
third party training providers to license emission related training courses transmitted via satellite or
the Internet for the purposes of repackaging that information for aftermarket service providers. This
provision is part of the light-duty service information and to date, there has been no demand from
third party training providers to utilize this provision of the regulations and we believe there we be
little or no demand on the heavy-duty side as well. Therefore, EPA will remove §86.010-380(11)(iii)
from the regulations.
IV.D.2 Availability of Classroom Training
Comments:
Caterpillar commented that they currently provide extensive online service training through
our service information website. On an annual basis, the number of people attending Caterpillar-run
onsite training is less than 1% of the number of people attending online training. Based on this
information, Caterpillar believes the aftermarket demand for onsite training of qualified service
technicians, if any, is extremely limited. Therefore, we recommend that the EPA not require heavy-
duty engine manufacturers to make classroom training available to the aftermarket.
Agency Response:
85
-------
EPA did not propose any provisions that would require manufacturers to open their class-
room training to aftermarket service providers. EPA did propose, and will finalize, provisions that
require manufacturers to make their training materials and information available, but there is no
requirement regarding the availability of class room training.
IV.D.3 Duplication of Training Information
Comments:
Cummins commented that section 86.010-38 (j)(11)(i) requires manufacturers to video tape
or otherwise duplicate and make available for sale on manufacturer Web sites within 30 days after
transmission any emissions-related training courses provided to manufacturer franchised
dealerships via the Internet or satellite transmission. Cummins commented that duplication of
transmitted emissions-related training courses should not be required if anyone engaged in the
repairing or servicing of heavy-duty engines has the opportunity to receive the Internet or satellite
transmission, even if there is a cost associated with the equipment required to receive the
transmission.
Cummins recommended adding the following regulatory language to this section:
"The manufacturer shall not be required to duplicate transmitted emissions-related training
courses if anyone engaged in the repairing or servicing of heavy-duty engines has the
opportunity to receive the Internet or satellite transmission, even if there is a cost associated
with the equipment required to receive the transmission."
Agency Response:
EPA agrees that it is not necessary for manufacturers to duplicate training courses that are
provided to dealer networks via the web or satellite if those same courses are made directly
available to the aftermarket at a fair and reasonable cost. However, EPA will require that, to the
extent that manufacturers may already duplicate these training courses for other purposes, these
duplicates must be made available to for purchase from the manufacturer at a fair and reasonable
price.
IV.E Service Information for Third Party Information Providers
Comments:
EMA commented that third party information providers have not played a large
role in the heavy-duty service information segment in the past, except for the limited case of
cross-over models that used light-duty cabs and components. Heavy-duty engine and vehicle
manufacturers directly sell service publications to their customers. There is no need for provisions in
the rule related to third party information providers.
Caterpillar commented that they did an inspection of websites for current major 3rd party
providers of service information which shows that these providers do not currently provide heavy-
duty service information. Many of these companies have been in the service information business for
decades, far before any service information regulation was ever introduced. If these companies have
not developed a business case for selling heavy-duty service information, it is unlikely that any new
service information regulation will cause this to situation to change. Caterpillar believes that market
86
-------
conditions do not support a need for mandating engine manufacturers to make service information
available to 3rd party providers.
Agency Response:
EPA did not propose an absolute requirement that manufacturers MUST provide their
information to third party information providers. Rather, the intent of this provision is to ensure that
manufacturers would be willing to consider working with third party information providers who were
willing to enter into proper licensing and/or other types of business arrangements to allow for the
consolidation of service information by third party information providers. While there may not
currently be any demand by third party information providers to enter into these arrangements with
manufacturers, EPA believes it is appropriate to maintain these provisions in the final rule to ensure
that manufacturers reasonably consider these arrangements should they ever be requested to do so
by third party information providers. EPA will revise the regulatory language to make this clearer.
IV.F Recalibration Information
Comments:
EMA commented that EPA should not proceed with finalizing requirements that
manufacturers provide recalibration and reprogramming information to third party service providers
under the Proposed Rule. If, however, EPA ultimately decides to proceed with that aspect of its
Proposal, then any requirements that heavy-duty service tools be made available to third party
service technicians (including any and all provisions of 0(13)) must not be required until at least
2013.
The ARB Service Information Rule (13 CCR 1969), currently in the amendment process to
incorporate provisions specific to heavy-duty, would require reprogramming capabilities to be
provided for heavy-duty engines beginning in 2013. Although originally proposed for 2010, ARB
changed that date to 2013 in recognition of several factors.
One of those factors ties directly in to OBD standardization requirements. The heavy-duty
OBD standardization requirements, which direct compliance with certain standards for OBD and
provision of service, do not become effective under either the ARB heavy-duty OBD Rule or the
proposed EPA rule until 2013. Thus, there is no requirement that heavy-duty engines use
standardized protocols until 2013. The timing of service tool and information requirements, if any,
must follow the OBD timing.
Agency Response:
EPA does not agree with commenters that we should not finalize provisions for the
availability to recalibration information to aftermarket service providers. This provision requires that
manufacturers must make available any emissions-related recalibration or reprogramming events
that may impact emissions. EPA considers a recalibration or reprogramming event to be a pre-
determined ECU software update that a manufacturer issues to address routine issues such as
driveability or some other routine update - only emission-related recalibration or reprogramming
events must be made available under these regulations. These events are generally made available
via CD to dealer networks, and while they may not occur very often, EPA believes these events,
when they are needed, are a critical repair procedure to which the aftermarket must have access
without having to return the vehicle to a dealer or authorize service network. EPA does agree that it
is appropriate to require implementation of this provision beginning with the 2013 model for
87
-------
consistency with CARB as well as the other standardization requirements of the OBD rule, which
take effect in the 2013 model. The regulatory language will be updated accordingly.
IV.G Pass-through Reprogramming Capabilities
Comments:
Caterpillar commented that the process to prepare a vehicle for programming and the
programming event itself is a procedure that is much more complicated in heavy-duty than what
light-duty manufacturers are required to implement. On a heavy-duty vehicle, there are always
multiple devices on the data link that communicate with one another. These devices are accessible
through the OBD connector. It is likely that more than one communicating module will be an OBD
component. For example, the instrument cluster that drives the MIL (malfunction indicator lamp) is
often a separate OBD-related control module that communicates on the data link. For Caterpillar,
programming of a control module involves public data link commands that communication with
external modules on the data link as well as proprietary command that execute the reprogramming.
Because reprogramming is intended to be a secure process, to inhibit tampering, some of these
proprietary programming commands are not standardized and may not be communicated through
standardized pass-through tools. Requiring manufacturers to support standardized pass-through
programming will make the reprogramming process more complicated, thus more costly, and less
secure from tampering. Caterpillar agrees with the EPA's comment that for light-duty vehicles, a
reprogramming event "does not occur very frequently", but would like to add that reprogramming is
even less frequent for heavy-duty engines. If the EPA required heavy-duty engine manufacturers to
support standardized pass-through programming, based on information already discussed above,
Caterpillar estimates the volume of pass-through programming use in the heavy-duty aftermarket will
be less than 1% that of the light-duty aftermarket.
Essentially this would be 1% of an occurrence rate that the EPA already recognizes is very
small to begin with. Also, the reason the EPA has indicated why standardized pass-through
programming is needed is "to prevent the need for aftermarket service providers to invest in
expensive OEM-specific or specialty tools to complete an emissions-related repair". This is simply
not accurate for heavy-duty. Caterpillar's proprietary programming adapter can currently be
purchased by the aftermarket at a very reasonable price of approximately $1000.
Considering the extremely low rate at which reprogramming will occur in the heavy-duty
aftermarket, and the potential for tampering if compliance is required to a standardized pass-through
programming tool, Caterpillar urges the EPA to not make pass-through programming a requirement
for heavy-duty service information.
Agency Response:
EPA does not agree with Caterpillar comments that we should not require standardized
pass-through reprogramming in this regulation. It is not clear to EPA why Caterpillar would oppose
this requirement since EPA believes that Caterpillar, and all HD engine manufacturers currently
support pass-through reprogramming through RP1210B, which is the equivalent of the light-duty
SAE J2534 standard for pass-through reprogramming. Codifying this requirement is intended to
ensure future consistency among all manufacturers and does not place any additional burden or
require the manufacturers to do anything differently than they are doing today.
88
-------
IV.H Availability of Generic and Enhanced Information for Scan Tools for Equipment and Tool
Companies
IV.H.1 Generic Information
Comments:
EMA commented that the requirement to make available generic scan tool information
should be deferred until 2013, due to lack of standardization requirements as well as the time
manufacturers need to build in appropriate safeguards to ensure the proper transfer and use of tool
development information. From 2010 through 2012, engine manufacturers propose to continue to
sell their manufacturers' tools to meet the service industry's needs for diagnostic service tools.
Cummins commented the requirement for bi-directional control information should be
changed to be required in 2013.
Agency Response:
Given that OBD systems will not be fully implemented on all engines until 2013, EPA
believes it is reasonable to require that all generic and enhanced scan tool information be made
available at that time as well and will change the regulatory language accordingly.
IV.H.2 Enhanced Information
Comments:
Caterpillar commented that OBD requirements mandate that OBD modules communicate
diagnostic information using J1939 (or ISO15765). Caterpillar's implementation of J1939 is far
beyond that which is required for OBD, and includes much of the information that would normally be
regarded as "enhanced data link information" for light-duty vehicles. Most engine operating
parameters and status information are provided using standardized data link protocols. This is a
practical requirement driven by the non-vertically integrated heavy duty vehicle industry.
The services in J1939 are extremely comprehensive for the development an aftermarket
service tool. If a company decides there is a market for such a heavy-duty service tool, the capability
and information is already available on standardized J1939 to create a full featured diagnostic tool.
From a diagnostic/repair perspective, there is little reason why aftermarket service tools would use
enhanced data link information. We recommend the EPA review the information that is available
from J1939 standards to understand why there is not a need to provide enhanced data link protocols
for developing service tools in the heavy-duty vehicle market.
Providing "enhanced" data link information to the aftermarket would provide the basis to
create service tools which do have the ability to make permanent changes to engine configurations.
Manufacturers must be allowed to keep all proprietary data link information confidential in order to
prevent tampering and misconfiguration. Because of J1939 standardization, full-featured service
tools can be created using standardized protocols. Caterpillar urges the EPA to not require engine
manufacturers to provide enhanced data link information to aftermarket tool manufacturers.
Agency Response:
EPA interprets Caterpillar's comments to be recommending that EPA not require
manufacturers to make available any enhanced diagnostic information to equipment and tool
89
-------
companies because that information contains the data that would allow equipment and tool
companies to incorporate the ability to reconfigure engines into aftermarket scan tools. EPA
understands the concerns that manufacturers have with releasing reconfiguration information to
equipment and tool companies. In fact, EPA believes that scan tool companies have no desire to
license reconfiguration information from the manufacturers because devising "generic"
reconfiguration capabilities is not possible. However, EPA does not believe that this means that we
should not require the availability of any enhanced information at all for those equipment and tool
companies that are willing to license it from the manufacturers. Therefore, EPA will finalize
provisions that require the availability of enhanced diagnostic information to equipment and tool
companies, but will add language to the regulations that this enhances diagnostic information does
not include any information related to reconfiguration capabilities.
IV.I Availability of OEM-Specific Diagnostic Scan Tools
IV.1.1 General Provisions for Tool Availability
Comments:
EMA commented that Service tools for heavy-duty are much different for the heavy-duty
service industry versus the light-duty service industry. EMA commented that heavy-duty scan tools
are much more complex than for light-duty, which is necessary to be able to provide the
customization and performance features that are required in a horizontally-integrated market. The
recalibration (or re-flashing) and reprogramming (or reconfiguration) of heavy-duty emission-related
engine control modules are more complicated than light-duty and require reconfiguration of many
engine, vehicle and customer features. The reprogramming and calibration process is time
consuming and requires a higher level of training and more powerful electronic tools compared to
light-duty. Current heavy-duty reprogramming and calibration tools (in the form of software) have the
power to change the horsepower and torque on an engine, which are some of the very features
which engine manufacturers sell. The nearly impossible challenge for manufacturers under the
Proposed Rule is how to make such tools and systems secure, yet "open" enough for third parties to
be able to use them.
EMA further commented that, if EPA proceeds to require the sale of recalibration and
reprogramming tools, manufacturers would need time to invent new mechanisms to change the way
service and tools are currently provided. Providing reasonable controls to prevent the misuse of
service calibrations by the general public needs more lead time than allowed by a 2010 requirement,
in part, due to the systems approach that must be taken to 1) modify engine control systems to
accept the new calibration file concepts, 2) create new tools to guide the selection of a calibration file
and upload it into engine controller; and 3) change existing or create new service support systems
that maintain engine (and vehicle) configurations.
In current products, data included in reprogramming and recalibration information includes
data that selects the power and torque characteristics of the engine, which are revenue drivers for
engine and vehicle manufacturers. Manufacturers also need more time to find ways to protect the
engine and vehicle manufacturers' revenue streams and warranty exposure before these powerful
tools can be placed in the hands of third parties. Creating a duplicate tool system that is powerful
and effective for the customer, yet "novice-safe," is extremely burdensome.
Novice-safe calibration systems would cost engine manufacturers thousands of dollars per
calibration performed. Manufacturers need time to make those systems as cost-effective as practical.
90
-------
Those systems would duplicate existing systems, which must be maintained to service existing
engine electronic control systems.
Indeed, in order to assure that manufacturers' proprietary and trade secret information is not
divulged to third-party service providers and scan tool makers, heavy-duty engine manufacturers
would have to undertake substantial re-designing of current software. For example, with respect to
the requirements governing availability of datastream and bi-directional control information,
manufacturers first would need sufficient time to undertake the extensive work that would be
necessary to organize electronic control module (ECM) software code so that access to proprietary
code is adequately secure and cannot be obtained by unauthorized parties.
Manufacturers would need time to develop and test whether the new controls and
infrastructure are viable with their own service providers. Once that has been accomplished,
manufacturers then would need additional time to determine how the information could be extended
to the aftermarket service industry. Making available to the aftermarket reprogramming and
calibration methods would require engine manufacturers to modify existing business practices and
try to find ways to minimize the risk that engine control system software changes could be
misapplied to engines.
In other words, with the introduction of OBD requirements in new engine products beginning
in 2010, how heavy-duty engine manufacturers would provide service tools to dealers and
authorized service networks while maintaining the necessary security controls is not yet clear, let
alone how they would provide such information to the aftermarket industry. The entire infrastructure
governing the provision of service information today would need to be re-built so that neither
authorized dealers and service networks nor third parties can obtain access to manufacturers'
proprietary information.
In fact, even with changes to the way the heavy-duty service tools and infrastructure work
now, engine manufacturers have significant concerns with regard to the possibility for tampering that
may arise when the service tools and information required by the Proposed Rule are provided to
non-authorized service providers. As aftermarket providers are given the tools not just to service, but
to calibrate and reconfigure engines, there is a possibility that inadvertent or deliberate mis-
configuring may occur.
EMA appreciates the fact that the Proposed Rule includes provisions that would allow heavy-
duty engine manufacturers to require training of those using their service tools, and we support such
a condition. However, while third party service providers can, in theory, be trained to do the same
checks as factory-authorized service facilities, the fact that they typically service several different
brands of engines, each with their own idiosyncrasies as far as configuration variations, means that
they may be "less specialized" and more likely to make mistakes than factory-authorized outlets that,
in many cases, focus on servicing engines from a single manufacturer. As described more fully
below, such training is not sufficient to address the concerns raised by the tool availability
requirements of the Proposed Rule. EPA must make further significant changes to the Proposed
Rule for it to be workable and cost-effective for the heavy-duty industry and to limit manufacturers'
liability for third party actions under the Proposed Rule.
EMA also commented on EPA's proposal to make reprogramming and recalibration tools
available to aftermarket service providers. EMA commented that engine manufacturers recommend
that EPA exempt heavy-duty engines from reprogramming and calibration tool requirements. Sales
of reprogramming and calibration tools place engine manufacturers at greater risk that their
proprietary access schemes for secured engine programming will be compromised by increased
public exposure. Reprogramming and calibration tool techniques cannot be devised to provide what,
as a practical matter, is a nearly "hands off method for reprogramming and calibrating heavy-duty
91
-------
engines that would ensure that tamperproof features for engine power settings, idle shutdown
systems, and vehicle speed limiters will never be compromised. Changing manufacturers' existing
methods for supporting engine control system replacement and reprogramming needs - in order to
accommodate a requirement for publicly-available reprogramming and calibration tools - is not cost-
effective or justified in terms of the benefits to commercial vehicle owners. Requiring aftermarket
availability of reprogramming and calibration tools also creates conflicts with existing regulations for
limiting engine idling (specifically, California's idle shut-down requirements) and a proposed federal
regulation for limiting vehicle speeds.
In addition, recent regulatory actions seek to regulate these features by specifying a fixed
time that vehicles are allowed to idle or by specifying a maximum vehicle speed limit setting. In 2008,
a recently-adopted California regulation (13 CCR 1956.8 (a)(6)) will require heavy-duty engines to
stop idling after 5 minutes. This is achieved through a required idle shutdown system on the engine
system produced by engine manufacturers. ARB regulations further require manufacturers to
prohibit vehicle owners from disabling (i.e., make "tamperproof) the fixed idle time system by turning
it off or by extending the idle time beyond 5 minutes. Further, NHTSA is currently considering a fixed,
or maximum, vehicle speed limit of 68 mph. as discussed in docket number NHTSA-2007-26851.
Engine manufacturers anticipate that a national vehicle speed limit setting of 68 mph would require
manufacturers to prohibit vehicle owners from turning vehicle speed limiting off or increasing the
vehicle speed limit setting above 68 mph, in a manner similar to that which is required by the ARB
idle shutdown rule.
The ARB regulation exempts many vehicle applications from the fixed idle shutdown limit,
including buses, emergency vehicles, and military tactical vehicles. Engine manufacturers must then
provide an engine control system that is both capable of shutting down the engine after five minutes
and capable of omitting the engine idle shutdown feature on exempt vehicles, because the same
engines are sold in both exempt and non-exempt applications. Since duplicate engine control
modules are not efficient or cost-effective and cannot be relied upon to control access for non-
exempt engines, engine manufacturers have proposed to use existing proprietary access control
schemes to address exempt vehicles. These schemes - and ARB's regulation - require
manufacturer participation to authorize or enable the change, and permit manufacturers to exercise
control over undesirable changes. Since the same engine is sold in exempt and nonexempt
applications, the purchase and installation of an "exempt" replacement part on a nonexempt vehicle
would defeat the fixed idle shutdown requirement, if a duplicate engine control system were used.
Manufacturers view the existing proprietary access schemes as cost-effective measures that
minimize risks that vehicle owners could turn off idle control systems on nonexempt vehicles, without
requiring duplicate engine control systems.
EMA also commented that even though recalibration/re-flashing events may occur more
frequently than reprogramming/reconfiguration of heavy-duty engines, making recalibration tools
available to the aftermarket is not appropriate or justified. Typically, re-flashing of software occurs to
address emissions issues and performance complaints arising during the warranty period. When
such repairs or service are done during the warranty period, they are usually undertaken by
manufacturers or authorized dealers. Moreover, in most cases, even if re-flash could be undertaken
to address emissions issues, reconfiguration or reprogramming of the engine (by the manufacturer)
also is necessary to make a vehicle drivable. In other words, it is difficult, if not impossible, to
separate the tools to flash and calibrate an ECM from the tools to reconfigure or reprogram an ECM.
In many cases, an entirely new calibration must be flashed in (not just limited to certain parameters).
ATD commented that the proposal would require heavy-duty OEMs who make
reprogramming available through dealerships either to offer reprogramming tools or to provide
aftermarket tool and equipment companies with reprogramming information. The Engine
Manufacturers Association indicates that this will present heavy-duty engine OEMs with
92
-------
considerable challenges and thus should not be mandated. AID agrees. Fortunately, all service
providers will have the option to sublet for emissions-related reprogramming with dealerships and
other OEM-authorized service providers that have the capability. AID is confident that all heavy-duty
engines in need of OBD-related emissions reprogramming will be able to get the service.
Agency Response:
Based on these comments, EPA understands that there are multiple types of OEM-specific
scan tools used by heavy-duty engine manufacturers. In addition to enhanced diagnostic scan tools,
some heavy-duty engine manufacturers have also developed separate tools for
recalibration/repgramming/reflashing as well as separate tools for reconfiguration. For clarity EPA
believes it is necessary to define all of these terms clearly. On the light-duty side, EPA and the
light-duty industry generally use the terms recalibration, reprogramming and reflashing
interchangeably to indicate a software update of an ECU with default vehicle parameters and
configurations determined and distributed by the manufacturer. These updates are not in any way
changeable by the end-user. They are merely loaded into the vehicle, either with an OEM-specific
scan tool or an aftermarket pass-through tool. The heavy-duty industry uses the terms recalibration
and reconfiguration interchangeably to indicate software that can be adjusted by the end-user for a
particular vehicle and/or application. To ensure that the terminology is consistent, EPA has adopted
the CARB definitions for recalibration (which now also means reprogramming and reflashing) and for
reconfiguration and included these definitions in the regulatory language.
In response to the comments regarding the aftermarket availability of recalibration tools, EPA
believes that these tools must be made available to aftermarket service providers since the
functionality they provide is essential to the proper completion of emissions-related repairs. Even
though EPA is finalizing provisions for generic pass-through recalibration tools, there will likely be
aftermarket service providers who will want to purchase the OEM-specific recalibration tools and we
believe that these tools must be available. In addition, it is EPA's understanding that there are
several manufacturers who are currently making these tools available without the need for any
significant additional controls before making them available to the aftermarket, nor have any these
manufacturers reported any issues misuse with the aftermarket currently having access to these
tools. Lastly, EPA has included provisions recommended by EMA to optionally require training of
aftermarket service providers as a condition of the purchase of these tools as well as the additional
liability language requested by EMA Therefore, EPA believes that there is no compelling reason to
preclude aftermarket availability of recalibration tools.
With regard to reconfiguration tools, EPA appreciates EMA concerns regarding the
aftermarket availability of reconfiguration tools. It is EPA's understanding that reconfiguration is not
always necessary for the completion of an emissions-related repair. We understand that
reconfiguration in most situations permits end-users to change engine configurations to meet the
needs of the individual purchaser of the engine/vehicle. While EPA understands that the ability to
reconfigure an engine may not always be directly tied to an emissions related repair, we are also
aware that there are manufacturers who combine their recalibration functionality and reconfiguration
capability into the same tools. In order to address the manufacturers concerns about the release of
reconfiguration tools, EPA will require manufacturers to release reconfiguration tools only to the
extent that those tools are needed for the completion of an emissions-related repair, including, but
not limited to recalibration functions.
In response to ATD comments, EPA believes it is not reasonable to require aftermarket
service providers to return the dealer or to an authorized service network in order to have an engine
recalibrated. EPA believes this is contrary to the intent of the Clean Air Act to place an aftermarket
service provider in the position of having to rely to on a dealer to complete an emissions related
repair. As discussed above, we believe the requirements we are finalizing with regards to
93
-------
aftermarket access to recalibration procedures are consistent with the vast majority of manufacturers
who are currently making available to the aftermarket the tools and data necessary for performing
recalibration functions.
IV.1.2 Requiring Training as a Condition of OEM-specific Scan Tool Purchases
Comments:
EMA commented that they support provisions that would allow engine manufacturers to
require training as a condition of sale of manufacturer-specific diagnostic tools. Engine
manufacturers currently require their own authorized service providers to attend training at
manufacturers' training centers. Allowing manufacturers to require training as a condition of sale of
their service tools to aftermarket service providers is essential, as it provides some protection
against misuse of those tools. Training can provide some assurance that purchasers and users of
heavy-duty service tools will be knowledgeable regarding the proper use of the tools. With such
knowledge, they can substantially avoid misuse that could result in improper engine configurations,
possibly leading to increased emissions, engine damage, or other injury.
AAIA, et. al. commented the aftermarket continues to oppose any training requirement for
purchase of a scan or diagnostic tool because it believes such training is unnecessary and could be
used to unfairly limit or restrict the ability of individual aftermarket providers to obtain OBD-related
information. No showing has ever been made that to use a scan or diagnostic tool (as opposed to a
recalibration or reconfiguration tool) special training only available through the engine manufacturer
is necessary. Such a training requirement could impose unnecessary costs on after-market service
providers and create long delays between the time a service provider requests a tool and the time it
"qualifies" to purchase one. We request that EPA eliminate any training requirement for purchase of
manufacturer specific scan tools and in fact add language that prohibits a manufacturer from
imposing such a requirement before a scan or diagnostic tool can be purchased.
AAIA, et.al, further commented that, if training is required, the rule must establish more
specific parameters for how that training must be provided. Without some enforceable guidelines, we
are concerned that some manufacturers could delay providing training or make training too
expensive or inconvenient to discourage aftermarket service providers from obtaining its tools. We
would encourage EPA to adopt limitations similar to those in Section 1969 (h)(1)(A) of the California
regulations. These require that training only can be required if it is required by the manufacturer for
purchase of the tool by its dealers or members of its authorized service network; that the training
must be substantially similar to that required of the dealers or members of the authorized service
network in terms of material covered and length of classes; that the training must be provided at a
fair, reasonable and nondiscriminatory price; that it must be available within six months after a tool
request has been made ; and that it must be available at a location reasonably available to the
requesting aftermarket facility. Without adding these or similar limitations to the rule, EPA is opening
the door for abuse of the training requirement.
Agency Response:
While EPA agrees with AAIA, et.al. that there has been little evidence on the light-duty side
of any misuse or other significant issues with regard to making available OEM-specific tools
available to the aftermarket, we do also understand that heavy-duty scan tools provide significantly
more capability than light-duty tools. In addition, heavy-duty engines are highly configurable to meet
the needs of individual customers and agree with EMA that some safeguards should be put in place
to ensure that aftermarket service providers can use these tools effectively. Therefore, we will put in
place provisions that allow manufacturers to require training as a condition of the sale of a scan tool
to an aftermarket service provider. EPA does agree with AAIA, et.al. that EPA must ensure that
94
-------
manufacturers do not misuse this requirement as a means to unreasonably deny access to
aftermarket service providers who may wish to purchase OEM-specific tools. Therefore, we will
adopt language similar to that found in the CARB heavy-duty service information regulations which
place strict guidelines under which manufacturers may require this training to ensure that there is
reasonable access to this training.
IV.J Reference Materials Being Proposed for Incorporated by Reference
Comments:
EMA commented that EPA has proposed the use of SAE J2534 and TMC RP 1210Afor
standardization of engine diagnostic tools and engine reprogramming and calibration tools. EMA
supports the use of those standardized tools (beginning in 2013), but notes that a new version of
TMC RP 1210 (121 OB) is completing its approval process within TMC. EMA will inform EPA when
the TMC RP 121 OB revision is complete and published by the American Trucking Association.
Agency Response:
EPA agrees with commenters that we should refer to latest version of TMC RP 1210. EMA
submitted their comments in May of 2007 and TMC RP 121 OB was finalized in June of 2007.
Therefore, EPA will Incorporate by Reference TMC RP 1210B in this final rulemaking.
IV.K Other Comments Received
IV.K.1 Costs Associated with the Rule
Comments:
EMA commented that EPA has not provided any estimate of the costs that engine
manufacturers would incur to comply with the requirements of the Proposed Rule. Such a lack of
cost information is a significant concern in light of the substantial nature of the changes to the heavy-
duty service industry that the Proposed Rule would require.
When California looked at the costs related to its proposed service information rule for the
heavy-duty industry, it estimated a heavy-duty engine manufacturer's start-up costs under its rule
would be likely to reach as high as $1.5 million per manufacturer, with yearly maintenance costs of
approximately $70,000 per year per manufacturer (Rulemaking To Consider Proposed Amendments
To Regulations For The Availability Of California Motor Vehicle Service Information (June 22, 2006),
Staff Report: Initial Statement of Reasons For Proposed Rulemaking, p.11). Engine manufacturers
believe those estimated costs could well be higher, particularly for yearly maintenance of Web sites
and tool availability. ARB Staff at that time suggested that engine manufacturers could recover some
of those costs from the sale of tools and information. But considering the sales volumes in the
heavy-duty industry, and the number of independent service outlets in the industry, heavy-duty
engine manufacturers simply cannot recoup those costs by selling their tools and information.
Based on information available from public industry sources, light- to heavy-duty vehicle
sales volume is approximately 40 to 1. As noted above, there are a limited number of heavy-duty
engine products, but an enormous number of variations on how those products may be configured
and calibrated. Adding together those factors, heavy-duty engine manufacturers have little
opportunities to spread out and recover the costs of the Proposed Rule.
95
-------
Furthermore, engine manufacturers make their tools and information available now to
anyone who wishes to purchase them, yet there is no great demand for them. And manufacturers do
not anticipate any great increased demand for their tools and information. In fact, for purposes of
comparison, we have recently obtained from light-duty manufacturers their experience with requests
for service information. One light-duty vehicle manufacturer has received through its Web site over
the course of one year only 43 requests for year-long subscriptions to service information and only
55 requests for month-long subscriptions from service providers nationwide. Another light-duty
vehicle manufacturer has received only 147 year-long subscription requests and only 27 month-long
requests nationwide. Those subscription unit sales are from two of the three primary U.S.
manufacturers of light-duty vehicles reporting nationwide data. Using that information, if there
existed a heavy-duty engine manufacturer with the same volume of subscription sales as those two
light-duty manufacturers combined, then calculating the 40 to 1 light- to heavy-duty volume ratio, that
heavy-duty engine manufacturer could be expected to sell just 4-5 year-long and 1-2 month-long
subscriptions per year. The costs of the Proposed Rule so outweigh its anticipated benefits that the
Rule cannot be justified.
Added to the above very high costs, the recovery of which is not likely, is the fact that the
Web site requirements necessitate duplicate channels and maintenance. Some manufacturers have
manufacturer Web sites that are not available to the public because of concerns about their ability to
protect and keep certain non-emissions information secure. Once information was required under
the California rule to be made available to the public, some manufacturers created separate public
Web sites to meet those current ARB requirements. Manufacturers would anticipate revising those
public sites as necessary to meet EPA requirements. But, both the manufacturer and the public Web
sites must be maintained, leading to additional costs.
Even though EPA cites Congress's intent that manufacturers be able to recover "reasonable
costs" in connection with the provision of service information, EPA has not provided any data that
would demonstrate the extremely high costs manufacturers are anticipated to incur. EPA has failed
to provide any cost information at all, and has failed to establish that the costs of the Proposed Rule
are justified by any benefits that may be gained.
Moreover, service information fees charged to authorized networks are part of a negotiated
franchise arrangement, in which the authorized dealers have had to invest resources and meet other
conditions as part of the license for what they do. EPA must take such factors into consideration
when reviewing price information, including the fact that many manufacturers currently subsidize the
costs for providing service information to their authorized net works.
Agency Response:
In response to the EMA comments, EMA did not provide any data that would demonstrate
the extreme high costs that manufacturers anticipate they would incur as the result of these
regulations. EMA did not provide any data or information to EPA to demonstrate the need to
develop separate web sites and the significant changes they say will be needed to their tools in
order to prevent misuse or tampering. In fact, EMA suggested throughout their comments that their
member companies already make available essentially the same service information tooling that
EPA is addressing in this rulemaking without issue to the aftermarket.
EPA does agrees that many elements of the proposal would have required manufacturers to
make changes to their existing information and tool distribution infrastructures (e.g. tiered web site
access). However, EPA has finalized the vast majority of the service information provisions in such
a way that allows manufacturers to continue to use their existing infrastructures to distribute
information and tools to the aftermarket without any significant additional burden and in such a way
96
-------
that will allow the manufacturers to recover at least some of those costs through reasonable web site
access charges, training fees, and licensing agreements with equipment and tool providers.
IV.K.2 Alignment with Other Agency and State Requirements
Comments:
EMA commented that EPA must align its service information requirements with California's in
certain respects. ARB has regulations in place requiring heavy-duty engine manufacturers to provide
certain service information to the aftermarket through Web sites. ARB also is in the process of
adopting additional regulations requiring heavy-duty engine manufacturers to provide service tools
and limited tool information to the aftermarket beginning in 2013. EPA's Proposed Rule includes a
number of provisions that are more burdensome than ARB's and which should be revised to align
with the ARB requirements. EMA worked extensively with ARB on the development of ARB's service
information rules. Many, although certainly not all, of the issues that we raised were addressed in the
ARB rulemaking. To the extent ARB resolved those to the mutual satisfaction of ARB and industry,
EPA should align. In some cases, ARB's rule did not adequately address issues associated with the
requirements. In other cases, new regulations have come into play that require significant changes
from the approach taken by ARB and similarly proposed by EPA. In those cases, EPA must revise
the Proposed Rule, and ARB also will need to make further changes to account for those issues.
As discussed in more detail below, there are adopted or pending regulations in California
and other federal agencies that will have an impact on some aspects of servicing heavy-duty
engines. In both cases of which engine manufacturers are aware - California idle shut-down
requirements and federal vehicle speed limiter requirements - certain aspects of an engine system's
electronic controls must be made non-programmable, or programmable only to a limited extent.
Because the goal of service information, including service tool availability, requirements is to allow
aftermarket service providers to service heavy-duty engines, there are potential areas where the
service information requirements and the requirements for nonprogrammable, or "sometimes"
programmable, systems may conflict. EPA must assure that its Proposed Rule does not require
manufacturers to meet requirements that would in any way conflict with other regulatory
requirements and provisions that are in place or currently under consideration.
Agency Response:
In response to EMA's comments, although EMA has not provided specific recommendations
as to where we should specifically harmonize with the CARB regulations, EPA has attempted
wherever possible and appropriate to ensure that manufacturers do not have conflicting regulations
with which to comply. In some cases, EPA will be implementing less stringent provisions than the
CARB regulations. For example, the CARB regulations require manufacturers to make training
information available in full-text on their web sites. EPA's regulations only require that
manufactures make their training materials available for purchase.
In response to EMA's comments regarding other Agency requirements such as idle shut-
down and federal vehicle speed limiters, we don't believe that any of the provisions we are finalizing
today would interfere with the mandates of other state or federal Agencies.
IV. K.3 Scope of the Rule
Comments:
97
-------
EMA commented that EPA's Proposed Rule would require heavy-duty engine manufacturers
to provide service information and tools related to heavy-duty engines used in vehicles over 14,000
pounds. EMA supports the scope of the rule, which does not attempt to cover or require engine
manufacturers to provide transmission or other non-engine system information.
EMA further commented that most engine manufacturers produce engines, not
transmissions and not vehicles. When an engine is sold to a vehicle manufacturer, the engine
manufacturer has no control over what transmission it is paired with or what transmission information
is available. That choice is up to the customer and the vehicle manufacturer. Engine manufacturers
do not and cannot provide information for components over which they have no control. EPA's
proposed heavy-duty OBD rule, on which service information requirements are based, recognizes
that fact. The service information that heavy-duty engine manufacturers are required to provide must
be limited to only engine information as well.
AAIA, et al. commented that a major problem with the proposed rule is that because it is
only to apply to engines, references in the light duty rule to "vehicle" or "vehicles" have been
changed to "engine" or "engines" without regard to the consequences of such a change. While many
of these changes are correct in light of the more limited scope of the heavy duty rule, others are not
and unless changed back will unintentionally deny the aftermarket of necessary repair information or
will otherwise affect the rights provided to aftermarket service facilities under the rule.
As an example, AAIA, et al. points to the definition for "data stream information" which limits
its scope to information originated within the "engine", however, such information might originate in
the fuel, ignition or other system monitored by the OBD system. Therefore, we believe that the
language of this definition has to be returned to the same language of the light duty rule which
defines it as information originated within the "vehicle". The substitution of "engine" for "vehicle" in
the definition of "enhanced service and repair information' is another example because that change
would exclude any information related to other systems, such as the fuel system and ignition system,
which may be necessary to complete the repair. The reference in this section should also be to the
vehicle not just the engine.
Similarly, in Subsection (j)(6)(iii) of the rule, "required information" only includes information
related to any system, component or part of an engine, but this again would exclude information
related to the systems, such as the fuel system and the ignition system, which may be vital in
making the repair. The reference here also needs to be changed from "engine" to "vehicle".
Also the language of Subsection (g)(5)(iv) of the light duty rule regarding information on
systems that may effect emissions in a multiplex system has not been included in Subsection 0(6)
of the new rule . While we agree that this language would have to be modified in the heavy-duty rule
in situations in which the engine manufacturer chose not to monitor a system for OBD purposes, we
believe that, like the California rule, if such information is monitored, it should be provided. Therefore,
we request that this language, modified as may be appropriate, be inserted in the proposed rule.
Finally, changing the word "vehicle" to "engine" in one part of Subsection 0(8) also could
unintentionally adversely affect the cost of the information to the aftermarket. Subsection 0(8) sets
forth the relevant factors that a manufacturer may consider when establishing its cost for the
information to be provided under the rule. The similar section of the light duty rule has language that
was designed to exclude any costs incurred in designing and implementing, upgrading or altering the
OBD system or any other vehicle part or component. In Subsection 0(8)(B) this has been modified
to "any other engine part or component". This modification would allow the manufacturer to pass on
any such costs not related to the engine itself. Thus, under the current language, if a portion of the
fuel, ignition or other system required change, this cost could be passed on to the aftermarket. The
aftermarket should not be reimbursing the manufacturer for such costs and this is certainly not what
98
-------
EPA intended. This language should also be changed back to refer to the parts and components of
the vehicle not just the engine.
Agency Response:
In response to the wording changes recommended by AAIA, et al., EPA does not believe it is
necessary to change the word "engine" to "vehicle". We believe making this change may have the
unintended effect of making the scope of the rule broader than intended since the OBD requirements
being finalized only apply to engines, not to transmission or other non-engine systems. However,
EPA agrees that we need to be clearer on which systems need to monitored and for which service
information needs to be made available. Therefore, to address the comment that not making this
wording change would lead to the interpretation that manufacturers could exclude information
related to systems such as the fuel system and the ignition system or multiplexed systems, EPA has
added the following definition to 86.010-2X of the regulations:
Engine or engine system as used in §§86.010-17, 86.010-18, 86.010-30, and 86.010-38
means the engine, fuel system, induction system, aftertreatment system, and everything that
makes up the system for which an engine manufacturer has received a certificate of
conformity.
We believe that adding this definition to the regulations will ensure that all information needed for
making emissions-related repairs is available.
IV.K.4 Timing
Comments:
EMA commented that EPA has proposed that almost all requirements - availability of
emission-related information (manuals, OBD system information, service bulletins, etc.) on
manufacturer Web sites, availability of recalibration and reprogramming tools to the aftermarket, and
tool development information for tool makers - be made available beginning in 2010. Emissions-
related training information is not required to be made available until 2011. EMA generally supports
requirements for Web site availability of information beginning in 2010 and training information in
2011, but bi-directional control information, software tools with recalibration and reprogramming
capability (if ultimately required) and tool development information should not be required until 2013.
Engine manufacturers already are devoting enormous resources to meeting the 2010 emission
standards. They will devote substantial resources to meeting the 2010 and later OBD requirements
that have been adopted in California and are being developed by EPA. EPA must take all possible
steps to assure that the imposition of any new SIR requirements on heavy-duty engines in 2010 and
2013, in particular, will not impede successful implementation and marketplace acceptance of the
2010 and later heavy-duty exhaust emission standards and OBD requirements.
Agency response:
EPA is requiring the availability of bi-directional control information, software tools with
recalibration and reprogramming capability and tool development in this final rule. To this extent,
EPA agrees that it is reasonable to require compliance with the provisions being finalized for bi-
directional control information, software tools with recalibration and reprogramming capability and
tool development beginning 2013, which coincides with the timing of full OBD implementation on all
heavy-duty engines.
99
-------
IV.K.5 Persons Entitled to Access
Comments:
EMA commented that EPA has proposed that engine manufacturers must provide service
information and tools to "any person engaged in the repairing or servicing of heavy-duty engines."
There are no limitations or qualifications on who may receive these very powerful tools. If EPA does
not exempt heavy-duty reprogramming and calibration tools from the Proposed Rule, then in addition
to allowing manufacturers to require training, EPA should add language to the rule that would allow
engine manufacturers to require purchasers to demonstrate some level of qualification to do heavy-
duty service repairs prior to selling the information and tools. Heavy-duty engines are commercial
products used in commercial service in business-to-business relationships. They are not "do-it-
yourself-type products and should not be treated casually. Requiring heavy-duty service information
to be made available to entities without regard for their qualifications could lead to improper repairs
by under-qualified entities. Such threshold qualification could include information such as name,
address, business licensing information, and other information necessary to demonstrative minimum
capabilities. Manufacturers' goal is not to be exclusionary or to erect "artificial barriers to access" but
to give manufacturers confidence that the persons requesting heavy-duty service tools have the
ability to work with these complex tools and systems, and to avoid unintentional mistakes as well as
deliberate misuse.
AAIA, et al. commented on this issue as well. They commented that the light duty rule
requires access for any aftermarket service provider who was engaged in the diagnosis, service and
repair of motor vehicles or engines. The heavy duty rule does not give access to providers who
service vehicles but only those which service engines. While we hope that EPA did not intend to limit
this access to those providers who only service engines and that this language also covers providers
who service engines as part of servicing the entire vehicle, this more limited language creates a
question as to whether those servicing the whole vehicle are included among those with access to
the information . More importantly, however, there are many aftermarket service providers who work
on systems which are auxiliary to the engine and which are likely to be monitored by the OBD
system, such as the fuel system and the ignition system, who would not be considered engaged in
the diagnosis, service or repair of motor vehicle engines . Therefore, the language must be
broadened to cover service providers who do not specifically service engines. We believe that the
best way to do this is to allow access to the information for any aftermarket provider engaged in the
diagnosis, service and repair of heavy duty vehicles not just engines.
ATD commented that the proposal inappropriately uses the term aftermarket service provider,
defined as any individual or business engaged in the diagnosis, service, and repair of a motor
vehicle or engine who is not directly affiliated with a manufacturer or manufacturer dealership. 49
CFR §86 .010-38(j)(3)(ii)(A). The appropriate term, used in the statute, is any person engaged in the
repairing or servicing of motor vehicles or motor vehicle engines. This latter language is designed to
help ensure that any service information rule applies equally to all persons, including dealerships,
engaged in such activities.
ATD commented that heavy-duty truck dealerships typically handle more than one chassis
and several engine franchises, each of which requires large investments in special (including
emissions-related) tools, in training, and in information. Heavy-duty dealerships also sell out-of-line
used vehicles that may be worked on by other service providers. Without question, franchised
dealerships are very concerned about the ever-increasing costs and complexities of doing service
and repair business. Therefore, EPA should delete all references to aftermarket service providers in
order better to reflect the statute's broadly neutral language and intent.
100
-------
Agency Response:
In response to these comments, EPA does not feel it is necessary to clarify who is entitled to
access to the service information required by this rule, beyond the general limitation that the person
be a "person engaged in the repairing or servicing of heavy-duty engines." We do not believe it is
necessary to add limits or qualifications, as recommended by EMA, as to who is entitled to access.
While not explicitly outlined in the light-duty rules in effect today, EPA has provided flexibility for
manufacturers to reasonably deny access in instances where they had significant concerns about
providing access to an aftermarket service provider. For example, EPA has permitted a
manufacturer to deny access to a specific individual who was purchasing their tools and reselling
them on Ebay. In addition, while EPA recognizes that heavy duty engines are commercial products
that are generally maintained commercially, EPA has added regulatory language recommended by
EMA that limits their liability to any third parties who may misuse their information and tooling so we
believe that further restrictions are unnecessary. EPA is concerned that adding specific language
regarding limits and qualifications could be unintentionally limiting and we do not believe that is
necessary. In response to the AIAM, et al comments, EPA also does not believe it is necessary to
broaden the language any further as to who is entitled to access. We have clarified the language
regarding what systems are covered by these requirements, and believe that service providers not
engaged in service of those systems need not be covered by this rule. Nor does EPA believe it is
necessary to use a term other that "aftermarket service provider" to describe whom this rule applies
to as recommended by AID. The regulations provide that manufacturers must make available the
required information to all service providers. However, EPA must have specific definitions for
aftermarket providers because they are distinct entities from franchised dealerships and authorized
service networks, and therefore, our regulations need to direct manufacturers specifically regarding
providing information for these entities.
IV.K.6 Liability Concerns
Comments:
EMA commented that EPA must include language in the final rule that would specifically limit
engine manufacturers' liability for use of tools by third parties. As discussed previously, engine
manufacturers have significant concerns with regard to the possibility for misuse that may arise
when the service tools and information required by this rule are provided to non-authorized service
providers. As aftermarket providers are given the tools to service and repair engines, there is a
possibility that inadvertent or deliberate mis-configuring may occur. While allowing manufacturers to
require training when those tools are sold to third parties alleviates that concern in part, independent
service providers may service several different brands of engines and may have less "specialized"
knowledge and be more likely to make mistakes than factory-authorized dealers that focus on
servicing engines from a single manufacturer.
In addition, EMA commented that they have greater concern for the provision that requires
engine manufacturers to make available to all equipment and tool companies all information
necessary to read and format all emission-related data stream information and to activate all
emission-related bi-directional controls. The provision is designed to ensure that independent tool
manufacturers have the information necessary to produce and make available for sale to service
providers diagnostic tools with bidirectional controls. Although bi-directional controls cannot be used
to permanently change an engine calibration, they give a service technician the ability to temporarily
control the engine.
EMA concludes that EPA must include specific language in the regulatory text that confirms
engine manufacturers will not have any emissions warranty, in-use compliance, defect reporting or
recall liability for service on a heavy-duty engine that is not undertaken by the manufacturer, for any
101
-------
damage caused by their own tools in the hands of independent service providers, or for the use and
misuse of third party tools.
Cummins and Volvo commented that EPA should revise the language of the Prohibited Acts,
Liability and Remedies section (86.010-38(j)(19)) of the regulatory language. The current language
refers to personal liability and Cummins and Volvo believe the reference should be to corporate
liability.
Agency Response:
EPA believes it is reasonable to add language to the regulations that would limit an engine
manufacturers' liability for use of tools by third parties and have added language accordingly.
Regarding the Cummins and Volvo comments, section 203 pf the CAA speaks in terms of
personal liability, so EPA believes that this language is appropriate. However, as a point of
clarification, EPA treats corporations as "people". Therefore, corporations can be the entity that is
liable, and we do not believe it is necessary to make any changes to the existing regulatory
language.
IV.K.7 Compliance Flexibility
Comments:
EPA has proposed in paragraph (j)(2) to allow engine manufacturers of heavy-duty engines
subject to the rule to alternatively comply with service information and tool provisions for 1996 and
later vehicles under 14,000 pounds GVWR. EMA supports that provision, as it provides needed
flexibility to manufacturers in cases where the same engines are used in similarly-sized vehicles.
EMA states that EPA also should extend that flexibility to engines in the 8,500 to 14,000-
pound range. The proposed flexibility would permit manufacturers to provide service information and
tools that follow the industry standards and practices that are most familiar to the type of service
providers that will work on the vehicles. There is no dis-benefit to providing this flexibility.
EPA should adopt the following additional compliance flexibility language (as sub-section (ii)
):
(2)(ii) Upon Administrator approval, manufacturers that produce engines
for use in vehicles between 8,500 and 14,000 pounds may, for those engines,
alternatively comply with all service information and tool provisions in CFR
section 86.010-38(j) that are applicable to 2010 and subsequent model year
vehicles over 14,000 pounds. Implementation dates must comply with the service
information provision dates applicable to engines in vehicles between 8,500 and
14,000 pounds.
Agency Response:
EPA agrees that is reasonable to add this additional language to the regulations to allow
more flexibility for manufacturers who produce engines across light-duty, medium-duty, and heavy-
duty classifications.
IV.K.8 The Service Information Requirements Must Comply with the CAA and Federal Law
102
-------
Comments:
EMA commented that Section 202(m) of the federal Clean Air Act (42 U.S.C. 7521)
establishes the framework under which EPA may regulate and require service information availability
in connection with emissions-related diagnosis and repair of engines and vehicles. EPA proposes to
regulate the cost of service information, requiring that manufacturers make information and tools
available "at a fair and reasonable price" and providing for approval of pricing based on a number of
factors delineated by the Agency.
Moreover, EPA must ensure that no provisions of its Proposed Rule lead to or become an
unconstitutional taking of property without due process and compensation. The Proposed Rule
would require manufacturers' proprietary and confidential business information - long recognized as
property that is protected from being taken without adequate compensation - to be turned over to
the aftermarket. To avoid an unconstitutional taking, manufacturers must be able to charge
reasonable prices for that property and not be unreasonably restricted by pricing requirements in the
rule. EPA must ensure that manufacturers can indeed recover reasonable costs for providing service
information to third parties.
ATD commented that the statute is permissive regarding an OBD mandate for heavy-duty
vehicles and engines, as the Administrator may, in the Administrator's discretion, promulgate
regulations requiring manufacturers to install such on-board systems on heavy-duty vehicles and
engines. 42 USC §7521 (m)(1). This arguably renders the information availability section of the
statute inapplicable, as it was intended for the light duty rule prescribed by 42 USC §7521 (m)(1).
42USC§7521(m)(5).
Second, EPA's cursory citations to highly suspect data supposedly demonstrating a growth
trend regarding "independent" repair facilities doing work on heavy-duty vehicles is grossly outdated,
makes no effort to break out engine-related (let alone emissions-related) service work, and thus is
irrelevant to this rulemaking. 72 Fed . Reg . at 3266. Importantly, no effort was made to forecast
where heavy-duty operators likely will have OBD-related emissions work done 5 or so years from
now, assuming a heavy-duty OBD mandate takes effect. Such a forecast should be done, taking
into account applicable emissions warranty periods and characteristics unique to the heavy-duty
service and repair industry, i.e., dealerships, engine distributors, fleet shops, government shops,
truck stops, "independent" shops, rebuilders, etc . Third, and most importantly, EPA does not
suggest that any segment of the heavy-duty service and repair industry does not now or will not in
the future be able to obtain the information necessary to conduct OBD-related emissions service. No
permissive rule should be promulgated unless and until the need for it can be demonstrated and it
can be shown to be cost-effective.
ATD further commented that the proposal contains several unnecessary and inappropriate
provisions given the language and intent of the Clean Air Act's information availability section. 42
USC §7521(m)(5). Importantly, several provisions in the proposal potentially could undermine
accurate and effective emission system repairs and the air quality benefits associated with heavy-
duty OBD.
ATD also commented that many non-dealer service facilities have a competitive advantage
in the heavy-duty marketplace, given their lower overhead and unbridled discretion to purchase and
use whatever parts, tools, information, and equipment they choose. In any event, the Clean Air Act
does not require or allow EPA to:
• Establish a welfare program for certain segments of the heavy-duty service industry.
103
-------
• Subsidize automotive maintenance and repair facilities that are unwilling or unable to make
necessary business investments.
• Govern the relationship between engine manufacturers and anyone other than any person
engaged in the repairing or servicing of motor vehicles or motor vehicle engines (e.g., parts
companies, tool companies, information providers, technician trainers, etc).
• Involve itself in issues of intra-industry competition.
Agency Response:
EPA believes that all of the provisions being finalized for heavy-duty service information
comply with Clean Air Act and federal law. While EPA would differ with EMA characterizations
regarding the issue of confidential business information, EPA does not disagree with the statement
that manufacturers must be permitted to charge reasonable prices and recover reasonable costs for
supplying service information to third parties.
In response to the AID comments, First, EPA believes that that section 200(m)(5) does give
EPA the authority to regulate heavy-duty service information. The language of section 202(m)(5)
does not restrict EPA's authority regarding provision of service information to any subset of motor
vehicles and engines. Indeed, unlike section 202(m)(1), the mandate to EPA applies to repair and
service of motor vehicles and engines, without qualification. While EPA did not believe it was
appropriate to require manufacturers to provide service information for heavy duty engines prior to
institution of OBD requirements for such engines, section 202(m)(5) certainly authorizes, if not
requires, EPA regulations requiring provision of service information now that OBD is required for
such engines. Moreover, the mandate under section 202(m)(5) contains no language with regard to
EPA finding need or cost-effectiveness, indicating that Congress had already made its evaluation of
the need or appropriateness of these regulations and intended these regulations to apply without
relying on any EPA determination of need.
To further respond to the ATD's comments, EPA does not believe that these regulations will
adversely impact or otherwise undermine the effective completion of emissions-related repairs, nor
does EPA believe that any of these provisions create a welfare program for the aftermarket or
otherwise impede industry competition in any manner. These provisions are intended to ensure that
manufacturers do not impose artificial barriers to access to tools and information. In addition, the
provisions being finalized today allow manufacturers reasonable parameters under which to require
aftermarket service providers to make the necessary business investments to properly service and
repair heavy-duty engines.
IV.K.9Tier 1 Suppliers
Comments:
EMA commented that "Third party" suppliers do not play a role in the heavy-duty market.
This is another way in which the light-duty and heavy-duty service industries differ. The heavy-duty
engine manufacturers to be regulated by the Proposed Rule are Tier 1 suppliers for the vehicle
manufacturers. Traditional technology suppliers for engine manufacturers have been fuel system
suppliers, turbocharger suppliers, and control system suppliers. More recently, suppliers of EGR
components and controls, and of aftertreatment components and controls, have joined traditional
technology suppliers. In general, technology suppliers do not provide service information directly to
heavy-duty vehicle owners - their recommendations are included in the engine manufacturer's
service information.
104
-------
AAIA, et.al. commented that EPA has requested comments on whether the requirement that
information related to the service, repair, installation or replacement of parts or systems developed
by Tier 1 suppliers and made available to the dealers, which is a part of the light duty rule, be made
part of the heavy duty rule. The commentersees no reason why it should not. By allowing this
information to be disseminated to the engine manufacturer's dealer or authorized service provider
network, the supplier has already agreed that it could be used to service the manufacturer's products.
Therefore, it should be made available to all facilities who do such service on the same basis.
Agency Response:
In response to EMA comments, EPA believes that the examples provided by EMA in their
comments fit EPA's view of a Tier 1 supplier. EPA considers a Tier 1 supplier to be a party other
than the manufacturer who has developed and/or supplied a system to the manufacturer that is an
integral part of the engine manufacturer's OBD system. Generally this also means that the service
information for those systems is developed by the third party, not the engine manufacturers
themselves and we understand that the engine manufacture has no direct control over this
information.
While the engine manufacturer is not directly responsible for this information, we do believe
it is reasonable that the manufacturers provide some general information on their web sites
regarding these Tier 1 one systems and where aftermarket service providers can go to find the
corresponding service information. Because it appears that manufacturers do in fact install systems
from parties that EPA would consider Tier 1 supplies, we believe it is necessary to retain the
provisions governing access to this information.
The regulations clearly state that service information from third party suppliers are not
required to be made available in full-text on manufacturer web sites. The regulations only require
that manufacturers make available on the manufacturer web sites an index of the relevant
information and instructions on how to order such information. The regulations also allow
manufacturers to create a link from its Web site to the Web site(s) of the third party supplier.
IV.K.10 Adding References to "Authorized Service Network"
Comments:
AAIA, et al commented that anywhere in the rule where there is reference to the
manufacturer's dealers that the reference include the members of a manufacturer's "authorized
service network". Some engine manufacturers use such networks for repairs more than dealerships
and for that reason the "authorized service network" language was included in the California rule.
For clarity, completeness and consistency with the California rule, we would request that reference
to these networks be made in addition to dealerships.
Agency Response:
EPA agrees that it is reasonable to include this suggested language for clarity as well as
consistency with the CARB service information regulations.
IV.K.11 Errors in Section References
AAIA, et al, commented that the proposed rule also contains a number of errors in section
references which should be corrected in the final rule. These are:
105
-------
Section Incorrect Reference Correct Reference
G)(6)(i) 0(3) (twice) (j)(4)
Agency Response:
EPA agrees with these comments and will make the necessary changes for the final rule.
V. What are the Emissions Reductions Associated with the Proposed OBD Requirements?
V.A Emissions Reductions Associated with the 2007HD Highway Rule
We received no comments pertaining to this section that require analysis.
VI. What are the Costs Associated with the Proposed OBD Requirements?
Comments:
One commenter provided the following list of general comments pertaining to the cost
analysis:
In general, the implementation costs do not account for the technological risks in
available aftertreatment and sensor technology, and presume that manufacturers
have a clear technology path to meet the regulation.
Discounting 2004 constant dollar estimates for inflation understates the present value
of the costs to be incurred.
Material costs for key inputs are above the industry's producer price index (PPI)
trend.
Labor rates are unburdened, excluding the costs of benefits from anticipated
compensation for algorithm development and test execution.
Consumption of key resources is understated (labor and economic capital)
HDOBD development is portrayed as a one-time event
o The cost to prepare HDOBD technology and calibrations for the 2016 phase-
in requirements is not estimated.
o The ongoing fixed cost to maintain HDOBD calibrations beyond the 2016
phase-in is not estimated.
HDOBD technologies are assumed to smoothly scale across all diesel engine
displacements.
The declining growth engine sales model understates the number of engines sold.
Conservatively, the combined effect of these factors will double or triple the industry costs for
diesel engines used in vehicles over 14,000 pounds.
106
-------
The commenterthen provides an analysis of EPA's net present value (NPV) discounting
methodology, the use of constant 2004 dollars to estimate future costs, and the use of certain costs
adjusted using the PPI index calculated by the Department of Labor's Bureau of Labor Statistics.
The commenter claims that constant dollars understate the actual cash flow since inflation is not
considered, and that the PPI understates price increases since the underlying commodity prices
have increased at a faster rate than inflation since 2004. A table is provided showing, according to
the commenter, that producer prices have increased anywhere from 22 percent to 123 percent from
December 2003 to December 2006, and that platinum prices have increased over $300 per Troy
ounce since December 2004.
EMA, 2005-0047-0026, Appendix C, p. 12
TMA notes that little of the OBD costs for vehicles over 14,000 Ibs. GVW appear to be
attributed to vehicle manufacturers for the engines they purchase in 2010 and beyond. OBD-related
development costs attributable to having to monitor the performance of new air induction systems,
charge air coolers, cooling systems, exhaust systems, instrumentation, and wiring will be incurred by
vehicle manufacturers to integrate OBD-equipped engines into vehicles. Vehicle manufacturers that
use engines supplied by separate engine manufacturers must expend resources to integrate those
engines into their products. EPA's cost proposal should account for these costs.
Truck Manufacturers Association, 2005-0047-0028, p. 4
Agency Response:
Regarding EMA's comment about technological risks, we believe that the final requirements
have appropriately considered these risks by decreasing the stringency of some of the near term
thresholds. These changes align the thresholds with sensor technology. As for aftertreatment
technology risks, we believe those risks have been proven small given that DPFs and SCR systems
are already being sold into the market place both domestically and in Europe and Japan. In any
case, those systems will be used to meet the underlying exhaust emission standards for these
engines, not to meet the OBD requirements in this rule.
Regarding discounting 2004 dollars for inflation, we do not understand this comment. We
chose 2004 dollars based solely on the timing of the original analysis. The discounting simply
accounts for the fact that a dollar tomorrow, or next year, is worth less than that dollar is worth today.
If the comment was meant to suggest that we should include inflation adjustments for future dollars,
we disagree. The attempt is to show what the rule costs in present value terms. Adjusting for
inflation would confuse that effort. The commenter also argues that development costs for 2016
have not been included. We do not understand what the commenter refers to given that our
certification requirements do not change in 2016, although some in-use flexibilities do. We do not
expect manufacturers to develop systems with those in-use flexibilities in mind so do not believe that
any increased development costs would be incurred. Perhaps the commenter speaks of the CARB
HDOBD requirements which do, in fact, have increased stringency for DPF monitors in 2016 on
some engines. This increased stringency is not part of our proposed or final regulation. As regards
the PPI adjustments used in our draft cost analysis, we agree that using PPI adjustments in this way
may or may not be a good way to adjust costs from one year to another. We have begun to
consider that perhaps the Consumer Price Index (CPI), which is the overall inflation adjuster, is a
better adjuster to use. That said, we only used the PPI adjustment in the context of the HD
2007/2010 program costs to adjust those costs from 1999 dollars to 2004 dollars. The adjustment
had no impact on the costs estimated for HDOBD. For the final cost analysis, we have adjusted all
costs developed for the draft analysis from 2004 dollars to 2007 dollars using the Consumer Price
Index (CPI). As for platinum prices, any rulemaking we do that includes use of platinum group
metals runs up against the issue of what cost/ounce is most appropriate - a recent spot price, a 10
107
-------
year moving average, etc. - and we know of no "best" metric to use. In the end, the price of
platinum is of little concern for HDOBD. Platinum costs would impact only the costs associated with
limit parts in the context of HDOBD. Platinum represents roughly one-third of our DPF limit-part
costs which represent less than one-half of the total estimated hardware costs for limit parts. In turn,
the hardware costs for limit parts represent less than one percent of the estimated total costs. In the
end, platinum constitutes less than 0.3% of our estimated costs. Therefore, the cost/ounce used for
platinum has little impact on the resultant cost estimate.
TMA argues that we did not consider costs they would incur to integrate OBD systems into
their vehicles. However, we do not believe that any new integration would occur. OBD systems are
already used on heavy-duty engines and are being integrated into vehicle designs today. The real
development work - that being for monitors that go beyond their current level of sophistication in
order to meet our requirements - will be done by engine manufacturers and/or their suppliers.
VI.A Variable Costs for Engines Used in Vehicles Over 14,000 Pounds
Comments:
The EPA cost estimates assume that there is no cost increase for sensors to improve their
ability to measure or for changes to aftertreatment systems to improve their ability to be measured.
Industry analysis hypothesizes that sensors may cost more than originally projected to provide the
level of discrimination suggested by the proposed thresholds.
Warranty rates are likely to increase in 2013, not decrease as EPA suggests due to the
increased number of extrapolated ratings in 2013. Those engines complying in 2010 may
experience a decrease in warranty rates by 2013, due to experience and learning, but that
experience may not transfer over to the newly extrapolated engines in 2013.
HDOBD requirements accelerate certain onboard computer related costs by displacing
existing computers for which the engine manufacturer continues to experience costs due to the need
to ensure that those old computers are supported throughout their life.
EPA's cost analysis assumes a certain number of future engine sales with those sales
growing at a two percent rate in early years but only a one percent rate in later years. However,
freight tonnage hauled by motor truck is expected to increase at the same rate as the economy,
which would support a constant growth model. The lower number of engines sold understates the
program's costs.
EMA, 2005-0047-0026, p. 14-15
Agency Response:
We believe that the sensor costs we have estimated are still appropriate and have heard
nothing from sensor suppliers to suggest otherwise. We know of no reason why aftertreatment
devices will have to be changed to improve their ability to be measured, especially given the
changes we have made in the final rule relative to the proposed rule (e.g., elimination of NMHC
catalyst monitoring against an emission threshold). We have revised the warranty related costs as
suggested by EMA given that many engines will be newly adding OBD in 2013. Rather than
warranty costs decreasing in 2013, we have now delayed that decrease until 2016. This has little
impact on the cost of the rule. We agree with EMA that computer costs will increase and have
accounted for that increase in our final cost estimate as we did in our draft cost analysis. As for
108
-------
sales projections, we have used a method consistent with our emissions modeling approach which is
the most appropriate way for us to project sales.
VLB Fixed Costs for Engines Used in Vehicles Over 14,000 Pounds
Comments:
The R&D Fixed Costs presented in Table 3 of the Technical Support Document (EPA420-D-
06-006) assume that "once an algorithm has been developed, it can ... be used over and over again
with only minor changes." This statement assumes that the technology available scales equally well
across all engine displacements (from 5 to 16 liters) covered by the rule. This assumption may be
acceptable below 14,000 Ibs GVWR, but is not believed to be valid above 14,000 GVWR, where
differences in fuel systems and aftertreatment technologies across product lines, will require that
algorithms for specific monitors will need to be developed according to the technology used by the
particular engine displacements.
There will be additional effort above the average assumptions (e.g. 30 weeks for a threshold
monitor) to develop generalized (or scaleable) algorithms across the full product line for diesel
engines. The technology to be used for HD Gasoline engines has a much longer history. There is
significantly more experience in developing threshold, functional, and rationality monitors for3-way
catalyst technology. The technology development effort for less mature HD diesel engine
technologies (reference Table 3, p. 18) is assigned the same 30-week level of effort as more mature
HD Gasoline technologies (reference Table 4, p. 21). Since algorithm development for HD diesel
engines entails more risk, a higher average effort should be applied for diesel engines.
The assumption that only 13 threshold monitors, 20 functional monitors and 15 rationality
monitors need to be developed per manufacturer ignores the additional effort that arises from
multiple technologies. Where it is not practical, due to differences in technology, separate 30- week
efforts will be needed to develop the additional threshold monitor algorithms. IRAFs and differences
between the emissions DF rating and the OBD rating can also significantly increase test cell time
needed to develop threshold monitors.
It is assumed that the technician and engineering labor rates do not include overhead factors
for health insurance, workman's compensation, and other similar unit costs of labor. The
unburdened unit costs and labor rates understate the development costs by omitting benefit costs
typically accorded to professional employees and technicians.
There are no costs in Table 5 of the Technical Support Document for 2013 and beyond. This
is carried forward to Table 9 (page 26) which shows no ongoing development costs in Table 9
beyond 2013. While it may not be required to demonstrate HD OBD performance each year, it is
required to demonstrate HD OBD performance when improvements are made to individual
components or subsystems subject to threshold, functional, or rationality monitoring. These fixed
costs will need to be recovered in the prices received for engines, and are not included in the fixed
cost per engine estimates.
Accordingly, test cell time will be needed from 2013 to 2016 and beyond. In 2010 new NOx
control technology will be installed on HD engines and vehicles. Lessons learned from 2010- 2014
will be applied to future engines in 2016 as refinements to the technology and control systems. Also
in 2016, manufacturers are faced with full in-use liability for all engine ratings across all engine
families. Thresholds will need to be confirmed and compared with the extrapolated projections to
establish sufficient confidence that HD OBD algorithms did scale as intended. From 2007-2013, test
109
-------
cell demand is understated due to the factors given for pp. 16-10 regarding the number of weeks
needed to develop algorithms, the number of algorithms to be developed, and the risks in adapting
algorithms across all product lines.
$700 per hour is a competitive estimate for an emissions test cell. The $100 per hour test cell
rate for HD OBD development is unrealistic. The weighed average cost (as driven by the 30% and
10% utilization factors) is dramatically understated by this figure. Either the $100 test cell will need to
be upgraded to measure constituents accurately, resulting in an increased cost per hour, or
additional emissions test cell capacity will need to be used. A $100 per hour test cell will include a
low precision dynamometer such as a water brake. It will not include constituent measurement
capability that is needed for threshold monitoring design and development. Measurement capability
requires capital equipment investment to achieve. The capital investment needed to make
"development quality" measurements will make test cell costs approach the $700 per hour figure.
The pricing methodology used for certification limit parts (twice the anticipated production
costs) is at the lower end of the range. Prototype parts often cost 3 or 4 times the anticipated
production costs. Prototype part prices sometimes do not include costs that are instead amortized
over the duration of the production contract. Individual piece prices for HD diesel engine parts are
generally higher, due to the lower volume of HD diesel engines, when compared to LD gasoline
engines.
Aging parts on engine dynamometers past their full useful life to the point where they can be
used as threshold demonstration parts creates an order of magnitude multiplicative factor on the
cost of threshold parts. Fuel and test cell costs for aging parts dwarf their initial purchase costs. The
cost of fuel alone to age parts to a 435,000 mile useful life will be
435,000 miles / 5 MPG * $ 3 / gallon = $ 261,000
Assuming that 750,000 miles is a reasonable MTTF design goal for a 435,000 mile useful life
requirement, the cost of fuel alone to age parts to a MTTF of 750,000 miles would be
750,000 miles / 5 MPG * $ 3 / gallon = $ 450,000
Test cell time at $100 / hour would cost
435,000 miles/30 MPH*$ 100 / hour =$ 1,400,000
750,000 miles / 30 MPH * $ 100 / hour = $ 2,500,000
Sharing these costs over an assumed inventory of 20 threshold parts for the 2010 HD OBD
parent engine adds $ 70,000 to $ 150,000 to the value of the each part.
Assuming that monitor verification engines can be obtained for negligible costs may not be
appropriate. Because it should not be assumed that manufacturers can recover the full price of a
new engine. Assuming that 50 % of the engine price can be recovered is more reasonable.
Depending on the amount of use, i.e. the amount of disassembly and reassembly activity to install
monitor demonstration parts to mechanically implant failures that a PVE test engine receives as a
part of'Monitor Verification', it may not be proper to then sell the engine as a "new" engine.
Negligible costs also presume that there are no shipping costs to ship the engine to the test facility,
where the test engineer and/or technician reside.
EMA, 2005-0047-0026, p. 15-19
110
-------
EPA's cost estimate for an "OBD threshold" part, estimated at two times the cost of a new
production part, is low. Moreover, EPA estimated only two threshold parts. By contrast, some
manufacturers estimate at least seven parts would be needed to arrive at a correct threshold part.
EPA should revise its estimate to be more in line with actual costs. The commenter provides an
appendix to its comments which discusses the EPA cost estimates and its flaws. In that, the
commenter suggests that commodity prices such as fuel, steel, platinum, have increases 20 to 123
percent in the time between December 2003 and December 2006 (see Appendix C, page 13 of
submitted comments). This is used, along with an estimate of the cost of fuel burned while engine
bench aging aftertreatment devices, as part of the basis for threshold, or limit, parts costing far more
than estimated by EPA.
EMA, 2005-0047-0026, p. 34-35
Agency Response:
Regarding the R&D costs and our assumption that the OBD technology scales well across all
engine sizes, we have attempted to account for costs associated with developing the basic approach
as an "Algorithm Development Cost" and then the costs to apply that to various engine families (i.e.,
engine sizes) via the "Application Costs to Each Family." As such, we believe we have properly
considered the costs for applying technology to different engine sizes.
Regarding the comment that we should increase the time required for diesel algorithm
development based on our equal estimate for both gasoline and diesel combined with the fact that
gasoline OBD is much more developed, we consider this to be an argument for decreasing the
gasoline time rather than increasing the diesel time. We have chosen not to undergo that change for
our final analysis since leaving the gasoline timeframe as estimated in our draft analysis represents
a worst case estimate.
Regarding our estimate that development of threshold monitors ignores multiple technologies,
we believe the comment speaks to manufacturers that may offer both a NOx adsorber and a SCR
approach to NOx control. While the OBD algorithm development for these two technologies may
differ, it is by a given manufacturer's choice to offer these two technologies since, for example, SCR
could be used for all HD applications. Importantly, we believe that the monitoring approaches
should transfer well between NOx adsorber and SCR systems, with the exception of the urea-
specific monitoring requirements that exist for most SCR systems. As regards labor rates, we have
used labor rates consistent with other recent HD rules. We believe there is value in being consistent
with those rules where possible and have not changed anything for the final analysis.
As for fixed costs beyond 2013, we expect the OBD systems developed to comply with this
rule to be sufficient for future applications as well. Should manufacturers choose to fundamentally
redesign their OBD systems, they will generally choose to do so only to reduce costs. We have not
estimated a cost or savings for such efforts as part of this rule.
Regarding test cell time, we do not agree that increased test cell time will be required due to
the increased level of in-use liability in the 2016 model year. This comment suggests that
manufacturers may be using the in-use liability provisions as a design target which is certainly not
the intent behind them. OBD monitors should be developed to the certification OBD thresholds
regardless of the in-use liability provisions and, as such, the major development work should be
complete by 2013.
Regarding test cell costs/hour, the commenter agrees that $700 per hour is a good estimate
for a certification cell but that $100 per hour for development is not realistic. The commenter then
argues that upgrades will be needed to measure constituents. We consider the $100 cell to be a cell
111
-------
that has no emissions measurement capability, at least not anywhere near the regulatory
requirements. The cell we envisioned was simply a dyno with proper ventilation and computer and
instrumentation capabilities and perhaps ppm emission measurement capability. We believe that
most of the algorithm development will entail testing of this sort and very little development testing
will entail full certification testing (i.e., emissions per unit work emissions measurement according to
40 CFR Part 1065 will be done only when algorithm development is nearly complete).
As regards limits parts, we believe that our estimate of double the production part price is a
good estimate. In general, these costs, in constant dollar terms, will probably decrease rather than
increase in the future.
As regards the aging of limit parts, we did not properly characterize this cost in our draft
analysis. Using the approach suggested by EMA, and using our final aging approach which requires
aging to represent full useful life (on average, we have estimated this to be roughly 80 percent of
useful life as manufacturers will be able to show, we expect, that this is representative of full useful
life for OBD), and using a sales weighted useful life of 335,000 miles and sales weighted MPG of 7,
and adjusting 2004 dollars to 2007 dollars using the Consumer Price Index (1.10), the fuel costs
would be:
(335,000 miles x 80%)/7 MPG x $3/gallon = $115,000 per parent engine
Test cell time would cost:
(335,000 miles x 80%)/30 MPH x $100/hrx 1.10 = $981,000 per parent engine.8
The total being $1.1 million per parent engine. We have added these costs to the costs of
limit parts in the final analysis. In the final cost analysis, limit part costs consist of the limit part
hardware and the limit part aging. Note that we consider this to be an overestimation of the costs to
age parts since we believe manufacturers will choose to conduct some sort of rapid aging of parts
via exposure to very high temperatures or very extreme conditions. Such aging would more quickly
degrade the parts for OBD use than simply bench aging for a useful life's worth of operating hours.
As regards costs for production engine testing, we believe we have properly characterized
these costs. These costs are a small portion of the cost of the program.
As regards the comment that seven limit parts would be required per engine per monitor
rather than our estimated two, we disagree, especially in light of our removal of IRAFs from the
threshold determination. Removing IRAFs from threshold determination should make limit part
generation and limit part aging a much less complex process.
VI.C Total Costs for Engines Used in Vehicles Over 14,000 Pounds
Comments:
The ongoing fixed costs per engine decline to near zero in Table 13 of the Technical Support
Document (EPA420-06-006). This suggests that in the future, on-going benefits are "free," because
the costs have been fully amortized. In fact, the ongoing costs will continue to be borne by
manufacturers as they design new components to meet on-going customer demands for
transportation capital goods.
The CPI for 2007 was reported as 207 while that for 2004 was 189 (the 1982 to 1984 timeframe=100, see
www.bls.gov/cpi/). Therefore the CPI adjustment for 2004 dollars to 2007 dollars is 207/189=1.10.
112
-------
EMA, 2005-0047-0026, p. 19
EPA's estimate of $50 per diesel engine is rather suspect considering the current state of
sensor development and the likely need for a separate platform design being required. If OBD has a
cost impact that far exceeds $50, EPA should be aware of the impacts this additional financial
burden will have on accelerating the anticipated pre-buy and post low-buy of engines surrounding
the 2010 model year.
American Trucking Association, 2005-0047-0029, p. 5
Agency Response:
Regarding the comment that fixed costs will be ongoing, please refer to our response under
section VLB. Regarding HDOBD costs and impacts on pre-buy, we do not believe that OBD
systems will result in any significant pre-buy or post low-buy of engines. If anything, we believe that
OBD equipped engines will be considered more valuable rather than less valuable as this comment
suggests.
VI.D Costs for Diesel Heavy-duty Vehicles and Engines Used in Heavy-duty Vehicles Under
14,000 Pounds
Comments:
Concerns for diesel costs under 14,000 Ibs. GVWare similar to those discussed above for
vehicles over 14,000 Ibs. GVW.
EMA, 2005-0047-0026, Appendix C, p. 20
Agency Response:
We have changed the final analysis for under 14,000 pound costs to be consistent with
changes made for the over 14,000 pound costs. Please refer to our responses under sections VI.A
through VI.C.
VII. What are the Updated Annual Costs and Costs per Ton Associated with the 2007/2010
Heavy-duty Highway Program?
VILA Updated 2007 Heavy-Duty Highway Rule Costs Including OBD
We received no comments pertaining to this section that require analysis.
VII.B Updated 2007 Heavy-Duty Highway Rule Costs per Ton Including OBD
We received no comments pertaining to this section that require analysis.
113
-------
VIM. What are the Requirements for Engine Manufacturers?
VIM.A Documentation Requirements
We received no comments pertaining to this section that require analysis.
VMI.B Catalyst Aging Procedures
Comments:
EMA supports use of "de-greened" (i.e., aged to 125 hours) engines for OBD demonstration
testing, but does not support any requirement that aftertreatment devices be aged to their full useful
life for such testing. Aging of aftertreatment devices to 125 hours provides sufficient assurance that
the components will be representative of emissions in-use. Aging these devices to full useful life
(e.g., 435,000 miles for heavy heavy-duty engines), is costly, time-consuming, and would impose
undue burdens on manufacturers. Rather, appropriate deterioration factors would be applied to the
baseline engine/aftertreatment system to achieve a representative system appropriate for
certification demonstration of the OBD system. The deterioration factors determined for emissions
certification should be allowed for OBD certification.
EMA, 2005-0047-0026, p. 34
Cummins Inc., 2005-0047-0031, p. 5
National Automobile Dealers Association, 2005-0047-0034, p. 3
Agency Response:
As of this writing, we are developing new emission certification durability procedures that will
require manufacturers to demonstrate emission compliance with systems aged beyond the 125 hour
aging used for years in engine compliance demonstration. Because that issue remains unresolved,
we have not changed our final HDOBD regulation. Therefore, manufacturers will be required to use,
for OBD compliance demonstration, engines aged fora minimum of 125 hours and aftertreatment
devices aged to represent full useful life aging (see §86.010-18(l)(2)(iv)).
VIII.C Demonstration Testing
Comments:
It is essential that EPA and CARB coordinate carefully to ensure that engine manufacturers
are not required to undertake duplicate testing or expend resources where test results can be shared.
EMA, 2005-0047-0026, p. 32
Agency Response:
We agree with this comment and this is our intention.
VIII.C.1 Selection of Test Engines
114
-------
Comments:
EPA's definition of the "parent" engine rating differs from CARB's definition in that the CARB
definition is based on California sales while EPA's is based on U.S. sales. In some cases, this may
result in different engines being chosen as the parent and, thus, double testing. Such an outcome
should be avoided and the regulation should be revised to ensure that such an outcome will not
occur. In fact, some manufacturers have already chosen their parent rating based on the CARB
definition.
EMA, 2005-0047-0026, p. 33
Agency Response:
Both agencies have a motivation of having parent ratings be those having high sales in their
area of interest. For the most part, we believe that will be the case by default. However, it is not
guaranteed. For the final rule, we have inserted a provision that allows the Administrator to accept
alternative engine ratings (see final §86.010-18(o)(1)(i) and compare to proposed §86.010-18(o)(1)).
VIII.C.2 Required Testing
Comments:
For certification demonstration to emissions standards, one certification cycle is required
while, for demonstration to OBD requirements as proposed, 7 to 10 and perhaps 14 to 20 such
cycles would be required. This testing places a huge workload burden on manufacturers. While
some level of OBD certification demonstration testing is reasonable, manufacturers must be allowed
to carryover engine testing requirements from year to year. It appears from EPA's preamble, that
EPA could be considering requiring new parent engines and, thus, new certification demonstration
testing every year beginning in 2016. This is unreasonable. There is little, if any, added benefit in
requiring OBD system certification demonstration on more and more engines each and every year.
Moreover, production evaluation testing already requires ongoing testing after engines are produced.
EMA supports EPA's proposed rule where it eliminates the requirement to conduct testing in 2011-
2012 and 2013-2015 and the ability to carryover engines from yearto year. EPA should eliminate
any requirement that new parent ratings could be determined from year to year beginning in 2016
and for those "new" engines to be fully tested for certification demonstration. There should be
language that would allow the Administrator to reduce the number of certification demonstration
tests beginning in 2016 and later.
EPA must clarify in the final rule that double-testing (for CARB and for EPA) will not occur,
and that EPA will coordinate the selection of test engines with CARB to ensure that double-testing
will not occur. There must be a reasonable cap on the number of engines required for testing by
manufacturers certifying under both EPA and CARB rules to ensure no double-testing.
EPA should work with CARB to align CARB's certification demonstration requirements with
those that EPA has proposed, particularly with respect to eliminating additional testing in 2011-2012
and 2014-2015.
EMA, 2005-0047-0026, p. 32-33
Currently, emissions standards are demonstrated using an emissions deterioration factor
(DF) test engine aged to its useful life. That DF test engine is chosen using a different set of criteria
(i.e., worst case emissions) than that for which the OBD parent rating is chosen (i.e., highest sales
weighted by useful life mileage). For 2010-2015, EPA should reduce the anticipated substantial
115
-------
impacts on engine manufacturers' test cell and development burden by allowing manufacturers to
use the OBD engine rating for emissions DF demonstration.
EMA, 2005-0047-0026, p. 35
Agency Response:
Regarding comments about carry-over and/or requiring new OBD certification demonstration
every year, it is important to note that we intend to implement the OBD requirements in a manner
consistent with our implementation practices for emission standards. In other words, when carry-
over of data is acceptable we intend to allow manufacturers to carry-over that data. However, if we
consider there to be some reason that carry-over is not acceptable, we intend to require new
certification demonstration. We do not intend to require new OBD demonstrations when such
demonstration is identical to a previous year. No one would benefit from such a practice.
Regarding double testing, we agree with this comment and do not intend to require any
double testing. We have added a provision to our regulation that allows the Administrator to accept
other ratings than the rating otherwise required by our regulation (see §§86.010-18(o)(1)(i) and
86.010-18(o)(2)(ii)(B)).
Regarding the comment on OBD demonstration and DFs, please refer to our response under
section VIII.B.
VIII.C.3 Testing Protocol
We received no comments pertaining to this section that require analysis.
VIII.C.4 Evaluation Protocol
We received no comments pertaining to this section that require analysis.
VIII.C.5 Confirmatory Testing
Comments:
EPA and CARB must assure that manufacturers are not subject to double-testing to satisfy
the OBD confirmatory testing requirements of both agencies. The agencies must confirm that they
would not conduct confirmatory testing in the same year. Even a temporary loss of a test engine for
confirmatory testing that would otherwise be used for internal development imposes a cost burden
on the manufacturer.
EMA, 2005-0047-0026, p. 36
Agency Response:
Regarding double testing, we agree with this comment and do not intend to require any
double testing. We have added a provision to our regulation that allows the Administrator to accept
other ratings than the rating otherwise required by our regulation (see §§86.010-18(o)(1)(i) and
86.010-18(o)(2)(ii)(B)).
116
-------
VIII.D Deficiencies
Comments:
EMA supports the proposal to adopt deficiency provisions which have proven to be a
necessary component of OBD regulations given their complexity. CARB granted a number of
deficiencies for medium-duty OBDII (13 CCR 1968.2) DPF threshold monitors which have a
threshold of the standard+0.09 for the 2007 model year. While deficiencies should not and cannot
be used as a substitute for a thorough and appropriate analysis of technological feasibility, it is
important that deficiency provisions be available in the rule.
Production evaluation testing (see proposed §86.010-180) would greatly expand
manufacturers' testing requirements to ensure that software errors and other production glitches are
discovered and corrected early on rather than years later. The ability to grant a post-production
deficiency (i.e., a "retroactive" deficiency) for less egregious issues discovered by manufactuers
during production evaluation testing would provide the ability to correct issues in a manageable time
frame with less disruption to existing resources. Such is the case with the CARB OBDII and HDOBD
programs which specifically allow for "retroactive" deficiency determinations. With EPA proposing to
adopt the same production evaluation testing requirements, it is equally important that EPA not
initiate enforcement actions against a manufacturer for similar, "less egregious" issues. While EMA
understands EPA's desire to address in-use enforcement issues via defect reporting, where CARB
grants a retroactive deficiency it is essential that EPA use reasonable discretion to not initiate
enforcement actions. The CARB regulation only allows retroactive deficiencies for less egregious
issues with the more egregious issues falling under their enforcement regulation.
EMA, 2005-0047-0026, p. 40
Agency Response:
We agree with this comment in that it provides some level of incentive to manufacturers to
find and correct potential errors in their OBD systems early in their life-cycle. It also provides
incentive to bring such issue to our attention rather than hoping we do not become aware of them.
Therefore, we have added a provision to our production evaluation testing requirements that would
allow, on a case-by-case basis, what could effectively be called a "retroactive" deficiency (see final
rule §86.010-180)).
VIII.E Production Evaluation Testing
Comments:
Although still having concerns with the production evaluation testing requirements, EMA
raises no further objections to these requirements provided that EPA and CARB assure that
manufacturers will not be required to conduct double-testing under the California and the federal
HDOBD requirements. In that regard, EPA should revise the proposal to clarify that EPA will not
require "double-testing" by engine manufacturers meeting both the California and federal OBD
requirements.
EMA, 2005-0047-0026, p. 35-36
Cummins Inc. 2005-0047-0031, p. 8
Agency Response:
117
-------
Our intent is to avoid the double testing possibility mentioned by these commenters. We
intend to work closely with CARB staff in an effort to ensure that it does not occur. We do not believe
it is necessary to state this in our regulatory text.
VIII.E.1 Verification of Standardization Requirements
Comments:
The text envisions that data link standardization requirements will be conducted at vehicle
assembly plants. Since HD vehicle assembly plants are not co-located with HD engine assembly
plants, the amount and extent of testing that can reasonably and routinely be conducted at vehicle
assembly plants is limited. In general, tests should demonstrate the integration of data link
communications, but cannot reasonably exercise specific threshold monitors or performance monitor
ratios.
TMA members believe it is unnecessary to test all possible details of data link
communications at the physical and data link layers, as many of these services are provided by
integrated circuits marketed to the automotive engine, vehicle, and tool industries. The performance
of these devices is tested by independent agencies as a part of the license requirements for CAN as
defined by ISO 11898. Key aspects of other communication capabilities can be readily incorporated
into tests for specific services. For example, transport protocol capability is readily demonstrated by
requesting data that require multiple CAN frames to accomplish. For example, both freeze frame and
VIN require the sender to appropriately support transport protocol.
TMA members support the limited numbers of test vehicles proposed and will collaborate
with engine manufacturers to achieve the desired test volumes. Members agree that industry
standard software will be developed that is similar to SAE J1699-3 in scope, but will be designed to
suit the HD OBD provisions defined in SAE J1939-73 (September 2006). A proposed standard draft
has been developed that will meet the needs of 13 CCR 1968.1, and is being revised to satisfy 13
CCR 1968.2 (2006) and 13 CCR 1971.1 (2006).
Truck Manufacturers Association, 2005-0047-0028, p. 4-5
Agency Response:
The first concern expressed by TMA goes beyond the scope of the proposed requirement.
In fact, the requirement to verify standardization requirements is to demonstrate the integration of
data link communications, not to exercise specific threshold monitors or performance monitor ratios.
This is what is suggested by TMA. As regards the second concern, testing all possible details of
data link communications, we believe that the requirements we proposed and are finalizing are
appropriate especially given the limited number of vehicles required to be tested. The level of
communications verification is very important since without proper communications the OBD system
is of little value.
VIII.E.2 Verification of Monitoring Requirements
We received no comments pertaining to this section that require analysis.
VII I.E.3 Verification of In-use Monitoring Performance Ratios
We received no comments pertaining to this section that require analysis.
118
-------
IX. What are the Issues Concerning Inspection and Maintenance Programs?
IX.A Current Heavy-duty I/M Programs
We received no comments pertaining to this section that require analysis.
IX.B Challenges for Heavy-duty I/M
Comments:
The USEPA should make information available to the States regarding the OBD systems
and components being used and monitored by the manufacturers, or essentially the "OBD profile",
for each engine configuration. As proposed, the manufacturers will be allowed to tailor the OBD
system to each engine configuration and that will include monitoring exemptions/alterations, creating
a myriad of different OBD configurations. In order to perform the OBD scan correctly, the states
need to know what monitors are included in each system to compare those configurations to the
information being captured with the OBD scan during the inspection. Further, several components
are proposed to have the capability to be switched off under certain conditions. The OBD profile
would identify those components and their normal on/off state in a specific configuration.
Under the current light-duty OBDII system, several manufacturers were granted
exceptions/exemptions to OBD monitoring by USEPA and the states were not notified of these
exceptions leading to problems after states implemented their light-duty I/M programs. It has been
difficult for the states to obtain the information regarding the OBD monitoring exceptions/exemptions.
Access to this information must be made readily available to the states that implement OBD I/M.
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 6
Agency Response:
We appreciate the time and effort put into providing these comments and will consider them
should we move forward with developing an I/M proposal based on HDOBD checks.
IX.C Heavy-duty OBD and I/M
Comments:
This rulemaking will provide an option for states to consider in meeting their SIP
requirements. However, the current Mobile6.2 model and the new mobile model under development,
Motor Vehicles Emissions Simulator (MOVES), do not provide any credit for conducting diesel
testing. The mobile model should be designed so that states are able to take SIP credits for
implementing I/M programs fordiesels (light-duty and heavy-duty) and heavy-duty gasoline.
Texas Commission on Environmental Quality, 2005-0047-0020, p. 1
Maryland Department of the Environment, 2005-0047-0022, p. 1
NY State DEC, 2005-0047-0033, p. 2
Unfortunately, there is a significant problem in this proposal, as acknowledged by EPA. The
current version of EPA's MOBILE model assumes zero deterioration of emissions for most heavy-
duty diesel engines over their lifetime. In order to appropriately account for emissions from this
119
-------
sector in their State Implementation Plans and assess the cost-effectiveness of heavy-duty
inspection and maintenance (I/M) programs, it is critical that EPA update the MOBILE model to
reflect the technology changes introduced by the new standards and their impact on emissions from
affected engines and equipment. We therefore urge EPA to expeditiously develop the necessary
technical tools and policy guidance to enable states to determine the deterioration offset benefit from
an OBD program.
NESCAUM, 2005-0047-0024, p. 2
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 3-4
EPA is encouraged to move forward with a HDOBD-based I/M program. EPA should make
HDOBD as similar to the light-duty side as possible in order to minimize hardware/operating system
adjustments and upgrades needed to include HDOBD testing in current I/M programs. EPA is
strongly encouraged to ensure standard protocols, connectors, data formats, and technical
specifications to the maximum extent possible. EPA is also strongly encouraged to provide guidance
for I/M states that identifies areas where standardization has been achieved, and where differences
exist.
Maryland Department of the Environment, 2005-0047-0022, p. 1
National Automobile Dealers Association, 2005-0047-0034, p. 3-4
We also take note of the suggestion in the proposal that HDOBD I/M programs may be fleet
or corporate-based, rather than following the traditional state models used for light-duty OBD I/M
programs. Individual states must have the ability to design I/M programs that best fit their particular
circumstance. Consequently, EPA should provide the appropriate technical and policy resources to
accommodate these diverse needs. We therefore urge EPA to work with program personnel in the
state agencies to develop model I/M program guidance which, among other things, addresses the
emissions benefits achievable under I/M programs of varying configurations. EPA should establish
emission control warranty requirements tied to possible HDOBD I/M programs.
NESCAUM, 2005-0047-0024, p. 2
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 3
NY State DEC, 2005-0047-0033, p. 3
Since HD diesel vehicles have two data link connectors (DLCs), will I/M states need to
access both DLCs for testing purposes, or are the emissions-related data all accessible through one
OBD connector? States need significant assurance that when the MIL is illuminated it is only for
emissions related failures. Will HD gasoline and HD diesel vehicles use the same OBD test
procedures to obtain test data from their OBD systems? Given the growing application of wireless
technologies to OBD testing and information gathering on the light-duty side, will such technologies
be compatible with HDOBD systems?
Maryland Department of the Environment, 2005-0047-0022, p. 1
EPA should investigate, in a future rule, what it would take to completely specify a HDOBD
based I/M program. We recommend that if a HDOBD I/M check is determined to be needed, then
industry should be allowed to satisfy the requirements by using a process that is already being run
today, such as a periodic preventive maintenance event or a Department of Transportation
inspection event.
Cummins Inc., 2005-0047-0031, p. 7
120
-------
There should be a simple means for state agency personnel, service technicians, and
owners (particularly second and later generation owners) to determine, by examining a serial
number placed on the chassis (from the VIN itself or a separate label provided by the engine
manufacturer) displaying, what engine has been installed, if the installed engine is OBD-equipped,
and if so equipped, what type of OBD system is present. The OBD system also should be readily
identifiable through a scan tool reading. Recognizing that different engine configurations are likely to
have differing OBD monitors, EPA should require engine manufacturers to make their OBD profiles
available to state enforcement agencies as a means to verify that all monitors are reporting data.
NESCAUM, 2005-0047-0024, p. 3
ATA views the diagnostic data that identifies the malfunction and is stored in the engine's
computer as a fleet's proprietary information. The integrity of the OBD regulatory framework and
vehicle inspections can be compromised if Mils light up when there is no failure of an emission
systems or component. Such information should only be used by fleets to indicate to them
malfunctions of the engine emission control systems. MIL readings should not be used by entities
as a basis of commencing enforcement actions or pursuit of emission violations.
American Trucking Association, 2005-0047-0029, p. 3
The California Air Resources Board (CARB) used their emission model (EMFAC) to estimate
the emission benefits for future model year vehicles (2010 and subsequent model year), and
determined that there could be substantial emission benefits from expanding their current HD diesel
I/M program to add an element to ensure that the new generation of ultra-clean HD diesels
maintains their low emissions long after being placed in-service. This data supports the growing
body of information showing that in-use emissions control performance of HD diesels are now
subject to the same factors as light-duty vehicles, if not more so. Without the ability to show emission
reduction credits for implementing a HDOBD I/M program, the states may have difficulty justifying
implementing such a program. Within the EMFAC model for the heavy-duty fleet, tables exist that
allow the user to input various emission component malfunction rates and the associated emission
rates with each of those component malfunctions. CARB modified several of the existing
components to better reflect the technology that is expected to be used on 2010 and subsequent
engines. Specifically, CARB added malfunction categories for particulate matter (PM) filter leaks,
missing/tampered PM filters, NOx aftertreatment system malfunctions, and NOx aftertreatment
control sensor malfunctions. CARB has estimated that when these aftertreatment devices
malfunction, emissions can rise dramatically. For example, for PM exhaust filter leaks and PM filter
missing/tampered, CARB estimates PM increases of 600 percent and 1000 percent, respectively
and a loss of feedback control (either a NOx sensor for Selective Catalytic Reduction or an air/fuel
ratio sensor for an NOx adsorber) would result in significantly lower NOx conversion rates for an
emission increase of 200 percent (to a tailpipe emission level of 0.6 g/bhp-hr NOx) (see California
Air Resources Board, Staff Report: Initial Statement of Reasons for Proposed Rulemaking,
Malfunction and Diagnostic System Requirements for 2010 and Subsequent Model Year Heavy-Duty
Engines (HD OBD), June 3, 2005).
New Jersey Department of Environmental Protection, 2005-0047-0027, p. 4
Owners of trucks under rental and lease agreements do not control the daily operation of
their vehicles. Driver response to an illuminated MIL is the responsibility of the carrier operating the
leased or rented truck. TRALA opposes any attempt to place enforcement penalties or liability on
the owners of rented or leased vehicles resulting from a driver's failure to respond to an illuminated
MIL.
Truck Renting and Leasing Association, 2005-0047-0032, p. 2
121
-------
TRALA members request EPA's consideration of an OBD program that uses information
gained from regular periodic maintenance performed by our own maintenance technicians to insure
that integrity of emissions control systems. Existing inspection procedures could serve as a check
on the success of these maintenance practices.
Truck Renting and Leasing Association, 2005-0047-0032, p. 3
Agency Response:
Several commenters supported using heavy-duty OBD as a means for testing vehicles
equipped with such systems in inspection and maintenance (I/M) programs that include heavy-duty
vehicles. Most of these commenters noted that for states to have a realistic chance of adopting such
programs, EPA must first provide a means for quantifying the benefits of such programs, similar to
the way that the current mobile source emission factor model, MOBILE6.2, quantifies benefits for
light-duty I/M programs. In discussing the MOBILE model, a handful of commenters pointed out that
the current model does not account for heavy-duty diesel deterioration, and therefore, there are no
excess emissions in the model for an I/M program to reduce. In requesting that the model be
revised to quantify a benefit for heavy-duty OBD-based programs, these commenters acknowledged
that the model would also need to be revised to include a deterioration curve for heavy-duty diesel
vehicles. Lastly, several commenters requested that in addition to quantifying the benefits from an
I/M program based on testing heavy-duty vehicles using OBD, that EPA also develop guidance on
how to implement such a program.
With respect to comments about modeling, EPA is working on a new mobile source emission
factor model - MOVES - that will replace MOBILES, and will quantify and project deterioration from
heavy-duty diesel vehicles. EPA is not intending to make further modifications to the MOBILE model.
EPA agrees that with the incorporation of emission control systems on heavy-duty vehicles, there is
the prospect that some of these systems will fail in-use and emission rates will rise as a result.
However, given the newness of these technologies and the unfamiliarity of there deployment on, for
the most part, commercial heavy-duty vehicles, there is no sound basis for estimating the incidence
of such failures, the emission impacts of such failures, the impacts of repairs on such failures, or the
costs of repair. Under EPA's heavy-duty in-use compliance testing program, EPA will be getting
substantial quantities of data on in-use performance of these OBD-equipped vehicles once they
enter the fleet. After accumulating sufficient data, EPA anticipates being able to conduct a
quantitative analysis of the incidence of failures and the emission impacts.
With respect to other comments, we appreciate the time and effort put into providing these
comments and will consider them should we move forward with developing an I/M proposal based
on HDOBD checks.
X. Statutory and Executive Order Reviews
We received no comments pertaining to Section X of the proposal preamble that require
analysis.
X.A Executive Order 12866: Regulatory Planning and Review
122
-------
X.B Paperwork Reduction Act
X.C Regulatory Flexibility Act (RFA), as amended by the Small Business Regulatory
Enforcement Fairness Act of 1996 (SBREFA), 5 DSC 601 et.seq.
X.D Unfunded Mandates Reform Act
X.E Executive Order 13132: Federalism
X.F Executive Order 13175: Consultation and Coordination with Indian Tribal Governments
X.G Executive Order 13045: Protection of Children from Environmental Health and Safety
Risks
X.H Executive Order 13211: Actions that Significantly Affect Energy Supply, Distribution, or
Use
X.I National Technology Transfer Advancement Act
XI. Statutory Provisions and Legal Authority
We received no comments pertaining to this section that require analysis.
XII. Other comments
Comments:
Are there any plans to make the MIL codes visible to the consumer? Currently the consumer
has to go to a repair technician to have the codes read. At some point will the codes be able to
appear so the consumer knows what codes have been triggered?
Maryland Department of the Environment, 2005-0047-0022, p. 2
Agency Response:
This is allowed by the regulation (see §86.010-18(b)(1)(iv)). At this time we do not have any
intention of requiring manufacturers to provide this information. Should we include OBD checks in a
possible future heavy-duty I/M rule, we may revisit this issue.
123
------- |