MY EPA HQ OAR ; FRL ; NHTSA

Transcription

1 Alliance of Automobile Manufacturers Comments on the Proposed Rulemaking to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards for MY EPA HQ OAR ; FRL ; NHTSA Submitted February 13, 2012

2 Alliance of Automobile Manufacturers Table of Contents EXECUTIVE SUMMARY... 1 THE ALLIANCE SUPPORTS THE PROPOSAL TO INCLUDE AN IN-DEPTH MID-TERM EVALUATION ADHERENCE TO THE MID-TERM EVALUATION PROCESS AND TIMING IS CRITICAL THE AGENCIES SHOULD CONDUCT PERIODIC TECHNICAL CHECK-INS THE AGENCIES MUST MAINTAIN THE CURRENT CAR AND TRUCK VEHICLE CLASSIFICATION FRAMEWORK THE PROGRAM FLEXIBILITIES IN THE NPRM WILL HELP MANUFACTURERS INTRODUCE NEW TECHNOLOGIES THAT PRODUCE CONCRETE ENVIRONMENTAL AND FUEL CONSUMPTION BENEFITS THE ALLIANCE SUPPORTS THE CREDITS FOR OFF-CYCLE GHG EMISSION REDUCTIONS AND FUEL ECONOMY IMPROVEMENTS THE AGENCIES SHOULD ENSURE THAT THE NHTSA REQUIREMENTS ARE FULLY HARMONIZED WITH THE EPA REQUIREMENTS ADDITIONAL DOCUMENTS THAT ELABORATE ON THESE COMMENTS ATTACHMENTS: APPENDIX 1 APPENDIX 2 APPENDIX 3 APPENDIX 4 APPENDIX 5 APPENDIX 6 APPENDIX 7 CONTENT AND PROCESS OF THE MID-TERM EVALUATION PROPOSED CHANGES TO THE OFF-CYCLE PROGRAM MOBILE AIR CONDITIONING CREDITS FOR DUAL FUEL E85 GASOLINE VEHICLES AND OTHER ALTERNATIVE VEHICLES COMPLIANCE WITH N2O REQUIREMENTS UPSTREAM EMISSIONS ACCOUNTING MISCELLANEOUS ISSUES (ADDITIONAL ADVANCED TECHNOLOGY VEHICLE ISSUES; BASE TIRE DEFINITION; GHG AND CAFE REGULATORY FRAMEWORK ISSUES; BACKSTOP STANDARDS; POLICE AND EMERGENCY VEHICLES; PROPORTION OF RECOVERED BRAKING ENERGY FOR HYBRID ELECTRIC VEHICLES; DRIVER SELECTABLE MODES; MISCELLANEOUS CORRECTIONS) TABLE OF ACRONYMS

3 Alliance of Automobile Manufacturers Executive Summary The Alliance of Automobile Manufacturers is an association of 12 vehicle manufacturers including BMW Group, Chrysler Group LLC, Ford Motor Company, General Motors Corporation, Jaguar Land Rover, Mazda, Mercedes-Benz USA, Mitsubishi Motors, Porsche, Toyota, Volkswagen Group of America and Volvo Cars of North America. Together, our members represent approximately three-fourths of new car sales in the United States. Two years ago the Alliance testified and commented in support of the model year (MY) greenhouse gas (GHG) and corporate average fuel economy (CAFE) rule and encouraged EPA, NHTSA and the California Air Resources Board (CARB) to continue the single National Program beyond MY We continue to support having a single National Program and appreciate the agencies efforts to pursue this goal. Of course, much has changed since For one thing, automakers today are driving this country s economic recovery. Autos represent the largest manufacturing sector in the United States, and auto sales are viewed as a leading economic indicator. Today, our industry supports eight million American jobs, $500 billion in annual compensation, and $70 billion in personal tax revenues. This year and next, automakers and suppliers are forecast to add 88,000 jobs per year in the United States. Another significant change is that automakers are offering more fuel-efficient choices than ever before models that achieve 30 miles per gallon or more on the highway. This is a 65% increase over model year The unprecedented effort over the coming 13 years to further our country s energy and environmental goals will succeed only if consumers buy the fuel-efficient vehicle technologies that will be offered. The Alliance comments provide detailed analysis of policy and technical aspects of the joint rulemaking. Shown below is a short summary of key issues and recommendations. 1. The Alliance supports the proposal to include an in-depth mid-term evaluation. This rulemaking reaches an unprecedented 13 years into the future. A mid-term evaluation process will allow the agencies to review a broad range of factors and make appropriate adjustments. It will provide better data and insight on a range of issues relevant to the appropriateness of the MY standards, including consumers willingness to buy the vehicles that are required to comply with the standards; future fuel pricing; and technology and raw materials costs. The Alliance comments on the mid-term evaluation include additional topics that the agencies should review. We recommend that, in addition to the proposed formal mid-term evaluation, the agencies continue their open dialogue and also conduct a series of smaller, focused technical evaluations - or check-ins - on the key assumptions of the proposal. The Alliance also requests a more specific description of the mid-term evaluation process and the specifics to be reviewed, including the timeline and procedures for assuring that the studies the agencies rely on are appropriately peer reviewed. 1

4 2 Alliance of Automobile Manufacturers 2. The agencies must maintain the current car and truck vehicle classification framework. The standards (i.e., footprint curves) that have been established, and the goals that have been placed are all based upon the current and known set of harmonized definitions. Any changes to the definitions during MYs necessarily would require a reevaluation of the appropriate level of stringency, cost, necessary flexibilities and final standards for any and all years in which a change would apply. 3. The program flexibilities in the Notice of Proposed Rulemaking (NPRM) will help manufacturers introduce new technologies that produce concrete environmental and fuel consumption benefits. The Alliance supports the flexibilities in the proposal and understands the needs of lower volume, limited line manufacturers. The program flexibilities in the NPRM will encourage early investment in technologies that produce concrete environmental and fuel consumption benefits that will be necessary to meet these challenging and increasingly stringent standards over the longer term. 4. The improved off-cycle technology framework for MY 2017 and later years should be made available for MYs The Alliance supports the additional detail and improved processes proposed for capturing off-cycle fuel economy and GHG improvements. This facet of the MY regulation recognizes improvements in fuel economy and GHGs that are not captured in current laboratory tests but do have real-world benefits. Recognizing the real-world improvements that these technologies achieve and how challenging it will be to place these technologies in the market, the agencies should allow automakers to apply all aspects of the revised off-cycle framework to MYs The proposed fleet penetration requirements and credit cap could slow new technology implementation and should therefore be removed. Throughout the NPRM, the agencies suggest that an automaker be required to apply advanced technologies to a minimum percentage of its fleet before receiving any level of credit. That would be the case even when the addition of an advanced technology to a single vehicle results in measurable, real-world GHG emission reductions. We propose that all actions be recognized, as they historically have been, on a pervehicle-so-equipped basis. This is an equitable and efficient approach, under which every vehicle built with the required technology for our customers receives credit. The prerequisite of specific penetration rates and imposition of a credit cap are economically inefficient and inconsistent with the goals of the rulemaking and may well have the unintended consequence of delaying the introduction of these technologies. 6. The agencies should assure equivalent program stringency. The proposed EPA and NHTSA requirements are coordinated, but not fully harmonized. The Alliance believes that adjustments to the NHTSA program are needed to ensure that it properly harmonizes with the EPA requirements under the differing statutory authorities provided to the agencies. 7. Automakers should not be required to account for utility GHG emissions. The proposed rule indicates that the agencies expect electric vehicles to become an

5 Alliance of Automobile Manufacturers increasingly large part of the car park. Yet the rule leaves open the possibility of requiring manufacturers to account for upstream emissions from electricity generation in the event that the Administration is unable to control these emissions through other channels. In other words, automakers may now be called on to not only make an unprecedented investment into vehicles with lower GHG emissions, but to also fill the void between this rulemaking and a comprehensive national energy policy. If Americans agree that programs to address upstream GHG emissions are appropriate, then such programs should be put in place through appropriate regulation of electricity generators, not by imposing additional burdens on vehicle manufacturers. 8. The performance-based Mobile Air Conditioning (MAC) efficiency test needs additional technical analysis and testing, and the Alliance stands ready to work with the agencies to address these concerns. The proposed MAC efficiency test is likely to interfere with the achievement of maximum credit levels for improved system efficiency that were fully included in the agencies feasibility analysis. The Alliance comments describe these concerns in depth and suggest that the test not be established as a strict requirement. Instead, we propose that the agencies continue to allow use of the credit menu, and that manufacturers work with EPA and NHTSA to provide reasonable verification of this progress through selected vehicle testing and other methods. Such verification will show that the menu amounts are appropriate, and that commensurate real-world progress is achieved. 9. The Alliance does not support additional requirements on Low Global Warming Potential (GWP) refrigerant systems (i.e., the high leak disincentive ). Some manufacturers have invested millions of dollars to redesign their vehicles and assembly plants for transitioning to new low-gwp refrigerants. These companies have counted on a specific level of credits in exchange for making this transition earlier than they might otherwise have done. This new proposal would potentially reduce the amount of credit, unfairly penalizing early adopters. 10. Additional time is needed for development of a method for measuring nitrous oxide (N2O). EPA has recognized the difficulties and complexities of evaluating, procuring and installing the equipment that would be needed to measure N2O. But, as our comments explain, EPA still has not provided sufficient time for manufacturers to incorporate accurate and robust N2O measurement capabilities into their test sites. The deadline for measuring N2O should be extended until the N2O measurement issues are resolved, and N2O measurement capabilities should be reevaluated during the mid-term and interim evaluations. By so doing, EPA would be providing manufacturers with sufficient time to evaluate appropriate test equipment and would be aligning possible N2O regulatory changes with possible subsequent changes to other light-duty GHG regulations. 11. Fuel quality improvement can help further the program goals. As EPA has requested, the Alliance plans to make its substantive comments about market fuel quality specifications within the context of the pending EPA Tier 3 proposed rule (and perhaps independently as well, separate from the Tier 3 rulemaking) and will 3

6 Alliance of Automobile Manufacturers not elaborate on them here. We want to note, however, that fuel quality can have a significant impact on fuel efficiency and GHG emission reductions. The Alliance Supports the Proposal to Include an In-Depth Mid-Term Evaluation. The MY GHG proposal includes provisions requiring EPA to conduct a mid-term evaluation of the MY light-duty GHG standards to determine whether those standards remain appropriate in light of technological and other changes that may have occurred since the time of proposal. 1 This evaluation process will be coordinated with NHTSA's effort to set final, binding CAFE standards for the model years. The mid-term evaluation will include consideration of up to date information, a holistic assessment of all of the factors considered by the agencies in setting standards and the expected impact of those factors on the manufacturers ability to comply. 2 To facilitate the evaluation, EPA (along with NHTSA and CARB) will publish a draft Technical Assessment Report (TAR), which will be peer-reviewed and made available for public comment. 3 EPA also will request comment on whether the MY standards remain appropriate under section 202(a) of the Clean Air Act (CAA) and whether the standards should be made more or less stringent. 4 No later than April 1, 2018, EPA will make a final determination whether the MY standards, as adopted in 2012, are appropriate. This process also is intended to guide NHTSA s decision-making regarding its MY CAFE standards. If EPA concludes that the standards are not appropriate, the agency will then initiate a rulemaking to adopt standards that are appropriate under section 202(a). Both EPA and NHTSA have stated that that they would issue a joint rulemaking at least 18 months prior to the beginning of the 2022 model year, consistent with the statutory directive in the Energy Policy and Conservation Act of 2005 (EPCA). The Alliance consistently has advocated that a mid-term evaluation is more than just appropriate; it is a critical component of this rulemaking package if these GHG and CAFE standards are to be successful. This rulemaking will govern vehicle production 13 years from now, a particularly long time period when predicting technologies, costs, infrastructure, fuels and consumer behavior. It comes on the heels of a five-year rulemaking that will, according to the agencies, cost automakers almost $52 billion the highest cost of any rulemaking imposed to date on the auto industry. 5 The agencies estimate the additional GHG reductions and fuel economy gains from this rule will cost automakers an additional $ billion, bringing the combined cost of the MY rules to somewhere between $185 and $209 billion. This unprecedented effort and expense will further our country s energy and environmental goals, but only if consumers choose to purchase these fuel-efficient, climate-friendly vehicle technologies Fed. Reg , (Dec. 1, 2011). 2 Id. 3 Id. 4 Id. 5 Final Rulemaking to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards, Regulatory Impact Analysis, EPA-420-R , April 2010, Table

7 Alliance of Automobile Manufacturers By necessity, the GHG and CAFE standards proposed here are predicated on significant assumptions regarding the future - including such factors as the pace of technological innovation, deployment of supportive infrastructure for alternative fuels and advanced vehicles, rates of market penetration for new vehicle technologies, future costs of emerging technologies, fuel cost and availability and consumer acceptance. The agencies have attempted to make reasonable projections based on recent data. Nevertheless, the proposed standards cover an unusually long time horizon, governing the production of vehicles over a decade into the future. The mid-term evaluation will allow the agencies to determine whether the CAFE and GHG standards should be adjusted as a result of customers willingness to buy vehicles that are required to comply with the standards, developments in technology, costs, safety, fuels, infrastructure and other relevant factors. Thirteen years into the future, consumer purchasing patterns will be the biggest unknown. Besides fuel economy, we know that consumers demand affordability, safety, convenience, performance and utility. One challenge we face is that fuel economy considerations often rank below these other attributes. Fuel prices, which are especially difficult to project, have a huge impact on how consumers weigh fuel economy at the dealership. All of this explains why the final rule should include a rigorous mid-term evaluation. Adherence to the Mid-Term Evaluation Process and Timing is Critical. EPA has proposed that the MY GHG standards will remain in effect unless and until EPA changes them by rulemaking. 6 EPA has not specifically provided for expedited judicial review of the results of the final mid-term evaluation or any final rule setting revised MY GHG standards. The Alliance would like to stress how important it is that both agencies follow the mid-term evaluation process laid out in the regulations, including strict adherence to the deadlines. Following the process as proposed should enable the agencies to consider all relevant issues, make an informed decision about the appropriateness of the MY standards, and allow sufficient time for the promulgation of different standards and/or judicial review, if necessary. The purpose of the mid-term evaluation provision is to ensure that the assumptions underlying the MY standards remain valid; to the extent that the assumptions are incorrect and the standards are inappropriate, the burden is likely to fall primarily on vehicle manufacturers. If EPA fails to follow the mid-term evaluation process or fails to meet the deadlines, it is probable that EPA will not have complied with the Section 202(a)(2) mandate to provide adequate time for the development and application of the technology required to comply with such standards. 7 Moreover, failure 6 Id. 7 See, e.g., International Harvester v. Ruckelshaus, 478 F.2d 615 (D.C. Cir. 1973) (Reversing EPA s refusal to temporarily suspend EPA s 1975 vehicle emission standards, the D.C. Circuit held that when EPA s support for its standards is tenuous and the agency has not provided a reasoned decision for its conclusions about the availability of technology, the agency has not met its burden of providing adequate lead time for the development of technology). 5

8 Alliance of Automobile Manufacturers to conduct the midterm evaluation or to meet the deadlines would constitute a failure to perform a nondiscretionary duty and/or final agency action. In making this comment, we wish to stress that the Alliance does not assume that EPA or NHTSA intend to deviate from the mid-term evaluation process or ignore its deadlines. We believe that all parties, including the agencies, will work in good faith to follow the process. We merely wish to stress that the success of the mid-term evaluation depends on close adherence to the process and the deadlines. If anything is allowed to undermine or delay the process, it creates a significant potential for disputes and difficulties in the future, something we all hope to avoid. The Agencies Should Conduct Periodic Technical Check-Ins. In the time leading up to the mid-term evaluation - and following the completion of the evaluation - the agencies should continue to check the validity of the assumptions upon which their standards are based. We suggest not only one formal mid-term evaluation, as the agencies have proposed, but also a series of smaller, focused, technical evaluations of, or check-ins on, the key assumptions of the proposal. These check-ins will allow the agencies to consider the latest relevant technical information, and thereby help the agencies keep the program on track and produce the best long term results. By having these check-ins the agencies will be better prepared to begin their formal mid-term evaluation and to make appropriate adjustments during the second half of the period covered by these regulations. The Agencies Must Maintain the Current Car and Truck Vehicle Classification Framework. In section IV.H. of the NPRM, NHTSA discusses the existing regulations governing the classification of cars and trucks. The agency states, NHTSA continues to believe that the definitions as they currently exist are consistent with the text of [the Energy Independence and Security Act of 2007 (EISA)] and with Congress' original intent." 8 Nevertheless, NHTSA requests comment on the possibility of changing the vehicle classification definitions, citing the long time frame of the rulemaking. First, we agree with NHTSA s assessment that the existing definitions for classifying vehicles are consistent with EISA and the original intent of Congress in EPCA. In past rulemakings, NHTSA has made some adjustments to the classification rules and clarified its interpretation of certain aspects of the rules. These efforts have accomplished their intended objectives by clearing up ambiguities and leveling the playing field. The Alliance is not aware of any further systemic problems with respect to the interpretation of the rules, and we do not believe any further changes need to be made. 8 See Supra note 4, at

9 Alliance of Automobile Manufacturers NHTSA requests comment on whether the current definitions might create an incentive to manufacturers to game vehicle designs in a way that would reduce potential fuel savings in the future. We consider this risk to be minimal because vehicle designs are evaluated primarily based upon an assessment of consumer acceptance. In other words, the key issue for manufacturers is whether potential purchasers would find the design useful and appealing, not how the vehicle would be classified under various regulatory programs. Moreover, the advent of "Reformed CAFE" has reduced the incentives for manufacturers to attempt to reclassify vehicles from one fleet to another, given the fact that even larger vehicles can be "CAFE-positive" based on their status relative to their footprint target. Finally, the existing definitions simply do not lend themselves to gamesmanship, a testament to NHTSA's efforts over the years to improve and refine the classification rules. Having developed a robust set of classification rules, NHTSA's priority should be preserving the stability of those rules, rather than engaging in continual modification and experimentation. Changes to the classification rules could have unintended consequences. For example, they could create new ambiguities or open up new opportunities for gamesmanship. If the changes are overly restrictive, they could have the effect of discouraging the production of vehicles that American consumers want to buy. A decision to change the classification rules at this point would create other problems. The classification rules represent a fundamental building block of the single National Program, and any attempt to change the car/truck definitions would have far-reaching consequences. All of the analyses of the proposed standards that manufacturers have conducted to date have been based on the assumption that the existing car/truck definitions would be retained. If the definitions applicable to MYs were changed, it would require a complete reevaluation of virtually all other aspects of the proposed rules, including the stringency of the standards, the cost of compliance and the adequacy of the program flexibilities. Such a reevaluation would be essential because, as a practical matter, a change to the classification definitions can be equivalent to a major change to the standards themselves. An amendment to the car/truck definitions could easily mean the difference between compliance and non-compliance for many manufacturers. Therefore, amendments to the classification rules would necessitate a brand new, top-to-bottom reanalysis of the standards by all manufacturers as well as NHTSA and EPA. And it is highly probable that large portions of the rulemaking package would need significant readjustment as a result of that exercise. In light of the above, the agencies must maintain the current car and truck vehicle classification framework. The standards (i.e., footprint curves) that have been established and the goals that have been placed are all based upon the current and known set of harmonized definitions. Any changes to the definitions during MYs MY necessarily would require a reevaluation of the appropriate level of stringency, cost, necessary flexibilities and final standards for any and all years that a change would apply. 7

10 Alliance of Automobile Manufacturers The program flexibilities in the NPRM will help manufacturers introduce new technologies that produce concrete environmental and fuel consumption benefits. The proposed rules properly include various provisions offering manufacturers some flexibility in developing their plans to comply with the CAFE and GHG standards. Some of these provisions enable manufacturers to earn credits that can be used to satisfy part of their compliance obligations. While some may think the term "credits," as used here, connotes reduced stringency or even "loopholes," that is not the case. The objective of the CAFE and GHG standards is to reduce actual fuel consumption and actual GHG emissions from vehicles driven on American roads. In some cases, however, the laboratory testing used by the agencies to measure fuel economy and GHG emissions may not fully reflect the improvements built into a vehicle by the manufacturer, due to limitations of laboratorybased tests. And improvements to reduce MAC system refrigerant loss can reduce GHG emissions from vehicles while having little or no impact on fuel economy. It is important for the rules to properly account for such factors. Otherwise, manufacturers would be encouraged to focus solely on the test procedures, and opportunities for real-world GHG reduction and fuel economy improvement would be lost. The Alliance believes that the various credit provisions proposed by EPA and NHTSA are essential elements of the rulemaking package. Below we offer our specific comments on the details of these provisions. The Alliance Supports the Credits for Off-Cycle GHG Emission Reductions And Fuel Economy Improvements. The overall GHG emission reductions proposed in the NPRM are a formidable challenge that requires new, creative approaches to emission reduction and energy efficiency. Continuing the off-cycle credit program provides an incentive to manufacturers to introduce new technologies that produce concrete environmental and fuel consumption benefits, provides flexibility toward meeting the increasingly stringent standards and encourages investment in technologies that will pay off over the longer term. However, the 10% minimum penetration threshold and the credit cap are barriers to the success of this feature and could result in the level of credits being out of sync with the level of GHG reductions that is actually achieved. Regarding the EPA approval process for technologies not included in the pre-defined list, the Alliance welcomes the efforts to provide a step-by-step process and 60-day timeline for approval of new technologies. The process changes outlined below would further streamline the program and provide more certainty to manufacturers irrespective of which approval process is used. In offering comments on this section of the NPRM, the Alliance supports the following overarching principles: 8

11 Alliance of Automobile Manufacturers Substantial GHG improvements should be achievable in off-cycle conditions using new technologies. The off-cycle technology credit menu is a necessary addition to the off-cycle program to avoid administrative delays and burdensome credit application requirements. It is counterproductive and unfair to create a 10% sales threshold during the initial phase-in period before some technologies can begin earning off-cycle credits. It is counterproductive to cap off-cycle credit attainment at 10 grams of carbon dioxide (gco2)/mile. The procedures for earning off-cycle credits need to be kept simple. Finally, opportunities exist to streamline traffic flow, reduce congestion and reduce emissions through better driving. For example, there are technologies that provide the driver or the vehicle with information for improved routing, or that provide the driver or the vehicle with information for more efficient vehicle operation. GPS technology can play a role in improving both driver behavior and vehicle operation. The opportunities for improvements through these eco driving technologies are not sufficiently defined for the Alliance to propose specific credit definitions and criteria at this time, but the industry hopes that it can work with the agencies in the future to create off-cycle credits for these technologies. The improved off-cycle technology framework for MY 2017 and later years should be made available for MYs The Alliance urges the agencies to allow manufacturers to utilize the off-cycle pre-defined technology list and values for MYs Providing this program feature in the earlier years improves the usefulness of the credit program and encourages manufacturers to introduce the listed technologies sooner, in lieu of postponing them to MY 2017 and beyond. There is every reason to incentivize early adoption of these technologies, since this would result in real CO2 emissions reductions. Additionally, such action would provide manufacturers with the same planning certainty regarding available credits as will be provided in 2017 and beyond. This in turn would help encourage earlier investments in off-cycle technologies. The alternative that an OEM faces (pathway 2 or 3) otherwise would be to make the investment without the certainty provided by the list, which may result in postponing investment until Off-cycle technologies will have to compete with resource demands for other vehicle technologies, and knowing that the same credit for the technology available in MY 2017 would be available starting in MY 2012 would help the business case for earlier deployment. The result could be earlier availability of GHG-reducing technologies for consumers to buy. 9

12 10 Alliance of Automobile Manufacturers The minor proposed changes in terminology are directionally correct but do not alleviate the overwhelming need to streamline the process by incorporating the pre-defined technology list for MYs as well as the later model years. The proposed fleet penetration requirements and credit cap could slow new technology implementation and should therefore be removed. While the Alliance strongly supports the concept of a pre-defined list, two proposed limitations -- the 10% minimum penetration rate and the 10 g/mi cap -- will constrain its ability to incentivize technology application. The public policy goal of maximizing early introduction of these technologies is at odds with both of these limitations and the NPRM fails to provide a compelling justification for either restriction. Further, the 10% threshold and 10 g/mi cap add an element of planning uncertainty that discourages use of the off-cycle program. The threshold also unfairly withholds credit for actual, real-world emission reductions that are achieved in the early stages of technology roll-out, before a 10% penetration can be achieved. New, innovative technologies are customarily initially introduced at low volumes in order to demonstrate the benefits, reduce costs and work through technology problems before the technology is rolled out in larger volumes. To minimize warranty concerns and expense, automakers always try to phase-in new technology at a measured pace across their fleets (often during the course of major vehicle redesigns). Requiring large step changes to get widespread penetration i.e., above a 10% penetration - is unlikely even with these off-cycle incentives. Requiring a minimum penetration rate would discourage companies from offering a new technology on a limited basis to test the technology and gauge consumer acceptance before launching it more broadly. The 10% minimum penetration threshold or any other minimum penetration rate may also have the unintended consequence of delaying investment in some technologies, at least until they can be applied to higher-volume models. Similarly, the 10 g/mi cap on credits would discourage maximum adoption of the pre-defined off-cycle technologies. Manufacturers would have less incentive to introduce technologies that would take them beyond the cap, leaving untapped GHG emissions reductions on the table. We think it would be productive to have further dialogue with the agencies regarding these issues. CAFE Fuel Consumption Improvement Values for MAC Efficiency Improvements and Off-Cycle Technologies In the GHG portion of the joint rulemaking, EPA is proposing to allow manufacturers to generate credits for improvements to MAC systems that reduce GHG emissions. EPA is also proposing to allow manufacturers to generate credits for implementing off-cycle technologies that result in real-world GHG reductions not fully accounted for under the existing test procedures.

13 Alliance of Automobile Manufacturers In the CAFE portion of the joint rulemaking, EPA, in coordination with NHTSA, is proposing to allow fuel consumption reductions (also called fuel consumption improvement values ) equivalent to the GHG credits allowed by EPA. These would apply for the credit menus provided for MAC efficiency and the use of off-cycle technologies. The proposal makes it clear that in the CAFE program manufacturers would only get credit for improvements that lead to better real-world fuel economy; improvements that are aimed at other GHG reductions such as reducing or eliminating MAC refrigerant leakage are not tied to fuel economy and would not qualify for CAFE program incentives. The expected generation of these MAC credits is accounted for by both agencies in setting the level of the overall GHG and CAFE standards they propose, but the ability to generate off-cycle credits and fuel consumption reductions is not accounted for in the standards. The agencies seek comment on the proposals to allow manufacturers to estimate fuel consumption reductions from MAC improvements and off-cycle technologies. While the Alliance has some suggestions for modifying the details of the above-described proposals, we firmly believe that the proposals in general are both appropriate and necessary. As noted by the agencies in the preamble to the proposed rule, President Obama's Memorandum of May 21, 2010 requested that NHTSA and EPA work together to develop...a coordinated national program under the [Clean Air Act] and the [Energy Independence and Security Act of 2007] to improve fuel efficiency and to reduce greenhouse gas emissions of passenger cars and light-duty trucks of model years " 9 As the President directed, and as all stakeholders recognize, a primary benefit of the single National Program approach is that it provides for harmonized EPA and NHTSA regulations so that manufacturers can build one fleet of vehicles that complies with both sets of rules. In keeping with that directive, it is important for EPA and NHTSA to include common provisions in both sets of rules to the maximum extent possible. The Alliance has already expressed its support for the inclusion of provisions allowing for MAC credits and off-cycle credits in the context of EPA s GHG rules, and we have explained why the inclusion of such provisions will further the goals of the program. In the interests of promoting harmonization, it only makes sense for the agencies to include comparable provisions in the CAFE rules. This applies to MAC and off-cycle improvements, as well as to the implementation of game changing-technologies in full-size pickup trucks and other credits. Failure to do so would only lead to increased disparities between the rules, giving rise to the possibility that manufacturers would need to undertake different actions to comply with the GHG rule on one hand, and the CAFE rule on the other. Of course, in developing these provisions, the agencies must be mindful of differences in the statutes underlying the two regulatory programs. Here, the agencies are being careful to ensure that the CAFE adjustments are limited to the demonstrated fuel economy benefits of the MAC and off-cycle improvements, which is entirely appropriate given the scope of the CAFE program. 9 See Supra note 4 (emphasis added). 11

14 12 Alliance of Automobile Manufacturers In light of the above, the Alliance believes that the agencies should proceed to include provisions accounting for the fuel economy benefits of MAC improvements and off-cycle technologies in the CAFE program, providing equivalent fuel consumption and CO2 credit values toward both the GHG and CAFE programs. This step will help to further harmonize one of the many remaining differences between the two regulations. The Agencies Should Insure that the NHTSA Requirements are Fully Harmonized with the EPA Requirements. EPA and NHTSA mention several times in the NPRM that they have worked to develop strong and coordinated Federal GHG and CAFE standards so that manufacturers can build a single fleet of vehicles to satisfy requirements under both programs as well as under the California program. As the agencies explain, this helps to reduce costs and regulatory complexity while achieving significant energy security and environmental benefits. While we appreciate the agencies' efforts to harmonize the two programs, more work needs to be done in this area. Specifically, NHTSA should modify its CAFE program for MY to better harmonize with EPA's GHG program. NHTSA s proposed CAFE standards account for many of the same factors that EPA considers in setting its proposed GHG standards. However, the proposed CAFE program does not include all of the program flexibilities built into the GHG program. NHTSA s program does account for some of the EPA flexibilities, including off-cycle technology benefits, mobile air conditioning benefits and benefits for hybridizing large work trucks. However, there are other important flexibilities that are present in the GHG program, but not the CAFE program. These include the advanced technology volume multiplier, the difference in quantification for advanced technologies with respect to the treatment of electricity, natural gas fuel utility factors, unlimited credit transfers between fleets and the one-time carry forward of previous credits through MY By MY 2025, the difference between the EPA's proposed fleet average standard and NHTSA's proposed fleet average standard equates to about 4.9 mpg. However, this difference is not large enough to offset the benefits of the additional flexible mechanisms included in EPA's program. In order to bring the two programs into better alignment, NHTSA needs to either increase the program flexibilities offered under the CAFE program or modify its curves to better reflect the other differences between the two programs. While the impact of the program differences is relatively small in the early years of the program, it will increase with the passage of time, particularly as manufacturers rely more and more on vehicle electrification in order to comply with the standards. Unless this imbalance is corrected, it will result in significant disharmony in the middle and later years of the time period covered by this proposal. The Alliance recognizes that, with respect to program flexibilities, EPCA and EISA impose some restraints on NHTSA that the Clean Air Act does not impose on EPA. Nevertheless, the Alliance believes that increased harmonization between the two programs is both possible and necessary. The Alliance strongly recommends that NHTSA undertake further study of its ability to include additional, appropriate program flexibilities to provide for

15 Alliance of Automobile Manufacturers equivalent stringency between the proposed CAFE standards and the proposed GHG standards. 10 To the extent that NHTSA cannot fully provide for equivalent stringency through the addition of program flexibilities, NHTSA should adjust the proposed CAFE standards themselves to fully account for the differences in the two programs. Such an adjustment is necessary to ensure that the President's goal of coordinated, harmonized CAFE and GHG programs is realized, and to avoid potential future problems due to disparities in the stringency of the two programs. Additional Documents That Elaborate on These Comments Additional documents that elaborate on the comments above can be found in the following appendices: Appendix 1 Appendix 2 Appendix 3 Appendix 4 Appendix 5 Appendix 6 Appendix 7 Content and Process of the Mid-Term Evaluation Proposed Changes to the Off-Cycle Program Mobile Air Conditioning Credits for Dual Fuel E85 Gasoline Vehicles and Other Alternative Vehicles Compliance with N2O Requirements Upstream Emissions Accounting Miscellaneous Issues (additional advanced technology vehicle issues; base tire definition; GHG and CAFE regulatory framework issues; backstop standards; police and emergency vehicles; proportion of recovered braking energy for hybrid electric vehicles; driver selectable modes; miscellaneous corrections) 10 Please note that the term "program flexibilities" does not refer to the enforcement provisions of the two programs, such as the payment of fines. The agencies' harmonization efforts should focus on achieving equivalent stringency in the CAFE and GHG standards, regardless of any differences in the enforcement mechanisms for the two programs. 13

16 Alliance of Automobile Manufacturers Appendix 1 Mid-Term Evaluation TOPICS FOR THE MID-TERM EVALUATION... 1 ARE THE COSTS OF ADVANCED TECHNOLOGIES DECLINING AS PREDICTED?... 2 ARE RESEARCHERS MAKING THE KIND OF BREAKTHROUGHS THAT THE AGENCIES ANTICIPATED?... 2 WHAT IMPACTS ARE THE NEW REQUIREMENTS HAVING ON SALES AND EMPLOYMENT?... 4 WHAT IMPACT ARE THE NEW REQUIREMENTS HAVING ON GOVERNMENT REVENUES AND HOW ARE GOVERNMENTS RESPONDING?... 4 HOW ARE THE NEW RULES IMPACTING VEHICLE SAFETY?... 4 IS THE NEEDED FUELING INFRASTRUCTURE AVAILABLE TO ENABLE PHEVS, BEVS AND FUEL CELL VEHICLES TO PENETRATE THE MARKET AT THE LEVELS PREDICTED?... 6 ARE CONSUMERS PURCHASING THE TECHNOLOGIES NEEDED TO ACHIEVE THE GOALS OF THE RULEMAKING?... 7 PROCESS FOR CONDUCTING THE MID-TERM EVALUATION... 8

17 Alliance of Automobile Manufacturers Appendix 1 Topics for the Mid-Term Evaluation The Alliance understands that EPA's mid-term evaluation will take place concurrently, and in conjunction with, NHTSA's process for setting final CAFE standards for MY The agencies should jointly examine progress achieved towards compliance with the standards, and assess the latest information available on key assumptions and trends used to develop the standards, including the criteria set forth for determining maximum feasible fuel economy standards in 49 U.S.C (f). Factors that should be considered include, but should not be limited to: Development of powertrain improvements to gasoline and diesel-powered vehicles; Level of employment in U.S. automotive sector; Availability and implementation of methods to reduce weight while assuring compliance with state and Federal safety, emissions and equipment laws and standards, and maintaining acceptable performance in consumer information crash testing and manufacturer due care testing; Actual and projected combined sales of alternative fuel vehicles; Actual and projected availability of public and private charging infrastructure for electric vehicles; Actual and projected availability of low carbon and technology-enabling fuels and infrastructure, along with adoption and implementation of clean and renewable energy standards; Costs, including average costs of technologies to ensure compliance with the standards, such as vehicle batteries and power electronics, mass reduction, and alternative fuels, and anticipated trends in these costs; Current and expected availability of state and Federal incentives/subsidies for advanced technology vehicles; Average payback periods for any incremental vehicle costs associated with meeting the standards, as well as up-front cost and impacts on consumer affordability; Costs for gasoline, diesel fuel and alternative fuels; Total light-duty vehicle sales and projected fleet mix; Consumer demand for and customer acceptance of fuel-efficient technologies, and consumer valuation of fuel savings; End-of-life costs associated with advanced technology vehicles; and Any other factors that may be deemed relevant to the review. Some recent studies attempt to identify opportunities for cost-effective near-term fuel economy improvements but also raise important questions about longer-term conditions. These questions call for information that is not yet available to EPA, NHTSA, the California Air Resources Board or any other party, including automobile manufacturers. The Alliance recommends that the mid-term evaluation focus on the issues as detailed below. During the evaluation, the agencies should seek expert peer-reviewed data and analysis, including the input of the National Academy of Sciences (NAS), to answer the following questions, among others: 1

18 Alliance of Automobile Manufacturers Appendix 1 Are the Costs of Advanced Technologies Declining as Predicted? Future technology costs are among the most difficult things to predict for MY 2020 and beyond. Factors that can change significantly over time include the availability and price of materials and parts, the number of suppliers and the rate of progress toward the production levels needed to achieve economies of scale. Recently, the National Research Council of the NAS issued its Assessment of Technologies for Improving Light Duty Vehicle Fuel Economy ( NAS Report ). Although the NAS report provides future cost estimates for numerous technologies, the NAS warns that data from automobile manufacturers and Tier 1 suppliers suggests a wide range of estimated incremental costs that makes assessments of cost-effectiveness very approximate. 1 The NAS characterizes future technology costs as more difficult to predict than the impact of these technologies on fuel consumption. In some cases, the NAS cost estimates were significantly higher than those of the agencies. The uncertainty surrounding the costs of integrating new technologies and then reaching economies of scale is illustrated in NHTSA s discussion of how to assign markup factors for learning. The Preliminary Regulatory Impact Analysis describes steps that NHTSA would need to take to develop accurate historical learning costs estimates for seven new CAFE and safety technologies. NHTSA concludes: This initial analysis indicates that adopting a cumulative production basis for learning applications could produce cost estimates that are within 4-7% of those used in the NPRM by 2025, with less variation in earlier years. However, this analysis is based on a very small sample of technologies and the data required to more precisely evaluate this issue are currently unavailable. Further, these data may not be obtainable without an extensive research effort, if at all. 2 Are Researchers Making the Kind of Breakthroughs That the Agencies Anticipated? Examples of anticipated battery technology breakthroughs include energy storage and management as well as power electronics capabilities; new battery chemistries and materials; new types of charging and faster charging; and advances in smart grid technology. Additional anticipated breakthroughs include the emergence of new, lowglobal warming potential fuels; high-efficiency transmissions; new down-weighting technologies and light-weight materials. The agency should also evaluate the ability to 1 National Research Council, 2011, Assessment of Technologies for Light Duty Vehicles, p NHTSA, Nov. 2011, Corporate Average Fuel Economy for MY 2017-MY 2025 Passenger Cars and Light Trucks Preliminary Regulatory Impact Analysis, p

19 Alliance of Automobile Manufacturers Appendix 1 meet increasingly stringent criteria pollutant standards using new combustion technologies for advanced internal combustion engines. The importance of gasoline pricing (including tax policy) cannot be underestimated, particularly for the post-my 2020 timeframe, when policy makers are counting on a rapid transition to non-petroleum transportation fuels. The report from Resources for the Future noted that it is an open question whether carbon prices at the levels currently under discussion will be sufficient, by themselves, to bring new fuel efficiency technology into the marketplace. 3 Yet the Energy Information Administration projects that fuel prices will be relatively stable over the next 15 years, with gasoline prices rising by less than 4 cents per year. 4 Even assuming that the energy cost savings will far exceed the increased up-front vehicle costs, consumer response is difficult to predict. The NPRM describes what the agencies call an energy paradox whereby consumers appear not to purchase products featuring levels of energy efficiency that, according to some metrics, might appear to be in their economic self-interest: Of 27 studies, significant numbers of them find that consumers undervalue, overvalue, or value approximately correctly the fuel savings that they will receive from improved fuel economy. The variation in the value of fuel economy in these studies is so high that it appears to be inappropriate to identify one central estimate of value from the literature. Thus, estimating consumer response to higher vehicle fuel economy is still unsettled science Regarding consumer response to [fuel economy] labeling information on cost savings: Whether the new label will help consumers to overcome the energy paradox is not known at this point. 5 Given how little is known about the energy paradox, the Alliance supports NHTSA s proposal to develop a Consumer Vehicle Choice Model to inform the mid-term evaluation. Such a model should also look at the other factors identified in the Preliminary RIA as having an impact on consumer purchasing decisions: sales taxes, insurance costs, the additional cost of auto loans and changes in resale value. To have credibility, the model needs to use real-world data, be developed in a transparent manner with full peer review, and should assess uncertainties in its predictions. 3 Kopp, Raymond J., Nov. 2007, Policies to Reduce CO 2 Emissions from the Light-Duty Vehicle Fleet, Resources for the Future, p Fed. Reg , (Dec. 1, 2011). 5 Id. at

20 Alliance of Automobile Manufacturers Appendix 1 What Impacts are the New Requirements Having on Sales and Employment? Automakers today are driving this country s economic recovery. Yet, in light of the uncertainty over consumer valuation of fuel savings and other factors, the agencies have not included an estimate of sales or employment impacts in the NPRM or supporting documents. The agencies need to understand and take these impacts into account to assure that the standards being put in place for MY do not reverse the economic gains and environmental benefits that have come from the industry s recent recovery. What Impact are the New Requirements Having on Government Revenues and How Are Governments Responding? As gas tax revenues decrease due to rising adoption of electric vehicles (EVs) and improved fuel economy, both the federal government and the states will need to determine how to address budget shortfalls resulting from falling gas tax revenues. In February 2009, the National Surface Transportation Infrastructure Financing Commission released a study recommending a shift from the present reliance on federal fuel taxes to fund federal surface transportation programs to a federal funding system based on more direct forms of user pay charges, in the form of a charge for each mile driven (commonly referred to as a vehicle miles traveled or VMT fee system). 6 Recently, certain states have been considering a vehicle miles traveled (VMT) tax or an annual registration fee on EVs as an alternative method to raise revenue for the state s transportation system. Legislation has been introduced in several states attempting to recover these lost tax revenues. State legislators in Arizona, Washington, Oregon, Texas, Indiana and Mississippi have attempted to address declining revenues with bills that would place fees (i.e., a road usage charge) on EVs, and in some cases, hybrid motor vehicles. States are expected to intensify such efforts to maintain their transportation infrastructure in light of these declining gas tax revenues. If states begin to enact such legislation, the agencies will need to evaluate how this could impact consumers willingness to invest in advanced technology vehicles. How Are the New Rules Impacting Vehicle Safety? Automakers, in conjunction with NHTSA and others, continue to work toward a common goal of reducing the annual number of fatalities and injuries that occur in motor vehicle crashes. The Alliance supports a CAFE and GHG rule structured to allow automakers to balance competing requirements in a manner that furthers this progress. 6 National Surface Transportation Infrastructure Financing Commission, Paying Our Way: A New Framework for Transportation Finance, Feb

21 Alliance of Automobile Manufacturers Appendix 1 Alliance members recognize that highway traffic safety is a shared responsibility and strive to do their part through the continuous improvement of the safety performance of their vehicles. NHTSA recently announced that the 2010 road fatality rate reached an historic low of 1.10 fatalities per 100 million vehicle miles traveled. Fatalities declined in most categories in 2010, including for occupants of passenger cars, SUVs, minivans and pickup trucks. 7 We take pride in our contributions toward this historic achievement and continue to work toward future progress by developing additional crashworthiness enhancements and introducing crash avoidance technologies. As breakthroughs in advanced material and powertrain technologies become available and their associated costs meet customer thresholds for affordability, consumers will benefit through an increase in vehicle fuel efficiency and a decrease in greenhouse gas emissions. However, the Alliance is mindful that such improvements must be implemented in a manner that does not compromise the rate of safety improvement that has been achieved to date. Achieving the proposed CAFE and GHG standards will rely on the availability of commercially viable emerging technologies for manufacturers to adopt. Should these technologies fail to mature as anticipated, greater reliance on mass reduction and downsizing in order to achieve these standards could occur. The Alliance supports the proposed mid-term evaluation and urges EPA and NHTSA to continuously update the safety analysis as part of this review. Even though the current rulemaking extends well into the future, there is a possibility that many of the advanced technology and mass reduction projections may not be realized in the proposed timeframe. Thus, when the agencies conduct their mid-term evaluation, it is critical that the safety analysis is updated to reflect the most recent crash data and revised projections regarding mass reduction scenarios. The Alliance supports NHTSA s intention to examine safety from the perspective of both the historical field crash data and the engineering analysis of potential future Advanced Materials Concept vehicles. NHTSA s planned analysis rightly looks backward and forward. However, with respect to looking ahead and the evaluation of concept vehicles, the Alliance recognizes that it is not sufficient to only consider regulatory and consumer information crash tests. A comprehensive evaluation of vehicle safety must also take into account realworld impact scenarios and the special requirements of vulnerable populations (e.g., children and elderly). These must also be adequately accounted for in any agency policy decisions. Analysis of the Lotus and FutureSteelVehicle concepts indicates that although these concept vehicles can be designed in a virtual world to perform well in virtual Federal Motor Vehicle Safety Standards and virtual Insurance Institute for Highway Safety tests, there remain concerns that these concepts yield aggressively stiffer crash pulses that may be 7 NHTSA Press Release, Dec. 8, 2011, U.S. Transportation Secretary LaHood Announces Lowest Level of Annual Traffic Fatalities in More Than Six Decades. 5

22 Alliance of Automobile Manufacturers Appendix 1 detrimental to rear seat occupants, vulnerable occupants and potential crash partners. Given the Computer-Aided Engineering (CAE) crash modeling uncertainties with respect to advanced materials that may possibly be available for mass production in the MY time-frame, it is possible that the real-world crash behavior of these concepts may not match that predicted in those studies. Further, significant uncertainties exist with respect to both manufacturing and CAE crash analysis of potential future advanced materials. CAE capabilities for some potential advanced materials that manufacturers are researching are far less mature than for materials currently in common use. Progress in these areas is highly competitive and therefore varies throughout the industry. As such, it will take considerable time and investment for each manufacturer to develop this knowledge and experience. Because agency projections fail to adequately take into account the timing and cost for the introduction of advanced materials, these projections are likely overly optimistic. Given the considerable uncertainty about future technology development, cost and consumer acceptability, the proposed mid-term evaluation is essential in order to assure that the maximum feasible fuel economy benefits are obtained in a cost-effective and safety neutral manner. Is the Needed Fueling Infrastructure Available to Enable PHEVs, BEVs and Fuel Cell Vehicles to Penetrate the Market at the Levels Predicted? President Obama has set a goal to put one million plug-in electric vehicles on U.S. roads by To meet this goal, and to achieve even more ambitious targets for post-2015 electrification, the U.S. will need to invest heavily in electric charging infrastructure. The Boston Consulting Group recently estimated that $8 billion in electric vehicle charging infrastructure would be needed by 2020 to support the growing market for plug-in hybrid electric vehicles and battery electric vehicles. 8 In addition to cost, a variety of other electric mobility infrastructure challenges remain, including development of uniform state, federal and local standards and protocols. Last year, the Alliance and AIAM issued a paper identifying these specific barriers and proposing a series of recommendations for addressing these challenges. 9 Hydrogen infrastructure is also needed to support the commercialization of fuel cell vehicles. 8 Boston Consulting Group, January 2010, Batteries for Electric Cars: Challenges, Opportunities and the Outlook to Alliance of Automobile Manufacturers and Association of International Automobile Manufacturers (now Association of Global Automakers), April 7, 2010, Proposed Elements of a Federal Initiative to Promote Electric Vehicle Deployment. 6

23 Alliance of Automobile Manufacturers Appendix 1 Are Consumers Purchasing the Technologies Needed to Achieve the Goals of the Rulemaking? Of course, the ultimate question will be whether mainstream consumers will be able and willing to purchase the technologies needed to achieve this country s fuel economy energy security and environmental goals particularly as the federal and state governments phase out many of the financial incentives that are available today. 10 The proposed regulatory language does not include this single most critical factor. Will mainstream consumers be willing and able to purchase vehicles with more fuel efficient technologies? As the NPRM states, there is considerable uncertainty in the economics literature about the extent to which consumers value fuel savings from increased fuel economy. 11 In fact, consumer acceptance is the most critical and unpredictable component of advanced technology vehicle deployment. Manufacturers strive to understand how their advanced technology vehicles will be used and fueled and what combination of pricing, incentives and vehicle attributes are needed to convince mainstream consumers to invest in new technologies. Recognizing the critical role of customers in determining the viability of future vehicle technologies, the NAS wrote: Manufacturers will choose fuel economy technologies based on what they think will be most effective and best received by consumers. Customers also will have a central role in what technologies are actually chosen and will make those choices based partly on initial and operating costs. Subsidies and other incentives also can significantly impact the market acceptance rate of technologies that reduce fuel consumption. Finally, adoption of these technologies must play out in a sometimes unpredictable marketplace and policy setting, with changing standards for emissions and fuel economy, government incentives, consumer preferences, and other events impacting their adoption. Thus, the committee acknowledges that technologies downplayed here may play a bigger role than anticipated, or that technologies covered in this report may never emerge in the marketplace In 2011, when automakers offered 231 models that achieved 30 miles per gallon or more on the highway, these vehicles represented about 37% of U.S. sales. More than ten years after hybrids were introduced to the U.S. market, in a year when there were 38 different models of hybrids on sale in the United States, these vehicles represented only 2.1% of new vehicle sales. In 2011, the top-selling pickup truck outsold all hybrids combined by a factor of two to one. (Data in this footnote is computed based on data from and from WardsAuto ( 11 See Supra note 4, at See Supra note 1, at

24 Alliance of Automobile Manufacturers Appendix 1 Process for Conducting the Mid-Term Evaluation The NPRM indicates that a draft Technical Assessment Report will be completed by November 15, 2017, and that EPA will make a final determination by April 1, The Alliance believes that a more detailed description of the process would be helpful. In particular, the final regulatory language should indicate that the agencies intend to perform a thorough analysis of consumer purchasing behavior, the single most important factor that will determine whether the goals of the program are being met. The final regulatory language should also include the following important details: start date of the evaluation and a schedule for major milestones to assure that the review is completed in time for EPA to make a fully informed regulatory determination; specific studies the agencies plan to conduct; details of the peer review process; availability of a pubic docket; role of NAS in the mid-term evaluation; and roles of other departments and agencies that provide or regulate alternative fuels and emerging technologies. 8

25 Alliance of Automobile Manufacturers Appendix 2 Table of Contents Off-Cycle Credits PROPOSED OFF-CYCLE CREDIT MENU... 1 HIGH EFFICIENCY EXTERIOR LIGHTING... 1 ENGINE HEAT RECOVERY... 4 SOLAR PANELS... 4 ACTIVE AERODYNAMIC IMPROVEMENTS... 5 ENGINE STOP-START... 6 OCCUPANT THERMAL COMFORT TECHNOLOGIES (E.G., ELECTRIC HEATER CIRCULATION PUMP)... 7 ACTIVE DRIVETRAIN WARM-UP (E.G., ACTIVE TRANSMISSION WARM-UP)... 7 ACTIVE ENGINE WARM-UP... 8 THERMAL CONTROL... 8 Glazing... 8 Active Seat Ventilation... 9 Solar Reflective Paint... 9 Active Cabin Ventilation... 9 Passive Cabin Ventilation ADDITIONS TO THE OFF-CYCLE CREDIT MENU HIGH EFFICIENCY ALTERNATOR HVAC ECO-MODE BYPASS VALVE FOR TRANSMISSION OIL COOLER ELECTRONIC THERMOSTAT AND ELECTRIC WATER PUMP SUMMARY OF RECOMMENDATIONS FOR CREDITS AND DEFINITIONS PROCESS FOR QUALIFYING OFF-CYCLE CREDITS FROM NEW TECHNOLOGIES... 17

26 Alliance of Automobile Manufacturers Appendix 2 Proposed Off-Cycle Credit Menu While there are substantial emission reductions that can be achieved through off-cycle technologies, it will be essential that the off-cycle program function effectively if the overall emission reduction goals are to be achieved. The proposed off-cycle credit menu is therefore a great addition to the GHG reduction and corporate fuel economy programs, as the experience thus far with separately testing and applying for off-cycle credits on each model has shown that the administrative obstacles inherent in this approach prohibit an effective program. The pre-defined list will incentivize automakers to apply technologies earlier than they might have otherwise. It also offers manufacturers certainty about how much credit they will earn if they choose to apply one or more of the technologies on the list. The Alliance welcomes the agencies willingness to add further technologies to the list as additional information becomes available. The Alliance supports establishment of credits for all of the proposed technologies, but, in a few cases, recommends revisions to the proposed credit amounts. In this section of our comments, we review each of the technologies on the list contained in the NPRM and, in several instances, recommend updates to the proposed technology definitions. In the next section of our comments, we recommend several additional technologies and credit amounts that should be added to the off-cycle credit menu. High Efficiency Exterior Lighting The Alliance supports this proposed off-cycle credit, but with modifications. EPA s calculation of the feasibility of a 60 watt total reduction threshold to qualify for this credit contains flaws in the calculation. Because of these flaws, achieving a full 60 watt improvement from the lights impacted by the credit may not be realistic. The 60 watt calculation included benefits from high efficiency low beam and high beam headlights, even though these lights are not covered by the credit provision. Due to their high wattage, the net benefits from these two high efficiency sets of lights are approximately 9 watts of the estimated 60 watt improvement. Creating such an unrealistically high improvement threshold for this technology could render the credit provision ineffective. In order to make the incentive to implement this technology more functional, this improvement threshold should be reduced to no more than 50 watts for the listed package of exterior lights, assuming no other changes are made to this provision. The credit amount for this package of lights would need to be adjusted accordingly. Also, Center High-Mount Stop Lamps (CHMSL) and brake lights impact the two-cycle fuel economy test, and high efficiency CHMSL and brake lights should not be a requirement to qualify for this off-cycle credit. Additional changes to this credit provision are also attractive. Roll-out implementation may be speeded if portions of the credit were available for separate, individual lights, 1

27 Alliance of Automobile Manufacturers Appendix 2 rather than requiring that all exterior lights feature high efficiency before a credit can be gained. This would allow each light to be swapped at the earliest possible time, in order to gain credits as quickly as possible. The most attractive candidates for individual credits are the largest three savings opportunities on the EPA/NHTSA list: parking/position lights, tail lights and license plate lights. Scaling the 1.1 gco2/mile credit proposed in the NPRM to the proportional benefits of these individual lights would indicate that individual credits are warranted of 0.2 gco2/mile for substitution of each of these lights individually (i.e., 0.2 gco2/mile for the parking/position lights, 0.2 gco2/mile for the tail lights and 0.2 gco2/mile for the license plate lights). The remaining 0.5 gco2/mile could then be earned by applying the package of all the remaining listed smaller lights in the NPRM. Because of their low usage and corresponding low energy consumption, we do not feel that it is warranted to include the requirement for more costly high efficiency turn signals in this package in order to qualify for the high efficiency lighting off-cycle credit. Our experience indicates that an off-cycle credit for low beam lights would also be beneficial. The source used by EPA showed potential LED low beam benefits of only a few watts, whereas automakers recent development experience has identified potential low beam savings opportunities of at least 60 watts. Based on EPA s usage calculations, a low beam savings of 60 watts should justify an off-cycle credit of 1.1 gco2/mile. We therefore recommend that a 1.1 gco2/mile off-cycle credit be offered on the menu for low beam lights that achieve power levels consistent with a savings of 60 watts below the baseline halogen technology. Beneficial emission reductions can also be achieved if a credit is offered for high efficiency Daytime Running Lights (DRLs). Based on the calculation below, we estimate a credit of 0.6 gco2/mile is justified for application of LED DRLs. Although DRLs are not mandatory, the reality is that they are often implemented as standard equipment by many manufacturers to improve highway safety. Due to their safety impact, they are a socially beneficial technology that has been encouraged by public policies at both EPA and NHTSA. Given that they are widely used, and that more efficient LED DRLs are often prohibitively expensive, it makes sense to offer an off-cycle credit as an incentive for LED DRLs. More widespread use of LED DRLs will result in real-world energy savings and GHG reductions. Because they are illuminated such a large portion of the time, they are the single most important exterior light to target for an off-cycle credit. In support of our above analysis, we recommend that the definition of high efficiency exterior lighting be updated to: High efficiency exterior lighting means a lighting technology that, when installed on the vehicle, is expected to reduce the total electrical demand of the exterior lighting system when compared to conventional lighting systems. LED lights specifically qualify. Separate credit values may be earned for high efficiency lighting installed in the following components: parking/position, tail lights, license plate lights, low beam lights and daytime running lights. Credits may also be earned for a high efficiency 2

28 Alliance of Automobile Manufacturers Appendix 2 lighting bundle that is installed in the following components: front and rear side markers, and backup/reverse lights. Sample Calculation of Energy Savings from Daytime Running Lights: Assume a 30mpg rated vehicle with Reduced Intensity Bulbs that consume 113W, traveling at an average speed of 30 mph for 1 mile. Also, assume that for every 1J of electrical power consumed the vehicle must burn 3.0J of fuel energy (33% efficient at producing electrical power with engine and alternator). Time to drive 1 mile => 1mi/30mph*3600sec/hr = 120sec Energy Consumed by DRL => 113W*120sec = 13.6kJ Fuel Energy Used to Produce DRL Energy => 13.6/0.33 = 41.1kJ Extra Gas Burned due to DRL per 1 mile driven => 41.1kJ/43020kJ/gal = gal Calculate Gallons Fuel/mile per watt = gal/mile/113w = E- 06 gal/mile/watt For 40W Energy reduction of LED DRL s = E-06 gal/mile/watt * 40W = gal/mile Convert to Grams CO2/mile = gal/mile * 8887 = Gram CO2/mile Calculate Grams CO2/mile LED DRL energy savings for nominal DRL usage, conservatively estimated at 60% = Gram CO2/mile * 0.60 = 0.64 Grams CO2/mile credit 3

29 Alliance of Automobile Manufacturers Appendix 2 Engine Heat Recovery The Alliance supports EPA s analysis for this credit. This technology is not yet available for commercial implementation. Offering a credit is appropriate and could play an important role in bringing this technology into commercial use. However, we believe that the credit amounts should be determined by a scalable application of the metric that EPA proposed, rather than a step-wise function that awards credits only in increments of 100 watt capacities. Further we believe that it is appropriate to award credits for recovered heat that is converted to either electrical or mechanical energy to meet vehicle requirements. To simplify implementation and fairly reward each application, the credit should be made a linear function, based on the wattage generated. We recommend that the definition be updated to: Engine heat recovery means a system that captures heat that would otherwise be lost through the exhaust system or through the radiator and converting that heat to electrical or mechanical energy to meet the requirements of the vehicle. Systems obtain credits according to the following formula: Solar Panels Credit (gco2/mile) = (System watt Capacity / 100) * 0.7 The Alliance supports EPA s analysis for this credit. Based on rough theoretical calculations and experimental data, offering a credit as proposed for each 50 watt unit of electricity generation is appropriate. Because there is a wide range of potential sizes for these panels, this credit should be scalable to reward panels that are smaller or larger than 50W. Also, the credit should not be confined to panels installed on the roof, since they may be installed elsewhere on the vehicle. This credit should be available for all vehicles, not only for electric-propulsion vehicles, since all vehicles can benefit from the additional battery charging to power accessories, even where it is not used for propulsion. Therefore, we suggest the following update to the proposed technology name and definition: Solar Panels means the installation of solar panels on a vehicle to capture and provide energy to the vehicle (e.g., provide energy to an electric drive system via battery charging or provide power to an electric motor or 12V battery trickle charging or cabin ventilation, etc.). Credit levels are granted according to the following formula: Credit (gco2/mile) = (Equivalent watt Output / 50) * 3.0 4

30 Alliance of Automobile Manufacturers Appendix 2 Active Aerodynamic Improvements The Alliance supports EPA s analysis for this credit. Active aerodynamic technologies hold great promise and are already entering commercial usage. Offering sizable off-cycle credits will be very helpful to stimulate faster adoption. The tables on p of the Draft Joint Technical Support Document (TSD) show the relationship between aerodynamic improvement and credit amounts. 1 The credits offered by EPA in the credit menu should be scalable, based on the lines in this table, rather than simply using one point on the table. Application of multiple active aerodynamic technologies can result in a Cd improvement of over 3%, whereas the proposed credit amounts are based on a 3% improvement. Therefore, to maintain an incentive to maximize the use of these technologies (e.g., active grille shutters plus active air dams), higher credit amounts should be allowed on the menu for aerodynamic improvements above 3%. EPA acknowledges in the TSD that larger aerodynamic improvements are possible, but suggests using model-by-model testing and applications to EPA for situations where greater credit is sought for these larger aerodynamic improvements. 2 Case-by-case testing and applications are overly burdensome, and it would be much simpler to amend the credit menu to use the tables below, from p of the TSD, to award appropriate credit for higher levels of aerodynamic improvement. 1 Draft Joint Technical Support Document: Rulemaking for Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards, EPA-420-D , Nov Id. at

31 Alliance of Automobile Manufacturers Appendix 2 In practice, drivers often operate vehicles at sustained, steady-state, high speeds. This condition is barely represented in the drive cycles and weighting system used in the 5-cycle fuel economy calculations. For a Cd improvement of 3%, this actually warrants a credit in excess of the proposed 0.6 gco2/mile for cars and 1.0 gco2/mile for trucks. Therefore, EPA should consider the lines on p of the TSD as minimum possible credit amounts for these technologies. We recommend the definition be updated as follows: Active aerodynamic improvements means technologies that are actively controlled to improve aerodynamic efficiency. Credits are awarded according to the following formulas: Engine Stop-Start Car: Credit (gco2/mile) = (Percent Reduction in Aero Drag, Cd) * 0.2 Truck: Credit (gco2/mile) = (Percent Reduction in Aero Drag, Cd) * 0.33 The Alliance supports the creation of a substantial off-cycle credit for this important technology. The proposed potential credit of 2.9 gco2/mile for cars and 4.5 gco2/mile for trucks, as proposed in the NPRM, is a major concern. We believe a credit of 5.5 gco2/mile is warranted; credits for this technology below 2.9/4.5 gco2/mile could substantially undermine the ability of the industry to achieve the overall GHG targets. We recommend that the definition be modified slightly to: Engine stop-start means a technology which enables a vehicle to automatically turn off the engine when the vehicle comes to rest and to restart the engine with driver action (e.g., applying pressure to the accelerator or releasing the brake). Off-cycle engine stop-start credits will only be allowed if the Administrator has made a determination under the testing and calculation provisions in 40 C.F.R. part 600 that engine stop-start is the predominant operating mode. Various studies and agency literature suggest longer periods are spent at idle than would be indicated by the FTP cycle, and support a high off-cycle credit for the stop-start technology. 3 3 See Harvey Michaels, David Brzezinski, U.S. EPA, OTAQ, Sue Kimbrough, U.S. EPA, ORD, 2003, Consistency in On-Road Mobile Source Activity Modeling, with an Application to Parked Passenger Cars, Linda Gaines, Argonne National Lab, To Idle or Not To Idle: That Is the Question, Your Car and Clean Air: What You Can Do to Reduce Pollution, EPA 420-F , Carrico, et al., Costly Myths: An Analysis of Idling Beliefs and Behavior in Personal Motor Vehicles, 37 Energy Policy 2881 (2009), pp

32 Alliance of Automobile Manufacturers Appendix 2 Occupant Thermal Comfort Technologies (e.g., Electric Heater Circulation Pump) In 2011, based on actual vehicle tests and 5-cycle calculations, EPA awarded General Motors (GM) credits of 1.8 gco2/mile for GM s full-size truck hybrids and 1.5 gco2/mile for its 2012 Buick Lacrosse and Regal hybrids. These vehicles use the auxiliary coolant pump to keep the stop/start feature working in cold weather, while continuing to provide heat to the passenger cabin. In contrast, the proposed credits for this technology are only 1.0 gco2/mile for cars and 1.5 gco2/mile for trucks. Using the 5-cycle methodology, the technology simply provides continued operation of the stop/start feature during the idle portions of the cold weather test, and the amount of fuel savings should be fairly consistent between applications. The differential between the proposed credit and the actual test results is sufficiently large that automobile manufacturers may independently file separate credit applications for the larger credit amount justified by actual vehicle testing of each application using the 5-cycle provisions of the regulation. This would be a large and unnecessary testing and administrative burden for the automobile manufacturers as well as the regulatory agencies. We recommend that EPA avoid this unattractive situation by making the menu credit for this technology more consistent with actual test values. A menu credit of at least 1.5 gco2/mile for cars and 1.8 gco2/mile for trucks is justified. To encourage additional technologies that provide similar benefits, the Alliance also recommends that the proposed definition be broadened to include other methods of maintaining occupant thermal comfort during off-engine periods. Specifically, we propose the following technology name and definition updates: Occupant thermal comfort technologies means technologies or strategies that maintain occupant thermal comfort during off-engine periods in a stop-start equipped vehicle or in a hybrid electric vehicle or plug-in hybrid electric vehicle (e.g., PTC heater or electric heater circulation pump). Active Drivetrain Warm-Up (e.g., Active Transmission Warm-Up) The Alliance supports EPA s analysis for the active transmission warm-up off-cycle credit of at least 1.8 gco2/mi. EPA s proposed definitions for credits for these technologies should, however, be broadened to allow the inclusion other methods of driveline fluid warm-up as well as other sources of waste heat (perhaps using different credit amounts for other variations of this technology). For example, credits should explicitly be allowed for systems that use a coolant loop to transfer the heat from the exhaust system to the transmission and/or engine, since this may be more practical than directly heating engine oil or transmission oil in a heat exchanger in the exhaust system. Also, the performance of the system is not significantly changed by the use or non-use of coolant in the heat exchange process. 7

33 Alliance of Automobile Manufacturers Appendix 2 Provisions should also be made to provide similar credits for other technologies that hasten transmission warm-up and viscosity management (perhaps without using exhaust gases). Some of these technologies are discussed in the next section of our comments, including quantification of potential credit amounts. In addition, we recommend the following updates to the proposed technology name and definition: Active drivetrain warm-up means a system that uses waste heat or waste energy to warm-up driveline fluids quickly and reduces parasitic drivetrain (transmission, axles, PTUs, t-cases) system losses related to friction and fluid viscosity. In this category, active transmission warm-up would receive credit of at least 1.8gCO2/mi. As mentioned in on p of the TSD, it is not necessary to heat the differential in rearwheel-drive vehicles in order to qualify for this credit. However, we believe that heating the rear differential in these vehicles for viscosity management might provide an attractive additional credit opportunity, and urge EPA to re-examine this possibility. Active Engine Warm-Up Our research supports the 1.8 gco2/mile off-cycle credit proposed by EPA for this technology. However, EPA s definitions for earning this credit should be broadened such that a coolant loop may be used to transfer the heat. Provisions should be made to provide similar credits for any other technologies that hasten engine warm-up to provide similar benefits. We recommend the following updates to the proposed definition: Active engine warm-up means a system that uses waste heat, thermal storage, or waste energy to warm up targeted sub-systems of the engine such that frictional losses are reduced. It would allow a faster transition from colder operation to warm operation, decreasing CO2 emissions and increasing fuel economy. In this category, active engine warm-up would receive credit of at least 1.8 gco2/mi. Thermal Control Substantial benefits are available from thermal management technologies, and we support establishing off-cycle credits for these technologies. Glazing The Alliance supports including glazing as an available off-cycle credit. However, the most effective of these technologies (solar reflective) have relatively high cost, weight and functional impact hurdles that must be overcome to integrate these technologies into the vehicle. For simplicity and clarity, the Alliance recommends that EPA state that the glazing area to be used in the calculation is the total glazing surface area. Also, credit should be granted for all vehicles that utilize glazing better than 62% Tts (except roof lights), regardless of whether the improved glass is marketed as the standard glazing or an optional upgrade. 8

34 Alliance of Automobile Manufacturers Appendix 2 In addition, we ask that EPA and NHTSA consider the input of suppliers of alternative glazing technologies on amendments or additions to the proposed rule. Consideration of a broader range of technologies will provide the necessary flexibility to achieve the desirable air conditioning-related emission reductions, for example, based on reduced glazing thermal conductivity. Active Seat Ventilation The Alliance supports EPA s analysis for this credit. As a practical matter, only the front seats need to be ventilated to qualify for this credit, and this should be stated in the final regulation. The analysis EPA uses to quantify the credit is based on two ventilated front seats. Rear seats are used much less than front seats, and the cost attractiveness of this credit opportunity would fall dramatically if it were required that more than the front seats be ventilated. Also, EPA s definitions should specify that this credit can be earned for systems that either pull air into the seat or push air out. The impacts on occupant comfort and energy consumption are the same, and both approaches are used. This could be accomplished by making the following minor modifications to the proposed definition: Active seat ventilation means a device which draws (or pushes) air or transfers heat/energy from the seating surface which is in contact with the occupant and exhausts (or pushes) it to a location away from the seat. Solar Reflective Paint The Alliance supports this credit while noting that, as with glazings, the analysis of energy benefits may be optimistic because of the worst case test conditions used in the studies by the National Renewable Energy Laboratory. Active Cabin Ventilation Cabin ventilation can be attractive for reducing air conditioner energy consumption and improving comfort upon vehicle entry. Research on ventilation technologies has confirmed that interior breath level temperatures can be reduced to the levels that EPA used in its analyses for both active and passive ventilation. There are many approaches that can be used, such as ventilating through slightly open windows or sunroofs, existing air conditioner ducts, or new air flow passages with dedicated fans. The ventilation may also be continuous or pulsed. Since such a wide range of approaches can be applied, the definitions used for this technology should not be overly prescriptive. The draft EPA definitions for both active and passive ventilation satisfy that objective, since the definitions allow for many ventilation techniques. Also, we have found that it is very difficult to physically measure air flow through the vehicle to a tight margin, so it is not practical to set air flow thresholds in order to qualify for these ventilation credits. 9

35 Alliance of Automobile Manufacturers Appendix 2 The cost of these technologies can be high. For example, active ventilation fans may need to be coupled to a photovoltaic panel which powers the fans, so that the system does not drain the battery if the vehicle is parked for long periods. The system used in EPA s analysis featured a unique sunroof with several small fans to pull hot air out of the cabin. Because are all very costly items, the credit offered for these technologies needs to be ample in order to make the business case for their implementation attractive. The Alliance therefore supports the definition of active cabin ventilation as proposed. Passive Cabin Ventilation As previously stated in the discussion of active ventilation, cabin ventilation can be attractive for reducing air conditioner energy consumption. Research on ventilation technologies has confirmed that interior breath level temperatures can be reduced to the levels that EPA used in its analyses for passive ventilation. All of our other comments relative to active ventilation also apply to passive ventilation. For example, the system analyzed by EPA employed automatic sunroof features together with eight new floor vents in the vehicle. These are significant hardware changes to the vehicle which would require a significant credit in order to make an adequate business case for implementation. We propose to broaden the definition of Passive Cabin Ventilation slightly as follows: Passive cabin ventilation means ducts, devices or methods that utilize convective airflow to move heated air from the cabin interior to the exterior of the vehicle. Additions to the Off-Cycle Credit Menu In this section of our comments we discuss several additional technologies that provide offcycle GHG benefits. These are all well-understood technologies that have no technical barriers that would prevent much more widespread implementation. However, for various reasons, none of these technologies are currently in widespread use. Establishment of the recommended off-cycle credits on the credit menu could be expected to rapidly result in substantial GHG benefits through widespread implementation of these technologies. High Efficiency Alternator This was a good recommendation on EPA s initial off-cycle technology list, as contained in the EPA/NHTSA July 2011 Supplemental Notice of Intent. However, this technology subsequently did not appear in the proposed credit menu contained in the NPRM. We recommend that it be added back to the menu. The standard 2-cycle fuel economy test is performed with accessories off, and even the 5-cycle tests only activate some accessories, such as the air conditioner on the SCO3 test. In contrast, real-world driving has higher average electrical loads from a variety of accessories such as radios, lights, rear-seat entertainment systems, wipers, power window motors, etc. Conservatively, we estimate that at least a 20 amp average electric load differential exists in actual real-world driving 10

36 Alliance of Automobile Manufacturers Appendix 2 over the typical 20 amp load during the 2-cycle test. High efficiency alternators provide fuel consumption benefits for this extra 20 amp real-world-driving differential which are not captured on the 2-cycle test, and so these benefits should be eligible for an off-cycle credit. A traditional baseline alternator might have had an efficiency rating under the Verband der Automobilindustrie (VDA, the trade association represent German automobile manufacturers) test procedure of 60% to 64%, with high efficiency models having ratings above 68% VDA. To translate these differences into a GHG-equivalent, GM ran simulations of three different alternators on a range of four different vehicles using the NEDC drive cycle. The alternators were the Valeo SG11 (61% VDA), Bosch E6 (69% VDA) and Denso DSO (70% VDA). Each of these was simulated using the GM Unified Models for the Cadillac SRX, Chevrolet Sonic, Chevrolet Cruze, and the new GM Alpha platform. Actual performance curves for each of the alternators were used for one set of simulations. To simplify the comparisons, another set of simulations was done for the high efficiency Bosch and Denso alternators, wherein the actual performance curves were compared to alternator performance curves set to be exactly 10% lower than the actual curves. Also, a very simple set of simulations was done to compare a flat 70% efficiency alternator to a 60% efficiency alternator. The following chart shows the efficiency curves for the Bosch and Valeo alternators, and also shows the curve for the Bosch alternator modified to be exactly 10% below the actual Bosch data: Alternator Efficiency (%) Amp 20 Amp 30 Amp 40 Amp 50 Amp Bosch E6 Valeo SG11 30 modified Bosch E Alternator Speed (rpm) 11

37 Alliance of Automobile Manufacturers Appendix 2 Using this approach, the following CO2 savings were estimated for the extra 20 amps typical of real word driving, compared to the 2-cycle test s 20 amp load: Vehicle Alternator Comparison SRX Valeo (~61% VDA) Vs Bosch (69% VDA) 0.1 Sonic Valeo (~61% VDA ) Vs Bosch (69% VDA) 1.1 Alpha Valeo (~61% VDA) Vs Bosch (69% VDA) 0.8 Cruze Valeo (~61% VDA) Vs Bosch (69% VDA) 0.7 SRX Const 60% vs 70% 1.3 Sonic Const 60% vs 70% 2 Alpha Const 60% vs 70% 1 Cruze Const 60% vs 70% 1.5 G/mi CO2 Savings with High Eff Alternator SRX Sonic Alpha Cruze Bosch mod (~61% VDA) Vs Bosch (69% VDA) 1.3 Bosch mod (~61% VDA) Vs Bosch (69% VDA) 1.8 Bosch mod (~61% VDA) Vs Bosch (69% VDA) 1.1 Bosch mod (~61% VDA) Vs Bosch (69% VDA) 1.1 SRX Valeo (~61% VDA) Vs Denso (70% VDA) -0.4 Sonic Valeo (~61% VDA) Vs Denso (70% VDA) 1.2 Alpha Valeo (~61% VDA) Vs Denso (70% VDA) 0.7 Cruze Valeo (~61% VDA) Vs Denso (70% VDA) 0.7 SRX Sonic Alpha Cruze Denso mod (~61% VDA) Vs Denso (70% VDA) 0.8 Denso mod (~61% VDA) Vs Denso (70% VDA) 1.6 Denso mod (~61% VDA) Vs Denso (70% VDA) 0.9 Denso mod (~61% VDA) Vs Denso (70% VDA)

38 Alliance of Automobile Manufacturers Appendix 2 As would be expected, the complexity of vehicle operations results in a spread in results from this exercise. However, there are consistently CO2 savings, with a representative savings appearing to be approximately 1.0 gco2/mile. We therefore recommend that an off-cycle credit of 1.0 gco2/mile be established for vehicles that use an alternator rated at 68% VDA or better. Alternator loads are rising as more electric features are used in vehicles; this credit amount is conservative in that it does not account for this trend of increasing vehicle-generated electricity usage. HVAC Eco-Mode We appreciate the agencies willingness to recognize the real-world fuel economy and GHG improvements from driver-selectable technologies. We understand that expected usage data will be required as a basis for adding these technologies to the menu. Since 2011, GM has featured an eco button on the Chevrolet Equinox that allows drivers to select a driving mode which adjusts powertrain operation to achieve an improvement of approximately one mpg in combined city/highway driving. This has proven to be a popular feature with many customers, and GM has collected substantial data that documents high customer usage of this driving mode. In a two-week survey of 3,500 owners of the 2011 Equinox conducted through OnStar technology, the following usage information was collected: 50.3% of customers were using the eco mode for in excess of 90% of their driving, 57.4% of customers were using the eco mode in excess of 50% of the time, and 34% had never activated their eco mode. While the fuel economy benefits of the powertrain eco settings appear in the city and highway fuel economy tests, the benefits of the new HVAC eco settings do not. These HVAC eco features should therefore be candidates for off-cycle GHG emissions credits. The driver selectable HVAC eco mode initiates alternative air conditioner settings, such as reduced blower speeds and evaporator core temperatures, both of which reduce load on the compressor. During cold weather, the blower speed is also reduced, which reduces blower energy consumption while also improving powertrain warm-up. Below are tables presenting test data from six SCO3 tests and two Cold CO tests on a 2013 Equinox. This data clearly shows the different, energy-saving operating characteristics from normal mode to eco mode. Based on GM 5-cycle testing on the 2013 Equinox, GM calculated that the HVAC energy savings for the eco button are 1.8 g CO2/mile. Based on GM OnStar usage data, 50% usage is an appropriate adjustment for this credit, since at least 50% of the drivers are using it at least 90% of the time, with another 7% of drivers using it between 50% and 90% of the time, and 13% of drivers using it between 0% and 50% of the time. Multiplying 50% usage by GM s tested 5-cycle improvement of 1.8 gco2/mile yields a credit of 0.9 gco2/mile. The 13

39 Alliance of Automobile Manufacturers Appendix 2 Alliance therefore recommends that a 0.9 gco2/mile credit for an HVAC eco button be established on the EPA menu. Miles per Gallon Impact of HVAC ECO Mode Button SCO3 #1 SCO3 #2 SCO3 #3 Average Normal Mode ECO Mode Improvement Cold Bag 1 Cold Bag 2 Cold Bag 3 Cold CO Combined Normal ECO Mode Improvement Bypass Valve for Transmission Oil Cooler This is an alternative approach to faster transmission warm-up and viscosity management, without using exhaust gases. Many vehicles, especially large trucks, feature transmission oil coolers that provide increased functionality to operate under heavy loads. One drawback, however, of the traditional transmission oil cooler is that it continuously cools the oil, even under circumstances when it would be advantageous for fuel economy to have the transmission oil gaining heat more rapidly. Adding a bypass valve for the transmission oil cooler allows the oil flow to be controlled to provide maximum fuel economy under a wide variety of operating conditions such as cold weather. However, bypass valves are not currently commonly used with transmission oil coolers. The Alliance recommends that an off-cycle credit of 0.3 gco2/mile be established for vehicles that have a transmission oil cooler with a bypass valve. This 0.3 gram proposed credit is proportional to the 1.8 gram benefit observed for active transmission warm-up using exhaust gases, based on the benefits observed during GM engineering development work with these technologies. Also, the bypass valve is additive, and even synergistic to the benefits of using exhaust gases for faster transmission warm-up, so both of these active transmission warm-up credits should be available on a vehicle. For example, a transmission with an oil cooler, combined with the bypass valve and exhaust gas-assisted warm-up, allows fine tuning of viscosity management, since accelerated heating or cooling of the oil can both be accomplished, depending on operating conditions. 14

40 Alliance of Automobile Manufacturers Appendix 2 Electronic Thermostat and Electric Water Pump This technology provides both faster engine warm-up and tighter continuous control of engine block temperature. The variable electric water pump only performs as much work as is demanded of it at any given time, based on instructions coming from the thermostat. This allows the water pump to work less immediately after startup, thereby warming the engine faster, since less heat is taken from the engine by its cooling system. This is especially valuable on the cold CO test cycle. The pump also works less under other operating conditions, thereby reducing the parasitic drag on the engine when compared to a conventional mechanical belt driven pump. This provides for tighter control of the engine temperature to its ideal (which optimizes fuel economy), with less energy spent on engine cooling. In an analysis using a conservative 3% on FTP City combined cycle, the improvement from a mechanically driven water pump and a conventional thermostat produces a 1 gram CO2/ benefit using the 5-cycle calculation method. The vehicle used for this analysis was a 2.4L four cylinder SUV. Summary of Recommendations for Credits and Definitions Based on these considerations, we propose an off-cycle technology credit menu as follows: Technology Car Credit Truck Credit High Efficiency Exterior Lighting Up to 2.8 Up to 2.8 Engine Heat Recovery 0.7 per 100W 0.7 per 100W Solar Panels 3.0 per 50W 3.0 per 50W Active Aerodynamic Improvements 0.2 per 1% Cd 0.33 per 1% Cd Engine Stop-Start Occupant Thermal Comfort Technologies Active Drivetrain Warm-up Active Engine Warm-up Glazing * Up to 2.9 Up to 3.9 Active Seat Ventilation * Solar Reflective Paint * Active Cabin Ventilation * Passive Cabin Ventilation * High Efficiency Alternator HVAC Eco-Mode Bypass Valve for Transmission Oil Cooler Electronic Thermostat and Electric Water Pump *Maximum combined thermal control technology credit is 3.0 gco2/mile for cars and 4.3 gco2/mile for trucks. 15

41 Alliance of Automobile Manufacturers Appendix 2 High Efficiency Light Credits Lighting Location: Car Credit Truck Credit Low Beams Daytime Running Lights Parking/Position Tail Lights License Plate All Other Exterior (except Turn Signals) Total Potential Credit As noted above, we also recommend the following updates to the off-cycle technology definitions: High efficiency exterior lighting means a lighting technology that, when installed on the vehicle, is expected to reduce the total electrical demand of the exterior lighting system when compared to conventional lighting systems. LED lights specifically qualify. Separate credit values may be earned for high efficiency lighting installed in the following components: parking/position, tail lights, license plate lights, low beam lights, and daytime running lights. Credits may also be earned for a high efficiency lighting bundle that is installed in the following components: front and rear side markers, and backup/reverse lights. Engine heat recovery means a system that captures heat that would otherwise be lost through the exhaust system or through the radiator and converting that heat to electrical or mechanical energy to meet the requirements of the vehicle. Systems obtain credits according to the following formula: Credit (gco2/mile) = (System watt Capacity / 100) * 0.7 Solar Panels means the installation of solar panels on a vehicle to capture and provide energy to the vehicle (e.g., provide energy to an electric drive system via battery charging or provide power to an electric motor or 12V battery trickle charging or cabin ventilation, etc.). Credit levels are granted according to the following formula: Credit (gco2/mile) = (Equivalent watt Output / 50) * 3.0 Active aerodynamic improvements means technologies that are actively controlled to improve aerodynamic efficiency. Credits are awarded according to the following formulas: Car: Credit (gco2/mile) = (Percent Reduction in Aero Drag, Cd) * 0.2 Truck: Credit (gco2/mile) = (Percent Reduction in Aero Drag, Cd) *

42 Alliance of Automobile Manufacturers Appendix 2 Engine stop-start means a technology which enables a vehicle to automatically turn off the engine when the vehicle comes to rest and restart the engine with driver action (e.g., applies pressure to the accelerator or releases the brake). Off-cycle engine stopstart credits will only be allowed if the Administrator has made a determination under the testing and calculation provisions in 40 C.F.R. part 600 that engine stop-start is the predominant operating mode. Occupant thermal comfort technologies means technologies or strategies that maintain occupant thermal comfort during off-engine periods in a stop-start equipped vehicle or in a hybrid electric vehicle or plug-in hybrid electric vehicle (e.g., PTC heater or electric heater circulation pump). Active drivetrain warm-up means a system that uses waste heat or waste energy to warm-up driveline fluids quickly and reduces parasitic drivetrain (transmission, axles, PTUs, t-cases) system losses, related to friction and fluid viscosity. In this category, active transmission warm-up would receive credit of at least 1.8 gco2/mi. Active engine warm-up means a system that uses waste heat, thermal storage, or waste energy to warm up targeted sub-systems of the engine such that frictional losses are reduced. It would allow a faster transition from colder operation to warm operation, decreasing CO2 emissions, and increasing fuel economy. In this category, active engine warm-up would receive credit of at least 1.8 gco2/mi. Active seat ventilation means a device which draws (or pushes) air or transfers heat/energy from the seating surface which is in contact with the occupant and exhausts (or pushes) it to a location away from the seat. Passive cabin ventilation means ducts, devices or methods that utilize convective airflow to move heated air from the cabin interior to the exterior of the vehicle. Process for Qualifying Off-Cycle Credits from New Technologies EPA originally adopted this program for model years MYs as an optional credit opportunity for new and innovative technologies that reduce vehicle CO2 emissions, but for which CO2 reduction benefits are not significantly captured over the 2-cycle test procedure. The agency adopted the off-cycle credit option to provide an incentive to accelerate the introduction of these types of technologies that result in concrete reductions in CO2 emissions. However well-meaning this program, its actual use has been undermined by uncertainties over which technologies would be deemed eligible, how much credit would be provided, and the risks and burdens on manufacturers inherent in a cumbersome caseby-case approval process. The NPRM would require three complete sets of 5-cycle tests (with technology on and technology off ) for relatively large impacts of over 3% GHG reduction. For technologies with less than a 3% impact, manufacturers would be required to run five complete 5-cycle 17

43 Alliance of Automobile Manufacturers Appendix 2 tests (with technology on and technology off ), plus complete an analysis using EPA s Vehicle Simulation Tool. It can be expected that almost all of the off-cycle technologies will individually yield emission reductions of less than 3%, making mandatory the completion of five full test series and the simulation analysis. While an extensive off-cycle credit menu will result in the largest portion of off-cycle emission reductions, the regulatory agencies also need to have an open system for submitting and evaluating new off-cycle technologies, since an ongoing flow of new technologies can be expected. In this regard, simplicity and openness should be the goals, since the program benefits from bringing in new approaches that can be developed into future credit menu items for widespread implementation. As currently proposed, the provisions for off-cycle credit applications require very extensive vehicle testing and modeling. These proposed requirements lean towards a very high level of accuracy and proof, but would also serve to hinder the flow of new technologies by setting such a high administrative burden. Our expectation is that the vast majority of credits would be pursued through the menu, and that the process for alternative individual applications will represent a small amount of total credit. It may be counterproductive to require a high test and analysis burden, in return for small credits, since manufacturers may choose simply not to undertake such unattractive projects. This may discourage efforts which, if undertaken, could eventually grow beyond the initial low-volume stage into widespread usage, with corresponding broad emission benefits. This was, after all, the original goal of the MY off-cycle program. The Alliance proposes some simplifications to these provisions. Based on our members experiences over the past year evaluating technologies for off-cycle credit in the MY timeframe, we would expect that the credit menu would be used almost universally to gain credit for any technologies that are listed on the menu, rather than testing and quantifying slightly larger credits for each model under this section of the regulation. The overwhelming majority of credits would be expected to flow from the menu, and the relatively small amount of credits arising through this alternative section of the regulation does not warrant the proposed very high level of accuracy and documentation through repeat testing and modeling. In order to encourage development of new technologies under this section of the regulation, the Alliance recommends that only three five-cycle tests be required for all applications, with no requirement to use the EPA Vehicle Simulation Tool. This retains a high level of accuracy, but with lower administrative obstacles. We expect the mechanism for energy savings to be easily explained and, if documented through actual vehicle testing, do not see the need to also conduct simulations. Although we have no experience with this particular tool, we anticipate that any single simulation tool may not be compatible or easily adaptable to analyze specific, unusual technologies, especially technologies which address the proliferation of very diverse sources of off-cycle energy losses. 18

44 Alliance of Automobile Manufacturers Appendix 2 We also recommend that EPA retain discretion to approve applications which forego some of the 5-cycle tests if such testing is deemed unnecessary. In many cases, technologies would reasonably be expected to have no impact on certain test cycles. For example, cold weather technologies might be expected to have no impact on the SCO3 cycle. In these cases, it would be wasteful to require multiple tests for cycles which do not impact the credits. Finally, opportunities exist to streamline traffic flow, reduce congestion and reduce emissions through better driving. For example, there are technologies that provide the driver or the vehicle with information for improved routing, or that provide the driver or the vehicle with information for more efficient vehicle operation. GPS technology can play a role in improving both driver behavior and vehicle operation. The opportunities for improvements through these eco driving technologies are not sufficiently defined for the Alliance to propose specific credit definitions and criteria at this time, but the industry hopes that it can work with the agencies in the future to create off-cycle credits for these technologies. 19

45 Alliance of Automobile Manufacturers Appendix 3 Table of Contents Mobile Air Conditioning Credits INTRODUCTION... 1 CREDITS FOR INDIRECT (EFFICIENCY) IMPROVEMENTS... 1 IDLE TEST... 1 TEMPERATURE AND HUMIDITY TOLERANCES... 3 IDLE TEST ENGINE SIZE ADJUSTMENT... 4 IDLE TEST REASONABLE VERIFICATION... 4 AC17 TEST... 5 AC17 THRESHOLDS... 7 AC17 CORRECTIONS AND CLARIFICATION... 7 TEST VEHICLE SELECTION... 8 TEST BURDEN... 9 BENCH TESTING CONCLUSION CREDITS FOR REDUCING DIRECT EMISSIONS (LEAKAGE) USE OF UPDATED SAE J USE OF SAE J HIGH LEAK DISINCENTIVE REFRIGERANT LEVEL MONITORING IMPLEMENTATION OF NEW ALTERNATIVE REFRIGERANTS REFRIGERANT AVAILABILITY ENGINEERING RESOURCES NEEDED TO TRANSITION TO R1234YF... 18

46 Alliance of Automobile Manufacturers Appendix 3 Introduction The MAC credit provisions have become an essential piece of EPA s GHG program. Rapid progress is underway as a result of the MY provisions to improve MAC efficiency, reduce refrigerant leakage, and switch to a new low global warming refrigerant. In order to maintain the progress made in improving MAC, as well as the overall integrity of the broader GHG program, the MAC provisions must be continued in MY at unreduced credit levels. This was one of the foundations of the consensus to move forward toward the dramatic GHG reduction targets set through MY 2025, as the program stringency is designed on the basis of these alternative compliance mechanisms. Any obstacles that would prevent maximum attainment of MAC credits are therefore a fundamental threat to the achievement of the overarching program goals. The Alliance comments are directed toward making a success of the MAC provisions through efficient crediting processes that achieve real-world GHG reductions, commensurate with the credit levels granted. The Alliance stands ready to work with the agencies to address the concerns outlined below. Credits for Indirect (Efficiency) Improvements Idle Test The MY EPA GHG program uses a menu of credits for application of various MAC efficiency technologies. However, beginning in MY 2014, an idle test requirement is superimposed on the credit menu, such that certain emissions thresholds must be achieved on the idle test before credits can be granted for application of the technologies on the menu. The Alliance, as well as individual manufacturers, supported the credit menu, but commented during that rulemaking process on the inadvisability of superimposing the idle test. One key concern was that the most promising real-world MAC efficiency technologies - those that reduce compressor workloads at moderate ambient temperatures - would not have sufficient time to show their benefits during the idle test procedure. Although the idle test is not performed at a high ambient temperature, the idle test procedure calls for the systems to be operated as if they were responding to high ambient temperatures. Automatic systems are set to 9 o F below the 75 o F ambient temperature. Thus, in order to get to the unrealistically low 66 o F interior setting required by the test procedure, the automatic systems work at maximum cooldown through much of the 10-minute MAC-on portion of the test. Manual MAC systems are tested at maximum cooldown for ten minutes (as if responding to a high ambient temperature) and at a low fan setting for ten minutes. The SAE IMAC study, which was a major basis for EPA s idle test thresholds, demonstrated a 30% lifecycle energy efficiency improvement for a nationally representative mix of high, moderate and light MAC loads. But the energy efficiency improvements came primarily 1

47 Alliance of Automobile Manufacturers Appendix 3 from better moderate temperature technologies. At moderate and light loads, efficiency improvements of 40% or more were recorded. An energy efficiency improvement of only 5% to 10% was measured at high ambient temperatures. Since the development of the MY regulation, many idle tests have been run in automobile manufacturer laboratories, and the anticipated concerns, as well as other problems, have been confirmed. One of the most prominent issues identified for the idle test is that smaller displacement engines will receive significantly less EPA CO2 credit for the same HVAC technology content, compared to larger displacement engines. Manufacturer testing also shows that on small engines, even the most sophisticated MAC systems will probably receive little EPA CO2 credit because of their idle test results, thereby sharply reducing the incentive to apply MAC efficiency technologies to these vehicles. Test-to-test variability is large relative to the scale for getting credit, creating an additional element of regulatory compliance planning uncertainty, resulting in an additional barrier to MAC efficiency technology implementation. High variation is inherent in this test, in part due to the higher CVS dilution (low CO2 concentrations) of tailpipe exhaust gases for a vehicle at idle, which makes CO2 measurement highly variable being close to the limit of detection of the analyzer. Many of these preliminary idle test results and their implications have been communicated to EPA over the past two years, and these issues are discussed by EPA in the NPRM. Although the idle test has some relationship to MAC efficiency, it does not sufficiently get at the most important area for improvement, the moderate load technologies, since much of the idle test is conducted as if the vehicle were in a high ambient temperature, with corresponding high demands placed on the MAC system. The thresholds for obtaining full credit were established by EPA based on the premise of a 30% efficiency improvement over baseline MAC technologies. This is a challenging hurdle that can only be met if moderate and light load technologies are allowed to demonstrate their benefits to a substantially greater extent than the idle test allows. Based on the tests reported to EPA, no vehicle with an engine below approximately 2.5 liters in displacement would receive full credit for its MAC technologies, due to the idle test results (e.g., TSD p. 5-40). Unmodified, the idle test poses a potentially insurmountable obstacle to MAC improvements for credits on many vehicles, especially those with small engines which are anticipated to be predominant in the future. The test-to-test variability is an additional obstacle, since it introduces inherent uncertainty as to whether technology additions will ultimately result in credits. MAC operation is a fraction of overall vehicle fuel consumption, and the fuel consumption difference between a good, advanced technology MAC system and a bad MAC system is a matter of only a few grams of CO2 per mile (or per minute). On the idle test scale, the total range from maximum credit to no credit is only 6.4 g CO2/minute. Thus, significant test-totest variation (even as low as a gram or two) carries the potential to negate a large portion 2

48 Alliance of Automobile Manufacturers Appendix 3 of the planned credits for any program to improve MAC efficiency technology or to move the vehicle to the next level of credits. In testing reported at the United States Council for Automotive Research (US CAR), standard deviations of over 1.0 gco2/minute were consistently found from repeated idle testing on the same vehicle. Statistically, this shows that inherent test-to-test variation could negate a substantial portion of the MAC indirect credits for any vehicle. Idle Test Temperature and Humidity Tolerances The Alliance supports EPA s effort to broaden the ambient air temperature and humidity specifications for the idle test and the optional idle test, from the current requirements for humidity levels of 50 ± 5 grains/pound, average temperature 75 ± 2 F and instantaneous temperature: 75 ± 5 F. We also support the EPA proposal to relax temperature and humidity requirements in order to use test cells designed for FTP testing. 1 The data that manufacturers shared with EPA on June 30, 2011 (EPA Ann Arbor Meeting) showed that some tests run at a manufacturer s lab failed/exceeded these stringent specifications defined for the idle test on temperature and humidity. Automaker emission test facilities are not all designed for tight temperature and humidity controls such as are required for SC03 test chambers. Since the idle/optional idle test will not require solar loads, these tests will probably not be performed in SC03 solar test cells. The non-solar test cells are designed to run standard emission tests such as the FTP (EPA75), Highway Fuel Economy (HWFET) and US06 tests, where temperature and humidity specifications are less stringent as compared to the current A/C SC03 test specifications. If the temperature and humidity range is not widened, it will cause a large percentage of void tests that the manufacturers will be forced to repeat simply because of seasonal temperature and humidity variation or because the original specifications of test cell HVAC system were not designed around the idle test limits. This will add to the manufacturer testing burden and costs without any significant benefit to the accuracy of the test results. Therefore, the Alliance encourages EPA to widen these temperature and humidity specifications limits on the idle/optional idle tests, as proposed, so that they can be performed in the non SC03 chambers without added testing and cost burdens on manufacturers. The Alliance recommends the following humidity and temperature tolerances: (1) Ambient humidity within the test cell during all phases of the test sequence shall be controlled to an average of 50 ± 10 grains of water/pound of dry air. (2) Ambient air temperature within the test cell during all phases of the test sequence shall be controlled to 75 ±5 F on average and 75 ±10 F as an instantaneous measurement. Air temperature shall be recorded continuously at a minimum of 30 second intervals Fed. Reg , (Dec. 1, 2011). 3

49 Alliance of Automobile Manufacturers Appendix 3 Idle Test Engine Size Adjustment To partially address these concerns, EPA proposes an optional revised set of thresholds for performance on the idle test beginning in The revised thresholds are adjusted for engine size, so that smaller engines receive better scores and more ability to earn credits from application of MAC technologies on the credit menu. Although we believe all the idle test requirements should be discarded from the MAC program, if the idle test is kept, then it is useful to add this optional engine-size adjusted set of thresholds. Adding the engine size adjustment option to the MAC program is an improvement, but it does not solve the most fundamental problems with the idle test. The most important technologies for real-world energy savings, the moderate load technologies, would continue to show little benefit on the idle test, and the test-to-test variability would remain inherently high. We therefore believe that both the idle test and the engine size-adjusted idle test contribute little toward the goal of ensuring real-world greenhouse gas reduction, and that they will do little to encourage improved MAC technology implementation. Reasonable Verification The Alliance supports the EPA goal stated in the NPRM of reasonable verification that the technologies receiving credit from the credit menu are actually producing commensurate levels of GHG reduction. 2 Chrysler, Ford and GM have worked with EPA and CARB over the past several months at USCAR to evaluate procedures for MAC efficiency testing. This work has identified several key criteria and issues for MAC testing that directly relate to the goal of reasonable verification. This research has resulted in the draft AC17 test procedure discussed in the NPRM. Since the creation of the EPA MAC program, automobile manufacturers have believed that the MAC technology improvements on the EPA credit menu will result in actual GHG emission reductions that significantly surpass the amounts of the credits on the menu. This stems from the methodology used to quantify credits on the menu. EPA began with an overall inventory of estimated fuel consumption from MAC operation, then apportioned improvements from that inventory to the percentage improvements identified for various prominent MAC efficiency technologies, especially those used by the SAE IMAC cooperative research program. EPA estimated that MAC operation accounted for 14.3 g CO2/mile on average for each vehicle, representing 3.9% of national light duty vehicle GHG emissions. 3 The cap of 5.7 g CO2/mile was derived as 40% of the 14.3 g CO2/mile total, and the credit 2 Id. 3 Final Rulemaking to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards, Regulatory Impact Analysis, EPA-420-R , April 2010, p

50 Alliance of Automobile Manufacturers Appendix 3 for each technology was roughly calculated as a percent improvement of the 14.3 g CO2/mile total. 4 Automobile manufacturers commented at that time that the EPA inventory of fuel consumed for MAC operation was at the low end of the range of estimates by various researchers, such as studies by the National Renewable Energy Laboratory (NREL) and Northeast States Center for a Clean Air Future (NESCCAF)/CARB. For example, NREL estimates were that MAC usage consumed 5.5% of national light duty fuel usage. NESCCAF and CARB together estimated that MAC operation accounted for 5.3% of a vehicles fuel usage. 5 These alternative estimates are at least 70% higher than the 14.3 g CO2/mile figure ultimately used by EPA. In the MY Regulatory Impact Analysis, there is a complicated comparison of these studies, and EPA made numerous adjustments and assumptions in arriving at its ultimate figures, noting numerous uncertainties along the way. Without fully replaying that debate, suffice it to say that automobile manufacturers believed MAC compressors were engaged much more often and used more total fuel than EPA estimated, and that the higher estimates from relatively sophisticated analyses by NREL and NESCCAF/CARB were closer to the real-world MAC energy consumption. If the higher amounts of baseline MAC fuel consumption were used, the reductions from each MAC efficiency technology would be expected to be much greater than the figures used in the EPA credit menu. EPA finalized the regulation using its own (low) estimate of total MAC fuel consumption, resulting in credits on the menu that should be very conservative compared to actual vehicle usage, as measured by other researchers and industry data, and as used in the SAE IMAC program. This should be kept in mind when considering the need for thorough and precise verification procedures, since EPA s very conservative methodology in creating the credits results in a huge margin before realworld emissions reductions might fall short of the credited amounts. AC17 Test The auto industry has shared EPA and CARB s interest in furthering understanding of these issues, and the draft AC17 vehicle test procedure that we jointly developed through USCAR is a significant step to aid future research. We therefore strongly support adding an option to use the AC17 procedure as a reporting-only alternative to the idle test to demonstrate that a vehicle s MAC system is delivering the efficiency benefits of the new technologies from the credit menu in MY During the development of the AC17 procedure, it was shown that a very complicated and elaborate procedure would unavoidably be needed to accurately measure MAC energy consumption. A high level of vehicle instrumentation is needed, in part to understand what is happening throughout the test and identify voided tests where the procedure may have gone wrong. A high number of voided tests are to be expected. Also, the procedure cannot 4 Id. at Id. at

51 Alliance of Automobile Manufacturers Appendix 3 be conducted in a standard FTP test cell, but needs a climate controlled chamber with solar lamps that meets SC03 test specifications. At least four hours is needed for the test, due to the desire to include solar soak periods that attempt to comprehend the benefits of reduced thermal load technologies. In addition, even with the improved repeatability and added instrumentation of the AC17 test, testing has shown that a single AC17 test may not demonstrate the benefit of a single or a bundled set of technologies. In order to statistically verify the benefit, multiple tests may need to be run and statistically analyzed. In total, although this test is unlike any other used for emission certification (or any other regulatory certification program), the high level of complexity was arrived at and determined to be necessary in pursuit of the goals of an accurate, reproducible test that could distinguish a good MAC system from a bad one under representative ambient climate conditions, and which could validate the benefits of the technologies on the EPA credit menu. EPA proposes testing with the relevant technologies turned on then turned off in order to validate the improvements that are awarded from the credit menu. We agree with the principle of validating the menu credit amounts through actual vehicle testing, but note the difficulty of doing this on a comprehensive basis for every model of vehicle. In actuality, there is typically no baseline vehicle and baseline MAC system that is engineered and built without the improved MAC technologies for true apples-to-apples comparison of tests with the MAC technologies turned on, and then turned off. For example, if a variable compressor with the associated computer controls is engineered for a vehicle platform, it is typically applied across-the-board on that platform, and no systems are built with a fixed compressor or with the computer controls that would be necessary to get optimized performance from this fixed compressor. The exercise suggested by EPA is more appropriately viewed as a research exercise, rather than a traditional vehicle emissions certification program. For example, the IMAC program tested a baseline Cadillac vehicle, and then added various new MAC technologies, including new, smarter computer controls that improved vehicle integration. The improvements from each added new technology were measured as the IMAC program progressed. However, IMAC was a research program that cost hundreds of thousands of dollars, involved experts from approximately 40 corporations, government labs and agencies, and spanned approximately two years. To expect such a complicated procedure to validate the menu credit amounts for a large number of vehicles is simply not feasible and would violate the boundaries of reasonable verification. In some circumstances, it may be possible to obtain (or build) baseline vehicles which approximate the apples-to-apples technology on versus technology off comparisons that EPA seeks. A research program would seek to identify the best of these vehicle opportunities and, on a selected basis, use them to answer the research questions that are to be studied. Such a program would allow for multiple repetitions of the test to be run on the selected vehicles to statistically verify results. There are only approximately eight efficiency technologies on the MAC credit menu (including the two levels of reduced reheat). This is a manageable number of technologies to assess using the AC17 test (or 6

52 Alliance of Automobile Manufacturers Appendix 3 other methods) in a detailed way in order to validate the amount of credits provided by the menu. On a survey basis, it can also be used to show that the credit amounts from the menu are being consistently achieved, but it is not reasonable to require that this be tested on every model or platform. Due to the complexity of the required tests, the rarity of good baseline comparison opportunities, and the overall high test burden of this program in comparison to the amount of credits involved, the Alliance recommends that EPA to utilize the AC17 test solely to validate menu credit amounts and monitor progress on a sample basis, rather than as a mandatory certification test that must be run on every vehicle model or platform in order to achieve MAC efficiency credits. AC17 Thresholds EPA raises the possibility of setting an absolute required threshold for the AC17 test, as was done in 2014 with the idle test, rather than comparing improvements to case-by-case baselines. This raises a host of complex issues. Establishing these standards would be a very complex exercise, and two key issues emerge that show this to be a bad idea, even before entering into the particular issues related to the standard setting process. First, the test burden would be overwhelming. The AC17 test is much longer and more complicated than other emissions tests, and it cannot be used in the same way. Second, the planning uncertainties and implementation difficulties from such a complex program would become an insurmountable obstacle to getting better MAC technologies implemented. Rather than speeding progress in this area, the tests would slow or stop the progress that is being made. AC17 Corrections and Clarifications We note the need for the following minor technical corrections and clarifications in the AC17 procedure written in the NPRM: During the soak that occurs between the preconditioning and test cycle (for both the solar-on and solar-off portions of the test) we believe that instead of turning the cooling fan off, it should be set to 4 mph. The low wind level is a more representative real-world condition. Maintaining that fan speed impacts solar glazing technology. This change would impact 40 C.F.R (f)(8). A solar load tolerance of 850 W/m 2 +/- 45W/m 2 should be allowed, and the procedure should specify "solar off" for the MAC off test. If windows are partially cracked during the test in order to accommodate wiring or other test instrumentation, a piece of foam or other flexible insulation should be used to keep a tight seal when the window should be closed or when wires are 7

53 Alliance of Automobile Manufacturers Appendix 3 inserted through the gap. (Without this, the gap is too large where the various cords run through the window and the vehicle will not heat up consistently during the solar soak.) In the test procedure flow chart on p of the NPRM the word Nominal should be added to Time (Minutes) AC17 Test Vehicle Selection Because the AC17 test is so long, expensive and complicated, the number of tests should be minimized to a manageable level, no matter what the ultimate purpose of the test. These tests must be done in climate-controlled SC03 chambers, not in regular FTP test cells, and test capacity is very limited in SC03 chambers. MAC systems generally have consistent designs and specifications on each vehicle platform (except that some platforms now have hybrid powertrain models, which would usually have a very efficient MAC electric compressor), so performing one test per vehicle platform would give a good overview of MAC efficiency performance. However, vehicle platforms usually have multiple engine and transmission combinations available, so the number of tests would escalate rapidly if various engine and transmission combinations require testing. For example, for 2012 model certification, one manufacturer, GM, tested and certified approximately 20 vehicle platforms. There were approximately 85 GM platform/engine combinations, and this number then approximately doubles if transmission combinations are included. Clearly, if various engine/transmission combinations required testing, the MAC testing program would surpass the number of tests performed for tailpipe certification, which violates the standard of reasonable verification, considering the relatively small environmental impacts of MAC indirect GHG emissions. We therefore object to the criteria for test vehicle selection proposed in the NPRM, which defines a platform as a group of vehicles with common body floorplan, chassis, engine and transmission. 6 We propose the following platform definition, which is adapted from the current EPA definition for a carline : Platform means a group of vehicles within an OEM which has a degree of commonality in construction (e.g., body, chassis). Platform does not consider the model name, brand or marketing division, does not consider any level of decor or opulence and also does not consider characteristics such as roof line, number of doors, seats, or windows. A platform may include vehicles from various fuel economy classes, including both cars and trucks. This definition provides the flexibility to combine the large variations which occur within platform families that use the same MAC architecture. Intra-platform variation is based on floorplan such as two-door, four-door and wagon/crossover variants or SUV/pickup 6 See Supra note 1, at

54 Alliance of Automobile Manufacturers Appendix 3 variants, as well as powertrains. However, the benefit of the menu technologies should not be significantly affected by body style or powertrain. We recommend that wherever the term platform is used in MAC regulations, it be based on the flexible and inclusive definition proposed above by the Alliance. In addition, in 40 C.F.R (c)(6)(iii), EPA proposes that the highest selling subconfiguration within each platform be tested in the first model year for which a MAC system is expected to generate credits and then one additional sub-configuration must be tested in each subsequent model year until all sub-configurations within the platform have been tested. Given the fact that a platform will contain tens, if not hundreds, of subconfigurations, this proposal essentially eliminates the possibility for a manufacturer to carryover representative data from a prior model year and unnecessarily increases a manufacturers overall testing burden. Therefore, we urge EPA to allow the use of good engineering judgment when selecting a representative test vehicle for each platform and when determining whether carryover of data is appropriate. Also, if EPA were to persist on having sales figures be a part of the basis for test vehicle selection, sales projections should be clearly allowed as the basis for these test vehicle selections rather than waiting for actual sales figures to be finalized at the end of the year. Finally, we believe that 40 C.F.R (c)(6)(iv) is redundant to 40 C.F.R (c)(6)(iii) and should be removed. AC17 Test Burden The AC17 test takes approximately four hours, which is eight times as long as the idle test. Further, AC17 requires more technician time to set up the elaborate instrumentation, and it requires SC03 climate-controlled test cells. More voided tests are also expected with AC17, due to all the complications. The following are specific recommendations to reduce test burden: (a) Overview. The reference for humidity should be changed from 50 percent relative humidity to 69 grains of water / pound of dry air to be consistent with our recommendation for test cell ambient conditions (please see our next recommendation). 9

55 Alliance of Automobile Manufacturers Appendix (c) Test cell ambient conditions. The proposed A17 test procedure limits are extremely stringent. SC03 test facilities were not designed to operate at 77 F at 69 grains of water/pound of dry air humidity at 850 W/m 2 solar load. We recommend that the tolerances be widened to minimize test voids without significantly impacting testing accuracy as described below: (c) Test cell ambient conditions. The test cell ambient temperature and humidity recorded values should lie within the specifications at least 95% of the time (1) Ambient Air Temperature (i) Temperature = 77 ± 3 F air temperature on average and 77 ± 5 F air temperature instantaneous (2) Ambient Humidity (i) Humidity = 69 ± 5 grains of water/pound of dry air on average and 69 ± 10 grains of water/pound of dry air instantaneous (d) Interior temperature measurement. A thermocouple location tolerance should be added to (d) Interior temperature measurement. The current requirement is too restrictive for high volume production testing (language implies a location of exactly 30mm and 330 mm). The word nominally should be added before each measurement tolerance and OEM s should not be required to validate exact physical location by documenting dimensions. Also there may be some vehicles (like 2 seat sports cars or pickup trucks) where the distance below the roof or behind the headrest is not achievable due to physical constraints of vehicle (example; rear deck lids or window). In these cases language should be added to allow OEM s to use good engineering judgment to get a close as possible to these prescribed physical locations (e) Air conditioning system settings. The requirement for 6 volts at the motor is too unwieldy a specification to be reliably executed in high volume testing and may not even be achievable, given varying motor voltage configurations such as vehicles with systems other than 12- volt. We recommend that it be reduced to the setting closest to 6 volts at the motor or the blower switch position at 50% of maximum blower speed, or immediately below 50% if there are an odd number of positions. For example, position 2 if the maximum is position 4, or position 3 if the maximum is position 7. Also, the word nominal should be added to the requirement to provide 55 degf since (1) this temperature may change with different segments of the test (idles, accelerations, cruses, decelerations) or (2) with some vehicles may not achievable or stable at this idle period, and (3) implies a tolerance of ±0.49 degf which may be difficult to set in such a short period of time. Finally, on vehicles that 10

56 Alliance of Automobile Manufacturers Appendix 3 "default to recirculated air above 75 F," the OEM should have the option to let this feature function as intended and not be required to start in recirculated air and change to outside air at the first idle of the SC (f)(8) Procedures following the preconditioning cycle. Following the preconditioning cycle, the test vehicle and cooling fan(s) are turned off, all windows are rolled up, and the vehicle is allowed to soak in the ambient conditions of paragraph (c)(1) of this section for 30 ±1 minutes. The solar heat system must be turned on and generating 850 W/m 2 within 1 minute of turning the engine off. This requirement implies that the solar heat system must be turned on and achieves 850 W/m 2 within 1 minute of turning the engine off. This takes four events to accomplish: (1) turn engine off, (2) turn solar lamps (heat) load on, (3) set pyrometers up near the vehicle, (4) solar load lamps warm up to produce 850 W/m 2. One minute (total) is insufficient time to accomplish these tasks as just item #4 can take more than five minutes to achieve (produce 850 W/m 2 ). Since this short time requirement of one minute is really not a critical element to the test, we recommend changing it to: The solar heat system must be turned on within one minute of turning the engine off. The 30 minute soak starts immediately after the solar load has achieved 850 ± 45 W/m 2. Facility calibration data on solar lamp warm-up can be used to establish the start of solar soak time. Total soak time would be time to turn on the lights (e.g., one minute) plus lamp warm-up time (from calibration data) plus 30 minute solar soak. For example if the facility calibration demonstrates the lamps reach 850 W/m 2 within two minutes, the soak time can be standardized at 33 minutes after engine off for that facility (f)(10) Air conditioning off test. The air conditioning off test is identical to the steps identified in paragraphs (d)(1) through (9) of this section, except that the air conditioning system and fan speeds are set to complete off or the lowest. It is preferred that the air conditioning off test be conducted sequentially after the air conditioning on test, following a minute soak. We believe this to be a typo, set to complete off or the lowest (setting). Also during this SC03 and HFET portion of the air conditioning off test, (f)(8) requires all windows are rolled up. Provisions need to be made to allow the driver to get sufficient cooling, such as allowing the windows to be partially or fully opened during the air conditioning off portion of the test. The MY regulation currently allows substantial flexibility based on good engineering judgment to limit idle testing to one worst-case vehicle per platform, and carryover data could be used from one year to the next if no changes are made to a platform. Under the current regulation, although initially all platforms would need testing, 11

57 Alliance of Automobile Manufacturers Appendix 3 over time, the regular cadence of vehicle changes over (typical) five-year program lives would mean that only 20% or so of each manufacturer s platforms would need testing in any year. In contrast, the new proposal requires AC17 testing in each year on each platform that receives credit. Beginning in 2017, carryover data is effectively disallowed, since a different sub-configuration within the platform must be tested each year. 7 Also, the technology on and technology off testing effectively doubles the number of tests. Thus, the high test burden from attempting to test every platform every year with the AC17 test exceeds the objective of reasonable verification, since the test burden has grown exponentially from the original idle test. Disallowing carryover data effectively raises the number of tests approximately five-fold, the technology on versus off requirement then doubles the number of tests, and the AC17 test is at least eight times longer than the idle test. Although this is very rough math, it shows that the proposed approach would require approximately 80 times more test hours. The objectives of menu validation and monitoring real-world progress can be achieved with a much lower test burden than this. In view of the complexity of the AC17 test in comparison with the idle test as well as other emission certification tests, no manufacturer should be required to conduct AC17 testing on more than four platforms in any year. Bench Testing SAE procedures have been developed for bench testing of MAC systems at a range of steady-state speeds and for the calculation of the Lifecycle Climate Change Performance of the system using the steady-state results as input data. These methodologies were used for some analyses within the IMAC program. These procedures have strengths as well as weaknesses. The bench test data is accurate and reproducible, although the full battery of SAE tests is very expensive to run (e.g., $80,000 per model), and automobile manufacturers are not currently set up to run these tests. Importantly, the integration of the MAC system with the vehicle is not comprehended in a sophisticated manner by these procedures. Thermal load technologies could not be directly evaluated. As computer controls grow more sophisticated, they have become a major factor in reducing energy consumption of MAC systems, while also meeting acceptable levels of performance in other vehicle parameters. The bench test procedure does not include sophisticated consideration of these computer control algorithms, and it would not be a simple task to include this important variable. The bench test methodology would be no better than the AC17 methodology in achieving the EPA goal of reasonable verification, and it would probably have deficiencies compared to AC17. The bench test methodology was considered within USCAR, but a vehicle test 7 See Supra note 1, at

58 Alliance of Automobile Manufacturers Appendix 3 approach was selected instead because it was more comprehensive and the OEMs had facilities and experienced staffs in place for vehicle testing. Since the AC17 test has shown positive early results, it is preferred over bench testing as the basis for future work on these issues. Once again, even using a bench test approach, the questions to be examined more closely resemble a research program than a traditional vehicle emissions certification program. If necessary, reasonable verification of the menu credit amounts could probably be achieved by a research program using bench test data, and sample-based verification could be used to validate that real improvements were occurring on new vehicles. However, comprehensive testing of every platform or model using this approach would be enormously burdensome, and the bench test approach does not solve the problem of defining baseline performance or standards. Since bench testing offers no clear advantages, we recommend AC17 as the basis for future progress on MAC performance. Conclusion EPA set the stringency of the overall GHG standards based on maximum achievement by the industry of 5.0 g CO2/mile MAC efficiency credits for cars in 2017, followed by maximum achievement of the 7.2 g CO2/mile MAC efficiency credits by trucks in Clearly, this is an ambitious forecast, since it requires that no manufacturer encounter obstacles that preclude achievement of the maximum credit on any of its vehicles. However, the proposed efficiency tests have a high potential to interfere with the achievement of these maximum credit levels. In testing thus far, vehicles with engines below 2.5 liters in displacement have consistently shown only partial achievement of the idle test thresholds. It is not yet clear what achievement levels can be attained on the AC17 test, or how that test may be used. The potential testing burden and/or the planning uncertainties created by these tests may by themselves be sufficient to prevent maximum achievement of these credits. In view of these considerations, we ask that achievement of certain levels on the MAC efficiency tests not be established as a strict requirement in order to gain credits from the MAC technology menu. The credit menu is working, and we expect it to continue to generate significant progress, provided that these test requirements are not allowed to interfere. We will work with EPA and NHTSA to provide reasonable verification of this progress through selected vehicle testing and other methods to show that the menu amounts are not overstated, and that commensurate real-world progress is achieved. The mid-term review and the "check-ins" prior to the mid-term review will provide an opportunity for EPA to review whether a reporting-only AC17 test (instead of an AC17 test with thresholds) continues to be adequate. However, the number of tests to achieve this should be far less than proposed in the NPRM. 8 See Supra note 1, at

59 Alliance of Automobile Manufacturers Appendix 3 NHTSA should adopt regulatory language that provides equivalent levels of MAC efficiency credits in the CAFE program. We note, however, both the difficulty and the importance of keeping the EPA and NHTSA programs aligned. NHTSA CAFE standards will also be based on maximum achievement of these credits, while the ability to earn other compliance credits to offset MAC efficiency shortfalls will not be the same. Finally, we support including MAC credits in fleet averages in a manner consistent with the proposal for off-cycle credits. Credits for Reducing Direct Emissions (Leakage) Use of Updated SAE J2727 The Alliance supports the proposed adoption of the updated SAE J2727 procedure to calculate leak rates. However, at the time the NPRM was written, the updated SAE J2727 was still in draft form. Therefore, not all of the changes have been included in the NPRM. Now that the update is nearly approved, we recommend that it be fully incorporated into the final rule. Using a lower multiplier of 10 (instead of 125) for helium tested connectors makes sense to provide lower leak rates for tighter connectors. The Alliance also proposes that this methodology, along with all other SAE J2727 updates, be allowed for MY so that manufacturers will be encouraged to perform helium testing as soon as possible and develop testing methods in advance. This allowance will benefit the environment and support automakers ability to earn credits to meet the challenging standards of the future. Use of SAE J2064 The Alliance supports the adoption of the SAE J2064 procedure for MY 2017 and beyond for calculating hose leak rates. The Alliance also proposes to allow this calculation method for MYs High Leak Disincentive The Alliance does not support the proposal to reduce the MAC direct credits (up to 1.8 g/mile for cars and 2.1 g/mi for trucks) via the high leak disincentive if the refrigerant leakage rate is not reduced by half from industry average leak rates (in other words, generating debits if the MAC fails to achieve the low leak standards). Compared to the MY regulation, an unreduced program for MAC credits was the basis for the consensus GHG targets set for MY The major reasons why the high leak disincentive should not be implemented for MYs and beyond for the R1234yf refrigerant are: 1) Manufacturers have invested millions of dollars to change the designs of their vehicles and their assembly plants to bear the added cost of the new refrigerant in order to earn these credits. This new proposal to potentially reduce the amount of credit from a switch in refrigerant changes the cost-benefit equation 14

60 Alliance of Automobile Manufacturers Appendix 3 from making this switch. This unfairly penalizes early adopters by undercutting the value derived from the credits from their decisions to switch rapidly to a new refrigerant, when a slower changeover might have been preferred under the new proposed rules. 2) EPA is proposing the high leak disincentive should be implemented for a refrigerant that has a GWP below 150. R1234yf, the new refrigerant that many auto manufacturers plan to use in this timeframe, has a GWP of 4. The global warming impact of R1234yf is times that of R134a. The actual global warming impact of an average leak air conditioner vs. a low leak air conditioner is shown in the table below, where grams/mile of CO2e are calculated based on the delta between average leak and low leak. The data suggests that the actual difference in global warming impact between a vehicle with an average leak and the same vehicle with a low leak air conditioner is g/mile for cars and g/mile for trucks on a CO2e basis using the R1234yf GWP of 4. The high leak disincentives proposed by EPA are 1.8 g/mile for cars and 2.1 g/mile for trucks, which are not in line with the actual global warming impacts. The data below suggests that, assuming manufacturers report CO2 emissions to the nearest tenth of a g/mile so in these terms, excess CO2e emissions from these average leak vehicles will be 0.0 g/mile for cars and 0.0 g/mile for trucks. This analysis shows that the high leak disincentive proposed by EPA has no environmental basis, due to the de minimis environmental impacts of R1234yf leakage. Average vs low refrigerant leak air conditioners impact in real world on cars and trucks Assume average vehicle miles per year.. = 15,000 Average Leak - Low Leak Environmental Impact Leak Rate g/year Passenger Car Tailpipe CO2 g/mi Leak Rate g/year Light Duty Truck Tailpipe CO2 g/mi g CO2e/mile EPA "HiLeakDis" g CO2e/mile 3) Refrigerant leak rates are historically low for modern MAC systems because automobile manufacturers have improved quality in order to meet rising customer expectations and reduce warranty expenses. There is no evidence that manufacturers will be backsliding on these leak rates to save costs. In fact, given the higher cost of R1234yf - up to ten times the price of R134a - manufacturers have an increased incentive to further reduce leaks and thereby retain the expensive new refrigerant. 15

61 Alliance of Automobile Manufacturers Appendix 3 4) EPA has expressed concern about refilling R134a refrigerant in place of the R1234yf refrigerant. However, this is not easily achieved, since the service port fittings installed on new vehicles with R1234yf refrigerant are totally different from those using R134a. 5) It is also worth noting that the size of the proposed penalty greatly exceeds the actual scale of the GHG impact of potential leakage. Please also note that the closing bracket is placed at the wrong place in the formulas that calculate direct MAC credits in section (b)(2)(i) & (ii), both for cars and trucks. We recommend that EPA make the correction of these closing brackets (as noted below) along with removing the HiLeakDis factor in both formulas. Refrigerant Level Monitoring Emission warranty requirements are not appropriate for mobile air conditioners under the proposed rule. This is because in-use performance of MAC systems at levels comparable to a new vehicle is not needed to achieve the emission levels targeted by EPA. Warranty requirements were established for tailpipe pollutants, such as CO and NOx, because emissions of those pollutants would rise significantly if the pollution control devices such as catalytic converters fail. This would typically not be the case for MAC components. First, consider the case of indirect emissions from fuel consumed to power the MAC. In the vast majority of MAC failure modes, the system stops cooling and ceases operation - either because the critical moving parts stop moving or because the system is switched off - thereby actually reducing the indirect CO2 emissions. Emission warranties should not be required in relation to the indirect MAC emissions. The most significant item in EPA s proposed warranty coverage, the compressor, can cost over $1,000 to replace. It seems paradoxical and disproportionate to impose such high costs in an emissions recall scenario to replace this component, and thereby actually increase indirect emissions. Although manufacturer warranties may typically already be longer than the two-year period proposed by EPA in this NPRM, in principle there is no sound basis for emission warranty coverage to safeguard indirect emission levels, since indirect emissions go down when the system fails. Finally, it is worth noting that proper 16

62 Alliance of Automobile Manufacturers Appendix 3 functioning of these parts is not actually required to achieve the emissions levels set by EPA. Regarding direct emissions of refrigerant, there is only a negligible environmental impact if refrigerants below a GWP of 150 are released from the system, even if the entire charge (typically between 1-2 pounds) is released. Therefore, emission warranty coverage of joints, hoses, seals, etc. is certainly not needed to protect the environmental gains from application of low-gwp refrigerants. While the ultimate cost of the new low-gwp refrigerant R1234yf (also known as HFO-1234yf)) is higher than the R134a, it is expected to be at a level that would severely discourage motorists from repeatedly recharging a system with significant unrepaired leaks (e.g., any cost of over $30 per pound). Therefore, there is no emission-based reason to mandate warranty coverage to prevent leaks on low- GWP systems, and the potential costs of an emission recall would be disproportionate to any environmental impact of leakage of these refrigerants. Any emission warranty requirements should specifically exclude emission warranty coverage for systems using a refrigerant with a GWP below 150. This is consistent with EPA s position that no emissions warranty is required for zero emissions vehicles. The sole remaining MAC environmental impact would be from refrigerant leakage in the current R134a systems. Given the prospect for fairly rapid adoption of the low-gwp refrigerants in new vehicles during the time frame of this regulation, this would appear to be a very small basis on which to create an entirely new area of emissions warranty coverage and all the associated elements of an in-use program for air conditioners. EPA should not create a program of warranty coverage for MAC components in pursuit of such a small and temporary emissions impact. In conclusion, a properly structured MAC credit program can provide substantial low cost, near term GHG reductions. Our recommendations are provided to make the MAC program work as effectively as possible. Implementation of New Alternative Refrigerants As part of the Single National Program, the current regulations provide incentives to manufacturers to implement low-gwp refrigerants and reduce system leakage. Our members plan to make use of these incentives, with the degree of use depending on each manufacturer's model changeover plans, MAC technology implementation plans and capital investment schedules. In fact, we expect some automobile manufacturers to begin use of R1234yf on some models as early as Despite this promising news, there remain prohibitive barriers to achieving 100% use of low-gwp refrigerants. At this juncture, it would be premature for EPA to remove R134a from the list of acceptable substitutes for CFC-12 in MAC systems. We believe that the key to this transition is to instead continue the credits available under the MY National Program for regulation of light-duty vehicle GHG and fuel economy. This approach would help ease the transition and encourage earlier action, to the extent that such action is achievable and cost-effective. It would encourage low-leak R134a systems, so long as R134a systems are allowed on new vehicles. It would also encourage accelerated introduction of new 17

63 Alliance of Automobile Manufacturers Appendix 3 refrigerants if manufacturers were awarded credit for any usage of the new refrigerants in excess of mandated requirements. Should R134a become prohibited and R1234yf become mandatory for all vehicle MAC systems, it is critical that the MAC credits continue to apply throughout MY , since these credits are an essential part of manufacturers compliance plans. Refrigerant Availability Based on current knowledge, the best product that could meet a low-gwp requirement (at or below 150 GWP) would be R1234yf. This refrigerant is not yet available in commercial quantities, and it is unclear when a sufficient supply will be available in for the U.S. market or what the cost of this product will be. The single manufacturer of R1234yf a joint venture of Honeywell and DuPont recently announced that it would begin supplying the refrigerant in commercial quantities in the fourth quarter of According to current information, once production is permitted at pilot facilities at both Honeywell and DuPont, as well as an intermediate-scale facility in China at DuPont affiliate Changshu 3F Zhonghao New Chemical Materials, supplies likely will meet near-term EU requirements. 9 A single world-scale plant is expected to follow at an undefined later date as demand grows. Current EU regulations mandate that automakers switch to low-gwp refrigerants by This EU mandate was phased in over a seven year period beginning in EU sales of nine million MAC-equipped vehicles filled with 600 grams of refrigerant equate to 5400 metric tons of refrigerant required for new production. Service and repair will require additional refrigerant. Since these regulations are already in place, the ramp-up of R1234yf production and system design is expected to fill the EU market first. The supply challenge is exacerbated by the fact that Honeywell and DuPont retain exclusive rights to the manufacture of R1234yf, thereby limiting the opportunity for other chemical manufacturers to supply the future demand for added manufacturing capacity. Engineering Resources Needed to Transition to R1234yf In evaluating this issue, the agencies should consider both the availability of the new refrigerant and the significant automaker resources needed to provide the engineering, logistics, training and roll-out. The deployment of these changes over 100% of vehicle models will present considerable challenge. In particular, all new systems will require complete revalidation using the new lubricants required for R1234yf. This will take time and strain the engineering resources throughout the MAC industry. 9 American Chemical Society, Chemical & Engineering News, July 26, 2010: 18

64 Alliance of Automobile Manufacturers Appendix 3 In most cases, manufacturers will implement R1234yf with a revised component layout, including additional components that create packaging problems. Therefore, R1234yf is ideally implemented during vehicle major redesigns. Finally, there are significant changes to the assembly plants that are required to handle R1234yf (especially OSHA rules). These typically include relocation of the refrigerant charging area and relocation of refrigerant storage tanks. Extensive plant rearrangements such as this are very disruptive to plant operations, and are therefore typically performed as part of the changeover that occurs when new models are introduced, when extensive replacement of tooling and revised layout of production lines is typically required throughout the plant. Substituting R1234yf for R134a is fundamentally more complex than the change from R12 to R134a that was made during the 1990's, and the SNAP usage requirements are more significant. The changes in the assembly plants will also be much more significant than were the case for R134a. For these reasons, EPA should not delist R134a from the approved SNAP list of automobile air conditioner refrigerants, nor should EPA establish other policies based on the assumption that a comprehensive changeover to new refrigerants can occur within the foreseeable future. 19

65 Alliance of Automobile Manufacturers Appendix 4 Credits for Dual Fuel E85 Gasoline Vehicles and Other Alternative Vehicles Table of Contents DUAL FUEL E85 GASOLINE VEHICLES... 1 INTRODUCTION... 1 VOLUME OF ETHANOL USED IN MOTOR VEHICLES... 1 VOLUME OF ETHANOL AVAILABLE TO THE FLEX-FUEL FLEET... 2 VOLUME OF E85 AVAILABLE TO THE FLEX-FUEL FLEET... 2 FLEET OF ACTIVE FLEX-FUEL CARS AND TRUCKS BY MODEL YEAR... 2 MILES TRAVELED BY THE FLEET OF ACTIVE FLEX-FUEL CARS AND TRUCKS... 2 TOTAL GRAMS OF CO 2 EMITTED BY THE FLEET OF ACTIVE FLEX-FUEL CARS AND TRUCKS... 3 CO 2 EMISSIONS OF THE FLEX-FUEL FLEET ON E PROPORTION OF THE FUEL USED BY THE FLEX-FUEL VEHICLE FLEET THAT IS E CREDITS FOR OTHER ALTERNATIVE FUEL VEHICLES... 3 DUAL FUEL PLUG-IN HYBRID ELECTRIC VEHICLES (PHEVS)... 3 DUAL-FUEL CNG AND LPG GASOLINE VEHICLES... 4

66 Alliance of Automobile Manufacturers Appendix 4 Dual Fuel E85 Gasoline Vehicles Introduction In the final rulemaking for MY , EPA created regulations for MY 2016 ethanol flex-fuel vehicles (FFVs) that differed significantly from those provided for by EPCA. EPA ended the GHG emissions compliance incentives and adopted a methodology based on demonstrated vehicle emissions performance. For MY 2016, EPA proposed awarding CO2 credits upon demonstration of actual usage of E85. EPA now proposes extending MY 2016 approach to MYs In the MY rulemaking, EPA offered two options for automobile manufacturers to consider: (1) a default system based on 100% gasoline operation and (2) fuel economy weightings on national E85 use, or on manufacturer-specific data showing the percentage of miles that are driven on E85 versus gasoline for that manufacturer s ethanol FFVs. The Alliance supports the determination of CO2 credits based on national E85 usage. The idea of actual national usage would be in conjunction with an early issuance of guidance to manufacturers indicating the value of the F-factor so that manufacturers can develop their vehicle portfolios and GHG compliance plans. The F factor is used in the calculation of Carbon Related Exhaust Emissions (CREE) of flex-fuel vehicle and represents the relative usage of gasoline and E85. The calculation of F needs to take into account the following: The volume of ethanol used in motor vehicles. The volume of ethanol available to the flex-fuel fleet. The volume of E85 available to the flex-fuel fleet. The fleet of active flex-fuel cars and trucks by model year. The miles traveled by the fleet of active flex-fuel cars and trucks. The total grams of CO2 emitted by the fleet of active flex-fuel cars and trucks. The CO2 emissions of the flex-fuel fleet on E85. The proportion of the fuel used by the flex-fuel vehicle fleet that is E85. Volume of Ethanol Used in Motor Vehicles EISA makes clear the volumes of renewable fuels that are to be used in the United States through 2022 and then maintains this proportion in 2023 and beyond. For example, the total renewable fuel requirement for 2016 is billion ethanol equivalent gallons. EISA does not give guidance as to whether the fuels marketed or imported to meet these requirements will be ethanol, biodiesel or some other renewable fuel. Our recommendation is to use the Energy Information Agency Annual Energy Outlook Liquid Fuels Supply and Disposition forecasts volumes for ethanol, biodiesel and other biomassderived liquids. The ratios between the three fuels can be used to determine the amount of ethanol to be used in any given year. For example, in 2016, the 2011 reference case forecast is million gallons of ethanol, million gallons of biodiesel and 307 1

67 Alliance of Automobile Manufacturers Appendix 4 million gallons of other biomass-derived liquids. This gives a ratio of 9.54 gallons of ethanol per gallon of other fuels. Assuming that both the biodiesel and other biomassderived liquids have a RIN value of 1.5 and the ethanol has a RIN value of 1.0. The volume of ethanol required = 9.54 gallons of ethanol per RINs ( ) multiplied by billion RINs or billion gallons of ethanol. The corresponding volume of biodiesel plus other biomass-derived liquids would be 2.01 billion gallons. Volume of Ethanol Available to the Flex-Fuel Fleet This is determined by subtracting the ethanol volumes used for gasoline blending from the total volume of ethanol. This is done by determining the total volume of hydrocarbons used in motor gasoline (this can be determined using the EIA AEO Liquid Fuels Supply and Disposition forecasts volumes) and dividing by 9 to determine the ethanol used for blending E10. For 2016, this is billion gallons of ethanol used in gasoline blending. When this is subtracted from the billion gallons of ethanol to be used, the resulting volume of ethanol to be used in the flex-fuel fleet is 4.97 billion gallons. Volume of E85 Available to the Flex-Fuel Fleet The volume of E85 available to the flex-fuel fleet is determined by dividing the volume of ethanol to be used by the flex-fuel fleet by the fractional ethanol content of the E85 certification fuel for the model year being evaluated (currently 0.85). Fleet of Active Flex-fuel Cars and Trucks by Model Year This can be done by obtaining vehicle registration data from a commercial company such as R. L. Polk and screening for flex-fuel vehicles. Alternatively, reported manufacturer FFV production by model year can be used, with the volumes reduced using vehicle survival rates from Table 4-3, Survival Rates and Unadjusted Annual Miles Traveled (VMT) by Age for Passenger Cars and Table 4-4 Survival Rates and Unadjusted Annual Vehicle-Miles Traveled (VMT) by Age for Light Trucks from Final Rulemaking to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards Joint Technical Support Document, April 2010 (2010 TSD). Miles Traveled by the Fleet of Active Flex-fuel Cars and Trucks Multiply the number of flex-fuel cars and trucks of each model year by the appropriate miles traveled per year for each model years age obtained from Table 4-3, Survival Rates and Unadjusted Annual Miles Traveled (VMT) by Age for Passenger Cars and Table 4-4 Survival Rates and Unadjusted Annual Vehicle-Miles Traveled (VMT) by Age for Light Trucks from the 2010 TSD. 2

68 Alliance of Automobile Manufacturers Appendix 4 Total Grams of CO2 Emitted by the Fleet of Active Flex-fuel Cars and Trucks Convert the average fuel efficiencies of U.S. light-duty vehicles (available from the Research and Innovative Technology Administration, Bureau of Transportation Statistics, table 4-23) 1 to grams of CO2 per mile using the conversion factor of 8,887 grams CO2 per gallon of gasoline. For any model year too recent to be included in the Bureau of Transportation Statistics table, use the projected fleet-wide emissions compliance levels under the MY or MY final rules, as appropriate. Multiply the vehicle miles traveled for the cars and trucks of each model year by the CO2 emissions per mile of the cars and trucks of each model year. Sum the CO2 emissions for each model year of cars and trucks on the road. CO2 Emissions of the Flex-Fuel Fleet on E85 Multiply the E85 volume available by the grams of CO2 per gallon of E85 certification fuel for the model year being evaluated (currently 6,295). Proportion of the Fuel Used by the Flex-fuel Vehicle Fleet That is E85 Divide the CO2 emissions of the flex-fuel fleet on E85 by the total CO2 emissions of the flexfuel fleet. This fraction is the value, F, used in the CREE calculation. Credits for Other Alternative Fuel Vehicles In the effort to continue the development of advanced technology vehicles, the Alliance would like to show support of the following technologies, which will help drive our country down the road toward energy independence. Dual Fuel Plug-In Hybrid Electric Vehicles (PHEVs) The Alliance supports the continued use of the Society of Automotive Engineers (SAE) cycle-specific utility factor approach for PHEV compliance and label emissions calculations. This utility factor approach provides a method for predicting the fractions of total distance driven in each mode of operation. In this case the modes of operation would be wall electricity from the grid or conventional liquid fuel such as gasoline

69 Alliance of Automobile Manufacturers Appendix 4 Dual-Fuel CNG and LPG Gasoline Vehicles CNG and LPG vehicles are another option that our country has to diversify the vehicle fleet and use a domestically available energy source. The Alliance supports the development of a utility factor approach very similar to the SAE standard mentioned above for PHEVs. The Alliance is also in favor of the option to allow manufacturers to use the proposed utility factor-based methodology as a pull-ahead option for MYs Based on the added cost of the vehicle technology and the cost advantage of using CNG and LPG fuel relative to gasoline, customers that purchase a dual-fuel CNG or LPG vehicle will, to the extent possible, use the intended alternative fuel. Many companies may leverage global designs in developing dual-fuel CNG and LPG vehicles for the U.S. market. It is important that the variety of global design features available be allowed into the U.S. market. Rather than making specific design requirements in the rules, a better approach would be have these design features be factors in the calculation of the CNG and LPG utility factors. The Alliance would like to propose a work group to discuss the constraints mentioned in the NPRM for dual-fuel CNG and LPG vehicles. In the NPRM, EPA specifically requested comments on the merits of providing sales multiplier (similar to the EV/PHEV incentives) for dedicated and/or dual-fuel compressed natural gas vehicles. The Alliance believes CNG and LPG technology also deserve multipliers. 4

70 Alliance of Automobile Manufacturers Appendix 5 Compliance with N2O Requirements Table of Contents NITROUS OXIDE (N 2O) MEASUREMENT... 1 N 2O DATA... 4 N 2O CERTIFICATION... 4 IN-USE N 2O... 5 HEAVY DUTY N 2O... 5

71 Alliance of Automobile Manufacturers Appendix 5 Nitrous Oxide (N2O) Measurement EPA has recognized the difficulties and complexities of evaluating, procuring and installing the equipment that would be needed to measure N2O and has proposed that manufacturers be permitted to use compliance statements in lieu of test data through MY However, as explained below, EPA has not provided sufficient time for manufacturers to incorporate accurate and robust N2O measurement capabilities into their test sites. We propose that the deadline for measuring N2O be extended until the measurement issues are resolved. The N2O measurement capabilities should be reevaluated during both the mid-term evaluation of standards and the check-ins occurring prior to the mid-term. By so doing, EPA would be providing manufacturers sufficient time to evaluate appropriate test equipment and would be aligning possible N2O regulatory changes with possible subsequent changes to other light duty GHG regulations. The first issue with regard to N2O measurement timing is that there is currently no accurate measurement technology available that is suitable for high-volume testing. As one example, the gas chromatograph electron capture detector (GC-ECD) is not suitable for high-volume testing since it includes an off-line multi-hour long analysis and has robustness issues. EPA provided a technical study of the capabilities of currently available and potentially available future measurement technologies as a separate memorandum to the docket. 1 The study compares instruments by analyzing ambient air and diluted vehicle exhaust samples on a number of different vehicles and schedules (FTP and HFET). The EPA technical study highlights the continuing difficulties of the currently available measurement technologies to accurately measure N2O. In the study, EPA compared the Fourier Transform Infrared Spectrometer (FTIR) to the GC- ECD and concluded that the FTIR compared very well to GC-ECD, which is considered the gold standard, but the accompanying data actually demonstrates the opposite. The data shows significant differences and variability between these two instruments on the order of approximately -17 to +25 ppb N2O equivalent for ambient air analysis and approximately -27 to +85 ppb N2O equivalent for vehicle exhaust testing. These differences represent a significant error at an N2O standard level of g/mi. In addition, in comparing the FTIR to the Non-Dispersive Infrared (NDIR) analyzer, the study states Both the NDIR and FTIR analyzers performed well, however some questions regarding performance remain. The accompanying data, however, shows significant interferences with both analyzers, results which are similar to those of the previously supplied Alliance Technical Study (June 2011) Fed. Reg , 74994, footnote 239 (Dec. 1, 2011) (referring to Data from the evaluation of instruments that measure Nitrous Oxide (N 2O), Memorandum from Chris Laroo to Docket EPA HQ OAR , October 31, 2011). 1

72 Alliance of Automobile Manufacturers Appendix 5 The study also highlights a potentially promising new N2O measurement technology that is based on laser spectroscopy and is made by a few manufacturers. However, N2O analysis is so new that most of these instruments are still in the development stages and hence are prototypes. Although these instruments show promise for N2O analysis, questions remain as to their accuracy and robustness (i.e., reliability) at such low N2O standards. The study contains evaluations of two such laser instruments based on simulated exhaust gas (water, carbon monoxide (CO), CO2 and N2O). In comparison to NDIR and FTIR, the data for the first laser instrument shows interference errors that were typically lower, ranging from -29 to +10 ppb N2O equivalent. Although the errors appear to be reduced, they still represent a significant portion of the 10 mg/mi N2O standard. CO and CO2 measurements were likewise affected by interference gases but in the opposite direction, i.e., higher levels than that observed with N2O. For the first laser instrument the study concludes that the instrument performed very well and does not appear to show any susceptibility to CO, CO2, or water interference and Based on CO and CO2 measurement error, we believe that the bulk of the associated N2O measurement error is due to bag blending error. It would seem that if the N2O errors were due to gas blending errors, then CO and CO2 would be likewise affected, but the data generally shows an underreporting of N2O and an over reporting of CO and CO2. Looking at this limited data, the Alliance believes that although the instrument shows promise, it still demonstrates significant measurement errors which have not yet been accounted for. The second laser instrument evaluation showed similar N2O measurement errors in the presence of interference gases but this was attributed to the inability to properly zero/span the instrument after the initial zero/span at the start of the testing. In the conclusions it is stated that EPA intends to re-evaluate this instrument after the manufacturer has resolved issues Suffice it to say that until more studies are conducted by multiple facilities, including correlation vehicle testing between facilities, the true accuracy of laser based instruments is still to be determined. The second issue with regard to N2O measurement is emission development and certification timing. Taking all things into account, we estimate that it will take approximately 4.5 years to properly install a new N2O analyzer into a single test site. Below is a graphical representation of the estimated timeline for instrument procurement and installation. 2

73 Alliance of Automobile Manufacturers Appendix 5 3