3

Light-Duty Vehicle Technologies and Fuels

INTRODUCTION

This chapter covers the technology areas funded by the U. S. Department of Energy (DOE) Vehicle Technologies Office (VTO) and Fuel Cell Technologies Office (FCTO), which support U.S. DRIVE (Driving Research and Innovation for Vehicle Efficiency and Energy Sustainability) goals and targets. These topics are advanced combustion engines, fuels and emissions, fuel cells and hydrogen (onboard storage, production delivery and distribution, safety codes and standards), electric propulsion (electric drive systems, power electronics, electrochemical energy storage, electricity as an energy source, and the grid interactions), structural materials, and cradle-to-grave (C2G) analysis and implications.

The task of the committee is not to provide a detailed technical review of each of the projects funded by DOE, which are those that contribute to U.S. DRIVE Partnership goals, as this is accomplished through the DOE Annual Merit Review (AMR) meeting, but rather to review progress and program directions and an assessment of ongoing research activities, as well as how effective the Partnership is in identifying precompetitive research and development (R&D) activities that lead to the achievement of Partnership goals.

To this end, each of the subsections addressed in this chapter provides a brief background of the technology and its importance to the goals of the Partnership; a current status vis-à-vis Partnership goals, targets, and timetables; an assessment of progress and key achievements and program directions; a DOE budget overview; significant barriers and issues that need to be addressed; the response to recommendations from the National Resource Council (NRC) Phase 4 review (NRC, 2013); the appropriate federal role; and findings and recommendations (see Chapter 1 for the full statement of task).

Internal combustion engines (ICEs) have been the dominant vehicle power plant since the inception of mass-produced personal transportation and can in many respects be considered a mature technology. Yet there is still potential for their improvement in terms of lower tailpipe emissions, reduced fuel consumption, and a lower greenhouse gas (GHG) total footprint. Given the number of vehicles in this class, activities related to advanced combustion technologies of vehicles using liquid fuels and powered by internal combustion engines will be a major part of DOE’s technology pathway to achieving an 83 percent reduction in GHG emissions by 2050 (Sarkar, 2016). Efforts to reduce the fuel consumption of ICE power trains are directed at improving the engines’ peak efficiency and being able to achieve that maximum efficiency over the engines’ entire operational range.

In order to fully meet the Partnership goals and current and future standards, additional approaches will be required. Electric drive vehicles, while still a small part of the market, have emerged as strong candidates for meeting these requirements, provided that a green electric grid evolves in parallel. The introduction of hybrid electric vehicles (HEVs) in 1999 initiated a significant move toward vehicle electrification, with many HEV options now commercially available. Advances in battery technology have enabled the introduction of plug-in hybrid electric vehicles (PHEVs) and battery electric vehicles (BEVs), which reduce GHG emissions even further. As new battery materials evolve, greater range, reduced charging times, and reduced cost may be achieved. Hydrogen fuel cell vehicles (HFCVs) provide an attractive electric alternative that addresses both refueling time and range but are currently limited by cost and the lack of a hydrogen refueling infrastructure.

It is worth noting that during this review Toyota, Hyundai, and Honda have made available within the United States a limited number of fuel cell vehicle sales or leases to the general public. General Motors, a U.S. DRIVE Partnership member, has reported they plan on a 2020 rollout of their latest fuel cell vehicle, following years of extensive and successful road testing of their first-generation HFCV, the Equinox. The development and deployment of roadworthy HFCVs is a major accomplishment and one that will help to identify remaining technical, cost, manufacturing, and infrastructure challenges. Though the cars are still in the late stages of development, the fact that these cars have advanced to this point is due in part to R&D coordination by the Partnership and its prior organizations, as well as from decades of funding of pertinent research projects by the DOE and Partnership members.

The lack of a hydrogen infrastructure remains a significant challenge that will impact the degree of acceptance and success of the fuel cell electric vehicle, especially in the early years. This not only includes production and delivery of hydrogen but also cost on a gallon of gas equivalent (GGE) basis, if such vehicles are to be competitive with other electric vehicles as well as current internal combustion technologies. Furthermore, the generation of hydrogen from renewable sources must continue to be researched, developed, and ultimately adopted if the

HFCV GHG emissions are to be reduced below other vehicle technologies. The near-zero emissions potential when an HFCV is fueled with hydrogen generated from renewable energy sources would be a significant breakthrough in the quest to limit GHG emissions from the transportation sector.

Finally, the deployment of lightweighting materials and components and more efficient electric power trains will contribute to increased vehicle range and decreased emissions, both being key goals of the Partnership.

ENGINES AND EMISSION CONTROLS

Introduction

Regulations and future targets for GHG emissions have instilled new urgency to reducing the fuel consumption and total GHG footprints of the nation’s vehicle systems. This is spurring further development of ICEs, their associated power trains, aftertreatment systems, and the characteristics of the fuels used to carry the energy into the engine. The advanced combustion and emissions control technical team (ACECTT) guides the effort of improving the performance of internal combustion engines, fuels, and aftertreatment systems for the U.S. DRIVE Partnership.

In assessing the relevance and accomplishments of the engine, fuel, and aftertreatment research programs within the ACECTT portfolio, it is important to understand that the improvement path for efficiency in ICEs is not without limits. There are fundamental and practical limits to the efficiencies that can be obtained, just as there are for fuel cell and battery electric systems. In March 2010 DOE hosted a colloquium on Transportation Combustion Engine Efficiency at USCAR to discuss these limits (Daw et al., 2010). The purpose of the colloquium was to engage academics, researchers, and developers in a review of the underlying thermodynamics applicable to ICEs, to identify the unavoidable losses (irreversibilities, i.e., exergy destruction) that must be accepted when producing power from chemically reacting systems, and to opine as to what might be maximum practical efficiencies achievable with internal combustion engines.

Although a definitive number was not agreed upon at the colloquium, the outcome of the meeting indeed did provide stretch goals presented to the technical community—for example, the targets set for the Super Truck programs. For heavy-duty diesel engines, brake thermal efficiencies on the order of 55 percent are in the vicinity of maximum practical efficiencies for a stand-alone ICE. Peak engine efficiencies are typically achieved at high-load operating conditions. For smaller engines, because of less favorable geometric characteristics such as surface area to volume ratios and friction factors, the peak efficiency is less than for larger engines; however, as with the larger engines the highest efficiencies are achieved at high load. In general the efficiency of the engine decreases as its operating condition moves away from high loads into lower-load operating regimes.

Despite the fact that some engines are approaching their practical maximum efficiencies, there is still room for improvement. Big large-bore, long-stroke engines have peak efficiencies that indeed are very near their practical limits. There is little opportunity to improve their peak efficiencies further. Engines used in light-duty vehicles will have lower practical maximum efficiencies than big engines, and the current values of their peak efficiencies are not as close to their practical maximums. The opportunities for improvements in smaller engines lie in activities to push their peak efficiencies closer to their practical maximum limits and efforts to achieve these higher efficiencies over a large portion of the engine operating regime.

The ACECTT has set program targets of efficiency gains of 20 percent by 2020 relative to a 2010 baseline engine when operating over a prescribed set of operating conditions. Since the efficiency typically drops as the operating condition moves away from the point of maximum efficiency—that is, moves to lower loads—there is much that can be done to improve the overall cycle efficiency. If the high-load efficiency of the engine can be replicated at lower loads, the overall cycle efficiency of the vehicle will improve, which would result in significant reductions in fuel consumption and emissions. Furthermore, as engine evolution takes place, changes in combustion processes and technologies for overcoming practical constraints in these smaller engines—like being able to operate knock free at higher compression ratios—will improve the peak engine efficiency. These improvements in peak efficiency could then be promulgated over the entire operating cycle of the engine and act as an additional multiplier to improvements to the overall cycle efficiency. Incorporating technologies that accomplish these goals (ca. 2020) would occur via model year changes of vehicles and would thus penetrate the market at the rate of the vehicle fleet turnover. Given the magnitude of the vehicle fleet powered by ICEs, fuel consumption and emission reductions introduced through ICE improvements would be multiplied by millions of vehicles per year, so the reduction in fuel consumption and emissions would be significant.

As a transportation power plant the ICE is ideal for applications that demand sustained high power operation and long times or distances between refueling. However, a major challenge in using ICEs for mobility applications is that typical mobility duty cycles contain significant intermediate- and light-load operation, where the engine’s efficiency drops to levels that can be considerably below its peak value.

Consequently, efforts to reduce the fuel consumption of ICE power trains are directed at improving the engines’ peak efficiency, being able to achieve that maximum efficiency over the engines’ entire operational range, and reducing the auxiliary, accessory, and friction loads that detract from the power delivered to the wheels. However, even the most efficient engine will still need exhaust gas aftertreatment to ensure that the criteria tailpipe pollutants are below regulated limits over the entire vehicle operating domain. So, the engine and the aftertreat-

ment components must be developed as a system to achieve optimal performance. Accomplishing this requires understanding all the energy flows, transformations, and work-potential dissipations (exergy destruction) that occur in the power train—that is, starting from the fuel leaving the fuel tank to the power being delivered to the wheels and then guaranteeing that the aftertreatment system can function effectively at the exhaust conditions leaving the engine cylinder. Achieving improvements in these areas pushes the boundaries of the current understanding of almost all the physical and engineering sciences.

Pursuing such improvements is the focus of the engine combustion and aftertreatment research community. It is manifest in such activities as hybridization of the power plants, downsized boosted engines, variable valve actuation, cylinder deactivation, and advanced combustion processes. In each of these applications the technical community is working both to increase the peak efficiency and to bring high-load engine efficiencies to vehicle operation that does not require high load engine operation.

As the engine efficiency is increased more of the fuel energy is converted to work and there is less energy leaving the engine in the exhaust stream. The lower energy content of the exhaust thermodynamically translates into lower exhaust gas temperatures. This presents an additional challenge for the exhaust gas aftertreatment system. Not only is there need for higher effectiveness of the aftertreatment systems to reduce criteria pollutants, but also the aftertreatment systems need to achieve this effectiveness at lower temperatures.

In addition, the fuel cycle needs to be part of the focus. Biomass-derived fuels, either as blends or drop-in components, have the potential to reduce total life cycle (also referred to as C2G) GHG emissions. Research is under way to explore potential synergies between fuel refining processes and fuel characteristics selected to enhance the combustion process while trying to achieve optimal C2G GHG emissions for the system as a whole. For example, higher octane number fuels would allow engines to have higher compression ratios, which improves their efficiency and could eliminate the need for spark retard (a control strategy to avoid engine knocking that is detrimental to efficiency). However, the extent to which this can be done, and the implications in the trade-offs in GHG emissions between the additional processing and potential costs necessary to achieve the higher octane number during the fuel production versus the lower fuel consumption achieved during engine operation, needs to be determined.

Fuels will also play an important role in achieving the 2050 carbon dioxide (CO₂) reduction targets for the light-duty vehicle fleet by reducing the carbon footprint of the energy carrier itself. The advanced engine and combustion strategies under investigation by the ACECTT will enable significant CO₂ emission reduction; however, without an accompanying effort to reduce the carbon in the fuel it is unlikely that the 2050 CO₂ target can be met (Farrell, 2016). DOE started an initiative on co-optimization of fuels and engines or “Co-Optima” (DOE, 2016c) in fiscal year (FY) 2016 to address these challenges. Although the co-optimization

initiative is a separate DOE effort from the U.S. DRIVE Partnership, it is providing funding support for DOE research projects that are considered as supporting the goals of U.S. DRIVE.

ACECTT Research

The U.S. DRIVE ACECTT mission statement reads as follows:

Their activities focus on continued improvement of spark-ignition (SI) and compression-ignition engines working in conjunction with more effective aftertreatment systems, while continuing the development of more advanced, kinetically controlled, low-temperature combustion engine concepts.

The ACECTT activities are fundamentally dissecting and analyzing the energy flows within the engine-aftertreatment-power train system to address every energy transformation that occurs within the vehicle. This includes the energy flow in the fuel leaving the fuel tank, through its introduction into the cylinder, during the combustion and work extraction processes, through the aftertreatment system, and out the tailpipe. Attention is also directed at all energy flows that use work generated by the engine for functions other than driving the wheels—such as accessories, pumping and friction; or energy flows that leave the engine in a form other than work, such as heat transfer and exhaust flow. Maximizing efficiency and minimizing emissions over the operating map of the vehicle is a complex challenge that requires detailed understanding of the fundamentals that govern these energy transformations, the degradations of the work-potential of the energy associated with these transformations—that is, exergy destruction—as well as being able to measure them via sensors and developing advanced control systems to optimize the energy management for the entire system. As an enhanced understanding of these energy flows, and their associated exergy destruction, is developed, it enables researchers to determine which irreversibilities can practically be pursued for reduction, and the extent to which efficiency improvements are possible by perfection of the energy transformation in question. For example, the following have all been important in directing research and technology development activities: understanding the magnitude of the work that is lost due to the chemical reactions that release the chemical energy in the fuel, and realizing that this loss cannot be prevented with current technologies; or understanding the relative importance of the heat transfer from the engine during different portions of the combustion process; or being able to calculate the work that might be obtained by exhaust energy recovery systems; or understanding the trade-offs of different exergy destruction processes that lead to improved engine efficiency by keeping combustion temperatures low. These activities are made especially challenging because to

gain the maximum benefit they need to be understood accurately for transient operation.

To address the interactions between fuels and the engine system U.S. DRIVE has established a Fuels Working Group (FWG). A working group is less formal than a technical team. It does not have a roadmap or defined technical targets but instead offers structure for technical experts to discuss topics of interest. The ACECTT, interfacing with the FWG, has sharpened the focus of fuels testing within the ACECTT engine testing programs. The ACECTT, the FWG, and USCAR have developed a well-to-wheels study fuel set that identifies fuel formulations with a range of research octane numbers (RONs), octane sensitivities, and biofuel energy content that might be representative of fuels that could be made available in the near term (Farenback-Brateman et al., 2016). These fuels now serve as a common base for the experimental spark-ignition ICE programs within the ACECTT portfolio. The activities promoted within the ACECTT by the FWG are closely aligned with the objectives of the newly formed DOE Co-Optima initiative.

The ACECTT understands that advanced spark- and compression-ignition engines will be dominant in the near-term mobility fleet. The potential of advanced low-temperature combustion systems, also referred to by the ACECTT as chemical kinetics-dominated combustion, has been demonstrated in the laboratory (Ra et al., 2012); however, in the February 2016 committee meeting the ACECTT indicated that chemical kinetics–dominated combustion was not considered a near-term technology. The technical team expects that the engine types used during the time period of focus for this program will be a mix of naturally aspirated hybrids and advanced downsized boosted architectures. That is, the ACECTT expects that in the near term the engines will have conventional four-stroke architectures with improved but relatively conventional spark-ignited flame propagation or diesel–type autoignition combustion systems. The ACECTT specifically did not suggest dates for near- and long-term time frames. Furthermore, the ACECTT believes that within each of these engine pathways, the combustion strategies will be one of the following:

• Spark ignition, where combustion is flame propagation dominated and dependent on premixing of the air and fuel prior to flame initiation;
• Compression ignition, where combustion processes are mixing and diffusion dominated; or
• Chemical kinetics dominated, low-temperature combustion (LTC).

Each of these combustion strategies depends on the characteristics of the fuels being used. Fuel characteristics such as resistance or propensity to autoignite and how that propensity changes with the conditions in the cylinder, burning velocity, tendency to form particulate matter, and heat of vaporization will impact the effectiveness of the combustion strategy in converting the energy in the fuel to

power out of the engine. Through the Fuels Working Group, the ACECTT is also attempting to identify optimal fuel characteristics for the different combustion strategies and determining if the introduction of such fuels could be easily integrated with the current fuel infrastructure and legacy fleets. These research issues are also part of the DOE Co-Optima initiative; these are areas where Co-Optima funding is being directed toward projects that support U.S. DRIVE goals.

The ACECTT has identified specific strategies and research targets to promote progress toward their goals. For premixed flame dominated (SI) engines, pursuing dilute combustion is recommended. Combustion in near-term dilute SI engines will be dominated by the propagation of a flame front through reactants that are largely premixed. The dilution of the charge, either with excess air (lean) or with exhaust gas recirculation (EGR) promotes higher efficiency because the combustion temperatures are lower than those occurring in undiluted, stoichiometric¹ spark-ignited combustion. The lower combustion temperatures result in a ratio of specific heats, gamma, that is higher than would occur if the gas temperatures were higher. Thermodynamically this leads to a higher work extraction per unit of piston motion during the expansion stroke, as well as lower heat transfer and exhaust enthalpy losses. It is also beneficial because the cylinder-out oxides of nitrogen (NO_x) emissions are lower owing to the lower temperatures, so less NO_x reduction is needed from the aftertreatment systems. Particulate emissions are also low because the reactants are largely premixed, so particulate filtration requirements are less.

For the near-term engines, which are mixing and diffusion dominated (CI), the technical team advocates “clean diesel.” In diesel engines the combustion is mixing controlled, where burning takes place in conjunction with the mixing of fuel and oxidizer. Low emissions from diesel combustion can be achieved by using EGR with advanced mixing and injection strategies, which must then be coupled with effective aftertreatment systems. Enhanced understanding of the interactions among the EGR, advanced injection strategies, and in-cylinder fluid motion is needed to promote combustion conditions that minimize the in-cylinder regions of high-soot and high-NO_x formation.

To move beyond the near-term engine combustion technologies ACECTT encourages continued fundamental development of chemical kinetics-controlled combustion, or LTC. LTC is used here as the name for the generic combustion process that is largely flameless, volumetric autoignition that is controlled by chemical kinetics. Control of LTC is achieved by staging the autoignition of different spatial regions of dilute (lean and/or EGR) fuel-air mixtures at temperatures below those necessary for rapid flame propagation: temperatures typically below approximately 1750^oC but above approximately 1250^oC. Many approaches are being investigated to achieve and control this staging of the autoignition, and this

___________________

¹ Stoichiometric combustion is when the mixture of fuel and air is chemically correct, such that there is just enough oxygen so that all the carbon and hydrogen in the fuel could be converted to carbon dioxide (CO₂) and water (H₂O), respectively, with no oxygen left over.

has led to an alphabet soup of acronyms: HCCI (homogeneous charge compression ignition), PPCI (partially premixed compression ignition), RCCI (reactivity controlled compression ignition), GCI (gasoline compression ignition), and so on. However, in all of these combustion modes the underlying objective is the same: one wants to establish a somewhat premixed mixture of fuel-vapor, air, and diluent that is encouraged to progress through the air-fuel mixture’s autoignition chemistry in such a way that the mixture ultimately chemically reacts through a rapid sequence of autoignition events that occur volumetrically throughout the cylinder. Successfully creating this situation inside the combustion chamber depends on the state of the gases at the start of compression and the physical and chemical characteristics of the fuel. The litany of acronyms that has appeared in the literature is an indication of the different approaches that can be taken to achieving this combustion mode. The benefit of LTC is higher efficiency, for the same reasons cited earlier in the discussion of dilute gasoline engines, and very low NO_x and soot.

As the engine becomes more efficient, more of the fuel energy is being converted to shaft work, so less energy is leaving the engine as exhaust enthalpy. Consequently, the engine exhaust is at a lower temperature than it is from a less efficient engine. This presents an additional challenge for the aftertreatment system. Today’s aftertreatment systems require catalyst bed temperatures in the range of 400^oC to 600^oC (Johnson, 2016). Industry is aggressively pursuing and making progress in lowering the light-off temperatures and temperature window of full functionality for the aftertreatment systems. Ideally one would like aftertreatment systems that are fully functional at cold start temperatures. For this reason the ACECTT is advocating continued research into perfecting aftertreatment systems that work at lower temperatures. This is needed for all aftertreatment systems: three-way catalysts, lean NO_x traps, selective catalytic reduction systems for NO_x, and particulate filters.

To benchmark progress and motivate research efforts the ACECTT has established research targets for 2020 for both engine efficiency improvements and exhaust emission levels. The high-level statement of the Partnership’s research target, which was presented to the committee, is given as follows.

2020 Partnership Research Target: A 20% improvement in engine efficiency, compared to a 2010 baseline. Engine concepts shall be commercially viable and meet 2020 emissions standards.

The research targets have been broken down for the specific engine types found in the market today, projecting that near-term engines will be further developments of current state-of-the-art engines. These engine-specific research targets are shown in Table 3-1. The targets are stated in terms of improvements relative to a 2010 baseline for each engine type. As shown in Table 3-1, the technical team has highlighted the operating points for each of the different engine pathways

TABLE 3-1 Advanced Combustion and Emission Control Technical Team (ACECTT) Stretch Goals for 2020 for Different Engine Types Currently Being Used in the Light-Duty Mobility Market

^a Entries in percent brake thermal efficiency (BTE) that are equal to 1.2 times the corresponding baseline BTE.

^b Entries in percent BTE.

^c Entries in bar of brake mean effective pressure (BMEP).

^d Downsized boosted baseline engine used premium grade fuel and direct injection.

NOTE: Baseline is 2010 light-duty vehicle U.S. penetration (multivalve, 86%; port injected, 91%; variable valve timing, 86%; stoichiometric with three-way catalytic converter, 99% [for less than 8,500 lb gross vehicle weight]). Highlighted cell represents most relevant operating point for that technology pathway.

SOURCE: Solomon and Howden (2016).

that it feels are the most relevant. The highlights represent the most important operating conditions at which research engines of that type should be thoroughly evaluated. The research targets for emissions were established based on current and pending regulations, as shown in Figure 3-1.

At the request of the committee the U.S. DRIVE Partnership provided a list of the projects within DOE’s advanced combustion engine research portfolio in which they are engaged. This list consists of 40 projects and is given in Table 3-2. The total DOE funding for FY 2015 for the projects listed in Table 3-2 was $29.02 million. The DOE funding allocation for these projects in FY 2016 decreased slightly to $27.5 million: several of the projects were completed in 2015 or early 2016, which is the primary reason for the respective differences in the funding levels.

The ACECTT has identified the fundamental areas (barriers) for which enhanced understanding would facilitate progress toward their research targets for each of the projects listed in Table 3-2. Rather than describe the specific barriers being addressed in each of the projects listed in the table individually a more general discussion is presented below, where the needed fundamental understandings are described in terms of the phenomena that currently limit a particular engine’s performance. This discussion is presented in the context of the three individual combustion strategies, the development of enhanced computational fluid dynamic (CFD) capabilities, and the challenges of future aftertreatment systems. The entries in Table 3-2 have been color coded to show the connection of the projects in the table to the general descriptions of the challenges for the different combustion strategies, CFD development, and aftertreatment systems. If a project advances more than one category of the general description it is shown with multiple color codes. Further detail on the individual projects in the table can be found in the presentations for the VTO given at the DOE 2015 and 2016 AMRs with the project numbers contained in Table 3-2.

TABLE 3-2 List of Department of Energy Projects Addressing U.S. DRIVE Goals

NOTE: Green, premixed flame dominated (SI); orange, mixing/diffusion dominated (CI); red, low-temperature combustion; blue, CFD development; yellow, aftertreatment; multiple colors, a project indicates it contributes to multiple areas. Acronyms defined in Appendix D.

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a).

Premixed Flame Dominated (SI) Engines—Dilute Gasoline

Premixed flame dominated, dilute combustion engines are currently constrained by knock-limited operating regimes, lack of combustion robustness, low combustion rates, low exhaust enthalpy, gas exchange complexity (such as EGR and air handling), and high emissions of hydrocarbons (HCs) and NO_x. Lean or dilute combustion offers higher efficiency; however, as one pushes the limits of dilution it is more difficult to initiate and maintain the flame. Once the flame is established the traditional approach of enhancing the burning velocity by increasing turbulence is limited, because it is easier to extinguish the flame under the higher fluid shear that occurs when turbulence is increased. If the dilution is obtained through EGR, either internal or external, the lack of homogeneous composition of the EGR with the intake air can cause flame extinction much more easily than for stoichiometric flames. Similarly, the flame propagation is more sensitive to mixture inhomogeneous conditions that occur between the fuel and the air in the combustion chamber under conditions of high dilution. Cycle-by-cycle variations during the engine operation now become a more serious problem. Because the combustion is less stable and more easily extinguished, unburned hydrocarbons and carbon monoxide (CO) emissions can become excessive. As boost is used to increase the power density of the engine, knock limits are often reached during high load operation. This limits the potential improvement obtainable from the engine because actions taken to avoid knocking operation compromise performance. Finally, when operating under very dilute conditions the exhaust temperature becomes so low that it taxes the capability of the aftertreatment system. If the high dilution is achieved through lean operation, three-way catalysts are no longer effective for NO_x control, so lean NO_x reduction would need to be implemented. Projects that contribute to further understanding of these phenomena are color coded green in Table 3-2.

Mixing and Diffusion-Dominated (CI)—“Clean Diesel”

Diesel engines have a potential efficiency advantage over premixed flame-dominated engines but they also have significant challenges. The ACECTT has identified the most critical challenges being faced by engineers working on diesel engines as follows: in-cylinder NO_x and soot control, EGR and air handling, fuel injection and control systems, and the cost and complexity of the systems. The characteristics of the fuel injection process play an important role in achieving clean diesel combustion. The intricate details of the internal geometry of the injector and how it affects the fuel flow leaving the injector play an important role in the early entrainment and mixing of the fuel and air in the immediate vicinity of the injector tip. This also impacts the symmetry of the spray pattern among the individual injector nozzle-hole plumes. The early behavior of the individual fuel plumes leaving the nozzle determine the air-to-fuel ratio of the portion of the spray that is first to experience autoignition, which in turn has a high impact

on the initial particulate formation rate. The formation of the particulate at this early part of the combustion process subsequently dictates the characteristics of the soot-NO_x trade-off of the engine. EGR is an important aspect of the soot-NO_x trade-off of the engine. Since diesel engines typically do not have throttles in the intake manifold, the pressure difference between the intake and the exhaust manifold is minimal. Engine performance is easily compromised by efforts made to induce the exhaust gas to flow between the manifolds when there are either insufficient or adverse pressure gradients. Different approaches to getting the exhaust gas from the exhaust to the intake manifold, like high-pressure and low-pressure loops, along with the challenges this imposes on the operation of the boosting system and the resulting impact on the pumping work, are motivating much technical research and development. Projects listed in Table 3-2 addressing these fundamental challenges are color coded orange.

Low-Temperature Combustion

Engines operating using LTC processes have the potential for very high efficiency with low cylinder-out emissions. However, there are significant challenges to implementing this combustion mode in an engine. Chief among them are high combustion noise, achieving combustion robustness, high engine-out HCs and CO, transient combustion control and emissions, cold start ability, obtaining a wide speed and load operating range, and cost and complexity. The combustion initiation for LTC operation depends on the chemical kinetic autoignition of a partially premixed air-fuel mixture. There needs to be a certain degree of nonuniformity within the air-fuel mixture in the cylinder to obtain acceptable combustion. If the mixture is too uniform the entire mixture autoignites at once and the rate of combustion is excessive. If the mixture is too stratified the combustion can either become excessively long or fail to go to completion. The nonuniformity of the mixture can be manipulated in many ways: nonuniformity in temperature, in air-fuel ratio, in oxygen concentration, or in fuel reactivity. In addition, the level of the nonuniformities in the cylinder necessary for good combustion is intimately linked to the characteristics of the fuel. This makes control of all aspects of the in-cylinder conditions extremely important, so engine controls and transients are a major technical challenge. Even if LTC modes were used in hybrid applications there would still be transients because of the stop-start nature of the engine operation in hybrid vehicles. Projects that contribute to further understanding of these phenomena are color coded red in Table 3-2.

Advancements in Computational Fluid Dynamic Capabilities

An important component of all of the ACECTT research effort is advanced computer simulation. New understandings from research are integrated into the simulations, which in turn are used to offer higher-fidelity interpretations of the

experimental results and to guide further research activities through prediction of more optimal operating conditions. In the quest of ever better engine performance the need for deeper understanding of the governing fundamentals continues to grow, which subsequently requires higher resolution and more precise submodels to be integrated into the simulations. For example, kinetic models capable of predicting the impact of different fuel compositions on engine performance; higher-fidelity numerical submodels such as fluid turbulence models simulating mixing and dissipation to ever smaller physical scales; combustion models that can capture the transition between flame propagation, mixing-controlled burning, and LTC energy release; better submodels for the heat transfer processes in the cylinder; more accurate emission models; more robust and faster numerical algorithms; and increased capability for program parallelization are examples of important advancements that would make the simulation efforts even more valuable than they are now. Projects listed in Table 3-2 addressing these fundamental challenges are color coded blue.

Aftertreatment Systems Advancements

As the engines are made more efficient the energy in the exhaust gases leaving the engine decreases, which requires the aftertreatment systems to function at lower temperatures. Furthermore, as the number of people living in urban areas increases, vehicle density in those areas increases. This results in a larger input of engine exhaust into the urban environment, which drives the need for even lower regulatory limits on criteria pollutants. Consequently, the exhaust gas aftertreatment systems need to be made more effective and capable of functioning at lower temperatures than current systems. This establishes the research priorities for exhaust gas aftertreatment systems. The aftertreatment systems are integral to the engine and power train; they need to be effective at temperatures below the current operating temperatures of 400^oC to 600^oC, reach light-off temperatures as quickly as possible, be resilient to poisoning from contamination, use minimal precious metals, and effectively filter particulate matter down to a size of 23 nm, the particle diameter above which the European number regulations are enforced, and they need to be easily regenerated. Projects that contribute to further understanding of these phenomena are color coded yellow in Table 3-2.

The deeper understanding of the thermodynamics, fluid mechanics, heat transfer, combustion physics, fuel chemistry, and catalysis that is sought in the research projects in Table 3-2 is necessary to develop better engine components and exhaust gas aftertreatment systems, but it is only part of the solution to achieve maximum performance with minimal environmental impact from the vehicle. The engine, exhaust gas aftertreatment components, and the power train must work as an optimized system, which presents complex challenges in total system control. These challenges not only push current capabilities in control theory but also

require advancement in the current state of the art of sensors, actuator technology, and onboard computer power.

The projects that the ACECTT advocates supporting strive to enhance the analysis capabilities and fundamental understanding of the technical barriers to more efficient and environmentally benign engines. As can be seen from the list of projects with which the U.S. DRIVE ACECTT engages (Table 3-2), research activities encompass a wide range of fundamental phenomena, which can be summarized topically as follows:

• Measuring and simulating phenomena occurring during the fuel injection process,
• Developing increased understanding and appropriate models for the combustion kinetics of different combustion strategies and their associated emissions,
• Developing sensors and new measurement techniques,
• Working on controlling combustion robustness and repeatability,
• Developing higher-fidelity turbulence models,
• Developing predictive simulations with high-performance computing,
• Developing combustion control algorithms,
• Working on aftertreatment modeling and analysis,
• Trying to enhance low-temperature catalysis, and
• Developing the next generation of an advanced modular computer program, which will facilitate inclusion of the new and higher-fidelity subprograms necessary to integrate predictive simulation into the design process.

In addition to the fundamental research activities, the ACECTT also engages in industry cost-shared demonstration programs. The purpose of these programs is to push the envelope of current product performance by integrating suites of advanced technologies from the research laboratory into a power train system and/or a vehicle. In doing this, valuable learning occurs as to the subtleties of interactions between the different system components and power train controls once they are completely integrated into a product. It also helps guide the manufacturers as to the time frame when different advanced technologies might be commercially viable. Within U.S. DRIVE four such activities have been identified:

1. ATP-LD, Cummins Next-Generation Tier 2 Bin 2 Diesel Engine;
2. Ultra-Efficient Light-Duty Power Train with Gasoline Low-Temperature Combustion, with Delphi;
3. Advanced Gasoline Turbocharged Direct Injection (GTDI) Engine Development, with Ford; and
4. Lean Miller Cycle System Development for Light-Duty Vehicles, with GM.

Overall the ACECTT is engaged in a broad array of research projects, ranging from fundamental laboratory investigations, to understanding subtle phenomena occurring within commercially available technologies, to exercises in advanced technology integration for market readiness assessment. With such a broad range of research activities it is important to make sure that efforts are synergistic and not duplicative, and that time horizons and risk factors of the research are appropriate to the organization performing the work. The advanced technology integration projects shown earlier can be very informative and are not something that industry would do when high-risk unproven technologies are being used. The national laboratories are not involved in these activities, which is appropriate. Within the ACECTT programs some of the research within the national laboratories involves use of already developed engine technologies. It is important that such work address underlying fundamentals that need to be better understood, and if understood would open up new performance possibilities. It should be work that industry would not or could not do on its own and that benefits the technical community at large. It appears to the committee that the ACECTT is doing a good job at maintaining this focus within their diverse program and that the DOE AMR serves as a good venue for getting feedback from the technical community.

Current Status vis-à-vis Goals and Targets

The principal barrier to achieving U.S. DRIVE’s research targets is lack of fundamental understanding of the relevant physics, chemistry, and thermodynamics. Consequently, the technical team’s status and progress toward reaching its goals is best measured by advances in knowledge, the effectiveness of the communication of that knowledge to the appropriate stakeholders, and ultimately the application of that knowledge to achieve an improvement in an engine-power train system.

The ACECTT has established a series of networks to facilitate communication and knowledge transfer within the appropriate technical communities. By interfacing with USCAR there is regular information exchange on fundamental needs as well as feedback on the programs considered to be part of the U.S. DRIVE portfolio. For example, industry representatives interface through USCAR to give critiques of the U.S. DRIVE projects presented at the DOE AMR each year. These critiques are in addition to the evaluations performed by the designated reviewers at the AMR meeting and are an important part of a checks and balances system of assessing if the work being performed by the different groups within the Partnership is appropriate to that group’s capability and time horizon.

Close collaboration with industry is facilitated through the Advanced Engine Combustion Memorandum of Understanding (MOU) led by Sandia National Laboratories. The MOU works to carry national laboratory research into products via crosscutting efforts between light-duty and heavy-duty engine R&D. Almost

all major engine manufacturers and energy companies participate in the MOU. University research is also integrated with the MOU.

The Engine Combustion Network is an excellent example of a predominantly web-based facilitation of data and CFD analysis exchange for better computer models of spray combustion. Over 16 international teams share experimental data, CFD approaches, codes, and actual model submissions for a well-characterized set of injectors operating under accurately prescribed conditions. This allows DOE/VTO investments to be leveraged many times over by complementary experiments conducted by other institutions. The collaboration efficiently transfers the research results to industry’s workflow through the development of engineering models. U.S. DRIVE estimates that the combined effort has accomplished 15 years of research progress in approximately 3 years (Howell, 2016).

As noted in the NRC Phase 4 report (NRC, 2013), the Cross Cut Lean Exhaust Emissions Reduction Simulation (CLEERS) is a technology focus group whose purpose is to promote development of improved computational tools for simulating realistic full-system performance of lean-burn diesel or gasoline engines and associated emissions control systems. CLEERS holds annual meetings as well as monthly phone conferences. Figure 3-2 shows the structure of the CLEERS focus group. It has proven to be a very effective venue for technical exchange between broad scopes of participants. The structure has proven so successful that the DOE Bioenergy Technologies Office (BETO) has adopted the model for their Computational Pyrolysis Consortium.

In addition to the preceding structured programs, which have regularly scheduled interactions, the ACECTT has established collaboration through special workshops, such as the following:

• Future Automotive Aftertreatment Solutions: The 150^oC Challenge Workshop (2012);
• Low-Temperature Oxidation Catalyst Test Protocol (2014);
• CFD for Engines Workshop (2014);
• Recommendations for Future Fuel Properties (2015); and
• Low-Temperature Combustion (LTC) Fuel Properties Survey (work in progress).

The ACECTT is actively engaged in following the research activities within the DOE VTO portfolio. Through the collaborations, working groups, workshops, and review participation they stay informed of the critical research needs of the industry partners and are effectively facilitating the knowledge exchange that supports the industry goals of more efficient, cleaner vehicles.

The assessment of the committee is that the U.S. DRIVE ACECTT is doing an excellent job at advancing the technical community’s understanding of the fundamental barriers limiting the improvement of engine efficiency and cleanliness and in disseminating that understanding to the relevant stakeholders.

Assessment of Progress and Key Achievements

When asked by the committee, the ACECTT submitted the following list of accomplishments for their combustion and aftertreatment activities:

• Combustion
  • Cummins ATLAS Diesel Engine Program Achieves 60% Cycle Fuel Economy Improvement over Gasoline Baseline—Cummins (2015);
  • Lean Downsized Boosted Engine Projected to Improve Combined-cycle Fuel Economy by 25%—GM (2013);
  • Gasoline Direct-Injection Compression Ignition Shows Potential for 39% Fuel Economy Improvement—Delphi (2014);
  • New Method Developed to Assess Engine Efficiency Opportunities—Oak Ridge National Laboratory (ORNL) (2013);
  • Unique Spray Measurements Enable Improved Models—Sandia National Laboratories (SNL) and Argonne National Laboratory (ANL);
  • Experimentally Validated High-Fidelity Simulations Enable Improved Gasoline Direct Injection Fuel Injection System Designs—ANL (2015); and
  • First Principles Simulations Reveal Details of Fuel Injection that Control Combustion and Emissions—SNL (2015).

• Aftertreatment
  • Study of GDI PM Size, Shape, and Composition Highlights Control Challenges That Differ from Diesels—ORNL (2013);
  • Low-Temperature Oxidation Catalyst Test Protocol—Advanced Combustion and Emission Control (2014);
  • Innovative Metal Oxide Catalyst Improves Low-Temperature Reactivity of Traditional Oxidation Catalyst—ORNL (2015); and
  • Promising Automotive Catalyst Provides Nitrogen Oxide Control at Temperatures Approaching 150°C—Pacific Northwest National Laboratory (PNNL) (2015).

Significant Barriers and Issues That Need to Be Addressed

The barrier to further improvement in engine efficiency and more effective aftertreatment systems continues to be the incomplete understanding of the detailed fundamentals of the thermodynamic, fluid mechanic, heat transfer, chemical kinetic, and combustion physics processes that occur within the engine and the aftertreatment systems. The tremendous advancements in engineering capabilities that allow designing and manufacturing with specifications on a molecular level gives confidence that as the engine community’s understanding becomes more complete, additional improvements in engine efficiency and cleanliness can be made.

Responses to Recommendations from the Phase 4 Report

___________________

² KIVA is the name of a 3-D CFD program that is being developed by DOE that can be used for engine combustion studies.

Committee Assessment of Response to 3-1. The committee thanks U.S. DRIVE for their efforts to improve the utility of the CFD modeling effort. The committee is aware that DOE held a high-performance computing workshop in August 2014 with invitations to all stakeholders. The outcomes of the workshop were a prioritized list of attributes heralded by industry as necessities, and plans for commercialization of the code DOE is developing with a goal of maintaining mostly open-source source code so university students can continue to modify and gain familiarity with the code for their own purposes and training. These outcomes address the concerns raised by the committee in Recommendation 3-1 of the Phase 4 review. The committee endorses the current program path and the continued development of KIVA-hpFE.

Appropriateness of Federal Role

The committee believes it is appropriate for the federal government to maintain a role in combustion and emissions control research. As described in the introduction to this section, there is still significant opportunity to reduce the fuel consumption and environmental impact of ICE-powered vehicles, so it is important to keep an active research program in this area. Developing the enhanced understanding and tools to do this pushes the state of the art in almost all physical and engineering sciences.

Findings and Recommendations

Finding 3-1. The ACECTT is a well-organized group within the U.S. DRIVE organization. They understand the technical barriers that need to be overcome to further increase the efficiency and environmental friendliness of ICE-powered vehicle systems.

Finding 3-2. To guide the technical work toward overcoming technical barriers to higher efficiency, the advanced combustion and emissions control technical team (ACECTT) has established stretch efficiencies goals for 2020 for peak and intermediate engine loads for the three types of engine power train systems they expect to be most prevalent in the near term: hybrid applications, naturally aspirated, and downsized boosted engine systems. The ACECTT is also engaged in research activities in chemical kinetic development, and promoting a more fundamental understanding of the interaction between fuel characteristics—such as Research and Motor Octane number, heat of vaporization, etc.—and different engine operating conditions. This work is aimed at facilitating the integration of advanced kinetically controlled combustion processes, i.e., low-temperature combustion, as part of the engines’ operating map, which is considered a longer-term technology.

Finding 3-3. The ACECTT focus for both near- and longer-term research is centered on conventional four-stroke engine architectures. However, work on alternative engine architectures is taking place. Some of that work is under DOE funding, and claims are being made in the literature of potential efficiency and environmental impact improvements for these different engine architectures (Redon et al., 2014; Chadwell et al., 2014).

Recommendation 3-1. The advanced combustion and emissions control technical team should be proactive in seeking out and assessing data on the performance of alternative engine architectures and concepts that will allow benchmarking against those within their current research portfolio.

Finding 3-4. The ACECTT has recognized that fuel characteristics can be a contributor to improving engine efficiency and reducing environmental impact. Because of the broadening of fuel feedstocks to nontraditional gas and oil as well as biomass, it is appropriate to include fuel characteristic considerations into ACECTT’s combustion research portfolio.

Finding 3-5. A Fuels Working Group has been formed and is actively engaged with the ACECTT, USCAR, and the DOE multilaboratory initiative Co-optimization of Fuels and Engines.

Finding 3-6. The ACECTT is engaged in a broad array of research projects ranging from fundamental laboratory investigations, to understanding subtle phenomena occurring within commercially available technologies, to exercises in advanced technology integration for market readiness assessment. There appears to be good collaborative engagement among the industry stakeholders (engine and vehicle manufactures, Tier 1 suppliers, and energy companies), USCAR, and the ACECTT. The committee believes that the ACECTT is doing a good job making sure the individual efforts are synergistic and not duplicative and that the time horizons and risk factors of the research are appropriate to the organization performing the work.

COMBUSTION ENGINE FUEL AND LUBRICANT TECHNOLOGIES

Petroleum

Background

The U.S. petroleum industry has evolved significantly since the NRC Phase 4 review of U.S. DRIVE (NRC, 2013). The development of hydraulic fracturing (“fracking”) and directional drilling techniques for production of both oil and natural gas has increased domestic energy supplies (EIA, 2016a), and reduced

energy imports to the point where it is predicted that the United States will be a net exporter of energy by 2040. Per the Energy Information Administration, “By 2040, total U.S. energy production is greater than total U.S. energy consumption, allowing for U.S. net energy exports equal to 4% of total consumption” (EIA, 2016a). As a result, the development of these alternative production methods has ensured the use of petroleum as one of the primary transportation fuels for light-duty vehicle applications for decades to come.

DOE and U.S. DRIVE interest in transportation fuel research is strongly influenced by three U.S. regulatory initiatives:

• EPA Tier 3 Emissions Regulations,
• Fleet CAFE [corporate average fuel economy] Standards for Light-Duty Vehicles of 54.5 mpg by 2025, and
• Renewable Fuel Standard for use of 36 billion gallons of renewable fuels (other than corn ethanol) by 2022.

The potential for ample amounts of petroleum-based fuel for use in transportation has increased concerns regarding the issue of GHG emissions and their effects on climate change. For that reason, it is critical that engine design and fuel composition be considered a system in developing new technologies for improving vehicle fuel efficiency and reducing CO₂ emissions. To this end, as described earlier in this chapter, the U.S. DRIVE ACECTT has included light-duty automotive fuel research projects within the mission of their research and development program (U.S. DRIVE, 2013h). This R&D effort has two overall goals:

These goals require identifying how drop-in fuels will impact advanced combustion and emissions control strategies as well as identifying practical, economic fuels and fuel-blending components with potential to directly replace significant amounts of petroleum (U.S. DRIVE, 2013h).

Parasitic losses in the engine and other driveline components are also a source of wasted fuel energy. Fenske et al. (2015) have calculated that for the 250 million “on-road” vehicles in the United States the daily loss in fuel energy due to friction is 1.5 to 1.8 million barrels of fuel. To address this issue, DOE is also funding research on advanced, low-friction lubricant formulations in support of U.S. DRIVE goals and objectives.

Petroleum fuel and lubricant research in support of the U.S. DRIVE goals is budgeted through the DOE VTO. Currently there are eight projects being con-

___________________

³ Fuels from alternative sources that are chemically equivalent to petroleum fuels.

ducted by DOE laboratories and supplier partners that specifically address fuel and lubricant research efforts. However, as described earlier in this chapter, there are additional advanced combustion projects that also include the evaluation of fuel formulations in meeting engine efficiency and emissions goals.

Table 3-3 provides an overview of DOE’s VTO projects specifically related to petroleum-based fuels, petroleum/bio-based fuels, and lubricant research activities.

The total DOE budget for VTO fuel and lubricant subprograms in 2014 was $15.5 million and in 2015 was $20.0 million, for a total budget over the previous 2 years of $35.5 million. A total fuel and lubricant budget of $37 million has been requested for 2016. These projects are distributed over a number of national laboratories, original equipment manufacturers (OEMs), and numerous suppliers. In the case of fuel research, the projects are focused on not only the combustion and spray characteristics of petroleum-based fuels (both gasoline and diesel) but also the characteristics of petroleum-biofuel blends. The use of biomass or nontraditional petroleum as either feedstocks or drop-in components to the fuel not only reduces the fuels’ carbon footprint, but it also introduces the possibility of changing the fuel characteristics in order to optimize the fuel and the engine together.

TABLE 3-3 Department of Energy Vehicle Technologies Office U.S. DRIVE-Related Fuel and Lubricant Projects List

NOTE: DI, direct injection.

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a).

Experimental data collected from individual projects are used to develop injector spray flow patterns and kinetic models of fuel combustion that are subsequently used in large CFD calculations of in-cylinder combustion. These programs are important in providing a methodology for efficiently evaluating a large number of potential options for fuel composition, injection strategies, and cylinder designs. The national laboratories have been able to develop new analytical techniques for identifying the combustion properties of fuel blends (e.g., Knock Index) in laboratory experiments using minute amounts of fluid, as well as new optical techniques for analyzing the formation of soot within an engine cylinder. In most cases the analytical facilities of the national laboratories greatly exceed the facilities at supplier and OEM laboratories and for this reason the use of the national laboratories for this work is appropriate in meeting U.S. vehicle fuel efficiency and emissions goals.

Status of U.S. DRIVE Fuel and Lubricant Goals and Targets

In support of the U.S. DRIVE goals for fuel and lubricant research, the VTO has directed a portion of the R&D they support toward achieving specific light-duty vehicle fuel and lubricant targets. As described to the committee (Howell, 2016), these targets include the following:

• Fuel Targets
  • Identify a spark-ignition candidate fuel “downselect” in 2017 (available at large scale, early 2020s),
  • Complete R&D on advanced conventional fuel/engine systems meeting 2025 CAFE/GHG target in 2020, and
  • Demonstrate optimized kinetically controlled engine/fuel system providing 30 percent GHG reduction versus “business as usual” in 2025.
• Lubricant Target
  • Demonstrate an engine/driveline lubricant package providing 4 percent fuel economy improvement relative to 2010 base fluids in 2020.

In order to achieve these goals, the VTO has funded the specific projects listed in Table 3-3. Although most of these projects are managed by individual national laboratories, the project teams include significant participation by U.S. DRIVE OEMs as well as other auto industry partners and suppliers. In addition, DOE has also funded supporting research at academic laboratories throughout the United States. For example, DOE recently announced a $7 million solicitation for research on advanced fuels for use in high-efficiency, low-emissions ICEs (DOE, 2016d). This solicitation was restricted to U.S. institutes of higher education. The final selection of academic research teams for this solicitation involves researchers from 14 universities. Using teams made up of governmental, industrial, and academic researchers, this research is intended to promote the rapid

commercial introduction of viable technologies developed by industry using the DOE research project results.

As described earlier in this chapter, U.S. DRIVE has recently established a “crosscut” activity titled the Fuels Working Group that includes not only U.S. DRIVE members and national laboratory personnel but also one non-U.S. DRIVE fuel company. Within the U.S. DRIVE organization, “working groups” are a less formal collaborative activity than are technical teams. Working groups do not conduct research according to a roadmap, nor do they have specific technical targets. Instead, they support the goals of related technical teams by coordinating research projects among interested parties both within and beyond U.S. DRIVE. In this regard, the FWG maintains close technical cooperation with the ACECTT. The specific mission of the FWG is to “evaluate potential properties of lower carbon fuels for future, high efficiency engines and combustion regimes working in close coordination with the advanced combustion and emission control tech team” (Cooper, 2016b).

To meet this mission, the FWG has identified in cooperation with the ACECTT four focus areas for their research programs. These focus areas are designed to evaluate potential properties of lower carbon fuels for future, high-efficiency engines and combustion regimes meeting ACECTT targets. These focus areas are as follows (Farenback-Brateman, et al., 2016):

1. Premixed, Flame Propagation, Spark Ignition Combustion Mode (SI)
2. Mixing/Diffusion Compression Ignition Combustion Mode (CI)
3. Chemical Kinetics Dominated Low-Temperature Combustion Modes (LTC)
4. New Combustion Quality Metrics
   • Anti-Knock for SI
   • Ignition Delay for LTC

As an example of how the FWG will address these focus areas, the ACECTT has identified a set of specifications for an advanced gasoline formulation shown in Table 3-4, which it believes would greatly enhance spark-ignition engine performance due to allowing higher compression ratios and reducing the need for spark retardation. To determine if these specifications will in fact meet the purposes identified in Table 3-4, the FWG is actively involved, as part of their focus area 1, in collecting engine data intended to define the benefits of each specification. These engine tests are being conducted within OEM laboratories and national laboratories.

In addition, the data collection efforts are being supported by several research projects being conducted within the Coordinating Research Council (CRC). The CRC is an independent research organization that is made up of OEM and oil company members. Recently, national laboratory personnel have also joined CRC working groups and contributed significantly to the identification of sur-

TABLE 3-4 Summary of Advanced Combustion and Emission Control Technical Team (ACECTT) Recommendations for Future Spark-Ignition Fuels

NOTE: MON, motor octane number; RON, research octane number.

SOURCE: Solomon and Howden (2016).

rogate fuel compositions that can be used to simulate commercial fuel combustion in advanced engines. The engine and fuel data collected from all of these sources will be used to model vehicle-level CO₂ emissions and fuel economy using “Autonomie Software” (Farenback-Brateman et al., 2016). At the same time Argonne National Laboratory will conduct “well-to-wheels” calculations of combined fuel production and vehicle efficiency using the GREET model. The data from all of these sources will be used to calculate a C2G analysis of the CO₂ footprint for each engine/fuel combination. The FWG will also identify and document infrastructure implications of new fuel formulations and identify technical roadblocks to implementation. All of this effort is directly related to determining the efficacy and benefits of the FWG focus area 1. Similar efforts and research projects will also be developed to quantify the benefits of the other three FWG focus areas.

As described earlier in this chapter, the FWG is also coordinating its efforts with the consortium within Energy Efficiency and Renewable Energy (EERE) called “Co-Optima” (Howell, 2016; Sarkar, 2016; Farenback-Brateman et al., 2016; Farrell, 2016). The Co-Optima initiative involves an External Advisory Board including not only membership of USCAR and the Truck and Engine Manufacturers Association but also membership of the American Petroleum Institute, the Advanced Biofuels Association, the EPA, the California Air Resources Board, and other industry groups. Current fuels are predominantly petroleum based, and refinery systems have been optimized for producing the current mix of gasoline and diesel. Their respective properties have traditionally been defined by meeting regulatory requirements and agreements reached within consensus organizations like the American Society for Testing and Materials (ASTM) and between members of the vehicle and energy industries. However, from the per-

spective of minimizing GHG emissions within the context of more advanced engine technologies with a broader base of fuel feedstocks, Co-Optima will evaluate nontraditional sources of petroleum and biomass-derived supplies to determine if it is possible to specify fuel characteristics that support more efficient engine operation. In addition, it is important to determine if optimization of fuel properties could further facilitate the introduction of advanced combustion processes, such as kinetically controlled, low-temperature combustion. However, it is important to note that identifying and ultimately implementing newly optimized fuels in the commercial marketplace is a huge task because any transition in fuel characteristics needs to be done while (1) maintaining compatibility with current infrastructure and legacy fleets, and (2) meeting current EPA/NHTSA fuel economy and GHG regulations as well as individual state fuel specifications based on industry consensus agreements such as those developed within the ASTM.

A recent status report on the developments generated within the first year of the Co-Optima consortium has been published (NREL, 2017). The introduction to this status report suggests that the objective to Co-Optima is

The status report summarizes recent research results that will significantly benefit USCAR OEMs in the development of advanced fuel-efficient, low-GHG engines. The status report maintains that the consortium will enable the introduction of new commercial fuel and engine technologies by 2025. As described in the following material, the committee understands that the Co-Optima initiative is a major undertaking and that meeting its stated goals, although well intended, will be difficult given the stated timeline.

Although the Co-Optima consortium is not a specific activity within U.S. DRIVE and therefore not a subject of this review, the existence of Co-Optima and its inclusion of many additional stakeholders with diverse interests raises several questions of importance to U.S. DRIVE such as how the initiative will be conducted, how decisions on experimental fuel properties and compositions will be made, and how advanced fuel specifications will be used. For example, if a set of fuel properties is identified to be useful in increasing engine efficiency or reducing C2G CO₂ emissions, will these properties be allowed or recommended for use in

• A laboratory fuel for advanced engine tests to demonstrate the benefits of co-optimized engine/fuel systems,
• A certification fuel that could be used by manufacturers in emissions testing of vehicles containing advanced engine technologies, or
• A commercial fuel available in the marketplace for general customer use?

There is also a question of how conflicts of opinions between the stakeholders directed at a common goal—engine/fuel combinations optimized for low emissions and high efficiency—will be adjudicated and resolved.

In addition, it is important to recognize that even if Co-Optima identifies an “optimum” engine/fuel combination for improving fuel efficiency and reducing GHG emissions, OEMs will require an extended amount of time for evaluating and validating the performance of such a fuel in their advanced development engines. Fuel effects on the durability of the fuel tank, pump, and injector systems will need to be verified. Fuel effects on the generation of criteria emissions including “toxics” and evaporative emissions will have to be determined over extended mileages. The fact that advanced engine/fuel combinations meet efficiency and emissions targets in national laboratory tests does not mean that OEMs will immediately adopt this proposed optimum for use in commercial vehicles. Correspondingly, energy companies will not introduce new fuels into the commercial marketplace without confidence that the engine systems that rely on such fuels will be available. Each of these issues will slow the introduction of new engine/fuel concepts despite the value of the advanced research being conducted and the new concepts being generated.

Progress and Key Achievements

There has been significant progress in meeting both fuel and lubricant goals in the 3 years since publication of the NRC Phase 4 review of U.S. DRIVE (NRC, 2013). In regard to fuel research accomplishments, the focus has clearly been on development of advanced fuel blends comprised of hydrocarbon and biofuel components. The research team led by Szybist at Oak Ridge National Laboratory has demonstrated a Reactivity Controlled Combustion Ignition (RCCI) engine using a gasoline/B20 dual-fuel strategy that meets the key ACECTT 2020 goal of 36 percent brake thermal efficiency (BTE) at 2,000 rpm and 20 percent peak load (Szybist et al., 2015). The RCCI operating range has been expanded to 75 percent of its theoretical maximum while maintaining low soot and NO_x emissions when using this fuel blend. This same team has also quantified the benefits of high-octane E30 (30 percent ethanol in gasoline) in downsized, turbocharged, direct-injection four-cylinder engines employing high compression ratios.

The team led by Zigler at the National Renewable Energy Laboratory (NREL; Zigler et al., 2015) has evaluated the derived cetane number of small fuel samples using an Ignition Quality Tester (IQT; ASTM Method D-6890). Over 388 pure compounds have been evaluated (122 by NREL using the IQT) and published in a compendium report (Yanowitz, 2014). The IQT has also been modified to provide kinetic data at gasoline combustion conditions of high pressures and temperatures. Octane reference fuels and gasoline/ethanol blends have been evaluated with the modified IQT for development of improved gasoline combustion kinetic mechanisms.

Sjoberg at Sandia National Laboratories has conducted a wide variety of analytical measurements in combustion engines operating on E0 to E30 gasoline/ethanol blends (Sjoberg, 2015). E30 blends are shown to be compatible with high-efficiency, boosted, direct-injection, stratified-charge engine performance. Sjoberg also concludes that ultralean SI operation requires end-gas autoignition for high-combustion efficiency.

A group at Lawrence Livermore National Laboratory (LLNL) led by Pitz has also developed a laboratory reactor that uses microliters of fuel to characterize the chemical kinetics of fuel combustion including gasoline, diesel, and diesel/biodiesel and ethanol/gasoline blends (Pitz et al., 2015). The “Micro-FIT” is a combustion tube that allows the evaluation of fuel ignition and flame extinction using only 5 to 20 µg/s of fuel. Multidimensional CFD simulations needed to understand fuel effects on lean/dilute direct injection, spark-ignition engines were conducted at Sandia National Laboratories (SNL) using data collected with the Micro-FIT at LLNL.

Mueller at SNL has conducted a detailed research program directed at reducing soot generation in diesel engines by modifying fuel composition and combustion strategies (Mueller, 2015). As part of a CRC program, a set of surrogate diesel fuels was developed, blended, and made available for laboratory test programs in combustion tubes and single-cylinder engines. An optical analytical technique, vertical-sheet, laser-induced incandescence, has been developed to analyze and quantify in-cylinder soot measurements. A diesel-biodiesel blend has been evaluated as an enabler for a leaner-lifted flame combustion strategy. Soot concentrations produced using this fuel/combustion combination were significantly reduced.

In regard to achievements in lubricant research, a DOE contract with Ashland Oil Corporation has led to the development of a heavy-duty engine oil formulation that provides greater than 2.0 percent fuel economy improvement relative to a CJ-4, SAE 15W-40 oil (Wu et al., 2015). SAE 5W-20 versions of this lubricant technology are predicted to provide greater than 2.0 percent fuel economy improvement in light-duty applications. Ashland has also developed an axle lubricant that can provide 0.7 percent improvement in fuel economy. These developments demonstrate significant progress in meeting the VTO goal of a 4.0 percent improvement in vehicle fuel economy provided by 2020 through reduction in driveline friction.

A research program led by Fenske at Argonne National Laboratory has focused on the development of methods for analyzing tribo-films that form on rubbing surfaces and affect friction and wear (Fenske, 2015). Techniques including focused ion beam electron spectroscopy and X-ray absorption near-edge structure techniques have been used to identify the degree of amorphous or crystalline nature of additive films on wear surfaces.

Significant Barriers and Issues That Need to Be Addressed

The DOE focus on petroleum fuels and lubricant research has developed into a well-structured portfolio of projects. There are no real barriers to conducting research on advanced fuel concepts under study at various national laboratories, suppliers, and OEM facilities. Instead there are only the challenges posed by DOE VTO management goals of identifying an advanced engine/fuel combination that reduces the nation’s dependence on petroleum and increases the use of low-carbon, alternative fuel components and developing a plan for introducing such engine/fuel combinations in the commercial market. An additional challenge is to reduce driveline parasitic losses through the development of advanced lubricant formulations. The research programs currently under way are well structured to meet these challenges, but whether the program will meet its challenges according to the timeline proposed by the VTO is unknown.

In the view of the committee, meeting the timing of the goals identified by the DOE for advanced engine/fuel combinations, although well intended, will be difficult. The committee was told that the VTO plans to “downselect” a specific set of candidate fuel properties by 2017 and to demonstrate an optimized kinetically controlled engine/fuel system by 2025 (Howell, 2016). This is a very aggressive set of objectives. Although the current portfolio of projects will provide technical data to aid in making the selection, the process of choosing an optimized engine/fuel system will be difficult. Presumably, the Co-Optima program will play an important, if not defining, role in selecting such optimized engine/fuel systems. However, reaching consensus within a diverse group of government and commercial interests is going to be a severe challenge. Each potential engine/fuel combination will have technical, economic, and political benefits and drawbacks. It is not too early in planning to identify the process and criteria for selecting an optimum system. Finally, it is also not too early to formulate the plans for promoting the use of such an engine/fuel combination in commercial vehicles.

Response to Recommendations from Phase 4 Review

There were no recommendations dealing with petroleum-based fuels or lubricants in the NRC Phase 4 review.

Appropriate Federal Role

It is entirely appropriate for the federal government to be involved in research that affects two independent, major U.S industries, auto and energy. It is the committee’s opinion that there is very little chance that a future advanced engine/fuel combination that meets national goals for efficiency and emissions will be universally accepted by all partners. The national laboratories can serve a critical role as arbiter of conflicting data and disagreements regarding potential benefits. The introduction of new fuels into the marketplace that must precede, or at least

be in conjunction with, the introduction of advanced engines will occur only after the development and agreement on commercial fuel-quality standards. Participation of national laboratories personnel in these standards-setting processes would add significant technical resources and expertise to this effort and accelerate the standards-setting process. The Co-Optima initiative may help in reaching consensus on these challenging technical and economic issues, although more information is needed on how this program will deal with such issues.

Findings and Recommendations

Finding 3-7. The portfolio of projects assembled by the U.S. DRIVE Fuels Working Group (FWG) and focused on the development of advanced petroleum-based fuels is well structured. Engine tests are being conducted or planned to evaluate the efficiency and emissions of petroleum/biofuel blends in dilute/lean spark-ignited (SI) engines, clean diesel engines, and low-temperature combustion engine designs.

Finding 3-8. Laboratory and dynamometer tests conducted at national laboratories are contributing to progress in meeting specific U.S. DRIVE ACECTT goals. They are generating important data in support of development of combustion models using advanced fuels, different injection strategies, and combustion chamber designs.

Finding 3-9. DOE has set an aggressive timeline for identifying an “optimized kinetically controlled” engine/fuel system. The Co-Optima program will presumably help in developing the data to establish such an “optimized” system, but the DOE has not yet addressed how such a system would be implemented in the light-duty vehicle fleet.

Finding 3-10. DOE needs to provide greater explanation of how the Co-Optima program will be managed. How are the research projects at the national laboratories set to meet the technical needs of the U.S DRIVE OEMs under the Co-Optima program?

Recommendation 3-2. Following on presentations made to the committee, the Department of Energy (DOE) should further explain how the Co-Optima program will lead to the introduction of an optimum engine/fuel system in commercial practice. The introduction of high-efficiency, low-greenhouse gas internal combustion energy technology into the marketplace may require fuel formulations that are different from today’s commercial fuels. Engine manufacturers will not introduce vehicles that utilize advanced combustion systems without the assurance that suitable fuels are available for the new combustion technology. Reaching consensus between the DOE’s Co-Optima program and U.S. DRIVE on the

concept of an optimum engine and fuel is necessary, but not sufficient. A plan for introduction of advanced combustion systems and fuels designed to increase transportation energy efficiency and reduce CO₂ emissions is required.

Biofuels

Background

As discussed in NRC (2015b) and other places, Congress established the Renewable Fuel Standard (RFS) in 2007, which set a goal of using 36 billion gallons of biofuels per year in transportation applications by 2022. To meet this goal, Congress has provided tax credits and incentives for biofuels production. These credits and incentives generally remain in effect.

Despite the federal government’s desire for increased biofuel use, the only significant source of biofuels in use today is corn-derived ethanol. This ethanol is added to gasoline in the United States, mostly at a concentration of 10 percent. In early 2011, the EPA expanded a waiver to allow up to 15 percent ethanol in gasoline used to fuel 2001 and later model year light-duty vehicles. The EPA cannot force fuel stations to provide gasoline blends containing 15 percent ethanol without the approval of Congress, which at this time, it does not have. The use of 15 percent ethanol has been opposed by some global OEMs due to concerns over fuel system durability in older engines designed for 10 percent ethanol in gasoline. To date very little gasoline containing 15 percent ethanol has become available at commercial fuel pumps.

The commercial production of cellulose-derived ethanol, envisioned in the RFS, is slowly being realized.⁴ Three different companies, Abengoa Bioenergy, POET, and DuPont have been producing ethanol from cellulosic feedstocks since either 2014 or 2015 (Abengoa Bioenergy, 2014; POET, 2014; Lane, 2014). The combined production of ethanol from these three plants could reach 60 million barrels per year. The plants are using technology that was developed in part at the NREL.

The use of bio-based butanol is also being considered for transportation use. Butanol has better gasoline blending and vapor pressure characteristics than ethanol, while still providing a significant octane boost. Butanol can be produced from renewable biomass, and it can be produced from refinery operations that produce excess C4 hydrocarbons. However, the production of butanol from petroleum light ends in the refinery would not be considered a renewable fuel.

The production of biodiesel (essentially fatty acid methyl ester [FAME] and other esters) is still minimal but continues to increase (National Biodiesel Board, 2016). Although the process for production of FAME is well defined, there is

___________________

⁴ Cellulosic feedstock is primarily sourced from the structural components of plants and trees (e.g., trunks, branches, stems, stalks, leaves, and roots).

significant research and development being conducted on the identification and production of other oxygenated diesel fuel components that could provide benefits in clean diesel combustion applications.

Renewable “drop-in” fuels (fuels that are derived from biomass feedstock but processed into commercial fuel components using refinery operations) are the subject of research in both national and industrial laboratories (Peckham, 2014). National laboratory research programs are defining the performance of both wood-derived bio-gasoline and bio-reformates in engine tests. Wood-derived gasoline is produced by gasifying wood residue, compressing the resulting syngas, and then catalytically reforming the components of the gas stream (Farenback-Brateman et al., 2016). Bio-reformate components are produced from starch, sugar, or cellulosic feedstocks that are subjected to catalytic reforming with extra added hydrogen. Although the technology for producing “drop-in” biofuels is known, further development is needed to reduce production costs per unit of energy produced (Brown and Capareda, 2015).

Status of U.S. DRIVE Biofuel Goals and Targets

The U.S. DRIVE biofuel goals and targets are essentially the same as those described earlier in this chapter for petroleum-based fuels. If there is any difference, it is in the emphasis on the validation of biofuel components that might be blended into gasoline or diesel fuel to promote improved efficiency and reduced emissions.

The projects listed in Table 3-3 describe research that focuses not only on petroleum fuel chemical properties and combustion characteristics but also on biofuel characteristics and petroleum/biofuel blends that provide specific properties to aid different advanced combustion strategies. The projects listed are intended to address issues that will facilitate meeting ACECTT and FWG objectives and goals. In addition, as with petroleum projects, the U.S. DRIVE groups work closely with oil company representatives, OEM members, and national laboratory personnel in Coordinating Research Council research groups and committees.

Progress and Key Achievements

As indicated previously, most of the key achievements in regard to VTO fuel development projects since the last U.S. DRIVE review have involved the use of petroleum fuel/biofuel blends. Many of these projects were listed in Table 3-2 as part of the ACECTT’s combustion, emission, and aftertreatment activities discussed earlier in the advanced engine combustion portion of this chapter.

One research area not covered is the work of McCormick and his team at NREL. This research project is investigating the possibility of using long-chain oxygenates derived from biomass as drop-in fuel components for blending

with either gasoline or diesel fuel (McCormick et al., 2015). The oxygenate candidates are being solicited from suppliers and are being evaluated in laboratory tests to determine fuel blend properties. In addition, the fuels are also being evaluated in both diesel and gasoline direct-injection (GDI) engine tests. Twenty-four different oxygenates have been considered. Partial results have demonstrated that phenolics and esters improve fuel lubricity. Ethers and ketones have no effect on lubricity. There is no effect of oxygenates on oxidation stability except for phenolics that act as antioxidants. Early engine test results have shown no effect of oxygen on particulate matter emissions. GDI tests have shown that particulate matter and particulate numbers for gasoline/oxygenate blends follow the same trends as for hydrocarbon fuels.

Significant Barriers and Issues That Need to Be Addressed

Cost challenges represent a substantial barrier to use of biofuels in commercial fuel blends. In the case of drop-in fuels derived from biomass, it has been reported that the cost of production is a major deterrent confronting commercialization of these fuel components (Brown and Capareda, 2015). In the NRC Phase 4 U.S. DRIVE review, it was stated that the DOE target for achieving cost-effective drop-in fuels was 2017 (NRC, 2013). There have been no data presented that would indicate that this objective will be met on time. It appears that this is a challenge that will continue into the future.

One issue that does not seem to be sufficiently addressed in many of the VTO research projects is the issue of engine and fuel system long-term hardware durability with any new commercial fuel or fuel blend. Obviously, first meeting goals related to fuel efficiency and emissions is paramount. Furthermore, it is the responsibility of each OEM to decide if their products will survive for acceptable periods when operating on a new fuel. However, early tests within participating research and development laboratories could provide indications of the significance of any durability issues. Early tests of oxygenated fuel effects on fuel lubricity are a good example of this effort (McCormick et al., 2015). EPA providing a waiver for 15 percent ethanol use in commercial gasolines is an example of the resistance that can be generated by both OEM and petroleum distributors when a new fuel blend is authorized without sufficient early durability testing.

Appropriate Federal Role

The programs funded by the VTO that include research on biofuels in support of U.S. DRIVE objectives are well organized. If the United States is to achieve its stated goal of using 36 billion gallons of biofuels in transportation applications, alternatives other than corn-based ethanol in gasoline need to be developed and commercialized. Since the federal government is the author of the Renewable Fuel Standard designed to reduce the use of petroleum-based fuels and green-

house gases from the transportation sector, the government has an appropriate role in conducting research and development in support of the achievement of its stated goals.

Findings and Recommendations

Finding 3-11. The use of ethanol as a blending component in petroleum-based fuels is being well researched and evaluated for use in advanced combustion concepts. There is less emphasis on the benefits of other low molecular weight alcohols (higher molecular weight than ethanol, but lower than C7 alcohols, e.g., butanol, pentanol, etc.). Although these alcohols would most likely be produced from petroleum feedstocks, they could provide benefits in fuel blending and advanced combustion strategies.

Finding 3-12. “Drop-in” fuels derived from biomass can be produced from recognized unit operations within a refinery, but cost is currently a challenge for use of these components in commercial fuel blends.

Finding 3-13. The effect of biofuel components on engine and fuel system durability is as important as their effect on engine efficiency and emissions.

Recommendation 3-3. It is recommended that in support of ongoing U.S. DRIVE, Coordinating Research Council, and Co-Optima fuel evaluation programs designed to improve internal combustion engine efficiency and reduce greenhouse gas emissions, other offices within the Department of Energy such as the Bioenergy Technologies Office continue to provide support for development of cost-effective manufacturing processes for converting biomass to liquid “drop-in” hydrocarbons compatible with gasoline and diesel fuel blends. Achieving lower cost manufacturing goals would provide U.S. DRIVE with a new pathway to meeting their objective of increasing the amount of biomass-derived fuels used in commercial fuel blends.

Recommendation 3-4. The Department of Energy’s national laboratories should systematically evaluate the effect of all potential biofuel components on engine and fuel system durability as well as their effect on engine efficiency and emissions.

Natural Gas

Background

A detailed review of the use of natural gas in medium- and heavy-duty trucks has recently been published by the NRC (2014). The reader is referred to this review for information regarding the justifications, availability, issues, and benefits

related to the use of natural gas as a transportation fuel. As described earlier in this chapter, recent developments in “fracking” and drilling technologies have greatly increased the supply and future estimates of reserve gas available in the United States (EIA, 2016a). The current production of natural gas using these advanced extraction techniques is so great that the Energy Information Administration has predicted that the United States will become a net exporter of gas by 2018 (EIA, 2016a). In addition, current natural gas supplies have driven the price of natural gas to its lowest level in recent history. This has resulted in much greater use of natural gas in electricity generation and slightly greater use as a medium- and heavy-duty truck fuel. In comparison, the use of natural gas in U.S. light-duty vehicle applications remains very small.

OEMs, both domestic and foreign, have been producing and marketing natural gas vehicles that meet fuel economy and emissions regulations for over three decades. Despite the requirement for large, expensive fuel tanks needed to provide acceptable driving “range” and that also create onboard packaging and storage challenges in vehicles, natural gas vehicles have achieved significant penetration in vehicle populations in numerous countries that have a suitable natural gas vehicle fueling infrastructure (e.g., Italy, Argentina, etc.). The reasons for limited sales of such vehicles in the United States can be attributed to three economic rather than technical issues:

1. New vehicle costs being above gasoline-fueled vehicles because of more expensive emissions control components and large bulky storage tanks;
2. Limited, high-cost natural gas refueling infrastructure for private citizens; and
3. Poor resale value for fleet owners that use centralized refueling facilities when they try to sell used natural gas vehicles to private citizens who have no access to refueling facilities.

With the exception of further materials research to lower the costs of emissions control components and storage tanks, there is little that the VTO can support in research programs that will solve these problems. For this reason, it is not surprising that there are no projects listed in Table 3-3 that focus on the use of natural gas as a light-duty vehicle fuel.

There is the possibility of using natural gas in a Fisher-Tropsch reaction to produce drop-in hydrocarbon fuels (either gasoline or diesel fuel). Currently there are large Fisher-Tropsch plants that are producing commercial products, but most of these are relatively high-cost lubricating base oils or chemical feedstocks. There are also mini Fisher-Tropsch plants that are planned for use at stranded gas deposits (not connected to a pipeline). Conversion of “gas to liquids” at these stranded sites allows liquid feedstocks to be transported by truck as opposed to by pipeline (Peckham, 2014). However, such mini Fischer-Tropsch plants are also challenged by unfavorable economics.

Status of U.S. DRIVE Natural Gas Goals and Targets

There are no known VTO initiatives in support of the U.S. DRIVE Partnership that utilize natural gas in their research projects. There are as many as 13 projects investigating natural gas as a transportation fuel within the DOE Advanced Research Projects Agency-Energy (ARPA-E) Methane Opportunities for Vehicular Energy initiative, but it does not appear that these projects are driven by U.S. DRIVE goals and objectives at this time (ARPA-E, 2012).

Significant Barriers and Issues That Need to Be Addressed

The largest barrier to the use of natural gas as a transportation fuel for light-duty vehicles is the lack of a refueling infrastructure for private vehicle owners. A population of refueling stations that would meet individuals’ needs for both local and long-distance trips could remove a good portion of the resistance to purchase light-duty vehicles fueled by natural gas, as it has in many countries around the world. However, it is worth considering whether the development of a hydrogen infrastructure and public refueling capability for use with fuel cell vehicles will create economic resistance to development of a similar natural gas refueling network for internal combustion engine-equipped vehicles. It is reasonable to ask if the United States will support the development of two public gaseous fuel infrastructure networks.

Response to Recommendations from Phase 4 Review

Committee Assessment of Response to 3-2. U.S. DRIVE also supplied the committee with extensive documentation that supports the summary statement given in their response. The committee thanks the U.S. DRIVE Partnership for the comprehensiveness of their response and is pleased that they have evaluated compressed natural gas in terms of its technical and economic viability for use in the light-duty mobility market.

Given that there are ample supplies of natural gas available in the United States at this time, coal is slowly being displaced by natural gas in electrical power generating stations because of its relatively low cost and its value as a method for reducing GHG emissions (EIA, 2016a) during electricity production. Because of the infrastructure needs that would be required for widespread use of natural gas in transportation applications, it is reasonable to think of natural gas, if it can be produced and transported without stray methane emissions, as the most clean and efficient fuel for use in electricity generation.

Appropriate Federal Role

It is entirely appropriate for the federal government to support research on methods for improving efficiency and reducing emissions related to the use of natural gas, if it believes that it is a viable option as at transportation fuel. There are valid applications for use of natural gas as a transportation fuel for medium- and heavy-duty trucks (Class 6 through Class 8, including metro buses, school buses, garbage trucks, and over-the-road tractor trailers) with access to dedicated refueling facilities.

Findings

Finding 3-14. There are no research projects at this time funded by the VTO in support of U.S. DRIVE goals and objectives that are focused on the use of natural gas as a fuel for advanced combustion systems. There are projects being conducted within ARPA-E focused on use of natural gas as a transportation fuel, although it does not appear that this focus is in response to U.S. DRIVE goals and objectives.

Finding 3-15. Liquid “drop-in” fuels could be produced from Fisher-Tropsch processes that use natural gas as a feedstock.

FUEL CELLS

Introduction and Background

Hydrogen fuel cell vehicles (HFCVs) have been in a development phase by the major automotive companies for decades. Their attractiveness when compared to current internal combustion engine technology is based on the direct conversion of chemical (hydrogen) to electrical energy via an electrochemical process, thereby resulting in reduced environmental impact provided that the hydrogen is derived from “green” primary energy sources. The efforts to develop HFCVs by the major automotive companies have been significant as is evident from the magnitude of the investments made by the individual automotive OEMs, the number of patents issued (DOE, 2016e), and the engineering accomplishments to date (U.S.

DRIVE, 2015). Notably, within this review period, a number of foreign OEMs (Toyota, Hyundai, and Honda) either have initiated membrane-based fuel cell vehicle sales or leases to the general public in the United States or have announced that vehicles will be available within the next few years. General Motors, Ford, and Fiat Chrysler, all three U.S. DRIVE Partnership members, do not currently have vehicle offerings, yet GM has been cited in the open literature as stating they are “on track” to produce their Gen 2 HFCV by 2020.⁵ Recent activities by the aforementioned OEMs, foreign and domestic, demonstrate that HFCVs are in the late stages of development and are, or are soon to be, ready for customer engagement, albeit at a modest level owing to limited production volume and hydrogen delivery and refueling infrastructure issues. The U.S. DRIVE Partnership, which is focused on coordination and communication of activities funded by the FCTO within EERE at the DOE and the OEMs, has been and remains fully engaged in all aspects of HFCV development.

A key interest in electric vehicles (battery or fuel cell powered) has been the potential of this technology to ultimately achieve near-zero carbon emissions (Satyapal, 2016) if renewable energy sources such as wind, solar, hydroelectric, or other non-fossil-fuel-based sources are used to generate the hydrogen needed to fuel the HFCVs. If achieved, this would facilitate meeting the low-emission mandates created over the last two decades. Even if fueled by hydrogen derived from natural gas, life-cycle emissions of carbon dioxide and pollutants would still be lower than those from current internal combustion engine emissions (DOE, 2015b).

Despite foreign OEM HFCVs entering an introductory customer-engagement phase, and with General Motors expected to follow in the next few years, challenges specific to fuel cell technology remain (Figure 3-3). Such challenges have been outlined in prior NRC program reviews (Phases 1 to 4) and many have been resolved. Meeting cost and fuel cell durability targets simultaneously remains the most critical barrier to overcome if HFCVs are to become viable, both technically and commercially.

Notable Changes since the NRC Phase 4 Report

In addition to the initial availability of HFCVs, another notable change since the NRC Phase 4 report is that the DOE has recently created a Fuel Cell Consortium for Performance and Durability (FC-PAD) to coordinate activities conducted at national laboratories, in conjunction with industry and universities, in the areas of electrocatalysts and supports, electrode layer, ionomers, modeling and validation, operando (i.e., operating conditions) evaluation, and component characterization. The overall goals of the consortium are to advance the performance and durability of proton exchange membrane (PEM) fuel cells at a precompetitive

___________________

⁵ See, e.g., http://www.usatoday.com/story/money/cars/2017/01/30/general-motors-honda-fuel-celldeal/97240096/ or http://www.autonews.com/article/20161011/BLOG06/310119999/fuel-cell-puzzlecomes-together.

level in order to facilitate their commercialization, develop the knowledge base, and optimize structures for more durable, high-performance PEM fuel cell components. Reducing costs and improving high current density performance at low platinum (Pt) loadings are also anticipated outcomes. The committee supports the creation of FC-PAD and its focus on key electrochemical electrode issues.

The newly formed DOE-facilitated ElectroCAT consortium is in its formative stages and like the FC-PAD, it is to be national laboratory-centric with projects involving industry and academia that focus research activities on non-platinum-group-metal fuel cell catalysts.

Budgets

Because the Partnership is not a funding entity, a direct link does not exist between the identified needs developed by the fuel cell technical team (FCTT) and the funds allocated for specific R&D by DOE. As a result, the entire FCTO fuel cell budget (Figures 3-4 and 3-5) should be viewed as contributing to technology development for the FCTT targets (see Table 3-6, later) set forth by the members of the Partnership. The current DOE budget for fuel cell development of approximately $35 million per year represents about 30 percent of the overall EERE FCTO appropriation and has remained stable in recent years. Additional funding of related technology takes place in other DOE offices such as Basic Energy Sciences, ARPA-E, and the Small Business Innovation Research (SBIR) program. Appropriately, as shown in Figure 3-5, catalysts and electrodes receive the largest allocation, yet funding of membrane topic areas has been decreasing in recent years. The most significant increase appears to be in the “Fuel Cell Operation & Performance” category followed by a slight increase in “Testing and

Technical Assessment.” It can be concluded that cost and durability issues are the primary focus areas. With that said, this conclusion draws attention to the delineation between near- and long-term projects incorporating the premise that the Partnership focuses only on identifying and addressing precompetitive concepts.

The ongoing DOE-funded projects that could impact Partnership-derived targets are presented in Table 3-5.

It is important to point out that in addition to the DOE funding, the OEMs are using their own internal financial resources to advance their respective proprietary technologies and specific needs. The DOE-supported precompetitive projects listed in Table 3-5 being conducted by national laboratories, academia, and private industry allow for critical resources at the OEMs to focus on getting a product to market.

Current Status vis-à-vis Goals and Targets

Three of the key 2020 technical targets set forth by the Partnership have been met as reported by DOE—power density, start temperature, and specific power (Table 3-6 and Figure 3-6). Though progress has been made in meeting the remaining targets, two continue to remain difficult to meet simultaneously, those being cost and durability (Wimmer and Gazelle, 2016). These were also identified in the NRC Phase 4 review report in 2013 (NRC, 2013). Since then, progress toward meeting the cost and durability goals has been minimal (Satyapal, 2016). It should be noted that the actual numbers achieved by the OEMs are generally not publicly available and therefore it is difficult to know or even monitor actual advancements relative to the FCTT targets. Furthermore, lifetime and durability testing requires significant blocks of time to perform; therefore, any currently reported data by the DOE may not be on the most recent generation of a given OEM’s HFCV. The most recent non-OEM reported data by NREL using on-road HFCVs, including one U.S. OEM (GM), generated from tests conducted between 2012 and 2015 (Kurtz et al., 2015), does cite an increase in lifetime, but still well shy of the 5,000-hour target. An on-site visit to GM’s fuel cell development operations in Pontiac, Michigan, by the fuel cell subgroup of the committee confirmed that significant advancements have been made toward the cost and durability targets.⁶ GM disclosed that they had achieved stack lifetimes (durability) exceeding DOE targets in laboratory tests. With respect to cost reduction, GM confirmed a press report (Truett, August 2015) of platinum group metal (PGM) stack loadings of 10 grams, reduced from 20 to 30 grams. It remains to be demonstrated that the durability reported to the committee can be achieved with the lower catalyst loadings.

Significant progress has been made by the national laboratories in gaining a better understanding of the fundamental chemical and materials science and engineering, which can impact cost (platinum loading) and durability.

___________________

⁶ The committee gathered this information during a visit to General Motors on June 21, 2016.

TABLE 3-5 List of Selected Department of Energy-Funded Fuel Cell-Related Projects

SOURCE: Cooper (2016a).

Cost

As in previous reviews of this Partnership, the cost target of $40/kW has proven to be difficult to reach, the current value being $53/kW (see Figure 3-7; Satyapal, 2016). Little progress has been made since 2010, further emphasizing that research must continue so as to offer new technical approaches that result in lower cost options. Such cost reductions must come from simplification of the system architecture as well as from new component solutions and operational approaches. System and stack costs are presented in Figure 3-8 as a function of volume; though interesting, it is not clear to the committee as to why low volume cost data (<500k units) are of any value.

As the state of maturity of fuel cell technology is still early, cost issues can be addressed by new technical solutions and approaches and less so by volume manufacturing. It is also important to emphasize that technical initiatives related to cost reduction and durability are inter-related, leading to long-term, complex,

TABLE 3-6 Fuel Cell System and Component Status and Targets

SOURCE: Masten and Papageorgopoulos (2016).

multivariable R&D projects. Many such efforts have been in place for years at the national laboratories, in academia, and private industry funded by the DOE. It is imperative that such efforts continue, as it is anticipated that ultimately, significant advancements will emerge from such work.

A challenging issue for the committee is that the cost estimates made by third parties to evaluate progress toward meeting the cost target may not accurately describe the actual OEM costs. To further add uncertainty to the cost estimates, the OEMs may have already developed their own proprietary solutions either

internally or together with the supply chain as DOE-funded projects are not likely to move at the same pace as the OEM internal development efforts. In a recent Strategic Analysis cost assessment (James and Spisak, 2012), selected components are used in the analysis, some of which are still being reported as development efforts by the private sector and are unproven. This adds a degree of risk to the credibility of cost assessments. As in the prior review, direct communication with one U.S. OEM indicated that even with this uncertainty, they are comfortable with the ultimate target of $40/kW. With that said, there were concerns regarding the differences between the OEM and the DOE-funded cost analyses based on the technology, systems, and operational approaches.

The sensitivity analyses reported by the DOE indicate that the catalyst loading, stack plates, and the air compressor are the components that have the largest impact on cost (Figure 3-9). Unique to fuel cells is that the current density and stack voltage can be controlled so as to impact performance and overall efficiencies. This characteristic will dictate design features that include cell area (plate area), the number of cells, as well as other components such as the radiator, air compressor, and blower sizing. As a result of the interdynamics of how the stack operates and behaves, the DOE has been appropriately funding studies and development activities in these areas, including system operating modes and systems architecture relationships, which to some extent depend on catalyst loadings and membrane and electrode characteristics. As presented in Figure 3-9, the number

one issue is the platinum content used in the catalyst layers. Currently, as reported by the DOE, approximately 20-30 grams of platinum per stack are used, a quantity significantly higher than the target. The objective is to reduce the Pt content to the target loading set by the Partnership of 0.125 grams of Pt/kW. Lower catalyst loadings work but at the expense of durability, as previously discussed. In response to this ongoing challenge, the DOE continues to fund a number of academic, industrial, and national laboratory catalyst, cell and stack performance, architecture, and systems analysis projects. Newly generated solutions will address these complex interactions that will ultimately translate to cost reductions without sacrificing durability. GM suggested to the committee fuel cell subgroup that a DOE-funded study to generate an understanding of durability issues of low Pt levels could provide valuable insight. The committee agrees with this. The recently formed national laboratory-based consortium described above will attempt to determine the membrane and electrode degradation mechanisms, and develop new catalyst solutions. The ElectroCAT consortium will address lower cost catalyst and electrode layer technologies using non-PGM catalysts. Both of these activities are warranted and long overdue and will hopefully impact cost.

In summary, the committee feels that the cost assessments by third parties must be addressed with the OEMs directly in a way that does not compromise their respective proprietary and confidential technologies and business plans. Cost estimates must be based on an understanding of system architecture, mode of operation, and especially end-of-life specifications as such knowledge will impact the component life assessments and cost requirements. At the current state of vehicle development, it is unlikely the three U.S. OEMs would be willing to share such sensitive information. Thus, it is not clear as to how third-party estimates are of value. Last, the DOE-supported analyses of very low volume HFCV production is of little value.

Durability

Stack durability (target: 5,000 hours) as presented by DOE in Table 3-7 and Figure 3-10, like cost, has been met but only at the expense of higher catalyst loadings. Since the NRC Phase 4 review, average stack life reported to the committee by the DOE based on on-road NREL data (Kurtz et al., 2015) has increased to an average of 3,900 hours from six HFCV on-road tests. This figure is up 2,500 hours from that reported in the NRC Phase 4 review. With recent OEM HFCV advancements, including new technologies, new modes of operation, and from proprietary accelerated testing by the OEMs themselves, it is not clear if the various non-OEM organizations performing lifetime and stack performance tests are using the same hardware and test protocols, performance, and end-of-life metrics as the OEMs, and hence these numbers may not be representative of the state of the art.

Repeated themes throughout this review are that stack components are complex and that how they are assembled into stacks and then operated impact stack

TABLE 3-7 Department of Energy Reported Membrane Electrode Assembly (MEA) Performance and Durability Metrics

NOTE: 5,000 hours of operation under simulated vehicle power cycling and shutdown/start-up cycling with less than 10 percent loss in rated power. Specifically, developing MEAs with state-of-the-art catalysts that demonstrate performance greater than 1 W/cm² with Pt loading less than 0.125 mg/cm².

SOURCE: Allendorf and Borup (2016).

durability. For example, membrane as well as catalyst ionomer hydration characteristics, electrode layer structure including porosity, composition, conductivity (electrical and ionic), reactant diffusion rates through the membrane, impurities, and many additional parameters can impact performance and durability. Operating modes of the stack can and will affect the chemical, mechanical, and physical nature of all of the above, thereby making control systems critical in developing a viable, long-life, durable stack. Each OEM is addressing these issues with their respective proprietary approaches. As reported to the fuel cell subgroup and not available to the public, GM, through extensive investments in infrastructure testing and analyses methods, has indicated significant advancements in stack durability. With that said, GM indicated that DOE support of new electrolytes for the membrane as well as the ionomer, and related topics, though long term, are warranted and should continue because membranes and catalysts are critical, and if possible, with increased levels of funding.

The most recent data in the public domain as presented to the committee are shown in Table 3-7 and Figure 3-10. Of particular interest is the durability with cycling data reported in hours in Table 3-7 and the average fuel cell operation data presented in Figure 3-10, also reported in hours. From these data, it is clear the durability of the membrane electrode assemblies (MEAs) of 2,500 hours is substantially above the average fleet durability lifetime reported (see Table 3-7 and Figure 3-10 references) to be approximately 2,000 hours. The fact that there is a difference in the two figures is anticipated, as actual vehicle data will include the effects of all aspects of the “system” on fuel cell stack performance, unlike data derived from single cell and stack testing in a laboratory. Although it is encouraging that the maximum fleet durability data are approaching 4,000 hours, there has been little change in the average fleet durability data since 2010 as presented in Figure 3-10. The information derived from the fundamental studies of MEAs will continue to provide valuable insight and guidance to the OEMs and the MEA suppliers as will the OEM vehicle testing from a systems and actual on-road perspective. It is anticipated that the recently formed FC-PAD consortium will contribute to the development a 5,000-hour-life MEA.

As previously mentioned, the FC-PAD consortium is chartered to address the durability question from a fundamental science and engineering perspective, with efforts from OEMs, industry, and academia. It will be up to DOE using the FC-PAD consortium to allocate funding in an effective way to make this a relevant program. At the minimum it will help coordinate the project portfolio.

Assessment of Progress and Key Achievements of the Partnership

Understanding fuel cell stack performance, the subsystem where the electrochemical processes take place, is complex, as it is a function of the membrane characteristics, catalysts, stack design, environment, and the overall fuel cell system architecture and mode of operation. Specifically, the durability, as stated

previously, is predominantly a function of the chemical stability of the catalyst, catalyst layer composition, membrane content, membrane physical and chemical stability, and operating environment. Work on carbon-free supported catalysts, chemically and mechanically stable membranes that can operate at a lower humidity, and the impact of ionomer properties and interfacial catalyst layers and membrane stability, is essential and has continued to receive support in this review period as recommended in prior NRC reviews.

In consideration of the preceding, there are a number of promising projects funded by the DOE that could make a significant impact on commercialization efforts by member OEMs including those listed here.

The FCTT of the Partnership recognizes that catalysts represent roughly 50 percent of the membrane-based stack cost and that reducing catalyst cost remains a key challenge to lowering the cost of fuel cells. With this in mind, the FCTT has encouraged partners to focus on reducing the PGM content of fuel cell catalysts, increasing the activity of PGM catalysts without sacrificing their durability, and exploring non-PGM catalysts and novel catalyst supports. The efforts toward these objectives have resulted in notable progress, as exemplified by the following illustrations:

• Argonne National Laboratory (ANL) has developed a method for mesostructuring PtNi thin films so that they exhibit 20x the specific activity of Pt nanoparticles supported on carbon.
• GM has found that slow de-alloying Pt₃Ni nanoparticles to remove Ni and create a Pt skin results in a mass activity higher than 0.44 A/mg_Pt for cathode loadings of approximately 0.1 mg_Pt/cm².
• ANL and Lawrence Berkeley National Laboratory (LBNL) have developed methods for producing Pt₃Ni nanoframes that when supported on carbon exhibit an oxygen reduction reaction activity of over 5 A/mg_Pt at 0.9 V in 0.1 M HClO₄ and show little loss in activity after 10⁴ cycles.
• Brookhaven National Laboratory (BNL) has investigated the performance of catalysts comprised of a Pt monolayer deposited over a PdNi core that exhibits good activity when incorporated into a membrane and electrode assembly.
• ANL has used structure-property modeling methods and new synthesis routes to produce new compositions of non-PGM catalysts.
• Los Alamos National Laboratory (LANL) has developed an engineered ionomer topology to facilitate proton transport, which when coated over Pt supported on multiwalled carbon nanotubes as a thin film led to enhanced oxygen transport to the Pt nanoparticles and targeted polarization curves at a Pt loading of 0.05 mg_Pt/cm².

Progress has also been made in developing a better understanding of the electrode structure, composition, performance, and degradation mechanisms as

exemplified by the following two citations. Myers (2016) continues to build on the knowledge regarding catalyst support degradation mechanisms as well as the behavior of platinum under all operating conditions. More (2016) has focused on the ionomer used in the electrode layer. Following the recommendations of the NRC Phase 4 report, activities in this area have been added to include projects addressing non-PGM catalysts, enhanced imaging methods of the ionomers in the electrode and gas diffusion layers, ultra-low catalyst loadings, especially for the cathode, continuation of core shell studies, and catalyst pretreatment methods. Of concern to the committee is that many of these programs have been completed in late 2015 and it is unclear at this time which projects will continue (even within the recently announced FC-PAD consortium).

The portfolio of research projects also includes a spectrum of efforts from topics including gaining a fundamental understanding of critical issues related to the stability of Pt-containing catalysts, the interactions of ionomers with Pt nanoparticles, and the role of the ionomer in controlling oxygen transfer to the surface of the catalyst, an issue that becomes critical as the loading of Pt decreases. Efforts devoted to cathode performance at the system level, that take into account the effects of oxygen and proton transport as well as ORR kinetics as a function of the properties and interactions of the catalyst, support, and ionomer, and the interactions of catalyst-support-ionomer interaction with the membrane as a complete “system” are of significant value. Given the national laboratory facilities for doing atomic- and meso-scale characterization of components and complex systems, future research devoted to these subjects will add to the critical understanding of all classes of catalysts used in the electrode. The new consortia are expected to address these critically important activities.

A notable characteristic of a number of the preceding examples and many others reported at the DOE AMR meeting is their close relationship to technologies now believed to be under consideration for use by the OEMs, both foreign and domestic. This characteristic of the work is a consequence of requests from industrial partners in U.S. DRIVE for assistance in looking at specific issues related to in-house technology development and the desire to leverage high-end surface science and other spectroscopic characterization techniques and methods available at the national laboratories.

The discussion presented here highlights that communication of technology requirements and project results among all interested parties, Partnership related or not, must remain a priority. The DOE uses the AMR very effectively to accomplish this, including peer-review ratings of each project that are then used to assess future funding and support of a given project. A concern to the committee is that a number of the projects listed earlier at the national laboratories are ending or will soon be completed.

Significant Barriers and Issues That Need to Be Addressed

As in previous NRC reviews of the Partnership, cost and durability remain the critical barriers. With cost and durability still key issues, and knowing that the cost issue actually translates to new technical solutions, the committee understands the degree of effort and time required for the OEMs to meet the targets. With that said, the committee sees that appropriate DOE-funded projects are already in place. The FC-PAD presents an opportunity to address communication, coordination, program management, and teaming, and minimize duplication of effort and best use available resources, which will improve the chances of success in overcoming these challenges. Durability issues are predominantly materials-based and a function of system architecture and operating mode. New materials and solutions will result only from the continued focus on understanding the fundamentals of degradation mechanisms coupled with the continued investigations of carbon-free supports.

Water management in fuel cells has a dramatic impact on cost and durability as well as overall systems architecture. Though solutions exist in the current generation of vehicles, advanced humidification and water management schemes could reduce the complexity of the fuel cell if membranes could be developed that have a lower water requirement for proton transport. This will assist in simplifying system architecture (cost) and possibly allow for higher-temperature operation resulting in smaller and more efficient cooling systems. Electrode layer architecture and composition would also be positively impacted. The current funding level of new electrolytes for both the membrane and ionomer used in the electrode layer may not be sufficient and the DOE should consider rebalancing its portfolio to ensure that these areas are given proper emphasis.

Near-Term versus Long-Term Implications of DOE-Funded R&D

As has been pointed out previously, there is not a quantitative mechanism to assess the value and success of the DOE-funded efforts and the advancement of vehicle technologies via the Partnership. With that said, in recent years the DOE has funded approximately 56 projects, 42 related to fuel cell hardware, and 14 related to systems analysis, as reported at the 2015 AMR. Such projects address a spectrum of R&D efforts (e.g., at different technology readiness levels [TRLs]) that are predominantly long term (low TRLs) in nature—for example, new membranes, electrodes, catalysts, compressors, plates, operating performance under varying environmental conditions, to name a few. The longer-term projects align well with the objectives of the Partnership—that is, the projects are precompetitive, long term and high risk. Yet in any development effort, nearer-term learnings are natural fallouts and have the potential to bring immediate value to the OEMs as well as other parties, including the supply chain. It is important to distinguish between the two and maintain the proper perspective on the primary objective of such R&D efforts.

Appropriateness of Federal Funding

With the U.S. vehicle market moving toward near-zero-emission electric vehicles, the committee feels that it is appropriate that there is a portfolio of electric power generation sources under consideration, including those that utilize fuel cells. Therefore, the committee feels that federal funding is appropriate to facilitate technical solutions through support of R&D activities related to transportation. Though it is still premature, eventually a return of investment analysis⁷ and an assessment of the benefits of the federally funded R&D by the DOE will be performed so as to assess the value of these programs to the United States.

Responses to Recommendations from the NRC Phase 4 Review

Committee Assessment of Response to 3-3. Based on the data presented in Figure 3-5, the committee feels that the FCTO has been responsive to NRC Recommendation 3-3 of the Phase 4 report, especially for funding next-generation catalysts and electrodes. With respect to cell and other stack hardware and membrane electrode assemblies, funding levels saw an increase (2012-2014) in activity but then were essentially zeroed out in 2016. Last, funding for membranes and electrolytes have dropped significantly from the 2013 period to 2016, as previously shown in Figure 3-5. This is a concern of the committee.

___________________

⁷ See Chapter 2, the section “The Role of the Federal Government.”

Committee Assessment of Response to 3-4 and S-4. The committee feels that this area has been addressed within this review period, especially with the systems analysis group, yet durability and cost (from a technical perspective) solutions have not yet emerged from such efforts. The committee questions the value of performing low-volume (100,000 units per year) manufacturing analyses.

Committee Assessment of Response to 3-5. The committee agrees that there has been good coordination of fuel cell development activities among the different offices as determined from the report outs at the Annual Merit Review meetings. The committee also agrees that innovative topics, as recommended in the NRC Phase 4 report, have been supported as cited in the Partnership’s response.

Committee Assessment of Response to 3-6. The committee feels the response has been appropriate.

Findings and Recommendations

The overall assessment based on the review is that the U.S.-based OEMs, with significant input from the Partnership, although in different states of development, have advanced fuel cell technology to the point where at least one U.S. DRIVE Partnership OEM (General Motors) is anticipating a rollout of its HFCV in 2020. With cost and durability challenges remaining, the DOE appropriately, and in response to the Partnership’s mission, continues to support longer term, precompetitive R&D projects. Furthermore, the DOE is adding additional activities at the national laboratories with the creation of consortia which will help focus and coordinate the R&D.

Finding 3-16. Since the NRC Phase 4 review, Toyota, Hyundai, and Honda have made available within the United States a limited number of fuel cell vehicle sales or leases to the general public. U.S.-based OEMs, with significant input from the Partnership, although in different states of development, have advanced fuel cell technology to the point where at least one U.S. DRIVE Partnership OEM (General Motors) is anticipating a rollout of its fuel cell vehicle in 2020. The development and deployment of roadworthy fuel cell vehicles is a major accomplishment and one that will help to identify remaining technical, cost, manufacturing, and infrastructure challenges. Though the cars are still in the late stages of development, the fact that the cars have advanced to this point is due in part to R&D coordination by the Partnership and its prior organizations, as well as from decades of funding of pertinent research projects by the DOE and Partnership members.

Finding 3-17. With the U.S. OEMs in different states of fuel cell vehicle development, and with competitive dynamics emerging, selected Partnership (fuel cell) goals and targets are relevant to only some of the OEM members (e.g., Pt loadings). Furthermore, it appears that there is a fine line between what might be considered near- and long-term projects based on the state of development of a given OEM’s technology.

Recommendation 3-5. The Partnership should evaluate projects for their near-term or long-term potential impact and assign technology readiness levels to them. The Partnership should continually assess its process for prioritizing projects and should continue to address the longer-term, precompetitive (lower technology readiness level) objectives and should update and set longer-term targets.

Finding 3-18. DOE is in the process of forming the Fuel Cell Performance and Durability consortium and the ElectroCAT consortium that will focus on fuel cell electrode durability science and electrochemical engineering, results of which should positively impact fuel cell performance and lifetime (durability).

Recommendation 3-6. The newly formed Fuel Cell Consortium for Performance and Durability and Electrocatalysis consortium should have an active oversight committee comprised of a panel of non-national laboratory experts, including representation from the original equipment manufacturers, to ensure that relevant problems are being addressed. Department of Energy-funded projects should focus on developing understandings of the fundamental issues relating to reduction in platinum loading to fuel cell durability.

Finding 3-19. Significant funding resources continue to be directed at carbon-supported Pt-based electrodes. Projects related to non-PGMs are in early stages of emphasis.

Recommendation 3-7. The Department of Energy should increase its focus on non-carbon-supported platinum catalysts and noncarbon catalyst support materials. In addition, the Department of Energy should increase the effort on developing non-platinum-group-metal catalysts and electrodes.

ONBOARD HYDROGEN STORAGE

Background

The mission of the hydrogen storage technical team (HSTT) is to “accelerate research and innovation that will lead to commercially viable hydrogen-storage technologies that meet the U.S. DRIVE Partnership goals” (U.S. DRIVE, 2013g). HSTT is one of two joint vehicle-fuel technical teams with a line to both the Vehicle Operations Group and the Fuel Operations Group within the U.S. DRIVE Partnership. The HSTT members are from the three car companies (GM, Ford, and FCA) plus representatives from the DOE and the national laboratories. The technical team has one associate member from a university. They have key interactions with several other technical teams, namely fuel cells, hydrogen production, hydrogen delivery, fuel pathway integration, hydrogen codes and standards, and vehicle systems and analysis (see Figure 2-1 in Chapter 2). These interactions

involve such issues as fuel quality, materials compatibility, storage capacity, and safety.

Vehicle driving range and fueling time are important customer attributes for fuel cell vehicles. The objective is to achieve a driving range of at least 300 miles for a full range of light-duty vehicles and at the same time meet performance, packaging, cost, rapid fueling time, and safety requirements.

A dual approach is being taken to address these goals. The near-term technology focus is on the 700-bar compressed hydrogen system. The longer-term approach is cold/cryo-compressed hydrogen storage systems and materials-based storage systems. Early in the program (2005-2010) materials R&D was conducted through three hydrogen storage centers of excellence, each managed by one of the national laboratories and each one focused on one of three technology areas: metal hydrides (led by SNL), chemical hydrogen storage (led by LANL), and sorbents (led by the NREL). These centers facilitated communication among research groups, developed best practices for research and a mechanism to evaluate progress, and were succeeded by the Hydrogen Storage Engineering Center of Excellence (HSECoE) led by the Savannah River National Laboratory, which ran from February 2009 through December 2015. This center was focused on engineering analysis and system integration modeling. The design and build of subscale prototype systems was used to validate the models and improve designs, and as a predictive tool to define material properties to meet the DOE system targets. These centers of excellence are now closed.

Recently a new consortium has been formed called the Hydrogen Materials—Advanced Research Consortium (HyMARC) which couples basic and applied R&D with advanced computational materials design. The approach taken is to have a core laboratory team comprised of SNL (lead), LLNL, and LBNL that conducts fundamental science and develops resources for the R&D community. The other component is comprised of individual projects selected through Funding Opportunity Announcements/Lab calls. Material characterization and validation resources are at NREL (lead), LBNL, Pacific Northwest National Laboratory (PNNL), and the National Institute for Standards and Technology. The long-term focus will be on multiscale simulations. The goal is to have an improved and faster approach to identify promising materials. The laboratory teams make use of DOE through the Office of Science (the Basic Energy Sciences [BES] and Advanced Scientific Computing Research [ASCR]) user facilities.

The U.S. DRIVE HSTT roadmap was last issued in June 2013 (U.S. DRIVE, 2013g). This roadmap contains detailed information on key issues and challenges, technical targets, technical barriers, and R&D strategies to achieve the technical targets.

DOE funding for the onboard hydrogen storage activities are bundled in the hydrogen fuel R&D line item, which includes hydrogen production and delivery R&D and hydrogen storage R&D. The DOE funding for hydrogen fuel R&D was $41.05 million for FY 2016 (Satyapal, 2016).

The DOE hydrogen storage projects as reported at the annual AMR reviews comprise a significant part of the preceding funding, which has remained constant at $15.6 million from 2014 through 2016 annually (Stetson, 2014, 2015, 2016). The requirement that projects be fully funded at the start of the project (a recent change) makes year-to-year comparisons difficult.

The current status of onboard hydrogen storage technologies is shown in Table 3-8. The storage targets are based on 5.6 kg of usable hydrogen and an 80 kW net power fuel cell. This status can be compared with the storage targets established for 2020 and the U.S. DOE ultimate targets. The technical targets have been revised since the NRC Phase 4 report was issued (NRC, 2013). Progress has been made in both near-term and long-term technologies. No-go decisions made by the DOE have provided for a focused effort on new and promising technologies.

Assessment of Progress and Key Achievements

The projects that address onboard hydrogen storage include both near-term and long-term technologies. The near-term technologies include issues related to cost and performance of the carbon fiber-based hydrogen storage vessels. The longer-term technologies include technologies related to the cryo/compressed storage vessels and storage materials with emphasis on meeting the U.S. DRIVE targets. The U.S. DOE Fuel Cell Technologies Office provided the committee with

TABLE 3-8 Current Status of Hydrogen Storage Technologies

^a Projected to 500,000 units/year.

SOURCE: Stetson and Veenstra (2016).

a list of projects related to onboard hydrogen storage that it funded in recent years (Table 3-9). The committee sorted these by topic for clarity.

The carbon fiber resin and matrix represent more than one-half of the cost of the compressed storage system for the carbon-fiber-based hydrogen storage tanks. Progress has been made through the use of low-cost textile-based polyacryrlonitrile (PAN) precursors to reduce the cost of carbon fiber. The PAN precursor represents an 18 percent reduction in carbon fiber cost which brings the carbon fiber price down to $10.6/lb from a 2013 baseline cost of $13/lb. Both costs assume an annual production of 25,000 tons. Progress has also been made on reducing system cost based on balance of plant component integration. There is a strong synergy between the carbon fiber interests of the onboard hydrogen storage technical team and the interest in lightweight materials of the materials technical team.

Although the progress in cost reduction of compressed hydrogen storage tanks quoted by the DOE is promising, it should be noted that the Toyota Corporation has published a recent technical paper (Yamashita et al., 2015) in which it claims that its latest compressed hydrogen storage system is lighter in weight (15 percent reduction) due to the use of advanced carbon fiber, new winding designs, and simplified balance of plant components. This tank design, currently used in limited production vehicles, is reduced in cost compared to the previous 2008 tank used in prototype fuel cell vehicles. It has also been validated for drop and fire safety regulations. Additional details on the absolute numbers for weight and cost are needed in order to fully assess these claims.

There are numerous reasons for the DOE to invest in R&D programs directed at improving carbon and glass fiber strength and reducing cost for applications not only in the automotive industry but in other industries requiring improved efficiency. However, improving the design of the compressed gas storage tanks (project ST101) and reengineering the balance of plant (projects ST005 and ST113) is an effort that is currently being conducted by both OEMs and suppliers, making it unnecessary for the DOE to conduct programs directed at these objectives. Allowing the engineering development to be done by industry would allow the DOE to focus on developing solid-state storage materials and advanced composites.

Cold/cryogenic compressed hydrogen storage is being pursued as a long-term option for compressed hydrogen storage. Potential advantages are lower operating pressure, higher energy density, less carbon fiber for tank construction, and lower system cost depending on the approach taken (see Figure 3-11). The two approaches are as follows: The cryogenic compressed system has a target volumetric capacity of >50 g/L hydrogen and >9 weight percent (wt%) H₂ system capacity at 700 bar and 40 K in contrast to 24 g/L for the 700-bar system at 295 K. The same amount of usable hydrogen could thus be stored in about 50 percent less volume (a smaller tank and at lower cost). The cold compressed hydrogen storage system has a target gravimetric capacity of 8-9 wt% H₂ at

TABLE 3-9 U.S. Department of Energy Fuel Cell Technologies and Vehicle Technologies Offices Active Project List Related to Onboard Hydrogen Storage That Supports U.S. DRIVE Targets

NOTE: Acronyms defined in Appendix D.

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a).

500 bar and 200 K compared with 4.4 wt% H₂ for the 700-bar system at 295 K. The same amount of usable hydrogen could thus be stored with 50 percent less system mass because of the lower pressure. During the past year the focus of the cryo/compressed project included pressure vessel design, fabrication, and testing, including electricity consumption on refueling. The energy efficiency of the total system operation (including the energy required for hydrogen liquefaction) is a major factor that needs to continue to be addressed.⁸

Materials-based storage continues to hold promise in the long term; however, many of these efforts have been redirected as they will not meet the 2020 targets. Work on the metal hydride system (e.g., sodium alanate) was discontinued based on HSECoE model projections. The materials under study lacked suitable thermodynamic, kinetic, and gravimetric properties. Progress was reported on an electrochemical process to reduce alane synthesis cost. Work on the chemical hydrogen storage system (e.g., ammonia borane) was discontinued due to lack of liquid phase materials with suitable regeneration processes. The committee supports and commends these programmatic decisions, which were based on model projections and clear insurmountable barriers.

Work continues on the sorbent system (e.g., MOF-5) for cold 80-160 K, 100-bar storage. Work in progress includes heat exchanger concepts and prototype testing. The HSECoE has developed system models that can be applied to materials, storage tank mass, and cost. The system models that capture the hydrogen storage systems, the fuel cell system, and model at the vehicle level are available on the HSECoE website.⁹ The work scope of several projects is supported by the HSECoE.

The Highlights of Technical Accomplishments Report (U.S. DRIVE, 2015) cites two areas of progress in the area of hydrogen storage. These are (1) Adsorption-Based Hydrogen Storage System Validated and (2) Lower-Cost Hydrogen Storage System. In (1) HSECoE subscale prototypes were used to validate the models and improve their predictive capability. Two prototype systems were used to characterize materials performance, internal heat exchange capability, and thermal management. This approach was applied successfully to confirm the robustness of the (MOF-5) adsorbent material. The second accomplishment cited (2) is the lower cost PAN precursor-based textile processing mentioned above.

Onboard hydrogen storage is cited in the DOE program reviews to be of interest to several technical teams, even beyond HSTT. For example, the materials technical team is addressing hydrogen storage tank materials, the hydrogen codes and standards technical team is addressing safe deployment, and the hydrogen delivery technical team shares needs for hydrogen storage. Collaboration among

___________________

⁸ Estimates of the energy required for liquefaction vary and can be found in the following DOE document: https://www.hydrogen.energy.gov/pdfs/9013_energy_requirements_for_hydrogen_gas_compression.pdf.

⁹ The website for the Hydrogen Storage Engineering Center of Excellence is http://hsecoe.org, accessed September 12, 2016.

the technical teams is increasing as the various technologies and commercialization needs mature.

HSTT participates in and monitors international hydrogen storage activities. These activities include the International Energy Agency—Hydrogen Implementation Agreement (Task 32) and international conferences.

Significant Barriers and Issues That Need to Be Addressed

While significant progress has been made toward the 2020 targets, the major barriers that need to be addressed are gravimetric density, volumetric density, and system cost for the projected 700-bar-type system. Practical limitations prevent the 700-bar system from meeting all of the onboard hydrogen storage targets. (The use of larger or multiple tanks in order to carry more hydrogen on board the vehicle is viewed as a solution for some applications.) Targets must be met across a wide range of vehicle platforms and especially for those platforms where the benefits are greatest. (Yamashita et al. [2015] cites gravimetric density exceeding the 2020 target.)

Significant barriers that are the focus of ongoing R&D are specific to each technology area.

• The near-term 700-bar compressed hydrogen system requires lower-cost carbon fiber, improved composites, innovative winding designs, conformable designs, and lower-cost balance of plant.
• The cold/cryo-compressed hydrogen storage system requires advanced insulation, improved dormancy, and composite development.
• The materials storage systems (metal hydrides, sorbents, and chemical H₂ storage) all need lower system cost and higher material capacity. Fill time and onboard efficiency is an issue for the metal hydrides (e.g., NaAlH₄). Dormancy and well-to-power plant efficiency is an issue for the sorbents (e.g., MOF-5). The chemical hydrogen storage systems (e.g., NH₃BH₃) need lower-cost off-board regeneration.

The need to overcome these barriers does not preclude vehicle introduction because advancements in overall power train system efficiency and vehicle weight reduction can make up for a shortfall in hydrogen storage capacity.

Responses to Phase 4 Recommendations from Phase 4 Review

Committee Assessment of Response to 3-7. The committee considers that the actions taken by the Partnership regarding high-pressure onboard hydrogen storage and systems analysis of materials-based technologies to be fully responsive to the recommendation.

Committee Assessment of Response to 3-8. The committee supports the involvement of the other technical teams in addressing the relationship between the onboard hydrogen storage tank pressure and the hydrogen infrastructure and concludes that it is fully responsive to the recommendation.

Committee Assessment of Response to 3-9. The committee believes that the activities regarding the partnerships with other government agencies and DOE offices regarding lower-cost aerospace quality carbon fibers are responsive to the recommendation.

Committee Assessment of Response to 3-10. The committee considers that the actions taken regarding the development of rigorous and comprehensive design and testing protocols to ensure the safe design and manufacture of compressed hydrogen tanks in conjunction with industry and government stakeholders are responsive to the recommendation.

Committee Assessment of Response to 3-11 and S-5. The committee believes that the coordination of activities within the DOE with BES and ARPA-E with respect to hydrogen storage is a good first step and looks forward to a more robust interoffice initiative to pursue this critical task.

Appropriate Federal Role

The DOE hydrogen storage materials projects are an appropriate role for federal funding given the benefits of fuel cells including high efficiency and low greenhouse gas emissions when the hydrogen fuel is made from renewable energy sources. Advanced hydrogen storage technologies need to be developed that meet system cost targets and onboard storage targets for volumetric density and gravimetric density. DOE-funded precompetitive R&D can lead to both new fundamental understanding and new technologies in order to meet the cost and capacity targets. The national laboratories, university laboratories, and selected industry laboratories are well equipped to contribute to this effort. Tank design and development is being conducted at this time by OEMs and suppliers and is being implemented in production vehicles. As such these are now competitive activities.

Findings and Recommendations

Finding 3-20. The 700-bar hydrogen storage tank will be the technology used for the immediate future and this technology, although continuing to evolve, will be used in production vehicles. Materials-based storage has been extensively researched in a well-organized DOE program. Materials that still hold promise and new discoveries could have a high hurdle if they are to displace technology in place. The new DOE Consortium HyMARC will address scientific gaps blocking the advancement of solid-state storage materials. The HSTT roadmap was last updated June 2013 and significant progress and changes have been made in recent years, including technical targets.

Recommendation 3-8. The Department of Energy should not be involved in design, development, or testing of production tank components, but the following actions are needed:

1. The mission for the hydrogen storage technical team should be made more explicit in terms of meeting critical needs and a plan should be developed for reaching these goals and targets.
2. The hydrogen storage technical team roadmap should be updated by the Partnership again in light of recent developments.

3. The management structure for the Hydrogen Materials—Advanced Research Consortium should provide for coordination, prevention of duplication, and decision-making authority over the consortium projects.

Finding 3-21. All the goals for onboard hydrogen storage have not been met, and basic scientific research has not produced an easy solution to date. Yet, onboard hydrogen storage is an issue for several technical teams and working groups beyond the hydrogen storage technical team—for example, the materials technical team, the fuel cells technical team, the hydrogen codes and standards technical team, and the hydrogen delivery technical team. As the technologies continue to mature, the need to merge activities can be expected to increase because vehicle performance parameters might be achieved through a wider range of options than gravimetric and volumetric hydrogen storage density alone.

Recommendation 3-9. The hydrogen storage technical team should increasingly work with the other technical teams even beyond those areas where overlap currently exists.

HYDROGEN PRODUCTION, DELIVERY, AND DISPENSING

Background and Introduction

The supply and refueling logistics, associated technologies, as well as the cost of hydrogen at the pump are critical elements of the transformation process in which HFCVs displace fossil fuel-powered internal combustion engines. At the time of the writing of this report, there were approximately 25 fueling stations in the United States, mostly in California, based on a number of technologies. In selected areas, mobile refuelers are used by some OEMs. Up until now only a very small number of HFCVs have been on the road, yet with the projections by the OEMs, this number will increase rapidly in the years to come. Though in its infancy, the hydrogen refueling infrastructure must be addressed in its entirety, from generation to delivery, as well as the fueling station itself and ancillary activities such as safety, codes, and standards.

This section reviews the budgets, goals, targets, and accomplishments of the U.S. DRIVE Partnership related to hydrogen production, delivery, and dispensing. A discussion of issues and barriers and the role of the government in overcoming such barriers is also included along with committee recommendations.

As mentioned in the introduction, one of the focus areas of the U.S. DRIVE Partnership is to provide a coordination mechanism by which OEMs interested in HFCVs can communicate precompetitive technical issues and target specifications to its members. Among the operations groups defined in the U.S. DRIVE Partnership Plan, the Fuel Operations Group (see Chapter 2) to date has almost entirely been focused on hydrogen—its production, delivery to fueling stations, dispens-

ing, fuel pathway integration, and onboard storage. With the Partnership guidance and based on other inputs, the DOE has been funding a number of production and delivery projects as a means to begin the learning process for when the fuel cell vehicles start to appear in the market. However, while the U.S. DRIVE industry partners appear to be sufficiently engaged in vehicle-based developments relevant to the Partnership’s goals and targets, their involvement in activities related to the hydrogen supply chain development, including production, delivery, and dispensing, appears to be limited, especially since the NRC Phase 4 review.

Introduction of HFCVs in the market and presence of an adequate hydrogen infrastructure are often seen as a “chicken-and-egg” problem—that is, who will invest in hydrogen fueling infrastructure if there are too few HFCVs on the road and, conversely, who will offer HFCVs if there are a limited number of hydrogen fueling stations. The DOE program goal relevant to U.S. DRIVE is “to enable a commitment by automakers no later than year 2015 to offer safe, affordable, and technically viable hydrogen fuel cell vehicles in the mass consumer market” (Satyapal, 2016). This does not seem to have been fulfilled by U.S. DRIVE auto industry partners, except perhaps GM, although Toyota, Hyundai, and Honda are often quoted by DOE as examples of automakers with commercial plans. Hyundai and Toyota are already offering commercial HFCVs in the United States. Toyota and Honda are also engaged in refueling activities. Toyota has already begun a rollout of its fuel cell vehicles; Honda has begun leasing vehicles (Clarity) and together with GM are targeting to introduce their vehicles by 2020 (Greimel, 2016). U.S. auto manufacturers have not publicly announced similar plans as yet, although GM so far has the most number of HFCVs (119 Equinox) with more than 3 million miles of on-road driving experience (General Motors, 2014, 2016).

As pointed out before, of the total DOE EERE annual budget for hydrogen and fuel cells, which has been more or less stable over the last 3 to 4 years at about $100 million (± 5 percent), roughly $35 million is spent on hydrogen production, storage, and delivery, which is about the same as that for fuel cells (Sarkar, 2015).¹⁰ Furthermore, production and delivery budgets, including those for FY 2016, are approximately $12.5 million each. There is an additional budget of $7 million each for technology validation activities and for Safety, Codes, and Standards (SCS) for hydrogen as a fuel (Satyapal, 2016). Much of the hydrogen R&D effort is appropriately directed to reducing the cost of dispensed hydrogen by focusing on key steps in the value chain. However, as shown by the fuel pathway integration analysis results with the current technologies, it is challenging to meet the cost targets by the prescribed dates (Joseck and Verduzco, 2016). More efforts will need to be focused on new concepts and next-generation technologies.

Since hydrogen is the lightest gas it has a very low volumetric energy density. It therefore requires compression to high pressure (e.g., 700 bar) or liquefaction (20 K) to store and transport as well as store on the vehicle before use. Thus,

___________________

¹⁰ Also see Peterson and Farmer (2016).

hydrogen supply to an HFCV involves a number of steps from production to the vehicle tank. One example of the supply chain for gaseous hydrogen is shown in the schematic in Figure 3-12.

Hydrogen production from fossil-based feedstocks, especially by reforming natural gas, is commercially well established. In the reforming process, all the carbon in natural gas is ultimately converted to CO₂ and released to the atmosphere. To avoid emitting the CO₂ in this pathway would require capture and storage and/or utilization of the CO₂ (CCSU). This of course would add cost and make the hydrogen more expensive. The other current pathway for hydrogen production is through the electrolysis of water; the associated GHG emissions in this case are dependent on the source of electricity. Further reduction in GHG emissions requires use of renewable resources, such as a biomass feedstock and/or use of wind- and solar-based electricity. Other long-term approaches also being explored include solar thermochemical, photoelectrochemical, and microbial biomass conversion.

While the Partnership is focused on the long-term, precompetitive R&D, it should be borne in mind that the near-term needs of hydrogen are met by existing, established production technologies, primarily using natural gas reforming, that meet the DOE production cost target of $2/kg H₂. Purification will likely be a small part of the cost. However, current delivery and station costs estimated at $11-14/kg H₂ (Miller, 2016) are high, and are mainly responsible for the high

total cost of dispensed hydrogen of $13-16/kg H₂ compared to the 2020 DOE cost target for untaxed dispensed hydrogen of $4/kg. High-volume usage is expected to bring these costs down to the $5-7.5/kg H₂ range. But the challenges to meet the ultimate cost targets are related to the cost of “renewable” hydrogen production and more importantly the cost of delivery and dispensing (station cost). Therefore, priorities between long-term production R&D and near-term developments related to hydrogen delivery and dispensing are often difficult to balance, given limited funding available. DOE-funded programs have made progress to improve technologies for high-pressure gaseous hydrogen storage, delivery, and dispensing by addressing and optimizing the value chain. Another related area that is getting attention from a fueling station perspective is SCS. However, it is not clear if the costs associated with the SCS requirements (e.g., for additional equipment and insurance) are adequately accounted for in the cost estimates for dispensed hydrogen.

Critically important are infrastructure build-out activities necessary to support a commercialization pathway for HFCVs and to get real-world experience with these vehicles, which will help identify further R&D needs. GM, for example, which is actively developing fuel cells and fuel cell vehicles, expects states to incentivize fueling station build-out and considers existence of a viable fueling infrastructure a prerequisite to vehicle introduction.¹¹ Thus, due to uncertainty of demand and volume, hydrogen infrastructure is highly dependent on federal and state funding; automotive and energy industry investment toward these efforts is minimal. Without adequate funding/investment in hydrogen infrastructure, commercialization of HFCVs will be pushed even further back beyond 2020. The lack of a robust plan for a viable national hydrogen infrastructure is a significant deterrent to the deployment of HFCVs.

Given the high risks and costs associated with initial introduction of HFCVs, for-profit companies are reluctant to participate on their own, and therefore federal and state roles in funding such activities are critically important and appropriate. The cost challenges combined with practical challenges of installing a hydrogen infrastructure imply that a paradigm shift is required if HFCVs are to become a realistic option for consumers.

Current Status of U.S. DRIVE Goals and Targets

With the limited number of HFCVs on the road today, resulting in low demand for hydrogen, economy of scale advantage is not realized. As a consequence, at present, the cost of hydrogen at the pump is very high ($13 to $16/gal gasoline equivalent (GGE) or kg H₂) compared to the DOE cost target of less than $4/kg H₂ necessary for HFCVs to be competitive with other options, specifically HEVs (Sutherland and Joseck, 2015). Of the total cost, the current produc-

___________________

¹¹ The committee gathered this information during a visit to General Motors on June 21, 2016.

tion cost is relatively low, since most of the hydrogen is sourced from existing large production facilities using natural gas as the feedstock; a major portion of the cost is for delivery and dispensing. While U.S. DRIVE has a cost target for onboard hydrogen storage, it does not have a cost target for dispensed hydrogen; it is instead considered within the scope of the U.S. DOE R&D program. The DOE target is based on a calculation of threshold costs which shows that the total cost of dispensed hydrogen needs to be less than $4/GGE—same as a kg of H₂ on energy content basis (in 2007$)—to be competitive with other transportation options expected in 2020 (Ruth and Joseck, 2011). As shown in Table 3-10, of the total cost of dispensed hydrogen, roughly half is associated with production, and the other half with delivery and dispensing (including fueling station cost).

The calculation makes certain assumptions with regard to the incremental cost of HFCVs based on projected improvements in vehicle technologies (both HFCVs and the competing HEVs), as well as the projected gasoline price in 2020 and fuel economy (mileage for HFCVs and HEVs). Since Ruth and Joseck’s report was published in 2011, there have been changes in the base case values. For example, the calculation uses fuel economy estimates of 59 miles/GGE for HFCVs and 42 miles/gal for HEVs. However, as per the recent presentation to the committee by Toyota, the HEV (Prius) fuel economy has improved from 42 mpg in 2010 to approximately 55 mpg in 2015 (Wimmer and Gazelle, 2016). The threshold cost calculations will need to be revised based on expected future improvements in the fuel economy of HEVs.

To avoid the high cost of delivery, on-site reforming has been evaluated in multiple demonstration projects as an option. However, current low demand has precluded a favorable cost benefit that could be realized through high-volume manufacturing and standardization of components. Also, on-site hydrogen generation by reforming may face operational challenges, and there may be greater need to store excess hydrogen at the fueling station to manage supply/demand balance.

Electrolysis, seen as the ultimate pathway for renewable hydrogen production, is still an expensive option. The basic technology for the water electrolysis

TABLE 3-10 Apportioned Target Costs for Centralized and Distributed Hydrogen (H₂) Production and Delivery in 2020

^a Based on a maximum hydrogen threshold cost of $4.00/GGE.

^b Dispensed, but untaxed.

NOTE: GGE, gasoline gallon equivalent.

SOURCE: Weil et al. (2012).

pathway is commercially well established. Current efforts to reduce the capital cost and improve the overall efficiency of electrolyzers to meet the production cost target are focused on improvements in the MEA, flow fields, gas diffusion layers, bipolar plate design, catalysts, membranes, etc. (Satyapal, April 2016; presentation to the committee, slide 21). The DOE Multiyear R&D plan (e.g., FCTO MYDD Production, 2015, Table 3.1.2) also points out materials development and improvements in manufacturing processes, as well as reduction in cost of electricity needs to meet the cost targets for renewable hydrogen production by electrolysis. This is challenging, since the scope for capital cost improvement is limited by performance and safety requirements and the efficiency improvement is limited by thermodynamic and other system constraints. Current electrolysis units are already highly efficient (at approximately 50 kWh/kg H₂ compared to a thermodynamic limit of 39.73 kWh/kg H₂), and any gains will be only incremental. As a consequence, even with the DOE long-term efficiency target of 43 kWh/kg H₂, the cost of electricity to run the unit is the most critical (greater than 50 percent) cost factor. Furthermore, electrolysis plants do not enjoy the same economy-of-scale benefit as is possible for chemical plants, such as for natural gas reforming or gasification of biomass and/or coal, due to its modular additive nature. This is evident in the capital cost projections for central and distributed electrolysis plants (Ainscough et al., 2014). As a result, as shown in Table 3-11, the projected future hydrogen costs for the base case electrolysis plants are about the same ($4.23/kg H₂—forecourt; $4.20/kg H₂—central), and are significantly higher than the 2020 target of $2.30/kg H₂.

With increased use of renewable electricity, hydrogen production by electrolysis and its storage and use may be an economical pathway to manage the irregular availability of renewable sources like solar and wind. If so, the use of such “off-peak” electricity could make hydrogen produced by electrolysis cost competitive.

The cost target for the delivery portion of the total hydrogen cost is less than $2/kg H₂ by 2020. The 2015 status, as per the hydrogen delivery technical team (HDTT) report, suggests that the lowest cost option at present is gaseous

TABLE 3-11 Hydrogen (H₂) Production High-Volume Cost Projections for Polymer Electrolyte Membrane Electrolysis Cases

SOURCE: Ainscough et al. (2014).

hydrogen (GH2) tube trailers with projected cost of $3/kg H₂ (Gupta and Soto, 2016). However, GH2 has payload limitations. So, as the demand grows, its utility is limited and other options need to be developed further to make them cost competitive. With the liquid hydrogen (LH2) pathway, the current liquid tanker delivery cost is projected to be $3.80/kg H₂, but hydrogen liquefaction is the most energy consuming and consequently the most expensive step. It is therefore the focus of the DOE R&D goals and targets. Novel liquefaction technologies are being developed to reduce the cost of liquefaction.

However, with the high intrinsic energy demand for hydrogen liquefaction requiring very low temperature (20 K), it is a challenging task. Another approach being pursued under the DOE program, called cryo-compressed storage, uses the higher density of hydrogen at lower temperatures to store more hydrogen by compressing and keeping cold hydrogen in the vessel. If compressed cryogenic onboard hydrogen storage is successful and the chosen option, the same technical approach would apply to hydrogen delivery.

Currently, a major part of the delivery cost is the cost of the fueling station with 700-bar capacity, and of that, the compression cost constitutes a significant portion, as shown in Figure 3-13, which presents the components of cost. DOE R&D efforts for this near-term approach are therefore focused on reducing the cost of compression and other station costs, such as ground storage.

Based on the progress to date and the remaining gaps to achieve the ultimate cost targets, it is apparent that novel technologies are necessary to overcome the

challenges to reduce the cost of hydrogen storage and delivery. One approach mentioned in the HDTT Roadmap relates to liquid hydrocarbon carriers (U.S. DRIVE, 2013f), although there is not much discussion of the pros and cons of such an approach. There were some early efforts within DOE, notably a concept developed by Air Products using carbazol compounds, which was later found to be impractical for technical reasons. However, this type of technology approach has not received sufficient attention within U.S. DRIVE, and a detailed techno-economic analysis along with in-depth exploration of the chemistry of this approach may be worth revisiting. According to a recent report, a German company called Hydrogenious Technologies (HT) has launched a commercial H₂ storage and logistics system using an innovative liquid organic hydrogen carrier technology (CryoGas International, 2016). Details of chemistry used, storage capacity, cost, and so on, are not provided. To enter the U.S. market, HT has signed an agreement with a U.S. hydrogen distribution company, United Hydrogen Group, for supply of these systems.

Another approach being investigated is to inject hydrogen in existing natural gas pipelines, which can then be withdrawn at desired locations after separating it from natural gas. This approach has its own challenges in terms of pressure management, pipeline integrity (materials compatibility), hydrogen separation and purification, public safety, and the associated costs (Melaina et al., 2013). This topic has not been addressed by the HDTT.

Progress and Key Achievements

U.S. DRIVE provided a list of DOE-funded hydrogen production and delivery projects relevant to the Partnership (see Table 3-12). The committee reviewed these projects presented at the 2015 and 2016 AMRs and has summarized the findings below along with a general discussion of the U.S. DRIVE technical team role and activities.

Hydrogen Production

The HPTT of the Partnership meets regularly and appears to have well-coordinated efforts to evaluate various production pathways and assess techno-economic viability. HPTT monitors a portfolio of technologies, as shown in Figure 3-14, which addresses near- and long-term needs. HPTT has identified appropriate technology areas that require focused R&D, and with their guidance along with other inputs, DOE is following a go/no-go methodology to downselect projects for continued funding. However, the committee feels that DOE could be more stringent about such decisions at various levels such as in technology areas and for specific projects. Examples of areas that may need more scrutiny are electrochemical compression, high-pressure electrolysis, and to some extent high-temperature electrolysis, as discussed later in this subsection.

TABLE 3-12 Department of Energy Projects on Hydrogen Related to U.S. DRIVE Goals

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a).

According to the HPTT report, hydrogen production with natural gas reforming, centrally or distributed, can meet the production cost targets with current low natural gas prices (Chapman and Randolph, 2016). This pathway also results in significantly lower GHG emissions (approximately 50 percent) per mile driven compared to petroleum-based fuels. Further reduction in GHG emissions with

this pathway would require CCSU since CO₂ is also produced in the natural gas reforming process, as mentioned before. Alternatively, pathways are needed to use with renewable resources, such as biomass conversion and electrochemical conversion using renewable electricity (e.g., wind, solar). Current costs of these pathways are much higher, and technology advances are necessary to bring the costs down. These are the conclusions of the extensive techno-economic analyses conducted by DOE (Chapman and Randolph, 2016). As a result, with the limited funding available, DOE program efforts are focused in these areas.

Based on the summary presentation at the committee meeting in February 2016, there has not been much progress reported in hydrogen production and delivery areas with respect to cost since the NRC Phase 4 review (Satyapal, 2016). Based on techno-economic analyses using the DOE H2A model, it was concluded that distributed natural gas reforming can meet the production cost target of less than $2/kg H₂ with a natural gas (NG) price of up to $5-6/GJ. To achieve lower GHG emissions with natural gas reforming it needs to be coupled with CCSU, which will add cost and has its own technical and economic challenges. Moreover, CCSU is only practical with large central plants; hence, distributed NG reforming is only a short-term solution. HPTT has evaluated the cost projection for CCSU (Chapman and Randolph, 2016). Accordingly, it constitutes a significant portion ($0.79-$0.96/kg H₂) of the total cost of hydrogen by reforming ($1.58-2.25/kg H₂). Moreover, the CCSU cost estimates are yet to be validated as the technology is still not fully developed, and the long-term viability of CCSU needs to be confirmed.

Among renewable hydrogen pathway options, only central biomass gasification is projected to achieve production costs closer to but more than $2/kg H₂ (the 2020 target). Other pathways, mainly electrolysis, continue to have significantly higher cost projections ($4-5/kg H₂). A key achievement in this area within the DOE program is a 40 percent reduction in PEM electrolyzer capital cost compared to a 2011 baseline as a result of bipolar plate innovations (Chapman and Randolph, 2016). However, the analysis shows that this cost reduction has no significant impact on the total hydrogen production cost since the major factor in the overall cost is still the actual cost of electricity (OPEX) to split water, which is outside the scope of any electrolyzer developments. As a result, slide 7 of the same presentation (see Figure 3-15) shows that the estimated cost of H₂ production with PEM electrolysis has not changed much from 2010 to 2015 for distributed or central plants.

Recognizing this challenge, two approaches are being investigated—electrolysis at high pressure to save on compression costs and electrolysis at high temperature to save on electricity cost. High-pressure electrolysis is an ongoing effort, and various companies have in the past attempted to commercialize this approach. There is a limiting pressure beyond which the benefits of high pressure are offset by the higher capital costs and safety issues. Thus, high-pressure electrolysis efforts have shown limited success. Since the NRC Phase 4 report, no significant progress has been reported on this approach within the DOE program.

An ongoing project by Giner, Inc. (AMR—pd117) has already concluded that high-pressure operation is generally uneconomical.

The other option proposed is high-temperature electrolysis (solid oxide electrochemical, or SOEC), wherein available, cheaper thermal energy is expected to reduce the electric energy requirements and hence the corresponding cost. The high temperature operation poses materials challenges and results in higher capital costs compared to PEM fuel cells. Hence, materials development is the focus of this program. The basic technology for this approach is the same as that for solid oxide fuel cells (SOFCs), which are being developed under the Solid State Energy Conversion Alliance program within the DOE Fossil Energy Office. Historically

SOFC developments date back more than 30 years, and a few years ago efforts to develop large-scale SOFC systems (multiple MW scale) were suspended by major developers due to insurmountable material stability issues.

James et al. (2015) of Strategic Analysis, Inc. (SA) provide cost projections for selected technologies, specifically PEM electrolysis, SOEC, and bio-fermentation, based on a detailed analysis. The results should help guide DOE decisions regarding focus areas and funding levels. Even with some aggressive assumptions with regard to potential reduction in capital cost and increase in efficiency for a large SOEC system (50 ton/day central plant), H₂ cost is projected to be near $4/kg, and it is shown that the key cost factor is the cost of electricity, making up more than 50 percent of the total cost, same as with PEM electrolysis. In view of the prior experience with such systems and the cost estimates, renewed efforts to conduct R&D of SOEC with federal funding are questionable and need to be reevaluated. Nevertheless a recent funding announcement from DOE shows that they are still funding three new projects (Ceramatec, FCE, and Giner) for the same general technology approach. One option to address the high cost of electricity required for hydrogen produced by electrolysis and to significantly reduce the cost of renewable hydrogen production is to use “excess” electricity produced by wind or solar, which is much cheaper; for example, it is estimated at $0.02/kWh or less compared to the average U.S. electricity cost of ~$0.11/kWh (see the section “Electricity as an Energy Source and the Grid,” later in Chapter 3). This approach, generally known as power-to-gas, is being pursued more aggressively in Europe, especially in Germany. In the United States, the first such project started in mid-2015 with NREL teaming with Southern California Gas Co. and National Fuel Cell Research center. ARPA-E recently announced funding of $2 million to Dioxide Materials for a similar project using their proprietary catalyst-based electrolyzer technology. A greater emphasis on this area would help overcome the electricity cost barrier to electrolysis mentioned earlier. A recent initiative with DOE leadership, called H2@Scale, includes this approach, along with a cost analysis, and shows the potential to reduce the cost of hydrogen by >45 percent to approach the target of $2/kg (HTAC presentation, April 6, 2016; IPHE Meeting, May 20, 2016).

A similar analysis of a biofermentation plant has resulted in an estimated H₂ production cost of $4.62/kg, assuming by-product credits. However, revised analysis by SA presented at the 2016 AMR showed the projected H₂ costs to be $58.53/kg and $5.65/kg, respectively, for the current and future cases. The main difference between the two cases is the assumption that the corn stover concentration of 175 g/L can be achieved in the future compared to 5 g/L for the current case. In this case, while the major cost contributor for the current case is the cost of heat required, capital and feedstock costs are the major factors for the future case with higher efficiency and heat generation by burning lignin. Given the aggressive assumptions in the preceding analyses, it seems challenging to achieve even the projected costs, and perhaps impossible to meet the 2020 DOE cost target of $2/kg with any of these options.

Based on these analyses, it is evident that a novel technology is needed to meet the cost targets or the cost target needs to be revised to reflect realistically plausible cases. In either case, a paradigm shift is required. In 2015-2016, DOE funded development of two new technology concepts, and SA has undertaken analysis of these to estimate corresponding costs. The two technologies are (1) bio-oil reformation using a monolith catalyst in a dual bed system with swing operation; and (2) NG reformation in a molten carbonate-based electrolyzer operated at approximately 600°C temperature, with simultaneous CO₂ removal to directly obtain approximately 98 percent purity H₂.

The committee understands that these are high-risk technology projects. However, lessons learned from the past should not be overlooked. For example, in the case of bio-oil reforming, it should be noted that the basic concept of conducting reforming and regeneration in a cyclic mode was extensively investigated before (e.g., GE, ExxonMobil, BOC). Even with natural gas feed, there were operational issues with thermal management, stability, contamination, and so on. Similarly, the reformer-electrolyzer-purifier approach is a variation on past efforts to remove CO₂ in situ in a reformer, which have not been so successful. SA projections should help DOE to determine viability of these approaches in terms of potential to meet the cost target.

Other long-term production pathways include wind-based electrolysis, photoelectrochemical (PEC), solar thermochemical hydrogen (STCH), and photobiological (PB) approaches; these are in early stages of development, with low efficiencies and high costs, and face more daunting challenges in terms of achieving the cost targets. Efforts in these areas are primarily focused on improving conversion efficiency, which is currently the main barrier. If target efficiencies are achieved, the pathways can be attractive since all of these use renewable energy and therefore lead to very low GHG emissions. Pathway analysis conducted by NREL shows that the projected cost of hydrogen production using the PB approach is very high (more than $9/kg) compared to other approaches mentioned (e.g., STCH: less than $4/kg) (Ramsden, 2015). Novel reactor and material concepts being developed for STCH at Sandia National Laboratories (McDaniel and Ermanoski, 2015) and by Professor Weimer at the University of Colorado Boulder (Weimer et al., 2015) show promising results. The approach is based on a two-step process wherein metal oxide is first heated by concentrated sunlight to high temperatures to reduce it and produce O₂. In the second step, the reduced material is exposed to H₂O to reoxidize it to its original oxidation state and produce H₂.

The materials and process conditions are being optimized in these projects to develop a more reliable and cost-effective process. Weimer’s approach consists of operating the reactor in a circulating mode similar to a fluid catalytic cracking reactor. This is challenging in terms of optimum reaction kinetics (material selection), material stability, and thermal management, but if successful, it has many benefits in terms of scale-up and economics.

If the final analysis confirms the preliminary results and if the scope for further improvement is limited with some of the approaches like PB, it would make sense to stop or scale back efforts on those that appear less promising and focus more on promising approaches like STCH. Another attractive approach is the microbial biomass conversion as it presents the opportunity to purify water simultaneously. This approach has not been included in the NREL study, but is being developed at NREL in collaboration with Penn State (Maness and Logan, 2015). There is also a new project begun in 2016 at Oregon State University (PD129) to develop a similar technology. These projects are not included in the list of U.S. DRIVE relevant projects. These are still early stage developments and should be pursued to advance the technology, albeit with an eye on economic viability. These findings are consistent with those of the HPTT.

Hydrogen Delivery

DOE-funded R&D projects are focused on improvements in high-pressure component performance and materials for low-cost fabrication. Efforts to reduce cost are largely through developments of new and improved methods for producing carbon fibers, which are used for the fabrication of lightweight, high-strength vessels for high-pressure hydrogen storage and transport as well as pipelines for hydrogen transport as a long-term option. Since high-pressure hydrogen transport has limitations in terms of pressure and payload capacity, there are also efforts to develop alternative approaches such as LH2 delivery. More recent efforts in hydrogen storage are focused on cold compressed hydrogen, called cryo-compressed, and use of adsorbents, especially in combination with low temperatures. Significant progress has been made in developing metal-organic framework (MOF) materials which, when used at cryogenic temperatures, can store high amounts of hydrogen. More discussion on hydrogen storage technologies can be found in the section on onboard hydrogen storage of this report. The same technologies are likely applicable for cost-effective hydrogen transport and delivery.

As mentioned earlier, the current cost of hydrogen at the pump is high ($13-16/GGE), and while high-volume usage will bring the costs down, to meet the ultimate cost targets advancements are required in technologies for hydrogen storage, delivery, and dispensing. Even with future projections, the cost of hydrogen delivery is a major contributor to the total cost of dispensed hydrogen. The delivery includes hydrogen processing (e.g., compression) and handling at the production facility, its transport to the fueling station and fueling station itself, which encompasses storage, further compression, and metering the hydrogen into the vehicle tank. All of these steps operating in tandem present operational complexities and add significant costs. DOE has conducted an analysis of key cost factors to identify and focus R&D efforts on critical areas. While the overall budget for the delivery area may be limited, the committee feels that the funds are being used appropriately with prioritization of R&D topics within the current

delivery approaches, and steady progress is being made. However, to meet the ultimate cost targets, a radically different approach may be necessary to overcome challenges and limitations of the high-pressure storage and transport pathway.

Within the current framework, a key achievement since the NRC Phase 4 report relates to a 25 percent reduction in the cost of gaseous hydrogen delivery from approximately $8 to approximately $6/kg H₂ with the development of high-pressure (500-bar) tube trailers by Hexagon to replace the conventional ones with less than 200 bar pressure, thereby increasing the payload, which results in the reduction of cost per kg H₂ (U.S. DRIVE, 2013f). Another accomplishment in the delivery arena is the demonstration by Argonne National Laboratory of a tube trailer consolidation strategy to lower forecourt compressor capital cost by about 60 percent (Gupta and Soto, 2016). Also Oak Ridge National Laboratory (ORNL) is developing a steel/concrete composite vessel for hydrogen storage on site at distribution terminals as well as at fueling stations, which can help reduce the overall delivery cost.

With the LH2 delivery pathway, seen as a near- to mid-term option, although delivery of LH2 is the most cost effective, liquefaction is the most energy consuming and expensive step. No significant progress has been reported since the NRC Phase 4 review for this pathway. But two new projects have been funded to develop novel concepts. One at NREL (Ainscough et al., 2014) involves use of a vortex tube for separation and simultaneous refrigeration to increase the efficiency of ortho/para hydrogen conversion, which is the most energy consuming step in hydrogen liquefaction. The other project at PNNL (Holladay, 2016) involves magnetocaloric refrigeration to improve the efficiency of liquefaction. This concept was originally developed by Prometheus with DOE funding. It remains to be seen whether the improvements envisioned in the new projects lead to a successful technology with potential to significantly reduce the cost of liquefaction.

A DOE-funded project (2007-2015) led by Fuel Cell Energy (FCE) attempted to develop electrochemical hydrogen compression (EHC) to reduce compression cost. While they improved cell durability, capacity, and efficiency, the final pressure achieved under practical conditions was limited to 3,000 psi (approximately 200 bar). Also, with an electrochemical cell as the building block, when scaled up, the compressor does not have the same economy-of-scale benefit as the mechanical compressors. Thus, the compressor cost can be a challenge. In the NRC Phase 4 report, it was suggested that the Partnership should evaluate this approach and compare it with conventional compression (NRC, 2013, p. 129). However, this has not been mentioned anywhere in the recent Partnership reports and presentations. Yet, surprisingly, there is a recent new DOE funding award announced to Giner, Inc., for developing EHC. Another project awarded to Greenway Energy is based on a novel approach combining two technologies: EHC and metal hydride compression (MHC). Development of MHC was attempted before by Ergenics and others; lessons learned from those projects should be helpful in developing the new concepts.

Independently, without any DOE funding, Linde has developed a novel hydrogen compression technology using ionic liquids to replace the mechanical piston, which has wear and tear issues (Mayer, 2014). The ionic liquid can be pumped to high pressures with much less energy, which results in lower energy costs. Also with the use of an immiscible liquid, the liquid can be in direct contact with hydrogen and the compression can be nearly isothermal and without any contamination issues. As a result, the compressor has high efficiency and throughput, and it can compress hydrogen from ambient to 900 bar with a single five-stage compressor. The current model IC90, for example, has a maximum delivery rate of 33.6 kg/hr H₂ and maximum pressure capacity of 1,000 bar. With input H₂ at 5 bar, power consumption is 75 kW, with specific energy consumption of only 2.7 kWh/kg H₂ (a 40 percent energy saving compared to a conventional dry piston compressor). With mass production, the current capital cost is expected to be reduced by almost 50 percent. Initially the technology was used in compressed natural gas fueling stations, and then extended to hydrogen compression starting around 2010. Linde has now commercialized this technology, and it is used in their standard hydrogen fueling station design (Mayer, 2014). Linde also recently started a serial production line for these products. The DOE is well aware of and familiar with this technology and the potential benefits. It would be useful for the Partnership to review and compare this compression technology with the ongoing DOE-funded programs to calibrate future R&D needs and direction.

Current DOE R&D efforts aimed at reducing fueling station costs include development of a linear motor reciprocating compressor to lower maintenance costs by minimizing wear parts and of coatings for compressor seals to improve durability in high-temperature, high-pressure hydrogen. R&D efforts are also focused on improving hydrogen dispensing equipment at the station, such as durability of the dispensing hose at cryogenic temperatures. The projects highlight challenges related to materials compatibility, compression efficiency, durability, maintenance, and associated costs. Some new approaches are emerging, such as underground or canopy storage, and high-pressure delivery and storage to eliminate compression at the fueling station, thereby reducing station footprint and cost. As pointed out before, there is clearly a need to strive for novel approaches to make hydrogen an affordable and practical fuel of choice.

In the DOE program, the long-term delivery cost projections are based on medium-pressure hydrogen gas delivery from a large central plant via pipelines. Various efforts are under way to reduce the cost of pipeline delivery (Gupta and Soto, 2016). These include development of high-volume centrifugal compressors and fiber-reinforced composite pipelines to reduce the distribution cost as well as safety-related studies to understand hydrogen embrittlement of structural steels. Hydrogen embrittlement is well known in the industry and is well studied. It may be worth making sure that there is no duplication of efforts, so that funding can be used for other more pressing matters.

Significant Barriers and Issues That Need to Be Addressed

The key remaining barriers and issues identified for the hydrogen production and delivery area discussed above are summarized here:

• A critical barrier to meeting the hydrogen production cost target with electrolysis is the cost of electricity to run the unit, which cannot be resolved with incremental improvements in capital cost and efficiency targeted with the current R&D.
• In the long term there is need to produce hydrogen with no or very low GHG emissions. This requires use of renewable sources such as biomass or solar energy, directly or indirectly through electrolysis of water. Efficiencies of various renewable pathways are currently low. They need to be significantly higher for cost-competitive hydrogen production.
• While high-pressure (350 and 700 bar) storage is currently the only viable option, it is a high-cost option when combined with the required delivery and dispensing infrastructure from a gaseous or liquid hydrogen source. The high cost of materials such as carbon fibers, required for high-pressure storage, and the high cost of compression continue to be the significant barriers that need to be addressed.
• Pipeline delivery of high-pressure hydrogen is envisioned as the long-term approach. This entails high initial investment costs as a key barrier along with land access and public safety as well material compatibility, durability, and costs.
• One critical issue with hydrogen as a light, colorless, odorless gas is that it can easily leak without easy detection from piping, fittings, valves, equipment, etc. especially under the high-pressure conditions. This can result in loss as well as unsafe conditions due to the flammable and explosive nature of hydrogen. This leads to the use of more expensive equipment.
• An obvious key barrier for the hydrogen production and delivery is the inability of the current technology options to meet the cost targets to be competitive with other transportation options.
• An overarching barrier is inadequate funding for the hydrogen production and delivery program to accomplish the stated goals within the desired time frame to make an impact. As a consequence, there is insufficient focus on novel generation technologies to overcome the technical barriers.

Responses to Recommendations from the NRC Phase 4 Review

Committee Assessment of Response to 4-3. The Partnership states that DOE has funded efforts to develop CCS in conjunction with NG reforming (SMR). The cost estimates for this option show that CCS adds substantial costs, even at a large-scale production plant. Further efforts in CCS development may bring these costs down. However, viability of CCS remains uncertain.

Committee Assessment of Response to 4-4. The Partnership has pointed out relevant projects and significant progress made toward reducing the capital cost of an electrolyzer. However, as discussed in this section, this cost reduction has not resulted in any significant reduction in the overall cost of hydrogen produced by electrolysis because of the fact that the major cost contributor is the cost of electricity.

Committee Assessment of Response to 4-5. In the response, the Partnership pointed out ongoing DOE-funded projects conducted by Giner and Proton Onsite. It also mentions a Fuel Cell Energy project on electrochemical compression. Since the last report it is not clear how much progress has been made in this area, and the techno-economic viability of electrochemical compression seems uncertain. Yet, DOE has recently awarded funding to Giner for a new project using such a technology. The ongoing analysis by Strategic Analysis, Inc., should help make the determination. The committee feels that efforts should focus on exploring other novel approaches to compression. The recently funded project to Greenway combining EHC and MHC may be a step in the right direction.

Committee Assessment of Response to 4-6. The suggested approach, generally known as power-to-gas, is being pursued more aggressively in Europe, especially

in Germany. In the United States, the first such project started in mid-2015, with NREL teaming with Southern California Gas Co. and National Fuel Cell Research Center. ARPA-E recently announced funding of $2 million to Dioxide Materials for a similar project using their proprietary catalyst-based electrolyzer technology. A greater emphasis on this area would help overcome the electricity cost barrier to electrolysis mentioned above.

Findings and Recommendations

Hydrogen production by natural gas reforming is currently a cost-effective option for the near-term hydrogen requirements, and it also provides a pathway to reduced GHG emissions. To further reduce GHG emissions, the use of renewable sources of energy, such as biomass, wind, and solar, is required. Development of such technologies is the focus of the long-term R&D. However, delivery and dispensing of hydrogen is still prohibitively expensive and requires technological advances to meet the overall cost targets for the HFCV option to be viable in the future. Currently the accepted option for onboard storage is 700-bar compressed hydrogen. The delivery and dispensing of hydrogen needs to meet the corresponding requirements—that is, even higher pressure (e.g., 875-900 bar) at the pump. Thus, the R&D focus has been to develop low-cost compression technologies and materials and concepts for high-pressure hydrogen storage and transport. There are several hurdles with this approach, as pointed out, and alternative new concepts need to be continually developed.

Finding 3-22. U.S. DRIVE does not have a cost target for dispensed hydrogen; it is instead considered within the scope of the U.S. DOE R&D program. The DOE cost target for dispensed hydrogen of less than $4/kg H₂ is based on their calculation of threshold cost.¹² Since DOE calculated this cost in 2011, there have been changes in the base case values such as the fuel economy for hybrid electric vehicles.

Recommendation 3-10. The hydrogen threshold cost calculation, published by the Department of Energy in 2011, should be revised by taking into consideration the advances in competing hybrid vehicle technologies as well as any progress made with vehicular hydrogen fuel cells. This should be carefully assessed and addressed by the appropriate U.S. DRIVE teams as well as the Executive Steering Group to incorporate the implications in the Partnership plans.

Finding 3-23. Hydrogen production by natural gas reforming at large central plants can meet the DOE production cost target. However, to meet the greenhouse gas reduction requirement, it needs to be combined with carbon capture

___________________

¹² Threshold cost is calculated so as to be competitive with other transportation options that are expected in 2020 (Ruth and Joseck, 2011).

and sequestration, which adds significant cost and particularly entails uncertainty with regard to the long-term viability of CO₂ sequestration.

Finding 3-24. Even with aggressive assumptions with regard to potential reduction in capital cost and the increase in efficiency of an electrolyzer (including high-pressure and high-temperature versions), hydrogen production cost with large scale (50 ton/day H₂) or small-scale on-site plants is projected to be near $4/kg H₂ at best, and it is shown that the key cost factor is the cost of electricity, making up more than 50 percent of the total cost, just as the situation is with PEM electrolysis.

Recommendation 3-11. U.S. DRIVE in conjunction with the Department of Energy should assess if the focus and prioritization of the electrolyzer research and development is appropriate considering the fact that the main cost factor is the cost of electricity and incremental improvements in capital cost and efficiency may not have as much impact on the cost of hydrogen produced.

Finding 3-25. Liquid hydrocarbon carrier technology for hydrogen delivery has not received sufficient attention within the U.S. DRIVE delivery technical team.

Recommendation 3-12. U.S. DRIVE should conduct a detailed analysis of the liquid hydrocarbon carrier option for hydrogen storage and transport, including potential new chemistries, and compare it with other pathways to determine if it merits development.

Finding 3-26. With the current pathways, hydrogen compression is a major cost item, which is being investigated. In addition to improvements in mechanical compression technologies, DOE has funded electrochemical compression projects. However, it has not been evaluated by the Partnership as suggested in the NRC Phase 4 report. Also, other competing options, such as ionic liquid based compression, have not been critically assessed by the Partnership to identify the scope for improvements and future R&D needs to meet the DOE cost targets.

Recommendation 3-13. As suggested in the National Research Council Phase 4 review, electrochemical compression should be evaluated and compared with other existing and emerging options (e.g., Linde’s ionic compressor) to decide on the appropriate future development direction for compression R&D.

Finding 3-27. Although considerable efforts have already been expended to reduce the costs of hydrogen production, delivery, and dispensing with current pathways, it is clear that there are significant hurdles to meet the current cost target for dispensed hydrogen. Therefore, there is a need to continually seek novel technology approaches and innovative pathways.

Recommendation 3-14. In view of the relative costs of hydrogen production and delivery, research and development (R&D) progress to date and the remaining barriers discussed, in identifying R&D needs, the Partnership should address the need for increased emphasis on novel and innovative approaches to hydrogen delivery and dispensing technologies to reduce the overall cost of hydrogen.

needs, targets, direction, and potential ramp-down for consideration by DOE in the management of the DOE FCTO. The FPITT is aligned with the Partnership Goal 2: “Enable reliable fuel cell electric vehicles with performance, safety, and costs comparable to or better than advanced conventional vehicle technologies, supported by viable hydrogen storage and the widespread availability of hydrogen fuel” (U.S. DRIVE, 2016).

In view of the preceding, the FPITT goals are as follows (U.S. DRIVE, 2013d):

• Assess techno-economic and environmental benefits of integrated hydrogen production, delivery, dispensing, and use pathways and compare them to other fuel pathways;
• Identify and bridge knowledge gaps that limit the ability to evaluate implementation scenarios for fuel cell technology pathways;
• Promote transparency in analysis activities by providing feedback on the communication of analysis parameters and results;
• Identify technical and institutional barriers to implementation of the pathways; and
• Promote consistency between goals, analysis, and modeling efforts of various technical teams of the Partnership to ensure consistent overall targets and effective use of resources.

As noted in the FPITT roadmap (U.S. DRIVE, 2013d), hydrogen transportation fuel pathways will need to compete against an existing petroleum-based fuel infrastructure that is well understood, highly efficient, and has widespread consumer acceptance. With the continual introduction of new vehicle technologies, the market conditions in which hydrogen must compete are continually evolving. Therefore, one of the most important technical challenges in implementing hydrogen as a transportation fuel is estimating the costs, benefits, and risks of potential hydrogen fuel pathways. The FPITT has done an excellent job of identifying gaps and barriers, which are listed as follows:

• Early transition station cost analysis;
• Well to wheels and life-cycle assessment: (1) energy, petroleum use, and greenhouse gas emissions for hydrogen pathways with 10,000 psi (700 bar) onboard storage; and (2) more complete life-cycle assessment of energy and emissions for hydrogen pathways;
• Marginal abatement cost of carbon;
• Hydrogen dispensing pressure optimization analysis; and
• Technology status/TRL and early to mature H₂ markets scenario analysis.

DOE has provided funding to undertake tasks to address some of these gaps and barriers. Progress has been made in many of the areas. A methodology was developed to evaluate complete hydrogen production, delivery, and dispensing pathways and to report on life-cycle hydrogen cost, energy use, and greenhouse gas emissions (Chapman and Randolph, 2016). NREL has completed evaluation of several current and future pathways, and the results have been published. FPITT along with ANL and ORNL evaluated trade-offs between hydrogen dispensing pressure with associated station costs and emissions and customer convenience. The study has not been conclusive and requires further analysis.

The Partnership provided a list of DOE-funded projects as those relevant to U.S. DRIVE. However, many of the projects listed were completed in 2015 with inconclusive results (see Table 3-13). Models and tools are well developed, but analysis requires reliable data, which are lacking.

The program overview states that while the 2020 at-the-pump H₂ cost target is still less than $4/kg, a new early market cost target of $5-10/kg has been introduced (Joseck, 2016). The cost analysis on a life-cycle basis shows that many of the options, except for BEVs, have levelized driving costs in the range $0.35-0.40/mi (BEV: $0.50-0.55/mi). Also, on a $/mi basis, H₂ fueling infrastructure costs are close to but higher than those for BEV charging. It is projected that the HFCVs can be cost competitive with conventional ICE vehicles on $/mi basis by 2025-2030 time frame, if DOE cost targets for hydrogen and fuel cells are met. However, as mentioned in the previous section, the comparison basis used for H₂ production and delivery threshold cost is the HEV. There is a need to have a common basis to compare different results. Nevertheless, the combined envi-

TABLE 3-13 Department of Energy Projects Related to Hydrogen Pathway Issues Relevant to U.S. DRIVE Goals

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a).

ronmental and life-cycle analyses demonstrate the need for a portfolio approach. With respect to hydrogen, fuel availability is a major concern for early adopters. Therefore it is evident that while more efforts are required to bring down hydrogen fueling costs, funding and innovative financing efforts are particularly needed to promote hydrogen infrastructure during the transition phase. Also there is a need to identify and promote specific situations and scenarios with viable business cases to enable gradual growth of the infrastructure.

One example of some success with the use of fuel cells in vehicles is that of forklift trucks, primarily due to productivity gains and battery versus fuel cell life and maintenance costs. The financial benefits (e.g., labor and time savings) of using fuel cells instead of batteries are such that it seems to support the current high cost of dispensed hydrogen, albeit in a limited number of cases. BMW, for example, now operates more than 350 forklifts at their production facility in Spartanburg, South Carolina, using hydrogen fuel cells. Fuel cell ownership cost appears to be still an issue that is prohibiting widespread use of hydrogen and fuel cells in this application.

For passenger vehicles, which are the focus area of the U.S. DRIVE Partnership, commercialization is even more challenging. First, the vehicles are still quite expensive and have only just started to appear in the market with limited availability. Even more important is the lack of availability of hydrogen fuel at affordable cost for public use. Nevertheless, some progress is being made in establishing a hydrogen infrastructure in limited areas. A notable example is that of funding by the state of California for fueling stations. Similar efforts are now under way in the Northeast states, such as Connecticut and New York. But in both cases there is little, if any, involvement of the Partnership members. Toyota and Honda are supporting the infrastructure build-out by providing funding. For example, Toyota ($7.6 million) and Honda ($13.8 million) announced funding to First Element Fuel for installing fueling stations in California. Air Liquide (AL) recently announced development of a network of fueling stations in the Northeast corridor using LH2 produced in their plant in Quebec using mostly hydroelectric (fossil-free) power (Edwards, 2016). Under the special circumstances, including available low-cost LH2 in the region with greater than 100 mile transportation, use of existing facilities, and collaboration with Toyota for certain guaranteed hydrogen demand, AL has been able to have a viable business case. Efforts like these are helpful in transitioning to a growing market for HFCVs and should be sought for and encouraged. Nevertheless, widespread use of HFCVs will require availability of cost-competitive hydrogen anywhere in the country where vehicles are to be sold.

Complete fueling station cost analysis was conducted using factors developed for economies of scale and for the learning curve (Joseck and Verduzco, 2016). Based on the current single 100 kg H₂/day station cost, and using these factors reflecting capital cost and O&M cost savings, station cost for a scenario with 100 stations at 1,000 kg H₂/day each, is estimated to be just under $4/kg, which is still

significantly higher than the target of $0.70/kg. This gap needs to be addressed with R&D which is a major challenge. The FPITT has also initiated an analysis, in conjunction with the C2G working group, to assess the marginal cost of carbon abatement for the different pathways. Similar other activities to address the gaps identified, such as 2 and 5 listed above, are in progress. Results of these analyses are critical to provide feedback to DOE for funding targeted R&D, to understand the viability of hydrogen pathways, and to assess commercialization timeline.

Japanese and Korean OEMs have begun introducing vehicles in the U.S. market, but lack of adequate hydrogen infrastructure represents a major roadblock in the market introduction and proliferation of HFCVs. Goal 2 in the U.S. DRIVE plan (U.S. DRIVE, 2014) quoted above has not received the attention it merits as the core industrial partners (e.g., the energy companies) do not seem to be sufficiently engaged in activities related to hydrogen production and delivery pathways. For this critical requirement, the U.S. DRIVE program is almost entirely dependent on the DOE and national laboratory efforts. However, the hurdles identified above suggest that apart from economies of scale and learning by doing, significant technological advances are necessary to meet the cost target for the dispensed hydrogen. This suggests that the current level of funding for the hydrogen program (production, delivery, station development) may not be adequate to meet the challenge. In particular, with regard to infrastructure development, the DOE budget of $2-3 million per year is focused on analysis. While this is an essential activity, it is not sufficient to promote the necessary infrastructure build-out activities. While some electric utility companies, representing electricity infrastructure, are part of the core U.S. DRIVE team, the same is not true for hydrogen; industrial gas companies, with hydrogen infrastructure expertise, are represented in the U.S. DRIVE Partnership only as one-off associate members of the various technical teams. DOE has initiated another partnership called H2USA, which has a wider membership and is almost entirely focused on hydrogen infrastructure. These efforts are complementary and helpful in promoting the U.S. DRIVE agenda.

Significant Barriers and Issues That Need to Be Addressed

A critical barrier to the widespread deployment of HFCVs is the lack of a hydrogen infrastructure. Although a few select states have been funding and promoting installation of public fueling stations, there are very few stations operating today. While companies like Toyota and Honda are providing additional funding for fueling stations, similar participation of domestic auto manufacturers and energy companies is lacking.

Responses to Recommendations from NRC Phase 4 Review

Committee Assessment of Response to 4-2. The FPITT efforts are well coordinated and the results of the analyses are being used to direct R&D efforts to focus on appropriate critical factors and to prioritize the project areas. However, as suggested, this may be taken a step further to make some bold decisions with respect to the ultimate viability of different approaches to update the roadmap and timeline for future transportation scenarios.

Findings and Recommendations

While electricity infrastructure required for the initial introduction of BEVs, PHEVs, and related vehicle technology options already exists, hydrogen infrastructure is practically nonexistent. Therefore, market introduction of HFCVs faces a daunting challenge. Moreover, with ongoing improvements in engine and battery technologies, competition for HFCVs is getting more intense. Nevertheless, HFCVs are being rolled out and made available, especially by automakers not part of U.S. DRIVE, like Toyota and Hyundai. But there is a lack of hydrogen infrastructure, which can derail deployment. Some states, like California, Connecticut, and New York, along with companies like Toyota and Honda, are promoting infrastructure build-out by providing funding.¹³ But to date there are very few operating fueling stations to support the projected market for HFCVs. High station cost is an obvious barrier.

Finding 3-28. Technology developments for hydrogen fuel cell vehicles are not enough to bring them to market; there is a critical need for hydrogen infrastructure development.

___________________

¹³ See, for example, the California Hydrogen Business Council at https://www.californiahydrogen.org/content/chbc-releases-report-private-financing-hydrogen-fueling-stations. A 2017 report by the California Energy Commission also addresses the time and cost for 100 hydrogen fueling stations and can be found at http://www.energy.ca.gov/serp.html?q=CEC-600-2017-002&cx=001779225245372747843%3Actr4z8fr3aa&cof=FORID%3A10&ie=UTF-8&submit.x=13&submit.y=9.

Recommendation 3-15. The Executive Steering Group should address issues (e.g., how will fueling stations be installed and by whom, who operates them, who will produce hydrogen, how will investments occur in fueling infrastructure without sufficient fuel cell vehicles on the road and vice versa, etc.) related to hydrogen infrastructure and assess U.S. DRIVE’s role to formulate an action plan to address the issues and barriers.

Finding 3-29. Although industrial gas companies currently have the most experience with hydrogen production, delivery, and infrastructure, they are not core members of the Partnership; some of them serve as associate members of technical teams, but not on a consistent basis.

Recommendation 3-16. U.S. DRIVE should consider having industrial gas companies involved in hydrogen infrastructure activities as permanent members rather than as temporary associate members.

SAFETY, CODES, AND STANDARDS

In addition to the importance of developing a hydrogen refueling infrastructure, global codes and standards play a critical role in laying the groundwork for commercialization of hydrogen and fuel cell technologies. It is important to coordinate the SCS efforts with international bodies to harmonize the regulations, codes, and standards. The hydrogen codes and standards technical team (CSTT) of the U.S. DRIVE is focused on “efforts to enable and facilitate appropriate R&D to support technology readiness for the development of safe, performance-based technical codes and standards that support the 2015 commercialization decision for widespread consumer use of fuel cells and hydrogen-based technologies with commercialization by 2020” (U.S. DRIVE, 2013a). Although CSTT efforts are focused on hydrogen, the current membership does not include any industrial gas company; efforts are under way to add one as an associate member. The CSTT coordinates its efforts with those of the other technical teams for consistency and synergy. The CSTT roadmap, updated in June 2013, provides information and detailed discussion of specific R&D, testing, and analysis in the areas of materials compatibility, risk assessment and behavior, accelerated testing, component performance, and fuel quality (U.S. DRIVE, 2013a). CSTT provides support for the development of HFCV standards as well as hydrogen fueling station standards.

The CSTT Roadmap and the R&D priorities identified in it are incorporated in the DOE multiyear research, development, and demonstration plan. The DOE-funded projects to support SCS development, which are considered relevant to the Partnership, are listed in Table 3-14.

The SCS programs at DOE support and facilitate the development and publication of essential codes and standards and of domestic and international regulations. The DOE conducts R&D to provide critical data and information

TABLE 3-14 Department of Energy Projects Related to Hydrogen Codes and Standards

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a).

needed to define requirements in developing codes and standards. For example, a data-driven science-based approach enabled an update to the hydrogen bulk storage separation distances used in key codes (e.g., National Fire Protection Association [NFPA] codes 2 and 55). As a result, required separation distances were reduced by as much as 50 percent in some instances. Direct DOE support has saved 3 to 5 years in the development of relevant codes and standards, according to Dr. Will James, who heads the efforts at DOE (Green Car Congress, 2016). CSTT achievements during 2012-2016 include support for the development of Global Technical Regulation (GTR), development of critical capability at NREL

for component safety and reliability testing, and hydrogen sensor field trials at California fueling stations. In 2015, PNNL developed the Hydrogen Tools Portal¹⁴ for disseminating hydrogen safety knowledge to critical user groups (U.S. DRIVE, 2015). This should be very useful in hydrogen fueling commercialization efforts.

Jesse Schneider (BMW) mentioned at the SAE World Congress (Green Car Congress, 2016) that in addition to automakers teaming up for fuel cell vehicle commercialization, work on infrastructures is under way in three continents: Asia (Japan), North America (United States), and Europe (Green Car Congress, 2016). In 2016, each of these areas plans to have more than 50 stations and hundreds more within the next 5 years. These stations are being standardized to NFPA/International Code Council code and are helped by the ISO Technical Specification 198801, which was published in 2016, giving a baseline of safety and performance.

Significant Barriers and Issues That Need to Be Addressed

While significant progress is being made in the area of SCS for safe use of hydrogen as a fuel, a critical technical barrier to the widespread use of hydrogen is usage and access restrictions, such as in parking structures, tunnels, and other areas. Other remaining barriers include high cost and limited availability of hydrogen components and equipment, and corresponding reliability data needed to develop codes and standards. Other nontechnical barriers related to SCS include limited availability of insurance, lack of knowledge and training for officials, lack of consistency in regulations across states, and limited participation of businesses.

Responses to Recommendations from the NRC Phase 4 Review

___________________

¹⁴ The website for the Hydrogen Tools Portal is http://h2tools.org, accessed September 7, 2016.

Committee Assessment of Response to 2-4. The Partnership response is appropriate and satisfactory. One agency that controls some of the relevant regulations, and is a member of the team, is DOT. Through DOT, the Partnership should be able to influence acceleration of regulation finalization.

Committee Assessment of Response to 2-5. While the response is adequate, and necessary work is being conducted as per the update provided in June 2016, it does not mention follow-up implementation actions that utilize the results of these tests. Also, acceleration of these programs would be helpful.

Committee Assessment of Response to 2-6. Considering the Partnership scope, the response seems appropriate. While the Partnership has worked on the development of fuel quality standards, the committee feels that it would be complementary if the fuel cell team initiates a program to improve tolerance of fuel cells to impurities.

Committee Assessment of Response to 2-7. As per the update provided in June 2016, DOE has ended the funding on the referenced safety sensor. The LANL/LLNL solid-state electrochemical sensor is now ready for commercialization, and DOE continues to fund NREL’s sensor validation laboratory to support industry.

Committee Assessment of Response to 2-8. The Partnership’s response is satisfactory. The issue of depressurizing a damaged tank identified in the NRC recommendation is important, and CSTT should continue to work with appropriate organizations to support relevant activities within their scope.

Findings and Recommendations

Finding 3-30. While testing and analysis of high-pressure tanks is progressing well within the DOE-funded program, there is a need to accelerate these activities. Also there is no mention in the codes and standards technical team’s plan for follow-up implementation actions that utilize the results of these tests.

Recommendation 3-17. The codes and standards technical team should assess the results of high-pressure tank testing and prepare a plan for implementation actions that utilize the results of testing.

Finding 3-31. While the Partnership has worked on the development of a fuel quality standard, it is also apparent that high-quality hydrogen leads to higher cost of hydrogen. One potential approach to avoid this cost escalation is to improve the tolerance of fuel cells to impurities.

Recommendation 3-18. U.S. DRIVE should explore the possibility of improving fuel cell tolerance to impurities in hydrogen so that a lower-quality, and hence lower-cost, hydrogen may be used.

ELECTRIC DRIVE SYSTEMS AND POWER ELECTRONICS

Background

Under global environmental pressures to reduce tailpipe emissions and dependency on petroleum as a source of energy for ground transportation, most automotive OEMs have been working for decades on the development of electric and hybrid power train systems with zero or ultra-low tailpipe emissions, respectively. Recent advances in battery and electric drive technologies made it possible for vehicle manufacturers to commercially deploy electrified hybrid and electric vehicles (cars, trucks, and buses). Despite continued environmental pressures and evidence of global warming, market penetration of electrified vehicles is growing at a slow rate for a variety of reasons, including cost, electric-only range, and battery charging time, in addition to a noticeably weak consumer acceptance of the technology. The electric drive system (consisting of an electric motor and an electronic controller) is a critical part of electrified power trains for light-duty vehicles. Therefore, a key objective of the U.S. DRIVE Partnership is the development of technologies addressing the electric drive components’ cost, weight, and size to help expedite electrified power train market penetration. The Partnership has established cost and technical targets for the electric traction system for 2010, 2015, and 2020 as listed in Table 3-15.

The DOE budgets for this electric drive technology R&D were approximately $24 million for FY 2014, $21 million for FY 2015, and $38.1 million in FY 2016.

A variety of electric drive technologies have been investigated over the years by the industry, academia, and government laboratories, including brush and brushless direct current (DC) and alternating current (AC) motors as they vary in their performance, reliability, and cost. Because of the critical impact of the drive efficiency and power density (weight) on the electric range of electrified vehicles, the leading contenders have been the following: (1) brushless permanent magnet DC motor (BLDC); (2) induction motor (IM); (3) switched reluctance motor (SRM); and (4) synchronous reluctance motor (SyRM). The brushless permanent magnet motor using rare earth (RE; NdFeB) magnets has been the choice for many OEMs (GM, Ford, Toyota, Nissan, Honda) because of

TABLE 3-15 Electric Traction System Technical Targets

SOURCE: Data from El-Refaie (2016).

its superior performance (efficiency, power density, and torque density), which enabled them to reach a reasonable commercial range despite the limited energy density offered by available lithium-ion batteries. Only a few, including Tesla Motors and some conversion companies, opted for IM technology, as a low-risk choice, since induction motors have served as the workhorse of the industry for decades. Furthermore, methods for induction motor control for operation from a battery source (DC) are known and have also been established for a long time. The switched reluctance motor offers great advantages in reliability and cost but suffers from an inherent audible noise problem that researchers have not been able to resolve to the level required for commercial implementation (Omekanda, 2013).

However, the preceding analysis and situation have been disrupted by two factors: (1) The cost of rare earth magnets skyrocketed a few years ago, sending shock waves across the globe to users and causing deep concerns about long-term availability. This has prompted most users to look for alternatives. (2) Battery technologies have made significant progress and battery energy densities showed considerable improvements, so a slightly lower motor efficiency or power density is less critical. Therefore, the focus of the U.S. DRIVE electric drive R&D activity is to develop technologies aimed at the reduction or elimination of rare earth magnets in motors, and simultaneously, the adoption of low-loss wide bandgap (WBG) semiconductors to further boost electric drive system efficiency (Dawsey and Rogers, 2016). The effort to eliminate RE magnets is exploring two possibilities: (1) maintaining the Brushless Permanent Magnet (PM) motor type, but developing advanced high-energy non-rare earth magnets (AlNiCo, Ferrite, or dysprosium-free RE magnets) to replace the NdFeB magnets having the super-expensive Dysprosium content currently used; and (2) reconsidering other non-PM motor types (IM, SRM, SyRM, wound field excited motors) but incorporating innovative structures/assemblies, and effective thermal and noise management techniques as lower-cost alternatives.

Current Status

Progress was made by the Partnership both on the research and the development fronts pushing the system cost and performance closer to the set targets listed in Table 3-16 (U.S. DRIVE, 2013b).

Estimating drive system performance and cost based on promising motor concepts (projects edt044 and edt045 listed in Table 3-17) and advanced inverters (project edt040 in Table 3-17) that have been evaluated point to a drive system that can meet the 2015 set targets, with potential for further improvements toward the 2020 targets.

TABLE 3-16 U.S. DRIVE Electric Drive Targets

SOURCE: U.S. DRIVE (2013b).

TABLE 3-17 Department of Energy Electric Drive Projects Related to U.S. DRIVE Goals and Targets

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a).

Assessment of Progress and Key Achievements

The U.S. DRIVE focus on electric drive technologies is supported through projects funded by the DOE VTO, which, as noted previously has a budget of $38.1 million in FY 2016. Table 3-17 lists DOE projects related to meeting U.S. DRIVE’s goals and targets in the electric drive area.

Electric Motors

Excellent progress has been made in assessing various motor topologies and advanced materials:

• In project number edt044, UQM has developed, built, and tested a new PM motor with an innovative magnetic design and incorporating new non-RE magnets (AlNiCo) developed by Ames Lab. The proof-of-concept (POC) motor demonstrated performance very close to requirements with off-the-shelf magnet material. The work on increasing magnet properties looks promising and would reduce magnet content and cost. Motor build will demonstrate the feasibility of the approach to meet or exceed the 2020 DOE requirements with Ames-designed magnets and optimized cooling methods from NREL (Ley, 2016).
• In project number edt045, GE has completed the experimental evaluation of several proof-of-concept motors: (1) with reduced RE (Dy-free flux switching type), (2) non-RE (Spoke-Ferrite), (3) no-magnet (DC-biased switched reluctance-SRM), and (4) SyRM using a special magnetic laminate patterned with nonmagnetic regions and a new high-temperature slot liner material. Results of the work of item 4 on the SyRM look most promising and seem to approach the 2020 DOE set targets (El-Refaie, 2016).
• In project numbers edt062 and edt065, the ORNL and its partners have also studied motors without rare earth permanent magnets—namely, Concentrated and distributive wound ferrite machines, brushless field excited (BFE), and synchronous reluctance (SyRM)—while using advanced soft magnetic materials with reduced magnetic loss to enhance performance (Burress, 2016; Ludois, 2015). The simulation results were encouraging. The surprising result is that two distributive wound ferrite machine designs are expected to reach DOE 2020 motor targets and have similar performance to rare earth PM motors despite the large difference in energy product, which is a measure of magnet strength, between ferrite and RE magnets (RE magnets are 5-10 times better than ferrite). The details of the designs were not revealed at the June 22-23, 2016, committee meeting due to pending patents.
• Most exciting results came from the ORNL’s work on a multi speed motor concept (Dawsey and Rogers, 2016), which uses electronic switching to reconfigure the stator windings at three different motor speeds potentially resulting in a 24 percent reduction in drive cycle loss (Dawsey and Rogers, 2016). While the concept is not new, it is the ability to use only a small number of power switches to accomplish the switching that is innovative. If successful this approach will help boost motor efficiency and reduce cooling needs, thus reducing motor size and weight.

Power Electronics

Significant progress was reported by the Partnership in power electronics. This was achieved by using innovative packaging and integration of classic inverters and converter configurations and also by exploring the use of WBG devices for automotive power electronic systems. Key achievements are summarized below:

• In project number edt040, which was completed in December 2015, General Motors in collaboration with ORNL, NREL, and suppliers has developed and tested its next-generation inverter, demonstrating a capability for scalability, high efficiency, and high power density, which is projected to meet all U.S. DRIVE 2020 targets for performance and cost. This was achieved by an innovative design that integrates active components and reduces/eliminates supporting components (Zhao, 2016).
• Many projects aimed at increasing the level of understanding of WBG devices, while others focused on developing high-frequency circuitry and high-temperature components to sustain the high-temperature and high-switching frequency environment enabled by the WBG devices. These include WBG device characterization, as in project number edt53, as well as evaluating converters and inverters using WBG devices (Chinthavali, 2016). In project number edt53, a 55-kW traction drive WBG silicon carbide (SiC)-based inverter was designed and will be built for testing in 2017. Another WBG traction inverter investigation was conducted by APEI and Cree in project numbers edt58 and edt73, leading to necessary fundamental learning (Olejniczak, 2016; Casady, 2016). Also, in project number edt67, Delta Products is evaluating a gallium nitride (GaN)-based onboard charger (Zhu, 2016). An integrated SiC-based onboard charger and DC-DC converter demonstrates double the power density with a 40 percent reduction in cost compared to the state-of-the-art non-WBG silicon-devices (Dawsey and Rogers, 2016).
• NREL has demonstrated ribbon electrical interconnects to replace the current round wire interconnects, which enables higher power densities and lower parasitic inductances that are particularly critical for WBG devices (Dawsey and Rogers, 2016). Ribbon interconnects of various geometries and materials were tested under elevated temperature, temperature cycling (thermal shock), corrosion and highly accelerated life test (combined vibration and thermal cycles) and exhibited similar reliability to wire interconnects under these conditions.
• Thermal stack-up enables full potential of WBG Devices. New SinteredSilver (Ag) Interconnects technology, project number edt063, developed by ORNL, is showing promise and is under intense testing to understand and demonstrate the ability to consistently and effectively create reliable sintered joints (DeVoto, 2016).

Barriers and Issues That Need to Be Addressed

There are two key barriers to closing the gaps to Partnership targets: (1) component cost, particularly that of rare earth permanent magnets, which, in some cases, can be more than half the motor cost; and (2) the high power loss associated with electronics, which leads to heavy and expensive cooling systems (U.S. DRIVE, 2013b). By improving the total drive system efficiency, the thermal management system can be reduced leading to overall reductions in volume, weight and cost.

Response to Phase 4 Recommendations

Committee Assessment of Response to 3-14. The committee appreciates the intense activities and focus of the DOE-APEEM program on WBG devices aiming at improving the efficiency and reducing the size and weight of the power electronic system. It is expected, however, that through these investigations one could determine which of the WBG devices (GaN versus SiC) is the more appropriate and cost effective for automotive applications and possibly redirect efforts to accelerate its readiness for implementation.

Committee Assessment of Response to 3-15. The committee agrees that the APEEM programs are designed to address the issues associated with using magnets with less or no RE material. The concern is that much of the focus was directed toward developing new AlNiCo or ferrite material with improved performance but still far below that of NdFeB magnets, which as the preceding response indicates would make it difficult to meet all the DOE targets. The approach followed by Honda (Williams, 2016), which appears to be successful, is to eliminate the need for only the super-expensive RE additives (dysprosium) used for protecting the magnet against demagnetization at high temperatures, by developing a new manufacturing process that gives the high-energy NdFeB magnet a natural protection at high temperature.

Committee Assessment of Response to 3-16. The committee agrees that the current DOE programs follow the above recommendation with a potential for a positive impact.

Appropriate Federal Role

Despite continued environmental pressures and evidence of global warming, market penetration of electrified vehicles is growing at a slow rate partially due to cost, size, and weight (impacting range) of the electric drive, which is the core of all electrified power trains. There are, however, fundamental challenges in dealing with issues of size, weight, and cost of the electric drive, discussed in this report, which require a concerted national effort to address. Addressing issues such as finding a replacement for the costly rare earth magnets used in the motor or finding means to produce WBG devices at affordable cost would require resources beyond the capabilities of individual OEMs and are likely to exist at the U.S. national laboratories. Therefore, in the committee’s view, the support of the federal government in forming this Partnership and involving the national laboratories is essential for a successful outcome, which would not only expedite electrified power train market penetration and help with national competitiveness but would also benefit other industries of national interest (e.g., defense).

Findings and Recommendations

It is the committee’s opinion that the APEEM project portfolio pursued by the U.S. DRIVE Partnership, and the supporting DOE-led projects in the motors and power electronics areas, are appropriate to address cost, size, and weight challenges. Several motor configurations and design variations are under investigation to address the high cost of RE magnets. Significant progress was made toward meeting the Partnership targets in the power electronics area, achieved through

innovative packaging and integration of inverter and converter components with the use of WBG devices (GaN and SiC). Given the inherent cost advantage of GaN devices grown on Si-substrate compared with SiC on SiC substrate (due to the much higher cost of SiC compared to Si), it is expected that GaN will ultimately be the winner among these two competing technologies for automotive applications. Historically, SiC devices have been the focus of research for many years, as they possess higher voltage and temperature capabilities than GaN devices. Operating at these high levels of temperatures requires other circuit components to be also capable of these temperatures, which is cost prohibitive for automotive applications but not so for other cost-tolerant applications, such as for defense.

In addition to meeting the DOE targets for cost and power density, propulsion drives must meet other criteria of automotive applications, such as electromagnetic interference, audible noise level, and instant torque availability. For example, switched reluctance and PM flux-switching types are known to exhibit a level of audible noise beyond what would be acceptable for automotive use. Also, wound-field excited machines exhibit a time delay in torque generation as it takes time for the excitation current and the magnetic field it produces to build up to the level required for full torque generation. Automotive propulsion requires instant torque availability upon demand.

An important element of the DOE VTO project management process is the go/no-go decision used to downselect from many considered options to the few that hold the most promise for meeting program objectives. For example, both ORNL and General Electric have programs that consider a wide range of motor technologies. The ORNL go/no-go decision point follows a thorough and extensive analysis phase, while General Electric carries many technologies to the build and test phase before a go/no-go decision. Considering the availability of accurate analytical tools for almost all motor technologies, it is recommended that DOE follow the excellent and cost-effective ORNL approach, mentioned above as an example, to eliminate options that could not meet fundamental automotive requirements stated earlier.

Finding 3-32. Some of the considered motor technology options, mentioned earlier, have been pursued for an extended length of time despite known inherent limitations. They could have been eliminated earlier on the basis of analysis.

Recommendation 3-19. The U.S. DRIVE Partnership should take into account those specific criteria of the automotive propulsion applications such as electromagnetic interference, audible noise level, and instant torque availability, while searching for the best solution to meet the Department of Energy-set targets for electric drive.

• The U.S. DRIVE Partnership should assess the audible noise produced by some of the motor configurations under consideration (e.g., switched

• reluctance, PM flux switching types) and determine if it would be commercially acceptable. If noise is found to be truly unacceptable, as anticipated, work on these machine types should be terminated and resources devoted to more promising configurations.

• If wound-field excited machine technology is still under consideration (El-Refaie, 2016), the U.S. DRIVE Partnership should address the field build-up delay in this type of motor, as torque is expected to be available instantly in automotive propulsion applications.

Finding 3-33. Only a few projects are exploring GaN, with the majority focusing on SiC. Given GaN’s potential cost advantage, it is expected to ultimately be the preferred choice for automotive applications.

Recommendation 3-20. The U.S DRIVE Partnership should increase the focus on the advancement of gallium nitride technology in order to accelerate its readiness for commercial implementation.

Finding 3-34. There is a tendency within the automotive industry to maintain similar gear-up ratios in order to avoid the additional cost associated with redesigning the transmission for higher motor speeds. Operating the motor at a higher speed will reduce the size, weight, and cost of all its active components including the magnets. However, there are major cost, noise, and reliability challenges to increasing the motor speed, which could be difficult to resolve.

Recommendation 3-21. The U.S. DRIVE Partnership should explore new and innovative alternative gearing concepts, as operating at higher motor speeds translates directly to a smaller and lighter motor, with reduced material cost including magnet cost, which might offset the increased cost of gearing. A trade-off study to assess the total impact on the cost and weight of the total propulsion drive including the gearing system should be undertaken.

ELECTROCHEMICAL ENERGY STORAGE

Background

As noted in Chapter 1 and in the Partnership Plan (U.S. DRIVE, 2016), the overall mission of the U.S. DRIVE Partnership is to “accelerate the development of precompetitive and innovative technologies to enable a full range of efficient and advanced light-duty vehicles, as well as related energy infrastructure.” Electrochemical energy storage is a critical component for these advanced vehicles. Electrochemical energy storage (including batteries and supercapacitors) is used in all electric drive vehicles including HEVs, PHEVs and BEVs, and HFCVs. (Note that PHEVs and BEVs are both plug-in electric vehicles that can

recharge their batteries from the electric grid.) The electric system for plug-in electric vehicles and HEVs operates at a high voltage, ranging from 200 to 400 V. Recently there is renewed interest in lower voltage systems of 12 and 48 V for start-stop (automatically shuts the internal combustion engine off when the car is idling and restarts when the foot is off the brake) and mild hybrid vehicles (do not have all the functionality of a full HEV).

The Toyota Prius, the first commercial vehicle to employ significant energy storage in an HEV, was introduced in 1999 in the United States. In 2015 there were over 50 models of HEVs available yet they represented only about 2 percent of new light-duty vehicle sales of the automotive market (DOE, 2016a). Initially HEVs used nickel metal-hydride batteries; however, the newer models are incorporating lithium-ion battery technology. Substantial performance improvement and the lower cost of lithium-ion battery technology have also enabled the successful introduction of BEVs and PHEVs in addition to higher-performing HEVs. In 2015 there were 1.26 million plug-in electric vehicles on the road worldwide, over 400,000 of which are in the United States, yet they represented only about 0.7 percent of new light-duty vehicle sales of the automotive market in the United States in 2015 (IEA, 2016; DOE, 2016b). Thus, there is much room for improvement in battery performance and cost reduction to increase market penetration of electric drive vehicles.

High cost remains the main impediment to significant market penetration of plug-in electric vehicles that utilize large batteries. There is also a need to improve battery performance characteristics—that is, energy density, specific energy, operation at extreme temperatures, charging and discharging rates, cycle, and calendar life. These improvements in performance and cost reduction need to be realized while addressing the inherent safety issues associated with lithium batteries, particularly as battery size increases. Lithium-ion battery performance, cost, and safety are being addressed by the DOE and other government entities, automotive OEMs, and battery manufacturers.

The DOE VTO has organized the electrochemical energy storage technology program into five subactivities:

1. Battery technology development;
2. Applied battery research;
3. Battery materials research;
4. Battery testing, analysis, and design; and
5. Manufacturing and process development.

Subactivity (1), battery technology development, efforts are directed toward support of the domestic battery industry to develop battery module and system hardware and related activities. The VTO works closely with the U.S. Advanced Battery Consortium (USABC) in support of cost-shared contracts with developers. In addition, it also directly supports battery and materials suppliers via

contracts administered by the National Energy Technology Laboratory aimed at BEV, PHEV, and HEV applications. Subactivity (2), applied battery research, is directed by the national laboratories with ANL in the lead role. It focuses on the next generation of high-energy batteries, optimizing systems incorporating new battery materials and diagnosing and mitigating issues that impact battery performance and life. Subactivity (3), battery materials research, is involved in the development of newer materials and electrochemical couples and the fundamental understanding of specific electrochemical systems for lithium batteries. It also studies electrochemical systems beyond lithium ion with a potential for higher energy and power. This exploratory work is directed by LBNL and conducted by many academic groups with participation of national laboratories and industrial research groups. Subactivity (4), battery testing, analysis, and design, is complementary to battery technology development (1) and is involved in life and abuse testing of deliverable cells and batteries, benchmarking systems from industry, test procedure development, modeling, and battery materials recycling. Subactivity (5), manufacturing and process development, is complementary to applied battery research (2) and is involved in development of kilogram-scale advanced materials and customized scaled processes. It is run by the national laboratories with industry partnerships to develop novel electrode/cell manufacturing technologies.

Individual DOE projects in support of the various subactivities (1)-(5) are listed in Table 3-18. The projects are sorted by the energy storage subactivity and were collated from the 2016 VTO AMR presentations.

In addition, the VTO supports several SBIR contracts to develop new ideas and concepts. As discussed in NRC (2013), longer-term energy storage objectives are also being pursued by DOE-funded R&D in the ARPA-E organization that is aimed at high-risk, game-changing technologies beyond the projected capabilities of lithium-ion batteries. Fundamental R&D activities on materials are also pursued through the DOE BES. The VTO coordinates efforts with other government agencies: the DOE Office of Electricity Delivery and Energy Reliability; the Chemical Working Group of the Interagency Advanced Power Group; the Department of Transportation/National Highway Traffic Safety Administration (DOT/NHTSA); the EPA; and the United Nations Working Group on Battery Shipment Requirements. In addition, it includes interactions with the International Energy Agency’s (IEA’s) Implementing Agreement on Hybrid Electric Vehicles (IA-HEV), the eight-nation Electric Vehicle Initiative (EVI), and the Clean Energy Research Center (CERC).

While the U.S. DRIVE Partnership does not have a budget or program per se, it provides goals, targets, and roadmaps via the electrochemical energy storage technical team (EESTT) to those entities that do have budgets—for example, the DOE through the VTO, the ARPA-E, SBIR program, and BES. The DOE VTO largely provides funding that supports U.S. DRIVE goals. The USABC establishes its own goals and targets, which are closely coordinated and aligned with U.S. DRIVE, and funds projects that address these. The DOE VTO FY 2016 budget

TABLE 3-18 Department of Energy’s Energy Storage Projects Supporting the U.S. DRIVE Mission

NOTE: Acronyms defined in Appendix D.

SOURCE: Project numbers are from the Department of Energy’s 2016 Annual Merit Review, see Cooper (2016a) and DOE (2016f).

for energy storage technologies and related activities is $103 million, up from $83 million in FY 2015 and $76 million in FY 2014. The 2016 budget is allocated as follows: battery technology development (32 percent), applied battery research (21 percent), battery materials research (17 percent), and solicitations (30 percent) (Howell and Snyder, 2016).

Current Status Versus Goals

In 2012 the EESTT, which consists primarily of members from DOE and USCAR OEMs, and includes investigators from the national laboratories, the battery industry, and universities, established a new set of goals and targets for electrochemical energy storage technologies. The Electrochemical Energy Storage Technical Team Roadmap was published in June 2013 and provides a summary of the targets for various plug-in electric vehicle applications and their approach of attaining those goals (U.S. DRIVE, 2013c). A detailed list of the targets can be found on the United States Consortium for Automotive Research (USCAR) site, including goals for PHEVs, BEVs, 12V start-stop battery systems, high-power low-energy storage systems, ultracapacitors, battery separators, and electrolytes (USCAR, 2014). These goals, targets, and guidelines are not only used for proposal solicitations but can also play an important role in providing directions to investigators for future research (see Recommendation 3-23).

Since the NRC’s fourth review of the U.S. DRIVE Partnership, the focus of electrochemical energy storage R&D has shifted from high-power batteries for HEVs to higher energy systems for PHEVs and BEVs. This is an appropriate shift, since there are almost 4 million HEVs of over 50 models in the United States. Also, many of the technical advancements realized for high-energy systems can be optimized for high-power systems required for HEV applications.

PHEV40 Battery Meets Cost Target but Falls Short of Energy Density Goals

Significant progress has been made in the cost reduction of the PHEV40 battery and the 2020 price target of $3,400 was achieved (see Table 3-19).¹⁵ This cost projection is derived from ANL’s Battery Production and Cost model (BatPaC) using material costs and cell and pack designs input. The cost projection is based on a production volume of at least 100,000 batteries per year. This cost represents more than a 50 percent reduction in system production price of $6,850 in 2012 (Howell and Elder, 2012). The committee has determined that these costs are not totally out of line with values available in the open literature.

The batteries also meet most of the performance targets; they far exceed the power requirements, meet the cycle life, and are projected to meet the calendar life, but are deficient in the 2020 energy density targets by about 40 percent.

BEV Battery Falls Short of Cost and Most Performance Targets

A major U.S. DRIVE research target has been to reduce the production cost of a BEV battery to $125/kWh by 2020. The current cost projection based on the BatPaC model is estimated to be less than $268/kWh. This represents a significant

___________________

¹⁵ Note that PHEV40 designates a vehicle that can drive in all-electric mode for 40 miles before the battery is depleted and the vehicle has to be powered by its internal combustion engine.