Eric Parker, Hydrogen and Fuel Cell Technologies Office:
Hello, everyone, and welcome to September’s H2IQ hour, part of our monthly educational webinar series that seeks to highlight research and development activities funded by the U.S. Department of Energy’s hydrogen and fuel cell technologies office, or HFTO, within the Office of Energy Efficiency and Renewable Energy. My name’s Eric Parker and I’m the HFTO webinar lead. As always, we’ll be announcing more topics like this soon, as well. This Webex call is being recorded and will be posted on the DOE’s website and used internally. All attendees will be on mute throughout the webinar, so please submit your questions via the Q&A box you should see in the bottom right there on your Webex panel. Be sure to use that and not the chat box. We will cover those questions during the Q&A portion at the end of today’s presentation. And, with that, I’d like to introduce today’s DOE host, Neha Rustagi, to tee up the topic and today’s speaker. Thanks, Neha.
Neha Rustagi, Hydrogen and Fuel Cell Technologies Office:
Thank-you, Eric. I’m really excited for this webinar. We’re presenting a cross-office analysis of market segmentation of medium- and heavy-duty transportation across fuel cell, battery, and combustion engine power trains. So, this effort was led by Chad Hunter. Chad was the team lead and assistant analyst at NREL’s Center for Innovative Mobility Sciences, and he has nearly 10 years of experience across the transportation and energy sectors on economic modeling and also emissions analysis on road transportation technologies, all the way from electric scooters to heavy-duty trucks with advanced power trains. And then, it’ll be co-led by Michael Pena. Michael comes to NREL with eight years of industry experience at Platt Power as a fuel cell systems engineer, and then for the past 14 years, he’s been supporting our office on analysis of hydrogen production, supply chain, energy storage, and transportation techno-economic analysis. So, with that, I will turn it over to Chad.
Chad Hunter, National Renewable Energy Laboratory:
Awesome, thank-you, Neha. Good morning and good afternoon, everyone. Thanks for joining. So, my name is Chad Hunter, and I will be the main presenter today, but, obviously, this is definitely not just my work but rather a collaboration with Mike Penev, Evan Reznicek, Jason Lustbader, and Chen Zhang here at the National Renewable Energy Lab. And, today, we’ll be talking about the spatial and temporal analysis of the total cost of ownership of Class 8 tractors and Class 4 parcel delivery trucks. And this report was recently published. A link is right here, and also provided at the end of the deck. If you have any questions, or can’t get to it for some reason, feel free to reach out. We’d be happy to provide it. We’ll jump into it.
So, first, just to level set a little bit, and provide some context, the National Renewable Energy Lab is one of the Department of Energy’s national labs, and it’s an applied national lab that focused on energy efficiency and renewable energy research, with around 3,000 employees, world-class facilities, and we really work with and across industry and academia and government to partner with the leading institutions in the U.S. and abroad to understand and research these energy efficiency and renewable energy challenges, and help provide guidance to the U.S. Department of Energy on the ways to accelerate the renewable energy transition.
And then, jumping into this specific analysis, quick thanks to all the reviewers and collaborators of this report and project. Obviously, this report and analysis would not be able to be completed without the funding from the U.S. Department of Energy, EERE’s Hydrogen and Fuel Cell Technologies Office. So, obviously a major thank-you there, as well as a major thank-you to all of the collaborators that have helped pull this cross-coordination and collaborative effort off; in particular, the Vehicle Technologies Office and Jake Moore, the analysis lead there, as well as the Deputy Assistant Secretary for Transportation’s office, Michael Berube, Rachel Nealer and Matteo Muratori have been really, really helpful to help coordinate and align the assumptions and analysis framework within this work, as well as there is a variety of national laboratory reviewers across Argonne, Oak Ridge, and NREL that have really been instrumental in helping us with this project. We’ve also brought in a variety of external industry and organizational reviewers and stakeholders, including OEMs and other organizations, such as the 21st Century Truck partnership, which have been, again, really helpful in making sure our analysis and assumptions and modeling efforts do reflect the real world so we can provide the most up-to-date and useful analysis as possible.
So, jumping in, we’ll go over a quick kind of executive summary of the analysis. So, that’s a full 120-page report that’s available online. We don’t have time to go through all of it today, so we’ll just hit some of the highlights, which’ll primarily focus on high-level discussion of the approach, some of the results and key results there, and then some of the key conclusions to wrap it up. So, first, jumping into a high-level overview of the approach to do the advanced powertrain total cost of ownership modeling, we basically coupled together two of the National Renewable Energy Lab’s models together into a new modeling framework, which we’ve titled T3 Scale. The two existing models that we’ve kind of integrated together are FASTsim and SERA. The FASTsim model stands for Future Automotive Systems Technology Simulator, which is our vehicle power train cost modeling tool. And that allows us to assess both the current and future vehicle technologies and understand in a dynamic simulation how they will perform and how to size those powertrains, as well as what the cost of that powertrain will be as a function of time and across the different powertrain archetypes.
We then feed a lot of that information into the total cost of ownership model, our SERA model, Scenario Evaluation Regionalization and Analysis model. So, that allows us to do spatially and temporally total cost of ownership, and they include a variety of exogenous different costs and different datasets into that model. It’s a very flexible model. And then the output of kind of both of those is a final total cost of ownership, and there are a variety of different scenarios that we’re defining in this work to help us understand the different market segments and the relative attributes of advanced powertrains and how they fit into those market segments. And we’ll kind of go through those different scenario definitions a little bit later. So, jumping a little bit more deeply into the powertrain optimization tool, FASTsim, that we have.
So, FASTsim is an automotive powertrain simulator, which includes technology cost data and technology performance data. So, that allows us to evaluate both current technology status as well as future technology trends and expectations. We can combine that with real-world drive cycle and duty cycle data from NREL’s fleet DNA database, and that allows us to not only size the powertrain components to match the acceleration and observe the acceleration needed for the different vehicle Classes and the different vehicle locations, and also allows us to estimate, from a component level, what the total cost of that vehicle will be, and then we can simulate that vehicle on that duty cycle to understand what is the real world fuel economy given the different components of those vehicles, different drive coefficients, different technologies statuses, such as, you know, energy density of the battery or things like that.
And so, that really allows us to estimate the fuel economy, the overall weight, and the vehicle cost of these different – and advanced future vehicles. And, in particular, for this analysis, we focused on kind of three different temporal years, so the 2018, what we were referring to as current status, which is a little – couple years ago now. Then we have a 2025, we have near-term status, and then we have an ultimate future status of kind of the technological potential or goals for some of these different powertrain technologies. And then, across the powertrain map, we’re looking at six different powertrains, all within the same consistent framework. And those powertrains are the diesel, conventional diesel, the diesel hybrid, compressed natural gas, bio-electric, fuel cell-electric, and then the plug-in hybrid powertrain.
So, moving from the powertrain optimization into our total cost of ownership modeling in the SERA model, the SERA model includes a variety of exogenous data, kind of in three different buckets. There’s cost data, there’s financial data, and then there’s vehicle operations data. On the cost data side, so, obviously, a big component of TCO is the up-front purchase price, which is estimated by FASTsim for these future vehicles. We also include fuel price as one of the direct costs of these vehicles. So, we aggregate a variety of data sources, including the AEO outlook, EPRI and Tesla data, as well as some hydrogen-related data from fuel cell electric plus evaluations out in California, as well as future targets for the hydrogen price. We also include operating and maintenance costs, which are primarily based on literature surveys and the fuel cell electric bus evaluations.
We also wanted to specifically look at, in this analysis, and understand the potential payload capacity costs and dwell time costs. And so, those may be unfamiliar, so I’ll define them real quickly. So, payload capacity costs, as we’re defining it, is effectively thinking about if the advanced powertrain or electric powertrains are marginally heavier than a diesel powertrain, some of that will – may impact the total cargo capacity. And so, what is the opportunity cost of that lost cargo capacity, and is that important from an economic perspective for the total cost of ownership? And then, on the dwell time cost, it’s a similar thing, which is diesel trucks today fuel very quickly because of very energy-dense liquid fuels. So, for advanced power trains such as battery electric or fuel cell electric, would increased downtime due to either re-fueling or recharging negatively impact the economics of these vehicles?
Because, if the vehicle is recharging or refueling for long periods of time, you know, how much does that really cost the fleet and the fleet operator-owner, and is that material? And, if it is, let’s understand what are the technology challenges there and what are the opportunities for the Department of Energy and the national labs to really focus their R&D to minimize some of that economic impact and really help accelerate renewable energy transition and decarbonation of these sectors. And so, that was a particular focus of this work, and we’ll talk a little bit more about that in the scenario definitions in the following slides. But, before that, we’ll quickly cover the financial data, which is just understanding, you know, the tradeoff between current costs and future costs related to different discount rates, and then a variety of vehicle data that characterize the operation of the vehicle, right?
There’s the fuel economy and weight, which we get from FASTsim. There’s also the VMT curves, so, basically, how much, you know, the vehicle, or the truck, is driven as a function of time or each year. And so, different driving profiles have different fuel expenses and fuel consumption profiles, as well as different attributes, such as lifetime of the vehicle. And the final output, really, from SERA is a total cost of ownership across the three different time statuses, the 2018, 2025, and the ultimate. And then, for each of those six powertrains, again, all within this apples to apples comparison. And a particular point of note here is that there is a lot of uncertainty in a variety of these different datasets, and so, we used a – effectively incorporated a low, medium, and high value for all of the different data provided here, and that allows us to effectively say what is the total cost of ownership and what are the error bars around that TCO, and then how do those error bars change over time, and how do the error bars of one technology compare to basically the uncertainty of the TCO in another technology? Can we make some robust decisions around that moving forward?
So, that was the analysis framework. So, from a scenario design perspective, as I kind of alluded to earlier, we were really looking to design scenarios that reflect the typical business operating models that are observed today. And the four scenarios that we looked are shown in a two-by-two matrix that shows payload limitation on the vertical axis and operating shift on the horizontal axis. At the top left, for example, we have a single shift limited operation, which is really a way of representing a truck that is only operating for, you know, maybe a single shift, maybe eight to 12 hours per day, has ample time to refuel or recharge overnight, and, additionally, it is volume limited. So, it’s carrying a very light payload. So, any potential marginal increase in weight of your powertrain due to electrification of that wouldn’t have a material impact on the cargo capacity itself from a weight perspective.
And so, basically, you could have a slightly heavier powertrain and not really worry about any lost cargo capacity from a weight perspective. At the bottom right is kind of the opposite scenario, where you have a multi-shift, weight-limited operation. And so, that’s more reflective of maybe a bulk hauler or heavy hauler with team driving. So, that would be, you know, where, for example, a Class 8 tractor that is up two or three shifts, kind of around-the-clock operation. So, any downtime would have an economic impact. You’d have drivers in the truck idle while it’s trying to recharge and refuel. And so, we can calculate the economic impact of that. And then, also, it’s weight-limited operation, which means that it’s – the cargo that it’s hauling is very heavy. You’re basically up near that 80,000 pound weight limit. And so, by having a slightly heavier powertrain on the tractor, you’re reducing the potential cargo capacity on the trailer.
And so, if we did that, we do want to account for that economic cost, and we want to understand is that a material cost, is it very important? And, if so, you know, what are the main R&D levers that we have to reduce that economic cost and really help understand how we can decarbonize that sector or that market segment? And so, those are the four different business operating models and scenarios that we’ve evaluated. And so, we’ll show the total cost of ownership for each of those independently and kind of walk through some of the key drivers and key differences across the powertrains. But, before we do that, we’ll quickly jump into kind of our assessment of the medium- and heavy-duty commercial vehicle space and kind of the amount of vehicles and fuel consumption that we’re covering within this specific analysis.
So, again, we focused, primarily, on Class 8 tractors and Class 4 parcel delivery trucks to really bracket the commercial vehicle space from an analysis perspective. And everything shown here in gray is kind of not covered within this work, which are, you know, a subset of the Class 8 market that is not the Class 8 tractor and then you’ll see there’s Class three, Class 5, six, and seven, which we didn’t evaluate. But, within the Class 8 and Class 4 sectors or Classes, we can look at the different shadings of blue and orange. So, those reflect the different operating scenarios that we just discussed, whether it’s single shift volume limited or multi-shift weight limited, or anything between. And you can see the relative breakdown, and effectively, what we’re looking at, especially in the Class 8 tractor, is that each of these shifts is relatively important.
It is geared a little bit more toward single-shift operation is the way we’re defining it. But, either way, the multi-shift operation is also very important, and a large component of the overall medium- and heavy-duty vehicle fuel consumption. And I will note that this data is based on the kind of pretty outdated segmentation of the VIAS data, the Vehicle Inventory and Use survey, which is based in 2002. We did do a bit of work to update that data with more recent HIS market or previously-pulled registration data. So, this was a slightly updated representation of that commercial vehicle space. And so, overall, we’re covering a fair bit of it. And then, zooming in to just the Class 8 and the Class 4, so you can see, again, the breakdown across the different scenarios. And then we also compare kind of the VIAS data here, shown on the left, with some other data sources that look at, you know, what fractions or segment, or what portions of the market are really weight limited for Class 8 and Class – or for Class 8 tractors and what proportion of the market is kind of multi-shift operation for Class 8 and Class 4.
And so, you know, based on the Polk and VIAS adjusted analysis, we see roughly around 33 percent of Class 8 tractor fuel consumption is used by tractors that could be in this weight limited Class. NACFE, back in 2015, estimated maybe closer to two to 10 percent of Class 8 tractors without any rigorous analysis, and then Schoettle et al. did a survey that indicated that it was potentially up to 54 percent of Class 8 tractors may weigh out. So, definitely a significant difference across the data sources that we have, but I think the overall conclusion there is that a very large or potentially substantial portion of the Class 8 tractor market could be operating in this weight limited scenario or operating model. And so, it is very important to understand, you know, what are the potential payload capacity costs, are those very important, and, if so, what technologies really make the most sense and how can we improve technology performance and status to overall reduce the impact of that advanced powertrain on cannibalizing cargo capacity.
Then, on the multi-shift side, again, we see some divergence in the datasets that we have to estimate kind of how many vehicles are really in multi-shift operation. VIUS datasets maybe indicate up to 13 percent of Class 8, maybe one percent of Class 4 parcel delivery, whereas the Schoettle et. al survey kind of pointed closer to maybe 6.2 percent, where long-haul team drivers – so in that multi-shift operation, and then closer to a third of all trucks have long-haul and overnight stays. So, again, multi-shift operation does appear to be very important, especially for Class 8, so we definitely wanted to use this analysis to better understand some of those economic drivers, those payload and dwell time costs, in particular for Class 8. So, that was a quick overview of the approach, the analysis framework, as well as kind of the need for this analysis and why it’s important.
So, we’ll jump into the results, and we’ll break down the results according to each different vehicle definition. We’ll start out with Class 8 short haul, which we’re defining as kind of a Class 8 tractor with around a 300-mile range. We’ll then talk about a Class 8 long haul with a little bit longer range. And then, we’ll also talk about a Class 4 parcel delivery very briefly at the end, and then we’ll jump into conclusions. So, for the Class 8 short haul, 300 mile range, we’ll first walk through kind of the initial scenario. So, we’ll have one slide for each of those four scenarios in that matrix that we discussed before. And a couple parameters that are key to this analysis are summarized here at the top left where we have a Class 8 short haul focused in the mid-Atlantic region just for simplicity, annual VMT around 60,000 miles per year, or roughly 230 miles per day. So, again, it’ll be able to be used for the day and have ample time to refuel and recharge overnight.
With a million mile lifetime, that would be correlating to roughly around 16 or 17 years, and then we’re kind of using middle of the road estimates for everything else. And the bar chart shown here at the bottom is the total cost of ownership in dollars per mile across three different time periods, 2018, 2025, and ultimate, and across the six. And, for each time period, there’s the six different powertrains within. There’s diesel, CMG, hybrid, plug-in hybrid, battery electric vehicle, and fuel cell electric vehicle. And so, within each bar, then, you have the dark gray, which is kind of a general operations, cost, which we assume to be consistent, and the same across all the different powertrains, which is effectively a base case – driver wages, insurance, permits, licenses, and tolls. Basically, what you need to operate the vehicle. And then, on top of that is really the differentiator across different powertrains, which is the light gray, which is, you know, upfront purchase price of the vehicle, which is reflective of the powertrain components costs; the green, which is O&M costs, which are different across different powertrains; the reddish color here is the fuel expense; and then blue and yellow will reflect the dwell time costs and payload capacity costs, if they are incurred.
So, for the single-shift volume-limited scenario shown here on this slide, they don’t have any payload or dwell time costs because, again, we’re assuming there’s ample time to refuel and you’re carrying a very light cargo, so there really isn’t that impact of a slightly heavier powertrain on the overall cargo capacity. And so, with that, we can kind of jump into some of the results, and you can see the error bars here represent uncertainty in fuel prices in O&M costs. And so, that’s a way of kind of gauging, OK, within the bounds of uncertainty, how do these different powertrains compare to each other and compare across temporal time periods. And, basically, what we see is that with current technology status, or at least 2018 technology status, the advanced powertrains are marginally more expense than conventional. I don’t think that’s a big surprise.
But, certainly, by 2025, in particular, the battery-electric short-haul tractor can be very, very competitive with diesel and CNG and, potentially, even lower cost, depending on what fuel prices are realized or what electricity prices or realized. And, in particular, by the ultimate time period, that Class 8 short-haul tractor really can become the lowest cost, because it’s, you know, a very efficient powertrain and overall the on-board battery isn’t substantially large. And so, with a reduction in battery prices, and with low electricity prices, you can become a very, very efficient and low cost, from a total cost of ownership perspective, electric tractor. On the fuel cell electric tractor side, you know, the primary components of the costs are the up-front purchase price, the light grey bar, as well as the maroon bar. So, that’s the hydrogen price.
And so, we can see as that hydrogen price comes down over time, from $10 per kilogram in 2018 to closer to $4 per kilogram in the ultimate scenario, the fuel cell electric powertrain also becomes very competitive with the diesel powertrain. So, moving into the second scenario, we can look at the single-shifts and weight limits. So, this is, again, the Class 8 short haul, 300 mile range. The only thing that changes here from the previous slide is the fact that it’s now weight limited. So, this is an operating model in which you have ample time to refuel and recharge overnight, but you do have a very, very heavy cargo capacity. And so, if you do have a slightly heavier powertrain than the conventional diesel, then there would be a slight economic cost to that. You basically wouldn’t be able to haul as much cargo and so you’d have to either, you know, increase the number of routes or increase the number of trucks.
And so, that’s what we’re trying to quantify here. And, in particular, I think the main conclusion here is that that yellow bar, which is really only very, very small on the 2018 side of this charge for the battery-electric truck really becomes non-existent by 2025 for the battery-electric truck, and really is non-existent for any other powertrains. So, for this 300-mile range Class 8 short haul, weight limit operation doesn’t really have an economic impact, and the loss of payload capacity costs are really not significant for this market segment, and that’s primarily driven by, you know, the 300-mile range truck not needing a significant amount of battery or energy storage on board, and, also, the 2000-pound federal weight limit exemption for zero emission vehicles, for non-diesel powertrains in particular. So, again, we see a very similar story, which was an interesting outcome of this analysis.
Moving into the third scenario, which is the multi-shift volume-limited scenario, so this is reflective of a Class 8 short haul tractor that needs to be, you know, maybe operating near continuously throughout the day, but also uses a fairly light payload or something that’s not payload limited. And so, here you can see the blue bars start to show up primarily around the battery-electric tractor, a little bit on the CNG tractors and fuel cell tractor. But, pretty much, predominantly in the battery-electric tractor, just because of the longer refueling or recharging time, we assume there. And so, for this analysis, we did assume a 500-kilowatt charging rate for this mid case, which is what we’re showing here with a high potential of one megawatt charger, which will help us understand, you know, do we need to – from the national lab and DOE perspective, do we need to increase the amount of charge – or the power ratings of some of these charges, and what is the economic impact from a total cost of ownership perspective?
But, here you can see that for this particular market segment, that multi-shift operation, you know, the fuel cell electric truck, dwell time is really not significant, slightly lower than the CNG, just because of the lower onboard storage requirement because of the higher fuel economy or fuel efficiency of that fuel cell electric powertrain. And by 2025, the battery-electric vehicle and the fuel cell electric vehicles can be competitive with diesel if low fuel prices are realized. And certainly by the ultimate time period, the fuel cell electric truck powertrain could be very, very competitive and, effectively, the lowest cost powertrain, depending on what fuel costs are realized. And then, scenario four is the last scenario for the Class 8 short haul 300-mile range, and this is reflective of multi-shift weight-limited operation. And, again, that’s for a Class 8 tractor that needs to be up all the time, but also uses, or carries a very heavy cargo capacity.
And so, for this scenario, we’re assuming kind of 1,000 or 100,000 miles per year operating, so that’s around 380 miles per day. So, you would have that refueling or recharging in the middle of that shift or middle of that day. And then, again, we will see payload and dwell time costs incurred here. And so, this is kind of like a superposition of the previous scenario. So, we have that blue and that yellow bar showing up in particular for the CNG and battery electric vehicle powertrains, and the dwell time costs are pretty close to zero for the fuel selection powertrains because of the high and faster fueling of the hydrogen into these trucks. But, as we saw before, kind of as the technology improves, as the battery energy density improves, the battery electric truck dwell time costs does go down a little bit, and the payload costs effective go to zero by 2025.
But in the ultimate time period, there’s still that marginal cost for a battery electric vehicle. So, it may be hard to decarbonize this particular market segment with a battery electric powertrain. However, the fuel cell electric powertrain is complementary to that, and will allow, you know, the fast refueling as well as long-range capability, and still be very economically competitive with conventional diesel vehicles today. So, that was the Class 8 short haul. So, now we’ll jump into the Class 8 long haul, so a 750-mile range vehicle. So, really, just increasing the amount of onboard storage required to achieve that longer range, as well as having a slightly different duty cycle, which results in, you know, different fuel economy and a different fuel expense than what we saw before. But, again, we’ll walk through the four different scenarios for this particular market segment.
So, in particular, this is a Class 8 long haul 750-mile range in the mid-Atlantic region. For other census division regions, you can take a look at the report. They’re all summarized there. In terms of VMT, this Class 8 long haul truck is assumed to travel about 150,000 miles per year or roughly 580 miles per day, based on the typical operating day or fractions of the year. And so, there would be ample time to refuel and recharge this vehicle overnight. Total lifetime would be just under seven years, and then we would use mid cost or kind of other primers for these bar charts. And, again, the plots themselves look very similar to what we had before, total cost of ownership on the vertical axis across the three different years, and the six different powertrains. And we’ll be zooming in on some of the results.
On the 2018, we can kind of see with existing technology, cost, and performance data, you can see the high up-front purchase price of the battery-electric truck for a 750-mile range is primarily driven by the large battery that has to be on board to achieve that 750-mile range. And so, with current battery costs, that results in just a very high up front or just price of that battery-electric tractor. And then, conversely, the fuel cell electric vehicle is a little bit more competitive from an upfront purchase prices. However, the high kind of VMT profile and high hydrogen price ultimately result in a pretty large fuel expense from a total cost of ownership perspective for the fuel cell electric powertrain. However, by 2025, both the battery-electric vehicle and the fuel cell electric powertrain become much more economically competitive with the conventional diesel and CNG powertrains, primarily because the battery electric tractor price declines as the battery price itself, as battery prices come down.
And, in particular, for the fuel cell electric powertrain, not only does hydrogen storage price come down, but the hydrogen price itself decreases from, you know, $10 per kilogram down to $7 per kilogram assumed in this study. And those really result in a pretty significant decrease in the total cost of ownership for these tractors. And by the ultimate time period here at the far right of the plot, the battery-electric tractor and the fuel cell electric tractor both become very competitive with the diesel electric powertrain, and the fuel cell electric tractor is just marginally lower in the middle of that error bar code at the top of the bar itself is the lowest cost zero emission vehicle for the long haul segment. And, again, the battery electric vehicle total cost of ownership and really upfront purchase price is really driven by that large onboard battery to achieve that 750-mile range.
So, moving to the second scenario for this Class 8 long haul, so the single-shift weight limited operation. So, again, single shift kind of meaning a daily VMT of 580 miles a day, so below the 750 mile range requirement. And so, you’d have ample time to recharge, but it is a weight limited operation, so you’re carrying a heavy cargo. So, any lost or any marginally heavier powertrain would reduce your cargo capacity very slightly, and so, what would be the economic cost and is that important from a total cost of ownership perspective? And that’s really shown here as yellow kind of bars within these bar charts, and from a TCO perspective, the weight limited operation really only impacts the battery-electric vehicle or the battery-electric tractors. And that’s, again, kind of what we saw before, which was that large onboard battery ends up creating a very heavy powertrain, which, then, has that economic cost of reducing your payload capacity, even accounting for that 2000-pound federal exemption for non-diesel powertrains.
However, as battery technologies improve, as the energy density of the batteries improve over time in 2025 and the ultimate scenario, the payload capacity cost really does, you know, shrink to half in 2025 and is nearly zero in the ultimate scenario, which is very important. So, even in this long 750-mile range tractor, you know, if technology progress for battery packs continues to progress, the lost capacity cost for that battery electric powertrain could be pretty minimal in the ultimate scenario. For the fuel cell electric powertrain, in particular, kind of what we saw before was that the, you know, the hydrogen price is a huge component and a huge factor for the total cost of ownership in the 2018 scenario. And as that hydrogen price decreases over time into 2025 and the ultimate, the fuel cell electric powertrain for the long haul tractor becomes very, very competitive. And when the ultimate targets are achieved with that $4 per kilogram hydrogen, the fuel cell electric powertrain is the lowest cost zero emission vehicle in this market segment.
So, moving into scenario three, so, this is the multi-shift volume limited operation. So, this is sort of what we saw before, but this is just the multi-shift operation. So, again, this a Class 8 long haul tractor that really wants to be continually used, have high utilization, whether it’s autonomous driving or just team shift driving. We assumed a VMT around 200,000 miles per year, so that’s 770 miles per day. So, again, just above the daily mileage range of these vehicles, so there would be a refueling or recharging event during the shift or during the day that the truck is being utilized. So, it results in roughly a five-year life, assuming a million miles. And, again, the bar chart here shows the total cost of ownership across the different powertrains, and the difference here, again, is just this blue chart or the blue block, which shows us the dwell time impact and the relative economic cost of that refueling or recharging event.
So, if the truck is sitting idle while it’s recharging or refueling, you know, there’s drivers on standby, the truck isn’t being utilized, so there’s the economic cost there. And so, this analysis shows the relative magnitude of that cost and how it impacts the total cost of ownership of the different powertrains. And we effectively see only a material cost for the CNG and battery-electric vehicle powertrains, again, driven primarily because of the large onboard energy storage requirement of the 750-mile range tractor. And, in particular, again, as technology improves, as battery pack energy density improves from 2018 through the ultimate, that dwell time cost does decline.
However, even in the ultimate scenario, that is a nonsignificant – or, is a significant cost and a significant TCO driver, which may make battery-electric vehicles very hard to decarbonize this particular segment of the Class 8 long haul market, and in particular that’s where the fuel selection vehicle powertrain is really complementary, again, with that faster refueling and higher onboard energy density of hydrogen allows the fuel selector tractor to be economically competitive with diesel, potentially in 2025, if low hydrogen prices are realized, and certainly by the ultimate time period, when we assume that that $4 per kilogram hydrogen is realized.
And then, the fourth scenario is kind of the maybe one of the arguably hardest to decarbonize and electrify scenarios, which is a multi-shift weight limited operation for a Class 8 long haul with a considerable range, a 750-mile range. And, again, the assumptions here are that the VMT is around 200,000 per year, so you do have that refueling and recharging event during each day. And, yeah, we are incurring payload and dwell time costs. And so, this is, again, mostly a superposition of the two previous scenarios with some slight differences, but, effectively, the TCO has not only the direct costs of fuel, O&M, and upfront purchase price, but also these indirect costs that are being quantified as dwell time costs and payload costs. And, similar to before, we really see that, you know, as the battery pack technology improves over time, that payload and dwell time costs do decline.
However, even in ultimate scenario, where the DOE technology, or, the assumptions are achieved and technology performance improves over time, there still is a material impact on the total cost of ownership for the battery-electric vehicle powertrain. And so, the fuel cell electric powertrain is, really, the lowest cost zero emission vehicle in this particular market segment, and, again, it’s just predominantly driven by the fact that hydrogen can be refueled very quickly, or fuel cell electric trucks can be refueled very quickly, and then they have high onboard energy density because of the hydrogen stored.
And so, that was – so we covered Class 8 short haul, we covered Class 8 long haul, and now just jumping in quickly into Class 4 parcel delivery, and kind of the relative merits there. And so, for this, we only looked at the scenarios from a operating shift perspective. So, we didn’t in particular look at the impacts of heavier powertrains on potential lost cargo capacity, assuming that, effectively, from a Class 4, you would just bump up into a Class 5 and account for that on your chassis design.
And so, we’ll just look at two slides here, one summarizing single shift operation, which is shown here in this slide, and then the next slide, which’ll summarize multi-shift operation. And so, this is the Class 4 parcel delivery. Again, just in the mid-Atlantic region. Other regions, they’re all included in the report. They VMT profile is assumed to be about 25,000 miles per year, roughly 80 miles per day. So, you’d have ample time to refuel or recharge this delivery truck overnight. And a lifetime would be roughly around 12 years, and, you know, we can jump right into the results. And I think one of the most important pieces here is that even in 2018, you have the battery electric powertrain for this Class 4 delivery truck could be competitive with diesel today because of the very efficient powertrain and potentially low electricity prices. So, the primary driver is going to be upfront purchase price of that battery-electric truck.
But, as, you know, if battery pack prices could be purchased at a lower price or as long as electricity prices are very cheap, or if diesel prices are expensive, battery electric vehicles for Class 4 parcel delivery trucks could be very competitive, even with diesel current technology statuses, and that’s kind of being demonstrated inn the market with a lot of large-scale fleets moving towards battery electric delivery vans and delivery vehicles. And, certainly by 2025, as the prices of these battery packs continues to decline, the battery electric vehicle option really becomes the cheapest and certainly in the ultimate, it really is the cheapest because of the low upfront purchase price of the vehicle, kind of low overall O&M costs, as well as that potential for low electricity prices and overall low fuel expenses because of the efficient powertrain.
And the fuel cell electric powertrain also is competitive in the ultimate scenario, and it really depends, as we’ve seen before, on that hydrogen fuel price. Certainly by 2025 and the ultimate, really the hydrogen fuel price can be the dominant factor in determining whether that powertrain is economically competitive with conventional powertrains in the market today. So, moving and shifting gears a little bit to multi-shift operations, so for the Class 4 parcel delivery. So, there might be a market segment, so based on the – shown here on the top right is just that tree map. Based on the VIAS and Polk data, perhaps the market segment for multi-shift operation isn’t very substantial today, but with autonomous driving and other kind of higher utilization scenarios in the future, maybe a multi-shift operation is very important for our Class 4 parcel delivery truck.
So, we did want to evaluate that. And so, some of the scenario assumptions here are effectively the VMT is a little bit higher than we saw before, which is around 50,000 miles per year, so 160 miles per day. So, again, you would have the refueling and recharging event in the middle of that shift. And so, what is the economic cost of, you know, having that downtime during the shift and is that important from a TCO perspective? And so, what we see here in the plot shown at the bottom is that really that recharging and refueling event is pretty minimal for all the powertrains except for the battery electric powertrain. And so, the way to think about that is, you know, there’s obviously the downtime for the battery electric parcel delivery van. And so, that would be, you know, at a certain cost for each of those downtimes.
And so, one thing we didn’t do in this study was effectively do a total cost of ownership-based optimization where you look at and design the powertrain to be maybe, for example, for the Class 4 parcel delivery BEV, you have a slightly larger range, so you have fewer refueling events. You’d have marginally longer refueling time, so it might – but it might not shake out. But, anyways, in this particular scenario, the way that we evaluated it here, you would have a fairly material impact to the total cost of ownership for the battery electric truck, and even as technology improves through the ultimate scenario, that overall downtime does impact the total cost of ownership in a material way. However, the fuel cell electric truck in this particular market segment could be a very viable and attractive alternative from an economics perspective because of that faster fueling, because of that high amount of uptime potential. The fuel cell electric Class 4 parcel delivery could be very economically competitive, certainly by the ultimate time period, and it’s really driven by those hydrogen prices, as we saw before. So, it’s really key to lock in a low hydrogen price to refuel those fuel cell electric powertrains.
So, that was a lot of very similar-looking slides, so maybe we’ll just summarize some of the main conclusions, and then we’ll open it up for Q&A. And so, I think, at a very high level, we really identified the different market segments and opportunities for BEVs and fuel cell electric vehicles to be complementary solutions to decarbonizing the particular medium- and heavy-duty markets segments that we evaluated. We did note that battery and fuel cell electric powertrains could be economically competitive with diesel powertrains under, you know, a variety of conditions as early as 2025, especially for those shorter-range applications. And if diesel prices are high, which, you know, they currently are today, they could be even economically competitive, more competitive, sooner.
In particular, battery electric powertrains tend to be best suited for potential shorter range applications or when dwell time is not a concern, and, really, they’re complemented by the fuel cell electric powertrains, which, you know, have a higher energy density for onboard energy storage, so that’s better for longer ranges, and they can _____ having market segments that _____ hydrogen fuel prices are _____ trucks. And powertrains _____ that _____ Class 8 _____ battery EVs, the lost payload capacity is actually pretty small, and that’s primarily driven by that 2,000-pound exemption, as well as just the overall technology performance and status. And a few other points that we noticed, basically, is that in the Class 8 short haul, that 300-mile range segment, and the Class 4 parcel delivery, 120-mile range, battery electric vehicles are the lowest cost zero emission vehicle if dwell time costs are not incurred, and those ultimate targets are achieved, so those battery pack prices come down and the energy density improves over time.
And if those dwell time costs are incurred, then fuel cell electric vehicles are the lowest cost zero emission vehicle for the Class 4 parcel delivery and the Class 8 short and long haul. And then, for the Class 8 long haul 750-mile _____, fuel cell electric vehicles are the lowest cost _____ zero emission vehicle if the targets are met, and that’s kind of regardless of dwell and payload costs. So, really, that longer range almost necessitates that fuel cell electric vehicle _____. And with that, I will stop there and, again, point people to the full report, about 120 pages there, and if you have any questions while you’re reading the report or want to talk about anything, feel free to reach out. Appreciate everyone’s time.
Neha:
Great, thank-you, Chad. So, we have a whole bunch of questions. I’ll just start with one of the – several of the questions were around the error bars. So, if you could speak to why the error bars are unsymmetrical, generally skewed toward higher costs, except for with the FEV, that’d be helpful.
Chad:
Yeah, I can do that. So, the error bars, again, reflect uncertainty, you know _____ reflect uncertainty in the fuel prices and O&M costs. And so, I would say that they’re roughly symmetrical or most of the powertrains except for maybe the BEV. And so, for diesel, CNG, HEV, hybrid electric vehicles and plug-in hybrids, those are – the error bars are predominately driven by the different fuel price, or different diesel price assumptions that we had, which are based on the AEO outlook 2021 data, which basically, like, low oil case from – and high diesel price cases. And so, those are directly a result of the AEO outlook cases. For the battery electric truck, we assumed a low electricity price of around $0.07 per kilowatt hour, kind of in line with the stated values from Tesla. The bar itself represents, like, the actual colored bar itself, represents the AEO reference case for transportation electricity, which is fairly low and is now put up the _____ model.
And then in the high itself is pretty sustainably high, and that’s based on effectively DC fast chargers in the market today as well as some analysis here at the National Renewable Energy Lab that looked at kind of high powered charging for some vehicle segments. And we do note in the report that, you know, the upper bar is likely a very conservative value and that a lot more research and detailed analysis should be done. And I think once that work is done, which is currently underway, we’ll be able to tighten up those error bars, in particular on the battery electric truck scenarios. And just to close it out, for the fuel cell electric tractor, we assumed the median price of $10 per kilogram of hydrogen for the fuel cell electric tractors, and that’s based on kind of current observed prices out in California today.
And then, assuming that won’t be getting any higher, rather they’ll be getting lower as the technology improves, as those refueling stations expand, you get better economies of scale, better economics. And so, those prices really go down towards $7 per kilogram and, ultimately, $4 per kilogram of hydrogen. And so, those are slightly skewed potentially a little bit lower, just because we expect that the really small economies of scale on hydrogen refueling stations out in California will only improve over time as fleets begin to adopt larger fueling stations, and you get better unit economics for each station, better utilization, better capital recovery there.
Neha:
Great, thank-you. Other questions we got were – this is, I guess, two questions in one, so whether the kind of salvage value for the truck at the end of its life _____ the battery to be used in energy storage was considered, and then also whether you would expect different results if we had considered battery swapping rather than a 750-mile truck.
Chad:
Great, yeah, great questions. So, the salvage life we didn’t include. Effectively, we said that all the trucks would have the same salvage value of zero at the end of their lifetime, and there wouldn’t be a potential cost for recycling or, you know, disposing of _____ or a potential revenue from, you know, other things of, yeah, selling, like, for example, the battery into a secondary energy storage market. That’s definitely something we could look at from a sensitivity analysis perspective of something we didn’t include in here because we don’t see that secondary market really taking shape, especially if battery prices and packs continue to decline in prices aggressively as they have. Then, you know, a battery pack being sold today or being used for 10 years and then being sold is probably going to be a very low price, and then discounted back is going to be even less valuable.
But, again, that’s definitely something we can include in the analysis framework, but it’s something we didn’t have for this particular project. And then, for battery swapping, we, again, that’s something we could definitely look at. It’s not something we included in this analysis as the technology isn’t mature or isn’t being really looked at. For some of these large battery packs, for example, ones required for a 750-mile range Class 8 tractor, it’s close to maybe two to three megawatt hours, and so it’d be a pretty substantial battery pack swap. So, again, it’s something we could look at from an analytical perspective, but I’m not _____ just talk about the relative feasibility and applicability of whether that is truly achievable in this market segment. But, definitely something that would be really interesting to look at and see if that is an area to think about how to improve R&D moving forward.
Neha:
Great, thank-you. Can you explain, again, how you did the dwell time and the volume and space limitations cost attribution, if it was a dollar per mile metric or some other manner?
Chad:
OK, yeah, so I think there’s two questions there. First was dwell time, second one was payload capacity. So, for dwell time, we – I’m trying to think if we have a slide for that – we do, so let me – yeah, so, for dwell time, when you effectively have different refueling rates, and we have to calculate how many refueling events are accounted for over the lifetime of the vehicle, so you can take the total VMT and divide it by the total range and – or the total daily driving requirement, and you can get a total number of refueling events. And then, for each refueling event, you can calculate how much time is required for that refueling event, based on the refueling rate going into each powertrain, whether it’s CNG, fuel cell electric vehicle, or EV. And then, based on that total time, you can then scale it by the, or, a dollars per hour unit that we used, which was $75 per hour. We have the references in the report on that. And that ultimately gave us a total cost for each operating year, and then we discounted that back to get to kind of a net dwell time impact on the total cost of ownership.
From a payload capacity perspective, I think we ended up looking at the marginal cost associated with effectively adding a marginal truck to the fleet to move those lost pounds. So, we didn’t do anything from a payload volume perspective, but, rather, just looked at it from a payload cost perspective. So, there are other efforts across the national labs looking at packaging and 3D design for volumetric considerations. This analysis was specifically focused on kind of a capacity and weight cannibalization cost, and so, from that, we effectively said, OK, if you lose, for example, 1,000 pounds of cargo capacity, you know, you would need some fraction of an additional alternative, you know, fuel cell electric vehicle or a battery electric tractor. And so, we would effectively just use a fraction of the total cost of ownership as an adder, which is a simplified way of doing it. And so, it wasn’t a constant across all of them, but rather based on effectively adding an additional advanced powertrain vehicle to that fleet to account for that lost cargo capacity. So, apples to apples, all of the cargo would be moved by all these powertrains. It’d just be an economic way of evaluating that.
Neha:
Great, thank-you. So, the diesel, I guess, diesel prices are projected to go up over time, but in our analysis, fuel costs remain constant. So, is that entire difference explain just by the efficiency improvements?
Chad:
Yeah, so that’s a really good question, and that’s exactly kind of what we’re seeing here play out is both an increase in diesel price over time as projected from the AEO outlook scenarios as well as the improvements in both the diesel engine, the diesel powertrain, as well as reductions in drag, rolling resistance, and everything else. And so, I don’t want to flip through the slides again, because it’s probably distracting, but, effectively, we see the diesel fuel economy go from kind of close to eight to 10 miles per DGE today up to closer to, I think, around 14 to 16 miles per DGE. So, yeah, you do increase your fuel efficiency of that diesel powertrain over time, which, then, offsets that increase in diesel price. So, that’s a really good observation.
Neha:
And then, I think this is the last question we’ll have time for. So, what level of confidence do you have in our method of characterizing pounds of lost cargo? And, I guess, were there other ways that you thought of depicting that?
Chad:
Yeah, so I think there’s maybe two things. There’s characterizing the amount of lost cargo capacity in terms of pounds, and then there’s quantifying the economic value of that. And I will say that we spent a lot of time thinking about the different approaches to quantifying the economic value of lost cargo capacity. We reviewed – we kind of broke it down into four different ways of quantifying it. We reviewed that in partnership with the VTO total cost of ownership analysis effort, and got a bunch of stakeholder feedback on that, and ultimately selected what we thought was the best option and the most reasonable option to quantify the economic impact of it from that marginal trucking perspective kind of. And then, on the weight perspective, again, we’re using a bottom-up analysis, which demonstrates, okay, as technology components improve over time or as they get lighter over time, what are the implications on the powertrain?
And so, we’re really looking at it from a systems component level. And so, we think that, you know, it’s a really good way of doing it. I think, obviously, there’s a lot of uncertainty with that, but we’ve tried to address that as well with the different scenario analyses that we’ve done. So, long story short, I think we have a relatively high degree of confidence in the approaches that we’ve used, but I’m sure we can always do better. And so, if there are ideas an suggestions on ways to improve our quantification in this realm, feel free to reach out. We’d love to do it. I think this is – that was one of the key areas we wanted to investigate in this analysis. So, definitely looking to continue to improve our methodologies moving forward.
Neha:
Great, thank-you, Chad. So, at this point, I’m going to turn it over to Eric Parker to close us out.
Eric:
Thanks, Neha, and thanks, Chad, for the presentation. I know we have way more questions than we had time for today, so we’ll be sure to capture all of those, and you can see some contact info there. We’ll work with the presenters on follow-up afterwards for those _____. And so, with that, I’ll conclude today’s H2IQ hour, and thank everyone for joining. As a reminder, the recording and slides will be available on the HFTO site at energy.gov in the coming weeks, so please check back soon. I saw a few of you asking that. In addition, I want to remind everyone that hydrogen and fuel cell day is just around the corner, coming up next Friday, October 8. We’ll be doing a full week of celebration, so please be on the lookout for announcements and blogs and articles we’ll be posting. I encourage everyone to sign up for our newsletter to get those updates as they come in, including the marquis 1.008 mile virtual walk or run on October 8 at 10:08 a.m. So, follow our social media for that, and we’ll be putting an article out on Monday to kick off the week to show how you can get involved in specific ways and celebrate hydrogen and fuel cell day. So, with that, I’ll wish everyone a great weekend, and goodbye.
[End of Audio]