Below is the text version for the "Hydrogen Fuel Dispensing for Medium- and Heavy-Duty Vehicles" H2IQ Hour webinar held on March 26, 2024.
>>Kyle Hlavacek: Hello, and welcome to this month's H2IQ Hour webinar. Today we have an overview of hydrogen fuel dispensing for medium- and heavy-duty vehicles from Shaun Onorato at the National Renewable Energy Laboratory. My name is Kyle Hlavacek with the Department of Energy's Hydrogen and Fuel Cell Technologies Office. I support stakeholder engagement and other outreach activities.
Please be aware that this webinar is being recorded and will be published online for our H2IQ Hour webinar archives. If you experience technical issues today, please check your audio settings under the Audio tab. If you continue experiencing issues, please send me a direct message. There will be a Q&A session at the end of the presentations and attendees have the opportunity to submit questions in the question box.
You can access the Q&A feature by clicking on the "Panel option" or "More options" button depending on your operating system. This is the button with the three dots at the bottom right corner of your window. To ensure that your questions are answered, please utilize the Q&A feature and do not add your questions to the chat. If you have questions using the Q&A feature, please view the link in the chat that I'm about to drop in for any help with the issue. I'm now going to go ahead and hand it directly over to our presenter. Shaun, it's all yours.
>>Shaun Onorato: Thank you. I'll share my slides.
All right. Is that showing up OK?
>>Kyle Hlavacek: Yeah, it's not in the presenter mode, though. There you go.
>>Shaun Onorato: Great. Thank you.
All right. Thank you. So, my name again, I'm Shaun Onorato, systems engineer and researcher here at the National Renewable Energy Laboratory. And so, this will be a discussion on hydrogen fuel dispensing for medium- and heavy-duty vehicles and research that's occurring at NREL. I'd like to make the introduction that this is a team effort. The team is represented by several co-PIs, Dr. Taichi Kuroki, Dr. Jamie Kee, Krishna Reddi at Argonne National Lab, and Lauren Matta at NextEnergy, and many others that helped to enable this research. So, next slide.
So, the presentation overview—we'll start with an overview of medium- and heavy-duty dispensing technology, where we're at in terms of fueling protocols and hardware. And so, this will kind of set up the stage so we're able to talk about some of the research and development efforts that are occurring at NREL, discuss some of the capabilities and achievements that we've achieved to date, current heavy-duty activities, a brief overview of modeling and analysis activities between NREL and Argonne. We could probably have an entire presentation on that work but we'll do a brief overview there and conclude with a summary.
So, in he general overview of medium- and heavy-duty dispensing, the current deployments and demonstrations are really relying on light-duty fueling infrastructure, codes and standards that were written for light-duty deployments, and these things are—currently we're waiting for—as we're revising faster-flow fueling concepts, we're waiting for those to be evaluated, introduced, and implemented.
And so, this is really slowing the—so, this results in slow and partial fueling events for the larger vehicles. This is treating current infrastructure as well as transferring some of the poor reliability and performance we've seen on the light-duty side to the heavy-duty—medium- and heavy-duty market. And so, this could delay fleet deployments and it overall hinders the adoption of potential—of the medium- and heavy-duty hydrogen fuel cell electric truck market.
And so, here we have kind of the core items that we're relying on for the deployment of medium- and heavy-duty dispensing, and so that's fueling protocols—and so, relying primarily on SAE J2601 for 350 and 700 bar pressures. This includes category A through D table-based fueling. And so, that category D really is enabling fueling vehicles greater than ten kilograms. And this is also—we also have an MC formula-based fueling for the light-duty J2601.
And then, in terms of dispensing—sorry, I should start my video there that's available. In terms of dispensing hardware, dispensers that are configured for J2601 category A through D and nozzle assemblies that are designed for 350 and 700 bar with mass flow rates around 60 grams per second peak.
In terms of communications, we're relying on SAE J2799, which is an IR-based communications system, as well as non-communication fueling, which is—kind of relates back to that table-based approach under J2601. And in terms of station architecture, compressions systems that are really based on light-duty bulk hydrogen storage systems that aren't large enough to support medium- and heavy-duty vehicles, as those have much larger onboard storage systems, perhaps gas management panels and precooling systems that aren't sufficient. And this will be a recurring theme for my presentation, but codes and standards that are really lagging in medium- and heavy-duty deployment and development.
So, recently there are—there's been a number of efforts to develop new medium- and heavy-duty fueling protocols and these have been—the development of these have been international. It's an international approach. So, through SAE, EU PRHYDE. There are several efforts in Japan and South Korea. And these are all aimed at conforming with the industry and federal targets to support vehicle deployments.
So, as we're developing these new fueling protocols for faster flow, these protocols require validation. And some of them have validated at—in a limited sense at laboratory—in a laboratory setting as well as in models, but we really need validation, real world validation with fueling components to assess pressure drop, flow coefficients, general ideas, whether—or, I guess feedback on how the protocol can be implemented with hardware that's available on the market, and then identify hardware that may need to be developed to support these faster flow fueling protocol concepts. And so, hardware really is lacking for implementation of these faster flow protocols, things like flow control valves, valves, general high-pressure components that would go into dispensing systems, as well as nozzle assemblies. So, when I say nozzle assemblies that means nozzles, hoses, breakaways, and receptacles. We also have a gap in vehicle-to-dispenser communication systems beyond SAE J2799 for some of the more advanced fueling protocols that are being proposed. And all of these things are being developed in parallel. So, we have a codes and standards gap here where things are not aligned and we're basically taking an iterative approach where we're developing these things in parallel.
So, in general, for the medium- and heavy-duty fueling protocol metrics we're targeting 70 MPa fueling and 35 MPa fueling. So, we refer to those as H70 and H35. We have a variety of mass flow rates that we're now looking at, so ranging from 60, 90, 120, and 300 grams per second. The research at NREL is very focused on this 300 grams per second target. We want to have our hydrogen delivery temperature down to -40℃ but looking at temperatures that are much warmer, up to an ambient condition, so that we can help bring those precooling temperatures a little higher, reduce costs in the station, improve efficiency. And in terms of CHSS capacity—so, that's combined—compressed hydrogen storage system, we're targeting around 100 kilograms, although some of the protocols can go as high as 200 kilograms.
We're looking for vehicle-to-station communications that is both communications-based and default to non-communications-based if that occurs during a fueling event. Right now, the focus is on infrareds, IrDA-based fueling communications methods, but there are some efforts looking into some advanced concepts for bidirectional communication for some of the new fueling protocols. And if we're using a table-based and MC formula calculation we want to have a fueling time around ten minutes. So, that's based on a metric to meet diesel targets—so, a 100 kilogram transfer in 10 minutes translates to 10 kilograms per minute mass flow rate.
So, right now there are a few efforts, as I described. One is the SAE J2601-5 that's been released. We also have the EU PRHYDE protocol that was released. And then, there's others—so, JPEC's dual nozzle approach and the RTS-HFP from South Korea and others that I may not be aware of. And these are all being presented to the ISO group for acceptance into a standard under Working Group 24.
So, recently, the J2601-5 technical information report was released, and that is available at this website that I've included, the SAE website link shown here. And this is very recent. This was February 23 of this year.
>>Kyle Hlavacek: Hey, Shaun, we lost your audio.
Hey, Shaun?
[Silence from 0:11:25 to 0:12:30]
>>Shaun Onorato: All right. Do we have audio back?
>>Kyle Hlavacek: Yep. Yep. You're good now.
>>Shaun Onorato: All right. Apologies for that. Not sure what happened there.
>>Kyle Hlavacek: We lost you right around when you mentioned that the report was released on the 23rd of February.
>>Shaun Onorato: Oh, wow. OK. Apologies. So, yes, this was released on February 23. And so, now this is fractured into category D, high-flow fueling—so, that's a table-based approach, and then MC formula high-flow general, which is MC formula high-flow G abbreviation here, for pressure classes of H70 and H35—so, 70 MPa and 35 MPa. And so, as I mentioned, this is—we are now looking at different flow rate classes, so 60, 90, 120, and 300, denoted by FM60, FM90, FM120, and FM300.
So, the coupling type—there's going to be different coupling types that will be required. So, when I say "coupling type" that's nozzles and receptacles that will be required to accommodate the different pressures and flow rates. And then, we're using communication—so, comm and non-comm. And this will be IrDA-based with some optional data that will be transmitted for these larger volumes of onboard storage for the trucks.
So, here we're really focused on the SAE MC formula high-flow G section for the protocol under -5. So, H35 being up to 120 grams per second and then—so, 350 bar here, and 700 bar fractured into a flow rate of 60 and 90 grams per second depending on how much data you can transmit as well as the geometry of the nozzle-receptacle interface. And then, really, the focus of the NREL research is the FM 300—so, 300 grams per second. So, this will be much physically larger hardware to accommodate these flow rates. And so, the MC formula high-flow G is taking a very dynamic approach. So, it's control based on actual fueling conditions. And then, the static approach under the traditional category speed fueling that we've been using for medium- and heavy-duty vehicles, which is the table-based approach.
I wanted to take a quick—or, make a quick note here about communications. And so, for non-comm you're going to get a fueling speed that's pretty average. So, this is going to be your IrDA-based SAE J2799 approach for communicating between the dispenser and the vehicle. So, for non-comm you'll get an average fueling speed with an ending vehicle state of charge that might be low. So, if you perform a communications fill for 2799 you'll get a fueling speed with pretty good with a vehicle ending SOC that's not too bad. And now, as we're looking into some of these faster fueling protocols where we're sending optional data you'll get a fueling speed that's optimized with a very high vehicle state of charge.
So, I'd like to kind of make a quick overview of dispensing components. So, these dispensing components—so, nozzles, hoses, breakaways, and receptacles—are being developed in parallel to meet the requirements of the fueling protocols as well as in some applications we're stretching light-duty infrastructure. So, codes and standards—this is going to be a reoccurring theme to the presentation—codes and standards are lagging product development and deployment. And so, metrics are being defined in organizations like ISO for nozzle assembly geometry. We're looking at things like pressure drop, flow coefficients, and thermal mass. These are very important items that we feed back into the fueling protocol to ensure that things like fueling tables, equations are implemented correctly. So, it's kind of a circular dependency here.
Preproduction prototypes are under evaluation for performance and reliability. So, there are three heavy-duty SM300-based nozzle-and-receptacle assemblies and hoses that are evaluation—under evaluation at NREL, which I'll talk more about during this presentation, as well as two hose designs, heavy-duty hose designs.
And so, really, harmonization is required with the fueling protocols and the dispenser systems. And some of these designs are really competing to set the standard. And so, we're evaluating the components in parallel and sending some of the performance information to ISO to help make them—to help them make decisions. And so, some of these—a quick note about hoses. These are upsized for heavy-duty applications. They're based on their light-duty counterparts but usability is really a concern for these devices. They're a little bit heavier. They become very stiff during operation when exposed to the -40℃ precooling temperature. So, usability is definitely a concern there.
This quick graphic shows a hydrogen nozzle and hydrogen receptacle and these are from—images from WEH. Now, these might be a standard nozzle and receptacle set that you would see on a light-duty vehicle when you approach a light-duty station. And so, as I mentioned with the fueling protocols, the future for light-, medium- and heavy-duty fueling is really fractured into two primary pressure classes—so, 350 bar and 700 bar. But now because of the different flow rates that we're looking into we're looking at several nozzle and receptacle classes.
And so, we have H35 for 60 grams per second, which would use an H35 normal flow receptacle, and then we're also looking at H35 high flow, so that's that 120 grams per second that would connect to an H35 high flow receptacle. And then, as we look to 700 bar we're fracturing into several different flow rate classes—so, normal flow, the dual nozzle approach that would be—that the Japanese colleagues are looking into for fueling their vehicles. We have H70 medium flow and then H70 high flow. And so, this is really the area, the emphasis that NREL is investigating, which would be a physically large nozzle, large receptacle to accommodate that 350 grams per second flow rate.
Now I'd like to make a few notes about dispensers and components for the medium- and heavy-duty market. So, again, dispensers are being developed in parallel to meet the requirements of the new fueling protocols but they're really using light-duty-based hardware. And this is a problem because we are seeing some issues with poor reliability and durability. That's been an ongoing issue. We have long lead times on components as well as high cost. There are some significant flow control technology gaps for heavy-duty dispensers. These are physically larger valves that are used for this application. I know perhaps some of you have seen hardware that's in a light-duty dispenser. It's quite a bit smaller. Packaging constraints aren't really a concern. And so, for heavy-duty we're moving to much larger—physically larger lines supplied from the station. We need much larger valves to achieve these flow rates. And so, packaging is definitely a concern for heavy-duty dispensers as well as just availability of these larger components.
Filters and fittings are required for line sizes of 9/16", 3/4", 1". And so, here again codes and standards are lagging product development.
And so, there are a few dispensers that are being prepared for commercial deployment. You can see on here. This is the Bennett heavy-duty dispenser that was installed at NREL as part of a DOE HFTO project. And we also have the NREL heavy-duty dispenser shown on the right side here attached to the vehicle simulator. And so, there's really a lack of test facilities to evaluate the dispensers, the nozzle assemblies, and the fueling protocols and really address overall general reliability and performance.
As I mentioned, advanced vehicle-to-dispenser communications really are not ready and those standards are being released in parallel, and so we'll need to retrofit commercial deployments of dispensers in the future. So, SAE does prescribe of the optional data that should be sent in the IR communications but the J2799 needs to be updated. And for future protocols, like perhaps EU PRHYDE where we're moving to a bidirectional dynamic—much more dynamic approach, we'll need brand new communications methods.
In terms of dispenser designs, they remain somewhat constrained to the number of fueling systems. So, as I mentioned, the components are much larger for these heavy-duty dispensers, and so we're really constrained to a single fueling position, which really isn't' what we're used to in terms of medium- and heavy-duty fueling where perhaps you might have a dual dispenser located at a station.
And in terms of consumer usability, there are some concerns about the larger nozzle designs, especially for the FM300, 300 grams per second flow rates, the hoses in terms of ergonomics. The noise that we've observed during heavy-duty fueling events might be perceived as maybe alarming to customers. And so, these are all things that we're considering as we're validating heavy-duty dispensers and nozzles with heavy-duty fueling protocols.
Some quick notes on station design. So, there's really a number of factors that contribute to medium- and heavy-duty stations. The architecture is really difficult to estimate while a number of these key factors are under development and harmonization is occurring across these various technology areas. So, again, codes and standards are lagging, and this is a problem because we have rapid deployment—development and deployment for vehicle demos going on now. So, OEMs are looking to get their pilot vehicles out on the roads, get them engaged in demonstration events, and so really the heavy-duty stations are either not available or are being created in parallel to some of these efforts that are ongoing with protocols and dispensers and fueling components, which is an issue.
The availability of major subsystems and components and parts is definitely lagging. Some of these items currently just don't exist for heavy-duty stations, one example being large high-throughput compressors. And so, capacity and throughput is dictated by many factors. So, we have vehicle types—so, the medium- and heavy-duty market has a number of vehicle types composed of different fleet sizes, demand profiles, fueling protocols that are suitable to those vehicle sizes and duty cycles, station layout, pressure classes—700 bar, 350 bar, or other pressures that are included in the standards.
And subsystem and component reliability really remains a major issue for stations, as we're seeing a lot of station downtime on the light-duty side. And so, here just kind of wanted to show some of the NREL-collected National Fuel Cell Technology Evaluation Center data. You can find this at the website here. But showing that dispensers, compressors, chillers, and storage systems really kind of contribute to the maintenance items on the light-duty station side. And in terms of failure modes, dispensers, compressors, and chillers really coming in on top here, and so these are expected to be a problem for medium- and heavy-duty stations.
All right. So, we're going to switch gears here and talk about NREL's research in the heavy-duty fast flow area. And so, this is the—I provided a presentation in 2022 kind of describing some of this work under the Innovating Hydrogen Stations Project, so we could include that link to that original presentation. So, I'm describing the creation of the station, the research-based station, and the capabilities there. But I'll do a quick review.
So, this is a first-of-its-kind experimental research capability focused on assessment of new heavy-duty fueling protocols, components, and systems in a real world environment supplemented by modeling and analysis tools. And so, the system is located—or, the research-based station is located at the Energy Systems Integration Facility on NREL's South Table Mountain campus in Golden, Colorado. And so, we really have—I'll describe some of the components on the next slide, but this is supplemented by some great modeling and analysis work in terms of thermal physical R&D with CFD. We have our advanced fueling models, which includes the HD-H2FillS model, techno-economic assessments, and cost of ownership studies with our analysis group. And the idea here is to provide industry stakeholders with validation data and publicly available modeling tools.
As so, you can see the station, research-based station here on the right image. This photo was taken in fall of 2023 with the new heavy-duty fast flow facilities located on the east side of the research station.
So, a quick overview of our fast flow facility. The fueling capability is really designed for pressures up to 70 MPa. And so, we can go down to lower pressures. A precooling temperature down to -40℃ up to warmer and ambient conditions. And we're targeting mass flow rates of 10 kilograms per minute average, 20 kilograms per minute peak.
So, we have—the system is comprised of a heavy-duty dispenser. So, this is configurable, has a configurable flow path with its own nozzle assembly as well as a hard piped line directly to our heavy-duty vehicle simulator. So, our dispenser is designed so that we can swap out components, mass flow meters, valves, other components that might be of interest to industry, to research in the heavy-duty fast flow area. Our heavy-duty vehicle simulator is an 80-kilogram fill mass, so it's equivalent to one Class 8 truck. I have a slide on this following this particular slide that we can dive more into the design of the heavy-duty vehicle simulator. But it is configurable volume heavily instrumented with sensors for pressure, temperature, as well as safety.
Our bulk gas storage system is 650 kilograms divided into a low-, medium-, and high-pressure system. We use the medium- and high-pressure systems for heavy-duty fills, so I want to highlight that as we jump into some of the data here shortly so you'll kind of see the difference in flow rates and pressure in some of the charts. We are limited to back-to-back fueling capability. We can really only do one or two heavy-duty fills into the 80-kilogram system a day. We recycle our gas through our heavy-duty gas management panel back to the station for reuse and experiments. And our precooling system is a brine-based R404a chiller circuit with a custom microchannel heat exchanger.
So, you can see some of the components around the image here. General station with the Bennett dispenser in front, NREL heavy-duty dispenser in the back with its nozzle assembly connection. Heavy-duty gas management panel. The dispenser. Heavy-duty vehicle simulator. Heat exchanger. And brine precooling.
So, again, here is our heavy-duty vehicle simulator. So, it represents one Class 8 truck. It's composed of nine tanks for a mass transfer of around 93 kilograms. This is configurable mass, so it's composed of seven Type IV tanks, around 68 kilograms total if you were to fill that entire mass, two Type III tanks, or 20 kilograms. The Type IVs are rated to an internal temperature of 85℃ and the Type IIIs are 121℃. I want to highlight this because we'll be looking at some of the data here shortly, as we want to confirm that the fill events are not exceeding these internal tank temperatures.
Each of these tanks uses an automotive-style on-tank valve with an integrated bulk gas temperature sensor and a TPRD—so, the thermal pressure relief device. We have triple point sensors installed in two tanks—so, one in a Type IV and one in a Type III—as well as pressure transducers installed through the tank so that we can monitor pressure of individual tanks during the fill. This just shows the location of the triple point sensors, pressure transducer, and then the configuration of the vehicle system, Type IV shown in blue and then the green are Type IIIs.
All right. So, this is a quick review of our achievements so far at the heavy-duty fast flow facility. So, research was completed—this research was completed under the Innovating Hydrogen Stations CRADA Project. That project concluded in 2022. And so, the idea here was to construct the heavy-duty fast flow facility and then demonstrate the feasibility of meeting DOE and industry targets for heavy-duty fast flow. So, this is without a fueling protocol and answering the question of can you perform a heavy-duty fast flow fill? How fast can you do it? And is this achievable? And can you do that within the tank temperature and pressure limits?
So, here our goal was to transfer 60 to 100 kilograms in less than 10 minutes at 70 MPa at a precooling temperature down to -40℃. And so, with that 10 minutes, that's 100 kilograms, we're looking at an average mass flow rate of 10 kilograms per minute, 20 peak. And as I mentioned, we want to stay within those temperature and pressure limits.
And so, in August of 2022 we demonstrated a 61.5 kilogram fill in 4.7 minutes. That translated to 13.2 kilograms per minute, 18.7 peak, so very fast. This test was done only into the Type IV tanks to feed into our modeling efforts for the H2FillS model. And then, a second fill into the entire system, and that was 82.3 kilograms done in a very fast 6.6 minutes, which translated to 12.6 average kilograms per minutes and a peak of 23. And so, this met all industry and DOE targets and metrics for fast-fill tests into our vehicle simulator and confirmed that the overall concept is achievable and can be done safely.
And so, now we're sort of taking a new approach where we're dialing back how fast can we do it and we're constraining that by developing or evaluating new fueling protocols with fueling hardware and under a new CRADA, which is a cooperative research and development agreement. And this CRADA is titled "Assessment of Heavy-Duty Fueling Methods and Components." And so, this is with a new set of partners where we're looking into understanding the effects of the fueling protocol architectures on station design, vehicle design, functional safety requirements, and the implications on the total cost of ownership.
So, this project is really divided into three subgroups. We have a hardware group, a modeling group, and an analysis group. The hardware group is really focused on evaluations of real world heavy-duty fueling components—so, that's nozzles, receptacles, breakaways, and hoses—and then applying those and evaluating those with fueling protocols that have recently been released. So, here we're focusing on SAE-5 fueling protocol and then EU PRHYDE and we're evaluating those at the new fast flow research station.
So, we're generating reliability, performance, and usability data and we're also assisting codes and standards groups with standardization activities, feeding some of that information to the modeling group, who is validating the components in our models with—combined with CFD as well as NREL's H2FillS model and updating those models in terms of H2FillS specifically for public release.
And then, on the analysis side, performing the techno-economic assessments to determine the total cost of ownership for heavy-duty station concepts and vehicle architectures. And we're leveraging data generated from the hardware group as well as the modeling group to feed into existing models that NREL and ANL developed and are updating for heavy-duty applications.
Again—so, this is the partner structure here with NREL, NextEnergy, Argonne, and Chevron leading the project and NextEnergy representing a larger group of industry folks from Air Liquide, Shell, Nel, Hyundai, Nikola, and Toyota, and interfacing directly with SAE and EU PRHYDE.
So, here I'd like to talk about some of our fast flow protocol evaluation and flow control strategies for implementing heavy-duty fueling protocols with the heavy-duty fast flow components. And so, we're really moving beyond the fundamentals here to evaluate these new concepts and new hardware. So, we're one of the first to attempt to fully implement the new SAE J2601-5 fueling protocol. And this is definitely a hardware and model validation challenge. And so, we're taking this in a phased approach, and we're using preproduction nozzle assemblies. And so, again, as I mentioned towards the beginning of the presentation we're looking into three nozzle-and-receptacle designs from three different manufacturers and then two different hose manufacturers. And these are all devices that are designed for heavy-duty fast flow with the 300 grams per second, 700 bar fueling.
And so, here we need a validation of the fueling tables that were developed for the individual fueling protocols. We want—we have some test data that was generated from real world evaluations, and we really want to confirm the assumptions made by the architects of the fueling protocol. And so, we're looking to things on how does pressure drop influence the performance of the fueling protocol? What flow coefficients and things are we getting from some of these—evaluation of some of these components? And then, comparing that to just physical implementation of the protocol and how it conforms and providing feedback to those entities that developed them.
And so, here we're finding that current flow control technology has major gaps. There's really a dynamic response that's required from the heavy-duty fast flow fueling protocols, much larger components than the light-duty-based ones. And so, we're having issues with dynamic response and sufficient flow coefficients. So, we really need some new larger valve designs for heavy-duty dispensers that have much faster response, reliability, and lower cost, as well as improved flow characteristics.
So, some of our preliminary tests are showing that because of valve speed we're starting to slowly deviate from our hose target pressure, and that's really just due to valve actuation progressively falling behind. And so, we are investigating some options for bank-switching strategies, control algorithms, valve position prediction, and pressure drop reduction
And so, here—this is sort of a mess of lines but we're really trying to show you here that this is a very dynamic process. And so, we're starting with our medium-pressure system here shown in the blue line, station pressure, moving through the medium-pressure system. We're pausing the fill to allow the flow control valve to reset itself, so it's a very, very slow process. And then, we're starting again with our high-pressure system. And you can see this sawtooth pattern here, cycling through different variations of our banked high-pressure storage system to complete the fill. And so, the red line here is our target pressure dictated by the fueling protocol, and some of these other lines here show our hose—actual hose pressure and the flow control valve position.
So, the chart to the right here kind of simplifies that: a blue line dictating the hose target pressure, dotted lines are the lower pressure limit and the upper pressure limit. And you can see here that we are falling—almost immediately starting to fall behind due to valve speed and then have trouble keeping up—so, that sawtooth pattern corresponds to station bank switches as we move across the fill.
So, this is a recent test that we completed that was on March 12. This was a 73 kilogram fill. This was done in—so we're going to start to talk a little bit different here. We have total fill time and fueling time. So, the total time of the fill was around 7 minutes, but actual fueling time—so, we have that pause here and that's really not included in some of the calculations that are completed for the fueling protocol—so, that's 358.9 seconds. And based on the SAE-5 fueling protocol, our target here in the fueling tables was 330 seconds, so we're right around what we needed to be on target but perhaps just a little bit slower due to that valve speed.
We're achieving an average mass flow rate of 172.3 grams per second—so, that 10 kilograms per minute average mass flow rate that we're trying to achieve per the industry and DOE targets. Peak mass flow rate was around 29 kilograms per minute. Again, we're using that J2601-5 entity formula high flow fueling protocol. This was done into all nine tanks. And our ambient temperature of the day was around 13℃. So, we're a little bit below our ending CHSS state of charge at 92 percent. The protocol requires that we achieve around 95 percent or higher so we're going to try to continue to work through this filling process to try to achieve some of the things that I've already highlighted in terms of minimizing pressure drop, valve timing, bank switching, things like that.
But wanted to highlight here—this is the same chart that was shown on the last slide but we're using our medium pressure system first to start the fill. So, you can see the corresponding mass flow rate shown on the dotted line, pausing the fill for the valve to actuate, and then moving through our high-pressure storage system to complete the fill. And so, really, the hose target pressure again here is in red. Our hose pressure is shown in yellow. So, you can see it starting to fall behind due to that valve speed and then having a tough time catching up through the fill towards the end. However, we're staying within the pressure ramp corridor, so we're able to complete the fill. However, we have some features of the fueling protocol turned off, like some of these peaks here for mass flow rate. So, these are some issues that might cause a fault during the fill. But really, right now we're trying to investigate, can we complete an SAE-5 fill within the parameters? So, we're working through some of these challenges.
Here I just wanted to show the temperatures shown across the fill. So, this would be JT heating across the flow control valve and then the lower lines here of yellow, purple, and green are showing the hydrogen precooling performance across the heat exchanger and then through the hose inlet into the receptacle of the vehicle. So, pretty good fueling performance, staying down around -40℃ and not really dropping out of that particular corridor.
This next slide shows the HDVS internal tank pressures. And so, this is the—pretty uniform pressure ramps across all the vehicle tanks. And then, really, our concern here is the heavy-duty internal tank temperatures. And so, we're wanting to stay below that 85℃ mark on the Type IVs and the dotted lines shown here are the Type IIIs, well below their 120℃ internal temperature. Slight variation between each of the tanks due to some number of factors: where the tanks are located, exposure to the sun, and things like that.
Wanted to make a quick note to some of our modeling work and capabilities. So, again, we could provide a presentation on its own for the large amount of work that's occurring under this project, but we have our hydrogen-filling simulation model, which is H2FillS. So, this is a 1-D physics-based thermal fluid model simulating the real world fueling process between the station and the vehicle. So, the light-duty version of this model is available at the link shown here. The heavy-duty version is expected to be released this spring. So, we hope to have that out very, very soon on our website for access.
And so, we've been using the H2FillS model to model the station as well as our vehicle system and validating that model against real world test data, as well as programming the SAE-5 and PRHYDE protocols into the model and running fueling tables for both SAE and PRHYDE.
I want to make a quick note about our computational fluid dynamics work. So, we're leveraging our high performance computing system, our supercomputer, to run 3D CFD simulations in Anays Fluent. And this is really moving beyond the 1-D-based model and allows us to explore various pressure ramp rates, hotspots, and thermal stratification. And so, we could—as I mentioned, we could do a pretty comprehensive presentation on this CFD work. If you're able to, please tune into our presentation at the DOE HFTO AMR where we will be providing a more comprehensive update on our CFD and modeling work.
We're running a little bit short on time here but I wanted to make a quick note about our analysis work. And so, NREL and ANL have been working on a combined legacy modeling structure to inform TCO and—perform TEA and TCO on the medium- and heavy-duty work that's been completed under this project. And so, this is—the goal of this is to provide a publicly available model structure that can be used by the public for kind of optimizing the overall—applying fueling protocols to station design, vehicle design, and generating a TCO.
So, here we're using H2FillS as an input into HDRSAM, and then HDRSAM providing input into FASTSim and T3CO. So, H2FillS is evaluating fueling protocols as well as the physical conditions—so, ambient temperature, precooling temperature, fueling time. HDRSAM is looking at things like the overall station design, cost of hydrogen. And then FASTSim and T3CO are very focused on vehicle metrics and financial data surrounding the vehicle to provide that total cost of ownership.
So, some quick takeaways here. So, industry supplied us with some metrics to validate the combined model structure. So, they've asked us to specifically look at 70 MPa fueling, -40℃ precooling, and use the J2601-5 fueling protocol as well as the EU PRHYDE fueling protocol. They are very interested in Class 8 trucks as the priority, followed by Class 6 and 4 vehicles. They've noted that demand profile is very critical, but as I mentioned, there's a wide variety of vehicles and use cases for medium- and heavy-duty applications, and so data is really lacking in this area. So, if—we welcome—if you have any contributions that you could make to this analysis, we would welcome demand profile data.
And then, station environmental conditions were an area of concern. So, that would be hot or cold regions throughout the world. And we really need updated costs for some of these components and subsystems, so we welcome your feedback there as well.
In terms of the TEA/TCO analysis side, I just wanted to provide some takeaways from the analysis that's completed so far. So, really, the analysis is investigating the influence of fueling protocol selection on station design and cost. So, this is specific to the HDRSAM model, which is evaluating the cost of heavy-duty fuel cell electric vehicle fleets over various fueling station configurations and demand profiles. And so, compressors and cryopumps remain the largest source of station costs, and that's for gaseous stations and liquid stations, followed by bulk hydrogen storage, and then precooling costs.
It appears that individual protocols—so, we're running these analyses over different fueling protocols—they don't really seem to have a large effect on the fueling time. So, therefore, the cost difference is not that significant between the protocol selection on the station. We are seeing levelized refueling costs generally be cheaper for liquid hydrogen versus gaseous stations. And refueling costs really depend on the individual contributions of different station components. So, here, again, we really could use some updated cost data for HD station components. If you would be willing to provide that information, we'd be happy to speak with you.
In terms of the analysis side on the TEA/TCO for FASTSim and T3CO—so, here we're looking at total cost of ownership and analysis investigating the influence of the fueling protocol on the vehicle side as well as the—including the station design. And so, this is calculated using inputs from HD-H2FillS and HDRSAM into FASTSim and T3CO, which are NREL models. So, the analysis methodology incorporates vehicle performance, leveraging some of NREL's legacy and new data collected for heavy-duty trucks. And so, here we're seeing that fuel price is really the largest cost driver here and the fueling protocol seems to be a pretty minor effect on TCO. Certainly longer fueling times incur but that's really just a modest—or, a modest increase in fueling costs. So, liquid fueling stations lead to lower dispensed costs and lower TCO. And as the technology—and as the fuel efficiency improves for the trucks we are seeing that it could be that hydrogen trucks could become competitive with diesel under various key assumptions.
All right. So, quick summary slide. So, validation of fueling protocols and fueling components are really critical for the success of the medium- and heavy-duty market. Further real-world test data is required to address technology gaps and assess the risk. We really need to validate the models that were required to update fueling protocols, fueling tables, and revisit key assumptions. We need to support codes and standards activities to accelerate their decision-making processes and harmonize efforts across technology areas—so, conducting the necessary validation of components, assemblies, and subsystems. We want to supply decision-making entities with the supporting data that they need.
We need to address component system poor reliability. We really have a lot of high costs for these components. Some of them are just not available on the market with very, very long lead times. Investments in existing test facilities to evaluate components. So, here we really want to focus on exposure to hydrogen at temperature and pressure, and so we've been receiving a lot of feedback from industry that we just really have a lack of facilities to achieve that. We really want to focus on manufacturing and materials R&D to improve durability and reliability. And then, target innovation and development in critical areas where gaps are identified in these areas highlighted above. So, particularly—oops—compressors, flow control valve technology, nozzle assemblies, advanced communications, and other areas.
So, with that I'll conclude the presentation and I'll jump to questions. Thank you.
>>Kyle Hlavacek: Awesome. Thanks, Shaun. We have had several questions come in. First, yes, all the slides and the recording of the webinar will be posted on the HFTO webinar archives in the next week. As far as when you talk about noise, what type of noise occurs during filling? Is it a whistling sound? A pump sound?
>>Shaun Onorato: It's—I would say a high-pitched rushing sound. It's not particularly like a whistle or a… it's—
[Crosstalk]
>>Kyle Hlavacek: Go ahead.
>>Shaun Onorato: It's just louder.
>>Kyle Hlavacek: How heavy are the nozzles? And will there be a limit on the weight?
>>Shaun Onorato: Unfortunately, I can't comment on the specific weight of the components we're evaluating, but they are quite a bit heavier than their light-duty counterparts. ISO is looking into potentially limiting the weight of the components but that is decisions that are still—still need to be made.
>>Kyle Hlavacek: OK. What is meant exactly by "poor hose ergonomics"? Is it stiffness? Weight?
>>Shaun Onorato: Yeah, as I mentioned, weight is certainly a factor but the hoses are just physically larger diameter. So, that equates to some more material that's used to fabricate the hose. The materials become quite stiff when exposed to hydrogen at -40℃. So, before the fueling event they're pretty easy to manipulate but after the fueling event they kind of almost lock into the position that they were in when the fueling event occurred. And so, they're pretty difficult to manipulate and hang back up on the dispenser, things like that. So, usability, being able to manipulate the nozzle easily before and after a fill.
>>Kyle Hlavacek: OK. For your fueling test, where does the hydrogen come from? Is it a reformer on-site at Golden?
>>Shaun Onorato: So, we have a bulk gas storage system, which is that 650 kilograms of low-, medium-, and high-pressure cascade approach. Our hydrogen is generated from on-site renewable electrolysis, which we supplement occasionally with tube trailer fills.
>>Kyle Hlavacek: OK. Does the flow rate drive the max tank temperature/pressure limit you were talking about?
>>Shaun Onorato: I'm sorry, can you repeat that one?
>>Kyle Hlavacek: Yeah. Does the flow rate drive the max tank temperature/pressure limit you mentioned?
>>Shaun Onorato: Does the flow rate…? So, yes. A flow rate contributes to increased—I guess a more rapid increase in tank temperatures. And so, that's the reason for doing some of the fundamental research. And now, looking into the fueling protocols to ensure that we aren't overheating the tanks for these fast flow fueling events. I think hopefully I answered that?
>>Kyle Hlavacek: Yeah. And then, what is the—what diameter is the piping to accommodate these fast flow rates?
>>Shaun Onorato: So, I can't comment on the nozzles, receptacles, and hoses. That's protected information. However, for the station side it's a combination of 1" stainless steel high-pressure lines fed from the station that fracture into various combinations of 3/4" and 9/16" on our particular system. But I think for general you could probably say 1" and 3/4".
>>Kyle Hlavacek: OK. Was the sawtooth pattern on fills mainly from latent valve response or pressure bank switches?
>>Shaun Onorato: The sawtooth pattern is caused by accessing our high-pressure storage system. And it depends on how we combine the high-pressure storage system into number of bank switches. So, those sawtooths are bank switches. So, that's cycling from sets of high-pressure storage tanks to the next.
>>Kyle Hlavacek: OK. Perfect.
>>Shaun Onorato: Which contributes to valve speed and actuation. But yeah, that sawtooth pattern is from accessing the system itself.
>>Kyle Hlavacek: OK. Perfect. Thank you, Shaun. So, that will conclude our H2IQ Hour for today. Once again, I'd like to thank Shaun for today's presentation. The slides and a link to the recording of this webinar will be available in the coming weeks in the H2IQ Hour archives. Also, a quick reminder to register for the HFTO Annual Merit Review in Alexandria, Virginia, taking place Monday, May 6 through Thursday, May 9. Also, be sure to subscribe to HFTO News to stay up to date. Thank you for attending and we look forward to seeing you at our next H2IQ Hour.
>>Shaun Onorato: Thank you.
[End of Audio]