July H2IQ Hour: ARIES Flatirons Campus MW-Scale Hydrogen System Research: Text Version (2024)

Below is the text version for the "ARIES Flatirons Campus MW-Scale Hydrogen System Research" H2IQ Hour webinar held on July 25, 2024.

>>Kyle Hlavacek: Hello and welcome to this month’s H2IQ Hour webinar. Today we have an update on hydrogen systems research from the National Renewable Energy Laboratory’s Advanced Research on Integrated Energy Systems, also known as ARIES. My name is Kyle Hlavacek with the Department of Energy’s Hydrogen and Fuel Cell Technologies Office, supporting stakeholder engagement and other outreach activities.

As always, please be aware this Webex webinar is being recorded and will be published online in 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 also be a Q&A session at the end of the presentation, and attendees have the opportunity to submit questions using the Webex Q&A feature.

You can access the Q&A feature by clicking on the panel options or more options button depending on your operating system. This feature is located in the bottom right corner of your window, using the three dots icon. To ensure that your questions are answered, please utilize the Q&A feature, and do not add your questions into the general chat. If you have trouble using the Q&A feature, please view the links in the chat for help with this issue.

I’ll now turn it over to Brian Hunter, DOE tech manager for HFTO’s Systems Development and Integration subprogram, to introduce our topic and presenter. Brian, it’s all yours.

>>Brian Hunter: Yes, thank you Kyle. And I’d like to thank everyone for joining this H2IQ Hour webinar. As Kyle mentioned, my name is Brian Hunter. I’m a technology manager for the DOE Hydrogen Fuel Cell Technologies Office, and I’ll be moderating today’s webinar. It’s my pleasure to introduce our speaker, Daniel Leighton. Daniel is a research engineer and the infrastructure and end use team lead with the National Renewable Energy Lab.

And he’ll be providing an overview of NREL’s Advanced Research on Integrated Energy Systems, or ARIES, initiative, as well as the associated megawatt-scale hydrogen research and testing capabilities being established at NREL’s Flatiron Campus. Daniel, I’ll turn it over to you. Please take it away.

>>Daniel Leighton: Thank you very much, Brian. I appreciate it. Yes, so as was said, I’m Daniel Leighton with the National Renewable Energy Laboratory. And I’d like to explain a little bit about I guess what being the infrastructure and end use team lead means. I’m part of our Hydrogen Production, Power, and Storage Group, and within that group we work on up to megawatt-scale electrolysis, production, storage and compression, vehicle dispensing, fuel cell power, safety and sensors, and other technologies. And the infrastructure and end use portion of it is more focused on our hardware side, because we do modeling and hardware together.

So, I’m really our team lead on our big hardware systems, in summary. And our group is part of our Energy Conversion and Storage Systems Center, so we’re looking at a wide variety of technologies, and particularly grid energy storage, which is a lot of what I’ll talk about today. And as was introduced, I’m going to talk about our Advanced Research on Integrated Energy Systems, or ARIES, research initiative. And more specifically talk about our Flatirons Campus hardware capabilities at the megawatt scale for hydrogen systems.

And I appreciate your time today. I’m obviously the presenter here. I’m also the PI on the project, but I do want to shout out to the entire team at NREL both within the Hydrogen Production, Power, and Storage Group as well as outside of that group. It takes a lot of people, a lot of work, to make all these projects a reality. I’m not doing it by myself, I’m just the presenter today. So thanks to the team there.

So, stepping up, I guess one really higher level really quickly, what is NREL? So NREL is about 4,000 folks strong at this point. And we are a federally funded research and development center, which means we are not federal employees. But the NREL campus is owned for instance by the Department of Energy, in this particular case, and we do research, right? That’s our primary focus. And of course being the National Renewable Energy Laboratory, we work on renewable energy topics.

But that also does span into things like sustainable transportation and fuels, energy efficiency of buildings and industry, and then energy systems integration, which is a really big piece now. How do you bring all of these different technologies together? On the control side, on the integration side. Looking at topics of resiliency and cybersecurity, and a whole big exhaustive set of other things. And I won’t go through all of them, but that gives at least a flavor.

We’re primarily funded by the Department of Energy, but we actually have an enormous number of partnerships with other entities as well, including a lot of different industry folks. People from academia, and other government organizations. Both other federal and other state and local government organizations. You can see on the kind of pie charts on the right here, just some breakdown on the number of agreements and funding type we have from partnerships.

So partnerships are a big part of what we do. In some cases, it’s joint partnerships between—in my particular case, the Hydrogen and Fuel Cell Technologies Office and a partner in industry, for instance, would be an example of a partnership, we may do a cooperative research and development agreement together or a variety of other potential structures. And so a big part of what we do is agreements. We do want to get our information out there into the public. We’re also looking to work together to help solve these challenges.

Last note here is we do have multiple campuses. Our main campus is in Golden, Colorado, here in Colorado. But actually, the campus we’re going to talk about primarily, or I’m going to talk about in this topic, is our Flatirons Campus, which is in Arvada, Colorado, but it’s up closer to Boulder, basically between Golden and Boulder. Between Denver and Boulder.

So, diving a little bit further into what NREL does. So, the high-level NREL vision is helping to create a clean energy future for the world. And there are three big focus areas within that, that I’ll elaborate a little bit on. One is integrated energy pathways, which is looking at that integration piece. The grid resiliency side. How these different renewable assets and energy storage assets work together on the integrated energy systems of the future. That’s a big part of what I’m going to talk about in today’s presentation.

Another piece is electrons to molecules, so relevant to this presentation would be things like actually using water and electrons, electricity, to create hydrogen and then using that hydrogen to build other molecules, so you can do things like sustainable aviation fuel or upcycling other fuels, green steel, a lot of different types of technology approaches to use the molecules for other industrial and chemical building uses.

And then finally, circular economy for energy materials. So, this is really from the beginning of the development of these technologies, looking at ways to make sure that they are sustainable in the long run. That there’s a plan for the end of the life cycle of these technologies, and the ability to either reuse or recycle them where appropriate, and making sure that the overall picture is taken into account. So those are three kind of really big technology areas that we focus on as an entire institute here.

And then one last higher-level NREL slide, because this kind of ties into what I’m doing. I really want to emphasize that NREL does, in hydrogen—and I’m going to use hydrogen as just one of many technology area examples, but of course it's the most relevant for us here today. But we do chemistry, benchtop chemistry, at the single cell level in developing new membranes and coatings and technologies, all the way up to the megawatt and soon to be multi-megawatt scale, pre-commercial deployments of integrated commercial or pre-commercial systems.

And so that spans from a very low technology readiness level, or TRL level, up to a very high, from the left to the right, on this plot here. And I’ll highlight on the right side, that’s where the Advanced Research on Integrated Energy Systems has been placed, because that’s really looking at that final integration piece as we scale up, and how these technologies work together to help, and trying to de-risk their deployment into the wider grid. And so, my team focuses on that high TRL level, but NREL as a whole spans across it in many different technologies.

So now, diving down further into the details. The Advanced Research on Integrated Energy Systems, or ARIES, vision, is at least in good part being able to take actual individual technologies at the real megawatt scale—so, on the bottom left here, and I will turn my pointer on here, so on the bottom left here, we have physical assets that are being put in at the campus, such as hydrogen electrolysis, which we’ll talk about. Wind turbines, solar field, PV arrays. Battery storage, et cetera, and being able to integrate those on a real grid, up to a 20-MW substation that has connection back to our utilities.

And we’ll talk a little bit about the architecture of that. But being able to actually demonstrate how these technologies connect together on that real grid, as well as some advanced configurations of those technologies and experimental scenarios to evaluate them. And then doing things like emulating power signatures. So we may not have a real data center. Or actually we do have real data centers at our other campus. But in the case of the grid here, we may not have a real data center on the grid. But we can emulate it, we can create physics-based models that actually go through a digital real-time simulation system, which can create those power signatures on the grid.

We can also do things like emulate offshore wind—because obviously in Colorado we’re landlocked, we’re not going to have an offshore wind turbine, but we can emulate one. So the vision is being able to both take hardware, software models, and then run actual experiments at relevant megawatt to multi-megawatt grid scale, to answer questions around that integration and energy storage and grid resiliency piece of the question that you know, we believe is so critical to help de-risk and deploy these technologies into the future at the relevant grid scales, which starts to go up three orders or magnitude to gigawatt scale, right? Or more.

So, here’s some actual hardware details. So, on the left, this pie chart we have here, this actually represents real hardware on our Flatirons Campus today. And these are actually assets, physically on campus that can be connected together. I like to start by highlighting the three rings that you see here, one red and two blue. These are actually 13,200 volt AC three phase medium voltage grids. So any time I’m referring to our grid research or our medium voltage grid, this is what we’re talking about.

Each device on here, such as the PV arrays, has these dots. And these are four-way switches. One of the ways is the actual device itself. But the other three ways are switches so that you can connect it either to, for instance, the building bus, which is the red one that runs back to the substation and out to our local utility. Or you can interconnect it and island it with any other technology connected on a four-way switch to either of the two blue research grids.

So, that lets us set up research cases that include specific technologies in specific combinations we’re interested in doing experiments on, and actually either island them or not and run those experiments in a way that is safe, right? Because if you do, say, a line-to-line fault to test the resiliency of your electronics, the local utility doesn’t want to see that. And so that’s an example where we would use our controlled grid interface, one of the two of them, or CGI devices, that can actually help isolate that from the grid.

So, you can even send that power back to the grid, so we don’t waste it; however, all the utility sees is nice clean power. And on our side, we have an islanded grid effectively by having a DCD coupling in between. I’ve had these CGI devices described to me. Since I’m not an electrical engineer the best way I understand it is that they’re very advanced, very large variable frequency drives, or VFDs. And so they have the ability to inject strange harmonics into the grid. Waveforms that are out of normal voltage or frequency, sagging or surging, or again line-to-line faults, line-to-ground faults, among a host of other capabilities.

And so, these assets really let us test that grid resiliency piece in a closed research environment that’s a safe way to test these technologies before they get deployed into the wider grid. And so, we have megawatt-class wind turbines and PV arrays. We have batteries. We have a load bank. We’ve got connection to our high performance computing. We have a number of other projects in the works. And then the one we’re going to talk about today of course is our megawatt-scale hydrogen system.

So that’s one of the nodes on here. And that’s graphically represented on the right side here. And I like to describe this in a very simple way as a big chemical battery, right? So we actually can take in AC electrons, store energy as chemical hydrogen energy, and then put AC electrons back onto the grid. And so today what that looks like is connection to that medium voltage grid, that goes to a 1.25-MW PEM electrolyzer that produces hydrogen to a dryness specification suitable for vehicle fueling, if we produce in the future.

Also, it puts out that hydrogen at 30 bar or about 435 psi of pressure. That’s then mechanically compressed up to 3,000 psi or 200 bar, which then goes into an existing ground storage system today, which can store up to 600 kilograms of storage in compressed gaseous tanks. That will source a number of end uses in the future, which we’ll get into a little bit later. But right now, the main end use is a 1-MW PEM fuel cell system, which converts that hydrogen, those hydrogen molecules back into electricity, converts it back to AC, and connects back to the grid. So that’s that AC electrons to molecules to AC electrons "battery" functionality. It does other things, but that’s one big thing we’ll focus on today.

So the last slide really demonstrated the hardware side of this grid system, or the multi grid system. And now that entire set of rings has been simplified as this dashed red line here, which just shows everything connected together. Very simple system. But the purpose of this slide is to actually show you the controls architecture side, because it does get complicated a little bit. And it’s an important part of optimizing and getting these systems to work together. And it’s representative of things that you would see out in the field, which is that you might have a hydrogen plant that includes an electrolyzer and fuel cell.

Each of those units may have an individual controller that does all the safety and data acquisition and monitoring controls, that communicates with the higher level hydrogen SCADA system, or supervisory PLC system. And then that talks to a hybrid plant controller. So this is really the grid controller. This would be the grid owner, that can actually do things like measure the wind turbine production and give a power command for what’s desirable to feather the wind turbine blades at times and then send those power signals over to the electrolyzer to opportunistically produce hydrogen when needed. And that’s just one case example.

Obviously we have connection to all of these devices. You can make it much more complex than that, but that’s the core of what it does. This is that hardware controls system that enables connection between all these devices to actually make it into a real functional grid. And then one thing I want to note there is we can actually inject another layer of controls into that plant controller so we can do things like test different control schemes, work on the optimization of the controls, use it for different end purposes, for different research.

And so it’s a really flexible platform that lets us do that relatively rapidly by reconfiguring the hardware side, and then deciding on a control scheme, and then operating the equipment together in a real grid-like fashion. So that’s a bit on kind of the highest level side, and diving into just one chain, which would be the electrolysis portion, to just give a little more information on how this works. You’ve got these software control, power hardware in the loop, experiments that you’ve set up and programmed.

You inject those into that hybrid controller, which is the real PLC hardware that does the data acquisition and fiber communication with all of the PLCs, all the major subsystems like the wind turbine or in this case, the hydrogen system. That hydrogen SCADA then talks with each subsystem SCADA such as the electrolyzer and the fuel cell and the compressor, which I’ve discussed already. And then it also does some direct communication with some safety valves and relays and other things. And then each of these PLCs can talk with, even in some cases sub-sub PLCs, for different subsystems within them or other direct sensor connections.

And so one important topic of research is quantifying what is the communication time between these devices. Working out those communication pathways to keep them fast and let them enable or have them enable experimental tests that we want to do, and make improvements as we go along and optimizations to how this whole system works. So I know there’s a lot of questions on the control side from some industry parties, so I wanted to talk through that a little bit.

So now, moving on kind of into a little bit of the hardware side of things. So I’m going to show, graphically I’m going to show this in a few different ways so hopefully everyone gets a good sense of what’s going on here. This is obviously a cartoon of our campus. We have real wind turbines, real solar fields, just like we said. We have specifically six grid integration research pads currently. Each one has different devices on it that are part of that research grid. We have batteries. We have a digital real time simulator. We have partner equipment. We have load banks. We have PV emulators. And then on the bottom left here, we have the hydrogen system that again includes that chain from AC to DC conversion, to electrolyzer, compressing hydrogen into storage, and then sending that hydrogen out to a fuel cell system, which creates DC power, and it’s converted back into AC power and sent back to the grid.

And then of course we’ve got auxiliary systems like thermal management for all of these units. Water quality, air, compressed air utilities, other kind of functional utilities to get the plant to work together. And so diving a little bit more into that hydrogen system specifically, this is a northern view of this. But it’s the same information that was on the last slide, but in a CAD iso view, simplified, but CAD iso view of what it looks like today. And so again, we have AC power coming into our conversion unit for the electrolyzer. DC power going to the electrolyzer, balance of plant. Dry hydrogen coming out, going to a mechanical compressor. Increasing the pressure, storage density up into our compressed storage through a gas management panel system. That then being able to be routed back out to any end use of which today we have the 1-MW PEM fuel cell generator, which uses oxygen from the air and the hydrogen coming in to create DC electrons, which go to the conversion electronics for that system, which converts it back to AC and completes that AC-to-AC grid.

And then again, like I mentioned, we have water polishing and heat rejection and supply and return trunks to be able to run all of this equipment in parallel so we have sufficient cooling to accommodate that. And then one thing I want to mention is the—we have aCRADAwith a research partner, GKN and SoCal Gas. The technology developer in this case is GKN, for a metal hydride ground storage system. This is going to be over 500 kilograms, over two 20-foot ISO containers. So we’re going to be demonstrating and evaluating that technology for storage. Because today we have functional compressed storage.

But we’re going to look at instead basically, you know, metal hydrides, which are essentially are, in a simplistic way, a metal sponge that absorbs hydrogen as a solid by binding with it when you reject heat from it, which is done with the thermal management system. And then when you add heat to it, it desorbs that hydrogen. And that hydrogen can drive the fuel cell generator for instance. And then we actually will pick up the waste heat from that fuel cell generator to aid in thatdesorption process. So that’s an exciting project that’s ongoing now, and in fact the team is out there pulling the last of the wires that we can energize this next week. It’ll be hydrogen in the coming month or two, but it will actually be energized next week to run everything. So that’s an exciting new capability that’s going in beyond the core capability I’m presenting here.

And I promiseda loton this being a real thing. I think you’ve seen it behind me maybe in the video here. But this is all equipment that’s on the site today. So this is a drone shot of the equipment. The hydrogen system is here down to the south. This was taken just a few months ago. We’ve got a 1.5-MW GE wind turbine owned by the Department of Energy in the background here. Solar photovoltaic field. We have a new control building going in, and then all of those grid integration pad research topics that I mentioned. And so this is that hardware today. Here we’ve got the fuel cell system, power conversion, electrolyzer power conversion, electrolyzer balance of plant, compressor system, gas management panel, compressed storage system, GKN metal hydride storage system, thermal management for that, and then we’ve got water polishing and auxiliaries and thermal management systems.

These containers are just storage systems, which have actually since been moved permanently, because we actually have a road that you can see is being installed through here. And this construction we show off to the side is actually a project, it’s a behind the meter storage project which is focused on DC heavy-duty vehicle fast charging topics. And that will be co-located with future construction of a hydrogen expansion area, which we will increase the size of our low temperature electrolysis capability up to the multi-megawatt level, which I have a few slides on that we’ll talk about. But showing how it all fits together, we expect to break ground in the coming months on, in this grassy area, grassy field here, on that future construction, which also will allow us to do other future research topics as well. It will provide more space for that, which I will again talk about in a few slides.

So, now getting into maybe what I’m calling some of the more interesting bits, right? So we’ve talked about what we built, what it’s doing today, and it’s running. And we’ve done extensive commissioning work. And I want to share a little bit of the commissioning data with you so that you can get a sense of how this system is performing today upon initial commissioning. And so, in this case, it’s a Nel Hydrogen 1.25-MW electrolyzer, which is equivalent to their NC 250 commercial unit that they sell today. We purchased one of the very earliest ones, and we now have it operating at our site.

And let’s take a look at some of the data here. I want to explain this graph. So, in light blue on the bottom here, we actually have the dry hydrogen mass flow rate out of the unit to the NREL side, measured with aCoriolisflow meter, so high accuracy measurement of the actual, produced hydrogen to the tanks, and that is in units of kilograms an hour on the right-side y-axis here, over time on the x-axis. And then in green, we have the internal hydrogen pressure, which is in bar, again on the right axis, and that’s a good measure of the control stability of the balance of plant of the electrolyzer.

And that’s one of the things that during commissioning, to be able to do really dynamic operation, we worked with Nel Hydrogen, which was the supplier and our research partner, to help tune that system to get better performance under dynamic operation. Minor software changes, and we were able to get very good dynamic following. And so I’ll show that on some of the coming slides. Obviously at steady state you can see slight perturbations from the PID controller doing its work, but it’s obviously quite stable. And then in orange here, you actually see the measured DC stack current with the 100% power point being 3,000 amps. So this is the 100% power point.

And that corresponds to the current DC current on the left of the y axes here. And there’s also in dark blue, the current command that we sent to the unit, which you cannot see, because it’s overlaid so well over the small power test. But we will see it in the coming graphs, so I just wanted to explain that it’s there.

These efficiencies that I note here at the different power step points, these are stack efficiencies. So this is using the lower heating value of hydrogen. And it is using the DC voltage and current to calculate the power being drawn. So it does not include balance of plant, and it does not include conversion losses. I just want to note that, so everybody’s on the same page as to what we’re looking at here, so this is stack efficiencies. And they obviously vary with power point. And these are just some early results showing a few different power points across the spectrum, varying from about 58% up to 66% in this particular set of data.

So, now we get into the interesting bits. This is where we start to get renewable, and it’s the same plot as before in terms of formatting. But now we’re looking at a solar PV profile. This is not actually directly coupled, yet—this was the initial step to run real data, collected at our site from a solar profile that’s been normalized to match our nominal power level of our electrolyzer, to just demonstrate, do the gain and slew rates of the power match the renewables closely enough to run future experiments where we do direct coupling.

Of course, we expected them to, but we actually have to measure it. We worked through, you know, the commissioning of this. This is an important step. And you can see this is a—what I would call a highly variable solar case for a one-hour test that we selected. And again, you can’t even see the power command relative to the measured current. It was very good following of the solar PV. Essentially the electrolyzer is faster than the PV moves in this case, and we actually had a very high efficiency in this case, because it’s a relatively low power point, so we had a 67% stack efficiency.

So moving on into more data, this is now a single wind turbine, which is a pretty variable way to measure it. It’s the 1.5-MW wind turbine on our site. Real data that we used. Picked one of the most variable sets of data I could find. This is only a half-hour profile. And you can see we’ve gone up and down to maximum, minimum power over and over again. Again, this is a prescribed power profile as an initial step. We’ve bounded it with maximum power and minimum power points for the electrolyzer from 100% to 20% in this particular case.

And now you can see the blue line of that power command and you can see differences between what we ask it to do and what it can do, because the gain and slew rates of the wind turbine are much faster than the solar case, and to get a better sense of this, I really want to zoom in on one of the most variable portions of this, to really get a good measure of it. And here you can actually see again all the same platform [inaudible], but you can actually see the lag, slightly, in actual response and measured power at the stack comparative to the power command set point. It was up to around 600 amps or three seconds on the way up, on the gain, and around 250 amps or around one second on the way down. So the electrolyzer’s a little faster on the way down than up.

And now this is just a baseline snapshot. And actually, we’re exceptionally happy with this performance. Really fast performance. This would represent very small energy storage needed in between, if you were to run a completely islanded system. And obviously your grid coupled. There’s other ways to manage this. But this means we really can follow even highly variable renewable profiles with the kind of commercial technology that we have today.

So this is an exciting result. Of course we’re not going to stop there. We’re going to do direct coupled experiments, and we’re setting up to do those in the very, very near term, as well as work with the vendor and others and do research internally to help even make improvements on the speed as well. We believe we can even go a little bit faster, although again it’s probably not even necessary. This is already matching quite well with the renewables.

So, that was the electrolyzer. And now I want to talk also about the fuel cell. And so the first thing I’ll note is the fuel cell is operational and commissioned. And the efficiencies that you see here are different than the electrolyzer. The electrolyzer ones were lower heating value. The DC stack efficiency. These are all in AC-to-AC grid efficiencies. Again lower heating value, but it’s using the actual AC power as the metric. And so that means already included in these efficiencies is all of the overall plant losses due to parasitic power such as compressor power, pump power, fan ventilation power, lighting, controls, et cetera, et cetera, et cetera, as well as all the conversion step losses.

So, you know, from the fuel cell to the AC side we have several steps of conversion of the power type, from DC to AC and everything, and from AC to higher level AC. And so all of those efficiencies are wrapped into this number. And in general, we operated this over a number of power points. We’ve even collected more data since then. But we’re getting around a 50% efficiency, all-in plant efficiency on that.

And when you take the two efficiencies—which is not an exact measure and we’ll do more detailed experiments on this—but just to get a rough estimate of what is that round-trip electron-to-molecule-to-electron efficiency, because that’s a question I get asked all the time, and I’ve promised it at the beginning of this talk. The answer I used to always give was about a third, and our actual measured data is about a third. So not too much outside of our expectations. It depends on power points and on profiles, because the efficiency varies due to a number of factors on all of the equipment based on these things. But it varies in this case as I showed on the bottom here from around 28% up to 35%. I suspect as we do more research we’ll be able to refine those numbers. But the answer is around 33%, around a third efficient round-trip efficiency. And not an unexpected result.

Of course, that is not as efficient as the round-trip electron to electron efficiency of, for instance, a lithium-ion battery. But if you look at the techno-economic cost analysis for hydrogen as you get to long duration energy storage—days, weeks, months, seasons—and you go to massive scales, gigawatt scales that are relevant for grid applications, the ability to cost effectively scale the energy storage portion of hydrogen using technologies like underground storage shows and indicates that it’s really a potentially extremely viable technology for those application spaces, i.e., it is cheaper than a lot of other possible technology choices. And so this is still a very overall exciting result, demonstrating how this works at the relevant megawatt scale and how this could be scaled up and deployed when you get to really high energy storage levels for long duration energy storage.

One more slide on the fuel cell is actually that not—the experiments I showed previously are just grid following experiments, but it actually also is a generator that can do black start and grid forming functionality. So this is just like a diesel backup generator that you can use for, for instance, a data center or a building or any other relevant application. And we did run an experiment and met a milestone specifically for demonstrating this capability and commissioning it. And we found that even with multi-MVA, multi-megawatt worth of transformers black started, we were able to get a nice smooth start of the grid in 200 milliseconds. Very minimal inrush current. Great performance out of the power electronic system.

And then we were able to run back to the AC grid, up to the megawatt power level, as well as a megawatt-hour of stored energy that we had stored for longer than a week, back to the grid to demonstrate how we could do long duration energy storage for the grid, and actually do black starting and grid forming. And so this is just a demonstration. First flag in the ground to say hey, this system works and we’ll of course do more advanced experimentation from here. In the interest of time, I won’t go through the details of all of this. But there is some detail available on, for instance, power versus cumulative energy output over time of the experiment, as well as some really high speed 50 kilohertz data on the wave form generation on the AC side.

So, the big message here is we built all this equipment. It’s commissioned. It’s running. And so we’ve, during the commission process, part of it is to actually start running some of the research use cases to ensure that the equipment works in those scenarios. And so we’re doing that today. And really the overarching goal here is to work with industry partners and academia and other institutes to answer the questions on the suitability of the current commercial PEM electrolysis and fuel cell technology at relevant megawatt scale, to help de-risk the larger scale deployments that are coming in the pipeline. And so, we can answer a lot of questions in a research environment quickly, safely, and help de-risk the cost and risk to the larger systems.

And so to that end, we’re working on a number of projects. As I’ve mentioned, there’s the GKN and SoCal Gas cooperative research and development agreement, or CRADA, that’s also sponsored by HFTO, that is going to evaluate the metal hydride technology and the suitability for direct coupling with electrolysis and fuel cell technology. We have a CRADA with EPRI and HFTO to look into topics of hydrogen production and grid integration and scaling for the future. We have a project with GE and Nel Hydrogen and HFTO for optimal wind turbine design for hydrogen production where we’ll run experimental validation for the modeling that was done. And then we have a project with EPRI and GTI, another CRADA with HFTO, for evaluating next generation hydrogen leak detection technologies.

And so that’s a project where we actually can do our other experiments at the site and then do measurement and monitoring of those experiments with that technology, which is a nice way to kind of get more bang for the buck there in terms of running multiple experiments at once and in a relevant scale. And then of course we have multiple other DOE projects. I’m not going to do an exhaustive list. We can always, if folks are interested, we can connect on those later or you can search around and find some of this information. But we have things like a FlexPower project demonstrating a lot of interconnection of these devices together in a relevant grid. We’ve got leak rate quantification and more.

And then we’re also working with some direct-funded industry projects that are more confidential, because we have mechanisms to work directly with industry and through different funding mechanisms. And so we’re working with a number of different partners there, including AES Clean Energies is one that has a lot of solar equipment on site, and I’ve been told I can talk about them a little bit. And so, we are doing, you know, research with them because they have coincident equipment connected to our grid as well, that we’re working with them with the hydrogen system.

So, so far I’ve been talking about what is going on today in terms of what exists today, and that’s in blue here. So that’s the megawatt-scale hydrogen infrastructure system. And this is a 2D view of a construction drawing for what we’re doing next. So in brown here, we have in development items. So these are funded and underway that we’re working on them today. We have that heavy-duty vehicle DC fast charging and heavy-duty fueling road that we mentioned that you can see in that photo with the dirt disturbed. And then what I want to talk about next is going to be the multi-megawatt low temperature electrolysis capability, including balance of plant for up to 6 MW of low temperature electrolysis, and then the ability to bring in up to 10-MW electrical systems.

And the reason for this is actually our campus. Even if you go and run the sink it’s trucked in water. We just don’t have a water connection today. And so there is investment on the infrastructure side going in. We already have acre feet reserved in a local reservoir. We’re refurbishing a pipeline and putting in a water treatment plant. Designs are done. We expect to start construction soon, although we’re waiting for some approvals from the state for building out a water treatment plant that will give, will source water for the entire campus but also the electrolysis capability that both exists today and the one I’m going to dive into next, which is our multi-megawatt low temperature research capability.

So, this is an expansion to help support the Bipartisan Infrastructure Law investment, billion-dollar investment into hydrogen electrolysis technologies. And there needs to be some capacity to assist in evaluating the performance of these technologies and how they couple with renewables and other topics. And so the Department of Energy is making an investment into the campus to expand the capability to evaluate larger stack systems at our site. Larger systems at our site. And multiple smaller systems in parallel. And so this is a little bit of a complicated slide. I’m not going to go through all the wording here, although you’re free to read it.

But we are going to build out a full balance of plant for up to 6 MW of PEM electrolysis technology, or liquid alkaline if the core balance of plant was provided. Because we’ll have the rectifiers and the hydrogen connection. Or, for liquid alkaline, we’ll have up to 2 MW of complete core balance of plant and 6 MW for PEM. And at this level of core balance of plant, what that means is actually forklifting in an electrolysis stack, and we stick it into our balance of plant and bay that we will have available. And that way we can very rapidly get back up and running and testing, and change between different partners stacks and evaluating different technology. And so that will include the full suite of balance of plant necessary.

And that might be an individual 6-MW stack, or it could be three 2-MW stacks, or it could be one 3-MW stack and two 2-MW stacks—or sorry, two 1-MW stacks. Any permutation of possibility that fits within our power supply availability and equipment availability, but we’ll have five bays in which to test and switch and reconfigure these stacks.

And then at the full 10-MW system level, we will have the electrical infrastructure, the thermal management, the water supply, the hydrogen offtake, and auxiliary power necessary to bring in a partner system that includes the balance of plant, drop it into the space that we have set aside for that, and be able to relatively quickly integrate it and test the entire completed system. So, we’re really trying to go for maximum flexibility to help make sure that this is future proofed for both the stacks coming in the near term, when we get this equipment online in a 2026 timeframe, as well as the further out stacks and technologies that are being developed and scaled up.

And so this project’s underway. I mentioned that design phase. We already have a design in place and we’re doing, we’re now going through the full detailed safety evaluation process so we can purchase all of the detailed equipment. In parallel we’ve actually purchased the major long-lead items like the power supply, transformers, construction, et cetera. And what we’re looking at here is that same 2D. This is that compression storage concrete pad. Now, an iso view. This is the 6-MW balance of plant area. This is the 10-MW system area.

And we have items like the PEM balance of plant shown here with the five bays and here’s a high level view of what that physical structure looks like. We’ve got transformer pads and DC conversion architecture. We’ve got a thermal management, pumping systems, heat exchangers, liquid alkaline balance of plant, dryer balance of plant, water balance of plant, and then again reserved space to be able to do things like bringing in a 10-MW system and connect our water utility and our thermal management system connections to it, and hydrogen system connections.

And so this is underway. It’s coming in the future. And of course we’ll be looking for partnerships actually as part of this project. We’re not buying any stacks. We are specifically building the entire balance of plant to be able to work with partners. So very interested in hearing from you, if you’ve got specifications you want to share to make sure that we’re trying to make targets that work for as many people as possible as well as just developing partnerships in the longer term, to help out the industry here.

So, I talked a little bit about underground storage previously and the relevance of scaling up for grid level seasonal storage. And so, it’s really exciting that we’re going to actually be demonstrating a 10-metric-ton underground storage capability. And this would be a small-scale unit, of a unit size that could be built into a much larger grid-scale system. But it’s what we can demonstrate today, to help develop and de-risk this technology and demonstrate the performance of it. And on a practical side, we’re going to have a big electrolysis capability. Everything I’ve talked about is in this renewable hydrogen production square here.

And in this area, we’ll actually create the hydrogen, and we’ll have that promised pipeline running through our campus, a bit over a half of a mile in length, to go down to what’s called a remote underground storage site. And that is all underway, and we are out in the bidding phase for the design portion of the design build project that we’re working on for that. And so the exact technology’s not selected yet. But an example of some of the possible technologies would be things like reusing existing capabilities out in industry today to do things like drill mineshafts or drill well bores. And you could do underground storage from hundreds of feet to thousands of feet or inches in diameter to feet in diameter. And exactly what it will be will be determined after we complete our selection of a technology and a partner for this, or a vendor for this.

But it’s really exciting to have this coming. And it will be a really cool demonstration as well as a great campus capability, because we’ll be able to actually do real long duration storage on-site using our electrolysis and then feeding future applications. This pipeline is strategically located and being planned such that it can be used for other future research activities that are potentially coming. And so these are unfunded things, but things we want to make sure we plan our infrastructure now, so we don’t have to dig our pipeline back up later and add a new tap. We want to have taps in the places that we think some of these technologies might be demonstrated out in the future.

And so my last slide here is potential future. Again, this is only potential. But we’re looking at options and partnerships for doing heavy-duty vehicle fueling station demonstration work here, coupled with our existing systems, other hydrogen power systems such as fuel cells, engines, and turbines, and we actually do have some of that already in the works and funded actually. And then we have molecule building topics.

This is a really big one. We’ve got a lot of other experts within NREL outside of my group that are working on topics like ammonia molecule building, green steel, methanol building, sustainable aviation fuel. Talking with other partners. And so, there’s a lot of potential in that space for using that electrons-to-molecules pathway to upgrade fuels. And so this is where we’re leaving space allocated on campus and hydrogen connections to be able to support these types of research activities in the future, if they do in fact come.

Another item we are evaluating, as I have mentioned, the three AC research grids that we’re doing today. And that’s really representative of where we are today with technology out at the grid. And that’s likely what at least some of our early deployments will be. But we do think based on some of our research and cost estimation—of course, it’s not too surprising, but there can be some benefits from DC integration potentially of these renewable and hydrogen technologies together.

And so, a good example of this would be a wind turbine, today has wild AC wave forms at least in the wind turbine we have on-site, created by the rotating turbine that generates a wild AC wave form that gets converted to DC and then converted back from DC to AC, and then in some cases stepped up in AC voltage to a medium voltage grid. Transferred across our grid to let’s say the electrolyzer. Stepped down in voltage in AC through another transformer. And then converted from AC to DC and then finally to the stack. And so that works. We’re going to demonstrate that today. But we’d like eventually to compare it to something like a DC grid, where instead you can do wild AC, convert it to DC, possibly a step up DC to DC, transferred possibly to, you know, DC to DC stepped down, and then electrolyzer. And so that has advantages potentially in capital cost, operational costs, efficiencies.

And so we, going into the future, are looking at the ability to potentially demonstrate these types of activities with the assets we have on campus. A lot of them are going to be able to accommodate that with some hardware changes and build out. Another big one that gets talked about a lot is liquefaction and/or liquid hydrogen storage, especially relevant to heavy-duty vehicle fueling stations potentially. And so, we have at least allocated space for setback distances and other things to be able—and high strength concrete—to be able to accommodate these types of systems going to the future so we don’t have to rebuild and start from scratch.

Again, it’s only a potential option. But it is something we’re potentially interested going into the future. And then another topic that comes up, and NREL does work on the techno-economic analysis and the modeling side of things, is natural gas blending. And even that could connect back to, for instance, engines and turbines that run on blended fuels. And so that’s also a capability we have at least on our radar that we would potentially be able to accommodate in the future. That one requires a little bit more hardware modification on the natural gas infrastructure side of our campus. But it’s at least an option that is potentially on the table there.

So those are just some of the potential future capabilities. It’s not an exhaustive list. We’ve got a lot of other ideas kicking around, and we’d be interested in hearing your ideas. We definitely love having partnerships where we work together to help demonstrate the technologies that other folks are developing and trying to evaluate. But that’s at least a flavor of what we’re looking forward to at this campus.

And so that’s my talk. I’d like to thank you for your attention. Hopefully you got something out of this that gives you at least an idea of our overall architecture for the grid at the campus and overall ARIES vision of how we’re going to integrate all these systems together to de-risk the scale-up of these technologies to the grid level by doing it on a research relevant megawatt scale and being able to do things like run hybrid plants and connect renewables and batteries and solar and wind and hydrogen technologies, among others, all directly connected together.

And so those are experiments that are underway today—was, you know, couldn’t share any results yet. But that’s in the works, and we’ll be able to do more advanced experiments as we go along. Controls optimizations, layers, introducing new technologies and new techniques. Other end-use cases. And so what we’re doing today really can, it was just going to continue going, and we’re happy to work together to do all this. So, yeah. Thank you very much, and I’ll take any questions I think at this point.

>>Brian Hunter: Thank you very much, Daniel. Very informative presentation. Great work. Great presentation. We do have a number of questions. You can see, people are giving you a bunch of applause in the webinar here, so. One of the—there were several questions about how to partner with NREL. You know, what’s the best method for connecting with you and you know, is there a potential for doing any site visits to see the Flatirons Campus there?

>>Daniel Leighton: So, the answer is, we love to do this, right? So we even have folks on staff that help coordinate this exact thing. I think you can actually go to our website, our public facing website. I know you can find my email address there, and you can find other relevant folks and reach out to somebody on our team. And depending on the question, we can help redirect to the person who can hopefully give an answer. We’re happy to accommodate additional discussions with any potential partnerships and evaluate whether it can work as a good fit for us.

There’s a variety of ways you can work with NREL. One big one of course is to watch out for potential Department of Energy and/or HFTO, you know, and HFTO funding opportunities for joint research. That’s something that, there are cycles of that, that come and go. And it depends on just timing. And you can watch out for those and look for opportunities that look relevant for a match between what we’re capable of and what you’d like to demonstrate.

We can also do directly funded projects if there’s interested parties in that. So the industry partners can come directly to us. And there’s a variety of agreement types that depend on what the funding mechanism will be. And basically we’ll work through that with you depending on what your interests are, if you can reach out and talk with us. And as part of that, we can of course accommodate tours as well.

I can’t make promises perfectly for everything, but we’re used to doing it. It’s something we do regularly. I think sometimes in the summer here when the weather is nice, there’s probably three tours a day coming through the site behind me, that’s in the picture behind me. And so it is something we do on a regular basis and we’re happy to have discussions on that front too, to get folks in to actually take a look at the equipment and help really develop those partnerships and then have meetings around that. So hopefully that answers some of the questions there.

>>Brian Hunter: Thank you, Daniel. A couple of questions about whether the presentation will be shared. This webinar is being recorded and we’ll post it on the HFTO website after this call.

>>Daniel Leighton: And we will publish, NREL will also publish the slide deck as well in a PDF format I believe. It should actually release after this I think, so it’s gone through our publishing pathway as well. So you’ll have both options.

>>Brian Hunter: Great, thanks Daniel. There was a kind of a related question about what operational data from this system will be publicly available.

>>Daniel Leighton: So, we have obviously some summary plots that we’ve actually released at the Annual—the Hydrogen and Fuel Cell Technologies Office Annual Merit Review in the past year. You can go back and find those slides there. It’s some of the data I’ve showed here as well as these slides coming out. And we are, we have a number of projects, in particular the CRADA projects that are jointly funded with HFTO. Because those are publicly funded projects, that means data gets published, which is great. We love publishing data out into the open whenever we can.

And so I expect there are AMR slides out there. There’s actually some data in some of those projects as well that has already been put out. And then we will continue to do research in those areas and expect to see additional publications and information come out of those. And then we’re also exploring some other options and avenues to potentially get more of this data out there publicly for everyone’s use. So, you know, stay tuned on those fronts. There should be more information coming. You know, we’re running it now. We’re collecting more data. And so expect to see more data over the coming years.

>>Brian Hunter: Thank you. I think this one was covered by your presentation, but slide six referred to 600 kilograms of ground storage. The question was whether that was a cavern or something else, and I think you showed that, you know, that was above ground storage tanks.

>>Daniel Leighton: Yes, so I’ll quickly just jump in and say we have 600 kilograms of above ground storage tanks operational today. We’ve got over 500 kilograms of above ground metal hydride storage coming in the next couple of months. And then we’ve got 10,000 kilograms of underground storage coming in the next few years. So that one’s a little further out. That’s obviously earlier in the design cycle. But we’ll have over 1,000 kilograms of storage operational by the end of this calendar year and then over 11,000, I think, in the 2026 timeframe or so.

>>Brian Hunter: Yeah, that was, there are several questions about the timing of the 10-ton underground storage. But also like, when do you expect some results or lessons learned from that installation?

>>Daniel Leighton: I won’t make any promises. I’m not the PI on that project. I’m working to help integrate it on the technical side, and I know it depends a lot on the technology selected and the vendors selected, and what the timelines are and construction availability. And it just becomes, you know, a lot of pieces. And I know there’s projections for it. I’m just not familiar with exactly what it will be. But we’re hoping to get it online roughly concurrently with the scaled-up electrolysis technology. So, again I’m, I think the 2026 timeframe. But take that with a big grain of salt. It really depends on a lot of factors, and I can’t remember exactly what the timelines are. I’m sorry I forgot the answer to that exact question. But it’s a couple years out, though, for sure. Right now it’s a dirt field, and we’re doing the design, we’re going through the design procurement and build phase. So that will take a couple of years.

>>Brian Hunter: Great. So, most of the storage you’re looking at right now is gaseous storage. Does NREL also have any capabilities or is considering liquid hydrogen storage?

>>Daniel Leighton: So definitely considering it. No projects currently underway, so nothing funded. That’s an important distinction there, right? And so, but we are at least setting up the high strength concrete, as I mentioned, and then the setback distances to be able to accommodate liquid storage systems in the future, co-located with that site. And so that’s important, because if you don’t prepare for it ahead of time with the site, it can be difficult to integrate it later and meet fire code and other requirements.

And so essentially, we’ve done an evaluation to make sure that we should, you know, we will be able to do it later or we should be able to do it later to the best of our ability and knowledge. Although you know, we’re projecting a target that we’re not exactly sure what it is. But we’ve built in margin for ourselves. And so yes, we’re talking about it. Yes, we’re thinking about it. It’s not funded. But we do have the ability to accommodate it at the campus from a safety and installation perspective and construction perspective.

>>Brian Hunter: Great. And can you speak to the—it talks about product purity. I assume they mean hydrogen purity, produced by electrolyzer and then it talks about otherwise downstream balance of plant items when coupled with solar and wind variable load demand. So, I mean, I guess you can talk about hydrogen purity and if the variable electricity input affects that at all.

>>Daniel Leighton: So, a couple things there. One is we do have the optional temperature swing absorption bed drying system for our unit, that came, you know, we optioned for that in the—kind of your commercial product. And that takes you to a 99.999, so three nine five, ultra purity specification that meets SAE 2719 hydrogen vehicle, fuel cell vehicle fueling specifications. So the gas is guaranteed to be quite clean. And the good news is PEM technology generally does produce very wet hydrogen gas that you have to dry up. And that’s part of the balance of plant. But otherwise very clean, high purity hydrogen.

You don’t have to really worry about a lot of sulfur contamination like you would in a steam methane reformation, or SMR, plant from natural gas, for instance. And so we have really high purity hydrogen. We have not quantified differences for variable operation versus steady-state operation. I don’t anticipate that there will be significant differences. We may have a little—there could potentially be, and this is always a question, what’s the effect on stack degradation.

But degradation in most cases ends up being contaminants on the water side that get filtered back out. Not so much pushing contaminants into the hydrogen side, that I’ve seen. But it’s a topic that we’re continuing to actually do research on. Doing things like in situ hydrogen quality monitoring and sensor systems that are capable of doing that. And I’m sure that we will deploy those ultimately at this campus as well sometime in the future. I guess I’m not sure but I wouldn’t be surprised if there’s interest in doing so, and so maybe more to come on that as well.

>>Brian Hunter: Great. And in addition to the PEM electrolyzers, is NREL testing any other types of electrolyzers, such as AEM?

>>Daniel Leighton: So—

>>Brian Hunter: Or alkaline systems.

>>Daniel Leighton: Yes. We have done at the smaller scale. So, we have done testing on AEM and alkaline technologies. That's at like the tens of kilowatts and smaller scale so far. But that multi-megawatt low temperature capability investment that’s going in, the intention is to start to scale those technologies up too so things like AEM and advanced, you know, future advanced alkaline, liquid alkaline systems up to that kind of 2-MW scale. And so the answer is yes, doing it at small scale today. Hoping and working toward deploying it at larger scales, and going into the future.

And then high temperature electrolysis actually. I’ll plug Idaho National Laboratory, have a lot of expertise on high temperature electrolysis. And they’re actually doing a similar project. They’re doing a multi-megawatt high temperature electrolysis capability buildout as part of a similar, let’s say sister initiative. And so I would refer to them for high temperature electrolysis testing capabilities that’s being built up there in parallel with ours. And I know there’s some information from AMR slides and other locations as well on that.

>>Brian Hunter: Excellent. Good recommendation there. A couple of different questions on how much water is needed, so one question on how much water is needed to produce one kilogram of hydrogen with your systems, and then kind of a related question about how much water you’re currently trucking in, a time and what frequency is required for, to support these systems ahead of the pipeline.

>>Daniel Leighton: So, the trucks can hold about 6,000 gallons of water before they’re over DOT regulations. So that’s about all the water truck you’ll ever get because that’s actually a hard over the road regulation. And so, we have 6,000-gallon trucks come to fill up our 8,000-gallon tank, which gives us buffer, right? To be able to fill it up. Because you don’t want to actually run it dry, because we continually polish it. And the frequency of delivery is relatively infrequent now, because we haven’t ramped up to really high utilization rates yet.

But we do have a number of research projects in the pipeline that are planning for much higher utilization. And so then that means the truck frequency will have to increase. And so that’s just part and parcel of how we’re going to have to end up having to do it until we get that water line established. And then as far as, I can’t, to remember this off the top of my head. I want to say it’s either 8 to 1 or 9 to 1 kilograms of oxygen versus hydrogen produced when you break H₂O apart. And so, it’s either 9 or 10 kilograms or liters of deionized water per kilogram of hydrogen produced. And so that’s the most basic level.

At a higher level, you can look at numbers. It depends on your process, if you’re doing like an RO/DI process, there is waste streams coming out of the RO process. Around a third of the potable water today, for today’s kind of basic industry standard RO/DI systems, is wastewater. Which again, we’re not doing at the campus. We’re actually trucking in deionized water right now. But we will convert over to that process once the pipeline is established.

>>Brian Hunter: Great. So you showed the efficiency of the various systems. But have you looked at or estimated the levelized cost of hydrogen for each kind of resource plus electrolyzer combination?

>>Daniel Leighton: So, we have a number of projects going on right now that are doing exactly that and have scopes of work that include significant techno-economic analysis. And what I’ve talked about for our system, and the data I shared, was really commissioning data. And so that wasn’t targeted at a specific project. But we are now doing the research for those projects to validate the models and help validate that the assumptions that were made are correct. So, we’re really feeding a lot of validation data and measurement data and performance data to those models that the projects that have a lot of component of techno-economic cost evaluation can use this system to help actually project what those are, and what changes can potentially improve and lower those costs.

And so that’s a lot of the research that we do. That’s the nuts and bolts of a lot of the projects that are ongoing. And we will continue to do demonstration of different approaches and different control schemes and other things that enable different, clever ways let’s call them, to try to reduce the cost of, levelized cost of hydrogen.

>>Brian Hunter: And I have one final question here. I know we’re running short on time. They’re asking if you have any chart or curve or any data to show the costs associated with implementing larger scale storage, and kind of how those costs may taper off with using larger scale versus smaller scale storage.

>>Daniel Leighton: Yes, I know—so I personally haven’t done any of that work myself. But I know that some of the techno-economic cost analysis folks here at NREL have worked on topics like that. And I would have to follow up on that. I think we’d have to have a follow-up question, I think. I don’t have the answer at this moment. But I know work has been done on projecting what those costs would be for the different technologies. And, I can say that today for demonstration number one, when we do something. And I know there are some other underground storage technologies deployed in the world. But there’s few to none in the United States as far as we know.

And it’s almost the first time, it’s maybe not quite. But there’s been very few iterations, and it might be a new technology approach. And so will the very first one be super cost effective? No, probably not. It’s research. But that’s just part of learning how to do it, and learning how to do it better, so that as you start to scale it the costs go down. Another big factor there too is where these types of technologies, if you make one well bore versus a thousand well bores, you’re going to have a lot of cost savings as you get to that scale, because you’ve mobilized the equipment.

You’ve got a design. And then you just bang, bang, bang, bang, knock in different storage systems underground, versus you know, doing just one as a demonstration, that's a lot of design work, a lot of early safety analysis, a lot of work, research work that will go into that. But I do know that we have made it, we do and can help—and again I don’t know exactly what has or has not been done on that front—but make projections for where that should go. And I know that information is out there. I just can’t recall it all off the top of my head.

>>Brian Hunter: Sure. All right, well thank you very much Daniel for fielding all of those questions. And again, great presentation. I’m going to go ahead and turn it back over to Kyle to wrap up the webinar. Thanks.

>>Daniel Leighton: Thank you everyone. Appreciate it.

>>Kyle Hlavacek: Thanks, Brian. And that concludes our H2IQ Hour for today. Once again, I’d like to thank Daniel and Brian for today’s fantastic ARIES presentation. The slides and a link to the recording of this webinar will be available within the coming weeks in the H2IQ Hour archives. 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. Thanks.

[End of audio]