Below is the text version for the "Thermodynamic and Economic Modeling of Boil-Off Losses in Liquid Hydrogen Handling Processes" webinar held on June 26, 2018.
Eric Parker, Fuel Cell Technologies Office
Good day everyone and welcome to the U.S. Department of Energy’s Fuel Cell Technologies Office webinar series. We’ve got another great presentation lined up this month from Guillaume Petitpas on thermodynamic and economic modeling of boil off losses in liquid hydrogen handling processes. My name is Eric Parker. I provide program support here at the Fuel Cell Technologies Office and I’m the organizer for today’s webinar. We’ll begin in just a moment but first I have a few housekeeping items to tell you about. Today’s webinar is being recorded and the recording along with the slides will eventually be posted and we’ll be sure to let you know.
All attendees will be on mute throughout the webinar so please submit your questions via the chat box you’ll see in WebEx. I’ll make sure to get those questions forwarded along and we’ll cover them during the Q&A portion at the end of the presentation. And with that I would like to introduce today’s DOE webinar host, Neha Rustagi who is joining us here at DOE headquarters. Hi Neha.
Neha Rustagi, U.S. Department of Energy
Hi Eric. Thank you. So as Eric mentioned our webinar today is on thermodynamic and economic modeling of liquid hydrogen infrastructure and identifying methods to mitigate boil off. Our presenter is Guillaume Petitpas from Lawrence Livermore National Lab. Guillaume has been working on cryogenic hydrogen for transportation for almost a decade together with Salvador Aceves. While his work originally focused on on board storage such as cryo complex pathways, he started working on dispensing technologies to propose affordable and practical solutions for the entire hydrogen supply chain.
Through DOE funded project’s Guillaume has engineered and built a testing facility for liquid hydrogen that includes a large dewar over 3,000 gallons as well as a high-throughput liquid hydrogen pump made by Linde. His experience with liquid hydrogen both theoretical and hands on gives him a unique perspective on this pathway. And with that, I’d like to turn it over to Guillaume for his presentation.
Guillaume Petitpas, Lawrence Livermore National Laboratory
Thank you, Neha. So I’m going to—so thank you everyone for joining us on this presentation. I am going to—so as you know the topic of the day is liquid hydrogen and so I’m going to go—I’m going to dive right into it. So why do we care about liquid hydrogen? There’s two really striking examples of that how, that shows how hydrogen is, liquid hydrogen is a really good option. So I’m showing here on this slide, on the left there’s the fuel cell forklift example we have today in the U.S. We have more than 14,000 fuel cell forklifts and most of them are supplied with liquid hydrogen. So even for early market applications we always see the need for liquid hydrogen at the fueling station.
The second example is of course AC Transit agency here in the Bay Area that owns the largest fuel cell bus fleet in the world and they – and even there are between 7 to 10 up to 12 buses as of now already need liquid hydrogen to supply their refueling stations. They have two liquid hydrogen based refueling stations here in the bat area. And so, the most important argument and maybe the single most important reason why we need liquid hydrogen is illustrated now at the bottom of this slide. When I show the loading capacity of the trailer carrying liquid hydrogen on the left and see how many trailers, I mean high pressure trailers we need to match that capacity, that loading capacity. And so, for example in the liquid hydrogen trailer on the left you have more than 4,000 kilograms of hydrogen. And it’s as many as 15, 190 bar high pressure trailer hydrogen or five, 350 bar composite high-pressure trailer. So you see that you can still fit 15 times more hydrogen in a liquid trailer as compared to the high pressure, to a high-pressure trailer.
Another reason here is as you can see on those numbers is the cost of carrying this hydrogen, the cost of the trailer in dollars per kilogram of hydrogen. We have at least one order of magnitude between what is being done today for high pressure trailer and what can be done today with liquid hydrogen trailer. Beyond this we also have other reasons that are really—that gives hydrogen a unique role in the deployment of—liquid hydrogen a unique role in the deployment of hydrogen for especially at large-scale.
So the first argument is of course the high density of liquid hydrogen that I just alluded to in the previous slide. And because we have high density we can minimize the footprint of the storage system and the transportation system but also the cost because basically we need less material per kilogram of hydrogen. Another important argument is the high capacity and the short transfer time so that we can minimize the delivery logistic and the scheduling at the station. This is also very important argument to make for liquid hydrogen. You know we have a very low potential burst of energy when you compare 20 K, 6 bar versus 300 K, more than 200 bar. And when you think about many trailers being on the road tomorrow when we have a large infrastructure of hydrogen you really want to make sure that if there is an accident that happens you want to be in the safest condition. And liquid hydrogen provides this very low potential burst energy, more than 20 times less than high pressure room temperature hydrogen.
Another argument to be made for liquid hydrogen is the high throughput that the liquid hydrogen pump can provide at the low distance cost. So when we talk about high-throughput as how many kilograms per hour and low distance in cost is a by kilowatt hours by kilogram—how many kilowatt hours of electricity that we need to move those molecules. And as an example, today liquid hydrogen pump can do more than 100 kilograms per hour versus the best compressors today can only do 40 kilograms per hour. At last there’s a very good argument that can be made for liquid hydrogen that we can see the high density of the molecule all the way to the storage on board the vehicle so that we can have very compact solution like cryo compressed or cold compressed hydrogen.
One way to show this is this—one way to show what it means really for a station is this graph, the figures that I’m showing here. When I compare the capex to the cost of the installed equipment for the station, pretty large station, 4,000 kilograms per day and with 80 trucks, which is kind of large for now. But really where we need to be if we want to make hydrogen happen. And I’m showing on this figure based on numbers that were completed by HDSAM, what is the cost of additional equipment. Each color here corresponds to an equipment at a filling station. So the green is installation. The red is the cost of the pump/compressor. And the blue is the cost of the storage and the dark blue is evaporator and so on.
As you can see on the right the last two slides are the gaseous delivery by [inaudible] which is on the left, those three bars. You see the liquid hydrogen is very [inaudible]. And you can see here that we have some model of difference between those two options yet again because we can afford very low cost for the storage—the light blue colors. And high-throughput and low-cost for the compressor or the pump which is the red color. And so, you see that the cryo compressed pathway for example is up to three times less [inaudible] than solution with the compressor and the pipeline.
So when we talk about liquid hydrogen one of the first things that comes to mind is the fact that we have to boil away a lot of hydrogen when we use molecules. And on this slide here I’m showing a picture that we took here at Lawrence Livermore of this kind of cloud that you see at the vent stack of the dewar which is what most of the people think when they think about liquid hydrogen. Here, we are wasting a lot of hydrogen and this is building here any cryogenic fluid. Any time we transfer cryogenic fluid you will have to boil off some. And the thing that this happens at many stages along the LH2 pathway from the liquefaction plant to the transmission to the station storage to the pump and into the car. Every time we have the likelihood of boil off.
And what we need to understand is how bad is it. What is actually – how much hydrogen do we need to boil off, to boil away to provide one kilogram of hydrogen to the car and is it really that bad? So that’s kind of the goal of the whole, the scope of the work that I’m presenting today. How can we quantify and understand what’s happening as far as boil off along the LH2 pathway so that we can have a better understanding of the whole process and how it can be harmful for the pathway?
So four different sections here, so the first one is I’m going to first talk about the thermodynamic model that simulates hydrogen transfer and that was based on some work by NASA. Second one I’m going to then talk about current practices and what are the requirements on the U.S. DOT perspective. Third, I’m going to look at – I’m going to present some data collection that we did at our filling facility at Lawrence Livermore to support these ideas and at last I’m going to show – I’m going to talk about what kind of boil off magnitude are we talking about and what is the best way to handle this boil off.
So first the thermodynamic model. So this work was based on the previous work that was done by NASA on loading their H2 rocket. At NASA they can transfer very large quantities of liquid hydrogen. And of course, again boil off is a big deal but also what is [inaudible] pressure variation when you load a very large volume of liquid hydrogen from those large fills—up to one million gallons of liquid hydrogen going to the rocket and how did that go and what is the best way to maximize the state of charge. So I took the work that was the basis for what I’m showing today and basically here is just a summary of how the code works and the code is just based on the interaction between two LH2 volumes. And it looks at the different phenomenon that happens during this transfer as far as condensation and evaporation between the liquid and vapor phase but also at the film and looking at energy balance.
And what’s really important here the heat transfer mode that you have based on the relative values of each temperature, the temperature of vapor, the temperature of liquid, and the surface temperature. Depending on the relative position you will have different phenomenon as far versus conduction versus convection. And this is really what the code captured.
So based on this work I had to do a few updates because for example they didn’t take into account the recategorization of states. They were assuming gas and a few of the things like that. So what I did is I took this work from them and I updated it with a bit of interaction with REFPROP for example. And I together mapped that code. And so, what I’m going to show now is a rapid overview of this math lab code. So I’m going to switch now to math lab and show – I’m going to first show the MATLAB code that was – I hope you can see it.
So here the MATLAB code basically is a few different sub buttons and all the sub buttons are actually detailed in a Word document that’s going to be available for download at the website at the link that is at the end of this transition. But I’m going to show you quickly what is – this is the Word document that’s going to be available and that gives more details as to what kind of version for MATLAB do we have and how do you link this code with REFPROP. And here is a description of – I hope you can see it. I guess you can. It gives a description of the different MATLAB files that make this code work.
And basically, what you want to – what’s really important is these files called "runNominal.m." This is what you’re going to have to run in order to compute the results. And this is the second most important, inputs is where you can modify the boundary, the boundary conditions for this code. And so, again here there’s a little bit more definition about where we’re coming from here. And I’m going to now go back to the MATLAB code if I can. So here’s what it looks like. This is the main file that you need to execute.
And you can modify here in this window here even the conditions. So for example, here you can modify the volume of the trailer, the volume of the receiving vessel, so of the dewar and a few things like that. And also of course modify the initial condition for the trailer and the initial condition for the dewar here. So pressure, temperature, and mass of hydrogen. And then you can look at the pressure devices and everything. And so here we have – and we can look at the time window. So once everything is set up then you just go there and you run your code from here. I would not run the code here because it would take about 20 minutes to execute.
But what I can show is that first what kind of output do we have. And the code itself generates all the outputs. So it looks like somebody from D.C. is calling me. So I can’t take this call right now but if you have something to let me know please let me know by over the phone, over this WebEx. Anyway, so yeah. So here on this figure I’m showing here – so the code would output those results when you have for example the height in the trailer as a function of time. You also have the height in the dewar as a function of time. So you can see that you start at very low eventually for liquid hydrogen and you go up to 90 percent fill. And what’s interesting here is of course the temperature as well.
This is your temperature and each volume. Top is a trailer. Bottom is a dewar. So red is a vapor. Green is the surface and blue, that blue is a liquid temperature. And you can see all those things are very dynamic. They change a lot during the fill. And we are capable of capturing what is a ratio of vapor and mass for each. We can see here for example the version of pressure here on this top figure here or Pv1 is the pressure in the trailer and Pv2 in the green is the pressure in the dewar. So all those things you can really capture and you can output in a text file that you can then use in Excel document or something. And you can really quantify what’s happening throughout the fill.
And here how much losses you have when you do this kind of fill. So you start at zero losses and as the transfer goes you lose some hydrogen. So here up to 25 kilograms of hydrogen during that transfer for those specific conditions. And here this is a flow for example of the [inaudible]. So I know this is a rapid overview but to give you a sense of what we can do and the level of details that we can capture with this code. So now I’m going to go back to the presentation and again I hope you’ll be able to see that. So I’m back to – so again if you cannot see it please let me know through the WebEx. Please speak up because I don’t have any feedback from anybody here.
Eric Parker
We’re good to go.
Guillaume Petitpas
We’re good? Ok. Thank you, Eric. I appreciate it. All right. So anyway, so once we have this kind of code and this kind of framework we can really simulate a lot of things. We can do what I just showed with this specific code. Again, this is a code that is going to be released open source and is going to be available for download very soon. And the link that we’ll give at the end of this presentation. So depending on how you can structure this code, you can simulate a lot of things. So you can simulate of course like I showed you transfer between the trailer and the dewar but you can also simulate just the boil off of the dewar alone being exposed to heat transfer so seasonal heat transfer from the environment. And this is what I’m showing on this slide. You can also look at the transfer between the large dewar and the trailer [inaudible] plant and you can also look at the boil off in the trailer. So you have a lot of options.
And so here, again here I’m showing here just the heat entry to a vessel just in order to make sure that the model is capturing accurately what’s happening. And so, what I did on the top left corner is I took the result that we got from our facility here, this is basically the inventory over time so how much liquid hydrogen is being measured over time. And then you see here it depletes over time because we have heat transfer from the environment. And what exactly does that mean in terms of heat transfer? And so, here we are able to capture the heat transfer profile. This is on the top right corner. Here we are able to see that when the liquid hydrogen dewar is almost full we have about 70 watt of heat transfer. As the inventory goes down we will have the heat transfer will decrease all the way to 30 watts, near zero liquid hydrogen left in the dewar.
And you can see here that depending on what outside temperature we have, summer versus winter, you will have a slightly different profile, heat transfer profile. And which makes sense because it makes sense that when you have a little bit less lower temperature in the winter you have little bit less heat transfer. Even though in the end you will have very similar. Because when you’re fully at 70 watts and when you’re next to empty you’re 30 watts.
What’s really interesting here is that with this model we can also look at the different temperature that we have across the dewar as a film, as a liquid, as a vapor and the wall temperature. And quite surprisingly the liquid is actually warmer than the film temperature. So this is a blue versus a green on this graph here which means that the actual density of the liquid is actually a little bit lower than what you would expect by just looking at the pressure set at 45 psi. And so that’s really interesting to quantify exactly what we have, density versus time in the dewar so we have a better understanding of the inventory in terms of mass of liquid hydrogen.
Here on this slide here I’m showing again some results from this model. But here just for the liquid transfer from the trailer to the dewar and on the left you see the inventory of mass of liquid and vapor and so blue and red and also the vented hydrogen in grey. And on the right, you see the variation of temperature and pressure. And you can see as you go through this fill you have less liquid in the trailer and more liquid in the dewar, which makes sense. And you also—and you still have some vented hydrogen. Here it says four but on the face it was 25 kilograms here that you see.
And so here is simulated that we had to empty the whole dewar. You have to read the pressure of the dewar at the end of the delivery so you have up to 100 kilograms of hydrogen being vented at the end of the delivery. So here I’m going to talk in a minute about why we do this vent more in details. But continuing on those results here for bottom fill, we can also look at the transfer of energy so the balance of energy between those two vessels. And I know there’s a lot of curves here but what matters here is the black curve is total heat entry to the volume of vapor at the bottom of the [inaudible] of liquid. And also, the colors are the different contributors to that total heat transfer going into this volume of vapor or liquid.
And what matters here is that the single most important contributor to the whole heat entry in the vapor is actually the pdV effort. So basically, what’s happening is that when you do a bottom fill you are going to compress the vapor. And this compression is going to be a pdV term and is going to be a big factor into winding up this volume of vapor and thus creating boil off. And here for the liquid, for the volume of liquid the single most contributor is of course the transfer of energy in blue here coming from the trailer to the dewar. And you can see here that the green curve is actually cooling down to the pdV is actually cooling down because you have some expansion of the liquid volume.
On the right here, I’m showing also the influence of pressure because it’s really important what kind of pressure you have in the dewar in order to be able to produce your so that you’re – this is specific to each application. You will have different set point for the PRD. And depending on that value you may have different results. And what I realized here that actually this value is not really, is not as sensitive as what I thought. I’m showing here a different value between 60, 80, and 120 psi. You can see that in the end of the value of the PRD don’t really matter. What really matters is initial pressure in the receiving vessel. You can have 30 psi versus 60 psi. You could also double your or threefold the amount of boil off that you could have from the vessel, from the dewar during the transfer.
And I’m showing you also that the influence of initial quantity of hydrogen, one percent versus 75 percent full. And yet again this is not nearly as important as the initial pressure in the receiving vessel. So this is really important to keep in mind that you want to minimize the initial pressure in your receiving vessel before the delivery so that you can minimize your boil off.
So this is it for all of the results from the model. And of course, there’s more to it but in the interest of time I’m moving on to current practices and DOT regulation. So again, as many of you when you think about liquid hydrogen transfer you see those large cloud of hydrogen at the station and this picture here, this picture here is taken at the end of the fill. So basically, your dewar here is completely full and but there’s still some pressure in the trailer. So what drivers typically do is they’re going to vent all of the vapor in the trailer to reduce the pressure so that they can go on the road.
So on the right here I’m showing on this figure the same simulation that I did with my code. And you can see that depending on how much, what is the initial mass of liquid hydrogen in the dewar and what is the pressure that you need to vent from. So from 40 black to 140 in blue psi you will have increasing amounts of venting losses from the trailer. And this is explained by the fact that when you’re completely full your volume of vapor is kind of small but as you are more and more empty or less and less liquid, you have more and more vapor so it takes more and more masses of vapor to be vented to reduce the pressure.
And you can see that this kind of very large number. Right? If you have only 1,000-kilogram hydrogen left in the trailer you could vent as much as 300 kilograms of vapor just to reduce your pressure from 140 to 20 psi. And so that alone could be a very strong argument against liquid hydrogen because if every time we did delivery we have to waste hundreds of kilograms just to reduce the pressure in the trailer, that’s kind of very difficult to make sense. Right?
So what I did is I asked myself why do we have to vent. And there was different answers depending on who I was asking to. But it all came down to the CFR, right, the code of federal regulation that take that really away from [inaudible] that controls all of this. And this code here says that actually what’s important is what is the pressure that you have to have in the trailer before it goes on the road. And it is especially the function of what is your inventory of your filling density in the trailer and also understand you have until your next delivery. And so, this is what I tried to highlight here.
I highlight here in yellow on that text. Again, this is just a copy paste from the CFR code. The CFR code and it says that again if you are completely full of liquid hydrogen and you’re going to have a long time until you do another delivery then you may have to vent some. And you as a driver you have to address your pressure as a function of those conditions. And that so in theory you have to – you have to vent some. Under specific scenarios you have to vent some hydrogen and those pictures here are taken from the empty truck that is coming to Lawrence Livermore and all those markings here are actually required by DOT. So that’s a full understanding of what the driver has to do depending of how much you have in your vessel, in your trailer.
And you can see on the right here this is what kind of PRD pressure we need to have as a function of liquid gauge. And when you’re full you need to have a certain psig but 20 or less your PRD can go up to 50 psig and 100 and all the way to 150 psig if you are next to empty. So really when you look at those numbers and you try to see or they make sense for delivery pathway and you realize that most of the time actually you – if the travel time is short enough which is going to be the case in the future when you have deployed H2 infrastructure. Or if your level of hydrogen in the trailer is low enough, then no venting is necessary.
And you can see an example here on this figure here that if you deliver 80 psi and your first delivery is 50 kilograms then your recommend [inaudible] is that you won’t need to vent a single gram of liquid hydrogen in order to go on the road. And so again for liquid hydrogen system that typical operate under typical condition you will not have to vent any hydrogen, which is a huge improvement over the current practice. And this is really here a good opportunity to optimize the amount of hydrogen that you need to use in order to make liquid hydrogen infrastructure happen.
So now this is the third section of this presentation. And here I’m showing some results that we collected in order to inform those boil off behaviors. So as some of you may know at Lawrence Livermore we have this liquid hydrogen facility that is, that includes fuel equipment. But most importantly we have this very high capacity dewar or high capacity for in the lab. For a station it’s maybe on the low side because it’s about 800 kilograms of hydrogen. But this is yet enough to make a lot of experiments. And we also have this liquid engine pump that was manufactured by Linde.
And with the system we can – we instrumented the system so that we can measure the boil off under a lot of different conditions. So when you start the boil off picture at the bottom of this – by the bottom I mean on the vent stack of this dewar and so that we can measure accurately what’s happening. And here on this slide I’m just showing the results from the bottom of meter in blue versus what we observe from just looking at the level of – so it's an inch of water, the level of liquid hydrogen in the dewar just from the [inaudible] gauge. And so, you see that so this [inaudible] gauge is recording only every hour as opposed to the boil off meter that records every second. So you have much more accuracy with the liquid with the boil off meter as compared to just the liquid gauge, right? So this is black. This is blue.
Again, this gives us a very accurate reading of what’s happening. And then when you use this boil off meter under typical condition you can really relate what’s happening for the entire system so the dewar and the pump. You can relate the boil off from the entire system with what’s happening with how you use your pump. So you can really see depending on how you use your pump you have [inaudible] boil off and you can relate that to—you can really correlate what is happening between those two things. So here I’m showing just the two examples of boil off.
So the boil off here is in black and we see that when it’s operating you have at the beginning a lot of boil off and then the boil off goes down over time. And when then the pump is inactive, this section here, so boil off slightly goes up and then it goes down again until the next cycle when you need to use it. But then you have a lot of boil off. And so again we are going to precisely measure what is happening at [inaudible].
What’s even more – what’s even more interesting here is that we can measure what’s happening during a transfer. So when we go from the trailer to dewar, when we have a delivery of liquid hydrogen to the dewar we can see what’s happening and – as long as we make sure that the venting is going through the boil off feature. We can measure what is happening during the fill. So and so – and this is really this effect is minimizing boil off during delivery is to look at top spray. So when you think about it you can use the two options to fill the dewar. You can do bottom fill or top fill. And back in the ‘90s NASA already identified this option as being a good way to minimize the boil off during transfer.
And on the left here I’m showing a figure that was taken from this paper by [inaudible] and they realized that when they do a fill from – when you do a top fill of the dewar at the beginning you have some flashing that happens because the system is not at equilibrium. But then you can see that the pressure in the receiving vessel would actually go down over time which means that the boil off is much reduced. It’s almost close to zero because what is happening is the vapor that is being sprayed, the liquid is being sprayed from the top is actually cooling down the vapor and actually collapsing the vapor and releasing the pressure and releasing the boil off.
What you have to make sure of is that you don’t – is at some point you don’t – you don’t overfill your vessel because otherwise you will have spike of pressure. And so, what we did was just try to reproduce this at our facility and this is a measurement that we did in February of this year. And in red you see the vapor pressure. And again, here we could by doing top fill only we could replicate the behaviors as observed. And we have initial spike here and then the pressure was going down which means that the boil off was going down as well. And so, in red is the pressure and in green this is the boil off as a function of, the as a function of time. And grey is cumulative boil off.
And in the end what we had, what happened is that we had – we could measure that for about 500-kilogram liquid hydrogen delivery only one kilogram of boil off could be measured, which is really much lower than what we’ve seen in the past doing bottom fill. And here is it yet again an opportunity to reduce your boil off and to really have a more efficient way of handling your liquid hydrogen transfer, liquid hydrogen transfer along the pathway.
So taking into account all of those measurements as well as the model we are now, we can now try to predict what will happen for the filling station design because we have some boil off from the pump, we have some boil off from the delivery and we have some boil off when the pump is idling. We can try to accumulate all of those boil off sources together, put them together and see what kind of boil off can we expect from the typical station size. Of course, this is, right now this is just simulation because there’s no station today that can do, that has many pumps and that can do that kind of throughput. So as an aside this is just a simulation and I have to make some assumption as to what would be the typical boil off from a system with multiple pumps for example.
But I want to say that those predictions are based on the measurements that we did at Lawrence Livermore with a pump from Linde and this pump starts to be a little old and maybe not really optimized for this kind of system. But yet I took those values as a baseline and it is likely in the future we can do much better. And actually, I had some conversation with Linde and they, a lot of room for improvement as far as boil off from the pump and how we could optimize pump idling boil off and even pump boil off when it’s running.
But nevertheless, here this is what we have and we have two different [inaudible], 350 bar or 700 bar refueling because indeed there is a big difference between those two conditions as far as how much boil off will the pump produce when it’s pumping high pressure. 700 bar this is mid pressure so we use 350 bar. So just something that I took here for this simulation is just to assume that we would need to have one pump on this dispenser and so using the [inaudible] and also assumption we can look at – we can then know how much, how many pumps or dispensers we need at certain station size. And you can see that if we have less than 500 kilograms per day we only need one pump. And when we go at 500 kilograms per day then we need a second pump and so forth until up to 10 pumps or 11 pumps for 500 kilograms per day.
And you can see here that as the station size increases the boil off decreases starting from up to 16 percent of boil off all the way to less than two percent boil off for larger stations. This is on those two figures here breaking down the comparative contribution of each factor of boil off. And you can see here that most of the boil off is from the pump idling and from the pump really and the boil off coming from the LH2 transfer is very small as compared to all the other contributors.
So the question then is what do we do because some boil off is going to be intrinsic. Once you optimize your boil off from the transfer if you do a top fill and everything, you really can minimize your boil off. But we still have some intrinsic boil off because when you use the pump you had some heat to the system and you would have to deal with that boil off. And here each of those boil offs they have a peak value and a mean value and of course you have to design this unfortunately for the peak value. So you need to find a way to capture whatever boil off you have and to capture it in an efficient way. So each of those modes of different peaks and up to three kilograms per hour peak of boil off needs to be captured from the pump.
And the question is then what do you do with this boil off? How do you recover the boil off? What can you do with it? Because if you use for example a compressor system to harvest this boil off then you will create a lot of hydrogen at the station and you could create 10 to 60 kilograms of boil off per day. And then you need to find a way to use that hydrogen because maybe it’s going to be low pressure hydrogen and you need to find a way to take, to send that hydrogen somewhat. Right?
So your second option is to use a cryo cooler that you can really collect the vapor which is a very neat way of reducing your footprint because you would just use the cryo cooler to cool down your vapor and collect the vapor so that you can reduce the boil off. But of course, this is a slightly more expensive option. You can also do a fuel cell and do net metering or provide local power to the surroundings. And the last option is to flare the hydrogen, to flare the boil off depending on the techno-economic viability and the safety aspect of course. There’s not too much information for flaring so I just put this option as an option, but I didn’t really explore this for this for this to work.
So then when you have – when you look at those options you can look at the cost of each option and you can try to see ok, what is the best trade off. What is the – how much electricity is going to be needed to capture the boil off and what does that compare between the cost of hydrogen and the cost of electricity and the initial capex that you need to spend for these solutions. And so here on this graph here I’m trying to show, to find a way to explore those tradeoffs and so each symbol is a different technology, electromechanical, mechanical or metalized compressor and then then triangle is cryo cooler. And you have four different lines for different combination of cost of hydrogen and electricity.
And basically, where you want to be of course is at the bottom here. You want to spend a very small amount of money to install your system and you want your system to use as little electricity as possible. And what it shows here is that it’s seen that one of those most interesting options from an techno-economic only perspective would be to use a mechanical compressor because that is really the best way to minimize your cost. But again, this is just from a techno-economic just looking at the capex only as an assumption. But it doesn’t take into account the cost of storing on site the hydrogen and then refilling this hydrogen.
Of course, today is you sell hydrogen in a bottle like 2,000 psi bottle you could find, you could make a good deal for that hydrogen. You could pay up to $50.00 or $60.00 per kilogram. And that could be a good way to get your money back. Right? But it really depends on the local market and everything. The demand for this kind of thing. So there’s other arguments that need to be taken into account. For example, if you really need to reduce a footprint maybe a cryo cooler may be the best option or it could also be the best way to reduce setback and everything. So if you look at just again techno-economic only just the cost of the capex then mechanical and electromechanical compressors may make the most sense. But other factors should be included to really make a better decision as to what is the best boil off recovery option.
And then yeah, when you examine those options you can reduce effective boil off from like 2.2 percent all the way down to less than one percent because you’re taking advantage or you’re valuing your boil off so in effect you’re capturing it. But you’re capturing your boil off but you’re paying extra cost for it. So you still have, you’re still going to have some effective cost that needs to be added to your typical cost. But you can reduce it to less than one percent if you have a good design and a good combination of hydrogen and electricity costs.
So we for this work we have been talking to really good people in the industry Linde and Praxair. And I really want to acknowledge this conversation with Martin, Willie, Kyle and Erik and also with Al Burgunder from Praxair. And so, the challenge now for beyond this work is that we, our thermodynamic models don’t stop the top spray, the top fill option and see something that is a limitation of our code right now because our code is only 0D and we need more if we want to capture the whole physics of the spray. And so, I think in the future we should look at safety codes to really capture this effect, this droplet interaction and everything. That needs to be properly done if we want to understand better spray options.
We publish already some work. We are publishing two papers in the International Journal of Hydrogen Energy and we publish a memo on the DOT regulations which is kind of what I alluded to concerning the CFR requirements. And we are distributing two open source codes. The first one is the code that I showed you before and the second one is one about how to evaluating the boil off [inaudible]. This code will be available soon in a matter of weeks online in the GitHub link.
So the concluding remark is that the boil off, on road boil off from the LH2 trailer is really negligible because a trailer can hold from 48 to 110 hours load without boiling any hydrogen. So that shouldn’t be a problem for a liquid hydrogen pathway. The second aspect that we need to follow the CFR requirement so that we can really reduce by a lot the venting that typical happens today at LH2 station. And it’s really a very important point here. So the real point is that we need to consider top fill as much as possible to reduce the boil off during transfer. This is limited by the requirement minimum pressure that you need to have for the compressor or the pump and that’s really where you need to pay attention because under certain scenarios you cannot use top fill all the way because then the pressure in the dewar will be too low which will reduce the efficiency or even will just stop the system if it goes beyond a certain point.
And we also need to make sure the station design will match our actual demand. I mean it’s easier said than done of course but this is really the best way to match all of these issues of idling and utilization. And if we can reduce this aspect by having a [inaudible] fleet and everything. This is really a good way to reduce your boil off. And again, here we, the boil offs really that we show here are for pumps that was done by Linde five to seven years ago and it seems that now they have much better results. And I encourage you to get in touch with them if you want to hear more about this.
And then the last point is that we can – the intrinsic boil off may be mitigated with a compressor or even a fuel cell or cooler. And the value of this option has to be looked at on a case by case basis to see if it makes sense on that. And there’s a lot of factors that come into play here. And yes, that’s it. We look at the benefits of hydrogen is really obvious for large-scale. We need to understand what’s happening from a boil off perspective.
And that’s why we develop this whole work putting together a model, collecting data at our site and trying to simulate all the date, all those things together to understand what does it mean for the entire pathway. And it seems that we reduce to less than two percent the boil off impact on the pathway and for this, for [inaudible] boil off is close to negligible as well. And I look forward to future work on it if needed. So thank you for your attention.
Eric Parker
Ok. I think with that we’ll move on to some Q&A. We have a couple of questions here we’ll get started with Neha.
Neha Rustagi
All right. So our first question is about comparing with liquid natural gas. So generally, in liquid natural gas the heat transfer rate increases as the amount of natural gas in the dewar decreases because the volume ratio goes up. And the question was is this the same conclusion for liquid hydrogen?
Guillaume Petitpas
That’s a really good question. I never thought about it that way. To be honest I’m going to have to think more about this to really answer this question because I don’t want to say something. I mean I’m sorry, but I don’t have the straight answer for this right away. I apologize on that, but I appreciate the comment.
Neha Rustagi
The next question was to confirm that liquid hydrogen tankers have a 48-hour hold time.
Guillaume Petitpas
So this is for the trailer. This is something – this is a marking that the manufacturer has to put on each trailer. So here’s what you see here on slide 14. I don’t know if you see that on the screen, but this is what they call the one-way travel time and so and basically here it says to go from 10 to 17. So if I look at the first line to go from 4 psig to 17 psig you would take 110 hours. And so, this whole time of course it depends on the initial condition and everything. But to be clear, the liquefication plant between four and six psigs so it would take 110 hours to go from this condition to 17 psig at those conditions.
So for dewar, a stationary dewar the whole time is again a function of the initial condition. And but typically what we say that the boil off at the maximum, when you’re at the maximum pressure is I think one percent, typically one percent per day of the total capacity so if you have 800 kilograms dewar you would do less than 8 kilograms per day, something like that.
Neha Rustagi
Ok. We had another question with respect to is one approach to managing liquid hydrogen boil off to agitate the liquid hydrogen before the tanker is used?
Guillaume Petitpas
So that’s a really interesting point as well. I’ve heard that many times that yeah, you shake the trailer. And this is kind of for drivers they’re like yeah, I’m not sure I want to do this because you in theory the trailer is rated for this kind of acceleration, but this is not something that they are really used to. What I can say is that this agitation argument is very varied though because I asked the drivers to let me know how the pressure in the trailer varies after they leave our station.
So basically, they would leave at 50 psig in the trailer and they would call me 20 minutes later and the pressure would be all the way to 10 psig already in 20 minutes. And just because just the vibration, just the trailer being on the road and just stopping at a stop sign or traffic light and accelerating and decelerating is enough agitation to really reduce the vapor pressure right away. So I guess agitation maybe is a strong word. It can be scary for a driver. But just turbulence or just typical vibration is typically enough to reduce your pressure while you’re on the road.
Neha Rustagi
Thank you Guillaume. One more question was if you can pull up slide 28. One of the audience members.
Guillaume Petitpas
Yes. I’m going there.
Neha Rustagi
So we can just maybe leave it on this slide for a little while. This is a good if anybody had questions on this. There were a couple of requests to show this slide again.
Guillaume Petitpas
Ok. 28, right? That one. The risks and challenges.
Neha Rustagi
Yes. Another question that came up was can you talk about ways of managing boil off in LNG? Was that something you explored?
Guillaume Petitpas
Not really. Not really. I know that LNG tankers, they use some of that boil off depending how they do it but they use some heat where they just burn the boil off for example or some tankers they use it for the engine or for APUs. But I’m more aware of big tankers, mitigation, boil off solution for big tankers. I don’t know about boil off solutions for trailers and LNG dewars, stationary dewars. And yeah, I apologize for that.
Neha Rustagi
All right. We can give it one more minute to see if anybody else has questions. Again, feel free to type it into your chat box. And in the meantime, Guillaume, maybe can you pull up the very last slide with the contact information just so folks know.
Guillaume Petitpas
Yeah.
Neha Rustagi
If they have more questions. All right. With that thank you very much for your presentation and we’ll get back to Eric.
Eric Parker
Yeah. Thanks everyone. Thanks Guillaume and Neha for the thoughtful and informative presentation and that’s going to conclude today’s webinar. So if we didn’t get to your question and you think of one later feel free to email either of them. And as I reminder we have been recording the webinar and we will be posting the full recording and the slides online so you’ll be able to access those soon. And I encourage everyone to sign up for the monthly newsletter that includes information on future webinars. So with that I will wish everyone a great rest of their week and goodbye.
[End of Audio]