Carbon Abatement Costs of Green Hydrogen Across End-Use Sectors
Joe Aldy: Welcome to the HKS Energy Policy Seminar. I’m Joe Aldy and I’ll serve as the host today. Thrilled to have everyone in the room as well as online joining us in this hybrid format. We’re thrilled to have with us today Roxana Shafiee for her talk, Carbon Abatement Costs of Green Hydrogen Across End-Use Sectors. She’s a Harvard environmental fellow through the Harvard University Center for the environment. She received her PhD in environmental research from the University of Oxford. Previously served as an environment, climate change and energy researcher at the Scottish Parliament following her dissertation research. And in 2022 was awarded the Stanback Postdoctoral Fellowship in Global Environmental Science at Caltech. Roxana, welcome to the Energy Policy Seminar. Roxana Shafiee: Thank you for having me. All good, the sound? Speaker 2: Yes. Roxana Shafiee: It’s great to be here. Great to round off a term of brilliant talks and a brilliant year of talks, actually. Today I’m going to be talking to you about some of the research that I’ve been undertaking with Professor Dan Schrag over at the Center for the Environment. This is one of our inaugural presentations of this research, so I’m very open to feedback, looking forward to hear what you have to say. There is significant optimism surrounding hydrogen. I’m sure a lot of you in this room are here because you’ve heard some sort of whisper about the potential that hydrogen has, either that or that it’s the last free lunch that you’re going to get of the year. This optimism is reflected in the investments into low carbon hydrogen across the world. There have been over 570 billion investments from the public and the private sector into hydrogen and 41 governments have invested in or have outlined some sort of hydrogen strategy or plan in order to implement hydrogen zed decarbonization goals. And since the passage of the Inflation Reduction Act in 2022, the US has really been leading the way in terms of financial incentives along with Europe for low carbon hydrogen. This is through the production tax credits programs where there’s the opportunity of up to $3 per kilogram for hydrogen production. The Congressional budget office estimates that this will probably amount to around 13.7 billion in investments into hydrogen as well as the Bipartisan Infrastructure Law, which earmarked 8 billion in grants for hydrogen hubs across the US. The focus of these investments and these policy targets are focused on low carbon hydrogen. There are two types of low carbon hydrogen. The first is green hydrogen, which is produced using renewable electricity with electrolysis to produce hydrogen. There’s also blue hydrogen which is produced using conventional fossil fuel methods, then capturing the CO₂ and storing it. For the purposes of this talk, I’m going to be focusing on green hydrogen because this is really where most of the focus is centered on and really is the gold standard when it comes to low carbon hydrogen. Some of you in this room might be questioning, "Well, I’ve heard about hydrogen hypes before," so I want to bring your attention to the fact that this article was published in spring 2004, so exactly 20 years ago, kind of outlining what’s going on in terms of the hydrogen world because we’re in a similar situation where there was a lot of enthusiasm and excitement. The difference now is that over the last 20 years, there’s been an unprecedented reduction in the cost of solar and wind energy. This has really put hydrogen back on the table because electricity is the main feedstock for hydrogen production. As this has come down, people have reconsidered hydrogen as a potential decarbonization strategy. There’s also another impetus for hydrogen. As we’ve increasingly electrified sectors of our society, it’s become increasingly evident that we need something to abate the 30% of emissions from the hard to abate or hard to decarbonize sectors. These are sectors such as steel-making, cement, backup power, heavy duty freight and aviation where electrification is currently prohibitively expensive or it’s technically challenging to electrify these sectors. But there’s a problem, as always. Hydrogen, even with these reductions in the cost of solar and wind is expensive to produce. So estimates, now bear in mind, there’s not currently a hydrogen economy. These are just estimates. Not a single ton of green hydrogen has been sold to date. Estimates put the cost of hydrogen around $5-$7 per kilogram in comparison with fossil-based hydrogen, which costs around $1-$2 per kilogram to produce. Speaker 3: This is without all the incentive? Roxana Shafiee: This is without all the incentive. That’s correct, yes. But many reports are optimistic that these production costs will fall rapidly. I’m showing you here two reports, one from Bloomberg and one from McKinsey that showed that by 2050 the cost of hydrogen production will have fallen to less than $2 per kilogram and then often in certain parts of the world could even be less than $1 per kilogram, the production. Why do they think that these costs will come down rapidly? There are a few different reasons. One of the reasons that it’s predicted is that electrolysers, the central technology for producing green hydrogen, will rapidly fall in cost. The second is that electrolysers will become more efficient. These two plots are showing you predictions of how the cost of electrolysers and the efficiency of electrolysers will become more efficient as time goes on. There’s also another prediction as well that solar and wind, as this plot shows you, is also predicted to come down increasingly in cost. This is potentially contentious but not something I want to dwell on right now. These three factors together are thought they’re going to bring down the cost of hydrogen with time. And the result of this will be extremely cheap abatement opportunities where hydrogen can be used across all different sectors of the economy, providing abatement opportunities at less than $200 per ton. Anywhere from use in cars to steel to shipping and space and water heating. Many reports are significantly optimistic that this will be possible as hydrogen comes down in cost. But among these reports and studies and analyses that are optimistic about these cost reductions, a number of fatal errors have started to gain prominence, which really call into question whether hydrogen will provide this widespread abatement opportunity. The first one is the fact that there’s an underestimation of current production costs by overestimating electrolyser utilization rates. Don’t worry, I’m going to break this down for you. In order to understand this error or this fallacy, you need to understand a bit about how hydrogen is made. There’s two scenarios in which green hydrogen can be made. The first one on the left is connecting your electrolyser, which is here, directly to your renewables, operating your electrolyser at the capacity factor of your renewables. Say around 20 to 40%. Then, thereby, having access to cheap renewable electricity. If it’s untenable to have such low utilization rates, you might instead choose to connect to the grid where you can run your electrolyser a hundred percent of the time. But this comes at a cost premium, so you have to pay higher electricity rates if you’re connecting through the grid. When we look at the cost components of these two different scenarios, on your left you have your off-grid scenario, which is mainly dominated by your capital costs because you’re running your electrolyser at low rates. And on the right-hand side you have a scenario in which you’re running your electrolyser at high rates, and therefore electricity costs dominate the bulk of your cost. But a third set of studies have emerged in which they combine high utilization rates of electrolysers with low-cost electricity, and this is, at the moment, just not possible. And what this does is lead to really low-cost estimates, around $2 per kilogram, that have combined essentially the best of both worlds. But we know from the way that renewables work is that this is just not possible. There might be niche scenarios where you can combine wind and solar to achieve high electrolyser utilization rates, but in terms of widespread opportunities, this will just not be possible. Secondly is exclusion of appropriate storage and distribution costs. Many of the costs I’ve showed you up to now are purely production costs, but we know that a green hydrogen supply chain will not only require production, it will require storage, distribution, and any infrastructure related to its end use. For instance, that might be a refueling station. To give you an example of how significant these costs can be, you can buy grey hydrogens or fossil based hydrogen at the pump In California. It costs $1-$2 per kilogram to produce, as I showed you previously. At the pump, it costs $15-$16 per kilogram of hydrogen. Now just to put this into context, one kilogram of hydrogen is equal to one gallon of diesel fuel in terms of energy equivalent. There are folks who are paying $15-$16 per gallon of hydrogen in California. These errors are also seen in the DOE’s Liftoff to Hydrogen report in which they’ve outlined that they have included storage and distribution costs as part of their modeling of the supply chain. This is their report with this very convoluted figure you’re looking at, and the different cost components. And as you see, they’ve also thought about the end use or end price of hydrogen. But again, there’s a problem. This report uses low values of $0.10 per kilogram for storage and $0.10 per kilogram for distribution. However, we need to understand a bit about what goes into the cost of hydrogen storage and distribution to understand why this is a mistake. The levelized cost of storage is the cost of storage that you would pay on top of the production costs and it’s formed of two components. The first component is your project costs: your capital costs, your operating costs, and anything cost associated with decommissioning. But this is a really important component that I want to bring your attention to. This is the component that’s the lifetime amount of hydrogen stored and you can see that that’s a function of the lifetime of your project, but also the rate at which you cycle hydrogen in and out of your stores. The same also goes for distribution. The cost of your distribution depends on the rate at which you’re distributing your hydrogen. We know that different sectors will have different demands for storage and distribution. I’m showing you here the natural gas consumption of the power sector versus the industrial sector. And you can see that there are large seasonal swings in the power sector, natural gas consumption. These are met with stored hydrogen, hydrogen that’s withdrawn at a low rate from stores, whereas the industrial sector, while it does have some changes, will unlikely need such significant stores of hydrogen. The same goes for distribution. Different end uses will have a different rate at which they’ll require hydrogen to be distributed to them. For instance, a chemical plant or a power plant will require a lot more hydrogen to be distributed to it for its final end use compared to a refueling station which won’t need as much. These different components as I showed you, I’ll just quickly go back, play into the levelized cost of distribution. Hopefully I’ve convinced you here that using a fixed value for the cost of distribution storage is just not the reality between different end sectors. Again, this lack of emphasis on storage and distribution is reflected in the investments. Around 75% of investments to date have focused on the production side of things, but very much less on the storage, distribution, end use, despite it playing such an integral role in supply of hydrogen. Finally, the final fallacy that’s gained prominence is the comparison of production costs with fossil fuel prices. Many trusted reports compare the production costs, and again you can see this projected decline in production costs with the cost of fossil fuels. Now as I’ve just shown you, the cost of hydrogen will depend also on storage and distribution. This is an erroneous comparison, comparing production costs with the cost of fossil fuels, which do include these storage and distribution costs. It’s not a like for like comparison that has gained prominence in these reports. So now considering all those areas, what did we seek to do in our study? In our study, we wanted to ask, if we take another look at these errors and correct for them, what is actually the delivered price of green hydrogen by end use sector? Then based on those delivered prices by end use, what is the carbon abatement cost of green hydrogen? So what is the cost per in tons of dollars of CO₂ to abate using hydrogen as an abatement strategy? The approach we took was to look at different and start with different end uses. We looked across three main categories where hydrogen is currently used or where hydrogen has been proposed for use. The first one being industrial uses, petrochemical refining and ammonia synthesis where hydrogen is already currently used, as well as for new usage as industrial heat to replace natural gas. We also looked at the power sector where there have been proposals, for instance by the EPA, to co blend hydrogen in with natural gas. And we looked at the freight sector, the heavy duty road freight. In these sectors, hydrogen would be abating. Green hydrogen would be abating grey hydrogen produced using fossil fuels, natural gas for industrial heat, natural gas within the power sector, and then of course for the freight sector, diesel. The reason that we chose these sectors, of course there are other sectors in which hydrogen has been proposed such as aviation, these are the closest to being, in terms of the technology readiness, to actually being implemented. In order to come to an estimate, a delivered price by entry sector, we’re constrained by the lack of empirical data. We needed to produce estimates for actually how much will hydrogen cost delivered to each of these different end uses. In order to do this, we performed an analysis of what is the storage and distribution requirements of these different end uses, specifically looking at the storage cycling rate and the distribution cycling rate. Distribution, sorry, rate. Now recall that the levelized cost of storage and distribution depends on these two factors which will vary by end use. This is why we took this approach to outlining costs. Our second part of our analysis in order to pin a cost for these different components was to compile storage and distribution costs from published cost values. And as you can see, as you increase your utilization of storage, here we have an increasing cycling rate so you have, for instance for your storage, an annual cycling rate where you’re only annually taking out the storage versus going all the way to daily. The same for distribution rate, increasing the rate at which you’re distributing hydrogen. As you would expect, as you increase the intensity of the utilization of your storage or your distribution capital, your costs come down. You can see that costs decrease with increasing utilization rate. That’s what we found when we compiled costs that have been published in different scenarios. Then the second part was just to bring your attention to the fact that the cost estimate that I showed you earlier for the DOE putting the costs of storage and distribution at $0.10 per kilogram is the very, very lowest cost estimate that you get when you look at actual storage and distribution costs based on extremely high cycling rates of storage and distribution infrastructure. The values that are being used now in these reports are values that are contingent on the most highly utilized storage and distribution capital. But this is just not going to be the scenario, the way in which storage is used. For instance, in contrast, the power sector we would expect to use, as I showed you earlier with the natural gas consumption, we would expect the power sector to use storage, for instance, seasonally or every few months. What that means is that they’re paying in the region of around $2-$3 per kilogram as opposed to $0.10 per kilogram of hydrogen. Again, for the distribution, because the power sector, we would expect to have a very high demand for hydrogen if it was going to use it in place of natural gas. The distribution costs are the lowest that they could be because of this high demand, whereas for instance if you are a refueling station, your demand is going to be a lot less on a per site basis, so the distribution costs are a lot more. This is how we gauge the cost structure of different components for different end uses, was looking at the storage and distribution cycling rate. So now what I’m going to show you are the delivered prices that we estimated for those different sectors that I showed you previously. Our analysis assumes that the delivered costs are production agnostics. They’re all getting their production and they’re all getting their hydrogen from the same supply, so production costs are the same. But we start to see differences when we look at the storage costs. Again, it’s really only the power sector, which we would expect to have significantly higher storage costs because of this low turnover of stores that are needed in order to mitigate those seasonal changes in demand as I showed you previously. Then because each of these sectors has a different demand per site for hydrogen, it also has a different cost when it comes to distribution. Then finally, there’s one thing that’s unique about the freight sector and that’s the need for refueling stations, which adds an additional cost to this sector. You can see that the costs that we estimate range anywhere from $7 up to $15. And you can see that there’s differences based on which the end use that hydrogen is used for. So now based on these delivered costs, we calculated the carbon abatement cost of using that hydrogen in place of incumbent fossil fuel systems. So natural gas, diesel or grey hydrogen. This is what we found. We found that the carbon abatement costs are typically over $500 per ton of CO₂ to use hydrogen across these, all different end uses, with the lowest costs in the sectors that already use hydrogen. That being ammonia and refining. And to put into context as to how high these carbon abatement costs are, if you were to translate this into an additional premium on your diesel, for instance, this would add an $8.39 per gallon of diesel sold. We also modeled the costs if the production costs come down in line with what would be expected, so $2 per kilogram, and we see that the abatement opportunities remain limited. They’re really only limited to ammonia, the ammonia sector where hydrogen is already used. In all other sectors, hydrogen remains a prohibitively expensive abatement strategy. This is really problematic because many reports and sectors are really banking on hydrogen to be the strategy for decarbonization. I’ve whizzed through this, which means that there’ll be plenty of time for questions and we can discuss, but the results of our work show that green hydrogen is a prohibitively expensive carbon abatement strategy and that current fossil fuel prices, the opportunities for abatement are over $600 per ton across a range of end sectors in the United States. And even with reductions in production costs in line with what’s predicted, the opportunities will remain highly limited, really only to ammonia production and potentially heavy duty freight. But there really needs to be significant reductions in the cost of storage and distribution in order for hydrogen to be a cost-effective abatement strategy. Based on these findings, we recommend that governments should really retain a broader strategy of decarbonization approaches. And while there may be niche opportunities and scenarios in which hydrogen can be low cost, for instance if you were to produce and consume your hydrogen on site and you are in close proximity to salt cabin storage, then there might be opportunities in which hydrogen is cost effective. But this is not going to provide a wide and a broad decarbonization pathway for the entirety of these hard to abate sectors. Really there needs to be emphasis on looking at different approaches to decarbonizing these sectors, things like biofuels or advanced electrification as well as continued investment into hydrogen, as well. Further support is needed for research into storage and distribution technologies. As we’ve shown you, storage and distribution are considerable cost components of the delivered cost of hydrogen. In order to bring down the abatement costs in line with, for instance, the social cost of carbon, there really needs to be additional support and investment into storage and distribution such that these costs also come down. I’ll leave it there and I’m happy to receive any of your questions. Thank you. Joe Aldy: Thank you. Before I go to the audience for questions, I guess I’ve got a couple of questions, but I’ll just start with one now. $600 a ton sounds expensive. It’s certainly more expensive than what we are willing to pay, at least in terms of policy in the US right now. But having said that, if I look at modeling scenarios that say we’re going to get to net zero by 2050, by the time we’re getting those last few tons out of the system, they’re expected to have quite high marginal abatement cost. This looks really expensive compared to putting in a little bit more solar into my power system. But it might be it doesn’t look quite so crazy when I think about the modeling analyses getting out to 2050 and getting to net zero. Help us think through a little bit, especially when you talk about some of these alternatives that look like they have potential. Is it because you really see them as those modeling scenarios that say we’re going to be north of $1,000 a ton, either real or shadow price of carbon are just too high and we think there may be other opportunities here? Or is it simply yes, there’s a role for this stuff, it’s really expensive, and that role is in 2030 and 2040 and beyond and we shouldn’t be subsidizing this deployment today? How should I interpret $600 a ton in a world in which I think we’re going to need really expensive abatement if we meet our net zero goals? Roxana Shafiee: Yes, let me think about that. I think it demonstrates it doesn’t have this broad de-carbonization strategy and it will be preserved [inaudible 00:25:30]. But at the moment that’s not the narrative. At the moment, for instance in the UK where I’m from, there have been proposals of co-blending hydrogen into natural gas pipelines, right? And we have other alternatives. We have heat pumps or efficiency measures. The thing there is really taking another critical look of where is hydrogen going to be necessary in terms of there is no alternative that we have to tolerate a high carbon cost, and where are there alternatives that could potentially be lower cost or needs or should garner more research development attention? [inaudible 00:26:10] William Hogan: This is a terrific and important work and I find it very interesting. I have one thing which is my pet concern about these kinds of discussions, is when you’re talking about grid connected hydrogen and improved electrolysis, we have divined, design, particularly in this country, our electricity markets so that they’re highly efficient. And they do all kinds of complicated substitutions across power plants over very large geographic areas. And one of the consequences of that is it’s extremely difficult to identify how much carbon emissions you are emitting from your plant if you don’t do the full blown analysis of the whole system. Those analyses of the whole system show that the range of how much when you consume, use more, reduce demand or increase renewable some place, what happens to emissions in the system? The answer is a factor of three to five difference depending on where you are and the time of day. It could be yet another factor on top of all of this in terms of you’re actually not reducing carbon emissions as much as you think unless you’re awfully smart about where you locate these things and when you operate them. It may mean not operating them in lots of periods of times because they’re actually not reducing carbon emissions, so it becomes kind of a scaling thing. This is a very hard problem. The only solution I know to this problem is the carbon tax because that handles all the same problems simultaneously. I think it’s something you should look into here because it’s actually a big deal. Roxana Shafiee: Yep. No, thank you. I think there’s been analyses that showed that if hydrogen is produced using grid electricity, but not in the ideal scenario, that the emissions can actually be higher than fossil based hydrogen for that reason. This is why the IRS is proposing to have such tough limits on the use of grid connected electricity in terms of it being time matched and backed up with renewable energy certificates. And you can go into what you think about that, but to prevent the problem of actually- William Hogan: It completely ignores it. Roxana Shafiee: Yup. William Hogan: [inaudible 00:28:42] they don’t look at the locational differences. Roxana Shafiee: Actually, of those requirements, it would need to be time matched hourly. It needs to be in the same location. It needs to be built in the last, I think, three months, but- William Hogan: Actually, when you do it, that’s when you say, "Then at this hour how much emissions are you losing?" very complicated question because of the whole history. My metric doesn’t solve- Roxana Shafiee: No, it doesn’t. William Hogan: [inaudible 00:29:10]. Roxana Shafiee: Yep. Thank you. Speaker 5: Yes, a excellent presentation. I followed very well until you got to the part about abatement cost. Could you explain what abatement cost you’re referring to and how you calculate it, or at least broadly speaking, how you calculate it? Roxana Shafiee: Sure, yeah. Speaker 5: And also is this paper or work available online? Roxana Shafiee: It’s not available online. We’re currently preparing it for resubmission. We calculated a marginal abatement cost, but put very simply, it’s the additional cost associated with the intervention. Here we’re just looking at the differences in the fuel prices. We’re assuming that there’s no major retrofits that are needed, but these would actually increase the carbon abatement costs. Then we’re looking at the reduction in CO₂ with that change and we’re assuming that the hydrogen is completely clean, so we’re assuming the best case scenario. These are almost lower bounds on estimates because if of course you look at your total lifecycle emissions of hydrogen, then your abatement costs are going to increase. And if you consider any retrofits associated with the abatement, then your abatement costs- Speaker 5: So those costs go on top of the production cost and the operational cost? Roxana Shafiee: Yes. Here we’re just looking at the difference in cost between the current incumbent fuel and hydrogen and then the emissions saving arising from that. But we’re not looking at the change in interventions. I know some methods, you can look at the first year versus the second and how it changes over time. This is a very straightforward- Speaker 5: Is there assumption on the carbon price there? Roxana Shafiee: No carbon price included in that. No. Does that answer your question? Speaker 5: Yeah. Roxana Shafiee: I’m happy to discuss it offline afterwards as well. We can go through the- Speaker 5: Certainly. Thank you. Roxana Shafiee: No worries. Thank you. Matt Floyd: Thank you. On the question of distribution, there’s also a question of distance which would impact the cost. I’m curious what assumption you made on distance. Some similar research that I’ve seen is focused on the concept of hubs and if you’re co-locating everything very close, you abrogate the need for the expensive distribution, so I’m wondering what… And also, are you looking at pipeline, are you looking at tube and trailer? What kind of distribution are you looking at? Roxana Shafiee: We’re looking at the two forms of distribution, which are the most mature at the moment, so pipeline and truck distribution. And the different distances, because this is going to vary on a site by site basis, our focus was mainly looking at what the difference is between end uses are as opposed to individual sites, but that’s actually encompassed within the error of these costs. Actually, demand is a bigger driver in terms of distribution costs and distance based on our analysis. Matt Floyd: Well, demand would impact the size of the pipe, which would impact the costs, but distance is also a function, right? Roxana Shafiee: Yes- Matt Floyd: It would definitely be expensive. Roxana Shafiee: … not a first order in the same way. If we go back to our calculation of the levelized cost of distribution, it’s the lifetime amount of hydrogen distributed. Yes, distance would play into that, but it’s mainly about the pipeline size and the volume you could actually distribute. Thank you. Speaker 6: Yeah, thank you for this. This is a good reminder to include the whole ecosystem when looking at a thing. One thing that I found interesting is this utilization rate slide that you showed. Maybe we could go to that. There was the CapEx and the OpEx graph. I think- Roxana Shafiee: You want the production costs? Speaker 6: Yeah, exactly. Exactly. Just on the production costs. I’d be curious. If we look at the off-grid production, which is basically I think something that you and others are pushing for, if we look at this larger CapEx component and assume then the numbers from the paper of Gunther Glenk that you referenced earlier, wouldn’t that have a higher potential to drive down the cost quite a bit? Because I think they have these 2030 projections that it would already quite decrease the CapEx. Roxana Shafiee: Yeah. This is why there’s a focus on decreasing those capital costs, is because of the low utilization rates. You really need a reduction in the cost of your electrolyser in order to be able to bring down costs. I’m not quite sure what your… Speaker 6: I don’t know, like this $4.55 is quite different from what Gunther mentioned in his paper. Is this now and his numbers are 2030? Roxana Shafiee: Yes. So that paper is referring to 2050 costs, but there are some problems with that paper too in that there’s a bit of an overestimation on learning curves in terms of bringing down the cost of electrolysers. Now learning curves typically apply to manufacturing, whereas the actual component, if you look at the cost components of an electrolyser, it’s only around 10%, which is actually manufacturing. Most of it is the critical elements of the stack which are needed. You need more innovation to bring down those costs. And in his study, he draws analogies with solar and wind coming down at that rate. And of course, as I’ve mentioned throughout this talks, production’s only one cost component. So even if your electrolyser comes down to zero, you’ve still got those other cost components in- Speaker 6: Yeah. Okay, cool. That’s helpful. And another question, maybe just one more. When you’re calculating the abatement cost, you’re comparing, for example, in steel to the gas prices. The natural gas prices in the US are quite low compared to Europe or Germany, in a sense. Have you run the analysis for coal-fired steel plants or coal-fired power plants? Roxana Shafiee: No. We’re predominantly doing it based on natural gas. And you are right in that if you were to do these analysis on European prices, the abatement cost would be lower because of the higher current cost of fossil fuels there. But because of the high delivered costs, they’re still untenably high above $500 per kilogram. But lower, you are right. Speaker 8: Hey, thank you for a great presentation. You mentioned California and the light duty market for hydrogen. Can you expand of that and if there’s any policy lessons for early adoption of hydrogen in light duty transportation? Roxana Shafiee: Yeah, so along with some of the safety issues that they’ve had, all I can really say about that is that it’s entirely gray hydrogen, so it’s not actually low carbon hydrogen. And even with production costs being as low as they are, $1-$2 per kilogram… Where did I show this? It’s still, yeah, the delivered price is $15-$16 per kilogram. The rest of that is all storage and distribution and the refueling station. We kind of have an indication from that, that these costs are going to be high as well for green hydrogen. Even higher because the production costs are higher too. If we can’t get it lower with gray hydrogen, then the chances for green are very low. Thank you. Speaker 9: Thank you very much for this wonderful presentation. At one point, you’re referring to the study from IRENA and where you’re showing how the scaling up of the electrolyser will drive down the cost. Your study is based on the American market here? What are you exactly comparing this with and which market did they consider for this study and what are you… I just wanted to be clear why you’re comparing your study with this one. Roxana Shafiee: The reason that I brought up this figure was to highlight the mistake that has been made in terms of comparing. Now the colored lines here show predictions of production costs, so we can say that they’re wrong or right, whatever. But really the erroneous thing here is that they’ve compared it with the cost of fossil fuels, and the cost of fossil fuels includes things like storage and distribution, whereas these production costs, like I’ve shown, are only one component of the entire supply chain. It’s comparing apples with oranges. It’s an incorrect comparison. This kind of makes it seem like hydrogen is closer to being cost competitive than it actually is by making this comparison. Does that answer your question? Speaker 9: Yeah, thank you. If I can add one more? I come from Nepal, just to gather your perspective. Actually, we have all hydro generation, our grid is all green. Bhutan has a similar all hydro-based generation, and there’s a country not that far, Laos, but we all… There is a potential for exporting. For example, Nepal sits next to India, Bhutan sits next to India, Laos exports majority of its production to Thailand. How are these kind of countries, where the generation is entirely the green, like hydro-based generations, are these countries… Or is it too early for these countries to see any kind of… To capitalize on this green hydrogen production where the input cost, the production cost would be much later? Roxana Shafiee: So even in theory, whether the feedstock, the electricity price could be low because of ample solar and wind, I’m not an expert on this, but there are other barriers to implementing hydrogen projects in those countries. Without seeing the specifics, I can’t say anything. But like you mentioned, if there are export opportunities, that comes at a cost. You’ve got to ship your hydrogen, you’ve got to build pipelines to move it, and then you have your last mile cost as well. Once it gets to the destination country, then you have to distribute it. It has to be stored somewhere. These costs all still apply regardless of whether you are making your hydrogen somewhere and exporting it. Speaker 9: Thank you. Joe Aldy: I’d like to get to the other question I had in mind, which was alluded to briefly when we talked about hubs and trying to think about geographically concentrating the users of hydrogen with where we’re going to be producing it. We now have eight or ten hubs identified through the Bipartisan Infrastructure Law that Department of Energy is putting money to help facilitate the build out of these hubs. Is one takeaway from this, when we sort of look at here’s the subsidy and the tax code, that despite all this sort of talk about all the proposed hydrogen facilities, should we expect to really only see them located in these hubs where we might think they have very low distribution costs because they’re right there at the potential users? And maybe depending on the nature of where they’re located, they might have relatively low storage costs? Some of these are in places where we’re already storing a lot of natural gas. There may be ways of repurposing some of those facilities. When you look out and think through the hype with this analysis, should we expect to just see this where we’re subsidizing the co-location or do you think there’s other places we might see some of the hydrogen come online? Roxana Shafiee: Yeah, that’s a good question. I think the challenge really with the hydrogen hub is going to be trying to get over the mismatch between where solar and wind might be ideal to achieve those really high utilization rates, but also in regions in which there’s demand for hydrogen. This has been one of the key challenges in the US has been there isn’t the similar scale of off-take and demand as there has been focus on production. You have this challenge of you are co-locating your wind and most of the projects now take a production centric view in that sites are chosen based on solar and wind. Actually, that might not match up with where the best salt cavern storage is or where the end use even is. There are lots of challenges in that. I don’t think I directly answered your question, but it’s a challenge trying to bring all those different factors together in one place. Joe Aldy: Do we think cost can meaningfully come down on distribution and storage? Roxana Shafiee: There’s no- Joe Aldy: For me, it strikes me that when we’re talking about the electrolysers, we’re talking about sort of a space that is I think relatively new, although there may be some people in here who would correct me on that. Whereas I feel like on storage and distribution, we’re talking about technologies that have been around for a very long time. Do we think there’s much potential for driving down those costs? Roxana Shafiee: Not in the technologies that already exist that are mature, but there is some potential, for instance, in metal hydrides which store hydrogen at high concentrations, for instance. But current projections don’t predict that the cost of storage and distribution will come down, which suggests that there’s really… Focus is needed on the innovation side of things. Joe Aldy: Any other questions from the audience? Yes, please. Matt Floyd: Shifting away from the methodology, more towards the recommendations at the end, I’m wondering broadly how useful it is to think of the cost of abatement relative to natural gas, as opposed to relative to other abatement solutions that you proposed and talked about? Because part of the reason that hydrogen has caught on, to your point earlier, is that if these are hard to abate industries where electrification is also prohibitively expensive or just not technologically feasible. So is it better to look at the relative cost of hydrogen abatement versus some of these other ones instead of the natural gas? What was the thinking there? Roxana Shafiee: Yeah, that’s a good way to contextualize it. I’ll give you this for context. The direct air capture is predicted to be around $600-$1,000. In some of these cases, hydrogen is, it’s cheaper to suck CO₂ out the atmosphere than it is to use hydrogen. Now there may be other strategies that are more sector specific where the abatement costs may be lower, and those should obviously be prioritized, but in terms of other strategies, hydrogen exceeds them. Henry Lee: One comment and one question. The comment is, it probably is worth doing a run of when you have this sort of dedicated solar and wind and you had a utilization rate of around 40%, and then you had one where you were buying it off the grid for a much higher price and you had a utilization of 100%. It seems to me that you might want to run what I call the Norway run, where you’re looking at a 100% of a dedicated hydro dam, for example. That’s a very, very isolated, locational specific, but it might answer some of the questions that other people have raised here. I guess as I look at all of this is that if you were looking at hydrogen down the road, is there… And you look at the IRA and what it has in it, I recognize that the carbon price still stays up. How much is the IRA going to do here or is this still going to be prohibitively for the individual investor? I’m not talking about the society. The government’s paying now part of the carbon fee. Roxana Shafiee: Yeah, that’s a different question because then it remains to be seen how actually close does the $3 per kilogram get the investors into economic viability and on a project by project basis? I don’t know the answer to that. That’s a different challenge. This is kind of more from the policy side of things, but yeah. Henry Lee: Okay. Roxana Shafiee: That’s a good point. James Stock: It’s very risky asking a question when I just walked in late, because you probably answered this already. And if you did, just tell me so, and if you didn’t, I’d be interested. An alternative approach instead of using hydrogen directly as a fuel is to use it as a feedstock for power to fuels. If you, say, combine this with CO₂ from direct air capture and maybe get those costs down, which is not where this is about, and then you go through some appropriate process and create a hydrocarbon change from that, then you can create SAFs and so forth. Now the thing about that is there’s not really transport necessarily involved. That all can be done on a fairly localized, on site facility. How do you think about that in terms of the costs? Roxana Shafiee: That’s not something we focused on just because it’s kind of further out in terms of technology readiness. But there has been a study come out of MIT which has looked at using hydrogen to make e-fuel, SAF, and their abatement costs are in the thousands. It’s still prohibitively expensive at this stage. And because of the early technology readiness level, we also just don’t know how robust the assumptions that go into those calculations are. That really remains to be seen too. It’s challenging, but people are thinking about this, yeah. Joe Aldy: One more question. One of the reasons why we do research is we want to inform people who make decisions in the real world. You talked some about the analyses that DOE has, they have these ambitious goals. Have you had a chance to share some of these insights with DOE? They may not necessarily be enthusiastic when you say, "We come to different conclusions," but I’m curious what kind of appetite you’ve seen out there in the policy community or among stakeholders when you’ve talked about this work. Roxana Shafiee: There’s increasing recognization that these are important costs that need to be considered. I think the uncertainty at the moment is what the actual costs are. There’s quite a range and there’s just uncertainty in terms of what the parameters in which the end-users are actually going to use this infrastructure and how it’s going to group. Is it going to be a cluster that’s served by a pipeline or individual sites? There’s increasing awareness that these need to be looked at seriously. But yeah, one needs to be done there. Joe Aldy: Great. Yes. Speaker 5: You researched this very well, so I don’t want to say you’re wrong, but Japan, as you no doubt know, has very aggressive policies in place. In fact, in my opinion, they’re more favorable to developing the supply chain infrastructure than anywhere in the world. But that’s just my opinion. They have a policy around the comparable price difference, the difference between what it would cost in the market for a green-produced hydrogen, and I assume it would include the infrastructure as well, the storage, distribution, versus comparable fossil fuel-generated hydrogen or natural gas of comparable power density. Have you looked at that and what do you think? What’s your opinion on the- Roxana Shafiee: The EU has that too, it’s called Contracts for Difference, and they’re primarily focused on industrial uses that already use hydrogen, petrochemical refining, ammonia synthesis. But those really aren’t the sectors in which they’re going to face the highest storage and distribution costs anyway, because there are a lot of proposals where hydrogen may be produced and consumed on site. It’s really the other sectors that storage and distribution infrastructure is more a significant component, but we haven’t looked at those policy side of things. We’ve been focused on- Speaker 5: So in Europe, you’re saying it applies to which sectors? Roxana Shafiee: It applies to, I believe petrochemical refining, ammonia synthesis, and I think there’s some indication that it might be used in steel making as well, as a replacement in natural gas in a blast furnace. Speaker 5: Has Japan’s policy come up in your work? Roxana Shafiee: No, but I’m aware that Japan is co-blending hydrogen, may even be with coal in power plants, but I haven’t looked into it specifically, no. Speaker 5: So it might be worth you looking to the same policy cost-price difference or CPCDF. My understanding, it’s across any industry, any hydrogen used in any industry, and it’s gathered a lot of interest among investors. Roxana Shafiee: Mm-hm. Thank you. Joe Aldy: Okay, so before we wrap up, I would like to thank Liz Hanlon, Paul Sherman for all of their work, making everything go incredibly smoothly over the course of the year in the seminar series. Thanks to both of you. Yes, we can applaud them. Thank you. I’d like to thank the Harvard University Center for the Environment, the Environment and Natural Resources Program, the Mossavar-Rahmani Center for Business and Government, and the Salata Institute for Climate and Sustainability for supporting our seminar series over the course of this year. I would like to thank my colleagues for suggestions over the course of the year, and especially Henry Lee for help in crafting this seminar series. Finally, I want to thank all of you for enjoying what is for you free food, but I hope also stimulating discussions of important energy and climate questions over the course of this year. This is our last seminar of the year. If you have suggestions for speakers for next year, shoot me an email. Over the summer we’ll start scheduling for the fall semester. Finally, please join me in thanking Roxana Shafiee for our presentation today. Thank you, Roxana. That was great.
An Energy Policy Seminar featuring Roxana Shafiee, Environmental Fellow in the Harvard University Center for the Environment. Shafiee gave a talk on “Carbon Abatement Costs of Green Hydrogen Across End-Use Sectors.”
For more information: https://www.belfercenter.org/event/energy-policy-seminar-carbon-abatement-costs-green-hydrogen-across-end-use-sectors