Eric Parker, Hydrogen & Fuel Cell Technologies Office:
Hello, and welcome to a special H2IQ Hour presentation. Today we have my colleague, Elizabeth Connelly, giving us a presentation on "Resource Assessment for Hydrogen Production." Welcome, Elizabeth.
Elizabeth Connelly:
Thanks, Eric. Great. Today I'll be presenting on some work that I and a team of researchers at the National Renewable Energy Laboratory completed assessing the potential for hydrogen production from domestic resources in the United States.
I'll start by going over some background and the objectives of this assessment. Then I'll review the assumptions and data sources. The results focus on three topics, mainly the technical potential of resources to produce hydrogen, as well as electricity requirements compared to the size of the grid, and water consumption impacts. Lastly, I'll mention directions for future work.
So, in 2016 the Department of Energy launched the H2@Scale initiative where H2@Scale is a concept that explores the potential for wide scale hydrogen production and utilization in the United States to enable affordable, reliable, clean, and secure energy across a number of sectors. As depicted in this graphic, hydrogen can be produced from a variety of resources, including fossil, nuclear, and renewable energy. At this time the US produces approximately 10 million metric tons of hydrogen, almost exclusively from natural gas, for purposes including oil refining, ammonia, and chemicals production. The resource assessment presented here is part of a multi-laboratory analysis effort addressing the potential for increased hydrogen production and utilization across more diverse end uses.
In addition to this resource assessment, the analysis effort is exploring US hydrogen demand potential across sectors that include transportation, chemical and industrial applications, as well as power generation and energy storage. The present work addresses the technical potential to produce hydrogen, though there is ongoing work analyzing the economic potential for hydrogen production and utilization across different scenarios of market and technology advancement assumptions.
The purpose of this assessment is to estimate the hydrogen production potential from a variety of energy resources. So, the goal is really to answer the question: Are there sufficient domestic energy resources to meet hydrogen demand potential, which is being estimated through other H2@Scale analyses? This assessment also looks into the quantity of energy resources required to produce an additional ten million metric tons of hydrogen by 2040, which would represent a doubling of the current market size. These results for 10 million metric tons are intentionally easily scalable to understand energy requirements for various scenarios of hydrogen demand across end use applications. We're also looking at how producing ten million metric tons of hydrogen would impact projected resource consumption.
Here I just wanted to give a preview of some of the results showing our estimates on technical hydrogen production potential from various resources. There are two tables, with the top showing the finite resources based on natural gas, coal, and uranium reserves. The second table shows five – shows results for five categories of renewable resources where the production potential is on an annual basis. The last column of these tables estimates the percentage of the resource technical potential that would be required to produce ten million metric tons of hydrogen. We'll come back to these results, but first I wanted to go through some of the assumptions and data sources that led to these results. I'd also like to take this time to recognize that this work relies on previous technical potential assessments, so I'd like to recognize the experts from across the Department of Energy and the national laboratories who helped in identifying the most current data sources to be used in this assessment.
This table summarizes the data sources used to estimate hydrogen production potential. As you can see, the data source for resource technical potential varies by technology. As a result, there will be some inherent underlying differences in the methodologies used to estimate technical potential. However, the sources used here were identified as the best data sources available at the time. The Annual Energy Outlook was used for the current and projected energy consumption values. In most cases the Department of Energy's hydrogen production models, known as H2A models, were used to determine the hydrogen production efficiency for different resources. The H2A coal gasification pathway was recently updated to match a 2010 NETL case study which is used directly in this assessment. Systems modeling at Idaho National Laboratory was used to determine the hydrogen production efficiency of a high temperature gas-cooled reactor. This reactor technology has a higher efficiency than today's available light water reactors but has yet not been commercially deployed due to economics. Finally, hydrogen production from gaseous biomass or biomethane is based on a 2018 NREL analysis report.
As shown on the previous slide, data from the U.S. Energy Information Agency, or EIA, was used to estimate the fossil and uranium resource potential. One thing I'd like to note here is that these reserve estimates are merely a proxy for actual resource potential. As technologies improve it's expected that additional reserves will be identified. For example, you can see an upward trend in the natural gas reserves in recent years due to improved technologies. In addition, I'd like to point out that uranium reserves are estimated at a forward cost only up to $100.00 per pound of uranium oxide, and as such are likely an underestimate of the true technical resource availability, which I'll discuss in more detail in later slides.
The previously listed data sources were used to create this graph comparing the technical potential of renewable resources. The generation potential has been converted to quads using EIA's fossil fuel heat rates to allow for comparison with the previous chart. However, it is important to note that the renewable potentials are on an annual basis. Zooming into the figure you can better see that the technical potential of each resource is broken down by technology. You will notice, for example, that the geothermal resource potential is dominated by enhanced geothermal systems, which have been proved technically but is yet not commercially employed. While the focus here is on technical potential, the actual future generation is of course based on economics of energy recovery and power generation.
This table shows the assumptions on hydrogen production by pathway. Again, these are mainly based on the Department of Energy's H2A models. One point I want to make on the coal pathway is that this is a NETL case study that was specific to one technology that converts bituminous coal to hydrogen. Due to lack of additional data, in this assessment we assumed that this production efficiency is valid for all coal resources. As you can see, steam methane reforming of either natural gas or biomethane has the highest hydrogen production efficiency. This is followed by low temperature electrolysis. However, it's important to note that the nuclear production efficiency is based on the thermal and electrical inputs and high temperature electrolysis actually uses less electricity than low temperature electrolysis.
So, now I'll revisit the results in more detail. This table summarizes the technical potential based on the reserve estimates previously shown and also lists the current and projected resource consumption. The amount of resource required to produce ten million metric tons of hydrogen is shown and put into context as a percentage of technical resource potential required to individually meet that demand. As you can see, across all of the resources less than one percent of the reserves would be required to meet this level of hydrogen demand. We also show the percent increase in projected resource consumption that would result from producing ten million metric tons of hydrogen. For example, there would only be a five percent increase in natural gas consumption. Because coal and nuclear are expected to contribute less to the energy mix than natural gas in 2040 these percent increases are relatively higher.
Here we show how the resource potential translates into hydrogen production potential. As previously shown, the coal reserves are the largest on an energy basis, and as a result show the greatest hydrogen production potential, of over 50,000 million metric tons. Natural gas reserve estimates have been increasing over time and are currently estimated as being able to produce 17,000 million metric tons of hydrogen. The EIA has revised their methodology for estimating uranium reserves, which has reduced the estimated resource availability and makes nuclear appear to have the lowest hydrogen production potential.
It's important for me to point out, however, that there are major uncertainties associated with the US uranium reserve estimates, as there hasn't been a thorough review or evaluation since 1980. It's expected that the resource potential is likely an underestimate and would increase as recovery and fabrication technologies progress. In addition, advanced reactor designs – for example, breeder reactors that can use spent uranium fuel – will improve the efficiency of uranium use. This is all said to specifically acknowledge that the technical hydrogen production potential from nuclear is expected to be an underestimate of the true long-term potential.
We have a similar table for the renewable resources, citing the annual technical potential and the amount of renewable resource required to produce ten million metric tons of hydrogen. As you can see, producing ten million metric tons of hydrogen would place the most strain on biomass and water-powered resources but would not exceed their technical potential. It's also important to point out how hydrogen production could increase the utilization of renewable energy as shown by the percent increase in energy consumption to produce ten million metric tons of hydrogen. For example, ten million metric tons of hydrogen would require only 2 percent of the geothermal technical potential but could increase resource consumption by over 900 percent.
Here the annual resource potentials are converted to annual hydrogen production potential for the renewable resources. This table breaks out the conventional hydropower and geothermal for more advanced technologies because both marine hydrokinetics and enhanced geothermal systems are at a relatively low technology readiness level, so we wanted to present a range here. As you can see, the range is quite large for geothermal, where including enhanced geothermal systems increases the technical potential by 100-fold. As shown earlier, solar power has the highest technical potential of the renewable resources and could potentially produce over 5000 million metric tons of hydrogen. That's over 75 percent of the approximately 6500 million metric tons of combined renewable hydrogen production potential.
Now I'll dive a little deeper into the renewable resources, starting with biomass. The Department of Energy's 2016 Billion Ton Study was used to estimate the solid biomass technical potential. Because there's a more recent source for gaseous biomass from waste resources, only the non-waste resources from the Billion Ton Study were included and a 2018 NREL analysis provided the potential of waste biomass. The total hydrogen production potential is 50 million metric tons, though it's important to emphasize that this work does not consider competition for resources. And we are not suggesting that a future hydrogen production potential would come from any one individual resource but likely a mix of energy resources.
I've also included these tables to show which biomass feedstocks were included in this analysis.
One of the main contributions of this work is the geospatial mapping of hydrogen production potential from renewable resources. These maps show the hydrogen production potential from biomass by county. Due to data limitations, only the gaseous biomass resource was able to be mapped for Alaska and Hawaii.
The wind resource potential, including both on- and offshore wind, is approximately 38,000 terawatt-hours per year. Please note, though, that the Wind Energy Technologies Office is currently updating these technical potential estimates, which may impact these results.
You can see that the hydrogen production potential from wind is highest in the center of the country. Then there's also significant potential along the East Coast and the Gulf of Mexico from offshore wind.
As previously mentioned, different data sources were used to identify technical potential by technology, so there are different references for the rooftop PV, the utility-scale PV< and the concentrated solar power. Combined, the solar technical potential is over 260,000 terawatt-hours per year.
Again, due to data limitations on rooftop PV, we were only able to map this resource by state. Here you'll notice the relatively high technical potential in California and Florida.
With utility-scale PV, the potential in California and Florida as not as high as that in the middle of the country, reflecting land availability constraints. Utility-scale PV represents about 70 percent of the total solar resource technical potential.
The potential from concentrated solar power is highest in the south and mid-to-western areas of the US.
As mentioned, a range is shown for the hydrogen production potential from water power, where the low end is based on conventional hydropower, including existing hydropower assets, non-powered dams, and new stream-reach development. The high end includes marine hydrokinetics from the sources listed in the table here. As you can see, the range for power production is from 690 to 2500 terawatt-hours per year.
Due to lack of more detailed data, the only potential from existing hydropower assets was able to be mapped by county for this assessment. New stream-reach development was mapped by state. And non-powered dams were mapped by hydraulic region. Geospatial data on marine hydrokinetics was not available for this analysis, but I do want to note that we're working with the Water Power Technologies Office to update these maps in the future.
Lastly is geothermal power. Energy availability was converted to generation potential based on the given assumptions used for flash and binary power plants. Again, a range is given for hydrogen production potential based on the exclusion or inclusion of enhanced geothermal systems. As mentioned, this range is relatively large from 3.5 to 480 million metric tons of hydrogen per year.
At the time of this analysis the map shown here was the most recent map of geothermal resources. Since this map is outdated we were unable to include a geothermal map in this report, but again, we're working with the Geothermal Technologies Office to produce an updated map.
Here we show a map of combined renewable hydrogen production potential. Only the resources for which county-level data was available are shown in the map, which represents about 90 percent of the renewable hydrogen production potential estimated in this assessment.
This figure shows the dominant renewable resource by county, and you'll see how solar tends to dominate in most of the country. In the Appalachian Mountains, however, the slope of the land precludes utility-scale PV, and thus we see wind dominating in this region. Hydropower and biomass are shown as dominating in Alaska and Hawaii, though this is likely just due to the lack of other resource data for these states.
This figure compares the current and future energy consumption from the EIA, with the dark blue bar representing the energy requirement to produce ten million metric tons of hydrogen from each resource. Note the units on the y-axis are in quads, where the EIA fossil fuel heat rate is used to convert the renewable electricity requirement to quads. This allows for comparison to the EIA values on the current and projected consumption but makes the blue bars for renewable electricity appear higher than that for coal, let's say, even though the production efficiency is actually higher. Again, you can see how there is an opportunity to increase renewable energy consumption through hydrogen production, shown as the percentages across the bottom.
Here we consider if ten million metric tons of hydrogen were to be produced through low temperature electrolysis. This would require over 500 terawatt-hours of electricity and would result in about a 10 percent increase in electricity generation in 2040. Because hydrogen can be produced from off-grid renewables, this does not necessarily imply that grid capacity would need to increase. Again, specifically considering renewables, producing ten million metric tons of renewable hydrogen could increase renewable power generation by almost 40 percent in 2040.
While a deep dive into water impacts was outside the scope of this assessment, we did want to include some information related to water consumption from various hydrogen production pathways. These values are based on Argonne's GREET model and a published record on water consumption for hydrogen production. Here you can see the direct water consumption for hydrogen production by technology. So, this is just the water that's used in the hydrogen production process. You can see high temperature electrolysis is shown as requiring the least amount of water, while coal gasification requires the greatest.
Including upstream water consumption for energy feedstocks gives more of a complete picture of the water impacts.
Eric:
Go.
Elizabeth:
As you can see, the wind and hydropower pathway include the least amount of water, though I should note we did not include evaporative losses, which may be significant for hydropower, especially when artificial reservoirs are used. Geothermal power has the greatest upstream water consumption, which is used for cooling purposes. The error bars on the geothermal bar represent the range between different technologies, including binary and flash power plants. Looking at the range of water consumption factors, the water required to produce ten million metric tons of hydrogen ranges from 29 billion to 645 billion gallons.
While that sounds like a lot, we wanted to put that into context. So, in 2015 water withdrawals in the US were over 100 trillion gallons. So, this requirement for ten million metric tons of hydrogen would represent from less than 0.03 of a percent to just over half a percent in water withdrawals. And it's important to point out that this calculation does not account for displaced water consumption. So, for example, if this hydrogen was to be used in vehicles, water consumption related to the displaced petroleum fuels should be subtracted in order to understand the full water impacts. So, in essence, these results are likely an overestimate of net water consumption.
So, that wraps up the results from this assessment, though there's still certainly more work that's needed. As previously mentioned, the economic demand and supply for hydrogen from a mix of energy resources will be further explored in the upcoming H2@Scale publications. In addition, the data presented here, in particular the geospatial data on resource potential, could be integrated into the NREL Scenario Evaluation and Regionalization Analysis model, which is being used to optimize hydrogen supply chains to meet various demand scenarios for the H2@ Scale effort. The SERA model default is least cost optimization that different criteria can be optimized around. For example, using geospatial water scarcity factors in this model could help address regional water constraints.
Finally, I just again want to acknowledge the DOE offices and the national laboratories who helped with identifying the data sources used in this assessment. And I especially want to acknowledge the Department of Energy's Hydrogen and Fuel Cell Technologies Office for funding this project. Soon this report should be publicly available online at NREL's publication page listed here, and more information on the H2@Scale initiative can be found at the second website listed. So, thank you for your attention.
Eric:
Thank you, Elizabeth, for the presentation today. And I'll remind everyone watching to make sure to check out our other H2IQ Hour videos on the HFTO website, and to make sure to use the hashtag #H2IQ on social media if you find anything interesting or want to join the conversation. And with that, thank you and goodbye.