Hydrogen is a zero-carbon energy carrier commonly used in the production of chemicals and fuels. Depending on the energy source and technology used for hydrogen production and delivery, there can be life-cycle greenhouse gas (GHG) emissions associated with hydrogen production and use. Clean energy sources and carbon management measures (e.g., carbon capture and storage) can be deployed to provide the largest reductions in GHG emissions associated hydrogen production and use.

Interest has grown globally in the economic production of clean hydrogen to replace the current carbon-intensive hydrogen production via reforming of fossil natural gas. In particular, the United States (U.S.) Congress passed several laws that incentivize the production and use of renewable and low-carbon hydrogen. The Bipartisan Infrastructure Law (BIL) passed in 2021 provided $9.5B to incentivize the deployment of clean hydrogen hubs in various parts of the United States and the development of electrolyzer technologies. To qualify for the BIL incentives, the produced hydrogen must meet specific GHG emissions standards on a life-cycle basis. Seven hydrogen hubs were awarded in 2023 in different regions of the United States. In 2022, Congress passed the Inflation Reduction Act (IRA), which provides tax credits of up to $3/kg depending on the carbon intensity of the produced hydrogen.

Due to the impact of hydrogen carbon intensity (CI) on the end use applications of hydrogen (e.g., fuel cell vehicles, ammonia, synthetic fuels and chemicals, metal reduction, etc.), many authors have examined the CI of hydrogen produced by various methods. According to the U.S. Department of Energy, steam methane reforming (SMR) of natural gas (NG) has been one of the most common methods of hydrogen production in the United States in 2024 (U.S. DOE undated). However, the CI of hydrogen produced via SMR is high. A recent study by Henriksen et al. (2024) reported life-cycle emissions of 10.6 kgCO₂e/kgH₂ while Zang et al. (2024) reported 9.9 kgCO₂e/kgH₂ including displacement credits due to the export of excess steam produced. Sun et al. (2019) studied SMR facility level GHG emissions and determined a U.S. national average of 9.3 kgCO₂e/kgH₂. In a report published by the National Energy Technology Laboratory (NETL) which modeled an SMR facility, they determined that the CI of hydrogen produced to be 12.4 kgCO₂e/kgH₂, which can be reduced to 10.2 kgCO₂e/kgH₂ if the produced steam was exported (Lewis et al., 2022). Khojasteh Salkuyeh et al. (2017) determined that hydrogen produced by SMR would have emissions of 11.5 kgCO₂e/kgH₂, which is supported by other authors determined that the life-cycle emissions of hydrogen produced by steam methane reforming of natural gas were 11.9 kgCO₂e/kgH₂, including embodied emissions from capital expenditure (CapEx) (Cetinkaya et al., 2012; Spath and Mann, 2000). Other studies report slightly lower emission values of between 10.4 CO₂e/kgH₂ (Susmozas et al., 2013). To reduce the CI of hydrogen produced from NG SMR, carbon capture and storage (CCS) systems were implemented. This approach resulted in a sharp reduction of GHG emissions to between 3.4 and 3.6 kgCO₂e/kgH₂ (Khojasteh Salkuyeh et al., 2017; Henriksen et al., 2024; Zang et al., 2024; Dufour et al., 2010).

Due to the high CI of hydrogen produced from NG SMR without CCS, many hydrogen production technologies have been studied as possible alternatives. One such method is autothermal reforming (ATR) of natural gas. A literature review determined that many authors reported GHG emissions from ATR with CCS in the range of 3.3–3.7 kgCO₂e/kgH₂ (Henriksen et al., 2024; Zang et al., 2024), although emissions as high as 3.9 kgCO₂e/kgH₂ have been reported (Hajjaji et al., 2013). In the absence of CCS, ATR GHG emissions increase to between 10.8 and 11.0 kgCO₂e/kgH₂ (Hajjaji et al., 2013; Oni et al., 2022). Gasification of carbonaceous material such as coal or biomass is another method of hydrogen production. Authors have reported emissions from coal gasification ranging between 11.6 and 20 kgCO₂e/kgH₂ without CCS (Cetinkaya et al., 2012; Henriksen et al., 2024; Verma and Kumar 2015), although emissions can reach as high as 25.3 kgCO₂/kgH₂ if lignite is used (Burmistrz et al., 2016). With the inclusion of CCS, Henriksen et al. (2024) determined that the CI of hydrogen decreases to 3.9 kgCO₂e/kgH₂, while Burmistrz et al. (2016) noted GHG emissions ranging between 4.1 and 7.1 kgCO₂e/kgH₂. To further reduce the CI of hydrogen produced via gasification, biomass may be used in place of coal as it is a biogenic source of carbon. However, reported emissions vary widely due to differences in the carbon content of the biomass gasified (Henriksen et al., 2024; Moreno and Dufour, 2013). Susmozas et al., 2013 investigated the CI of hydrogen produced from poplar gasification and reported GHG emissions of 0.4 kgCO₂e/kgH₂. With CCS, this value decreases to −14.6 kgCO₂e/kgH₂ (Susmozas et al., 2016). In contrast, Henriksen et al. (2024) reported GHG emissions of 5.3 kgCO₂e/kgH₂ without CCS and −15 kgCO₂e/kgH₂ when CCS is included. Kalinci et al. (2012) reported emissions of 8.5 kgCO₂e/kgH₂ using pinewood, while Li et al. (2020) calculated GHG emissions of 5.4 kgCO₂e/kgH₂ for hydrogen produced using straw.

Finally, the electrolysis of water presents the possibility for low CI hydrogen production should renewable electricity, such as wind, solar, hydro, or nuclear electricity be used. Low-temperature electrolysis (LTE) is typical for proton exchange membrane (PEM) and alkaline electrolyzer technologies, while high-temperature electrolysis (HTE) is typical for solid-oxide electrolyzer cell (SOEC) technology. While nuclear high-temperature gas cooled reactors (HTGR) have supply chain upstream emissions, they can produce hydrogen with CI as low as 0.4 kgCO₂e/kgH₂ (Giraldi et al., 2015). Karaca et al. (2020) reported similar values of 0.4 and 0.5 kgCO₂e/kgH₂ for electrolysis using boiling water reactors (BWR) and pressurized water reactors (PWR), respectively. These values are also supported by Bicer and Dincer (2017), whose study indicated GHG emissions between 0.5 and 0.6 kgCO₂e/kgH₂ for hydrogen produced by high-temperature electrolysis and electrolysis, respectively (Bicer and Dincer, 2017).

CapEx-embodied emissions of clean electricity contribute significantly to the CI of hydrogen produced via electrolysis. Iyer et al. (2024) reported emissions as high as 2.4 kgCO₂e/kgH₂ for electrolysis using solar power and 0.5 kgCO₂e/kgH₂ when nuclear power from LWRs is used. These results are supported by Gan et al. (2024) who reported GHG emissions of 2.1, 0.6, 0.4, and 0.3 kgCO₂e/kgH₂ for water electrolysis using solar, wind, hydro, and nuclear power, respectively. Henriksen et al. (2024) further distinguishes between the type of electrolysis methods, noting emissions of 1.8 and 2.2 kgCO₂e/kgH₂ for PEM and SOEC, respectively, when wind electricity is used, which increases to 2.8 and 2.9 kgCO₂e/kgH₂ when solar energy is used. Palmer et al. (2021) reported a similar CI of hydrogen produced via electrolysis using solar power of 2.3 kgCO₂e/kgH₂. Cetinkaya et al. (2012) reported GHG emissions of 1.0 and 2.4 kgCO₂e/kgH₂ for hydrogen produced by using wind and solar power, respectively.

As can be seen, many LCA studies of hydrogen have been carried out over a span of multiple years, covering different system boundaries, and from the perspective of different countries. A comprehensive life-cycle accounting of GHG emissions associated with hydrogen production is needed to determine the carbon intensity of hydrogen throughout its value chain. This paper aims to provide readers with the information necessary to understand hydrogen production in the United States and the life-cycle emissions associated with it using consistent system boundaries and most recent technology performance and supply chain data. In the United States, Argonne’s R&D GREET^® (Greenhouse Gases, Regulated emissions, and Energy use in Technologies) model has been widely used for hydrogen carbon intensity calculations. R&D GREET provides in-depth life-cycle simulations for a variety of energy and chemical products, including hydrogen technologies. This paper describes the major hydrogen technology pathways considered in the United States and provides data sources and carbon intensity results for each of the hydrogen production pathways.

In this section, we discuss hydrogen production technologies, their necessary process inputs, and the operating conditions in R&D GREET, 2023. We also highlight important parameters that have a significant impact on the well-to-gate (WTG) CI of hydrogen production and any accompanying data sources. Detailed data for each pathway can be found in the Supplementary Material. The co-product allocation method in GREET follows the ISO 14067 protocol, with system expansion and substitution method applied where feasible, and allocation based on physical attributes (e.g., by energy or mass) is otherwise applied.

While methane reforming and water electrolysis are established technologies and widely used methods of hydrogen production, there are several emerging technologies for hydrogen production that are at various technology readiness levels (TRLs).

The WTG emission results for the hydrogen pathways in R&D GREET are provided in Figure 1. These results do not include CapEx embodied emissions associated with construction of hydrogen production facilities. Conventional SMR without CCS and NGL cracking produce hydrogen with the highest carbon intensities at 9.3 and 8.5 kgCO₂e/kgH₂, respectively. The high carbon intensity of hydrogen production from NGL cracking comes from replacing energy in hydrogen separated from tail gas with natural gas combustion for NGL cracking process. While the production of hydrogen via coal gasification generates high GHG emissions at 18.6 kgCO₂e/kgH₂, this is not a common method of hydrogen production in the United States. Coal gasification with CCS significantly reduces GHG emissions associated with hydrogen production. The other methods of hydrogen production, such as water electrolysis using clean electricity, by-product hydrogen from chlor-alkali, biomass gasification, methane pyrolysis, and SMR and ATR with CCS have significantly lower hydrogen CI as shown in Figure 1. When CCS systems are implemented or RNG is used, ATR and SMR on-site emissions can be significantly reduced.

For many hydrogen production technologies, upstream (supply chain) emissions play a significant role in the CI of produced hydrogen. Upstream emissions for feedstock and energy supply to hydrogen production technologies can be subdivided into two categories: Upstream emissions associated with electricity generation and transmission, and upstream emissions associated with feedstock and process energy acquisition. Upstream emissions associated with electricity supply are the primary driving factor of GHG emissions for electrolysis, chlor-alkali, and methane pyrolysis pathways, accounting for at least half of overall emissions in the pathway. As a result, the hydrogen CI from these pathways can be significantly reduced if clean electricity is used, thus presenting a significant decarbonization opportunity. On the other hand, upstream emissions associated with feedstock and process energy acquisition are the key drivers of GHG emissions for coal gasification with CCS, biomass gasification, SMR with CCS, SMR using LFG, and ATR with CCS, as shown in Figure 1. However, unlike upstream emissions associated with electricity supply, these emissions are harder to mitigate by hydrogen producers. This is because their supply chain often involves multiple activities that are heavily reliant on fossil fuels, such as mining for coal, recovery, processing and fugitive methane emissions for natural gas, as well as transporting these feedstocks to the hydrogen production plant.

We examined the impact of electricity CI on several hydrogen pathways in Figure 2. For this sensitivity analysis, we examined electricity from three different sources: renewable electricity, United States grid mix (440 gCO₂e/kWh), and Hawaiian grid mix (870 gCO₂e/kWh). SMR, NGL, and biomass gasification are the least affected by the CI of electricity due to the minor role of electricity in these processes, accounting for 0.5, 2.6, and 11% of hydrogen CI, respectively, as shown in Figure 1. As the implementation of CCS systems demand additional electricity, hydrogen production pathways that use CCS, such as SMR, ATR, and coal gasification show higher contribution of electricity supply emissions to hydrogen CI. Water electrolysis and methane pyrolysis, however, show the largest impact of electricity supply emissions on hydrogen CI. This is because of their significant electricity consumption per unit of hydrogen produced, as shown in Supplementary Table S3. As the C-A process also involves high electricity use for the electrolysis of brine solution, the CI of its products are also expected to be largely dependent on the CI of electricity supply. Hydrogen CI produced via water electrolysis is directly proportional to the CI of electricity used.

We also carried out a sensitivity analysis on hydrogen production pathways that rely on natural gas. As methane is a potent GHG with a Global Warming Potential (GWP) of 29.8 (Intergovernmental Panel on Climate C, 2023), methane emissions in the NG supply chain can have a significant impact on the hydrogen CI. Sensitivity analysis was carried out using methane emission rates ranging from 0% (low) to 2% (high), with a default value of 1%. The results of this analysis are shown in Figure 3. SMR with CCS and ATR with CCS showed the greatest dependence on methane emission rate, with WTG emissions increasing by 1.2 and 1.1 kgCO₂e/kgH₂, respectively, at 2% methane emission rate, compared to default value. The increase in GHG emissions for SMR with CCS is higher compared to SMR, as more natural gas is required to meet the increased energy demands of the CCS. Biomass gasification was insensitive to the changed methane emission rate, as natural gas is not a primary energy source for the gasification process. While chlor-alkali and methane pyrolysis, which is sensitive to methane emissions, rely heavily on electricity, the allocation of emissions between hydrogen and other co-products reduces the sensitivity to the methane emission rate.

Figure 4 shows the gate-to-pump (GTP) emissions for each step after the hydrogen leaves the production plant up to hydrogen refueling station for hydrogen vehicle applications. These emission sources can be broken down into several steps: compressing gaseous hydrogen for delivery and storage, liquefaction, transportation to destination, and any additional compression and precooling required for dispensing or storage at HRS. Compressing hydrogen at the plant gate for tube-trailer loading or pipeline delivery is carried out next to production and thus, the electricity supply is assumed to be sourced from the regional grid around the hydrogen plant. Using U.S. grid electricity for this compression step contributes 0.8 kgCO₂e/kgH₂ to overall emissions for loading tube-trailers and 0.3 kgCO₂e/kgH₂ for pipeline delivery, constituting 63% of total GTP emissions for gaseous hydrogen transported via tube-trailers and 20% for pipelines. As the compression process for hydrogen delivery requires a dispatchable source of electricity, intermittent electricity sources such as solar PV (photovoltaic) and wind may not be suitable for this process without energy storage system to mitigate the intermittency. Thus, compression is assumed to rely on regional grid electricity, even though the facility may be using solar PV or wind electricity for hydrogen production. To reduce these emissions, other clean dispatchable electricity sources such as hydroelectricity and nuclear power can be used, as shown in Figure 4. However, this effect is less significant when hydrogen is transported through pipelines due to the higher electricity consumption at HRS. Using grid electricity, the pressurization requirements for storage at the HRS contribute 1.2 kgCO₂e/kgH₂ to overall emissions and represent 76% of GTP emissions for pipelines. While gaseous hydrogen delivery via pipeline is less energy and emissions intensive (80% lower) compared to tube-trailer delivery, it incurs higher GHG emissions at HRS due to the higher compression ratio at HRS. Alternatively, the gaseous hydrogen may be liquefied and subsequently transported to the HRS via cryogenic tanker truck. The liquefaction step is energy-intensive, contributing a significant 5 kgCO₂e/kgH₂ to GTP emissions when U.S. average grid electricity is used. These emissions can only be reduced by using a dispatchable source of clean electricity, e.g., hydro and nuclear power. For this analysis, we assumed that all hydrogen production methods used grid electricity unless otherwise specified.

While we did not include CapEx embodied emissions in the aforementioned analysis, our previous work on embodied emissions clearly showed that it has a significant impact on the CI of hydrogen produced via electrolysis using renewable energy (Gan et al., 2023; Gan et al., 2024). Therefore, it is important to understand the impact of embodied emissions on the CI of hydrogen produced from various technologies.

In Figure 5, we included embodied emissions coming from the feedstock and electricity inputs, such as the construction of coal mines and farming equipment, as well as the GHG emissions from the construction of hydrogen production facilities and electrolyzers (Iyer et al., 2024). In addition, we assumed that the emissions from SMR facility construction would be similar for ATR, methane pyrolysis, and gasification. However, these emissions are small and do not have a significant impact on the CI of hydrogen. Figure 5 shows the impact of CapEx embodied emissions on the CI of hydrogen produced via a variety of technology pathways, such as SMR, ATR, coal and biomass gasification, and water electrolysis. We observed that the CapEx embodied emissions for SMR, ATR, coal gasification, and methane pyrolysis comprise between 0.7% and 6% of total life cycle GHG emissions associated with hydrogen production. Embodied emissions have a large impact on hydrogen CI from biomass gasification, mainly due to the farming equipment. They also have large impact on the hydrogen CI from water electrolysis using renewable electricity, contributing approximately 2 kgCO₂e/kgH₂ for solar power.

This paper examines the CI of hydrogen produced via a variety of conventional and emerging technologies, as well as its packaging (compression and liquefaction), delivery and storage in the United States. The feedstock and energy supply chain upstream GHG emissions play a significant role in the CI of hydrogen production and thus using biogenic feedstock (e.g., biogas) and/or clean electricity (e.g., renewable or nuclear power), a low CI hydrogen production can be achieved. As clean electricity is currently the simplest approach of the two, many hydrogen producers may consider water electrolysis or other emerging technologies such as plasma pyrolysis of methane, both of which consume large amounts of electricity per unit of hydrogen produced. Also, onsite emissions can be mitigated via carbon management technologies such as CCS. Hydrogen packaging and delivery also contribute significant GHG emissions due to the high electricity requirement for hydrogen compression and liquefaction. Finally, embodied emissions can have a significant impact on the CI of hydrogen produced and cannot be ignored in some cases. Overall, mitigating feedstock supply chain GHG emissions (e.g., CH₄ emissions in the NG supply chain), utilizing cleaner energy supply (e.g., clean electricity for water electrolysis and methane pyrolysis, and RNG for SMR and ATR), implementing carbon management strategies (e.g., CCS), and using clean electricity for hydrogen compression and liquefaction can substantially reduce the life cycle carbon footprint of hydrogen supplied to various end use applications.