The many greenhouse gas footprints of green hydrogen

Green hydrogen could contribute to climate change mitigation, but its greenhouse gas footprint varies with electricity source and allocation choices. Using life-cycle assessment we conclude that if electricity comes from additional renewable capacity, green hydrogen outperforms fossil-based hydrogen. In the short run, alternative uses of renewable electricity likely achieve greater emission reductions.


Methods for greenhouse gas footprint calculation hydrogen
We calculated the greenhouse gas (GHG) footprint of polymer electrolyte membrane (PEM) electrolytic hydrogen for different electricity sources and multi-functionality approaches in kg CO 2 -equivalents (kgCO 2 -eq) per kg H 2 ( Figure 2). The inputs for the life-cycle assessment (LCA), using a functional unit of 1 kg of hydrogen produced, can be found in Table S1. These are based on Bareiß et al. 1 and background lifecycle inventory data on the Ecoinvent database version v3.7.1, using allocation at point of substitution. The ReCiPe2016 method (H) v1.05 was used at midpoint level to quantify the GHG footprints. The four electricity sources used to calculate the GHG footprint are detailed in Table S2. Details on how we applied different methods to deal with multi-functionality can be found in Table S3. The GHG footprints of the benchmarks grey and blue hydrogen are less dependent on the GHG intensity of the electricity source because they mainly require natural gas as input. Bauer et al. 2 showed the contribution of electricity to the GHG footprint of grey and blue hydrogen in a contribution analysis. Based on this, we harmonised the electricity used in the electrolysis process and the grey/blue hydrogen production processes, see Table  S4. We directly used the global warming impact of 0.132 kgCO 2 -eq per kg H 2 from Bareiß et al. 1 based on 3000 full-load hours in the case of using renewable electricity. For the grid mix cases, we scaled this value to be in line with 8000 full-load hours.

Multi-functionality approach Notes
Venting of oxygen The full climate impact is assigned to hydrogen.

Substitution of oxygen
To determine the GHG footprint of 1.0 kg hydrogen production, the GHG footprint of 8.0 kg of conventional oxygen production (via cryogenic air separation) is subtracted from the footprint of the overall production process. We used the Ecoinvent v3.7.1 process 'Market for oxygen, liquid {RER}' and adapted this process to ensure harmonisation of the electricity used in electrolysis and the substituted oxygen production process. We replaced the electricity source for the required 1.42 kWh/kg O 2 in the Ecoinvent process by the four electricity sources specified in Table S2. This approach avoids that benefits of substitution would be inflated by a replaced process that runs on more GHG-intensive electricity.

Economic allocation
We used the factors in Bargiacchi et al. 6 based on the 2020 prices of hydrogen (1.21 USD/kg) and oxygen (0.25 USD/kg). This led to assigning 37.7% of the impacts to hydrogen and 62.3% to oxygen. For a CH 4 emission rate of 1.5%, GWP100 and a 55% CO 2 capture rate Bauer et al. 2 found a GHG footprint of 6.63 kgCO 2 -eq/kg H 2 , of which 0.11 kgCO 2 -eq/kg H 2 is due to electricity use 2 . This equals 0.28 kWh/kg H 2 based on the indicated ENTSO-E GHG intensity (0.39 kgCO 2 -eq/kWh), which we replaced by the four electricity sources specified in Table S2 to find the harmonised blue hydrogen (55% capture) footprints. Blue hydrogen (93% capture) For a CH 4 emission rate of 1.5%, GWP100 and a 93% CO 2 capture rate Bauer et al. 2 found a GHG footprint of 3.61 kgCO 2 -eq/kg H 2 , of which 0.41 kgCO 2 -eq/kg H 2 due to electricity use 2 . This equals 1.04 kWh/kg H 2 based on the indicated ENTSO-E GHG intensity (0.39 kgCO 2 -eq/kWh), which we replaced by the four electricity sources specified in Table S2 to find the harmonised blue hydrogen (93% capture) footprints.

Methods for the comparison of green hydrogen against alternative uses of renewable electricity
We calculated the GHG emissions or savings for different ways of using 1 kWh produced from newly built offshore wind capacity (Figure 3 and Table S6) and from newly built solar PV in Europe ( Figure S1 and Table S7): for green hydrogen replacing grey hydrogen or blue hydrogen (55% or 93% CO 2 capture rates), for electric cars replacing petrol cars and for grid decarbonisation, replacing coal or natural gas electricity. We expressed the GHG footprints of using 1 kWh of additional renewable electricity in kgCO 2 -equivalents per kWh of electricity. We arrived at these values by dividing the GHG footprints per original functional unit by the electricity used in kWh per functional unit, for example 55 kWh per kg of H 2 in the case of green hydrogen. For green hydrogen it is assumed that oxygen is vented, and the values are based on blue and grey hydrogen are from Table S5. Emissions and emissions savings for electric vehicles and heat pumps in the EU are calculated based on Knobloch et al. 7 , where we took a weighted average of the emissions associated with electric vehicles and heat pumps for the 2022 EU-27 countries, based on absolute vehicle kilometres in each country and heat demand in each country, respectively. For grid decarbonisation we based on Hertwich et al. 8 .
The greenhouse gas emissions in Table S6 and Table S7 are the emissions associated with the use of the electricity (e.g., green hydrogen), and the avoided greenhouse gas emissions are those that are avoided (e.g., avoiding grey hydrogen production) and therefore include a negative sign. The overall greenhouse gas emission reductions are the sum of the two.