Fuelling the future: Assessing multifuel filling stations for hydrogen and other renewable fuels through life cycle analysis

Hydrogen could play an important role in reducing the climate impact of the transport sector. This study explores the possibility of using existing biomethane infrastructure to enable the accelerated roll‐out of hydrogen as a transport fuel in a Swedish context. The concept of multifuel filling stations for hydrogen and biomethane are examined based on four cases, where the hydrogen is produced either via electrolysis or biomethane reforming, at a smaller or larger scale, and through either centralised or decentralised production. The cases are compared using established life cycle assessment (LCA) methodology to establish their respective impact from a greenhouse gas (GHG) emission mitigation potential. The LCA results show generally good GHG performance for all production paths being studied with a range from −7 g CO2 eq./MJ hydrogen for hydrogen production based on biomethane via steam reformation (SMR) compared to +19 g CO2 eq. for production based on Swedish National Grid Mix via electrolyser. The SMR is the more efficient technology in mitigating GHG emissions, especially if system expansion is applied. In addition, sensitivity analyses also show that electrolyses production based on renewable wind power will decrease the impact significantly and vice versa that a European Average Electricity Grid Mix (EU – 28) would increase the impact significantly. The findings of this study underline the potential of the gradual introduction of hydrogen as a fuel for transport without the need for large investments in a dedicated fuel‐specific distribution system. The concept could contribute to overcoming the current chicken‐and‐egg catch of achieving both scalable and profitable supply of hydrogen for transport as well as the vehicles using it as fuel.


| INTRODUCTION
Hydrogen is expected to play an important role in mitigating the environmental impact of the transport sector and is already considered an important component across various modes of transport including road, sea, and aviation. 1 However, despite the wide use of hydrogen in industry, the uptake in the transport sector has remained limited, with little over 40,000 fuel cell vehicles in use worldwide as of June 2021. 2 Moreover, the refuelling infrastructure remains scarce, with a total of 540 stations in operation worldwide in late 2020, 2 primarily concentrated to Japan (140), Germany (90) and China (85). 3 In Sweden there are currently only five hydrogen filling stations, though the number is expected to grow rapidly since there are several dozens more planned for the forthcoming years. 4 The concept of a multifuel filling station includes offers refuelling of multiple fuels at a single point of fuelling and is one way to address the early market barriers to an increased uptake of alternative fuels such as hydrogen. 5 In general, the hydrogen production and distribution involves two main approaches. In a centralised system, hydrogen is produced in large-scale plants and then transported to the location of its use in dedicated systems (such as gas cylinders or possible gas pipelines for hydrogen only). In a decentralised system, hydrogen is produced at the site of its use or final distribution.
In this work, two options are considered for hydrogen production, where the gas is produced either via electrolysis of water or through reforming of biomethane, where biomethane in this project refers to already upgraded biogas. In the latter case, an already established distribution system for biomethane could be used, where the gas is led to a small-scale steam methane reformer (SMR) and in which steam reacts with biomethane in the presence of heat and a catalyst to form hydrogen, carbon dioxide and carbon monoxide. Similarly, already present distribution systems for electricity and water can be used to supply raw materials for hydrogen production through electrolysis at the site of its use. Alkaline electrolysis has been assumed for the present study representing commercially available technology. More advanced electrolysis concepts for hydrogen achieving higher system efficiency are under development (such as steam electrolysis 6,7 ), but have not been considered in the present study as their commercial potential still needs to be proven.
There is a rich and growing literature exploring the hydrogen technology as an alternative fuel for transportation applications. Khzouz et al. 8 investigate life cycle costing analysis as a tool to estimate the cost of hydrogen to be used as fuel for hydrogen fuel cell vehicles, comparing centralised and decentralised production systems and two hydrogen production methods, methane reforming and water electrolysis. The authors also provide an overview of recent studies of hydrogen economy for mobility applications. Including, for example, a study by Reuß et al. 9 which evaluates all parts of a future hydrogen supply chain, from hydrogen production to refilling using Germany as case. Reuß et al. 9 also in a later study investigates optimisation potential for hydrogen distribution.
The aim of this study is to explore the possibility of using already established infrastructure for biomethane to enable the extended and accelerated deployment of hydrogen as a fuel, through a systematic evaluation of the climate impact of four different setups that could supply a filling station with both hydrogen and biomethane.

| MATERIALS AND METHODS
This section presents the method applied to the life cycle assessment (LCA) of the greenhouse gas (GHG) emission mitigation potential of the studied system solutions.
In the selected system setup, hydrogen is produced via electrolysis in an alkaline electrolyser, for which model data is collected from. 10 During electrolysis, water is electrochemically split into hydrogen and oxygen using electricity. The assumed system efficiency of the electrolysis process is 62%. 11 The main inputs accounted to the electrolysis process are therefore electricity, water and auxiliary materials needed. Heat released (at 70°C) in the electrolysis step is modelled as waste heat released at and not as a cproduct in the baseline scenarios. Similarly, oxygen released from the processes is considered an emission. For this reason, no allocation is needed in the electrolysis base-case. The consecutive process steps during hydrogen production relate to adjustments of the pressure of the obtained gas to 20 bar(a), thus requiring mainly electricity for their operation.
The main inputs to the SMR process are biomethane, electricity, and water (in the form of steam), where hydrogen is produced from the biomethane in multiple process steps. The SMR process is modelled based on data from a supplier of small-scale reformer systems. 12 The feedstock to the biogas facility is assumed to consist of 22% food waste, 40% manure, 28% industrial waste and 11% slaughterhouse waste. 13 The feedstock is assumed to be transported 20 km to the biogas facility. During the reforming process, carbon dioxide and residual heat are obtained besides hydrogen. In the SMR-baseline scenario, it is assumed that there is no use for the carbon dioxide gas and the residual heat, which are released to the environment. For this reason, there is no need for allocation in this case. During biogas production, digestate is obtained that is assumed to replace mineral fertiliser, in accordance with the ISO methodology. 8,9 Some emissions are assumed to arise from methane-slip throughout the process chain: 0.3% from the biogas production plant, 4% from biogas flaring (and of which an extra 2% of these 4% are lost in methane slip in this process) and lastly 0.3% in the upgrading process, representing the average slip caused by amine and water scrubbers. 13 In the cases with off-site, centralised production of hydrogen, distribution of compressed hydrogen is performed by trucks in cylindrical steel or low-weight composite vessels, assuming an average transport distance of 150 km. It is further assumed that distribution is performed with biomethane-fuelled trucks.
The environmental impact of the studied system setups is evaluated from a life cycle perspective and by considering all key activities related to hydrogen production, distribution, and storage at the filling station. In this work, the focus of the assessment has been on the impact on climate change and the associated performance of different hydrogen production systems in terms of (reduced) GHG emissions.
The study applies two methods for estimating the impact of the studied systems. First, the widely adopted LCA methodology is applied. LCA quantifies the potential environmental impacts related to a product or a system during its whole life cycle, that is, from material extraction and manufacturing to use and end-of-life. The method is described in a variety of standards and guidelines, for example, ISO 14040 14 and ISO 14044. 15 Since this work is focused on the environmental performance in terms of GHG emissions, a streamlined approach is applied, focusing on the climate change or global warming potential (GWP 100) indicator (expressed in g CO 2 eq.). 16 The Renewable Energy Directive (RED) 17 offers a complementary method to assess the climate impact of transportation fuels. The RED was introduced by the European Union with the aim to promote the use of renewable fuels in the energy and transport sector. RED is not an LCA framework, but it has a life-cycle-based approach where the GHG emissions from transportation biofuels are accounted for. More specifically, the second method proposed in the revised version of the RED directive, 18 referred to as Renewable Energy Directive II (RED II), is applied here as a comparison to the standard ISO method. In both RED and RED II, the calculations of GHG emissions from biofuels and bioliquids consider all life cycle stages from raw material production to the final use of the fuel.
Data in relation to the studied systems were estimated in this work based on available literature. The LCA was modelled in GaBi 2018 (version 10). 19 Generic life cycle inventory datasets were used to model background processes. 20-22

| System boundaries
The four different cases considered have been described in short in the case description above. A more detailed description can be found in the Supporting Information Material.
The evaluation of the climate impact employs a wellto-tank lifecycle approach. In all four cases, the studied contains all key activities related to the production of hydrogen, the distribution of hydrogen and the storage of hydrogen at the filling station. This includes the processes of feedstock production and the acquisition of biomethane and electricity. Simplified flowcharts of the studied systems are shown in Figures 1 and 2. In the centralised cases (Figure 1), hydrogen is produced in central facilities and then distributed to the filling station via road transport. In the decentralised alternatives ( Figure 2), hydrogen is produced at the filling station by electrolysis of water or reforming of biomethane. Water is transported to the filling station in water pipelines, biomethane is transported to the filling station in gas pipelines or by road transport, and electricity is supplied through its established distribution system.
It is important to note that the different approaches to accounting for the coproducts in the hydrogen production process result in variations to the system boundaries. This mainly affects the biogas production process steps. According to the methodology proposed in the RED II and the ISO standard, biogas production receives a credit from avoided emissions of manure storage. Moreover, in the case of ISO standard 8,9 system expansion allows biogas digestate to receive a credit from avoided emissions resulting from the production of mineral fertiliser. Data for these credits are collected from. 13 The functional unit of the study is 1 MJ of compressed gas (hydrogen) at 20 bar (absolute pressure) stored at the filling station.

| Scenario analysis
In addition to the baseline cases for electrolysis and SMR presented above, two additional scenarios have been investigated. The scenarios are briefly described below: Scenario A: This scenario investigates different ways of modelling the heat that is released both during (i) electrolysis and (ii) biomethane reforming.
1. Heat released during water electrolysis in the central hydrogen production pathway is considered a coproduct, which makes electrolysis a multifunctional process. To account for the impact of hydrogen production only, energy-based allocation is used in RED II. The allocation factor for hydrogen was calculated to approximately 0.8 based on the energy content of hydrogen (120; MJ/kg) and heat (31 MJ/ kg). In the case of ISO, system expansion and substitution are applied. Therefore, the heat released F I G U R E 1 Process description of centralised (not at location of filling station) hydrogen production via water electrolysis and steam reforming of biomethane (SMR). For the baseline scenario, the heat released from the electrolysis and reformer step is considered as waste heat thus no allocation is needed. CGH2, compressed hydrogen; LP storage, low-pressure storage.
F I G U R E 2 Process description of decentralised (at location of filling station) hydrogen production via water electrolysis and steam biomethane reforming (SMR). For the baseline scenario, the heat released from the electrolysis and reformer step is considered as waste heat thus no allocation is needed.
was assumed to have an adequate temperature level of approximately 70°C 23 to replace an equivalent amount of heat produced in the district heating network in Sweden. The considered fuel mix in the district heating network, with 44% woody biomass and 25% of waste was based on data from the IVL Swedish Environmental Insititute internal database 22 which originates from data on district heating production during the year 2014 by Swedish District Heating Association, known as Swedenergy. 24 2. The heat that is released during the biomethane reforming was considered to be recirculated to the biomethane production and assume to reduce the needs for external heat by 25%.
Scenario B: This scenario investigates different options for the supply of electricity throughout the entire process chain: domestic wind power, Nordic Average Grid mix and European Average Grid mix (EU- 28). No other variation is adopted compared to the baseline cases. The datasets were obtained from Sphera, version 2021. 19

| RESULTS AND DISCUSSION
The GWP of the four hydrogen production alternatives over a 100-year period is illustrated in Figure 3, presenting the results in grams of CO 2 equivalents per megajoule of hydrogen stored at the filling station (g CO 2 eq./MJ of H 2 ). The results represent the baseline scenarios, where no coproduct allocation or benefit is considered, apart from the credit assigned to the biogas process digestate and manure used as feedstock in the biogas production.
In the case of hydrogen production through water electrolysis, GHG amount to 19 g CO 2 eq./MJ hydrogen, with very small variations between centralised and decentralised production. In line with previous studies, for example, [25][26][27] the electrolysis step contributes to the majority of emissions, despite the nine times lower carbon emission factor of the Swedish electricity mix compared to the European average (according to Sphera 19 using datasets for the electricity mix with a reference year of 2017).
Similarly, the comparison between the centralised and decentralised hydrogen production via biomethane reforming shows marginal variations with the total impact between −4.8 to −5.1 (RED II method) and −6.7 to −7.0 g CO 2 eq./MJ hydrogen when the benefits of digestate are taken into consideration (ISO method).
Comparing the four different alternatives, hydrogen production via the biomethane reforming pathway remains the alternative with the lowest impact, even without the manure and digestate credit. The net-negative results are however caused by the RED II and ISO methodology, where the biogas production using manure as feedstock is assigned credits from avoided emissions.

| Scenario analysis
The results from the additional scenarios (noted as A and B) are illustrated in the figures below. Figure 4 showcases the results of Scenario A, where the heat F I G U R E 3 Results illustrating the climate change impact indicator (expressed in g CO 2 eq./MJ hydrogen at the filling station) for the different hydrogen production pathways investigated in this report. The numbers on each bar indicate the total impact, or net impact in case of negative emissions. The dominant process step for electrolysis-based hydrogen is the electrolysis, while the reforming process causes the largest carbon emissions in the biomethane cases. released at the central hydrogen production is recovered. In the electrolysis case, the heat released is assumed to replace heat produced in a district heating system while in the case of biomethane reforming, it substitutes part of the external heat demand to the biogas production process. Both systems demonstrate reduced climate impact when the heat resulting from the hydrogen production is recovered, with electrolysis pathways showcasing emission reductions of 16%-19% and biomethane reforming pathways showcasing reductions of 6%-8%, depending on the LCA method employed.
In Scenario B, the baseline assumptions are maintained for all studied systems, but the Swedish electricity mix is substituted with alternative electricity mixes, including electricity from wind power but also the Nordic or European average mixes. This scenario is of great relevance to the electrolysis pathways where electricity dominates the total GHG emissions. As shown in Figure 5 and for the production routes where electrolysis is used, a reduction of 82% on the total GHG emissions is obtained compared to the baseline when wind power is used. In the case where a reformer is used, a reduction of 30%-48% can be obtained depending on if the system is centralised or decentralised and what LCA method is used. As for the other two alternatives, Nordic and EU electricity mix, the GHG emission levels become significantly higher, especially in the cases where electrolysis is used. This highlights both the importance of ensuring a low-carbon electricity supply for the hydrogen production via electrolysis and the significance of the representation of the electricity system, its system boundaries and the technology mix on the results on the climate impact indicator.
As previously mentioned, the concept of a multifuel filling station includes the provision of multiple fuels at a single point of fuelling. However, its purpose extends beyond this as it enables the gradual market introduction of a new fuel, such as hydrogen, by leveraging existing infrastructure and distribution networks. This concept offers several benefits, including overcoming the chickenand-egg dilemma of fuel-cell vehicles needing sufficient infrastructure to reach a higher market uptake and filling stations requiring a certain level of demand to be profitable. Additionally, establishing a hydrogen station at an existing biomethane filling station significantly reduces the time required for decision-making and realisation compared to starting from scratch. Colocating with existing filling stations, which are often located in prime areas, provides easier access to hydrogen as a fuel, benefiting drivers by reducing the time needed to locate a station and refuel their vehicles.
The availability of sustainable feedstocks such as waste or residue streams for biogas production is highlighted as a crucial factor in reducing GHG emissions associated with hydrogen production through reforming of biomethane. RED II and the updated calculation rules clearly benefit systems using manure as feedstock. Systems that employ manure as a feedstock benefit from the regulations outlined in the RED II and updated calculation rules. Although concerns regarding resource availability are often raised, recent studies indicate that the biogas potential of manure and other organic residues for biogas production in Sweden is estimated to be 6-8 TWh by 2030 and 2045, respectively. 28 Similarly, the use of renewable electricity as an energy source for hydrogen production via electrolysis is a decisive factor in reducing GHG emissions. Another approach to mitigating the climate impact of hydrogen is through the utilisation of coproducts generated during the electrolysis process. In the baseline scenario in this study, all climate impact of the electrolysis is allocated to the produced hydrogen. If heat and oxygen are utilised and are considered as coproducts, the overall climate impact of hydrogen production could be lowered as it will be allocated to all three outputs. This approach seems most realistic in the centralised scenarios. Only the utilisation of heat is considered in this study, while oxygen utilisation has not been included due to the uncertainties regarding its feasibility. The potential concept of using excess oxygen has been discussed in the literature and some industrial examples are also available including uses in the metal and chemical industry or in medicine. 29,30 Thus, investigating its impact of lifecycle assessments of the climate impact should be the focus of a future study.
Finally, it should be noted that centralised production systems have a higher potential for coproduct utilisation due to the ease of recovery and distribution. This characteristic contributes to improving the overall impact and energy efficiency of the process.

| CONCLUSIONS
The GHG emissions from the assessed alternatives range from −7 g CO 2 eq./MJ hydrogen to 19 g CO 2 eq./MJ hydrogen, depending on the LCA framework used for the baseline scenarios. Depending on the improvement strategy considered (i.e., use of heat, distribution via gas network or renewable electricity), significant reductions were estimated. The higher emission reductions were obtained when electricity from wind power was used for the electrolysis pathways reducing the impact from hydrogen production with over 80%.
The findings of the study indicate that steam reforming of biomethane is the alternative with the lowest net emissions. This is mainly due to the low impact of biogas especially when waste and residues are used as feedstock and when coproducts from biogas production can be used to replace fossil alternatives (such as mineral fertilisers). Hydrogen produced via electrolysis is an energy-intensive alternative. It would be a more competitive alternative if there is a market potential for all product streams, including oxygen and heat and if the produced heat replaces a fossil-based heat mix. In accordance with other studies, 25-27 the fossilbased carbon intensity of the electricity mix used, especially in the case of the hydrogen production via electrolysis and the possibility to offset the coproducts obtained can have a strong influence on the results. In the case of the biomethane reforming systems, the upstream impacts during the production of biogas used in the process would determine the overall environmental performance of hydrogen.
Methodological variations as a result of the LCA framework used and corresponding underlying assumptions are shown to affect the results obtained. Different frameworks, however, can be applied to different contexts (internal environmental work or for policy making), thus understanding the differences and interpreting the results accordingly is of vital importance (Table 1).
T A B L E 1 System setups for H 2 production and distribution described and compared in this study. 10 300

Systems
Note: Four setups are distinguished, where "a" and "b" denote subcategories of each main case.