Life cycle assessment of hydrogen production pathways to support hydrogen decarbonization policies in a Canadian context

Abstract

Hydrogen is a promising energy carrier that is expected to play a crucial role in helping Canada achieve its net-zero target by 2050. However, reducing the ambiguity in regulatory frameworks is essential to incentivize and facilitate international trade in hydrogen. To this end, regulators must agree on quantification methodologies that consider life cycle boundaries, process descriptions, co-product allocation, conversion constants, and certification units. Several studies have highlighted the importance of life cycle assessment (LCA) as a standardized, relevant method for estimating the carbon footprint associated with hydrogen production and evaluating its environmental sustainability. As such, LCA-based certification schemes could help create a transparent hydrogen market. The aim of this study is to validate the proposed harmonized LCA-based methodology for quantifying hydrogen production’s carbon intensity. This methodology follows a consistent scope and life cycle inventory (LCI) development criteria, alongside a rigorous data quality assessment. The well-to-gate carbon intensities of six hydrogen production pathways are compared, which range from 0.26 to 10.07 kg CO_2e per kg of hydrogen (kg CO_2e/kg H₂), against the hydrogen carbon intensity thresholds established by the Canadian Clean Hydrogen Investment Tax Credit (CHITC). For example, the biomass gasification with carbon capture (CC) pathway demonstrates the lowest carbon intensity, while thermochemical pathways, such as steam methane reforming of natural gas without CC, poses challenges to meeting the maximum CHTIC threshold of 4 kg CO_2e/kg H₂.

Graphical abstract

Keywords

life cycle assessment (LCA) / hydrogen production pathways / carbon intensity / hydrogen standards / hydrogen certification scheme

1 Introduction

At the United Nations Framework Convention on Climate Change (UNFCCC) COP 26, held in November 2021 in Glasgow, UK [1], a series of sector-specific common goals, known as the Glasgow Breakthroughs, were established to address the five key economic sectors: power, road transport, steel, hydrogen, and agriculture [1]. The hydrogen-related goals aim to make renewable and low-carbon hydrogen affordable and globally available by 2030. In a recent Breakthrough Agenda Report, the International Energy Agency, the International Renewable Energy Agency (IRENA), and the UN Climate Change High-Level Champions [2] highlighted their efforts to develop international standards and associated certification schemes for renewable and low-carbon hydrogen to address key issues such as greenhouse gas (GHG) emissions accounting, safety, and operational concerns, including leakage, which are critical for supporting high-quality demand commitments and trade agreements.

Hydrogen policies are essential tools for implementing national hydrogen strategies, establishing a sustainable hydrogen economy, and supporting various stages of hydrogen development, such as technology development, market penetration, and market growth. For example, transitioning renewable hydrogen from a niche energy source to a widespread energy carrier requires a policy approach that incorporates national hydrogen strategies, clear policy priorities, guarantees of hydrogen origin, and enabling policies and regulations [3]. When produced with low carbon intensity, hydrogen can replace hydrocarbon-based fuels in various energy-intensive sectors, potentially leading to a significant reduction in GHG emissions. However, a key barrier to scaling up low-carbon hydrogen production is the ambiguity in regulatory frameworks and certification schemes, including the need for harmonized standards for environmental criteria and policies across jurisdictional borders to incentivize and facilitate international trade.

Although the common color hydrogen classifications, such as “green” or “blue” hydrogen, provide an overview of hydrogen pathways, they fail to establish a clear relationship between the color assignation and the specific emissions level [4]. Therefore, there is a pressing need to develop harmonized standards, regulations, and certification schemes that focus on the carbon intensity of hydrogen production. These frameworks will ensure transparency and enable reliable investment decisions, thereby supporting the tradability of hydrogen [4].

This study aims to conduct a life cycle environmental assessment (LCA) of hydrogen production pathways by applying the previously developed LCA-based methodology [5], including defining system boundaries and developing life cycle inventories in a Canadian context. This study evaluates various environmental impacts, including global warming potential, toxicity, land use, water consumption, terrestrial acidification, fine particulate matter formation, and ozone formation, while focusing on evaluating six hydrogen production pathways in Canada. These pathways include steam methane reforming (SMR) with and without carbon capture (CC), autothermal reforming (ATR) of natural gas with CC, alkaline (AE) and proton exchange membrane (PEM) electrolysis using grid electricity, and biomass gasification using wood chips with CC.

1.1 National hydrogen strategies for decarbonization

By the end of 2023, forty-one countries had established their hydrogen strategies [6,7]. In 2017, Japan announced the world’s first national hydrogen strategy, and since then, many other countries have developed their own national hydrogen strategies to accomplish decarbonization goals. One way to compare these national strategies is by evaluating the regulatory stringency regarding decarbonization objectives. It is important not only to set ambitious targets and define the scope of hydrogen strategy development [8] but also to understand the implementation of decarbonization strategies through regulatory measures such as penalties for fossil fuels, certifications, and the promotion of renewable hydrogen production [1].

The Hydrogen Strategy for Canada [9,10] outlines clear policies, rules, and standards in the context of hydrogen framework in Canada. These policies are designed to support the development and deployment of clean hydrogen technologies, helping Canada achieve its net-zero target by 2050. As part of decarbonization goals, the Canadian Hydrogen Strategy intends to reduce GHG emissions by 45 million metric tons per year by 2030. Moreover, Canada’s federal regulations and incentives encompass the clean hydrogen investment tax credit (CHITC), which provides tax incentives for clean hydrogen production. These credits range from 15% to 40% of the project costs, depending on the lifecycle GHG emissions of hydrogen produced from less than 0.75 to 4 kg CO_2e/kg H₂ [11,12]; the clean fuel regulations aimed at reducing carbon intensity in fuel production and use; and the CC, utilization, and storage (CCUS) tax credit, which incentivizes the use of CCUS technologies in hydrogen production.

Together, these initiatives are intended to further incentivize the adoption of clean hydrogen technology and support Canada’s goal of achieving net-zero emissions by 2050, as outlined in the federal government’s 2030 Emissions Reduction Plan [13].

1.2 Regulatory context implications for standard and certification development

Current regulatory frameworks for well-developed hydrogen markets and for incentivizing both supply and demand rely on standard methodologies and certification systems. The International Energy Agency (IEA)’s Global Hydrogen Review report [14] indicates that several countries have made progress in launching regulations and certification schemes. For instance, Australia plans to implement a Guarantee of Origin scheme in July 2025. The European Council defined low-carbon hydrogen in the Hydrogen and Decarbonized Gas Package in May 2024. The UK released an updated version of its Low-Carbon Hydrogen Standard in March 2024. In the United States, the rules for the Inflation Reduction Act (IRA) Clean Hydrogen Production Tax Credit were issued in December 2023, addressing incrementality, temporal matching, and deliverability [14].

At present, two standard development organizations, the Canadian Standards Association (CSA) and the Bureau de normalization du Québec (BNQ), are collaborating on a Bi-National Carbon Intensity of Hydrogen Standard aimed at quantifying and verifying hydrogen production carbon intensity. More information on the current hydrogen standard and certification schemes available worldwide can be found in a recent review paper [15].

However, the methodologies established for these certification schemes are not always consistent. While many methods share common elements, such as limiting emission intensity to hydrogen production, there are differences, particularly in system boundary definitions and the specific emission intensity thresholds required. Establishing a uniform approach to define hydrogen based on its emission intensity is essential for ensuring compatibility between different regulatory frameworks and certification systems [4]. Harmonized certification schemes, along with an internationally agreed-upon emissions accounting framework, can foster a competitive hydrogen commodity market based on transparency and consumer trust [16].

The International Partnership for Hydrogen and Fuel Cells in the Economy (IPHE) proposed a methodology for estimating the GHG emissions associated with hydrogen production [17,18]. This methodology formed the basis for the development of the International Organization for Standardization (ISO) standard ISO/TS 19870:2023, titled “Hydrogen Technologies—Methodology for Determining the Greenhouse Gas Emissions Associated with The Production, Conditioning and Transport of Hydrogen to Consumption Gate” [19].

These harmonized standard methodologies for calculating the LCA of GHG emissions from hydrogen production are crucial for establishing: thresholds for qualifying hydrogen as low-carbon, clean, or renewable (expressed in terms of kg CO_2e/kg H₂ or kg CO_2e/MJ H₂); standardized categorizations of economic activities, such as hydrogen production, as environmentally sustainable; unified standards to support global certification schemes for hydrogen; and consistent international definitions for hydrogen to be applied throughout the industry value chain to aid in the commercialization of hydrogen technologies and hydrogen-based products [16,20].

Carbon emission levels for renewable or clean hydrogen produced from renewable sources, as well as low-carbon hydrogen from unabated fossil fuels, should be defined and classified according to their carbon intensity thresholds [4].

Different hydrogen production pathways generate hydrogen at different pressures and purities, and distinct end uses require specific purity and pressure specifications. However, not all existing schemes establish requirements for pressure and purity in the final hydrogen product [21].

Furthermore, the primary function of hydrogen certification systems is to differentiate hydrogen types by disclosing the renewable or low-carbon source of hydrogen, based on its production method, life cycle carbon intensity, and other sustainability aspects, such as water, land, and rare earth metal consumption [16]. Certification is crucial for enhancing the traceability of embedded carbon emissions and facilitating cross-border trade in hydrogen [1].

The Appendix section (Table A1) compiles and summarizes a comparison of certification schemes from various countries and jurisdictions. In Section 3.4, this study presents a comparison of the results with international standards and models to highlight the main similarities and differences with hydrogen LCA model proposed in this study.

2 Methodology

This study performed a LCA of various hydrogen production technologies within a Canadian context to estimate their life cycle carbon intensity and evaluate their compliance to the Canadian CHITC by adapting and applying the LCA-based methodological framework developed by Gonzales-Calienes et al. [5], which is based on a harmonized approach to hydrogen carbon footprint methodologies from relevant and current international standards and certifications schemes [18,22–27] and follows the life cycle assessment (LCA) normative principles, procedures, and recommendations outlined by the International Organization for Standardization (ISO) [19,28–34]. The harmonized methodology includes the adoption of a “well-to-gate” system boundary, a functional unit of 1 kg of hydrogen specifying minimum purity levels, and common life cycle inventory (LCI) criteria.

Fig.1 illustrates the four stages of the LCA-based methodology used in this study, as outlined below:

Goal and scope definition: The first stage defines the product system (technical specifications and functional unit), system boundary (carbon accounting scope), technological scope (type of hydrogen production pathway), and geographical scope (feedstock and energy source location).

Data collection: Primary data is collected by generating foreground data via process modeling and simulation.

LCI and life cycle impact assessment (LCIA): This stage involves constructing a hydrogen LCI database, evaluating data quality, and calculating the carbon intensity of various hydrogen production pathways.

Data hub development: This stage focuses on creating a public repository of hydrogen LCI data sets and implementing an openLCA collaboration server.

In addition to estimating the carbon footprint of hydrogen production, this study also evaluated several other environmental impact categories to assess the overall environmental sustainability of the hydrogen production technologies, including water consumption and land use.

The following sections provide an overview of the goal and scope definition, the methods used for collecting LCI data, the evaluation of LCI data quality, and the selection of the impact assessment method.

2.1 Goal and scope

The goal of this LCA is to assess the environmental impact of climate change, specifically focusing on carbon footprints of different hydrogen production pathways in a Canadian context. This is achieved by quantifying GHG emissions, expressed in kg CO_2e, along all “well-to-gate” unit processes, in accordance with the ISO standards, 14040:2006, 14044:2006, 14067:2018, and 9870:2023.

2.1.1 Product system

The function under evaluation is the production of hydrogen in its gaseous form for various applications in Canada. The functional unit and reference flow, which serve as the basis for normalizing the input and output data, are defined as follows: 1 kg of compressed gaseous hydrogen with a purity greater than or equal to 99.9% (industrial grade), a compression pressure that corresponds to the inlet requirements of the subsequent stage, and an energy content (LHV) of 119.9 MJ/kg. The minimum default values are used for hydrogen purity and a compression pressure of 62.2 bar. The lower heating value (LHV) is considered as the conversion constant for hydrogen energy content.

The functional unit or reference flow for our product system is established as 1 kg of hydrogen produced, which is applied to six hydrogen technology pathways in a Canadian context.

Life cycle GHG emissions are estimated by dividing the total kilograms of carbon dioxide equivalents (kg CO_2e) by the total of kilograms of hydrogen produced within the system boundary.

2.1.2 System boundary

Establishing system boundaries has a significant effect on emission-reduction incentives and the balance of international tradeability [35]. In this study, the carbon accounting scope is defined by adopting a “well-to-gate” system boundary for the hydrogen production system [23,25–27].

This harmonized system boundary for hydrogen production, illustrated in Fig.2, consists of both foreground and background systems. The foreground system includes three subsystems:

Upstream processes: This subsystem encompasses the upstream processes, which range from the extraction or production of feedstock, its transportation via pipelines and trucks, as well as electricity generation from grid mixes and renewable or non-renewable energy sources. It also includes electricity transmission and distribution activities.

Hydrogen production and purification: This subsystem comprises the hydrogen technology-specific production process, the hydrogen purification process, and, where necessary, the CC unit process.

Hydrogen compression: This subsystem includes the hydrogen compression process.

The background system consists of additional input processes that support the foreground system, such as non-feedstock-related materials and energy unit processes.

This system boundary follows a “well-to-gate” approach, which covers Scope 1, Scope 2, and partial upstream Scope 3 emissions. The characterization of these three emissions scopes is defined by the World Resources Institute (WRI) [28].

Emissions associated with downstream Scope 3 activities, such as business travel, employee commuting, upstream leased assets, and embedded emissions in capital goods (including the construction, manufacturing, and decommissioning of hydrogen facilities and electrolyzers), are excluded from this assessment.

2.1.3 Technological scope

GHG emissions throughout the life cycle of these production methods are influenced by the technology and energy source employed [36]. In this study, the hydrogen technology pathways are limited to those that meet the conditions of the Clean Hydrogen Investment Tax Credit and are eligible under the specified technology pathways, as follows: SMR using natural gas without CC; SMR using natural gas with CC; ATR using natural gas with CC; Gasification using biomass with CC; AE electrolysis using grid electricity; and PEM electrolysis using grid electricity.

2.1.4 Geographical scope

This study focuses on Canadian provinces where certain hydrogen production technologies have significant potential, examining the geographical relationship with low-carbon-emission electricity. The emission factor of grid electricity varies depending on the types and proportions of energy sources used in each provincial grid.

2.2 LCI data set development

The LCIAs were conducted using the LCI data set, which is a compilation of input and output data that adheres to the ISO 14040/44:2006 standard [29]. According to the International Reference Life Cycle Data System (ILCD), either primary or secondary sources can be used to create LCI data sets [31]. For the foreground systems in this study, the LCI data sets were developed by using primary data sources, such as process engineering models. The background system was modeled using the EcoInvent© database, a secondary source.

Fig.3 outlines the steps involved in developing and implementing an LCI data set for use in the LCIA of a hydrogen production pathway.

2.2.1 Literature review: collection of secondary data

A scoping review of relevant publications related to the LCA of hydrogen production was conducted to identify complete and reliable LCI data sets, built from process simulation results and/or industry data.

The literature review follows three steps. In the first step, a screening literature review was conducted to select up-to-date publications on H₂ LCA studies of the technologies under assessment, focusing on those with similar system boundaries. An overview of the product system, initial system boundary, and functional unit were gathered from these publications for each hydrogen production pathway. In the second step, a second screening was conducted to review existing hydrogen production LCI data sets. After this review, the inventory lists for the same hydrogen production pathways were integrated to compile a potential inventory list of input/output flows and values. The average values of these LCI input/output flows were then calculated. The final step involved gathering data and information related to process simulation parameters and assumptions.

2.2.2 Process modeling and simulation: collection of primary data

Due to the lack of accessible primary data from industry, process simulation results were used as data sources to generate primary data. The primary data for the foreground LCI data sets was derived from the mass and energy balance obtained from the process simulation.

To collect the necessary raw data, the inventory list of input and output names compiled in the second screening of the literature review was used as a model for the flows of inputs and outputs. These names were then matched with the corresponding inputs and output values generated from the mass and energy balances of the process simulation, after performing the necessary calculations and providing supporting information [37].

2.2.3 Building the LCI database

The attributional LCI modeling approach is adopted to build the LCI.

For the foreground system, the LCI of unit processes was completed using the mass and energy balance results from the simulation and optimization processes. For the background system, aggregated data sets from the EcoInvent database were used. According to the ILCD, converting mass and energy balance results to unit process data requires proper scaling of inputs and outputs to the functional unit (i.e., 1 kg of hydrogen produced and documenting unit conversions and assumptions [32].

The hydrogen production LCI database was generated by compiling final LCI data sets created in the openLCA desktop environment and then uploading them to our openLCA collaboration server.

2.2.3.1 1) Interim data quality control

Once the preliminary data sets were created using the mass and energy balance results from process modeling and simulation, an initial data quality evaluation was conducted to provide feedback on the process simulations for any necessary adjustments to the process modeling operating parameters [5,32]. An interim quality control check was performed on the preliminary LCI data set to assess completeness, sensitivity, and consistency. The preliminary LCI data set values were compared with the average values from secondary sources. If missing values/data gaps or values out of average ranges were identified according to their relevance in the unit process inventory, they could be addressed by calculations/correlations and data from similar processes, such as average values from secondary sources collected in the literature review, expert estimates, and background databases.

2.2.3.2 2) Validation and data quality assessment

The final step in completing the LCI data sets is assuring that adequate data quality was used in the generation of process flow values. While a robust methodology is crucial for quantifying GHG emissions, even the most rigorous methodology can yield inaccurate results if incomplete or inaccurate data is used, leading to significant deviations from the actual emissions [4].

This study employed standard data quality indicators (DQI), including reliability, completeness, and time-related, geographical, and technological appropriateness, to evaluate the data quality of each flow value, based on the pedigree matrix method. In addition, the overall data quality of the process was also evaluated using completeness and review type [38,39]. For flow-related DQIs, each flow was assigned a score from 1 to 5, based on its alignment with the goal and scope and its representativeness, which generated a matrix. The complete matrix and criteria were available in Wernet et al. [38,39]. Similarly for process-based DQIs, a score of 1 to 5 was given for each process based on process completeness and type of review, as elucidated by Edelen et al. [38,39]. A score of 1 represents the highest data quality, while a score of 5 represents the lowest quality in both flow and process based DQIs.

The DQI scores for each LCI data set used in this study are provided in the Electronic Supplementary Material. Following the flow-based DQI evaluation, the following scores are assigned to the LCI data sets in this study:

• Reliability: Score 3—Represents non-verified data, partly based on estimates.

• Completeness: Score 5—Data from small sites, mainly because many plants are still prospective and do not exist in Canada at an industrial scale.

• Time representativeness: Score 1–3—The data set correlates to parameters and assumptions from the past 3–10 years.

• Geographical representativeness: Score 2–3—Most plant data are adjusted to Canadian average conditions (e.g., natural gas quality, water quality), thus representative of average data from larger area or similar production conditions.

• Technological representativeness: Score 1–2—The plant and process parameters are based on similar technologies processing the same feedstocks, and in some cases are validated with industry experts, hence a higher technological representativeness.

Following the process-based DQIs evaluation, the following scores are assigned to LCI data sets in this study:

• Process completeness: Score 1—More than 80% of input and output flows of the process are identified and quantified.

• Process review: Score 4—The process data sets are reviewed by an internal reviewer.

2.2.4 LCIA

The “well-to-gate” GHG emissions accounting in this study is illustrated in Fig.4 in accordance with relevant normative and standards and norms [28,33]. The total emissions within the system boundary (“well-to-gate”) include all Scopes 1 and 2 emissions, along with partial Scope 3 emissions.Net GHG emissions for each hydrogen production pathway are estimated by subtracting the carbon capture removal and the allocated coproduct emissions from the total emissions.

The emissions inventory for the hydrogen production pathways evaluated in this study primarily includes emissions related to combustion, energy supply (electricity), and embodied emissions from upstream processes. Emissions associated with electricity use do not account for emissions embedded in capital goods. Additionally, a cut-off criteria was applied in this study’s hydrogen LCA model to exclude input and output flows that contribute less than 1% of environmental impact [40]. For example, chemicals such as catalysts and solvents were excluded from the system boundary and LCI since they were estimated to represent only 0.02% to 0.05% of the total carbon intensity, according to Antonini et al. [41].

Furthermore, ISO/TS 19870:2023 [19] addresses allocation procedures for multifunctionality and coproducts, specifying the allocation procedures to apply in attributional or consequential LCA approaches. The allocation methods for attributional LCA include partitioning; physical allocation based on mass or energy content; and economic value allocation.

Given that the goal of this LCA study is to estimate the GHG emissions from hydrogen production pathways, climate change is the primary environmental impact category considered. This was done using the Intergovernmental Panel on Climate Change (IPCC) 2013 method, based on the IPCC’s Fifth Assessment Report (AR5) [42]. This method employs Global Warming Potential (GWP100) characterization factors for all GHGs with a GWP of zero for biogenic CO₂ and a value of 1 for carbon dioxide resulting from direct land use change.

The life impact assessment results are expressed in kgCO_2e/kg H₂.

2.3 Developing and implementation of an open-source LCA database of hydrogen production

Data quality plays a critical role in the development of LCI data sets, as it directly impacts the reliability of outcomes and helps avoid inconsistencies between actual and estimated emissions. Poor data quality can lead to discrepancies, where projects may appear compliant with regulations and support schemes despite not meeting the necessary standards, undermining the credibility and transparency of the system [4].

To this end, the Research and Technology Organization (RTO), the National Research Council of Canada, developed and implemented an open-source hydrogen LCA database, which hosted LCI data set repositories on the recently proposed data hub platform [43]. Built on open-source tools like openLCA, this platform enables the linkage of data flows across multiple hierarchical data sets, making them accessible to engineers, laboratory technicians, LCA practitioners, LCI data creators and decision makers. The following levels of data have been considered:

• Technology and scientific data sets: These include data from experiments, laboratory trials, modeling and simulation, pilot plants, and commercial plants.

• LCI data sets: These are derived from technology and scientific data sets.

• Aggregated data sets: These include environmental product declarations (EPDs), which provide LCA methodological data and impact assessment results.

• Proxy data: This includes data from reports and peer-review publications, such as carbon footprints and organizational emissions.

• Technology and policy data: These data sets are used for decision making processes.

The availability of multiple levels of data improves the transparency and traceability of decisions that can be made from LCA studies, ensuring more informed and reliable assessments.

3 Results and discussion

The life cycle GHG emissions of the hydrogen production technologies listed in Section 2.1.3 were estimated using the open-source LCA software openLCA version 2.2.0, developed by GreenDelta©.

3.1 Well-to-gate carbon intensity of hydrogen production pathways in Canada

The life cycle carbon intensities of the six hydrogen production pathways were estimated based on the LCA-based methodology described in Section 2. LCIA results were obtained by using openLCA 2.2.0 and LCI data sets, which were built from data collected via process modeling and simulation using Aspen© and ProSim© software. Detailed life cycle inventories for these hydrogen production pathways are provided in the Electronic Supplementary Material.

Tab.1 summarizes the hydrogen technologies evaluated, including the emission scopes that depend on the nature of the feedstock, energy supply, and emission inventory related to each hydrogen pathway. The variability in the estimated range of carbon intensities is influenced by factors such as the emission factor of the electricity grid, the type and chemical composition of the feedstock, the location of the energy supply, the process flow configuration, and the CC rate.

The well-to-gate carbon intensities of the six hydrogen production pathways are illustrated in Fig.5. The carbon intensities for alkaline and PEM electrolysis are very similar, with values of 1.23 and 1.34 kg CO_2e/kg H₂, respectively, assuming that the energy supply comes from electricity generated from the grid in the Province of Quebec, Canada. Emissions from the use of electricity in the Quebec electricity system include emissions from the Quebec electricity generation mix, electricity imports, and electricity transmission grid losses.

The life cycle carbon intensity for the ATR with CC was calculated as 4.44 kg CO_2e/kg H₂, with the main contributors being the ATR technology itself, the chemical composition of the natural gas in Alberta, Canada, and the carbon intensity of the Alberta electricity grid. This grid affects the emissions from electricity consumption in the air separation unit used to produce oxygen gas. In this case, no surplus electricity is generated.

For the biomass gasification with CC pathway, the carbon intensity was calculated to be 0.26 kg CO_2e/kg H₂ because of the chemical composition and carbon content of wood chips in Canada (average value) and the economic allocation of the co-product CO₂ liquid.

In the case of SMR without and with CC, the carbon intensities were calculated as 10.07 and 4.89 kg CO_2e/kg H₂, respectively. The carbon intensity of SMR without CC is heavily influenced by the SMR technology itself, the direct CO_2e emissions from the process, natural gas production and pretreatment, and electricity consumption. Moreover, the carbon intensity of SMR with CC relies on the CC rate, technology type, energy consumption of the CC unit, fugitive emissions, and the carbon content of the chemical composition of the natural gas in Alberta. In this design, amine capture was used, as it is the most mature and advanced CC technology available.

Furthermore, CO₂ (liquid) is a co-product of both the SMR and ATR pathways, and its allocation also influences the carbon intensity of the hydrogen produced. In these cases, economic allocation was applied to capture the environmental burdens associated with the CO₂ liquid and the hydrogen produced. In the case of SMR and ATR with CC and economic allocation, the carbon intensity was estimated at 2.02 and 1.01 kg CO_2e/kg H₂, respectively.

3.2 Policy implications

In the 2022 Fall Economic Statement [11], the Canadian federal government introduced a refundable investment tax credit for clean hydrogen production investments, with the credit amount based on the life cycle carbon intensity of the hydrogen. The CHITC, as detailed in Budget 2023 [12], provides economic support ranging from 15% to 40% of eligible project costs, with the highest tax credit allocated to projects that produce the cleanest hydrogen (Tab.2). The CHITC offers a refundable tax credit for investments in eligible equipment used in hydrogen production in Canada via electrolysis or natural gas (with emissions controlled using CC, utilization, and storage units).

This credit is applicable from March 28, 2023, and is scheduled to phase out by 2035. The CHITC is projected to provide $17.7 billion in tax incentives to the sector by 2035, based on the expected uptake of the credit, rather than a funding cap or target.

Based on the policy incentive of the CHITC, Fig.5 shows that alkaline and PEM electrolysis comply with the carbon intensity threshold of 0.75 to 2 kg CO_2e/kg H₂. These electrolysis-based projects would qualify for at least an investment tax credit of 25%, as the emissions from electricity grid represent the maximum carbon intensity level for electrolysis. In contrast, the ATR with CC and SMR pathways, both with and without CC, could not claim these incentives because their carbon intensities exceed the 4 kg CO_2e/kg H₂. However, emissions from these pathways can still be reduced via different CC technologies with lower energy consumption requirements compared to conventional monoethanolamine (MEA) and methyldiethanolamine (MDEA) absorption technologies. Moreover, economic allocation would be applied to the CO₂ liquid co-product and the hydrogen produced in the SMR with CC, ATR with CC, and biomass gasification with CC pathways. In this case, SMR with CC and ATR with CC would qualify for at least an investment tax credit of 15% and 25%, respectively. The biomass gasification with CC pathway could qualify for a maximum of 40% of the investment tax credit

It is important to consider uncertainties in the key assumptions and parameters when implementing incentives-based policies. As shown in Fig.5, error bars indicate that water electrolysis would qualify for the maximum 40% tax credit if it uses more renewable energy sources instead of grid electricity, but would receive less if uses a grid electricity with more fossil fuel contribution in the generation mix. For the carbon intensities close to the tiers maximum limit (4 kg CO_2e/kg H₂), such as SMR with CC and ATR with CC both without co-product allocation, allocation procedures should be employed by following the current ISO LCA standards.

3.3 Sensitivity analysis

A sensitivity analysis was conducted for alkaline electrolysis, PEM electrolysis, and autothermal reforming pathways to evaluate the impact of critical parameters on carbon intensity of hydrogen production. Tab.3 presents the parameters that were varied for each of the selected pathways.

The sensitivity of well-to-gate carbon intensity for the selected hydrogen production pathways is shown in Fig.6. For the PEM electrolyser, the specific electricity consumption, and thus the system efficiency, is the key metric determining the final calculated carbon intensity of hydrogen. The system efficiency of a PEM-based hydrogen production facility, measured in kWh/kg H₂ (lower heating value), depends on the individual efficiencies of the electrolyser cell/stack and the balance of plant components.

According to the 2020 IRENA report [44], the stack electrical efficiency of the state-of-art PEM electrolyzers ranged between 47–66 kWh/kg H₂ (LHV), depending on the electrolyser design. In addition, various balance of plant components, such as water treatment, purifiers, cooling, thermal management, and others, consume power and can contribute to hydrogen escape, affecting the overall system efficiency. A range of potential system electrical efficiency from 50–83 kWh/kg H₂ (LHV) as reported in International Renewable Energy Agency [44] was used in the sensitivity analysis.

The carbon intensity significantly increased from 1.34 to 2.06 kg CO₂/kg H₂ when the higher electricity consumption value of 83 kWh/kg H₂ was used, representing a 54% increase.

For the biomass gasification pathway, when the Alberta grid mix was used (based on EcoInvent 3.9 database background data sets), the carbon intensity increased to 2.12 kg CO₂/kg H₂, compared to the baseline case that used the Canadian average electricity mix. This increase is mainly attributed to the higher consumption of electricity in the gasification H₂ production process.

In the case of ATR, the carbon intensity increased to 5.64 kg CO₂/kg H₂ when the production was located in Quebec, despite the availability of low carbon electricity. This is due to the fact that Quebec does not have local natural gas supplies, requiring transportation over long distances, hence increasing the overall carbon footprint of hydrogen production. The Canadian electricity grid mix and natural gas supply chain background data sets used in this analysis were based on EcoInvent database version 3.9.

3.4 Comparison with other LCA studies

The well-to-gate carbon intensity results from this study are compared with two global LCA hydrogen frameworks: the UK Low Carbon Hydrogen Standard [27] and Argonne’s R&D GREET® (Greenhouse Gases, Regulated Emissions, and Energy Use in Technologies) model [45]. The former is a GHG methodology and conditions for standard compliance released in December 2023, while the latter provides in-depth life-cycle simulations for a variety of energy and chemical products, including hydrogen technologies.

The hydrogen LCA framework proposed in this study aligns with the previous frameworks in various methodological options such as the well-to-gate system boundary, the hydrogen gas product system with a purity more than 99.9 % (mole fraction), the IPCC 2013 method for impact assessment, inclusion of cut-off criteria, exclusion of emissions embedded from capital goods, and the use of ISO standards for building the LCI data set.

However, there are also some divergencies such as the functional unit. This study and the GREET model use 1 kg of hydrogen gas (LHV), while the UK Low Carbon Hydrogen Standard uses 1 MJ of hydrogen gas (LHV). The carbon intensity result for water electrolysis diverged due to the electricity generation mix composition of each country. However, the carbon intensity for SMR and ATR pathways has similar values in the three frameworks. In the case of biomass gasification with CC, this study shows a significant divergence from both the UK Low Carbon Hydrogen Standard and the GREET model. While the UK and GREET frameworks consider carbon credits from CC and sequestration, this study treats the CO₂ captured by liquefying it under high pressure to obtain CO₂ in liquid state, which is then commercialized.

3.5 Other environmental impacts

Net GHG emissions are the primary focus of this LCA study, which is crucial for developing hydrogen standards and certification schemes in Canada. However, the same LCA-based methodology employed for assessing climate change impact can also be adapted to quantify other impact factors, such as water consumption and land use.

The selected impact assessment method is the ReCiPe 2016 v1.13 midpoint method using the hierarchic version [47], which evaluates 18 environmental impact categories. Normalization plays a crucial role in evaluating the overall environmental footprint, as it helps identify the most significant impact categories. This process ensures that attention is focused on the most relevant aspects of the environmental impact, thereby facilitating effective communication and decision-making [48].

Using normalization factors, specifically the World (2010) H set, the following seven impact categories are identified for comparison purposes: human toxicity (noncarcinogenic and carcinogenic), freshwater ecotoxicity, water consumption, land use, terrestrial ecotoxicity, fine particulate matter formation, and ozone formation (human health).

Fig.7 shows the environmental impacts of the nine selected impact categories.

3.5.1 Human toxicity, freshwater, and terrestrial ecotoxicity

Toxicity was assessed for three environmental impact categories: human toxicity (both noncarcinogenic and carcinogenic), freshwater ecotoxicity, and terrestrial ecotoxicity. The characterization factors at the midpoint level were based on the impacts of chemical emissions expressed in kg 1,4-dichlorobenzene equivalents (1,4DCB-eq) [47].

Among the pathways assessed, ATR with the CC pathway presents the highest level of human toxicity (both noncarcinogenic and carcinogenic) because of waste treatment and disposal during grid electricity generation, which is used in the ATR hydrogen production process. The impacts related to waste management during electricity generation contribute significantly to this higher toxicity level.

Moreover, the alkaline electrolysis and PEM electrolysis pathways result in the highest levels of freshwater ecotoxicity and terrestrial ecotoxicity because of the large amount of wastewater discharged during the electrolysis process, which require further treatment and contribute to the toxicity in water and soil ecosystems.

3.5.2 Land use

The midpoint characterization factor for land use impact is expressed in units of area occupied or transformed over a specific time period, often measured in square meters per year (m²·yr), which represents the area of land required to support various land use types such as pasture, permanent crops, mosaic farming, forestry, urban land, or annual crops and reflects the loss of species associated with that land use type [47].

Among the pathways evaluated, biomass gasification with CC presents the highest level of land use impact due to the use of softwood and hardwood forestry for biomass feedstock, which leads to significant land conversion and habitat disruption, resulting in the highest land use impact compared to other hydrogen production pathways.

3.5.3 Water consumption

The GWP100 for the environmental impact category water consumption is a characterization factor at the midpoint level, expressed in m³ of water consumed per m³ of water extracted [47].

Among the hydrogen production pathways evaluated, the alkaline and PEM electrolysis processes exhibit higher water consumption compared to thermochemical pathways. This is due to the substantial water requirements for the electrolysis reactions, where water is split into hydrogen and oxygen. In contrast, thermochemical processes like SMR and ATR typically consume less water, mainly for cooling and other ancillary processes.

3.5.4 Fine particulate matter formation and ozone formation (human health)

For the midpoint characterization factors of fine particulate matter formation, the human population intake of PM_2.5 was considered. The particulate matter formation potential (PMFP) is expressed in kg primary PM_2.5-eq. Similarly, the amount of ozone that people breathe in was used as a midpoint characterization factor for the environmental impact category called “photochemical ozone formation related to human exposure,” with the human health ozone formation potential (HOFP) is expressed in kg NO_x-equation [47].

The thermochemical pathways, e.g., biomass gasification, ATR, and SMR, result in high levels of PM_2.5e and NO_{x equivalents} emissions, the latter refers to the combined amount of nitrogen oxides (NO_x), specifically nitric oxide (NO) and nitrogen dioxide (NO₂), due to the variations in ambient concentration of ozone after the release of precursor compounds like nitric oxide, which is part of the flue gas stream in these processes.

Overall, trade-offs between environmental impact categories can provide a balanced perspective of the environmental impact for each hydrogen production pathway. For example, water electrolysis pathways can show low carbon intensities ranging from 0.19 to 2.06 kg CO_2e/kg H₂ (Fig.7) depending on the electricity consumption from renewable energy (wind electricity) or grid electricity. However, while water electrolysis offers low carbon emissions, it consumes a significant amount of water (about 1.1 m³ of water per kg H₂), which could strain water resources. Additionally, toxicity impacts on human health, freshwater ecosystems, and terrestrial environments may increase if emissions from capital goods are included in the life cycle inventory.

In the case of the biomass gasification with CC pathway, the carbon insensitive can range from 0.26 to 2.12 kg CO_2e/kg H₂ (Fig.7), depending on the share of renewable energy in the grid electricity mix. However, this pathway shows relatively high land use impacts compared to other pathways, primarily due to the significant space required for biomass cultivation. Meanwhile, using biomass for bioenergy production is relatively environmentally friendly, large-scale facilities would require significant land use that could lead to reforestation issues if not managed sustainably.

3.6 Conclusion and further work

This study performed a LCA of six hydrogen production technologies in a Canadian context by employing a carbon intensity quantification methodology aligned with harmonized LCA methodological options that allowed comparability among the six different hydrogen technology pathways and addressed the challenge of building life cycle inventories with a high level of data quality.

The findings demonstrate that the technology type, feedstock, and energy source have a significant influence on the well-to-gate GWP impact variability on the six hydrogen pathways studied. The well-to-gate carbon intensities of 6 hydrogen production pathways range from 0.26 to 10.07 kg of CO_2-eq per kg of H_2, which corresponds to the gasification and SMR processes, respectively. These results were compared against hydrogen carbon intensity thresholds established by the Canadian Clean Hydrogen Investment Tax Credit. The robustness of the LCA-based methodology, combined with the completeness, reliability, transparency, and accuracy of the data used for building inventories and estimating the GHG environmental impact of the hydrogen technology under assessment, will be critical in ensuring compliance with governmental incentives designed to promote investments in clean hydrogen technologies and contribute to the development of a sustainable hydrogen market.

Besides the GWP, seven other environmental impact categories were also estimated to provide a comprehensive evaluation of the overall environmental impacts of each hydrogen production technology. Both SMR and ATR thermochemical pathways without CC technology exceeded the 4 kg CO_2e/kg H₂ threshold, with the ATR pathway having the highest human toxicity impact due to waste treatment during grid electricity generation. On the other hand, biomass gasification with CC technology has the lowest carbon intensity (0.26 kg CO_2e/kg H₂) but posed significant land use impacts compared to the other pathways. In the same way, both the alkaline and PEM electrolysis pathways present carbon intensities below 2 kg CO_2e/kg H₂. However, they resulted in water higher consumption and have negative effects on freshwater and terrestrial ecosystems.

While this study primarily focuses on the environmental and economic impact of hydrogen production, it highlights the importance of integrating social indicators alongside environmental impact categories for a holistic and transparent sustainability assessment of the hydrogen economy.

Although this hydrogen well-to-gate LCA model is used to estimate the carbon intensity of hydrogen technologies eligible for the Clean Hydrogen Investment Tax Credit, this model is currently being expanded to include other potential technologies such as methane pyrolysis or applications like ammonia production, both in a Canadian and an international context considering various conditions apart from system boundary, technology scope, and LCI data sets, such as data quality, electricity sources, feedstock compositions, and carbon intensity thresholds, aiming to define a harmonized international definition of hydrogen.

Furthermore, to comprehensively evaluate the LCA of the whole hydrogen supply chain, the system boundary should be expanded to include downstream processes. A “well-to-tank” boundary (at the point of use) would encompass hydrogen conditioning, conversion (e.g., liquefaction, ammonia production, and storage), and delivery (e.g., hydrogen transportation via pipelines and trucks).

4 Appendix A

Table A1 shows a comparison of the most important global hydrogen certification programs, origin guarantees, and standards based on an up-to-date literature review. The following table provides a summary comparing 6 certification initiatives and proposed hydrogen standards.

References

[1] Cheng W, Lee S. How green are the national hydrogen strategies. Sustainability, 2022, 14(3): 1930

[2] InternationalEnergy Agency. Accelerating sector transitions through stronger international collaboration. The Breakthrough Agenda Report, 2023

[3] InternationalRenewable Energy Agency. Green hydrogen: A guide to policy making. 2020, available at the website of IRENA

[4] InternationalEnergy Agency. Towards hydrogen definitions based on their emissions intensity. 2023, available at the website of IEA

[5] Gonzales-CalienesG, KannangaraM, YangJ J, . Life cycle assessment of hydrogen production pathways in Canada. 2022, available at the website of Government of Canada

[6] International Energy Agency/International Renewable Energy Agency. Policies database. 2023, available at the website of IEA

[7] InternationalEnergy Agency. Global Hydrogen Review 2023. 2023, available at the website of IEA

[8] Chaube A, Chapman A, Shigetomi Y. . The role of hydrogen in achieving long term Japanese energy system goals. Energies, 2020, 13(17): 4539

[9] NaturalResources Canada. Hydrogen strategy for Canada—Seizing the opportunities for hydrogen. 2020, available at the website of Government of Canada

[10] NaturalResources Canada. Hydrogen strategy for Canada: Progress report. 2024, available at the website of Government of Canada

[11] Governmentof Canada. Fall economic statement. 2022, available at the website of Government of Canada

[12] Governmentof Canada. Budget 2023-tax measures: Supplementary information. 2023, available at the website of Government of Canada

[13] EnvironmentChange CanadaClimate. Canada 2030 emissions reduction plan: Canada’s next steps for clean air and a strong economy. 2022, available at the website of Government of Canada

[14] InternationalEnergy Agency. Global hydrogen review 2024. 2024, available at the website of IEA

[15] Gale F, Goodwin D, Lovell H. . Renewable hydrogen standards, certifications, and labels: A state-of-the-art review from a sustainability systems governance perspective. International Journal of Hydrogen Energy, 2024, 59: 654–667

[16] Hydrogen Council. Policy toolbox for low carbon and renewable hydrogen. 2021, available at the website of Hydrogen Council

[17] InternationalPartnership for HydrogenFuelCells in the Economy. Methodology for determining the greenhouse gas emissions associated with the production of Hydrogen. 2021, available at the website of IPHE

[18] InternationalPartnership for HydrogenFuelCells in the Economy. Methodology for determining the greenhouse gas emissions associated with the production of hydrogen. 2022, available at the website of IPHE

[19] InternationalOrganization for Standardization. ISO/TS 19870:2023 “Hydrogen technologies—Methodology for determining the greenhouse gas emissions associated with the production, conditioning and transport of hydrogen to consumption gate”. 2023, available at the website of ISO

[20] Hydrogen Council. Hydrogen decarbonisation pathways—A life-cycle assessment Hydrogen Council. 2021, available at the website of Hydrogen Council

[21] Gonçalves Dias Ponzi G, dos Santos V H J M, de Medeiros Engelmann P. . The hydrogen life cycle assessment methodology: An overlooked puzzle piece in harmonizing hydrogen certification and trade. Clean Technologies and Environmental Policy, 2024, 26(8): 2573–2596

[22] Departmentof Industry , Science, Energy and Resources, Australian Government.Discussion paper—A hydrogen guarantee of origin scheme for Australia . 2021, available at the website of Australian Government

[23] CertifHy. CertifHy Scheme. 2022, available at the website of Certifhy

[24] Officeof Fossil EnergyUSDepartment of Energy. Clean hydrogen production standard (CHPS) draft guidance. 2022, available at the website of US Department of Energy

[25] TÜV-Rheinland. Standard H2.21 renewable and low-carbon hydrogen fuels. Version 2.1. 2023

[26] TÜV-SÜD. Production of green hydrogen- TÜV SÜD Standard CMS 70. 2021

[27] Departmentfor Energy Security & Net Zero. UK low carbon hydrogen standard—Guidance on the greenhouse gas emissions and sustainability criteria. Version 2. 2023

[28] WorldResources Institute. The greenhouse gas protocol. 2004, available at the website of Greenhouse Gas Protocol

[29] ISO 14040:2006. Environmental management—Life cycle assessment—Principles and framework. 2006, available at the website of ISO

[30] ISO 14044:2006. Environmental management—Life cycle assessment—Requirements and guidelines 2006, available at the website of ISO

[31] EuropeanCommission ‒ Joint Research Centre ‒ Institute for EnvironmentSustainability: International Reference Life Cycle Data System. Handbook—Specific Guide for Life Cycle Inventory Datasets. 2010, available at the website of European Commission

[32] EuropeanCommission ‒ Joint Research Centre ‒ Institute for EnvironmentSustainability: International Reference Life Cycle Data System. Handbook—General Guide for Life Cycle Assessment—Detailed Guidance. 2010, available at the website of European Commission

[33] ISO 14067:2018. Greenhouse gases—Carbon footprint of products—Requirements and guidelines for quantification. 2018, available at the website of ISO

[34] ISO 14044:2006. Amd2:2020. Environmental management—Life cycle assessment—Requirements and guidelines. 2020, available at the website of ISO

[35] White L V, Fazeli R, Cheng W T. . Towards emissions certification systems for international trade in hydrogen: The policy challenge of defining boundaries for emissions accounting. Energy, 2021, 215: 119139

[36] European Commision. A hydrogen strategy for a climate-neutral Europe. 2020, available at the website of European Union

[37] SonnemannG, Vigon B.Global guidance principles for Life Cycle Assessment (LCA) databases: A basis for greener processes and products . 2011, available at the website of UN Environment Programme (UNEP)

[38] EdelenA, Ingwersen W.Guidance on data quality assessment for life cycle inventory data . 2016, available at the website of US Environmental Protection Agency

[39] Wernet G, Bauer C, Steubing B. . The ecoinvent database version 3 (part I): Overview and methodology. International Journal of Life Cycle Assessment, 2016, 21: 1218–1230

[40] EnvironmentChange CanadaClimate. Fuel life cycle assessment model methodology. 2021, available at the website of Government of Canada

[41] AntoniniCTreyer KStrebA, . Hydrogen production from natural gas and biomethane with carbon capture and storage—A techno-environmental analysis. 2020, 6: 2967–2986

[42] MyhreGShindell DBréonF-M, . 2013: Anthropogenic and natural radiative forcing. In: Stocker T F, Qin D, Plattner G K, eds. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the 5th Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge and New York: Cambridge University Press, 2013

[43] Zargar S, Kannangara M, Gonzales-Calienes G. . Data hub for life cycle assessment of climate change solutions—Hydrogen case study. Data, 2024, 9(11): 129

[44] InternationalRenewable Energy Agency. Green hydrogen cost reduction: Scaling up electrolysers to meet the 1.5 °C climate goal. 2020, available at the website of IRENA

[45] ElgowainyA, Vyawahare P, Ng C, et al. Environmental life-cycle analysis of hydrogen technology pathways in the United States. Frontiers in Energy Research, 2024

[46] Departmentfor Energy SecurityNetZero. Hydrogen emissions calculator for the UK low carbon hydrogen standard: Full Calculator for all eligible pathways. 2022, available at the website of Government of UK

[47] Huijbregts M A J, Steinmann Z J N, Elshout P M F. . ReCiPe2016: A harmonised life cycle impact assessment method at midpoint and endpoint level. International Journal of Life Cycle Assessment, 2017, 22(2): 138–147

[48] SalaS, Crenna E, SecchiM, . Global normalisation factors for the environmental footprint and life cycle assessment. JRC Technical Report, 2017

[49] SindhavK K, Verma A KV, JainA, .International hydrogen strategies: Opportunities for India . In: Proceedings of 2023 IEEE IAS Global Conference on Renewable Energy and Hydrogen Technologies (GlobConHT), 2023, Maldives

[50] Liu W, Wan Y M, Xiong Y L. . Green hydrogen standard in China: Standard and evaluation of low-carbon hydrogen, clean hydrogen, and renewable hydrogen. International Journal of Hydrogen Energy, 2022, 47(58): 24584–24591

RIGHTS & PERMISSIONS

The Author(s), corrected publication. This article is published with open access at link.springer.com and journal.hep. com.cn