Assessing the Hydrogen Supply Chain from an Eco-Design Perspective: Challenges and Opportunities for Sustainability

Abstract

Hydrogen (H₂) is widely regarded as a key energy vector for the transition to a low-carbon energy economy, yet its production, storage, and distribution remain associated with significant environmental burdens. This review analyses the entire low-emission H₂ value chain from an eco-design perspective, synthesising recent life cycle assessment (LCA) literature on water-splitting technologies (electrolysis, thermochemical cycles, and photocatalysis), biomass-based thermochemical and biological routes, and the emerging exploitation of natural (geological) hydrogen. Environmental performance is benchmarked against conventional steam methane reforming and coal gasification across multiple indicators, including global warming potential (GWP), terrestrial acidification, water scarcity, mineral resource scarcity, and human toxicity. The results demonstrate that no single pathway is universally sustainable. Electrolysis remains strongly dependent on the electricity grid mix, while biomass gasification coupled with carbon capture and storage (CCS) achieves deep carbon mitigation but exacerbates acidification concerns. Furthermore, photocatalysis and biological routes show theoretically low carbon footprints but remain severely constrained by low technology readiness levels (TRLs). Downstream, physical compression showed lower life cycle burdens than chemical carriers, and pipeline networks outperformed other transportation modes, in the studies reviewed; both findings are conditional on the distance, scale, pressure and utilisation assumptions adopted. Finally, the integrated economic overview reveals that environmental hotspots correlate directly to financial penalties, ultimately dictating the levelised cost of hydrogen (LCOH). This review consolidates critical technological gaps, highlights the necessity for harmonised LCA boundaries, multi-indicator reporting, and primary industrial data, and proposes strategic research directions to enable a truly sustainable H₂ economy where eco-design choices are matched to localised energy, water, and material realities.

1. Introduction

In the last few years, the global energy landscape has undergone significant shifts driven by the urgent need to mitigate climate change and reduce reliance on fossil fuels. The ambitious environmental commitments of the Paris Agreement [1] have accelerated the search for cleaner alternatives to replace non-renewable sources and avoid pollution. Among the various alternatives being explored, hydrogen stands out as a versatile and clean energy carrier with the potential to play a pivotal role in the transition to a sustainable energy future [2,3].

Global hydrogen (H₂) demand exceeded 97 million tonnes (Mt) in 2023 [4] and reached almost 100 Mt in 2024 [5], gaining worldwide attention. In response, several nations have developed H₂ roadmaps aimed at establishing and implementing H₂ projects. As of May 2024, approximately 1572 low-emission H₂ projects larger than 1 MW, including production from renewable sources and fossil fuels with carbon capture and storage (CCS), have been announced globally, representing a total investment of USD 680 billion through 2030 [6]. Around 28% of these projects (434) have reached the Final Investment Decision (FID) stage. According to the Global Hydrogen Review of 2024 [4], low-emission H₂ production could reach up to 49 Mt by 2030—a 30% increase compared to 2023. However, projects that have reached the FID stage or are under construction account for only 3.4 Mt of this total, with electrolysis representing 1.9 Mt (55%) and 1.5 Mt (45%) coming from fossil fuels with CCS. While demand for low-emission H₂ grew almost 10% in 2023, global production still relies on unabated fossil fuels. Natural gas accounts for approximately 60% of production, coal gasification for 20%, H₂ as a by-product from refineries and petrochemical industries for 15%, and less than 1% comes from low-emission production [5,7]. This indicates that renewable-based H₂ technologies are not yet widely adopted and have not been fully integrated into the global supply chain. As a result, significant efforts are being made to research and implement processes and technologies that enhance the efficiency and economic viability of low-emission H₂ production, making it more competitive with fossil-based methods. According to the Global Hydrogen Review 2024 [4], the cost of producing hydrogen from natural gas ranges from USD 0.8 to 5.7 per kg of H₂ and can exceed USD 6 per kg when coupled to CCS. In Europe, the European Hydrogen Observatory [8] reported that, by 2023, the average cost of H₂ production via steam methane reforming (SMR) was 3.76 EUR/kg H₂ (4.38 USD/kg H₂), increasing to 4.41 EUR/kg H₂ (5.13 USD/kg H₂) when integrated with CCS. H₂ produced via electrolysis using electricity from the grid ranged between 4.06 and 17.4 EUR/kg H₂ (4.73 and 20.25 USD/kg H₂), with an average cost of 7.94 EUR/kg H₂ (9.24 USD/kg H₂). In comparison, electrolysis powered by renewable energy sources had a narrower cost range of 4.13 to 9.30 EUR/kg H₂ (4.81 to 10.8 USD/kg H₂), with an average of 6.61 EUR/kg H₂ (7.69 USD/kg H₂) (EUR-USD conversion at 1.16 USD/EUR). A graphical representation of H₂ costs for the year 2023 can be seen in Figure 1.

Figure 1. Levelised cost of hydrogen ranges for fossil-based routes and water electrolysis for the year 2023. Diamonds represent average values reported by the European Hydrogen Observatory [8].

As the demand for low-emission H₂ grows, understanding the economic and technical challenges associated with its production becomes crucial. Prices are expected to decline as technologies advance, which could accelerate the adoption and implementation of low-emission H₂ projects.

In this article, low-emission H₂ refers to hydrogen produced through electrolysis powered by renewable energy, thermochemical water splitting, biomass conversion or biological processes. Although the combustion of H₂ produces only water, offering clear environmental benefits, low-emission H₂ technologies still face several challenges; these include the material demands for production equipment, low process efficiencies, and consumption of other natural resources, such as water. Furthermore, environmental concerns are emerging across other stages of the H₂ value chain, particularly regarding the energy-intensive systems required for storing compressed H₂. This review examines H₂ production and supply issues through the lens of key eco-design principles. These principles are based on different documents, including ISO standards (ISO 14006) [9], the European Eco-design Directive, and the Design for Environment (DfE) guidelines developed by Telenko and collaborators [10], which are listed and defined below.

• Resource Efficiency: optimising the use of raw materials in order to reduce waste and consumption, preferably selecting materials that are abundant and recyclable.
• Energy Efficiency: designing products and processes that consume less energy during their manufacture, use and disposal, involving energy-efficient technologies and renewable energy sources.
• Minimisation of Environmental Impacts: minimising pollution and emissions associated with manufacturing processes and product use. The quantification of the environmental impacts through life cycle assessment (LCA) is preferred.
• Design for Durability and Recyclability: creating products that have a long lifespan and whose materials can be recycled at the end of their life, ensuring the recovery of valuable materials and minimising non-recyclable ones.
• Life Cycle Thinking: considering the environmental impacts of a product system at every stage of its life cycle, from raw material extraction, production, distribution, use and end-of-life disposal.

The aim of this work is to analyse the entire H₂ value chain from a sustainability perspective through a state-of-the-art literature review, scrutinising the environmental and energy hotspots across production, storage, and distribution, and identifying areas where the ecological performance of the system could be improved. Since low-emission H₂ production is still in its early stages, this article also seeks to highlight the methodological and technological gaps that currently constrain its sustainability assessment, providing guidance for future research.

Building on this framework, the present review pursues four interconnected objectives:

• To synthesize the state of the art of low-emission H₂ production pathways, including water-splitting technologies, such as electrolysis and thermochemical water-splitting cycles, photocatalytic systems, biomass-based thermochemical and biological routes, and the emerging exploitation of natural (geological) hydrogen.
• To benchmark the environmental performance of these pathways against conventional fossil-based production: steam methane reforming (SMR) and coal gasification (CG), through a critical reading of recent life cycle assessment (LCA) studies, considering not only global warming potential but also acidification, water scarcity, mineral resource scarcity, and human toxicity indicators.
• To identify the eco-design hotspots and trade-offs along the entire H₂ value chain, from production and storage to transport and distribution, and to evaluate how these environmental burdens couple with the economic competitiveness of each technology.
• To consolidate the methodological and technological knowledge gaps that currently limit the comparability and reliability of sustainability assessments of low-emission H₂ and to propose research directions aligned with the eco-design principles described above.

These objectives are addressed, respectively, in the production sections (electrolysis, thermochemical, photocatalytic, biomass, and natural hydrogen—Section 3), the comparative analysis and eco-design trade-off section (Section 4), the storage (Section 5), transport (Section 6), and economic assessments (Section 7), and the concluding research-gaps synthesis (Section 8).

Following the Introduction, Section 2 describes the methodology employed for this review. Section 3 investigates current H₂ production processes, focusing on low-emission methods derived from renewable energy and resources. These pathways are then comparatively analysed through an eco-design trade-off lens in Section 4. Subsequently, the focus shifts downstream, examining H₂ storage methods in Section 5 and transport and distribution alternatives in Section 6. An integrated economic analysis is then presented in Section 7 to contextualise these technological findings. Finally, Section 8 consolidates the identified research gaps and proposes future directions, culminating with concluding remarks in the final section. For each phase of the value chain, the discussion is structured around key environmental issues identified from scientific studies published in the literature.

2. Methodology

This work is a narrative review supported by structured literature searches, appropriate to the multidisciplinary scope of the hydrogen value chain. The review adopts a systematic search-and-screening procedure to ensure that the most relevant recent literature is captured. The literature search was conducted using the AI-powered platforms Discovery and Semantic Scholar, together with Google Scholar. The primary search window was restricted to the period of 2020–2025 to capture the most current developments in low-emission H₂ technologies. In addition to the 2020–2025 core literature, a limited number of studies before 2020 were retained when the study provided an established benchmark for comparison, particularly for conventional fossil-based pathways (e.g., SMR and CG) and biological processes. These earlier references are clearly identifiable in the corresponding tables and discussions.

For each section of the review, dedicated keyword combinations were applied:

• For Production Processes: “Green Hydrogen Production” OR “Low-emission Hydrogen Production” AND “Review”/”Life Cycle Assessment”/”Sustainability”.

It is worth clarifying that in this article the term “low-emission hydrogen” is preferred over “green hydrogen”, since no universal consensus currently exists regarding hydrogen colour nomenclature. The term “green” was used in the database queries to ensure full coverage of the relevant literature.

• For Storage: “Hydrogen Storage” AND “Review”/”Life Cycle Assessment”/”Sustainability”.
• For Transportation and Distribution: “Hydrogen Transportation” AND “Review”/”Life Cycle Assessment”/”Sustainability”.

Because this work is a narrative review rather than a registered systematic review, the screening process was not recorded prospectively against a PRISMA protocol. The figures reported below were reconstructed from the search history and are intended to convey the scale of the screening effort and the criteria applied. The initial database queries returned on the order of hundreds of thousands of records (10⁵) across all queries and databases, before filtering. Records were then screened to retain peer-reviewed research published in English between 2020 and 2025, sorted by influence and relevance, which reduced the pool to approximately 2000 records. Subsequently, titles and abstracts were screened for direct relevance to the eco-design assessment of the H₂ value chain, prioritising studies reporting quantitative LCA results, energy efficiency data, or eco-design considerations; this reduced the pool to approximately 275 records.

Studies were included if they (a) addressed hydrogen production, storage, transport, or economic assessment; (b) provided quantitative environmental, energy, or economic data suitable for benchmarking; (c) reported in LCA studies transparent system boundaries and functional units; and (d) were published in peer-reviewed journals or internationally recognised institutions.

Studies were excluded if they (a) focused exclusively on fossil-based hydrogen without any low-emission comparison; (b) presented only qualitative data without quantitative environmental or economic indicators; (c) addressed fuel cell applications outside the scope of this review; (d) were duplicate publications or superseded by more recent versions; or (e) were conference abstracts, preprints, or non-peer-reviewed sources without institutional backing.

Full-text assessment against the inclusion and exclusion criteria detailed above led to the exclusion of a further 142 records, most frequently because system boundaries or functional units were not reported transparently enough to allow the results to be contextualised or because the study had been superseded by a more recent version by the same authors. The remaining 112 peer-reviewed studies (110 journal articles and 2 book chapters) are cited in this review. In addition, 21 technical reports, standards and policy documents from international agencies and institutions (IEA, IRENA, European Commission, Clean Hydrogen Partnership, and others) were consulted directly to ensure coverage of the most recent industrial and policy developments; these were not subject to the screening procedure described above and are cited separately.

3. Low-Emission Hydrogen Production Processes

This first part of the article will discuss the technologies and processes for H₂ production. They were divided into water-splitting technologies, which comprise electrolysis, thermochemical water splitting and photocatalysis; biomass conversion technologies, which include thermochemical and biological processes; and, lastly, natural hydrogen production. The detailed technical description of each H₂ production technology has already been extensively covered by previous review articles cited throughout this work [11,12,13,14,15,16] among others available in the literature. Therefore, this review provides a concise overview of each production pathway, focusing the main discussion on the key sustainability and eco-design aspects rather than on technological fundamentals.

3.1. Water Splitting

Water splitting is the process in which the water molecule (H₂O) is separated into its principal components hydrogen (H₂) and oxygen (O₂) by applying energy to break the chemical bonds between H₂ and oxygen (O₂). To produce gaseous H₂ from water diverse energy sources can be used, such as electricity (used in electrolysis), thermal energy (used in thermochemical water splitting) or photonic energy (used in photocatalysis) [12,17]. This determines the water-splitting pathway; electrolysis, thermochemical water splitting and photocatalysis are discussed in this section.

3.1.1. Electrolysis

Electrolysis is an electrochemical process that converts electrical energy into chemical energy by splitting water into its constituent molecules, H₂ and O₂ [12]. Chemical reactions take place at the electrodes (cathode and anode), in which H₂ and O₂ are produced, and a separator or membrane keeps the two products apart. Fundamentally, these components—electrodes, electrolyte, and separator—constitute the electrolytic cell stack, while the assembly together with the frame and interconnections composes the electrolyser.

Different technologies have been developed to perform electrolysis, namely: alkaline electrolysis (AEL), proton exchange membrane electrolysis (PEMEL), solid-oxide electrolysis (SOEL) and, more recently, anion exchange membrane electrolysis (AEMEL). These technologies differ from one another mainly by the electrolyte. The electrolyte is the medium that permits the transport of cations and anions from and to the electrodes [18]. Table 1 displays the reactions that take place at the cathode and anode, the materials used for the electrolyte and electrodes, and the performance parameters for the mentioned electrolysis technologies, complemented by Figure 2. In Table 2, the advantages and disadvantages for each electrolysis technology are listed.

Table 1. Operational parameters for different electrolysis technologies.

Parameter	AEL	PEMEL	AEMEL	SOEL	Ref.
Overall Reaction	${\mathrm{H}}_2\mathrm{O}\to {\mathrm{H}}_2+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2$				[19]
Anode Reaction	$2{\mathrm{O}\mathrm{H}}^{-}\to {\mathrm{H}}_2\mathrm{O}+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2+2{\mathrm{e}}^{-}$	${\mathrm{H}}_2\mathrm{O}\to 2{\mathrm{H}}^{+}+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2+2{\mathrm{e}}^{-}$	$2{\mathrm{O}\mathrm{H}}^{-}\to {\mathrm{H}}_2\mathrm{O}+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2+2{\mathrm{e}}^{-}$	${\mathrm{O}}^{2-}\to {\displaystyle \frac{1}{2}}{\mathrm{O}}_2+2{\mathrm{e}}^{-}$
Cathode Reaction	$2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2+2{\mathrm{O}\mathrm{H}}^{-}$	${2\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2$	$2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2+2{\mathrm{O}\mathrm{H}}^{-}$	${\mathrm{H}}_2\mathrm{O}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2+{\mathrm{O}}^{2-}$
Electrolyte	Alkaline solution: NaOH/KOH	Solid polymer electrolyte: perfluorosulfonic acid (PFSA)	Quaternary ammonium polysulfone (QAPS), distilled water or low-concentration alkaline solution	Dense ceramic material: yttria-stabilised zirconia (YSZ)	[20,21,22]
Electrode/Catalyst (Hydrogen side)	Nickel-coated perforated stainless steel	Platinum nanoparticles on carbon black	Large-surface-area nickel	Nickel/YSZ	[18]
Electrode/Catalyst (Oxygen side)	Nickel-coated perforated stainless steel	Iridium oxide	Large-surface-area nickel or NiFeCo alloys	Perovskite-type
Operating Temperature (°C)	40–90	20–100	50–70	500–1000	[18]
Operating Pressure (bar)	1–30	30–80	30–35	<30
Current Density (A/cm2)	0.2–0.8	1.0–2.0	0.2–2.0	0.3–1.0
Cell Voltage (V)	1.8–2.4	1.8–2.2	1.8–2.2	0.95–1.3
Electrical Efficiency, Stack (kWh/kg H2)	47–66	47–66	51–66	35–50
Hydrogen Production Max. (Nm3/h)	1–500	1–250	-	~1	[23]
Gas Purity	>99.5	>99.9999	>99.99	>99.9	[24]
Stack Lifetime (h)	60,000–100,000	50,000–80,000	>5000	40,000
Technology Readiness Level (TRL)	Mature and commercial: 8–9	Mature and commercial: 8–9	Demonstration: 7–8	Development of large prototype: 6	[25]

Figure 2. Schematic representation of water electrolysis cell configurations: (a) AEL, (b) PEMEL, (c) AEMEL, (d) SOEL, detailing anode/cathode half reactions, ion transport across the electrolyte and charge-carrier pathway.

Table 2. Advantages and disadvantages of water electrolysis technologies.

Technology	Advantages	Disadvantages
AEL	High maturity and robust technology Longer stack lifetime Industrial-scale applications Use of available and inexpensive catalysts for electrodes Non-platinum group metals (PGMs) required Low cost compared to other technologies	Low current densities, therefore, bigger space is required for the electrolyser Corrosive liquid electrolyte Mixing of gases (crossover) Low dynamic response Low operating pressure
PEMEL	Commercialised technology High current densities, more compact stack Higher hydrogen purity Rapid system response allowing flexible operation (especially with RESs) No need for aqueous electrolyte, increasing safety	Works in an acidic environment which is highly corrosive High costs of electrolyser components (membrane and catalysts) Use of noble metal catalysts Shorter lifetime Water quality sensitivity
AEMEL	Absence of corrosive liquid electrolyte Lower cost of electrolyser components (compared to PEMEL membrane) Use of non-noble metal catalysts Flexible operation due to fast response times	Low TRL Low current densities Low durability Membrane degradation
SOEL	Higher energy efficiency Less electricity consumption Use of non-noble metal catalysts	High temperature requirement Unstable electrodes due to exposure to high temperature and humidity Sealing issues Higher costs

RESs: Renewable Energy Sources.

Electricity Grid Mix vs. Renewable Energy Sources

In Europe, the electricity consumption across the 27 countries of the European Union (EU) and the United Kingdom (UK) is approximately 3200 TWh per year. The source of electricity for the 27 EU countries for 2022 is divided into 39% coming from fossil fuels, 22% coming from nuclear energy, 38.3% coming from renewable energy (wind, solar, hydro and bioenergy) and 0.2% coming from less conventional sources like geothermal [26]. The amending Directive EU/2023/2413, commonly referred to as “RED III”, sets an overall renewable energy target of at least 42.5% (with an aspirational goal of 45%) at the EU level by 2030 in the overall energy mix [27].

However, until these targets are met, the current reliance on fossil fuels poses a significant challenge for hydrogen production. For electrolysis to be considered a renewable method of producing H₂, it must be powered by clean energy sources such as solar or wind. In this scenario, the hydrogen’s carbon footprint would be low enough to meet emission benchmarks set at the EU level; for example, the Delegated Regulation (EU) 2023/1184, adopted under the Renewable Energy Directive (RED II) establishes a threshold of 3.38 kg CO₂-eq/kg of H₂, representing at least 70% GHG savings compared to fossil fuels [28,29]. Conversely, when connected to the grid, the environmental impact of electrolysis depends largely on the grid’s energy mix. If the grid relies on fossil fuels, electrolysis can result in higher emissions than steam methane reforming (SMR) or coal gasification (CG) processes. Table 3 compares the global warming potential (GWP) reported in various studies for hydrogen production via electrolysis, coupled to the grid or renewable sources, against fossil-based processes such as coal gasification and steam methane reforming. These values confirm that the composition of the electricity mix is one of the most influential parameters determining the environmental footprint of electrolytic hydrogen.

Table 3. Global Warming Potential from conventional hydrogen production pathways and electrolysis coupled to the grid or coming from renewable sources.

Process	System Boundaries/ Functional Unit	Technology Characteristics	Grid Geography or RES	GWP (kg CO2-eq/kg H2)	Ref.
Coal Gasification (CG)	Not disclosed/ 1 kg of H2	Without CCS		11.6	[30]
	Well to pump/ 1 kg of compressed H2 (700 bar)	Without CCS		22.8	[31]
Steam Methane Reforming (SMR)	Cradle-to-gate/ 1 kg of H2	Without CCS		12.3	[32]
		Coupled with CCS		7.6
	Cradle-to-gate/ 1 kg of H2 (200 bar, purity > 99.9% vol.)	Without CCS		~11.5	[33]
	Not disclosed/ 1 kg of H2	Without CCS		10.3	[30]
Electrolysis coupled to the grid	Cradle-to-gate/ 1 kg of H2	AEL	Germany’s 2019 electricity mix (33% share of renewable energy)	26.9	[34]
		PEMEL		27.9
		SOEL		23.5
	Well to pump/ 1 kg of compressed H2 (700 bar)	Electrolysis not specified	U.S. 2019 electricity mix (65.8% fossil-based, 20.3% nuclear, 6.5% hydropower, wind 4.9%, others 2.5%)	28.6	[31]
	Cradle-to-gate/ 1 kg of H2 (30 bar, 50 °C, purity > 99.99%)	PEMEL	Europe 2030 electricity mix (% share of renewables not specified)	2.37	[24]
		AEMEL		2.42
	Cradle-to-gate/ 1 kg of H2	AEL	U.S. 2022 grid mix (61% fossil-based)	>15	[35]
		PEMEL		>20
	Cradle-to-gate/ 1 kg H2 (925 psig, purity > 99.9% vol.)	PEMEL	U.S. 2016 grid mix (share not specified)	31.3	[36]
		SOEL		25.2
Electrolysis coupled to Renewable Energy Sources (RESs)	Cradle-to-gate/ 1 kg of H2	AEL	Solar PV + Wind (Germany)	2.90	[34]
		PEMEL		2.94
		SOEL		3.84
	Cradle-to-grave/Generation of 1000 tonnes of hydrogen produced over the lifetime of electrolyser system driven by dedicated solar/wind farms	AEL	Solar PV (Australia)	3.39	[37]
			Wind (Australia)	0.85
		PEMEL	Solar PV (Australia)	3.13
			Wind (Australia)	0.75
	Not stated/ 1 kg H2	AEL	Solar PV ** (Nordic countries)	2.7 *	[38]
			Wind ** (Nordic countries)	0.84 *
		PEMEL	Solar PV ** (Nordic countries)	2.8 *
			Wind ** (Nordic countries)	0.71 *
		SOEL	Solar PV ** (Nordic countries)	2.4 *
			Wind ** (Nordic countries)	0.71 *
	Cradle-to-gate/ 1 kg of H2	AEL	Solar PV (U.S.)	~2.4	[35]
			Wind (U.S.)	<1
		PEMEL	Solar PV (U.S.)	~2.3
			Wind (U.S.)	<1
	Cradle-to-gate/ 1 kg of H2 (3 MPa, 60 °C, purity 99.9%)	PEMEL	Solar PV (Saudi Arabia)	3.66	[39]
			Wind (Saudi Arabia)	0.76
	Not disclosed/ 1 kg of H2	AEL	Solar PV (geography not specified)	3.08	[30]
	Cradle-to-gate/500 Mt H2/year	PEMEL	Solar PV (Global 2019)	~3.9	[40]
			Wind (Global 2019)	~1.0
	Cradle-to-gate/ 1 kg H2 (925 psig, purity >99.9% vol.)	PEMEL	Solar PV (U.S.)	2.76	[36]
			Wind (U.S.)	1.83
		SOEL	Solar PV (U.S.)	2.93
			Wind (U.S.)	2.20

* Value calculated as an average from the minimum and maximum. ** Used an average emission factor (g CO₂-eq/kWh) of 30 and 10 g CO₂-eq/kWh for solar PV and wind electricity production, respectively.

Water Use in Electrolysis

Water intensity is a critical eco-design factor for renewable H₂ production. Electrolysis requires ultrapure (deionised) water to reduce electrical resistance, prevent mineral deposition [41] and contamination of the catalyst, diaphragm or membrane [42] and avoid rapid degradation of the stack components. From a chemical and stoichiometric perspective, producing 1 kg of H₂ requires 9 kg of water for electrolysis processes. However, industrial operation augments water consumption due to cooling requirements and water purification losses. Depending on the technology, electrolysis water consumption typically ranges from 18 to 24 kg of water per kg of H₂ [18]. When accounting for the entire balance of the plant, including water pretreatment (demineralisation) and cooling systems, the total water withdrawal intensity can rise to 34.6 kg of water per kg H₂, as observed in alkaline electrolysis estimations [43].

To contextualise the water requirements of low-emission hydrogen, Table 4 compares the operational water withdrawal of electrolysis against fossil-based pathways. Two distinct metrics are reported in Table 4 and must be carefully distinguished: water withdrawal refers to the total volume of water anthropogenically removed from a water body, either permanently or temporarily, while water consumption refers to the volume that is removed from the water body, is not returned to it, and is incorporated into the product, evaporated, transpired or displaced [44]. For H₂ production, comparing the values presented by [43], withdrawal is typically larger than consumption in the fossil-based pathways with significant cooling loads (specifically the pathways coupled to CCS). Meanwhile, the water demand for electrolysis remains lower than that of coal gasification, which stands as the most water-intensive process in that study.

Table 4. Water withdrawal and consumption intensities for the main H₂ production pathways to generate 1 kg of H₂. Table prepared by the authors based on data reported in [36,43,44].

Process Technology	Scope of Assessment (System Boundary and Processes Included)	Water Withdrawal (kg H2O/kg H2)	Water Consumption (kg H2O/kg H2)	Ref.
Fossil-Based
Coal Gasification	Coal–water slurry gasification. Data based on engineering design models obtained from the industry. Cooling included.	48.2–51.4	29.9–32.0	[43]
	Cradle-to-gate, including upstream coal production. Data based on industrial reports and process simulation.	-	21.8–32.7	[36]
Coal Gasification + CCS	Coal–water slurry gasification. Data based on engineering design models obtained from the industry. Cooling load for CCS system included.	73.8–87.2	46.5–52.5	[43]
	Cradle-to-gate, including upstream coal production and CCS. Data based on industrial reports and process simulation.	-	28.5–39.1	[36]
Steam Methane Reforming	Steam methane reforming of natural gas. Data obtained from process simulations, cooling included.	16.4–25.2	15.8–19.8	[43]
	Cradle-to-gate, including upstream natural gas production. Data based on industrial reports and process simulation.	-	12.2–14.5	[36]
Steam Methane Reforming + CCS	Steam methane reforming of natural gas. Data obtained from process simulations, cooling included.	29.8–47.8	24.1–39.0	[43]
	Cradle-to-gate, including upstream natural gas production and CCS. Data based on industrial reports and process simulation.	-	35.4–36.0	[36]
Electrolysis
Electrolysis (AEL) (Electricity source not specified)	Data based on engineering design models obtained from the industry, energy efficiency of 70%.	29.9–34.6	21.0–23.6	[43]
Electrolysis (AEL) + Australia’s Average Grid Electricity	System boundaries include electricity generation from either grid mix, solar or wind sources, supply and purification of water, the installation of electrolysis equipment and power plant.	-	130	[44]
Electrolysis (AEL) + Solar PV		-	43
Electrolysis (AEL) + Wind		-	20
Electrolysis (PEM)	Data based on engineering design models obtained from the industry, energy efficiency of 70%. (Electricity source not specified).	24.9–26.5	17.0–18.0	[43]
Electrolysis (PEM) + U.S. Average Grid Electricity	Data from literature, based in a commercial-scale PEM electrolyser. Includes water treatment, drying and purification of H2.	-	281–284	[36]
Electrolysis (PEM) + Solar PV		-	36.4–39.0
Electrolysis (PEM) + Wind		-	30.0–32.5

However, values reported for the same technology can differ by an order of magnitude depending on whether the assessment considers only the production process or extends the boundaries to include upstream processes like electricity generation, fuel extraction or water purification. This effect is visible when comparing the water consumption for PEM electrolysis: IRENA [43] reported a value range of 17.0–18.0 kg H₂O/kg H₂, when considering only the production process, but this rises to 281–284 kg H₂O/kg H₂ once the U.S. grid supply chain, water treatment and full cradle-to-gate boundary are included [36]. Similarly, IRENA [43] reported a range of 21.0–23.6 kg H₂O/kg H₂ for AEL, whereas the expanded system boundaries used by Shi et al. [44], which include electricity generation from the Australian grid and the supply and purification of water, raise the value to 130 kg H₂O/kg H₂. The values reported in Table 4 should therefore be interpreted as indicative ranges whose magnitude depends on the specific assessment scope.

Crucially, the water footprint of hydrogen extends beyond the electrolyser use to the power generation source. Mehmeti et al. [45] highlighted this by calculating the water scarcity footprint (WSF) for PEMEL and SOEL. The direct water requirement for electrolysis was 18.04 kg of water for PEMEL and 9.1 kg of water for SOEL to obtain 1 kg of H₂. It could be concluded that the WSF was dominated by the energy source. PEM electrolysis coupled with the grid resulted in a WSF of 9.60 × 10³ m³ per kg of H₂, whereas using wind energy reduced the footprint to 6.29 × 10² m³ per kg of H₂. A similar trend was observed for SOEL, with a WSF of 6.31 × 10² m³ per kg of H₂ for the electrolyser connected to the grid and 3.79 × 10² m³ with electricity coming from renewable energy (wind). The authors concluded that the impacts associated with water scarcity were determined by electricity consumption. Similarly, Shi and co-authors [44] assessed hydrogen production in Australia, concluding that grid-connected electrolysis exhibits the highest water consumption (0.13 m³ per kg of H₂). In contrast, coupling electrolysis with renewable sources significantly lowers the water burden; when PV is used the water consumption decreases to 0.04 m³ per kg of H₂ and it is even lower when wind electricity is used (0.02 m³ per kg of H₂).

In water-stressed regions the reliance on high-quality freshwater presents a significant barrier to the large-scale deployment of electrolysis. Consequently, integrating seawater or wastewater has been investigated for sustainable process design. For instance, Vazquez-Sanchez et al. [39] investigated hydrogen production via PEM electrolysis using desalinated seawater, a relevant study for water-stressed regions such as Saudi Arabia. The study estimated a seawater requirement of 28.6 kg of seawater to produce 1 kg of H₂, accounting for potential losses during the desalination process. When evaluating the entire life cycle emissions, the study found that PEM coupled with solar photovoltaic (PV) had a freshwater consumption (FWC) of 9 × 10⁻² m³/kg H_2, whereas PEM coupled to wind energy presented a lower water footprint of 7.2 × 10⁻³ m³/kg H₂. This disparity highlights that the freshwater footprint is largely due to water requirements of PV panel manufacturing, rather than the operation of the electrolyser. In comparison, while hydrogen production via SMR has a lower operational water demand, the upstream activities, such as the extraction of oil and gas, result in a substantial freshwater footprint of approximately 2.65 m³ of freshwater per kg H₂.

While electrolysis offers a reduced life cycle water footprint compared to fossil-based routes, its operational reliance on high-purity freshwater continues to be a critical eco-design challenge. Future sustainability depends on decoupling hydrogen production from freshwater stress by advancing technologies capable of valorising non-conventional streams, like seawater [46] and municipal wastewater [47].

Materials Used for Electrolyser Assembly and Recyclability

The material composition of electrolysers represents a critical eco-design hotspot, particularly given the projected hydrogen deployment.

Zhao et al. [20] conducted a comparative LCA of three electrolysis technologies (AEL, PEMEL and SOEL) focusing on stack manufacturing with a functional unit (FU) of 1 m² of stack area for each technology. In terms of GWP for the stack production, PEMEL exhibits the highest impact (1.76 × 10³ kg CO₂-eq per m²), followed by AEL (2.01 × 10² kg CO₂-eq per m²) and SOEL (1.09 × 10² kg CO₂-eq per m²). The trade-offs become more pronounced in the impact category mineral resource scarcity (MRSP) expressed as kg of Cu equivalent. PEMEL again presents the highest burden (4.59 × 10² kg Cu-eq per m²) due to the reliance on platinum, titanium, and iridium for interconnections and hydrogen and oxygen electrodes. In contrast, AEL (23.2 kg Cu-eq per m²) and SOEL (6.50 kg Cu-eq per m²) showed lower values for this impact category, with AEL impacts driven by nickel and for SOEL driven by stainless steel used in interconnections.

Emerging technologies such as AEMEL aim to mitigate this critical material dependence. Schropp et al. [24] analysed the assembly of an anion exchange membrane (AEM) electrolyser with a daily H₂ production of 1000 kg, using a QAPS membrane and a platinum group metal (PGM)-free design. The study found that membrane production contributes more than 90% of the ozone depletion potential (ODP) (4.52 × 10⁻² kg CFC-11-eq) due to emissions to air related to one of the steps considered for membrane production (chloromethylation of polysulfone). Furthermore, regarding manufacturing, the GWP presented a value of 5.04 × 10⁴ kg CO₂-eq, with 34% of emissions attributed to the electrode spraying process, which includes the use of isopropanol as a solvent. The production of the end plates contributed 31% of the GWP emissions, due to the high mass share of the end plates and their construction material (chromium steel).

A material criticality assessment highlights the supply chain risks associated with materials. For PEMEL, the material hotspot corresponded to iridium, titanium, and platinum, all of which are classified as critical raw materials (CRMs) by the European Commission [48]. In contrast, AEMEL and AEL systems present critical hotspots associated with materials with high mass shares, such as chromium and iron (coming from stainless steel) and nickel. While nickel is not a CRM, it is designated as a strategic raw material owing to its important role in the green transition [49].

To improve the eco-profile of these electrolysis technologies, the EU encourages the recyclability of metals. The recycling input rate (RIR) refers to the percentage of overall demand that can be covered by secondary raw materials. Currently, metals like copper (RIR 55%) and nickel (RIR 16%) have established recycling strategies, whereas PGMs lag behind with an RIR of 12%. Increasing the recyclability of critical materials used in electrolyser construction could improve the eco-profile of the different electrolysis technologies [48].

Ultimately, improving the eco-profile of hydrogen production via electrolysis requires tailored material strategies that address the specific maturity and limitations of each technology. For PEMEL, the main material challenge lies in the use of PGMs. Reducing the noble metal content and/or the replacement of these materials would enhance not only the environmental but also the economic performance of PEM [20]. In contrast, AEL already benefits from using non-noble metals or metal oxides as catalysts, yet ongoing research aims to optimise their catalytic activity further and cost-effectiveness to maintain competitiveness [50]. Meanwhile, emerging technologies offer major opportunities for eco-design innovation; for SOEL the state-of-the-art material is yttria-stabilised zirconia (YSZ), although research on alternative and innovative materials to achieve competitive H₂ production has been investigated [51], while AEMEL’s early development stage leaves room to optimise technical parameters and embed sustainable material choices before industrial standardisation sets in [52].

3.1.2. Thermochemical Water Splitting

Direct water thermolysis, the thermal decomposition of water into H₂ and O₂, requires a large quantity of energy provided by heat, since water can be decomposed at temperatures higher than 1700 °C. Additionally, the product is a gas mixture of H₂ and O₂ that needs to be separated later. While thermolysis requires extremely high temperatures, thermochemical cycles use chemical reagents to lower the required temperature, making the process more feasible for industrial-scale production [52,53].

Thermochemical water-splitting cycles (TWSCs) are another way to produce H₂ based on water decomposition through repetitive chemical reactions using high temperatures. The temperatures required in this process can range from 500 up to 2500 °C, and the source of heat can be derived from concentrated solar power or residual heat of nuclear power reactions [54]. Due to the high temperature required for a single step, thermochemical cycles using oxidation/reduction chemicals have been proposed, performing the cycles in various steps (three, four or even more steps). Cycles are categorised by the number of steps they involve, where increasing the number of steps generally decreases the maximum temperature needed [53].

Promising oxidation/reduction chemicals include copper–chlorine (Cu-Cl), sulphur–iodine (S-I), magnesium–chlorine (Mg-Cl), and magnesium–iodine (Mg-I). They can help to lower the reaction temperature; nonetheless, temperatures around or even above 800 °C are required. As depicted in Figure 3, the intermediate chemicals are recycled and re-used in the same process to create a closed-loop system, having H₂ and O₂ as final products [55].

Figure 3. Schematic representation of the solar-driven Sulphur–Iodine (S-I) thermochemical water-splitting cycle, illustrating high-temperature heat integration from a concentrated solar tower and the closed-loop chemical reactions (Sulphuric acid (H₂SO₄) decomposition, Bunsen reaction, and Hydrogen Iodide (HI) decomposition) for hydrogen evolution.

Environmental Performance and Hotspots

One of the most significant challenges for TWSCs is the heat source required to reach the working temperatures for water dissociation [14]. If coupled with a renewable heat energy source, thermochemical cycles could present lower values than the threshold established by the Renewable Energy Directive (RED II), as demonstrated in various LCA studies [55,56,57,58] displayed in Table 5.

Table 5. Global Warming Potential from TWSC coupled with renewable energy sources.

System	System Boundaries/Functional Unit	GWP (kg CO2-eq/kg H2)	Ref.
Renewable Energy Sources (hydroelectrical + wind+ solar) + Cu-Cl cycle	Energy production, energy required for the H2 production, H2 transportation, and ends with H2 utilisation by a Fuel Cell Electric Bus (FCEB)/1 kg of H2	0.87	[56]
Solar Tower + 4-step Cu-Cl cycle	Cradle-to-grave/ 1 kg of H2	0.94	[57]
Solar Photothermal + S-I cycle	Cradle-to-gate/ 1 kg of H2	1.02	[55]
Solar-based Hybrid Sulphur (HyS)	Cradle-to-gate/ 1 kg of H2	1.04	[52]
Concentrated Solar Power Plant + S-I cycle	Production of H2 */ 1 kg of H2	1.18 *	[58]

* The authors considered in the system boundaries after the production of H₂ via TWSC S-I cycle the hydrogenation into LOHCs, transport by truck/ship, and dehydrogenation of LOHCs, obtaining a total impact of 6.49 kg CO₂-eq per kg H₂, the value reported in the table corresponds only to the production of H₂.

Sadeghi and Ghandehariun [57] performed an LCA of a solar-powered Cu-Cl cycle, finding a global warming potential (GWP) of 0.94 kg CO₂-eq/kg H₂. This is primarily because a TWSC utilises direct solar heat rather than converting light to electricity first. However, in this case the hotspots shift from operation to construction. The study revealed that 90% of total emissions (0.853 kg CO₂-eq/kg H₂) are allocated to the solar subsystem (heliostat mirrors, tower, molten salt and storage tanks) due to the intensive use of steel, iron alloys, and glass. Similarly, Zhang et al. [56] analysed a solar-coupled S-I cycle, reporting a GWP of 1.02 kg CO₂-eq, and concluded that the construction of the solar energy plant and H₂ plant were the dominant environmental burdens. The materials required for the specialised equipment used to conduct TWSCs are another challenge that needs to be overcome, due to high temperatures and the use of corrosive and harmful chemicals [54,59].

The values presented in Table 5 could be lower than those of fossil-based pathways and some specific electrolysis cases (see Table 3). This comparison must, however, be interpreted with caution: the reviewed studies differ in system boundaries, modelling assumptions, geographical context, plant scale, technology maturity (TRL 3–4 for thermochemical cycles versus TRL 8–9 for commercial electrolysis), and data sources. In particular, thermochemical cycle values reflect scaled-up plant modelling rather than measured industrial data and therefore represent projected performance.

Corrosive Environments and Toxicity

While thermochemical water-splitting cycles (TWSCs) offer a low carbon footprint, a primary eco-design drawback is the aggressive nature of the intermediate chemical reagents. Sulphur-based cycles, such as the sulphur–iodine (S-I) and hybrid sulphur (HyS) processes, use sulphuric acid (H₂SO₄) and hydroiodic acid (HI). At the high temperatures required for decomposition, these acids create a hostile environment that can severely degrade reactor materials [53,60]. Researchers are investigating highly corrosion-resistant materials, such as silicon carbide, to withstand these conditions [61]. Similarly, the chlorine family (e.g., Cu-Cl, Mg-Cl, Fe-Cl, V-Cl) and even the two-step zinc oxide (ZnO/Zn) cycle rely on corrosive working agents and by-products [62] that pose significant material challenges for industrial scale-up.

Beyond equipment degradation, these processes carry substantial risks regarding human toxicity potential (HTP) and ecosystem health. Many thermochemical processes involve toxic chemical reductants or oxidants, and the literature specifically highlights the risk of toxic leakages from S-I hydrogen plants [61]. Furthermore, the HTP is heavily influenced by the chosen heat source. For instance, when the HyS cycle is driven by nuclear power, the HTP environmental indicator increases to 1.90 kg 1,4-DB-eq, due to the environmental burden and disposal of uranium waste, compared to 0.27 kg 1,4-DB-eq when the cycle is solar-powered [52].

In summary, although a TWSC provides a low-carbon pathway for hydrogen production, the process has not yet been developed at a large scale [54] and remains at a TRL of 3–4 [63]. The transition from laboratory scale to industrial viability will require dealing with these complex challenges: appropriate treatment of the toxic chemicals and by-products to maintain safety, avoiding toxic leaks, advancing robust reactor materials, and effectively coupling renewable heat sources.

3.1.3. Photocatalytic Water Splitting

The conversion of sunlight into chemical energy is the basis of photocatalytic hydrogen production. Generally, these systems require a reactant, in this case water, a light source, a photoreactor that allows light transmission, and a photocatalyst that absorbs the light. For efficient hydrogen production, an efficient interaction between light, catalysts and reactants is essential. Additionally, the presence of different organic compounds used as sacrificial agents can enhance hydrogen production and the formation of other high-value compounds [64].

Fundamentally, the photocatalytic production of H₂ operates through a three-step mechanism. First, photoexcitation occurs when the semiconductor catalyst absorbs light energy, generating electron–hole pairs (e⁻h⁺). Second, these charge carriers separate and migrate to the surface of the photocatalyst particles. Finally, the photogenerated charges participate in surface redox reactions, where the available electrons reduce protons to form molecular hydrogen (H₂) [65]. A simplified diagram can be seen in Figure 4.

Figure 4. Schematic representation of solar-driven photocatalytic water splitting, illustrating light absorption, photogenerated electron–hole (e⁻/h⁺) pair generation and surface redox reactions for hydrogen evolution.

Different systems of photocatalytic H₂ production have been studied, namely heterogeneous photocatalytic H₂ production systems (HETPHPs), in which a semiconductor acts as both catalyst and light harvester; homogeneous photocatalytic H₂ production systems (HOMPHPs) where an organometallic complex acts as the catalyst and an organic/organometallic compound acts as a photosensitiser; and hybrid photocatalytic H₂ production systems (HYBPHPs) which utilise homogeneous catalysts and heterogeneous semiconductor photosensitisers [66].

Catalyst Materials and Performance

Critical factors are assessed when selecting an optimal photocatalyst for hydrogen production. The material properties such as band gap, which determines the light absorption capability of a material, and charge carrier dynamics are of the utmost importance [67]. The main photocatalysts that have been evaluated for hydrogen production are divided into four categories: I. titanium-oxide-based photocatalysts, II. Cadmium-sulphide-based photocatalysts, III. zinc oxide/sulphide-based photocatalysts and IV. other metal-oxide-based photocatalysts [68].

Historically, titanium oxide (TiO₂) photocatalysts have been widely studied owing to their non-toxicity, low cost and photochemical stability; however, their large band gap limits absorption to the UV spectrum. To address this, coupling TiO₂ with visible-light active semiconductors such as graphitic carbon nitride (g-C₃N₄) has gained significant attention [69]. Conversely, cadmium sulphide (CdS) photocatalysts exhibit strong visible light absorption [67] and have shown higher production rates [68]. Other promising candidates include metal sulphides, because they possess more suitable conduction bands and offer a better response to sunlight [70], and tantalates (e.g., LiTaO₃), which can perform water splitting without the need for a sacrificial agent or co-catalysts [71]. Perovskite materials also offer outstanding optical performance and compositional flexibility [67]. Table 6 summarises selected catalysts evaluated under direct sunlight irradiation. As this review focuses on solar-driven applications, catalysts requiring simulated lamps are excluded (comprehensive listings can be found in [67,68,69,70,71,72]).

Table 6. Performance of different photocatalysts under solar irradiation.

Photocatalyst	Light Source	Sacrificial Agent	Hydrogen Production Rate (H2 mmol h−1 g−1)	Ref.
CdS/TiO2 (25%)/Pt (1%)	Visible Sun	Na2S/Na2SO3	68.0	[73]
ZnO-MoS2-graphene	Sunlight	Na2S/Na2SO3	28.6	[74]
Bi2O3 (0.4%)/TiO2	Visible Sun	Glycerol (5%)	26.0	[75]
CdS.ZnS	Sunlight	-	22.8	[76]

Environmental Assessment (LCA)

Assessing the environmental impact of emerging technologies like photocatalysis is challenging due to data scarcity and low maturity levels of the technology. Cruz et al. [72] proposed a novel methodological approach in the form of a scoring dashboard, designed for systems with a low maturity level, where the application of a full LCA is often unfeasible. Applying the methodology to a photocatalytic case study, the authors identified the sacrificial agent as the primary driver of the carbon footprint, suggesting the substitution of conventional methanol with biomethanol. While power consumption was less influential, the study highlighted that coupling the process with electricity coming from renewable sources such as wind energy, rather than the Spanish electricity grid, could lower emissions enough to meet the threshold of 3.38 kg CO₂-eq/kg of H₂ proposed by the Renewable Energy Directive (RED II).

Furthermore, Oh et al. [77] evaluated photocatalytic production against conventional SMR and electrocatalysis (ALK) for hydrogen production. The photocatalyst corresponded to the carbon-coated hollow tubular In₂O₃/ZnIn₂S₄ heterostructure [78]. The results revealed that SMR exhibited a significantly lower global warming potential (13.5 kg CO₂-eq/kg H₂) compared to both photocatalytic (63.1 kg CO₂-eq/kg H₂) and electrocatalytic processes (70.1 kg CO₂-eq/kg H₂). The high environmental burden of the photocatalytic process was attributed primarily to the catalyst synthesis stage, driven by the production of precursors and the substantial mass requirement (at least 40 times the catalyst mass used in SMR). The second contributor corresponded to the power consumption of the synthesis step. The authors emphasise that, when scaled up, the synthesis of catalysts becomes unsustainable.

Maurya et al. [79] developed a cradle-to-grave LCA for a theoretical utility-scale plant producing 5 tonnes of H₂ per day, comparing four different photocatalytic materials: TiO₂ nanorods (TNRs), fluorine-doped carbon nitride quantum dots embedded with TiO₂ (CNF: TNR/TiO₂), carbon nitride (g-C₃N₄) and a carbon nitride sheet–bismuth iodide composite (g-C₃N₄/BiOI). The results showed that the lowest GWP footprint was achieved with the g-C₃N₄/BiOI composite, with a value of 0.43 kg of CO₂ per kg H₂, followed by CNF:TNR/TiO₂ with 0.91 kg of CO₂ per kg H₂. The TNR and the g-C₃N₄ exhibited the highest values of GWP, with 1.49 and 2.08 kg of CO₂-eq per kg H_2, respectively. The assessment revealed that material extraction accounts for 83–89% of the total life cycle emissions. The lower values obtained for g-C₃N₄/BiOI and CNF: TNR/TiO₂ are due to their high response to solar radiation resulting in higher hydrogen yields, outweighing the additional material requirements.

The pronounced difference in reported GWP values for photocatalytic water splitting reflects differences in catalyst chemistry but, more importantly, methodological differences between studies. Oh et al. [77] conducted a consequential LCA, using laboratory-scale synthesis inventories without a detailed scale-up methodology, and compared this pathway (TRL 4) against SMR (TRL 9). Meanwhile, Maurya et al. [79] used laboratory data for the photocatalyst synthesis but performed a scale-up based on scale factors for sizing equipment and operational parameters based on the overall capacity assumed for the facility. Direct comparison with commercial processes that produce H₂ should therefore be interpreted with caution.

Solar-to-Hydrogen Efficiency

From a practical and economic standpoint, photocatalytic water splitting is heavily limited by its low solar-to-hydrogen (STH) conversion efficiency and low production rates. To make solar-driven hydrogen production commercially viable for industrial scale-up, an STH efficiency of at least 10% is required [80]. However, conventional single-component materials fall far short of this target under natural sunlight: unmodified titanium dioxide (${\mathrm{TiO}}_2$) achieves an STH efficiency below 1%, while visible-light-active materials such as graphitic carbon nitride (g-${\mathrm{C}}_3{\mathrm{N}}_4$) and cadmium sulphide ($\mathrm{CdS}$) typically yield efficiencies between 1% and 3% [67]. This low productivity is governed by three primary factors:

i.

Limited Light Absorption: Wide bandgap materials like ${\mathrm{TiO}}_2$ can only absorb ultraviolet (UV) light, which accounts for less than 4% of the solar spectrum.

ii.

Rapid Charge Recombination: The light-generated electrons and holes recombine in fractions of a second before they can reach the catalyst surface to drive chemical reactions.

iii.

Sluggish Surface Reaction Kinetics: Breaking water molecules apart requires multiple electron and hole transfers, which occur much more slowly than electron recombination.

Consequently, most laboratory-scale demonstrations achieve measurable hydrogen evolution rates (mmol H₂ g⁻¹ h⁻¹) only through the addition of organic sacrificial electron donors (SEDs, such as triethanolamine or methanol) [69], which consume non-renewable reagents and deviate from overall water splitting. When scaled up, these low volumetric production rates mean large and expensive reactor arrays requiring areas of many square metres to yield 1 tonne per day of H₂. Bound by these technological, kinetic, and material trade-offs, photocatalytic hydrogen production currently remains at a low TRL (3–4). Major advances in advanced material coupling (such as Z- or S-scheme heterojunctions), non-precious co-catalyst development, and light-harvesting efficiency are still required before commercial deployment is possible.

In summary, while photocatalytic water splitting offers an innovative pathway for direct solar-to-hydrogen conversion, it currently remains at a low TRL. Consequently, there is a sharp divergence in LCA results, underscoring a critical dependency on catalyst efficiency and synthesis scalability. Moving forward, research must prioritise earth-abundant materials and scalable, low-energy synthesis routes to bridge the gap between laboratory feasibility and industrial viability.

3.2. Biomass Treatment

Biomass represents an innovative and promising pathway for the production of H₂. Biomass resulting from technosphere activities is generally considered waste and can come from various sources, including agricultural crops, forestry residues (wood, grass), municipal solid waste, algae, and plant and animal waste [81]. Beyond its abundance and wide availability, a primary advantage of using biomass as a feedstock for other productive processes is that it does not compete with food crops or involve land occupation [82].

Converting biomass to produce H₂ has garnered significant interest because it transforms waste into a high-value product. This approach aligns with bioeconomy principles, simultaneously addressing waste management challenges and aiding climate change mitigation [82]. Biomass is considered as “carbon neutral”: in other words, even if carbon dioxide (CO₂) is emitted during the production of H₂, the quantity of CO₂ released is theoretically equivalent to the CO₂ absorbed by the biomass during its growth lifetime, thus resulting in a net-zero carbon process. To produce H₂ from biomass two primary pathways are considered: thermochemical conversion and biological conversion. The following sections discuss these pathways.

3.2.1. Thermochemical Processes

Thermochemical conversion processes utilise heat to perform chemical reactions that break down the complex polymeric structures of biomass into simpler molecules, primarily producing gas streams containing H₂ and other gaseous products. The most common thermochemical methods are gasification and pyrolysis.

Biomass gasification occurs at high temperatures (>700–800 °C), in the presence of an oxidising agent, such as air, oxygen or steam, converting the biomass into a gaseous mixture known as syngas [82,83,84]. The syngas is a mixture of H₂, methane (CH₄), carbon monoxide (CO), carbon dioxide (CO₂) and other gases. Gasification of biomass involves three main stages: I. Processing: reducing moisture content and particle size through drying and grinding. II. Gasification: releasing volatile products that react with oxygen to generate syngas and solid char residue. III. Gas purification and upgrading: removing impurities and solid residues obtaining a gas stream rich in H₂ and CO.

Due to its reliance on higher temperatures and the presence of an oxidising agent, biomass gasification is generally recognised as a more effective method for producing syngas compared to pyrolysis [82].

Pyrolysis is the thermal decomposition of biomass in the absence of oxygen, typically at temperatures between 350 and 600 °C [84,85]. The pyrolysis process yields three different fractions: a liquid fraction (bio-oil), a gas fraction (syngas) and the solid fraction (char). In most pyrolysis processes the main product corresponds to the liquid fraction, usually obtaining a bio-oil [86]. Pyrolysis can also be classified into slow and fast. Slow pyrolysis operates at temperatures between 400 and 450 °C with longer residence times (4–8 min). In contrast, fast pyrolysis operates at higher temperatures (450–950 °C) but exhibits significantly shorter residence times (1–5 s).

For both gasification and pyrolysis, the resulting syngas undergoes the water–gas shift (WGS) reaction, where carbon monoxide reacts with steam to further increase the H₂ content [14]. Then, H₂ is cleaned and purified from contaminants (like tar and sulphur), using technologies such as pressure swing adsorption (PSA), which can achieve purity levels of up to 99.99%, or membrane separation technologies [87]. A schematic representation of these thermochemical conversion pathways is displayed in Figure 5.

Figure 5. Thermochemical biomass conversion pathways for H₂ production: biomass gasification yielding syngas (top), and biomass pyrolysis producing bio-oil, biochar, and syngas fractions (bottom). Syngas from both routes undergoes the WGS reaction followed by downstream purification to obtain high-purity H₂.

Technical Barriers and Process Limitations

Thermochemical processes for H₂ production face significant technical challenges due to the complexity of the chemical transformation routes and technologies employed. Overall efficiency varies significantly depending on the selected technology, operational parameters, and the composition of the biomass [14].

Biomass steam gasification, in particular, struggles with the formation of tar and soot in the gaseous products [88]. Another technical hurdle is catalyst deactivation. This primarily occurs through coke deposition, poisoning and sintering [88], which degrade the catalyst, disrupting its stability and durability over time [89].

Furthermore, the feedstock itself introduces significant variability into process energy requirements. Each type of biomass possesses a specific H₂ content and different moisture levels, which can lead to different H₂ yields and different energy demand for required pretreatment or higher temperatures needed to produce H₂. This inherent feedstock heterogeneity results in varying volatile streams, making it difficult to design a universal reactor or catalyst system that performs optimally across different feedstocks. Ultimately, the performance and outcome of a thermochemical process are highly dependent on the specific biomass source used [90].

Environmental Indicators

Life cycle assessments suggest that the use of biomass for gasification and reforming processes can result in a higher terrestrial acidification potential (TAP) compared to alternative pathways. Mehmeti et al. [45] reported that biomass gasification exhibits a higher TAP (3.71 × 10⁻² kg SO₂-eq/kg H₂) compared to conventional SMR process (8.70 × 10⁻³ kg SO₂-eq/kg H₂) or PEM electrolysis coupled with wind energy (1.18 × 10⁻² kg SO₂-eq/kg H₂). Similar results were corroborated by Arfan et al. [87], who found a TAP contribution of 3.21 × 10⁻³ kg SO₂-eq/kg H₂ for sawdust biomass pyrolysis, attributing it to the air emissions of SO₂, linked to the production and combustion of natural gas used in the process. Furthermore, Khoo et al. [84] found that the gasification of rice straw presented a higher TAP value of 8.0 × 10⁻² kg SO₂-eq/kg H₂, compared to rice straw pyrolysis, as well as using other biomass sources such as wheat straw, switchgrass, sugarcane bagasse and corn stalk. This could be attributed to the chemical fertilisers used in agriculture.

Regarding operational efficiency, Buffi et al. [90] compared the conversion efficiency and cumulative energy demand (CED) of various biomass processes with SMR and electrolysis. Their results indicate that thermochemical conversion efficiencies are comparable to SMR and electrolysis. For wood gasification, the reported efficiency values ranged from 0.43 to 0.70, and for biogas steam reforming, 0.65 to 0.77. This compares to reference values of 0.74–0.85 for SMR and 0.60–0.80 for water electrolysis. When evaluating the CED (considering only non-renewable energy inputs), biomass-based processes demonstrated values below the benchmark of SMR and grid-coupled electrolysis, highlighting the energetic benefits of utilising biomass feedstocks [90].

Finally, in terms of carbon intensity, biomass processes expectedly exhibit lower environmental burdens than SMR (which ranges from 10.4 to 12.9 kg CO₂-eq/kg H₂) and grid-coupled electrolysis (ranging from 13.4 to 26 kg CO₂-eq/kg H₂)_. Wood gasification and biogas steam reforming showed mean values between 5 and 10 kg CO₂-eq/kg H₂, corroborating the environmental benefits of utilising biomass for H₂ production [90].

Negative Emissions for Biomass Gasification

The production of H₂ via thermochemical processes has the potential to achieve very low or even negative emissions if coupled with CCS [13]. For instance, Wu et al. [91] conducted an LCA for H₂ production via biomass gasification with chemical looping using rice straw. When air was used as the gasification agent, the GWP was 5.65 kg CO₂-eq/kg H₂, with the chemical looping stage being the largest contributor, followed by biomass pretreatment (shredding and grinding). However, when CO₂ capture was integrated into the model, it generated carbon credits, driving the GWP down to a negative value of −15.13 kg of CO₂-eq/kg H₂.

Similarly, Susmozas et al. [92] evaluated the environmental impacts of H₂ production via poplar biomass gasification with and without CCS. Without carbon capture, the GWP was 0.41 kg CO₂-eq/kg H₂. Conversely, coupling the gasification process with CO₂ capture resulted in a negative GWP of −14.6 kg CO₂-eq/kg H₂. The “poplar production and transport” phase contributed a favourable −32.8 kg CO₂-eq/kg H₂ due to carbon fixation during biomass growth. The primary contributors to environmental burdens were the H₂ production phase (12.7 kg CO₂-eq/kg H₂) and the carbon capture process itself (5.45 kg CO₂-eq/kg H₂). The authors, who modelled a plant capable of satisfying most of its own thermal and electricity demands, concluded that CO₂ capture represents a significant burden because it relies on external electricity consumption, which directly affects the evaluated impact categories.

Further supporting this, Khoo et al. [84] presented an overall GWP of −20.2 kg CO₂-eq/kg H₂, with −33.8 kg CO₂-eq of sequestration during the agriculture stage and 13 kg CO₂-eq/kg H₂ reported as energy used in the process of switch grass gasification. Lastly, Takeda et al. [85] proposed a novel hydrogen production process named solar-driven advanced biomass indirect gasification (SABI-Hydrogen). By replacing the burning of biomass with external heat from concentrated solar power (CSP), the process increases efficiency and reduces pollutants. The calculated GWP resulted in 1.04 kg CO₂-eq/kg H₂, offering a highly sustainable pathway to produce H₂ while abating CO₂ emissions. Table 7 summarises the GWP and AP values reported in recent literature for various thermochemical pathways and biomass feedstocks.

Table 7. Comparison of Global Warming Potential (GWP) and Acidification Potential (AP) for various biomass-based hydrogen production pathways.

Feedstock	System Boundaries/ Functional Unit	Thermochemical Process	GWP (kg CO2-eq/kg H2)	TAP (kg SO2-eq/kg H2)	Ref.
Sawdust	Gate-to-gate/1 kg H2 (purity 99.99% vol.)	Fast Pyrolysis	1.54	3.21 × 10−3	[87]
Sugarcane Bagasse	Gate-to-gate (starts with agriculture production stage and ends at 1 kg of H2)/1 kg H2	Gasification	2.6	1.0 × 10−2	[84]
Wheat Straw		Pyrolysis	9.3	<2.0 × 10−2
Rice Straw		Gasification	14.4	<7.0 × 10−2
		Pyrolysis	9.1	<5.0 × 10−2
Switchgrass		Gasification	−20.2	4.0 × 10−2
Biomass	Cradle-to-grave/1 kg H2	Gasification	2.67 *	3.71 × 10−2 *	[45]
Biomass	Cradle-to-gate/ 1 kg H2	Gasification	1.04	3.69 × 10−2	[85]
Poplar	Cradle-to-gate/ 1 kg H2	Gasification + CCS	−14.63	2.07 × 10−2	[92]
Rice Straw	Cradle-to-gate/ 1 kg H2	Gasification with Chemical Looping	−15.13	Not Specified	[91]

* Values for [45] are cited in [85] as the benchmark for comparison.

Direct comparison of biomass-based thermochemical pathways with fossil- and electrolysis-based routes requires particular caution. Biomass studies typically apply carbon-credit allocation for atmospheric CO₂ fixation during biomass growth, which shifts the reported GWP toward negative values and depends strongly on the allocation method chosen. Feedstock assumptions (type, moisture content, cultivation practices) and geographic scope also vary widely across studies, and system boundaries frequently exclude downstream compression and purification. As a result, the negative-emission values reported in this section should be understood as specific to each system.

Logistical and Supply Chain Challenges

Beyond conversion efficiency, the scalability of biomass-based hydrogen is heavily constrained by logistical and resource management challenges. A primary issue is the geographical mismatch between rural biomass production sites (such as forests or agricultural hubs) and industrial hydrogen end-users [93]. Because hydrogen possesses a low volumetric energy density, its transportation is both complex and cost-intensive (as will be described in a further section). This poses a challenge regarding transporting large volumes of bulky raw biomass or constructing smaller, distributed hydrogen facilities near the feedstock sources.

Achieving sustainable hydrogen production from biomass requires navigating a highly complex matrix of variables. Process designers must jointly optimise biomass characteristics, technology selection, and operational parameters, while also accounting for broader processes such as transportation logistics and CCS integration.

In conclusion, while biomass thermochemical processes offer a promising route for converting waste into a high-value molecule like H₂, challenges such as feedstock heterogeneity and the formation of undesirable by-products must be managed. Furthermore, securing a truly sustainable supply chain requires locating biomass processing facilities close to feedstock sources; otherwise, the extended transportation of the feedstock may induce additional environmental impacts that offset the carbon-neutral benefits of the biomass itself.

3.2.2. Biological Processes

Biological methods for H₂ production rely on the metabolic activities of microorganisms to convert water or organic substrates into biohydrogen (bio-H₂). The utilisation of organic waste materials for bio-H₂ production has garnered significant research interest, as it aligns with eco-design principles by coupling sustainable processes with municipal or agricultural waste minimisation [12]. These biological pathways are broadly categorised into light-dependent and light-independent mechanisms [94,95].

Biophotolysis is a light-dependent mechanism performed by photosynthetic microorganisms such as cyanobacteria and microalgae; this process combines biological and chemical mechanisms to split water into H₂ using solar energy [95,96]. The rate of H₂ generation depends on the carbon source, light intensity and the specific microbial culture employed [97].

Fermentation processes represent a major route for bio-H₂ and can be divided into two primary types:

Photofermentation is a light-dependent process commonly employing purple non-sulphur (PNS) bacteria under anaerobic conditions. These bacteria utilise light as an energy source to drive the conversion of organic acids into H₂ and CO₂ [98]. The conversion efficiency is strongly influenced by the bacterial strain, growth rate, and substrate type. A key advantage of photofermentation is its substrate versatility, allowing it to process organic acids directly from wastewater and biomass [95].

Dark fermentation is a light-independent process capable of utilising a wide range of organic waste and wastewater streams [98]. Operating under dark, anaerobic conditions, microorganisms produce H₂ alongside organic acids and alcohols [96]. However, the formation of these diverse by-products severely limits the H₂ yield, presenting a major drawback [95,96]. The maximum theoretical yield for dark fermentation is restricted to 4 moles of H₂ per mole of hexose consumed [99].

To overcome the yield limitations of isolated processes, researchers are exploring integrated system design. By coupling dark and photofermentation in sequence, the organic acid effluents from the dark fermentation, rich in volatile fatty acids (VFAs), serve as the direct substrate for PNS bacteria in the photofermentation stage, obtaining a higher H₂ yield [96,100]. As illustrated in Figure 6, this synergistic approach can theoretically boost the overall conversion yield to a maximum of 12 moles of H₂ per mole of glucose [97].

Figure 6. Schematic representation of a sequential two-stage integrated system combining dark fermentation and photofermentation, illustrating the valorisation of volatile fatty acid (VFA) effluents to maximise biological H₂ yield.

Microbial electrolysis cells (MECs) are considered an emerging technology in the biological process field (Figure 7). Microorganisms act as biocatalysts at the anode, oxidising organic matter to release protons and electrons [99]. The electrons travel through an external circuit to the cathode, where they combine with protons to produce H₂ [96]. MECs can be strategically coupled with dark fermentation to valorise its acidic effluent using electrical energy. Notably, MECs require significantly lower external voltages (approximately 0.2 to 0.8 V) compared to conventional water electrolysis [94], providing an opportunity as a hybrid approach.

Figure 7. Schematic diagram of a MEC for biohydrogen production, illustrating anodic organic matter oxidation, electron transfer through an external circuit, and cathodic proton reduction. Red components on the anode represent electrochemically active microorganisms acting as biocatalysts.

Table 8 presents a summary of reported biohydrogen production yields and by-products across various biological pathways and substrates. For comprehensive datasets and expanded reports, readers are directed to [98,101].

Table 8. Summary of reported biohydrogen production yields and by-products across various biological pathways and substrates.

Type of Biological Process	Primary Microorganisms	Substrate	By-Products	Reported Yield or H2 Production Rate (g H2/g Substrate)	Ref.
Direct Biophotolysis	Green algae (Chlamydomonas reinhardtii)	Water (H2O)	Oxygen (O2)	0.1411	[102]
Indirect Biophotolysis	Cyanobacteria (Anabaena)	Water (H2O)	Oxygen (O2)	0.7258
Photofermentation	Rhodobacter sphaeroides	Succinic Acid	CO2	0.0632	[98]
		Butyric Acid		0.0572
Dark Fermentation	Klebsiella oxytoca HP1	Sucrose	CO2, Volatile Fatty Acids (VFAs)	0.0210
	Sludge	Sucrose		0.0130
Coupled System (Dark Fermentation + Photofermentation)	Clostridium acetobutylicum + Rhodobacter sphaeroides	Glucose	CO2, VFAs, Alcohols	0.0696
Microbial Electrolysis	Urban Wastewater	Glucose	CO2	0.0957	[103]

Feedstock Pretreatment

Before biological conversion can occur, the pretreatment of feedstock is a critical prerequisite to remove unwanted components from industrial waste and wastewater. These pretreatment methods are generally classified into four categories: physico-mechanical, physicochemical, chemical, and enzymatic [94]. For instance, agricultural wastes with high cellulosic content must be mechanically and chemically pretreated to break down the rigid lignocellulosic matrix and increase the surface area available for microbial degradation [97]. Similarly, the presence of toxic substances in industrial wastewater, such as heavy metals and phenolic and aromatic compounds, could negatively affect the photofermentation process, requiring pretreatment before H₂ production [97].

From an eco-design perspective, pretreatment constitutes an energy-intensive step that can significantly increase the overall environmental footprint of the process. Furthermore, similar to thermochemical biomass processes, biological facilities must be strategically co-located with the feedstock source. Transporting high-volume, low-density organic waste over long distances introduces severe logistical emissions that can rapidly offset the environmental benefits of biohydrogen production.

Efficiency

The efficiency and yield of bio-H₂ production are highly variable and constrained by complex metabolic pathways. For example, the yield of dark fermentation typically ranges from 4 to 44 g H₂/kg of feedstock, limited primarily by the formation of diverse by-products (such as organic acids) and inherently low production rates. Photofermentation exhibits a slightly higher yield range of 9 to 49 g H₂/kg of feedstock, though it requires precise light management [14]. Technologies such as photofermentation and biophotolysis are often characterised by very low solar-to-hydrogen efficiencies and modest production rates compared to thermochemical methods [103,104]. MEC systems, however, present promising H₂ yields, achieving conversion rates ranging from 67% to 91% depending on the substrate. Nevertheless, the industrial implementation of MECs remains compromised by the high construction costs and scaling-up difficulties [103,105].

Gas Purification

Gas purification is a critical downstream operation for biological hydrogen production, as most biological routes yield a mixed biogas rather than a pure hydrogen stream. This separation step is essential if the hydrogen is intended for transportation or fuel cell applications, which demand stringent high-purity standards.

Almost all biological pathways require some degree of cleaning, though the extent depends on the process utilised. Dark fermentation typically produces a biogas mixture containing 50–60% of H₂ and 40–50% CO₂ [94]. It could also contain traces of methane (CH₄), hydrogen sulphide (H₂S) and occasionally carbon monoxide (CO), all of which must be separated from the H₂ stream [94,103]. Photofermentation produces a mixture of H₂ and CO₂ [98]. Biophotolysis simultaneously generates H₂ and O₂; the continuous removal of oxygen is necessary, as its accumulation acts as an inhibitor to the hydrogenase enzymes, directly affecting hydrogen production. In contrast, if operational conditions are optimised, MECs can produce pure H₂ gas without the requirement for clean-up [103]. From an eco-design perspective, gas purification necessitates additional downstream energy consumption, which can significantly enlarge the overall environmental footprint of these biological processes.

Environmental Performance

The environmental performance of biological hydrogen production exhibits extreme variability, as illustrated by the global warming potential (GWP) values summarised in Table 9. Reported values span from 0.7 kg CO₂-eq/kg H₂ (dark fermentation of sugarcane bagasse) to 21.8 kg CO₂-eq/kg H₂ (dark fermentation of wheat straw), reflecting the complex interplay of biological and operational variables. Differences in data sources, system boundaries, and reactor scale further complicate cross-study comparisons: Zheng et al. [106] combined laboratory data with information from a pilot plant located in China; Khoo et al. [84] relied on laboratory-scale studies and process simulations; and Chen et al. [107] used pilot-scale data. Given the low TRL of biological hydrogen processes, LCA studies must rely on scale-up assumptions to represent industrial-scale operation, and direct comparison with mature technologies is therefore not yet meaningful.

Table 9. Global Warming Potential (GWP) of various biological hydrogen production pathways.

Feedstock	Biological Process	System Boundaries/Functional Unit	GWP (kg CO2-eq/kg H2)	Ref.
Food Waste	Hydrolysis + Dark Fermentation	Cradle-to-gate/ 1 kg H2 (purity 99.99% wt.)	10.1 a–10.9 b	[106]
Wheat Straw WS-Df-1	Dark Fermentation	Gate-to-gate (starts with agriculture production stage and ends at 1 kg of H2)/1 kg H2	21.8	[84]
Sugarcane Bagasse	Dark Fermentation		0.7
Wheat Straw WS-Df-2	Dark Fermentation		4.8
Not Reported	Dark Fermentation + MEC	Cradle-to-grave/ 1 kg H2	6.60–16.29 c	[45]
Wastewater	MEC	Cradle-to-gate/ 1 kg H2	18.8 d	[107]

^a Economic allocation; ^b System expansion; ^c Dark fermentation + microbial electrolysis cell, with and without energy recovery; ^d Performance optimisation.

Key factors such as the chemical composition of the feedstock, the energy intensity of necessary pretreatments (e.g., mechanical milling vs. chemical hydrolysis), the efficiency of the microbial metabolic pathways, and the downstream energy required for gas purification have a strong influence on the final carbon footprint. Methodological choices in the LCA study, for example, how co-products (like volatile fatty acids) are credited via system expansion or economic allocation, also have a significant effect on the reported GWP, further limiting direct comparability across studies.

Future Outlook

Although biohydrogen offers the distinct advantages of operating under safe, low-temperature conditions and utilising waste streams (pollution-free valorisation), significant operational bottlenecks remain [82]. Issues related to bacterial performance, by-product formation, and fundamentally low production yields currently hinder the industrial scalability of bio-H₂. To make biological pathways competitive with conventional or electrolytic hydrogen production methods, future research must focus on genetically optimising microbial strains to synthesise H₂ at higher rates, engineering advanced bioreactor designs, and perfecting coupled systems (such as dark/photofermentation or MECs) to overcome the current operational limitations [104].

3.3. Natural Hydrogen Production

Natural hydrogen, often referred to as “white hydrogen”, refers to H₂ that has been produced through various geological and biological processes within the Earth’s crust, which subsequently becomes trapped by impermeable rock barriers and is not released to the atmosphere [108]. When produced independently of biological activity, it is classified as abiotic H₂. The primary mechanisms for natural H₂ formation include:

Serpentinisation: considered the most significant abiotic pathway, this process occurs when water infiltrates the Earth’s mantle and interacts with ultramafic rocks. Ultramafic rocks are low-silica rocks rich in reduced metals, primarily consisting of iron and magnesium minerals, with olivine being the main constituent. At temperatures between 200 and 300 °C, water oxidises the ferrous iron (Fe²⁺) in the rocks to ferric iron (Fe³⁺), forming minerals like magnetite and releasing as H₂ as a by-product [109].

Radiolysis: the radioactive decay of elements such as uranium, thorium, and potassium in the Earth’s crust emits alpha, beta, and gamma radiation that ionises water molecules, breaking the bonds between hydrogen and oxygen and creating free radicals that eventually recombine to form molecular H₂ [109].

Rock fracturing: mechanical stresses during tectonic activity or earthquakes can break chemical bonds in silicate rocks. This creates highly reactive free radicals on fresh mineral surfaces, which then react with infiltrating water to generate hydrogen [109].

Magma degassing: deep-seated magmatic systems operating at extreme temperatures (around 1200 °C) favour hydrogen in their chemical equilibria. This H₂ can be exsolved and released as the magma degasses during its ascent toward the surface [109].

Biological processes: in the deep subsurface, certain microorganisms produce H₂ via fermentation and nitrogen fixation as part of their metabolic cycles. However, the net contribution of biological activity to natural H₂ reserves remains poorly understood [108], as other subsurface microorganisms may also consume H₂ as an energy source [110].

• Extraction of Natural H₂ and Eco-Design Trade-Offs

The extraction of natural H₂ currently relies on adapting established technologies from the natural gas industry, utilising drilling rigs to penetrate the rock layers. Extraction occurs through pipelines that harness the natural pressurisation of the reservoir (often exceeding 500 bar) to pipe the gas to the surface to fill tanks for small-scale distribution [108].

While white hydrogen bypasses the heavy manufacturing footprint of electrolysers, or the carbon emissions of conventional methods, using the techniques already established may entail potential environmental impacts that cannot be overlooked.

The exploitation of deep underground reservoirs may interfere with natural hydrological cycles [110]. Such interference can trigger adverse ecological effects, including the accumulation of nitrates in groundwater. Furthermore, because hydrogen is highly chemically reactive, its presence in underground environments can induce unfavourable fluid–rock reactions that generate toxic and corrosive gases, such as hydrogen sulphide (H₂S) [110]. This introduces severe eco-design constraints regarding equipment corrosion, flammability, and human toxicity.

2.

Efficiency and Environmental Performance

Unlike thermochemical, electrochemical, or biological pathways, natural (geological) hydrogen production is a resource extraction activity. Performance is characterised by different metrics, namely well productivity (kg H₂/well/day, or Nm³/h), extraction energy and gas purity. Reported values derive almost exclusively from exploratory wells and modelling studies. The only continuously producing site currently reported in the peer-reviewed literature is the Bourakebougou well in Mali, which delivers hydrogen at approximately 98% purity [111].

Direct comparison with mature technologies is therefore not yet meaningful, and dedicated LCA frameworks for extraction pathways are urgently needed, as discussed in Section 8.

Although natural hydrogen may appear to offer environmental advantages over conversion processes, site selection is critical. Extraction infrastructure must be strategically managed to minimise local ecological and social impacts, particularly concerning proximity to protected natural areas, urban centres, and vulnerable water bodies [112].

4. Comparative Analysis and Eco-Design Trade-Offs

The transition toward a low-emission H₂ economy relies on replacing conventional, fossil-based processes (like SMR and CG) with alternative production methods based on renewable energy or raw materials. However, as some metrics demonstrate, every low-emission hydrogen technology presents a unique set of eco-design trade-offs, often shifting the environmental burden from CO₂ emissions to other environmental impacts such as material scarcity, water deprivation or terrestrial acidification. Table 10 synthesises the technical maturity, carbon footprint ranges, and primary eco-design bottlenecks of the hydrogen production pathways analysed in this review.

Table 10. Comparison of hydrogen production technologies: Maturity, Global Warming Potential, and Eco-design Hotspots.

Technology	TRL	Indicative GWP Range (kg CO2-eq/kg H2)	Eco-Design Hotspots	Mitigation Pathway
Conventional (fossil-based)
SMR/Coal Gasification	9	7.6–22.8	High carbon emissions. Upstream fossil fuel extraction.	CCS integration (although this increases water consumption).
Water Splitting
Water Electrolysis (AEL, PEM)	8–9	<1–26 *	Highly dependent on the composition of the electricity grid. Using critical raw materials for electrodes (PEM) and strategic raw materials (AEL). High reliance on ultrapure water.	Ensuring that electrolysis is coupled to RES. Material optimisation and research on alternative materials. Use of direct seawater/wastewater.
Thermochemical Cycles	3–4	0.87–1.18	Severe reactor corrosion. Toxicity potential from acidic intermediates	Use of different non-corrosive materials. Closed-loop leak prevention.
Photocatalysis	3–4	0.43–63.1	Catalyst manufacturing burdens Low solar-to-hydrogen efficiency	Use of earth-abundant catalysts Low-energy and scalable synthesis routes for catalysts.
Biomass-based
Thermochemical Processes (Gasification and Pyrolysis)	5–7	−20.2–14.4 **	Efficiency dependent on biomass composition (high feedstock variability). Geographical logistics. Higher acidification potential.	Process optimisation Decentralised facilities near feedstock source. Coupling with CCS can result in negative emissions.
Biological Pathways (Fermentation and MEC)	3–4	0.7–21.8	Inherent biological yield limits. Energy-intensive pretreatment of feedstock. Gas purification.	Genetic strain optimisation. Coupling sequential systems (DF + MEC).
Geological
Natural (White H2)	1–3	***	Subsurface toxic gas release. Hydrological disruption. Ecological footprint of drilling.	Rigorous site selection. Safe leveraging of existing natural gas infrastructure.

* Assumes 100% renewable electricity coupling. Grid-coupled electrolysis can exceed 26 kg CO₂-eq. ** Negative values achievable when coupled with carbon capture and storage (CCS) or when accounting for CO₂ sequestration during the agricultural stage. *** Highly dependent on the natural geological reservoir and extraction methodology. Note: The GWP ranges reported in this table are drawn from studies conducted under heterogeneous system boundaries, functional units, and geographical assumptions, as detailed in Table 3, Table 5, Table 7 and Table 9. Values should therefore be interpreted as indicative ranges rather than directly comparable point estimates. The TRLs reflect current state of the art at the time of writing.

A critical limitation applies to the comparisons presented throughout this review. The LCA results collected from the literature were computed under heterogeneous methodological assumptions: system boundaries range from gate-to-gate to cradle-to-gate, cradle-to-grave and well-to-pump. Functional units vary from 1 kg of H₂ at unspecified conditions to 1 kg of compressed H₂ at defined pressure and purity, up to plant-level units such as 500 Mt H₂/year or 1000 tonnes over the electrolyser lifetime. Grid geographies range from national averages (Germany 2019, U.S. 2019, U.S. 2016, U.S. 2022) to regional mixes (Europe 2030) and dedicated renewable configurations (solar PV, wind, or a combination of both). Technology maturity varies from TRL 8–9 (AEL, PEMEL) to TRL 3–4 (photocatalysis, thermochemical cycles, biological routes, natural H₂). These differences are explicitly reported, where available, in the “System Boundaries/Functional Unit” and “Grid Geography or RES” columns of Table 3, Table 5, Table 7 and Table 9. Consequently, the numerical ranges reported in this review should be interpreted as indicative of the order of magnitude and dominant environmental hotspots of each pathway, rather than as directly comparable point estimates.

Evaluating these technologies strictly through the lens of carbon emissions (GWP) is insufficient for sustainable scale-up. The literature reveals three major types of eco-design trade-offs:

• High-maturity technologies, namely AEL and PEM electrolysis, may be constrained by their supply chain. The deep decarbonisation offered by these technologies comes at the cost of high mineral resource scarcity potential (MRSP), relying heavily on iridium, platinum, and titanium (PEM) and on other metals such as nickel (AEL). Dematerialisation and recycling efforts are therefore an eco-design imperative for electrolysis. Moreover, without coupling to RESs, electrolysis cannot be considered a low-emission pathway to produce H₂.
• Biomass thermochemical pathways offer the highly attractive proposition of carbon-neutral (or even negative) H₂ production while simultaneously valorising organic waste. However, their physical implementation is severely constrained by logistics. Because these feedstocks are bulky and possess high moisture contents, their transportation and pretreatment introduce substantial energy penalties. On the other hand, biological processes present an interesting alternative, but constraints on production yields and process scalability remain a demanding challenge that must be overcome if this pathway is to be exploited.
• Thermochemical water splitting cycles (TWSCs) and natural hydrogen extraction drastically reduce the need for external electricity by utilising direct heat or geological forces. However, this energetic advantage is offset by severe chemical risks. TWSCs require managing highly corrosive and toxic intermediates at extreme temperatures, while natural hydrogen extraction risks contaminating groundwater aquifers with hydrogen sulphide. Both pathways require safe infrastructure to prevent their low-carbon benefits from being negated by ecological toxicity.

Ultimately, a resilient global hydrogen economy will not rely on a single pathway. Instead, it will deploy a combination of these technologies, matching the specific eco-design strengths of a process to the geographic and resource realities of the deployment site.

5. Hydrogen Storage

H₂ can be stored in different ways, and it can be classified into two main categories: physical-based and material-based storage. Compressed gas, cryo-compressed H₂ and liquid H₂ belong to the first category. The material-based category can be divided according to two different phenomena: chemical sorption and physical sorption [113]. In the following sections, these types of storage and their associated environmental hotspots will be discussed.

5.1. Physical-Based Storage

Within the physical-based category, H₂ can be stored in its gaseous form. However, one fundamental difficulty across all methods is hydrogen’s exceptionally low volumetric energy density. With a lower heating value (LHV) of 9.9 MJ/m³, it contains approximately one-third of the energy present in an equivalent volume of methane (32.8 MJ/m³). Consequently, massive storage vessels or intense compression/cooling processes are required, introducing significant energy and environmental hotspots along the supply chain.

Compression can increase the volumetric density up to 40 kg/m³ [114], but the process incurs energy losses of around 10% of the energy content [115]. Compressed gas is typically stored in four types of vessel.

• Type I vessels are heavy cylinders made of metallic materials. They have poor mass storage efficiency (the weight of the stored H₂ corresponds to 1 wt.%) [115] and operate at a working pressure of 200–300 bar.
• Type II vessels are also made from metals but feature a thick metallic liner with a hoop of fibre resin positioned around the cylindrical part. They are around 30–40% lighter than Type I cylinders [115] and do not present limits on the storage pressure.
• Type III and Type IV cylinders are made from composite materials. Type III utilises a carbon fibre composite with an aluminium liner, operating at pressures up to 450 bar [113]. Meanwhile, Type IV consists of a polymer liner, such as high-density polyethylene [116], withstanding pressures of up to 1000 bar [113] and storing around 4.8 wt.% of H₂.

While Types III and IV are the lightest options, they are also the most expensive. It should be noted that all cylinder types (I to IV) generally present poor mass storage efficiency.

For liquid storage, the gas is liquefied by cooling it below temperatures of −253 °C. A major advantage of liquefaction is achieving a volumetric density of 70.8 kg/m³ [114]. Nonetheless, it is an extremely energy-intensive process, consuming over 30% of the energy content present in the H₂ itself [117]. Furthermore, maintaining these cryogenic temperatures requires a constant energy input, leading to inevitable “boil-off” losses during both storage and transport of H₂ [116]. Even though liquefaction of H₂ is a highly energetic process with substantial energy losses, it remains one of the preferred alternatives for intercontinental transport of H₂ [115].

Cryo-compressed H₂ is an alternative for storing gaseous H₂ at cryogenic temperatures (−233 °C). It can reach a volumetric density of 80 kg/m³, even higher than that of liquid H₂ [113], and was developed to minimise the boil-off present in liquefied H₂ storage [115]. To store it, a Type III tank is necessary, equipped with an insulated jacket that limits the heat transfer between H₂ and the environment [115], which consequently increases the cost of the tank. Even though cryo-compressed H₂ seems like a suitable option to store H₂, the required energy input increases with respect to the previous storage methods, as it combines the two main processes described, cooling down H₂ and compressing it.

5.2. Material-Based Storage and Chemical Carriers

Within H₂ material-based storage, chemical sorption comprises the integration of H₂ atoms into the chemical structure of the material [113]. This category includes ammonia, metal hydrides, formic acid, and liquid organic hydrogen carriers (LOHCs).

In metal hydrides, H₂ atoms are dissociated and diffused into the solid surface of the material, creating a stable and reversible bond. The metal hydride material is subsequently heated, and H₂ is released, allowing the discharged material to be reused for a new H₂ cycle [116]. Chemical sorption with metal hydrides presents advantages such as the possibility to store H₂ under normal conditions and controlled release. However, major challenges like enhancing charge–discharge kinetics, preventing the formation of unwanted gases during desorption, and high costs need to be overcome [113].

LOHCs are organic compounds that can store H₂ through catalytic hydrogenation and dehydrogenation cycles. This permits H₂ storage at ambient conditions, avoiding the need for high-pressure compression or low temperatures, and offers the potential to reuse the liquid carrier [118]. Because current LOHCs still present low H₂ capacity, researchers are actively seeking compounds with enhanced H₂ storage capacities and optimal thermodynamic properties [116]. The maximum storage capacities reported (in weight-to-weight percentage, wt.%) for ammonia borane, metal hydrides, formic acid and LOHCs are 19.4%, 12.6%, 4.4%, and 7.2%, respectively [113].

On the other hand, physical sorption relies on porous materials and on the physical attachment of H₂ to its surface through weak van der Waals forces [118]. The process occurs at low temperatures and does not involve chemical reactions. Large surface areas and porous structures are fundamental characteristics of materials that bind with H₂, with metal organic frameworks (MOFs) and porous carbon materials being the most common. Compared to chemical sorption, these materials present faster charge/discharge kinetics, lower hydrogen binding energy, and lower material costs [113]. Disadvantages include the weight of the carrier materials, low-temperature and high-pressure requirements, and low volumetric H₂ density [113]. The maximum storage capacities reported for carbon materials and zeolites are 8 wt.% and 9.2 wt.%, respectively [113]. Ultimately, even if material-based storage provides easier handling compared to gaseous and liquid H₂, it requires chemical processes that can be highly energy-intensive [116].

Bottlenecks of Solid-State H₂ Storage

Despite offering high volumetric densities and safer ambient storage than compressed or liquefied H₂, solid-state hydrogen storage remains constrained by a set of tightly coupled bottlenecks that limit commercial deployment. The central challenge is not hydrogen capacity alone but the simultaneous achievement of high usable capacity, moderate operating temperature, fast sorption kinetics, efficient heat transfer, reversibility, long cycle life, and low system cost [119]. Improving one property typically compromises another, producing a materials trade-off that differs across storage classes. Metal hydrides such as magnesium hydride (MgH₂) offer a high theoretical capacity of 7.6 wt.% but require dehydrogenation temperatures of 300–350 °C due to their high enthalpy (~76 kJ/mol) [119]. Hydrogen uptake and release are also often too slow, making them unsuitable for rapid refuelling applications. Nanostructuring and catalyst incorporation can improve kinetics but add inactive mass that dilutes gravimetric capacity [119].

Two system-level constraints cut across all material classes. First, heat management is intrinsically difficult: absorption is exothermic and desorption endothermic. A storage reactor therefore needs to remove heat during charging and supply heat during discharging. If external heat is required for release, a fraction of the stored H₂ may need to be burned to drive the reaction, reducing the effective energy density of the system [120]; MOFs further exhibit low thermal conductivity (~0.3 W m⁻¹ K⁻¹), limiting charge–discharge rates. Second, cycling degradation is a persistent issue: at the high temperatures required for H₂ release, repeated absorption–desorption cycles can cause capacity loss, catalyst sintering, phase separation, irreversible oxidation, particle coarsening, or structural collapse [119]. A clear pathway from laboratory-scale molecules to industrially manufacturable devices is therefore still lacking and constitutes one of the principal research directions identified in Section 8.

5.3. Large-Scale and Underground Storage

For larger quantities of H₂, large-scale underground storage in salt caverns and depleted hydrocarbon reservoirs has been considered. These are viewed as a solution to seasonal fluctuations, especially with surplus energy coming from renewables, to help balance the grid by storing excess energy in the form of H₂ [116].

However, this method faces several difficulties. H₂ can react with minerals or microorganisms present in underground formations, and the quality of H₂ can be compromised by reacting with salts [121]. Additionally, the geographical inaccessibility of salt caverns in proximity to where H₂ is being produced can be a limiting factor [122].

From an environmental and technical hotspot perspective, underground storage faces two major hurdles: hydrogen embrittlement, which can cause leakage or failure in the steel components of the storage wells and extraction infrastructure, and brine disposal. Environmentally, the construction of salt caverns requires massive brine extraction. Because this concentrated brine cannot be discharged into freshwater ecosystems, site selection is heavily constrained, often forcing facilities to be located within 50 km of the coast to facilitate sea disposal [112].

• Efficiency of Different H₂ Storage Methods

Staudt and colleagues [123] conducted an analysis comparing five different ways of transportation of low-emission H₂ and derivatives, namely, liquid hydrogen (LH₂), liquid methane (LNG), ammonia (NH₃), liquid organic hydrogen carrier (LOHC) and methanol (MeOH). They evaluated energy utilisation, defined as how much energy is required to transport H₂ or its derivatives to an importing country, considering ship transportation, direct utilisation of the derivatives, and conversion back to H₂.

The energy utilisation rates reported were 73% for LH₂ and 77% for ammonia used directly (which dropped to 63% when NH₃ was converted back into H₂ in the importing country). For LNG, an energy utilisation rate of 76% was obtained in the direct utilisation scenario, compared to just 51% when LNG was converted back to H₂. The same trend was observed with MeOH, showing 83% efficiency with direct utilisation and 62% when reconverted into H₂. Lastly, LOHCs presented an energy utilisation rate of 68%. For options that require a CO₂ source (MeOH and LNG), having a point source available in the producing country was decisive; if direct air capture was required, the efficiencies would be lower than the ones presented (56% for MeOH and 57% for LNG with direct utilisation).

Risco-Bravo et al. [124] presented a similar comparison between LOHCs and LH₂ in terms of energy available in H₂ form along the conversion chain. It was concluded that approximately 41% of energy was lost during the LOHC process, whereas around 21% of the energy in H₂ form was lost for LH₂.

2.

Environmental Assessment of H₂ Storage

An LCA study by Altuntas-Vahapo and Erdemir [125] compared the material compositions, technical characteristics, and environmental impacts of various physical hydrogen storage tanks, using a functional unit of 1 kg of stored H₂. The system boundaries encompassed material extraction, processing, and tank production. The results demonstrate that, for all compressed gas tanks (Types I–V), embodied emissions scale directly with storage pressure, as higher pressures necessitate thicker walls and additional material reinforcement. Tanks incorporating composite materials (Types II–V) incur significant environmental burdens driven by the high energy intensity of carbon fibre production. Consequently, while Type I steel tanks are the heaviest, they emerged as the lowest-impact option for stationary applications where weight is not a constraint. Conversely, Type IV tanks offer an optimal compromise for mobile applications, yielding lower emissions than other composite variants due to the substitution of metallic liners with lightweight polymers. Liquid storage provides superior volumetric density (occupying only ~15 L/kg H₂) and exhibits moderate material emissions compared to high-pressure composite tanks; however, its total life cycle footprint increases dramatically when the highly energy-intensive liquefaction process is accounted for.

Shifting to chemical carriers, Alghool et al. [126] presented a comparative LCA evaluating various H₂ storage and transportation pathways. The findings indicate that the initial conditioning stage of compression (0.57 kWh/kg H₂) is the least energy-intensive, whereas ammonia synthesis (11.2 kWh/kg H₂), liquefaction (10 kWh/kg H₂) and CO₂ hydrogenation (10.9 kWh/kg H₂) demand substantially higher energy consumption. During the storage phase, methanol exhibits the lowest climate impact as it remains stable as a liquid at ambient temperatures, unlike liquid H₂, which requires continuous energy to maintain cryogenic conditions (−253 °C).

For transportation, denser carriers like methanol and ammonia enable the movement of larger quantities; however, their added mass inherently increases the fuel consumption of transport vessels. Ultimately, the reconditioning stage emerges as the primary environmental hotspot for chemical carriers. CO₂ dehydrogenation accounts for 56.6% of the methanol pathway’s total GWP, primarily because it relies on burning natural gas for thermal energy, while ammonia cracking contributes 29.7% to its respective pathway’s GWP. Overall, the compressed H₂ pathway was identified as the most environmentally friendly delivery method in that assessment, as it avoids the energy-intensive reconditioning processes required by chemical carriers. The values are conditional on the overseas shipping configuration and the electricity source for production assumed by the authors.

Taken together, Table 11 and Table 12 indicate that no single storage route is universally optimal: for stationary applications where mass is not a constraint, Type I steel tanks offer the lowest embodied emissions, whereas for mobile applications Type IV tanks represent the best compromise between weight and life cycle burden. Among chemical carriers, compressed H₂ (2.67 kg CO₂-eq/kg H₂) showed lower impacts than liquid H₂, ammonia, and methanol (3.93, 5.09, and 9.60 kg CO₂-eq/kg H₂, respectively), under the conditions assessed, as the latter incur their principal environmental penalty during the energy-intensive reconditioning stage. It should be noted that Table 11 and Table 12 each derive from a single source and apply different system boundaries, tank manufacturing only in Table 11, and complete production-to-delivery chains in Table 12, so values may be compared within each table but not across them.

Table 11. GWP for the production of different tanks used for H₂ storage. Table prepared by the authors based on data reported in [125].

Tank Type	Materials	Working Pressure (Bar)	Key Characteristics and Features	GWP (kg CO2-eq/kg H2) a
Type I	Stainless steel	100–200	Fully metallic; heaviest design; lowest cost; high thermal safety	81.6–142.8
Type II	Steel liner + carbon fibre composite	200–500	Hybrid construction; ~40% lighter than Type I; moderate cost	271.1–832.6
Type III	Aluminium liner + CFRP b	400–700	Lightweight; corrosion-resistant; good fatigue performance; high cost	308.4–849.1
Type IV	Polymer liner (HDPE/PA) c + CFRP	700–1000	Very lightweight; no metal liner; best gravimetric efficiency	231.0–588.6
Type V	Full CFRP composite (no liner)	700–1000	Ultralight; non-metallic; highest cost; limited commercial maturity	347.6–893.9
Liquid (cryogenic)	Stainless steel + multilayer insulation	1–10	Stored at −253 °C; highest volumetric density; requires thermal management	179.6–276.0

^a The range presented for Types I to V corresponds to the GWP of H₂ stored at 100 bar and 1000 bar. For liquid storage of H₂ the range corresponds to 1 to 10 bar. ^b CFRP: carbon-fibre-reinforced polymer. ^c HDPE/PA: high-density polyethylene/polyamide.

Table 12. Climate Change Potential of different H₂ supply network pathways. Table prepared by the authors based on data reported in [126].

Carrier	Conditioning Stage	Storage Stage	Transportation Stage	Reconditioning Stage	GWP (kg CO2-eq/kg H2) a
Compressed H2 (CH2)	Compression of H2 to 70 bar	Stored in compressed H2 storage tanks	Overseas transport via compressed H2 containers	None; delivered to consumers as compressed gas	2.67
Liquid H2 (LH2)	Liquefaction: H2 is cooled to a liquid state at −253 °C	Stored in liquid H2 storage tanks; requires continuous cooling; ~0.18% daily boil-off loss	Overseas transport via liquid H2 tankers	Regasification: converting liquid H2 to gas at the consumer end	3.93
Ammonia (NH3)	H2 is synthesised via the Haber–Bosch process to obtain ammonia	Stored in ammonia storage tanks; ~0.062% daily losses	Overseas transport via ammonia tankers	Ammonia cracking: chemically decomposing ammonia to release H2	5.09
Methanol (CH3OH)	CO2 hydrogenation: hydrogen is bound to captured CO2 to form methanol	Stored as a stable liquid at ambient temperature and pressure	Overseas transport via methanol tankers	CO2 dehydrogenation: high-heat endothermic process to release H2	9.60

^a GWP values correspond to H₂ production using concentrated photovoltaic thermal (CPVT) collectors coupled to electrolysis. Total impacts are higher if electrolysis is coupled to standard PV panels or photovoltaic thermal (PV-T) collectors.

In summary, the eco-design of hydrogen storage is defined by a compromise between volumetric density, material burdens, and energy hotspots. While physical compression requires large quantities of material, including composites to reduce weight, cryogenic and chemical storage methods achieve high densities at the cost of immense energy requirements for cooling, synthesis, and reconversion. Therefore, selecting the most sustainable storage medium depends largely on the end-use application, favouring heavy steel tanks for stationary use and compact chemical carriers for long-distance mobility. These storage constraints directly dictate the logistical challenges of the subsequent transportation phase.

6. Transportation and Distribution

To enable the deployment of the H₂ economy, transportation and distribution play an important role in delivering H₂ from the production site to the point of use or storage [127]. Three usual ways to transport H₂ are compressed gas in tube trailers, cryogenic liquid tankers, and pipelines. Depending on the quantity and distance, there are preferred methods of transportation; compressed gas is recommended for small volumes and short distances, while liquid or solid storage forms are preferred for higher amounts of H₂ and longer distances [124].

6.1. Compressed H₂ in Tube Trailers

When H₂ gas is compressed, various pressurised gas cylinders (in the range of 200–500 bar) are placed in trucks known as tube trailers [128]. The maximum quantity of H₂ transported is 540 kg per trailer for short distances (200–300 km) [124]. While this is a well-established technology for H₂ transport, it suffers from severe inefficiencies, transporting around 1% of weight as compressed H₂ [124].

6.2. Cryogenic Liquid in Tankers

When H₂ is transported as a cryogenic liquid, tankers and ships serve as transportation methods. Due to the low temperatures required (−253 °C), effective thermal insulation of the transportation method is critical to prevent boil-off [127], a phenomenon consisting of the gradual evaporation of liquid H₂ and the subsequent increase in vessel pressure [118]. Compared to tube trailers, a higher quantity of H₂ can be transported when it is a cryogenic liquid, around 4000 kg of H₂ per tanker can be moved over longer distances (4000 km) [124].

6.3. Pipelines

Lastly, pipelines are widely regarded as the most promising transport method where continuous, large-scale distribution is required, as they offer high throughput at low specific energy demand [124]. Currently, H₂ infrastructure consists of around 5000 km of pipelines already in operation worldwide, located mainly in the United States and Europe [7]. Compared to natural gas pipelines, with a length of operating pipelines of around 1 million km, it is evident that H₂ infrastructure is still too small to consider it for large-scale distribution, necessitating efforts to expand new infrastructure or repurpose existing pipelines.

• Blending H₂ with Natural Gas

An opportunity to use already built pipelines for H₂ transport has been proposed by blending it into natural gas infrastructure, especially as natural gas consumption is expected to decline. Yet, physical and chemical characteristics like molecular size, density, and flame colour complicate the introduction of H₂ within existing pipelines.

One of the main challenges faced with H₂ transport in existing pipelines is the process of embrittlement or H₂-accelerated fatigue cracking (HAFC). Owing to the small size and high diffusion coefficient of the molecule, H₂ atoms diffuse into the steel. Another obstacle relates to the volumetric density of H₂, which is a third of that of natural gas. To deliver the same energy given by a volumetric flow rate of natural gas, the volumetric flow rate of H₂ must increase, resulting in higher energy consumption by compression and therefore undermining the climate benefits of H₂ [121]. In addition, safety and security concerns regarding the transport of H₂ in pipelines have arisen because H₂ gas is odourless and its flame is invisible in daylight, making leakage and permeation critical aspects to consider.

Blending H₂ with natural gas has been tested to determine the maximum H₂ concentration before the integrity of the pipeline is compromised. In the U.S. natural gas pipeline system, only minor issues arise when blending H₂ in the range from 5–15% by volume [129] and up to 10% in Germany [124]. Generally, lower concentrations of H₂ can be blended without negatively affecting end-users or the pipeline infrastructure. The HyDeploy project [130] in the United Kingdom demonstrated that H₂ blending was feasible, safely operating domestic appliances with up to a 28.4% H₂ blend, resulting in a reported CO₂ emission reduction of up to 0.5%.

Evidently, blending is not the optimal approach to lowering GHG emissions. It is limited by its lower volumetric energy density relative to the gas that it replaces [121], and it adds a further separation and purification step to deliver high-purity H₂ to end-users, inevitably increasing energy consumption and further reducing the environmental benefits of low-emission H₂.

2.

Environmental Assessment of H₂ Transportation

Demir and Dincer [131] evaluated three scenarios for H₂ transmission and distribution: the first scenario considered liquid H₂ transported in liquid tankers; the second, H₂ compression and transport via tube trailers; and the last, a pipeline network. The analysis began after the first compression stage at the exit of the hydrogen production plant, meaning the reported values reflect the transmission and distribution stage only, excluding production burdens. Emission factors for the electricity consumed during compression, liquefaction, and transport were derived from GREET model projections for 2020, which assume a grid-average electricity carbon intensity of 204 g CO₂ per MJ. The authors reported pipeline transport as the most environmentally favourable, at 28 g CO₂-eq per MJ of H₂ delivered (equivalent to 3.36 kg CO₂-eq/kg H₂ taking the LHV of hydrogen), approximately 26 times lower than tube trailers and over two orders of magnitude lower than liquid tankers under the conditions modelled. Because this analysis excludes the production stage and relies on an average grid electricity for its energy demand, is not directly comparable to the production-to-fuelling station boundary applied by Wulf et al. [132] described below, where hydrogen production is included with wind power as the electricity source for electrolysis, and a future electricity mix for Germany used throughout the chain.

To comprehensively evaluate the environmental impacts of the entire hydrogen delivery network, Wulf et al. [132] conducted an LCA based on a functional unit of 1 kg of high-purity H₂ delivered to a fuelling station at 700 bar. The system boundaries accounted for hydrogen production (via electrolysis), intermediate storage (in salt caverns for gaseous H₂ or specialised tanks for LOHCs), transportation (via pipeline, compressed gas trucks, or LOHC trucks), and final compression at the fuelling station. The study highlights that the intermediate seasonal storage, whether utilising underground salt caverns for gaseous H₂ or specialised tanks for liquid organic hydrogen carriers (LOHCs), contributes minimally to the environmental burden, representing only 1.5% to 3.5% of the overall supply chain GWP. For the transportation phase, the authors evaluated three distinct pathways: dedicated pipelines featuring a special coating to prevent hydrogen diffusion; gaseous H₂ trucks, where the gas is compressed to 500 bar and transported in fibre-reinforced composite tanks; and LOHC trucks, where hydrogen is chemically bound to dibenzyl toluene.

Under the conditions modelled (production rate of 40 tonnes per day over a transport distance of 100 km), pipeline transport was identified as the most environmentally favourable option across almost all scenarios, benefiting from significant infrastructural scaling effects and lower energy demands for recompression compared to the diesel fuel consumption of transport trucks. Conversely, the LOHC pathway exhibited the highest GWP, primarily driven by the energy-intensive dehydrogenation stage. Within the LOHC transportation phase, approximately 83% of the generated emissions are exclusively attributed to the thermal heat provision required to release the H₂ from the chemical carrier. Table 13 presents the comparative GWP results for these hydrogen supply pathways.

Table 13. Comparative Global Warming Potential (GWP) of hydrogen delivery pathways. Table prepared by the authors based on data reported in [132].

H2	Storage	Transportation	Dispensing	GWP (kg CO2-eq/kg H2)
Gaseous H2	GH2 Cavern	Pipeline (100 bar)	Fuelling Station (700 bar)	1.52
Gaseous H2	GH2 Cavern	Truck (500 bar)	Fuelling Station (700 bar)	2.52
LOHC	LOHC Storage Tank	LOHC Truck	Fuelling Station (700 bar)	4.96

In conclusion, transporting and delivering hydrogen is a major bottleneck that can reduce the environmental benefits of clean production. Although dedicated or repurposed pipelines are among the most efficient and environmentally favourable delivery methods at high throughput and over short to medium distances, their limited global reach forces a reliance on higher-emission alternatives for now. Moving hydrogen via cryogenic ships, tube trailers, or natural gas blending networks creates severe life cycle penalties due to energy-intensive steps like liquefaction, transport fuel consumption, and downstream gas separation. To build a truly sustainable hydrogen economy, future planning must focus on establishing local, decentralised production hubs near end-users or rapidly expanding large-scale pipeline networks to eliminate these midstream supply chain hotspots.

7. Hydrogen Costs

A complete sustainability assessment of the H₂ value chain cannot be limited to environmental performance; the economic dimension is an equally essential pillar of sustainability. Levelised cost of hydrogen (LCOH) figures shape a technology’s eco-design profile, and a pathway that is technically promising but economically uncompetitive will struggle to achieve the deployment scale required to deliver its environmental benefits. For this reason, environmental and economic indicators should be considered jointly, particularly when guiding investment decisions or policy design.

For several components of the value chain, environmental impacts and economic constraints are closely interlinked. In low-emission electrolysis pathways, electrolyser equipment accounts for approximately 55–65% of the production LCOH [5]. This capital expenditure (CAPEX) is strongly influenced by the material criticality discussed in earlier sections: for proton exchange membrane electrolysis (PEMEL), the reliance on critical raw materials such as iridium and platinum simultaneously drives up manufacturing costs and increases the environmental footprint of stack production, which has been reported as significantly higher than that of alkaline electrolysis (AEL) or solid-oxide electrolysis (SOEL). On the operational side, electricity consumption is the single largest cost driver, mirroring the central role that the electricity grid mix plays in determining the operational environmental impact of electrolysis.

Downstream operations exhibit a similar entanglement between cost and impact. Storage, compression, and transportation introduce significant additional costs to the delivered price of H₂, while at the same time generating hotspots for life cycle emissions through liquefaction, transport fuel consumption, and downstream gas separation. Recognising both the convergences and the divergences between economic and environmental drivers is therefore essential to avoid overstating either dimension. Table 14 consolidates 2024 cost ranges and 2030 projections for the principal hydrogen production technologies and presents the corresponding cost and economic considerations for storage and transport options, mapping each cost driver to its eco-design hotspot.

Table 14. Integrated economic overview and eco-design hotspots across the hydrogen value chain (2024 and 2030 Projections). Data adapted from the IEA Global Hydrogen Review 2025 [5] if not stated otherwise.

Technology	2024 Cost Range (USD/kg H2)	2030 Projected Cost Range (USD/kg H2)	Integrated Eco-Design and Economic Hotspots
Natural Gas (w/o CCS)	0.8–4.6	0.7–3.7	Costs are highly sensitive to local natural gas prices. Low cost in gas-producing regions (USD 0.7–1) is offset by high direct GWP, facing steep financial carbon tax penalties in regions like Europe.
Natural Gas (with CCS)	~2.0–5.0	Dependent on regional variation	Regional variation is heavily driven by both the underlying gas price and the local cost of CO2 CAPEX driven by CO2 capture structure. While it reduces direct emissions it increases operational water withdrawal requirements.
Coal Gasification (w/o CCS)	<1.0–2.0	Remains competitive in areas rich in coal	Dominant and highly economical in China (USD 1.5/kg cheaper than local renewables). However, it stands as the most carbon-intensive and water-intensive pathway reviewed
Electrolysis (China)	~4.0–6.0	<2.0 (Optimal regions)	Lowest LCOH globally due to exceptionally low electrolyser CAPEX cost (USD 900/kW vs. USD 2300/kW globally) and cheap electricity cost. Operational emissions remain strictly tied to the grid mix profile.
Electrolysis (Europe)	~4.0–10.0	<3.5 a (<3.0 EUR/kg H2) [3]	Heavily constrained by regional electricity prices and price of critical raw materials. Costs range dramatically depending on the renewable source, from southern solar to northwestern offshore wind.
Electrolysis (Australia, Chile, Brazil, Middle East, United States)	Not defined individually	3–4	Excellent renewable resource profiles pair favourably with a low cost of capital. However, localised water-stressed regions often necessitate additional operations like seawater desalination, increasing costs.
Storage	Cost Range (USD/kg H2)	Economic Considerations	Integrated Eco-design and Economic Hotspots
Salt Caverns (Large Scale)	0.6 [133]	Low levelised cost over time	Provides excellent operational flexibility through frequent cycling. Compromised by the technical risk of hydrogen embrittlement and heavy localised environmental impacts from high-volume coastal brine disposal.
Transport	2024 Cost Range (USD/kg H2)	Economic Considerations	Integrated Eco-design and Economic Hotspots
New H2 Pipelines (Large Scale)	1.0/kg H2 b [133]	Cheaper option to transport H2 gas for distances below around 1500 km	Reported as the transport option with the lowest environmental footprint at high throughput over short to medium distances. Economies of scale are massive, but new pipelines require high capital; repurposing existing gas pipes cuts investment by ~80% but could increase downstream separation costs (if blended with NG).
Liquid Hydrogen Shipping	Liquid H2: 2.0 Ammonia: 1.2 LOHC: 0.6 [133]	Above 1500 km, LOHC and ammonia transport by ship become cheaper delivery options	Achieves excellent volumetric density but requires vastly more expensive cryogenic loading and storage systems (−253 °C) at port terminals, inducing a severe 21% energy loss.
Compressed H2 Truck Delivery	H2 gas: >1.5–2.0 Liquid H2: >0.5 LCOH: 2.9 Ammonia: 1.5 Transport distance: 500 km [133]	If pure H2 is required, the additional cost of extracting H2 from ammonia or an LOHC must be included	The cost of distributing LOHCs by truck is around USD 0.8/kg H2, however, the cost of extracting and purifying H2 at the point of use would be USD 2.1/kg H2. This highlights that both costs and environmental impacts increase with chemical carriers.

^a EUR–USD conversion at 1.16 USD/EUR. ^b Transporting H₂ as a gas over around 1500 km in pipelines, taking into account all capital and operating costs.

Ultimately, the cheapest hydrogen at the point of production can easily become the most expensive and carbon-intensive energy vector by the time it reaches the end-user. Bridging the gap between laboratory success and industrial viability requires project developers to look beyond isolated electrolyser costs and prioritise localised production loops or repurposed pipeline backbones that eliminate energy-intensive midstream conversions.

8. Research Gaps and Future Directions

Despite the increasing body of literature on low-emission H₂, several methodological, technological, and systemic gaps persist that hinder rigorous sustainability assessment and impede the transition from laboratory to industrial deployment. Table 15 consolidates the principal gaps identified across the production, storage, transportation, and economic dimensions analysed in this review, alongside the eco-design principle each gap relates to and the corresponding research direction proposed in the literature.

Table 15. Knowledge gaps and proposed research directions for a sustainable H₂ value chain.

Value-Chain Stage	Identified Gap	Related Eco-Design Principle	Proposed Research Direction
Electrolysis	Reliance on outdated or laboratory-scale inventory data for electrolyser materials; limited primary data from industrial-scale stacks.	Resource and energy efficiency	Collection of primary inventory data from operational industrial electrolysers. Transparent reporting of stack composition and balance of plant.
	Dependence on critical raw materials with low recycling input rates.	Resource efficiency and design for recyclability	Development of PGM-free catalysts or dematerialisation. End-of-life recovery strategies for electrolyser stacks.
	Reliance on ultrapure freshwater; water-stressed regions present impediments.	Resource efficiency	Direct seawater electrolysis, integration with municipal wastewater treatment, and coupling with desalination using renewable energy.
Thermochemical Water Splitting	Material degradation under highly corrosive acidic intermediates.	Design for durability	Development of corrosion-resistant reactor materials; closed-loop strategies; awareness of toxicity by the cycle selection.
Photocatalysis	Sharp divergence in reported LCA results, driven by catalyst synthesis intensity and lack of standardised LCA at low TRL.	Resource efficiency and life cycle thinking	Use of earth-abundant photocatalysts. Low-energy, scalable synthesis routes. LCA frameworks suited to low-TRL technologies.
Biomass Thermochemical	Feedstock heterogeneity and by-product formation reduce H2 yield and complicate process design; geographical mismatch between feedstock and end-user.	Resource efficiency and life cycle thinking	Process optimisation on catalysts and working conditions to minimise by-products and enhance efficiency. Decentralised H2 production. Adjacent feedstock production hubs.
Biological Processes	Low H2 yields, slow production rates, energy-intensive pretreatment and gas purification limit industrial scalability.	Energy efficiency	Genetic optimisation of microbial strains. Process optimisation to minimise energy requirements for pretreatment and gas purification. Integration of biological processes (dark/photofermentation and MEC systems).
Natural Hydrogen	Scarce LCA data; environmental impacts poorly quantified.	Minimisation of environmental impacts	Dedicated LCA studies for natural H2; rigorous site-selection criteria and environmental monitoring frameworks.
Storage	Trade-off between volumetric density (liquefied H2, chemical carriers) and energy/material burdens (favouring compression) not yet resolved through eco-design.	Energy efficiency and resource efficiency	Application-specific storage selection. Chemical carriers that do not require high energy for reconditioning stage. Catalysts based on abundant elements.
Transport and Distribution	Limited dedicated H2 pipeline network (~5000 km globally vs. ~1,000,000 km of natural gas pipelines); blending introduces downstream separation.	Energy efficiency and life cycle thinking	Strategic expansion and repurposing of pipeline infrastructure; localised production hubs to minimise long-distance transport.
Impact Coverage	Prominent focus on global warming potential; non-GWP indicators (acidification, toxicity, mineral scarcity, water deprivation, eutrophication) reported unevenly.	Life cycle thinking	Reporting all impact categories. Include life cycle sustainability assessment (LCSA) that integrates economic and social dimensions.
Economic–Environmental Coupling	Economic and environmental indicators are typically reported in isolation, despite shared drivers (material criticality, electricity cost/grid mix).	Life cycle thinking	Integrated techno-economic and environmental assessments (TEA-LCAs); explicit mapping of LCOH drivers to eco-design hotspots.

Recognising that H₂ production itself is an energy-demanding process, it is crucial that the entire supply chain is designed to maximise energy efficiency and minimise losses. To achieve a truly sustainable H₂ chain, it is essential that not only the production processes but also all stages of the value chain are aligned with reducing environmental burdens. Making use of different tools and software to acquire data related not only to environmental impacts but also to energy and resource efficiency will allow a case-by-case analysis of the optimal configurations for designing an environmentally friendly H₂ value chain. These gaps point to a common requirement: the future of sustainable H₂ assessment depends on integrating environmental, economic, and methodological perspectives within a unified eco-design framework, rather than treating them as individual exercises.

9. Conclusions

This review analysed the low-emission H₂ value chain through an eco-design lens, addressing the four objectives outlined in the Introduction: synthesising the state of the art of production pathways, benchmarking their environmental performance, identifying eco-design hotspots and trade-offs along the entire chain, and consolidating the methodological and technological gaps that currently constrain a rigorous sustainability assessment. The evidence collected from recent LCA literature confirms that no single pathway is universally “green”; each technology shifts the environmental burden along a different dimension—energy, materials, water, or land—and only a value-chain perspective can identify where meaningful improvements are achievable.

Among production pathways, electrolysis remains the most mature renewable option, but its sustainability is dictated almost entirely by the electricity source: GWP can exceed ~30 kg CO₂-eq/kg H₂ on a fossil-based grid and fall below 1 kg CO₂-eq/kg H₂ when coupled to dedicated wind or solar plants. Biomass thermochemical routes can reach negative emissions when combined with CCS (down to −15 kg CO₂-eq/kg H₂), yet face acidification and feedstock-logistics constraints. Photocatalysis and biological processes show low theoretical carbon footprints (0.43 and 0.7 kg CO₂-eq/kg H₂, respectively) but remain limited by catalyst synthesis burdens and low TRL. Thermochemical cycles and natural hydrogen offer compelling low-carbon alternatives but introduce material corrosion, toxicity, and subsurface ecological risks that demand careful site and material selection. Downstream, liquefaction consumes more than 30% of the H₂ energy content, while pipeline-based supply chain delivery (~1.5 kg CO₂-eq/kg H₂) showed lower impacts than cryogenic shipping and LOHC pathways in the studies reviewed, confirming storage and transport as critical hotspots of the value chain.

Transversal analysis reveals four eco-design priorities: decarbonising the electricity grid or using dedicated renewable energy plants for H₂ production, which will enable low-emission H₂ production from electrolysers; reducing dependence on critical raw materials through dematerialisation and improved recycling (currently below 16% for nickel and PGMs); prioritising direct H₂ utilisation to avoid cumulative energy penalties from compression, liquefaction, and reconversion; and recognising that environmental and economic competitiveness are closely intertwined.

As consolidated in the previous section, key knowledge gaps remain: harmonised LCA boundaries, systematic reporting of non-GWP indicators, primary industrial data, and dedicated assessment frameworks for low-TRL pathways. Addressing these will require integrated environmental and economic methodologies (TEA-LCA, LCSA) embedded within a unified eco-design framework. Ultimately, a sustainable H₂ economy will not rest on a single “best” technology but on matching the eco-design strengths of each pathway to the local energy, water, and material realities of its deployment context.