## Introduction

Plastic is a crucial material in many sectors, including construction, packaging, transportation, electronics, textiles, and others1,2. The last half century has witnessed the rapidly increasing demand and production of plastics1, resulting in considerable plastic waste due to the low plastic recycling rate. From 1950 to 2015, only 9% of the cumulative plastic waste generation (6300 million metric tons (Mt)) was recycled, compared to over 60% discarded (accumulating in landfills or in the natural environment)1. The landfilled or disposed plastic wastes and their fragments, i.e., microplastics and nanoplastics, have caused increasing environmental concerns3,4,5,6. Increasing plastic recycling is one essential strategy to reduce plastic waste disposal7. There are two common types of plastic recycling, mechanical (e.g., magnetic density separation) and chemical recycling (e.g., gasification)7. Recently, another type of recycling method, solvent-based recycling (or referred as physical recycling), is also attracting attention8. The challenges of plastic mechanical recycling include thermal-mechanical degradation (e.g., caused by heating and mechanical shearing of polymer)7, plastic degradation (e.g., caused by photo-oxidation process during lifetime), incompatibility among different polymers when recycling blended plastics9, and contaminations (e.g., coating, ink, additives, metal residues or cross-contaminations among different plastic streams)9,10. Some waste plastics are hard to be mechanically recycled due to low bulk density (e.g., films), lightweight (e.g., polystyrene (PS)), low economic value (e.g., PS), and carbon-black pigments that absorb infrared light and confound the sorting machine9,11. Hence, relying on the traditional mechanical recycling method alone is insufficient to address the increasing volume and variety of plastic waste. Compared to mechanical recycling, thermochemical methods, as one type of chemical recycling, have advantages in processing plastic wastes that are difficult to be depolymerized, or mechanically recycled due to economic or technical barriers7,12. Thermochemical processes include pyrolysis and gasification, which have potentials to treat waste plastics with high energy, carbon, and hydrogen content, and low moisture content13. Thermochemical processes can produce a variety of products, and hydrogen is one product with a mature and growing market14. Hydrogen is an important industrial gas widely used in the oil refining and chemical industries, it can also be used as a clean energy source for transportation15. U.S. Department of Energy (DOE) estimated the U.S. hydrogen demand as high as 22–41 Mt per year by 2050, given the enormous need for clean energy16. Currently, 96% of hydrogen production uses fossil fuel reforming (e.g., petroleum, natural gas, and coal)15. Converting MPW to hydrogen has the potential to reduce fossil fuel demand for hydrogen production and address worldwide challenges of rapidly growing plastic wastes17. For example, the U.S. DOE Hydrogen Program Plan highlighted “diverse domestic resources” including waste plastics as an important source of hydrogen production16.

Since most plastics are made from fossil fuels, it is necessary to mitigate the fossil-based carbon emissions during the thermochemical conversion of MPW to hydrogen1,15. Carbon capture and storage (CCS) is an important technology to mitigate climate change by capturing and geologically storing CO2 (ref. 18). Coupling hydrogen production with CCS offers a means to produce low-carbon hydrogen1,15. For the large-scale development and implementation of plastic recycling technologies, it is critical to understand the economic feasibility and environmental performance of plastic waste to hydrogen pathway with/without CCS and policy incentives, as well as to identify the key drivers and future improvement opportunities.

Techno-economic analysis (TEA) is one of the most widely used tools to assess the economic and technical feasibility of emerging technologies19,20,21,22,23,24; Life Cycle Assessment (LCA) is a standardized tool to quantify life-cycle environmental impacts25,26,27,28,29,30,31. Several studies have used TEA to evaluate the economic feasibility or LCA to assess the environmental implications of plastic wastes to energy products (see Supplementary Note 1 for literature review). However, few studies have explored the economic and environmental implications of MPW to hydrogen at a large scale with CCS, or investigated the drivers of the economic and environmental performance of MPW compared to single-stream recycled plastic.

To fill the knowledge gap, we conducted a TEA and LCA to evaluate the economic and environmental performance of hydrogen production from MPW and single-stream recycled plastic in the U.S. and identify cost reduction opportunities. A mechanistic process simulation model (see Fig. 1 for system boundary and plant process flow diagram, and see Methods for details) was developed in Aspen Plus32 to provide rigorous engineering estimation of mass and energy balance data used in TEA and LCA. The minimum hydrogen selling price (MHSP) was selected to assess the economic feasibility of the hydrogen plant33. For life-cycle environmental impact assessment (LCIA), TRACI 2.1 by U.S. Environmental Protection Agency (EPA) and Global Warming Potential (GWP) factors (100-year time horizon) by the Intergovernmental Panel on Climate Change (IPCC) 2021 (in the sixth assessment) are used34,35. Different scenarios were designed to examine the impacts of varied feedstock compositions, plant capacities, CCS adoption, and policy incentives. A sensitivity analysis was conducted to identify key drivers of production costs. Finally, an improvement analysis outlined the roadmap for reducing production costs by improving key technical and economic parameters. This study contributes to the fundamental understanding of the economic and environmental performance of MPW to hydrogen pathway, which will inform the waste management industry with economically and environmentally preferable system design and shed light on opportunities to reduce cost and environmental burden.

## Results

In this study, the scenario analysis was used to evaluate the impacts of feedstock types, plant capacities, CCS adoption, and carbon credits, as shown in Supplementary Table 1. Scenario 1 depicts the baseline cases without CCS; Scenario 2 describes the cases with CCS but no carbon credit available; Scenario 3 considers CCS and carbon credits for capturing and storing CO2. In each scenario, five different feedstock cases are included, namely one MPW case and four single-stream feedstock cases (polyethylene (PE), polyethylene terephthalate (PET), polypropylene (PP), and PS) (see Methods). Many studies have explored the thermochemical conversion of single-stream or mixed plastics, but few studies have compared the economic and environmental performance of hydrogen derived from gasifying the single-stream plastics and MPW13,17,36,37,38,39,40,41,42,43. For each feedstock case, the varied steam/feedstock ratios were studied to reach the optimal MHSP. Five plant capacities (100–2000 oven-dry metric ton (ODMT) per day of plastic fed in) are compared to explore the impacts of capacities on the MHSP. The capacities were selected based on the current estimation of plastic wastes landfilled in the U.S. The quantity of state-level landfilled plastic waste in 2019 exceeds 250,000 t year−1 (762 t day−1) in 35 states and exceeds 1000,000 t year−1 (3049 t day−1) in 12 states44.

### Economic competitiveness of plastic waste derived hydrogen

Figure 3 shows the MHSP of hydrogen plants at 2000 ODMT per day of plastic waste in three scenarios (see Supplementary Fig. 8 for the MHSP of varied capacities). The detailed results of capital investment and operating cost are available in Supplementary Notes 2 and 3, Supplementary Figs. 911. In Fig. 3, the feedstock cost range was collected from the literature (see Supplementary Note 4). Waste plastic-derived hydrogen is economically competitive when the MHSP is within the range of the current market price of hydrogen.

Without CCS, only the MPW case shows competitive MHSP (US$1.33–$2.00 kg−1 H2 for feedstock cost $0–$151 ODMT−1), compared with current fossil-based hydrogen, as shown in Fig. 3a. The economic competitiveness of other cases depends on the feedstock costs (except for PET and PS whose MHSP is always higher than the fossil-based H2). For example, PE needs a feedstock cost under US$236 ODMT−1 to be economically competitive; PP needs a feedstock cost under US$238 ODMT−1. These thresholds are towards the lower bounds of feedstock costs of PE, and PP, indicating the limited possibility of utilizing recycled single plastic streams for hydrogen production in most cases, given the high feedstock costs caused by expensive sorting and processing in MRF. Some strategies have been proposed in the literature to overcome the cost barriers, e.g., advocating for “design for recycling” to lower the recycling cost47, improving waste collection and separation infrastructures11, optimizing municipal waste collection systems prior to MRF48, and adopting cost-effective technologies (e.g., triboelectrostatic separation) in MRF49.

### Environmental impacts of plastic waste derived hydrogen

This study conducted the LCA to examine the environmental impacts of waste plastic-derived hydrogen. Figure 6 shows the normalized LCA results of ten impact categories in varied scenarios and feedstock cases under the optimal S/F ratios identified in Fig. 2. The LCA results of each impact category are normalized based on the highest value (on 1 kg H2 basis) of that impact across Scenario 1–3 (including 5 feedstock cases in Scenario 1 without CCS and 5 feedstock cases in Scenario 2 & 3 with CCS). The absolute values of LCA results are in Supplementary Tables 7 and 8 in Supplementary Data 1 (ref. 67).

In Fig. 6, MPW shows the lowest environmental impacts across all scenarios and impact categories (1–93% lower than the other four single-stream feedstocks) mainly due to the lower environmental burdens of feedstock collection, sorting (for single-stream cases), and transportation. Note that the burdens of producing plastic are assumed to be cut-off from the system boundary. Across most impact categories, feedstock collection, sorting, and transportation dominate the environmental impacts of hydrogen derived from single-stream plastic (27–94%), but they only contribute to 1–10% for MWP. The only exceptions are GWP and fossil fuel depletion that are dominated by energy, contributing to similar percentages of results for single-stream plastics and MPW (25–90%). MPW has 1–59% higher environmental burdens of chemicals and materials than that of PE, PP, and PS, given the additional steps in pretreatment and dechlorination. However, chemicals and materials overall only contribute to 1–32% of life cycle environmental impacts across all single-stream plastic feedstocks. Waste treatment has minor contributions to most impact categories except acidification and human health—carcinogenics, although MPW has 19–94% higher environmental burdens related to waste treatment than single-stream plastics. This is caused by the higher wastewater generation in pretreatment and dechlorination. Across single-stream plastics, PET shows the worst environmental performance, similar to TEA results for similar reasons – low hydrogen yields and high cost (environmental burdens) of sorting and processing plastic feedstock.

Adding CCS to the hydrogen plant increase all environmental impacts by 9–117% except reducing GWP by 42–67%, regardless of plastic feedstocks. The increased environmental impacts are attributed to the chemicals and energy consumption68,69, while the decreased GWP are contributed by CCS that removes carbon.

From the perspective of climate change, MPW-derived hydrogen without CCS has higher life cycle GWP (16.0–21.0 kg CO2e kg−1 H2, depending on S/F ratios, see Supplementary Table 9 for detailed values) than natural gas (9.0–12.3 kg CO2e kg−1 H2 (refs. 15,51,52,54,70)) but mostly lower than coal (20.0–26.0 kg CO2e kg−1 H2 (refs. 51,52,53,54)). CCS reduces the GWP of MPW-derived hydrogen to 5.1–6.2 kg CO2e kg−1 H2, which is much lower than fossil-based hydrogen without CCS. However, if CCS is implemented for fossil-based hydrogen in the future, MPW-derived hydrogen will have higher life-cycle GWP than natural gas-based hydrogen with CCS (1.0–4.1 kg CO2e kg−1 H2 (refs. 15,31,51)), and comparable with coal-based hydrogen with CCS (2.0–6.9 kg CO2e kg−1 H2 (refs. 51,52,53,54)) or biomass gasification hydrogen without CCS (0.3–19.2 kg CO2e kg−1 H2 (refs. 31,71,72,73)). MPW-derived hydrogen with CCS has lower life-cycle GWP than electrolysis hydrogen from global average grid electricity (25.5 kg CO2e kg−1 H2 (ref. 51)), although the GWP of MPW-derived hydrogen with CCS is higher than electrolysis hydrogen with clean electricity (0.9–6.9 kg CO2e kg−1 H2 (refs. 70,71)), or biomass gasification with CCS (−18.8 to −9.6 kg CO2e kg−1 H2 (refs. 31,71)). As most GHG emissions are attributed to energy consumption (Fig. 6b), future research should focus on improving energy efficiency and exploring alternative energy sources to reduce the life cycle GWP of MPW-derive hydrogen.

For other impact categories, this study compared Scenario 3-MPW with CCS with hydrogen made from natural gas using steam reforming and CCS (see Supplementary Fig. 12). The MPW with CCS is 2.4–80.3% lower than natural gas with CCS in acidification, fossil fuel depletion, ozone depletion, and smog formation. At the same time, hydrogen from natural gas with CCS is 26.8–53.6% lower than Scenario 3-MPW with CCS in carcinogenics, non-carcinogenics, ecotoxicity, eutrophication, and respiratory effects.

## Discussion

This study conducted a TEA and LCA to explore the economic feasibility and environmental performance of hydrogen production from gasifying MPW that commonly ends in landfill. The TEA and LCA were coupled with the process simulation model developed in Aspen Plus to determine the impacts of plant capacities, feedstock compositions, policy incentives, and process parameters on MHSP and life-cycle environmental impacts. It is economically feasible to produce US$1.67 kg−1 H2 from a 2000 ODMT per day hydrogen plant utilizing MPW without CCS, compared with the current fossil-based hydrogen price without CCS (US$0.91–$2.21 kg−1 H2). Incorporating CCS into the gasification plant increases most environmental impacts due to the additional chemical and energy consumption of CCS systems. The only exception is GWP given the carbon removal benefits of CCS. Adding CCS also increases the MHSP to US$2.60 kg−1 H2 (US$2.26–$2.94 kg−1 upon varied feedstock cost) for the same plant, and the economic feasibility of a CCS-coupled hydrogen plant depends on CCS cost and policy incentives. CCS is essential to ensure that MPW-derived hydrogen has lower life cycle GHG emissions than current fossil-based hydrogen, and this advantage may not hold if CCS is implemented for natural gas-based hydrogen in the future. Future research is needed to reduce energy-related carbon emissions to lower the life cycle GWP of MPW-derived hydrogen. The results show the economic and environmental advantages of using MPW over single-stream plastics (i.e., PE, PET, PP, and PS) in producing hydrogen via gasification, given the high feedstock cost and environmental burdens of sorting and processing single-stream plastics in MRFs at current stage and low hydrogen yield of some plastics (e.g., PET). Given the current high portion of MPW landfilled or discarded, more efforts are needed to prioritize MPW valorization in an environmentally benign and cost-effective way. Among single-stream plastics, PET is the least favorable in terms of both environmental and economic performance. This implies the necessity of exploring other high-value and feasible recycling methods for sorted single-stream plastics (e.g., replacing virgin materials)7. Increasing the plant capacity can reduce the MHSP across all feedstock cases. From an operational aspect, the steam/feed ratio directly affects MHSP, and the optimal steam/feedstock ratio varies by feedstock (e.g., 2.0 for MPW and 3.5 for PS) and generally increases as the feedstock cost grows. The improvement analysis exhibits possible pathways to decrease the MHSP of MWP-derived hydrogen with CCS from US$2.60 to US$1.46 kg−1 H2. If carbon credits are close to the CCS costs and MPW feedstock cost is low, the MHSP of utilizing MPW can reach US$1.06 kg−1 H2. To achieve the ambitious goal of US$1.0 per kg clean hydrogen in one decade, the roadmap highlights the need for simultaneous improvement of process economics and policy supports.

## Methods

### Feedstock compositions

Common plastic wastes include PET, high-density polyethylene (HDPE), PVC, low-density polyethylene (LDPE), PP, PS, and other plastic waste39,74. Supplementary Table 2 summarizes the composition data from the proximate and ultimate analysis of the plastics used in this study. Five feedstock cases were developed to investigate the impacts of different plastic waste feeds, particularly to compare the economic and environmental performance of single-stream plastic feed and MPW. Four cases use single-stream plastic waste, including PE (assuming 50% LDPE and 50% HDPE), PET, PP, and PS provided by sorting or recycling facilities. Pure PVC feed was not selected due to extremely high chlorine content causing safety and corrosion concerns75. One case was designed for MPW that was typically rejected from the mechanical recycling at MRF7. These MPW are commonly landfilled that need “tipping fee” or incinerated to generate power7,76. In this study, MPW contains 19.5% HDPE, 27.9% LDPE, 27.5% PP, 7.6% PS, 14.6% PET, and 2.9% PVC based on the data of landfilled plastic waste that is neither recycled nor combusted in the U.S. in the year 2018 by U.S. EPA76.

### Process simulation model of the hydrogen plant

A process simulation model was established in Aspen Plus to provide mass and energy data for TEA and LCA32. As shown in Fig. 1, the hydrogen plant comprises five main areas: feedstock handling and pretreatment, gasification, hydrogen purification, CHP plant, and utilities. The detailed process diagrams of Aspen Plus in each area are shown in Supplementary Figs. 1317. An example of summarized flow information is available in Supplementary Fig. 18 and Supplementary Table 10 in Supplementary Data 1 (ref. 67).

In this study, five different feedstocks (i.e., PE, PET, PP, PS, and MPW as shown in Supplementary Table 1) were fed into the simulation model to study the impacts of varied feedstock compositions. The plastic waste is assumed to arrive at the hydrogen plant in the form of bales7,77. The bales are then unloaded and transferred to the warehouse for storage. The first unit operation is the size reduction of the plastic waste in the shredder to around 152 mm (6 inches)78. After the initial grinding, the feedstocks are washed in the rotary drum washer to remove the entrained ash and other contaminates7,79,80. Different from the pure feedstocks (i.e., PE, PET, PP, PS) that have been sorted and processed, MPW will need another two washing steps in friction washers as a common practice7. Then the feedstocks are dried in the rotary drum dryer at 105 °C to reach a moisture content lower than 10% (dry basis)81,82. Followed by drying, feedstocks are further grounded in the secondary grinding to around 1–2 mm7,81, and are ready for gasification.

Before gasification, dechlorination is essential for removing toxic chlorine from PVC for safety and corrosion concerns. Based on the study by López et al., treating the plastic mixtures containing PVC at 300 °C in a nitrogen atmosphere for 30 min can efficiently remove 99.2% of chlorine in PVC75. In this study, the dechlorination process is conducted at the same condition before gasification75,83. The weight loss of PE, PP, PS, and PET in the dechlorination process is only 0.7%, 0.3%, 3.3%, and 0.8%, respectively75. Two-stage gasification was modeled in this study, including gasification followed by tar cracking which is essential for large-scale hydrogen plant operation84. This study uses a bubbling fluidized bed reactor for gasification, and a fixed bed reactor for tar cracking based on the literature84,85. For gasification, the operating condition was selected to be 850 °C and 3.5 MPa with steam as the gasifying agent for H2-rich production13,86,87. In Aspen Plus, the gasification was modeled with two reactors in sequence using RStoic and RGibbs, which is consistent with previous process simulations for gasification40,88,89,90,91,92,93. The RStoic reactor decomposes the inlet stream based on the feedstock compositions. Then the decomposed stream along with steam is sent to the RGibbs reactor that calculates the syngas composition using Gibbs free energy minimization method88,89. In RGibbs reactor, 12 reactions are considered based on the literature (see Supplementary Table 11 for detailed reactions)94,95,96.

This study uses steam gasification which is commonly used for generating H2-rich syngas as the presence of steam can increase hydrogen yield, reduce the tar concentration, and promote water gas shift reactions85,97,98,99,100. Previous studies show the importance of S/F ratio in gasification design and optimization40. The S/F ratio commonly varies from 1.0 to 4.040,85. The higher S/F ratio may lead to higher hydrogen yield, but at the same time can cause higher energy costs. To choose a suitable S/F ratio, this study investigated the S/F ratios from 1.0 to 4.0 in each feedstock case, and selected the S/F ratio with the lowest MHSP. The bed material is natural olivine with a diameter of 100–300 μm37,84. Natural olivine is a highly attrition-resistant catalyst to reduce tar formation37,84. For tar cracking, the fixed bed reactor operates at 800 °C and 3.5 MPa with additives that are 1:1.5 mixtures of calcined dolomite and activated carbon84. These additives can efficiently decompose the NH3 formed in gasification and reduce the concentration of HCl and H2S in syngas84. After the tar cracking, a cyclone is deployed to separate the solid phase (e.g., fly ash)101.

After the hot syngas is generated, the first step is to remove the impurities. The moving-bed granular filter with CaO is deployed to desulfurize and dechlorinate the hot syngas102,103. Then the remaining tar is removed by a Venturi scrubber at about 35 °C and a wet-packed column for fine tar removal101,104. To integrate the tar removing with other impurity removing, the Venturi scrubber washes with 10% NaOH solution to remove the remaining HCN, HCl, and H2S105,106. To further eliminate NH3, an acid wash column with H2SO4 solution at pH 5 is adopted107. The purified gas primarily contains H2, H2O, CO, CO2, and CH4. To separate hydrogen, the syngas is compressed to 13.7 atm and fed to a PSA which is assumed 84% hydrogen recovery with 99% purity106. All the off-gases are sent to the CHP plant for energy recovery106. To store the hydrogen, the purified hydrogen is assumed to be compressed to 700 bars through two-stage compressing108. 700 bar is a common pressure level for storage or for hydrogen stations to refuel the fuel cell108,109.

As this study uses steam gasification, the steam load in the gasification area is high. At the same time, the hydrogen plant consumes electricity in each area. Given the demand of electricity and heat, this study includes a CHP plant that recovers the energy in PSA off-gas and char to produce electricity and heat needed by the whole plant. If the heat supply is not sufficient by combusting the intermediate flows, natural gas will be combusted as a supplementary fuel. The boiler generates superheated steam at 62 atm and 454 °C with 80% boiler energy efficiency110. The superheated steam then goes through multi-stage turbines for power generation. In this study, the low-pressure steam at 13 atm and 268 °C from the first stage turbine is extracted for feeding the gasifier and providing heat to the dechlorination reactor and tar cracking reactor.

Plant utilities include electricity, cooling water, process water, chilled water, plant air system, and the storage of materials and products110,111. All of these utilities are included in the process simulation, TEA, and LCA.

This study includes scenarios with and without CCS. CCS captures and stores the CO2 from the CHP plant flue gas. The CO2 concentration in the cooled flue gas is around 23 vol.%. Post-combustion CCS was chosen because of its suitableness for capturing carbon from air-combusted flue gas with much lower CO2 concentration (commonly lower than 25 vol. % of flue gas112,113) than oxyfuel combustion CCS that uses pure O2 (ref. 114). The capture efficiency is assumed to be 90% for post-combustion CCS115. See Supplementary Note 5 for detailed technical information and Supplementary Note 4 for cost data.

### Techno-economic analysis model

This study focuses on the hydrogen plant with a capacity of 100–2000 ODMT plastic waste per day. The mass and energy balance data by Aspen Plus simulation were input to determine variable operating costs and capital costs. In TEA, the original purchased costs, installing factors, equipment scaling factors, and material and energy prices, and feedstock costs were collected from the literature and discussed in Supplementary Note 6 for capital expenditures and Supplementary Note 4 for operating expenditures. The MHSP, a widely adopted indicator describing the production cost under preset IRR, was selected to assess the economic feasibility of the hydrogen plant33. The MHSP was derived through the discounted cash flow rate of return (DCFROR) analysis as a widely used economic analysis method in TEA23. In the DCFROR analysis established in EXCEL, the MHSP was derived by setting the IRR to be 10% and the Net Present Value (NPV) to be zero23. The year of analysis is 2019 based on the latest data availability. Supplementary Tables 12 and 13 list the key assumptions and parameters of TEA based on literature data. The plant is assumed to have 40% equity-financed and take the remaining 60% on loan. The capital cost was assumed to be depreciated over 7 years by following the Modified Accelerated Cost Recovery System by the U.S. IRS116.

The total capital investment includes total installed equipment cost, other direct costs, indirect cost, and land and working capital. Total installed equipment cost is the sum of the installed equipment costs that were estimated by multiplying purchased costs with installation factors (see Supplementary Tables 1418). The purchased costs and installation factors used in this study were collected from the literature as shown in Supplementary Note 6. The economy of scale was considered using the scaling factors (see Supplementary Note 6) to scale the purchased costs found in the literature to the capacities explored in this study. Plant cost indices by Chemical Engineering Magazine were used117 to adjust equipment purchased costs collected from the literature to the year of analysis 2019 in this study. The detailed method of determining equipment cost is documented in Supplementary Note 6.

The operating expenditures include the variable costs of feedstocks, raw materials, waste stream charges, byproduct credits, and fixed operating costs (including labor cost), and other operating costs. The prices of feedstocks, raw materials, waste stream charges, and energy were collected from the literature and documented in Supplementary Table 3. If the price is not in the year of analysis (2019), the Producer Price Index for chemical manufacturing was used to adjust the original prices to 2019 (ref. 118). The details are available in Supplementary Note 4 and Supplementary Table 19.

### Life cycle assessment model

In this study, a cradle-to-gate LCA was conducted to display the environmental impacts of hydrogen converted from MPW. The life cycle inventory (LCI) data for the hydrogen plant were derived from the Aspen Plus simulation for different scenarios, including energy and material consumption (e.g., fuels, chemicals, water) and CHP plant emissions. AP-42 emission factors by U.S. EPA were used to estimate emissions from natural gas combustion (see Supplementary Table 20 for emission factors)119. The LCI data of upstream production of electricity and materials and treatment of wastewater and solid waste (e.g., ash) were collected from the ecoinvent database (see Supplementary Table 21 for the unit processes used in this study)120. The functional unit is 1 kg H2 produced in consistency with the TEA. LCIA uses the TRACI 2.1 method by U.S. EPA and 100-year GWP characterization factors by IPCC AR6 202134,35.