Technological evolution of large-scale blue hydrogen production toward the U.S. Hydrogen Energy Earthshot

The study addresses how technological learning and U.S. federal tax incentives can reduce the cost of blue hydrogen (fossil-based hydrogen with carbon capture and storage) to achieve the Department of Energy’s Hydrogen Energy Earthshot goal of $1/kg H2. In the context of the U.S. push for a clean hydrogen economy—targeting 10 MMT/year by 2030 and supported by the Bipartisan Infrastructure Law and DOE’s Regional Clean Hydrogen Hubs—the paper examines blue hydrogen as a near-term bridge option alongside green hydrogen. Given current debates around methane leakage and the relatively high cost of green hydrogen, the research evaluates whether learning-by-doing and Inflation Reduction Act (IRA) incentives (Sections 45Q and 45V) can drive blue hydrogen costs toward the $1/kg target, and identifies the breakeven cumulative capacity needed under varying market and policy conditions. The purpose is to provide quantitative guidance for technology development, investment, and policy design to accelerate progress toward the Earthshot goal.

Prior assessments indicate most hydrogen is currently produced by fossil pathways without CCS, with SMR dominating U.S. production. NETL benchmarked costs show CCS increases costs for reforming by >50% and for gasification by ~20%, with fuel cost dominating SMR-based LCOH and capital cost dominating gasification. Historical learning studies report: SMR capital learning rate ~11% with negligible O&M learning (Schoots et al.); gasification capital and O&M learning rates ~14% and ~12% (Rubin et al., IEAGHG); post-combustion CO2 capture learning rates ~11% (capital) and ~22% (O&M) (Rubin et al.). Despite many learning-curve applications, few studies holistically assess learning for blue hydrogen systems or quantify the effect of IRA tax credits (45Q, 45V) on technological evolution. This work fills that gap by constructing component-based learning curves for blue hydrogen subsystems and integrating policy incentives into cost trajectories.

The study uses empirical learning curves and a diffusion-of-innovation framework to project future costs of blue hydrogen (SMR+CCS for natural gas and oxygen-blown entrained-flow coal gasification+CCS) with and without IRA tax incentives. Key elements include: 1) Component-based learning curve model: The plant is decomposed into subsystems (e.g., SMR, PSA, CO2 capture, compression, transport and storage for gas-based; and gasification block, ASU, syngas cleanup, sulfur recovery, power block, CCS, etc. for coal-based). For each subsystem, initial total as-spent capital (TASC), total O&M, initial installed capacity, and learning rates (capital and O&M) are specified from NETL and literature. Costs evolve as C_i = a_i * x^{b_i}, with learning rates LR_i mapped to b_i via b_i = log(1−LR_i)/log(2). Plant-level cost is the sum across subsystems. 2) Cost metric: Levelized cost of hydrogen (LCOH) in real dollars is computed as the sum of levelized capital cost, non-fuel O&M, and fuel cost components, using fixed charge rate methods aligned with NETL guidelines. 3) Tax credits: The model incorporates levelized impacts of Section 45Q (assumed $85/tCO2 stored in saline for 12 years) and Section 45V (assumed $0.6/kg H2 for 10 years if life-cycle GHG < 4.0 kg CO2-eq/kg H2) to estimate LCOH with tax credits over the project book life. 4) Diffusion-of-innovation: An S-curve estimates the time-based diffusion of global low-carbon hydrogen capacities, enabling translation from cumulative capacity to calendar-year costs. 5) Data inputs: Initial costs and performance from NETL; initial installed capacities from IEA databases and sector sources; learning rates from Schoots et al., Rubin et al., IEAGHG; financial parameters per NETL guidance. 6) Sensitivity analyses: Parametric studies examine effects of natural gas price, CO2 removal system contingencies, learning rates (with particular focus on O&M learning for SMR, PSA, CO2 compression), and inflation (nominal-dollar analysis) on future LCOH trajectories and breakeven capacity to reach $1/kg.

- Baseline current costs and emissions: For plants producing 99.9% purity H2 at 6.48 MPa and storing CO2 at 15.3 MPa, LCOH (2018$) is $1.64/kg for SMR+CCS and $3.09/kg for coal gasification+CCS. SMR+CCS stack emissions: ~0.4 kg CO2/kg H2 vs ~1.4 kg CO2/kg H2 for gasification+CCS. Fuel cost share: ~50% of LCOH for gas-based vs ~14% for coal-based, underscoring natural gas price sensitivity. - Compared to gray SMR (no CCS), SMR+CCS reduces stack CO2 intensity by ~96% but raises LCOH by ~55%; CO2 avoidance cost ~ $65/tCO2. Green H2 by PEM electrolysis has LCOH ~$3.0–7.5/kg, with CO2 avoidance cost ~$212–689/tCO2 relative to gray, highlighting cost–emissions trade-offs. - Learning without tax incentives: At 10 MMTA cumulative capacity, capital and O&M cost reductions are ~20.0% and ~8.3% for gas-based; plant LCOH declines to ~$1.46/kg (−10.7%) for SMR+CCS and ~$2.75/kg (−10.9%) for coal+CCS—still above the $1/kg target. - With tax incentives: Assuming 45Q ($85/tCO2, 12 years) or 45V ($0.6/kg, 10 years), at 10 MMTA cumulative capacity SMR+CCS LCOH falls to ~$1.14/kg (45Q) and ~$1.26/kg (45V), representing −22.2% and −13.7% vs no-credit, respectively. 45Q delivers greater value to blue hydrogen than 45V. Coal gasification+CCS remains unlikely to achieve $1/kg even with credits. - Breakeven capacity sensitivity to gas price under credits: With 45Q, breakeven cumulative capacity to reach $1/kg is ~4.9 MMTA at $3.3/GJ gas and ~0.6 MMTA at $2.8/GJ; at $2.4/GJ, initial LCOH can be < $1/kg. With 45V, breakeven is ~9.8 MMTA at $2.8/GJ and ~1.2 MMTA at $2.4/GJ. Without credits, achieving $1/kg is difficult even with cheap gas absent substantially higher learning rates. - CO2 removal system cost uncertainty: Varying process and project contingencies for capture systems notably affects TASC and LCOH; at 10 MMTA, LCOH spans ~$1.45–$1.48/kg. To hit $1.46/kg, required cumulative capacity ranges from ~7–16 MMTA, highlighting the leverage of capture cost uncertainty on capacity requirements. - Learning-rate sensitivity: Increasing capital and O&M learning rates (especially assigning positive O&M learning for SMR, PSA, and CO2 compression at 5–10%) accelerates cost declines and reduces required cumulative capacity. Even so, without credits and with higher gas prices, reaching $1/kg by 2030 remains challenging. With 45Q and O&M learning ≥5% for key subsystems, $1/kg becomes feasible at large cumulative capacities (≥~20 MMTA depending on other subsystem learning). - Inflation effects: Estimating in nominal dollars, inflation (1–3%) raises LCOH and can offset learning-induced reductions; at 1% inflation, the cost reduction from 10 MMTA deployment (no 45Q) can be nullified. With 3% inflation, even with 45Q and up to 30 MMTA capacity, reaching $1/kg is difficult unless gas prices are low. - Scale and resource implications: Producing 10 MMTA H2 via SMR+CCS would consume ~1.9 billion GJ/year of natural gas (~17% of 2022 U.S. industrial use), withdraw ~0.30 km³/year water, and occupy ~25.1 km² land, underscoring regional resource planning needs. - Policy/design insights: Blue H2 projects derive more economic benefit from 45Q than 45V; extending the 45Q period (e.g., to 18 years) would materially reduce required cumulative capacity to reach $1/kg. Deep CCS (e.g., 99% capture) can approach net-zero site emissions with modest CO2 avoidance cost increases (3–13%) and enhance 45Q value. Methane leakage control is critical; leakage ≥3.5% can undermine blue hydrogen’s climate competitiveness. Overall, the breakeven cumulative capacity for SMR+CCS to achieve $1/kg is jointly determined by tax credits, natural gas price, learning rates, and inflation.

The findings show that learning-by-doing lowers blue hydrogen costs, but tax incentives are pivotal to approach the $1/kg Earthshot target, especially for gas-based SMR+CCS. The analysis quantifies how 45Q outperforms 45V for blue hydrogen economics and how natural gas prices and inflation can accelerate or impede progress. Coal gasification+CCS remains structurally less competitive due to higher capital intensity and higher residual emissions. The results inform policy by highlighting that extending 45Q duration, supporting CCS R&D (notably capture cost and energy penalties), enabling low-cost gas supply, and curbing methane leakage can materially reduce the capacity scale and time needed to hit cost targets. For market formation, hydrogen hub strategies with co-location of feedstock, capture, transport, and storage infrastructure and large-scale plants can enhance economies of scale and reliability. Retrofitting CCS onto existing SMR can lower upfront capital and near-term LCOH, bridging to larger greenfield projects. However, inflationary environments and financing uncertainties can erode gains from learning, suggesting the importance of stable, long-dated policy signals and offtake contracts to reach FID and enable diffusion consistent with the Earthshot timeline.

This study contributes a component-based learning and diffusion framework quantifying how technological learning and IRA tax credits shape the cost trajectory of blue hydrogen. It shows that SMR+CCS can approach or reach $1/kg H2 under favorable combinations of 45Q credits, low natural gas prices, and improved learning—whereas coal gasification+CCS is unlikely to meet the target. The analysis identifies tax credits, fuel price, learning rates (especially O&M for SMR/PSA/compression), and inflation as the dominant levers determining the breakeven cumulative capacity. Policy recommendations include prioritizing 45Q enhancements (e.g., longer credit period), strengthening methane abatement standards, accelerating CCS RD&D (including deep capture and solvent regeneration), supporting hub-based infrastructure, and facilitating retrofits of existing SMR. Future research should develop probabilistic learning models with richer empirical data, assess regional resource constraints and siting optimization, evaluate integration with by-product markets and energy systems, and explore next-generation blue hydrogen pathways (e.g., methane pyrolysis, dry reforming, chemical looping) for accelerated learning and cost reductions.

The analysis uses a single-factor, top-down learning-curve approach with constant learning rates, which simplifies complex innovation dynamics and may not capture step-change technological advances or deployment bottlenecks. Sensitivity analyses explore parameter ranges, but do not assign probabilistic likelihoods due to limited empirical data for several subsystems. Results rely on assumptions for initial installed capacities, subsystem cost allocation, and learning rates derived from analogous technologies. Inflation, market volatility, and policy implementation uncertainties (e.g., eligibility for 45V based on life-cycle emissions) can cause deviations from projected cost paths.