Integrated Systems-to-Atoms (S2A) Framework for Designing Resilient and Efficient Hydrogen Infrastructure Solutions
Mengyao Yuan, Giovanna Bucci, Tanusree Chatterjee, Shyam Deo, John R. Kitchin, Carl D. Laird, Wenqin Li, Thomas Moore, Corey Myers, Wenyu Sun, Ethan M. Sunshine, Bo-Xun Wang, Matthew J. McNenly, Sneha A. Akhade, “Integrated Systems-to-Atoms (S2A) Framework for Designing Resilient and Efficient Hydrogen Infrastructure Solutions”, Energy & Fuels 39(14), 7119-7128 (2025). https://doi.org/10.1021/acs.energyfuels.4c05903
This summary was written by a language model from the paper and has not been checked line by line. Rely on the published paper for details.
Note on the source text: the extracted text is the published article and is readable; figures are not available in the text, so results reported only graphically are described from the prose.
Question
Whether a hydrogen storage material is worth using depends on how it performs inside a whole delivery chain, yet catalyst models rarely see system-level costs or demand, and technoeconomic analyses usually treat material performance as a fixed input. The paper asks how material, device and system design for hydrogen storage and transport can be co-simulated, and what such a cross-scale view reveals for a liquid organic hydrogen carrier (LOHC), using formic acid as the demonstration case (Yuan et al. 2025).
Approach
The Systems-to-Atoms (S2A) framework, developed at Lawrence Livermore National Laboratory with collaborators at Carnegie Mellon, links three models with data passed between scales. At the material scale, microkinetic modeling with CatMAP, using first-principles energetics and a mechanism from earlier work, gives formic acid decomposition rates on Pd(111) and Pd(211). Arrhenius fits give an apparent barrier of about 89.5 kJ/mol for dehydrogenation and 156.6 kJ/mol for the competing dehydration path. Facet rates are combined into a nanoparticle turnover frequency using Wulff-construction facet fractions, with an effectiveness factor for mass transport (details in a companion paper by Deo et al.). At the component scale, a liquid-phase reactor mass balance with 5 wt% metal on carbon pellets converts the TOF into reactor volume and catalyst mass for 99.99% conversion. At the system scale, a bottom-up technoeconomic model, built on Argonne’s HDSAM v3.1 and other sources, computes the levelized cost of hydrogen. The scenario is California-specific: 10 refueling stations of 1000 kg H2/day each, 100-mile truck delivery, and dispensing at 700 bar. Formic acid delivery is compared with compressed and liquid hydrogen. Formic acid is made by CO2 electrolysis and purified, and the CO2 released at the station is separated by pressure swing adsorption and trucked back. Assumptions were deliberately chosen to favor the LOHC. Pd, Pt and Cu catalysts were then compared across 30 to 90 °C and 1 to 100 bar. The code and data are in LLNL’s systems2atoms repository.
Key findings
- The formic acid pathway costs about $32/kg H2, against about $6 to 7/kg for compressed and liquid hydrogen. Formic acid synthesis, purification and CO2 recycling dominate, and all are amplified by its low hydrogen content (4.4 wt%). Each $1/kg of formic acid corresponds to about $23/kg of H2, and the estimated production cost of $0.64/kg formic acid alone is about $15/kg H2.
- Pd activity rises by about three orders of magnitude from 30 to 90 °C, but this barely changes station cost (savings under $0.15/kg H2, partly offset by higher precooling demand), because the reactor and catalyst are less than 3% of station cost.
- Pressure up to 100 bar has almost no effect on Pd activity, yet it lowers station cost by about $2/kg H2 (up to about 5% of station cost). A higher reactor outlet pressure lowers the energy and equipment size of downstream PSA and compression, outweighing sturdier reactors and feed pumping. Most of the benefit appears up to about 20 bar, the PSA operating pressure.
- Pd and Pt respond little to pressure, but pressure increases the Cu rate by one to two orders of magnitude, so about 15 to 45 times less Cu is needed at 100 bar than at 1 bar. Cu is the cheapest catalyst to buy at 100 bar below about 45 °C, and at 100 bar Cu and Pd give nearly the same station cost. Because Pd and Pt are listed as critical minerals, Cu could be preferred for supply-chain resilience. System cost was also insensitive to catalyst lifetime.
- The authors conclude that short-distance delivery to high-pressure refueling stations is likely not an economic use of formic acid. They suggest screening carriers by hydrogen content, loading and release conditions, and byproducts.
Significance
The study shows that the best catalyst and operating point can change once system costs are included. In this case, a cheaper, less active, pressure-responsive metal can match a noble metal, and the cost lever is the downstream compression rather than catalyst activity. The framework is meant as a screening tool for other carriers, end uses and transport modes, and its cross-scale models can be optimized through surrogate forms, as in earlier decision-tree surrogate work by Sunshine et al. (Yuan et al. 2025). The Cu result rests on model predictions with assumed selectivity, so the authors call for experimental validation.