Grey hydrogen produced via steam methane reforming without carbon capture generates between 9 and 11 kilograms of CO2 per kilogram of hydrogen produced. Combusting that hydrogen in a gas turbine to generate 1 megawatt of net electrical output produces approximately 857 kilograms of CO2 per hour, compared to 577 kilograms per hour for natural gas combustion at the same output level.
A policy or infrastructure decision that switches from natural gas to grey hydrogen combustion does not reduce emissions. It increases them by roughly 50% while adding the cost and complexity of hydrogen production, storage, and delivery. This is the sharpest result from a thermodynamic scenario analysis published in Clean Technologies in August 2026 by researchers from Istanbul Atlas University and Istanbul Gelisim University, and it has direct implications for how hydrogen investment is being directed in industrial and power generation contexts.
The study compares five scenarios at a consistent 1 MW net electrical output boundary: natural gas combustion as the reference, grey hydrogen combustion, grey hydrogen in a proton exchange membrane fuel cell, green hydrogen combustion, and green hydrogen in a PEM fuel cell. The methodological choice to hold output constant rather than fuel input constant is consequential, because it forces a fair comparison of how much resource each pathway requires to deliver the same useful work. The results show that the pathway combining green hydrogen with PEM fuel cell conversion achieves both the lowest carbon footprint and the highest thermodynamic efficiency among all scenarios examined. But the path to that outcome runs through an infrastructure and technology deployment sequence that the hydrogen industry’s current investment priorities are not consistently aligned with.
What Exergy Analysis Adds to the Carbon Accounting
Energy efficiency, as measured by the first law of thermodynamics, tells you what fraction of input fuel becomes useful electrical output. The natural gas combustion system achieves approximately 35% energy efficiency; the PEM fuel cell system approximately 55%. Those figures already favour the fuel cell by a significant margin, but they do not capture the full picture of resource utilisation. Exergy analysis, grounded in the second law, quantifies how much of the maximum theoretical work potential in a fuel is destroyed by irreversibilities during conversion, principally through entropy generation in high-temperature combustion processes.
For the natural gas reference system at 1 MW output, the exergy input is 2.97 MW and the exergy efficiency is 33.7%, meaning 66.3% of the fuel’s theoretical work potential is destroyed rather than converted to useful electricity. For the green hydrogen PEM fuel cell system, the exergy input is 2.15 MW and the exergy efficiency is 46.5%, corresponding to 53.5% exergy destruction. The 42% reduction in exergy destruction between the two systems reflects the electrochemical pathway’s fundamental thermodynamic advantage: PEM fuel cells avoid the large irreversibilities associated with high-temperature combustion and the subsequent Carnot-limited heat-to-work conversion. They are not subject to the temperature differential constraint that sets a ceiling on the efficiency of any heat engine.
The practical consequence is that the green hydrogen fuel cell system requires approximately 36% less fuel energy input than natural gas combustion to deliver the same electrical output. At current green hydrogen production costs of approximately €5 to €8 per kilogram in Europe, that efficiency advantage partially offsets the cost premium relative to natural gas, though it does not yet eliminate it at prevailing carbon prices. The gap closes faster as carbon pricing rises and green hydrogen production costs fall, which is the trajectory that both the IEA’s Hydrogen Review and the European Commission’s updated hydrogen targets assume.
The Grey Hydrogen Trap
The most policy-relevant finding in the scenario comparison is not the superiority of green hydrogen combined with fuel cells. That result is expected and well-documented. The more troubling finding is how poorly grey hydrogen combustion performs relative to both the natural gas baseline and the grey hydrogen fuel cell alternative.
Grey hydrogen combustion, at 857 kg CO2 per hour for 1 MW of output, exceeds the natural gas baseline by 280 kg CO2 per hour. The mechanism is straightforward: producing hydrogen from methane via steam methane reforming without carbon capture is itself an energy-intensive, carbon-emitting process, and combusting the resulting hydrogen at 30 to 40% energy efficiency then destroys a large fraction of the calorific value that was extracted from natural gas in the first place. The only reason to pursue this pathway is if the combustion infrastructure investment is believed to be necessary before green hydrogen supply is available, and the existing infrastructure can later be converted to green hydrogen supply. That argument depends on the cost and timeline of green hydrogen availability and the credibility of the conversion commitment.
The grey hydrogen fuel cell pathway, at approximately 545 kg CO2 per hour, is marginally better than the natural gas baseline in total emissions but not because the hydrogen is cleaner at the point of use. It is better because the fuel cell’s higher conversion efficiency requires less hydrogen to be produced and burned for the same electrical output, so the upstream emissions from SMR are correspondingly lower. This is a weak form of emissions reduction that depends entirely on not accounting for the carbon capture that SMR without CCS fails to provide, and it disappears entirely if a well-to-wire lifecycle boundary is used rather than a simplified carbon assessment.
The Combustion-to-Fuel-Cell Gap Across All Feedstocks
The consistent pattern across all five scenarios is that fuel cell conversion outperforms combustion regardless of hydrogen colour. Green hydrogen combustion produces near-zero direct emissions but achieves only 30 to 40% exergy efficiency, wasting more than 60% of the thermodynamic potential in the green hydrogen that required substantial renewable electricity to produce. Green hydrogen produced via electrolysis at approximately 65 to 75% efficiency and then combusted at 35% efficiency represents a round-trip from renewable electricity to useful electrical work at roughly 23 to 26% overall efficiency. The same renewable electricity delivered directly to a consumer achieves 100% conversion by definition, and routed through an electrolyser and a PEM fuel cell achieves approximately 36% round-trip efficiency.
The implication for the gas turbine hydrogen blending programmes that many European utilities and industrial operators are pursuing is that they represent an investment in infrastructure and regulatory compliance that locks in thermodynamic inefficiency. A 20% hydrogen blend in a natural gas gas turbine achieves approximately 7% CO2 reduction from the combustion stage while improving the exergy efficiency of the conversion process by a negligible amount, because the thermal irreversibilities of combustion dominate regardless of the hydrogen fraction. If the goal is to reduce emissions and improve thermodynamic sustainability simultaneously, the investment case for incremental combustion modification is weak compared to the case for fuel cell deployment at scale.
Where the Real Efficiency Loss Occurs in Green Hydrogen Systems
The Sankey diagram analysis in the study reveals that even in the best-performing scenario, green hydrogen plus PEM fuel cell, exergy destruction is concentrated at the production stage rather than the conversion stage. Electrolysis itself involves significant thermodynamic losses: the energy efficiency of PEM electrolysers currently ranges from approximately 65 to 75%, and the exergy efficiency from approximately 55 to 70% depending on operating conditions. Large fractions of the renewable electricity input are converted to waste heat rather than chemical energy stored in hydrogen.
This upstream loss is not captured in the conversion-stage efficiencies that dominate public discussion of hydrogen technology performance. A PEM fuel cell achieving 55% electrical efficiency looks highly competitive against a gas turbine at 35%, but the relevant system boundary for decarbonisation assessment includes the full chain from renewable electricity input to final electrical output. On that boundary, the green hydrogen fuel cell system has a round-trip efficiency that makes sense only in specific contexts: where the electricity cannot be transmitted or stored in batteries at lower cost, where the hydrogen serves multiple applications including industrial feedstock or heat as well as power, or where seasonal storage requirements make hydrogen the only technically feasible option for multi-week energy shifting.
The study’s authors are explicit that hydrogen utilisation alone does not guarantee effective decarbonisation, and that energy-based evaluations alone are insufficient to identify which pathways deliver thermodynamic sustainability alongside emissions reduction. The extension of this framework to include exergoeconomic analysis, which integrates cost accounting with exergy destruction quantification, would allow a more direct translation of thermodynamic performance differences into investment and policy guidance. That integration remains future work, but the thermodynamic baseline established here already provides enough specificity to distinguish between hydrogen deployment strategies that are substantively effective and those that are primarily regulatory compliance vehicles.
