Climate change threatens human lives and ecosystems worldwide. Limiting the escalating adverse impacts requires transforming global energy systems, responsible for 78% of current greenhouse gas emissions, away from fossil fuels. While the rapid growth of low-cost renewable electricity technologies gives reason for hope, realising their full potential across the entire energy system requires integrating variable electricity supply with demand in all end-use sectors. This sector coupling consists of two main routes: direct electrification via technologies like electric vehicles and heat pumps, and indirect electrification through electricity-based fuels such as green hydrogen and synthetic hydrocarbon fuels. However, fundamental uncertainties remain about the realistic availability and competitiveness of green hydrogen, as well as about the role of power system flexibility in long-term climate mitigation scenarios. This thesis addresses these gaps in three complementary contributions, using a variety of methods that contribute to an improved understanding of future energy systems from different angles. Chapter 2 analyses the feasibility limits of scaling-up green hydrogen supply, using a probabilistic technology diffusion model parametrised with a global database of electrolysis project announcements. This reveals short-term scarcity and long-term uncertainty. Even if electrolysis capacity grows as fast as wind and solar PV, green hydrogen will likely supply less than 1% of final energy until 2030 in the EU and until 2035 globally. At such growth rates, both the ambitious EU 2030 policy targets and the required deployment in global climate mitigation scenarios remain out of reach. While a breakthrough to higher production volumes is possible, the timing and magnitude are highly uncertain. Only under unconventional growth rates, comparable to emergency deployment, could supply scarcity be overcome. These findings reveal the substantial risks of a fossil lock-in if policymakers neglect other transformation options and rely on uncertain green hydrogen availability. Chapter 3 structures past and future challenges of green hydrogen by defining and quantifying the green hydrogen ambition and implementation gap. This reveals a wide past implementation gap in 2023, as only 7% of global announced capacity was realised on time. At the same time, the 2030 ambition gap between project announcements and the requirements in most 1. 5 °C scenarios has been gradually closing in recent years. However, this has been accompanied by a widening future implementation gap between project announcements and realistic deployment given current policy support. A bottom-up technoeconomic cost model reveals a substantial cost gap between green hydrogen and fossil fuels across all end-use sectors. Without carbon pricing, realising all project announcements by 2030 would therefore require subsidies of US 1. 3 trillion, far exceeding real-world policy support. The implementation gaps indicate persistent challenges that limit the role of green hydrogen in the energy transition. The corresponding Policy Brief in Section 3. 1 provides recommendations for policymakers, arguing that green hydrogen expectations need to be grounded in reality. Chapter 4 analyses the role of electrification within flexible power systems as a counterpart and competitor to green hydrogen. The chapter introduces a novel modelling framework that soft-links the IAM REMIND with the ESM PyPSA-Eur. This addresses a critical modelling trade-off between the wide scope required for long-term transition pathways and the high spatio-temporal detail required to capture the costs and benefits of power system flexibility. Through a bi-directional, iterative and price-based coupling, the approach enables the joint optimisation of long-term investment decisions in REMIND and short-term power system operation in PyPSA-Eur. Results for Germany confirm the technical feasibility and cost effectiveness of a sector-coupled energy system with nearly 100% renewable electricity. While demand-side flexibility reduces average electricity prices, substantial differences persist across sectors, with flexible electrolysis benefiting from below-average prices, whereas heat pumps pay almost twice the average price due to winter peak loads. This demonstrates how power system effects unfolding on hourly time scales are critical for investment decisions on decadal time scales. Chapter 5 synthesises all findings, provides methodological reflections, and discusses future research and policy implications. Revisiting the analyses from Chapter 2 and Chapter 3 with recent data reveals that green hydrogen projects have continued to fall short of expectations in 2024. Methodologically, I argue that the fast-growing field of feasibility studies needs to advance by incorporating policies and competitiveness to deliver actionable insights. Future research directions range from continuous monitoring of the green hydrogen scale-up to an improved understanding of the impact of sectoral electricity prices on end-use transformation pathways. The policy implications of this thesis underline that betting on the large-scale availability of cheap hydrogen risks a fossil lock-in. To safeguard climate targets, policymakers should prioritise cheap, scalable, and readily available direct electrification wherever possible, while focusing hydrogen support on sectors that cannot be electrified, rather than relying on the distant vision of hydrogen as a silver bullet.
Adrian Odenweller (Thu,) studied this question.