Policy levers and technological opportunities to achieve decarbonisation in Europe at minimal cost

Abstract

Energy systems must shift their supply from fossil fuels to carbon-neutral technologies. At the same time, they are subject to substantial economic pressure, as only minimal to moderate consumer cost increases can erode most of the energy transition’s political support. Rapid cost reductions in solar photovoltaics and wind generators have rendered them the transition’s backbone, pulling electrification into transport and heating, but also increasing system complexity through inter-sectoral energy flows and non-dispatchable generators. Therefore, whole-system energy research, focused on minimising transition costs, also faces new complexities through the interplay of heterogeneous technologies, intermittency, and spatial constraints, all in the context of evolving policy frameworks. Models that can address these problems are essential for high-level decisions because technology competitiveness, infrastructure roll-out, and market designs may need, or are already undergoing, a fundamental overhaul as a result of the changing nature of energy supply.

Such energy-system research is sometimes conducted with proprietary, GUI-oriented software and associated workflows, which can be less well suited to large-scale, reproducible analysis. Two aspects are particularly relevant. First, these workflows do not always take full advantage of the highly optimised, programmatic tooling that has matured since the data-driven methods of the early 2010s. Second, they may be slower to benefit from recent advances in AI-assisted modelling, since large language models and related tooling tend to integrate most naturally with open, scriptable software ecosystems on which they can be trained and evaluated. Furthermore, where important parts of the modelling chain are closed, key assumptions and preprocessing steps can be harder to audit externally, which may shift influence towards a small set of actors and make independent validation more difficult. Recent open-source advances, led by the Python for Power System Analysis (PyPSA) ecosystem, offer a more transparent and resource-efficient alternative. PyPSA’s flagship sector-coupled model, PyPSA-Eur, optimises Europe-wide investment and operation at high spatio-temporal resolution and is, along with other open-source variants, increasingly adopted by industry and policy stakeholders.

However, despite these recent advances, important gaps remain in how widely modern, open, and reproducible modelling is applied; as a result, decision-making can be informed by analyses that are not as robust or detailed as they could be, which risks making the energy transition more expensive than necessary.

This thesis contributes to a cost-efficient transition through three individual research contributions and the “meta” contribution of pushing the application of transparent, open-source tools to new domains.

The first contribution relates to Enhanced Geothermal Systems (EGS), which leverage heat in deep geological formations to deliver low-carbon heat and power. However, current drilling costs render them largely uncompetitive.

Future cost reductions are plausible but remain uncertain, and EGS’s market opportunity in light of that uncertainty remains poorly scoped, particularly in Europe. This thesis finds that, in a carbonneutral, multi-sector European energy system, at today’s drilling costs, heat-generating EGS could supply about 20–30 GWth of district heating and low-temperature industrial demand. If drilling costs were reduced by roughly 60%, EGS could compete directly with wind and solar generation in electricity markets, expanding its market opportunity by an order of magnitude.

The second contribution assesses how smart-charging electric vehicles (EVs) and heat-pump-coupled thermal storage could help balance Great Britain’s increasingly renewable power supply. Under an optimistic 2030–50 pathway, domestic demand flexibility could unlock 20–30 TWh of additional renewable output annually and thereby avoid about 20 GW each of dispatchable generation capacity and distribution-grid expansion, reducing annual costs by around £5 billion. However, it is also found that half of those savings come from electrifying and smart-charging just 25% of cars; beyond that, additional EV or heating flexibility yields substantially smaller returns. For heating, most of its system-level benefit relies on expanded thermal storage lifting its temporal flexibility to around 12 hours, which is an optional and expensive add-on to a heat pump. Finding that even large-storage-coupled flexible heating offers limited cost benefits raises questions about the cost threshold at which thermal storage becomes economically viable.

Finally, Great Britain was considering moving from its current national electricity market to a zonal design. Stakeholders disagree about whether the efficiency gains of such a reform offset the downstream costs of an investment environment perceived to be more uncertain. The opposing findings are hard to reconcile, since they are based on different, proprietary modelling software.

This work, based on a novel, back-casting open-source electricity-market model, finds that a six-zone wholesale market for Great Britain would have lowered 2022–24 consumer costs by £9.4/MWh (≈ £2.3 billion per year) through better dispatch and less wind curtailment, but would have reduced revenues for northern renewables and nuclear plants by 30–40%. Targeted compensation could restore up to 97% of those revenues while still preserving £3.1/MWh (≈ £0.75 billion per year) in savings. As the net balance between those two effects, zonal pricing delivers net welfare gains of around £0.5 billion annually today, which are likely to rise to £1–2 billion per year after 2029.

Therefore, we conclude that the benefits of the reform outweigh the long-term investment risks.

In summary, this thesis makes three novel contributions regarding a cost-optimal energy transition. It first evaluates a firm complement for intermittent generators in the form of EGS on the generator side, and then assesses to what extent that complement could, alternatively, be provided on the demand side.

Finally, it assesses the suitability of different market designs to enable the economical deployment of these technologies.

Across this thesis, a unifying theme is the minimisation of consumer energy cost in the context of the ongoing energy transition. As decarbonisation progresses, its impact on energy prices will increase, and therefore ensuring the transition proceeds as efficiently as possible is crucial. The analyses are run on the back of modern or novel, opensource energy system models highlighting the potential of transparent, powerful and reproducible energy science.