Geopolitical and socio-economic uncertainties are putting the European and Global energy transitions at stake. These deep uncertainties affect the analytical assessments underpinning energy and climate policies. For example, the models used to inform energy planning rely on uncertain forecasts and assumptions for future energy demands, macroeconomic indicators, social acceptance, fuel prices, technology costs, and climate scenarios. Due to fundamental methodological, computational, and data challenges, this uncertainty is at best rarely considered in energy planning, which increases the risk of failing to meet our urgent climate targets. This makes accounting for uncertainty one of the major unsolved problems in energy planning.
UNITES addresses these limitations to enable a new paradigm for long-term energy planning. In contrast to current approaches, which try to accurately predict the future, UNITES’ ambition is a systematic integration of uncertainty in energy-climate models.
General aims
The project addresses key barriers to advancing uncertainty-aware energy modeling through three main objectives (O1–O3), supported by a fourth (O4). O1 focuses on developing a systematic method to characterize and quantify uncertainties in input data, including qualitative uncertainties, by assigning probability distributions. This will lead to a proof-of-concept tool and, ultimately, a comprehensive database to support energy modeling. O2 builds on this by integrating the quantified uncertainties into energy models using optimization methods, particularly targeting multi-stage decision-making to identify critical “here-and-now” choices for the energy transition. A secondary objective (O2/b) explores “non-intrusive” techniques to address computational challenges in large-scale models. O3 translates the complex outputs from these uncertainty studies into actionable policies using machine learning and AI, while incorporating human judgment. It will engage stakeholders through workshops and deliver tangible outcomes such as identifying “no-regret” and “must-avoid” decisions for EU climate policy. O4 underpins all objectives by providing a flexible, open-source EU energy model tailored for uncertainty and what-if analyses. This model will allow testing and validation of O1–O3 and will be enhanced to address broader EU policy concerns, including biodiversity. The project’s success hinges on delivering practical tools and models for decision-makers to navigate the energy transition effectively.
CMCC role:
Affiliated entity
Activities
The project is structured over five years (60 months) with a flexible yet well-defined framework comprising 4 work packages, 11 tasks (7 core and 4 ancillary), and 5 milestones. This organization allows for steady progress toward the project’s goals, with measurable indicators such as publications, software releases, and data outputs. The structure is designed to be adaptable, building on the PI’s successful experience in managing past fellowships. Milestones 1 to 3 (M1–M3) focus on core methodological developments: M1 covers the publication of the uncertainty characterization method and the creation of a proof-of-concept database; M2 formalizes the integration of quantitative and qualitative uncertainties into energy models; M3 delivers an AI-based framework for policymaking support. Milestone 4 (M4) applies these methods to identify “low-regret” and “must-avoid” decisions for the EU energy transition. Milestone 5 (M5) marks the release of the open-source EU energy model, incorporating biodiversity and climate dimensions. Most tasks are planned to run for about 24 months, allowing deep research without excessive overlap. Task 4.3 spans the full project to ensure continuous model development. Stakeholder workshops and the empirical study will begin in Year 3, allowing time for methodological progress and effective feedback integration. The Data Management Plan will be delivered by Month 6.
Expected results
The project will result in three fundamental contributions: (i) theoretical developments to consider different types of uncertainty in energy planning models; (ii) data to quantify and characterize input uncertainties; (iii) AI methods to streamline the complex outputs of uncertainty studies into interpretable policy decisions.
The transferable and interdisciplinary developments across the data science, energy, and social sciences domains, unique to Europe and not available in concert elsewhere in the world, will bring about a breakthrough scientific contribution by overcoming important barriers for uncertainty analyses in a wide array of disciplines. The project will also be highly impactful on society, bringing together for the first time quantitative and qualitative dimensions to provide decisionmakers with actionable policies to steer the energy transition, which will be central to ensuring that we can meet our climate neutrality targets in the face of a highly unpredictable future.
Partners
POLITECNICO DI MILANO and FONDAZIONE CENTRO EURO-MEDITERRANEO SUI CAMBIAMENTI CLIMATICI