In the ongoing global quest to decarbonize energy systems, the critical role of societal factors is gaining unprecedented recognition. A groundbreaking multi-country study has now unveiled a comprehensive, open-source power system model that intricately integrates these factors, shedding new light on the complex socio-technical underpinnings of energy transitions. This novel model, named STONES (Socio-Technical Outlook of National Energy System), represents a significant advancement in the field of energy transition modeling by capturing not only techno-economic parameters but also the nuanced dynamics arising from societal influences across European nations.
The STONES model fundamentally challenges traditional power system simulations that often rely solely on technical and economic data. By incorporating a set of six well-defined societal transformation factors, the model simulates national power generation and capacity dynamics in 31 European countries over a transformative thirty-year period from 1990 to 2019. This sophisticated approach acknowledges that infrastructures, decision-makers, and social contexts weave together a complex tapestry that shapes the trajectory of energy transitions, beyond what purely economic optimization can capture.
Central to the STONES model is its recursive bottom-up structure, employing a one-year time step to dynamically simulate the evolution of electricity systems at a granular level. Incorporating a detailed representation of 14 electricity generation and storage technologies, the model uses rich, country-specific input data — from technology costs and resource potentials to flexibility constraints — to output annual projections of installed capacities and electricity generation. This level of detail allows the model to bridge technology-specific features with broader system-level socio-technical interactions.
One of the model’s core modules is the electricity demand module, which generates load duration curves that reflect both observed and projected electricity consumption patterns. This module takes hourly demand profiles alongside renewable energy output characteristics, employing data aggregation techniques that condense annual time series into six representative days. This approach achieves computational efficiency while preserving essential temporal demand variability, a pivotal factor for reliable dispatch and capacity planning.
The dispatch module follows, employing a cost-minimizing algorithm to optimize electricity generation dispatch across the modeled technologies, ensuring system adequacy within the constraints imposed by generation and transmission capabilities. This optimization incorporates hourly demand constraints over the representative days and weighs variable operating costs of technologies, thereby illustrating the economic merits of diverse generation mixes within each national context.
The capacity expansion module completes the model’s yearly cycle by forecasting installed capacities over a five-year horizon. It estimates required system capacity based on peak demand projections and planned retirements while segmenting load profiles into peak, intermediate, and base load categories. This segmentation enables the model to allocate investments across technologies according to their cost-effectiveness for each segment, employing a multinomial logit function to realistically represent heterogeneous investor preferences and partial adoption pathways.
Societal factors are embedded within the model through sophisticated mechanisms that capture infrastructure inertia, actor heterogeneity, and social-political dynamics influencing technology adoption. For instance, the model represents infrastructure lock-in by permitting extensions of existing coal, gas, and nuclear capacity lifespans beyond nominal limits, reflecting real-world path dependencies and investment rigidities. Sensitivity analyses probe the effects of varying retirement rates and lifetime extensions, offering insights into how these factors either retard or accelerate energy transitions.
Actor heterogeneity is woven into investment decision-making by modulating cost sensitivities within the multinomial logit framework. This reflects the reality that diverse investors weigh technology costs differently, influenced by risk perceptions and strategic considerations. By calibrating parameters to replicate observed market behavior where cost savings do not translate linearly into market shares, the model adds a critical layer of realism commonly absent in techno-economic simulations.
Investment risk is explicitly captured through the incorporation of country- and technology-specific weighted average costs of capital, replacing uniform discount rates. This adjustment recognizes that financial parameters profoundly influence adoption speeds of capital-intensive technologies such as renewables, revealing how evolving financial landscapes impact long-term capacity choices in different national settings.
Social and institutional contexts exert further influence through constraints on public acceptance and governance structures. The STONES model leverages survey data on public perceptions to impose threshold-based restrictions on technology deployment, simulating, for example, the de facto moratorium on nuclear expansion in Germany during the 2010s or slowed wind development due to local opposition. Correspondingly, positive public sentiment is modeled to reduce investment costs for renewables, capturing feedback loops where societal support enables policy and financial incentives.
Governance variables—such as market liberalization and public ownership of utilities—are introduced via adjustments to investment costs for solar and wind projects, based on OECD market regulation indicators. These parameters illustrate the profound impact of regulatory environments on energy transitions, highlighting how entry barriers and state control shape technology diffusion patterns. The model’s sensitivity analyses explore a range of governance scenarios, elucidating how shifts in policy landscapes could modulate decarbonization pathways.
Leveraging an extensive, harmonized dataset spanning three decades, the study grounds the model in robust empirical evidence. Historical data on technology costs, generation, capacity dynamics, emissions intensities, and load profiles for 31 countries were meticulously assembled, ensuring consistency and completeness to fuel accurate simulations. This dataset, drawn from diverse open-access sources and national statistics, stands as a vital resource for replicability and further research.
To rigorously evaluate the model’s capacity to reconstruct historical energy trajectories, the researchers performed hindcasting simulations, testing scenarios with all possible combinations of societal factors represented or omitted. This exhaustive approach generated 64 hindcasting runs per country, enabling a granular assessment of each societal factor’s contribution to modeling accuracy. By comparing simulated outcomes to observed historical values for installed capacity, CO2 emissions, and renewable generation shares, the team quantitatively assessed performance.
The evaluation employed the Symmetric Mean Absolute Percentage Error (SMAPE), an absolute metric that avoids canceling effects of positive and negative deviations and facilitates aggregations across technologies. Results demonstrated that incorporating societal factors substantially improved the model’s fidelity, reducing errors in capacity and emissions projections. An analysis of variance identified which societal parameters most significantly influenced prediction variance, revealing vital levers for more precise energy system modeling.
Interestingly, the hindcasting highlighted that no single societal factor dominates; rather, their interactions collectively modulate the pace and direction of energy transitions. For instance, infrastructure inertia and public acceptance interplay to define the speed at which legacy technologies are phased out, while actor behavior and governance environments jointly shape renewable uptake dynamics. This intricate coupling of technical and social components underscores the necessity of integrated modeling frameworks.
The implications of this work are profound. As policymakers and planners grapple with design of energy systems resilient to climate change, models like STONES provide a richer, multidimensional basis for scenario analysis that accounts for social realities. This advances beyond techno-economic paradigms by quantifying the societal influences that have historically shaped energy transitions, offering pathways to anticipate future trajectories amid evolving social landscapes.
Moreover, the open-source nature of the STONES model invites broad collaboration and iterative refinement. By democratizing access to a sophisticated socio-technical simulation platform, the research empowers diverse stakeholders—from researchers to policymakers—to examine and adapt model assumptions in alignment with local contexts or emerging data. Such transparency is critical for fostering trust and accelerating innovation.
This study also encourages a paradigm shift in energy economics and modeling communities. Moving toward consistent inclusion of societal parameters challenges entrenched assumptions of purely rational, cost-driven decision-making, acknowledging the profound impacts of social acceptance, institutional settings, and investor heterogeneity. As energy systems become more decentralized and participatory, capturing these dimensions grows ever more essential.
Ultimately, this model offers a versatile tool to explore scenarios that were previously opaque—how varying degrees of public support, governance reforms, or market behaviors could accelerate or hinder decarbonization. By simulating past decades with high fidelity, the STONES framework equips decision-makers with a roadmap informed by the intricate reality of past transitions, enabling more grounded and adaptive energy planning.
In an era when the urgency of addressing climate change intersects with societal complexity, models such as STONES represent a leap forward. They challenge the field to embrace socio-technical intricacies as integral rather than peripheral, opening new avenues for holistic energy transition strategies that not only envision a low-carbon future but also understand the human and institutional pathways to reach it.
Subject of Research:
Not specified.
Article Title:
Not specified.
Article References:
Fisch-Romito, V., Jaxa-Rozen, M., Wen, X. et al. Multi-country evidence on societal factors to include in energy transition modelling. Nat Energy (2025). https://doi.org/10.1038/s41560-025-01719-7
Image Credits: AI Generated
Tags: complex interactions in power generationdecarbonization and societal influencesenergy transition modelingEuropean energy transition insightslong-term energy system simulationsmulti-country energy studiesnational power capacity dynamicssocietal factors in energy systemssocietal transformation in energy policiessocio-economic parameters in energysocio-technical dynamics of energySTONES power system model
