Impact of integrating societal factors on the accuracy of optimization-based electricity system modeling in 31 European countries

RENEWABLE ENERGY SYSTEMS 1
Grant no. 186834 (ACCURACY)
Impact of integrating societal factors on the accuracy of
optimization-based electricity system modeling in 31
European countries
Xin Wen 1
Qin Alexander Crebas 1, 2
Kenneth Bruninx 2
Evelina Trutnevyte 1
1 Renewable Energy Systems group,
University of Geneva
2 Policy and Management, Faculty of Technology, Delft
University of Technology
SUMMER 2025 Semi-annual ETSAP Meeting
June 9, 2025
Nara, Japan

Backgroud
Energy system models should integrate societal factors for feasibility:
• Techno-economic models often fail to capture the real-world dynamics (Geels et al., 2016;
Pfenninger et al., 2014; Schubert et al., 2015).
• Energy scenarios preferred by citizens and experts tend to highly deviate from the model-based
scenarios (Xexakis et al., 2020; Xexakis and Trutnevyte, 2021).

Backgroud
Energy system models should integrate societal factors for feasibility:
• Techno-economic models often fail to capture the real-world dynamics (Geels et al., 2016;
Pfenninger et al., 2014; Schubert et al., 2015).
• Energy scenarios preferred by citizens and experts tend to highly deviate from the model-based
scenarios (Xexakis et al., 2020; Xexakis and Trutnevyte, 2021).
Despite attempts to integrate the societal aspects into modeling, it is still unknown what are their impacts
on the quality of the model (Trutnevyte et al., 2019, Fisch-Romito et al., 2024).

Research question
By hindcasting exercises in 31 European countries, how accurate are the following model
versions when integrating one or multiple societal aspects in electricity system modeling?

Research question
By hindcasting exercises in 31 European countries, how accurate are the following model
versions when integrating one or multiple societal aspects in electricity system modeling?
• EU climate policy (emission targets) (Delreux and Ohler, 2019)
• Perceived seriousness towards climate change (Schubert et al., 2015; Devine-Wright et
al., 2007; Süsser et al., 2022)
• Investment risks (Li et al., 2017; De Cian et al., 2020, Stavrakas et al., 2019)
• EU climate policy with investment risks
• Perceived seriousness-adjusted climate policy with investment risks

* D-EXPANSE : Dynamic version of EXploration of PAtterns in Near-optimal energy ScEnarios (Trutnevyte, 2016; Wen et al. 2022)
Methods: Hindcasting by D-EXPANSE
(Jaxa-Rozen et al., 2022)
• EU climate policy
• Perceived seriousness towards
climate change
• Investment risks
Societal aspects

Hindcasting: Cost-optimization based
electricity sector modeling
• Techno-economic version
• Versions with societal aspects
31 national models (EU27, UK, Switzerland, Norway, Iceland)
D-EXPANSE* model
Least-cost pathways
Historical pathway 1990–2019
Projection year 1990
1991
2019
...
(Wen et al., 2022)
climate change
Societal aspects

Hindcasting: Cost-optimization based
electricity sector modeling
• Techno-economic version
• Versions with societal aspects
31 national models (EU27, UK, Switzerland, Norway, Iceland)
D-EXPANSE* model
Least-cost pathways
Historical pathway 1990–2019
1991
2019
...
(Wen et al., 2022)
Deviations between pathways
under four model versions
Accuracy quantification
Deviations of projections in different projection years
under six model versions
• sMAPE (Symmetric Mean Absolute Percentage Error)
• RMSLE (Root-Mean-Squared Logarithmic Error)
.
1991
.
2019
.
climate change
Societal aspects

Methods
• EU climate policy (emission targets) (Delreux, 2019)
Emission targets are set as emission constraints in the model.
• 1) Yearly emission targets
• 2) Cumulative emission credits
0
20
40
60
80
100
120
1990 2000 2010 2020 2030
Targeted
emission
level
(%
of
1990
level)
CO2 emission targets

Methods
• Perceived seriousness towards climate change (among EU citizens)
Emission targets are adjusted based on the quantified perceived seriousness scores and set as emission
constraints in the model.
Source: Eurobarometer, Climate change, and Europeans’Attitudes towards Climate Change, 2009–2019.
0
20
40
60
80
100
120
1990 2000 2010 2020 2030
Targeted
emission
level
(%
of
1990
level)
Adjusted emission target
based on quantified scores

Methods
• Investment risks with WACC values (Weighted Average Cost of Capital) (Polzin et al, 2021)
Technology costs are recalculated by considering country-specific and technology-specific WACC values.
Constant assumptions
before 2009
Uniform WACC values
Lines in color:
Country-specific WACC values in time seires

Results: Accuracy quantified by sMAPE (Symmetric Mean Absolute Percentage Error)
With societal factors:
• Emissions are less over-estimated.

Results: Accuracy quantified by sMAPE (Symmetric Mean Absolute Percentage Error)
With societal factors:
• Emissions are less over-estimated.
• Installed capacity and annual generation: No single
model version shows a clear improvement in accuracy
across all the countries.

Results: Accuracy quantified by RMSLE (Root-Mean-Squared Logarithmic Error)
• Inclusion of societal aspects increases the accuracy, albeit sometimes very little.

Results: Accuracy quantified by RMSLE (Root-Mean-Squared Logarithmic Error)
• Inclusion of societal aspects increases the accuracy, albeit sometimes very little.
• Inclusion of more societal aspects does not necessarily mean a better accuracy.

Results: Technology-specific accuracy
• For wind power and solar PV, there is no clear systematic tendency of accuracy improvement.
Renewable
technologies

• For wind power and solar PV, there is no clear tendency of accuracy improvement.
• The inclusion of investment risks tends to improve the accuracy of the generation for nuclear power (e.g.
Netherland and Latvia).
Renewable
technologies
Low-carbon incumbent
technologies, such as
nuclear power

• For wind power and solar PV, there is no clear tendency of accuracy improvement.
• The inclusion of investment risks tends to improve the accuracy of the generation for nuclear power (e.g.
Netherland and Latvia).
• When considering investment risks, gas tends to be favoured over nuclear power and hard coal.
Renewable
technologies
Low-carbon incumbent
nuclear power
Combustion-based
gas

Takeaways
• Integrating investment risks has the most substantial effect on improving the hindcasting performance among the
three societal aspects considered, although not always.
• The highest improvements are observed for gas, hard coal, nuclear power, net import, waste incineration, and
onshore wind power (when integrating investment risks).
• Strikingly, modelling of the historic EU climate policies rarely improved the hindcasting performance as
compared to the model version without policies.
• There is coevolution of multiple societal aspects, e.g. interaction between EU climate policy and investment
risks.
Future work
• We encourage further studies to test similar methods using hindcasting and accuracy quantification.
• More investigation on better modelling of climate policy.

Grant no. 186834 (ACCURACY)
Thank you!
Do not hesitate to reach out!
xin.wen@unige.ch
SUMMER 2025 Semi-annual ETSAP Meeting
June 9, 2025
Nara, Japan

Trutnevyte, E., 2016. Does cost optimization approximate the real-world energy transition? Energy 106, 182–193. https://doi.org/10.1016/j.energy.2016.03.038
Jaxa-Rozen, M., Wen, X., Trutnevyte, E., 2022. Historic data of the national electricity system transitions in Europe in 1990–2019 for retrospective evaluation of models. Data in Brief 43, 108459. https://doi.org/10.1016/J.DIB.2022.108459
Wen, X., Jaxa-Rozen, M., Trutnevyte, E., 2022. Accuracy indicators for evaluating retrospective performance of energy system models. Applied Energy 325, 1–30. https://doi.org/10.1016/j.apenergy.2022.119906
Geels, F.W., Berkhout, F., Van Vuuren, D.P., 2016. Bridging analytical approaches for low-carbon transitions. Nature Climate Change 6, 576–583. https://doi.org/10.1038/nclimate2980
Pfenninger, S., Hawkes, A., Keirstead, J., 2014. Energy systems modeling for twenty-first century energy challenges. Renewable and Sustainable Energy Reviews 33, 74–86. https://doi.org/10.1016/j.rser.2014.02.003
Schubert, D.K.J., Thuß, S., Möst, D., 2015. Does political and social feasibility matter in energy scenarios? Energy Research & Social Science 7, 43–54. https://doi.org/10.1016/J.ERSS.2015.03.003
Xexakis G, Hansmann R, Volken SP, Trutnevyte E. Models on the wrong track: Model-based electricity supply scenarios in Switzerland are not aligned with the perspectives of energy experts and the public. Renewable and Sustainable
Energy Reviews 2020;134:110297. https://doi.org/10.1016/j.rser.2020.110297.
Xexakis G, Trutnevyte E. Consensus on future EU electricity supply among citizens of France, Germany, and Poland: Implications for modeling. Energy Strategy Reviews 2021;38:100742. https://doi.org/10.1016/j.esr.2021.100742.
Trutnevyte, E., Hirt, L.F., Bauer, N., Cherp, A., Hawkes, A., Edelenbosch, O.Y., Pedde, S., van Vuuren, D.P., 2019. Societal Transformations in Models for Energy and Climate Policy: The Ambitious Next Step. One Earth 1, 423–433.
https://doi.org/10.1016/j.oneear.2019.12.002
Fisch-Romito V, Jaxa-Rozen M, Wen X, Trutnevyte E. Multi-country evidence on societal factors to include in energy transition modeling (under review). Nature Energy 2024.
Schubert, D.K.J., Thuß, S., Möst, D., 2015. Does political and social feasibility matter in energy scenarios? Energy Research & Social Science 7, 43–54. https://doi.org/10.1016/J.ERSS.2015.03.003
Devine-Wright P, Grubb J, Pollitt U. Reconsidering Public Acceptance of Renewable Energy Technologies: a Critical Review. Taking Climate Change Seriously: a Low Carbon Future for the Electricity Sector, Cambridge University Press;
2007.
Süsser D, Martin N, Stavrakas V, Gaschnig H, Talens-Peiró L, Flamos A, et al. Why energy models should integrate social and environmental factors: Assessing user needs, omission impacts, and real-word accuracy in the European Union.
Energy Research and Social Science 2022;92:102775. https://doi.org/10.1016/j.erss.2022.102775.
Li FGN, Strachan N. Modelling energy transitions for climate targets under landscape and actor inertia. Environmental Innovation and Societal Transitions 2017;24:106–29. https://doi.org/10.1016/j.eist.2016.08.002.
De Cian E, Dasgupta S, Hof AF, van Sluisveld MAE, Köhler J, Pfluger B, et al. Actors, decision-making, and institutions in quantitative system modelling. Technological Forecasting and Social Change 2020;151:119480.
https://doi.org/10.1016/j.techfore.2018.10.004.
Stavrakas V, Papadelis S, Flamos A. An agent-based model to simulate technology adoption quantifying behavioural uncertainty of consumers. Applied Energy 2019;255:113795. https://doi.org/10.1016/j.apenergy.2019.113795.
Delreux T, Ohler F. Climate Policy in European Union Politics. Oxford Research Encyclopedia of Politics, Oxford University Press; 2019. https://doi.org/10.1093/acrefore/9780190228637.013.1097.
References

Methods
0
20
40
60
80
100
120
1990 2000 2010 2020 2030
Targeted
emission
level
(%
of
1990
level)

• Perceived seriousness towards climate policy (among EU citizens)
Source: Eurobarometer, Climate change, and Europeans’ Attitudes towards Climate Change, 2009–2019
Methods
0
20
40
60
80
100
120
1990 2000 2010 2020 2030
Targeted
emission
level
(%
of
1990
level)

• Perceived seriousness towards climate policy (among EU citizens)
Source: Eurobarometer, Climate change, and Europeans’ Attitudes towards Climate Change, 2009–2019
Methods
0
20
40
60
80
100
120
1990 2000 2010 2020 2030
Targeted
emission
level
(%
of
1990
level)
Adjusted emission target based on quantified scores

Result: The Netherlands

Result: Latvia

Results: Italy

Results: Portugal

Finland

Sweden

Impact of integrating societal factors on the accuracy of optimization-based electricity system modeling in 31 European countries

More Related Content

Similar to Impact of integrating societal factors on the accuracy of optimization-based electricity system modeling in 31 European countries (20)

More from IEA-ETSAP (20)

Recently uploaded (20)

Impact of integrating societal factors on the accuracy of optimization-based electricity system modeling in 31 European countries