Beyond climate change. Multi-attribute decision making for a sustainability assessment of energy system transformation pathways

A multitude of pathways for decarbonizing energy systems have been formulated. In the development of these scenarios, the focus is often only on system costs at a given CO2 emission reduction. However, when assessing the sustainability of energy systems in a broader sense, many more aspects need to be considered: In addition to greenhouse gas emissions, energy systems induce further environmental and socio-economic impacts and must meet requirements for security of supply and cost efficiency. For assessing the compatibility of a future energy system with sustainability concepts, alternative pathways must be compared holistically using dedicated assessment methods, which is facilitated by multi-attribute decision making methods (MADM). With the target of identifying sustainable transformation pathways, we assess ten transformation scenarios for the example of Germany, using the three MADM methods weighted sum method, PROMETHEE II and TOPSIS. We find that top ranks are not completely stable across methods, but there are scenario clusters which rank high, medium and low for all methods. In the top ranks, there are both less ambitious scenarios that aim at reducing direct CO2 emissions by 80%, and more ambitious scenarios with a reduction by at least 95%. We conclude that scenarios with more ambitious climate protection goals are not necessarily more (or less) sustainable than scenarios that aim for a reduction of 80% only.

Automated summary

When assessing the sustainability of energy systems, it is important to consider not only greenhouse gas emissions, but also other environmental and socio-economic impacts, as well as requirements for supply security and cost efficiency. Multi-attribute decision making methods can be used to holistically compare alternative pathways. In the case of transformation scenarios for Germany, rankings varied across different assessment methods, but clusters of high, medium, and low-ranked scenarios were identified.