**5. Summary and Recommendations Concerning the Mainau GmbH**

The preceding analysis shows that an increased PV use by Mainau GmbH is ecologically beneficial, although GHG savings are partially offset by increased mineral resources and heavy metal emissions (cf. scenario S3 in Figure 6). The inclusion of the mobility sector in scenario S1 makes sense in terms of achieving climate neutrality. However, the continuous operation of the power to liquid system requires a substantial increase of PV capacity, the installation of a heat storage unit, and a battery, as well as the increased use of the wood gasifier. This leads to increases in most of the investigated impact categories that largely offset the GW reductions from substituting fossil fuels. On the other hand, the additional storage capacities entail a completely self-sufficient energy supply. This may be considered an additional advantage in terms of energy systems decentralization, since necessary system services and infrastructure are also partially decentralized. This perspective would put into question the functional unit and system boundaries chosen to compare S1 to the other scenarios.

**Figure 6.** Share of energy carriers and storage technology for several impact categories, LU = land use, CSIA = carcinogenic substances into air, MR = mineral resources, APP = main air pollutants and PM HMIW = heavy metals into water, HMIA = heavy metals into air.

The hot spot analysis reveals the main drivers of negative environmental impacts associated with an increased use of renewable energies: substantial increases in land use (LU), air pollutants (APP, CSIA), and heavy metal emissions (HMIW, HMIA). The latter is closely linked to the mining activities for raw materials for photovoltaic modules and storage technologies. Land use is increasing due to the use of wood as an energy carrier. These are potential levers for further improving the energy system design: e.g., by switching to CdTe cells instead of multi-silicon cells, which cause lower environmental impacts [31] or by installing PV modules from manufacturers that use secondary raw materials or pay attention to high environmental standards in the extraction of raw materials. For the wood gasifier and wood chip boiler, wood waste or wood from extensive forestry use could be used.

The present analysis suggests extending the set of considered scenarios. In particular, the commitment to local energy production seems unnecessarily restrictive. Imported wind energy could be a resource-saving alternative to photovoltaics. The current scenarios are based on an exclusive cost optimization. The integration of the environmental indicators derived from the ENsource ESM as optimization objectives or constraints could help to identify even more environmentally friendly energy supply scenarios.

#### **6. Conclusions and Outlook**

#### *6.1. Environmental Trade-O*ff*s of Renewable Energy Systems*

This paper presents an updated ESM for Germany, which includes mineral resource use of renewable energy systems. From a practical point of view, the ENsource ESM quantifies environmental trade-offs between different renewable technologies and scenarios and provides a single score for an unambiguous environmental ranking of different options. Eco-factors that integrate external normalization and distance-to-target weighting make different impact categories comparable in terms of their relevance for a specific decision context. Applying the ENsource ESM to the case of a decentralized renewable energy system confirms the increased demand for mineral resources found by other authors [3,6]. However, it also shows that the major contribution to the total environmental impact stems from pollutant emissions associated with the mining activities. In addition, increasing land use due to biomass combustion is emerging as a potential problem area. The shift in importance from mineral resources to pollutant emissions related categories and land use could be due to extending renewable energy system analysis from electricity production to heating and cooling as well as mobility. To explore the findings' generalizability from the case study, the ENsource ESM could be applied to other energy systems in Germany.

#### *6.2. Missing Targets—Limitations and Transparency*

As for other ESM [17,21], the presented environmental assessment is mainly limited by missing legally binding, quantitative targets for some impact categories (cf. Table 1). In these cases, provisional targets were derived in the most plausible way possible from suitable references, such as EU regulations or directives or governmental strategy reports. However, this clearly limits the democratic legitimacy, which is one of the most important arguments in favor of ESM in general [32]. This study suggests nevertheless that the ESM's mathematical simplicity makes it a particularly transparent approach that allows for a critical reflection and, if necessary, case-specific adaptation of weights for environmental impacts. It moreover makes it possible to involve practical decision-makers in this process by presenting and communicating key assumptions in a structured way. In that sense, we take in the following a closer look at the weights for heavy metal emissions and land-use, which have been identified in the case study as the main adverse environmental effects of renewable technologies, as well as the weight for global warming, which is without doubt particularly important and involves a legal zero-emission target.

With regard to heavy metal emissions into water, it can be objected that the ENsource ESM does not regionalize weights. This means that policy targets (*T*) and pollutant emissions (*A*) for the Rhine are effectively transferred locally to the Mainau and, more questionably, in terms of representativeness, to the entire world. The latter directly affects the validity of our results because pollutant emissions from mining processes occur outside Germany. As environmental standards are lower in many countries than in Germany or even do not exist (*T* = ∞), regionalized weights would be much smaller or even zero. This would improve the environmental score or the investigated renewable scenarios. However, we follow the argumentation of Frischknecht and Büsser-Knöpfel, who prefer in such cases to apply weights derived from German environmental standards in order to prevent "environmental dumping" [13].

The indicator for land use in the ENsource ESM refers to land occupation. Since land occupation is not limited by law in Germany, the corresponding weight in the eco-factor is based on a limit for land transformation (measured in ha/a) taken from Germany's sustainability strategy instead [26]. Other approaches to derive the weighting in a more consistent way would be conceivable: e.g., to relate the worldwide land use by German consumption (*A*) to the usable area available in Germany (*T*). Regarding our results, this would tend to increase the contribution of land use to the total impact but leave the ranking of the alternative renewable scenarios (S1, S2, and S3) unchanged. Even though methodically appealing, this approach has been abandoned for the present study, as it lacks democratic legitimacy.

The ESM's transparent way of defining weights is even an advantage where the legal basis is clear. Regarding policies aimed at climate neutrality, the fundamental methodological question arises as to how to deal sensibly with a zero-emission target. Mathematically, this leads to infinite weights (Equation (2)) and, thus, eventually to binary weighting. As pointed out initially, there are good reasons to refuse such a narrowing of the focus. At the same time, the practical way of dealing with this limit within the ENsource ESM provides evidence that a strong emphasis on climate change, does not necessarily lead to neglecting other environmental impacts.
