*3.2. Life Cycle Modelling of the Energy System Scenarios*

The functional unit to compare the different scenarios is the provision of 2275 MWh of electricity and 5960 MWh of heat (space heating and hot water) per year. It corresponds to the actual energy consumption in 2015, which the Mainau GmbH deems representative for the future. The system boundary is cradle to grave. The inventory analysis uses the software tools Umberto and openLCA, both relying on the database ecoinvent (v3.5, cut-off system model) for background system modeling [23]. In order to assure consistency, process parameters were, wherever possible, set to the same values as in the optimization model (cf. Table 3: Parameterized processes). For PV plants the site-specific yield is taken into account. The combined heat and power (CHP) plant uses a synthetic gas produced from a wood mix harvested in the surroundings of the island. The wood chips boiler uses the same local wood mix. Wood chips production is modeled using generic datasets. Since no wood gas-fired CHP is available in ecoinvent v3.5 the biogas CHP (adapted with site-specific emission data) is combined with a gasifier dataset (see SI\_3\_Parameter and modeling for modeling details). The allocation of impacts for heat and electricity follows the ecoinvent approach based on exergy. The battery lifetime is set to 10 years [29]. The energy density corresponds to a common lithium battery [30]. Import electricity is represented by the process "market for electricity, low voltage (DE)".



Because of missing inventory data, the power to liquid plant is not explicitly modeled in S1. Instead, a credit is given to the 870 MWh surplus electric energy in order to gain functional equivalency

between S1 and the other scenarios. As the surplus energy exclusively substitutes fossil fuel combustion in the company's car fleet, the credit is derived from the fuel consumption and CO2 emissions of a medium-sized diesel car. An estimated methanol production efficiency of 50% and the heating value of diesel are assumed for credit calculation.

#### **4. Results and Discussion**

## *4.1. Scenario Comparison Based on the Total Environmental Impact*

The application of the ENsource ESM leads to an unambiguous ranking of the energy supply scenarios according to their total environmental impact (Figure 4). The alternative renewable scenarios only slightly reduce the total environmental impact compared to BAU. However, this merely reflects the fact that Mainau GmbH's current energy system already comprises a high share of renewable energies. To obtain a more complete picture, the renewable scenarios (including BAU) are compared with a fictitious carbon scenario (CS), which corresponds to electricity from the German grid and natural gas-based heat generation. Compared to CS, all renewable scenarios have a significantly lower total environmental burden (up to 30% for S1). Furthermore, the relative share of GW in the total environmental impact is significantly smaller in the renewable scenarios (minimum 26% for S1) than in CS (70%). This underlines the necessity of a multi-dimensional environmental impact analysis when comparing the renewable scenarios with each other. Contrary to the partly significant reductions in almost all other impact categories, land use (up to a factor of 20.3 in the case of S1) and mineral resources (up to a factor 3.2 in the case of S1) increase in all renewable scenarios compared to CS.

**Figure 4.** Scenario Impacts with ESM ENsource, GW = global warming, LU = land use, APP = main air pollutants and PM, ER = energy resources, HMIW = heavy metals into water, CSIA = carcinogenic substances into air, HMIA = heavy metals into air, WP = water pollutants, MR = mineral resources, WR = water resources, ODP = ozone layer depletion, WTD = non-radioactive waste to deposit.

The renewable scenarios S1 and S3 have the lowest total environmental impact. The slightly more intensive PV use, the increased usage of the wood chips boiler, and the abandonment of the natural gas boiler in scenario S2 lead to a minor reduction in impacts compared to BAU. Scenario S1 is particularly interesting as it leads to the strongest GHG reductions and is the only scenario that explicitly includes the mobility sector. The following section examines which impact categories particularly contribute to the total environmental impact.

#### *4.2. Comparative Analysis of Environmental Impacts*

Based on the ENsource ESM, it is possible to analyze the environmental trade-offs between different scenarios. Figure 4 shows absolute differences between the investigated scenarios compared to BAU for the different impact categories. The global warming (GW) reduction in scenario S3 (−36.2 MEP = mega eco-points) is mainly counteracted by increases in land use (LU, 16.2 MEP), mineral resources (MR) and heavy metal emissions into air and other air pollutants (HMIA and APP, together +4.2 MEP), leaving a clear net environmental benefit (−17.9 MEP). The significantly higher GW reduction in scenario S1 (−49 MEP) is offset by correspondingly higher negative environmental impacts in other categories: land use (LU, +19.0 MEP), mineral resources (MR, +2.7 MEP) and heavy metal emissions (+6.3 MEP both into water and air) but also carcinogenic substances (CSIA, +3.6 MEP). Consequently, despite the clearly better climate performance of S1 its total environmental impact (+18.8 MEP) is only slightly lower than for scenario S3.

To better understand the adverse effects of decarbonizing Mainau GmbH's energy supply and thus to develop context-specific remedies, the following section takes a closer look at critical technologies and life cycle processes that significantly contribute to the increasing impact indicators.

#### *4.3. Hot Spot Analysis: Technologies and Life Cycle Processes*

Increasing environmental impacts are mainly due to the additional technical infrastructure required to generate and store renewable energy, which is particularly high for scenario S1 (cf. Figure 5). Precious metals are important drivers of mineral resources (MR): Main consumers are photovoltaic cells (silver for metallization paste), inverters (gold for circuits, copper for converter), the battery (gold for circuits), and the CHP (platinum for catalytic converter). However, the mining and refining processes needed to provide those metals involve other environmental impacts as well, especially heavy metal emissions into air and water (HMIA and HMIW). The battery disposal has significant carcinogenic environmental effects (CSIA). The wood chips boiler and the CHP (both natural and wood gas) contribute mainly to the categories CSIA and other air pollutants (APP). In contrast, increased land use (LU) clearly results from operating the wood chips boiler and, to a lesser extent, the wood gas CHP.

**Figure 5.** Absolute changes (as differences expressed in MEP = 1e6 EP) of the alternative renewable scenarios compared to BAU, only significant changes considered: GW = global warming, LU = land use, APP = main air pollutants and PM, ER = energy resources, HMIW = heavy metals into water, CSIA = carcinogenic substances into air, HMIA = heavy metals into air, WP = water pollutants, MR = mineral resources.
