*2.3. Estimation of the CO2 Saving Potential Using Life Cycle Assessment with openLCA*

The ICU concept offers the opportunity to save CO2 due to reduced transport of biowaste and food [40]. To quantify the amount of saved CO2, a life cycle assessment (LCA) with the open-source software openLCA (version 1.10.3) was conducted. This software considers the total energy consumption by all components at various levels. The used database was ecoinvent35\_Cut [41]. The system environment was divided into five phases: extraction of raw materials and energy sources, manufacture, use, transport and disposal (Figure 3A) [42]. The boundaries for the ecological assessment are shown in Figure 3A (grey dots). An average distance of 30 km for the transport of biowaste to the ADP in the conventional scenario was assumed. In the ICU concept, transport was neglected (Figure 3B). All flows and process data are found in supplementary file 4. The system contains specific elements providing the functional unit of 1 kg biomass for the complete life cycle. The input flow is 11 t with bio-degradable garden and park waste, food and kitchen waste from households, restaurants, caterers and retail stores, comparable waste from food processing plants, as well as forestry or agricultural residues, and manure. It does not contain sewage sludge or other biodegradable waste such as natural textiles, paper or processed wood.

**Figure 3.** System boundaries of the life cycle assessment study. (**A**): System environments of the product life-cycles based on DIN 15804 in four steps and system boundary (functional unit 1 m<sup>3</sup> biogas). (**B**): System boundary for the ICU concept in comparison to a building with transport of biowaste to a conventional biogas plant. (Energy: Available online: https://biogas.fnr.de/daten-und-fakten/faustzahlen (accessed on 22 July 2020)).

#### *2.4. Cost Calculation*

The costs of implementing the ICU concept were estimated by a life cycle cost analysis (LCCA, Equation (1)) for the example of a building with 100 residents. Therefore, the costs for acquiring and operating the fermenters, the soiling, the hydroponic systems, the technical staff and the building's space were considered taking into account the net present value (*NPV*).

$$NPV = \mathbb{C} + R - S + A + M + E \tag{1}$$

*C* = investment costs;

*R* = replacement costs;

*S* = resale value at the end of study period;

*M* = non-annually recurring operating, maintenance and repair cost;

*E* = energy costs.

Investment costs (*C*) were depreciated for a period of 20 years. Replacement costs were assumed with 10% of the investment costs after 20 years and maintenance costs (*A* + *M*) with annually 5% of the investment costs. Energy costs (*E*) were omitted since the system produces the required energy on its own. Additionally, the resale value (*R*) of the installations was assumed as "0" € as it was expected that the building's value remains at least stable. Additionally, ADP operation and plant cultivation require an experienced worker requiring at least 25 € per hour.

The production of in-house biogas generates energy in form of electricity and heat. This energy is reused inside the ICU building. If the generated energy were sold, the price for 1 kWh heat and 1 kWh electricity would be 0.024 € [43] and 0.1 € (§43 EEG 2017), respectively. A summary of results obtained is found in Table 1.


**Table 1.** Model output per m2 ground area and year.

\* 500 m<sup>2</sup> greenhouse is for industry the smallest size for implementation.

Against these costs, the value of the produced energy and food was considered. The benefit of the reduction in the disposal of biowaste and wastewater was neglected to avoid further complication of the calculation. Furthermore, the installation of vacuum toilets and separate black and grey water tubes is also cost-intensive. However, the cost of the installations compensates with the benefit of a reduced wastewater volume.

#### *2.5. Overview of Important Social-Cultural Aspects Required for the Implementation*

To assess social-cultural aspects for the implementation of the ICU concept in Germany, a literature survey was performed addressing the following questions:


While for most of these questions results and data from literature already exist, the question on the real estate owners' willingness to implement an ICU concept has not been addressed, so far. Therefore, an online survey to collect this data was performed. To obtain a comprehensive picture of the attitude towards this new concept, 235 real estate owners, about 15 from every Federal State, were selected. All owners received a short online questionnaire containing ten questions (Supplementary File 5) to rate to which extend different aspects of the ICU concept and its implementation are important to them. In the end, only 14 answers were received.

#### **3. Results and Discussion**

#### *3.1. Structure of Section 3*

This feasibility study evaluates the amount of energy (Section 3.1) and food (Section 3.3), which could be produced by implementing an ICU concept for a building with 100 residents. Before the final simulation, the ADPs' fermenter size and configuration were evaluated and the best scenario for house-internal food production was selected. A precondition for plant growth was converting NH4 in the digestate to NO3 (Section 3.2).

Based on the ICU-concept's best scenario, the costs were calculated (Section 3.4) and the potential CO2-savings (Section 3.5). Finally, social–cultural aspects were reviewed, including the laws required for implementing the concepts (Section 3.6) and potential addons for the ICU-concept (Section 3.7).

3.1.1. Utilization of Biowaste by Optimized Anaerobic Fermenters Enable to Cover 21% of the Annual Energy Demand of the Building

As the first step, the size and performance of CSTR, PFR, and 2sR ADP for processing of biowaste (Figure 4) and black water (Figure 5) were compared based on the energy content of the biogas. The volume ratio between the hydrolysis and the main fermenter of the 2sR was 1:50 as determined in the supplementary Table S6. For the PFR the sum of all five fermenters connected in series was assumed for the simulation.

**Figure 4.** Evaluation of reactor scenarios with biowaste fermenters. Energy content of the biogas produced annually (without losses). Only biowaste input. For PFR reactor the sum of all 5 fermenters and for 2sR is hydrolyse + main fermenter are considered. Marker shows the chosen reactor size.

**Figure 5.** Evaluation of reactor scenarios with additional blackwater fermenter. Energy content of the biogas produced annually (without losses). Blackwater input. For PFR reactor the sum of all 5 fermenters and for 2sR is hydrolyse + main fermenter are considered. Marker shows the chosen reactor size.

The first scenario was the calculation of biowaste input in one fermenter with an average amount of 33.6 kg biowaste per day (Figure 4). Depending on the reactor size 1 kW energy may be produced. The second scenario calculates the additional black water fermentation in a second fermenter with an average of 720 L black water per day (Figure 5). Black water may improve the daily energy yield to 3 kW energy, fitting to the values of studies with similar substrates [44].

For the first scenario, the simulation of PFR produces about 5.5% more energy than the two-step and 22% more than the CSTR fermenter. These magnitudes between the fermenter types were also shown by Bensmann et al. [45,46]. Additionally, a PFR is more robust against contaminants like plastic material in biowaste. The shape of the power to fermenter size curve is sigmoid, reflecting that too small fermenter sizes lead to acidification, whereas too large fermenter adds no further benefit (Figure 4). As optimal biowaste fermenter sizes were chosena4m<sup>3</sup> CSTR-fermenter, five fermenters connected in series with each 1 m3 for the PFR-fermenter and also 4 m3 for the main fermenter of the 2sR (Figure 4). The PFR-fermenter was selected as optimal because with 11.434 kWh calorific value annually it was able to produce the most energy. This amount of energy corresponds to 9.5% of the annual energy demand of 100 persons [47]. Since the reactors with their control units require less than 20 m2, installation in the technical center of a building is technically feasible. Production of heat and electricity would require an additional CHP unit of about 10 m2 size. Alternatively, the biogas can be used to cook and climatize the building. This scenario requires that the building have gas heating/heating pumps instead of an oil or electric system. Alternatively, the biogas could also be upgraded to biomethane and fed into the local gas grid [48,49].

In comparison the second scenario with an additional black water produce 25,855 kWh energy. This scenario is ecological more efficient to the first one because it can cover 21% of the yearly energy production (Figure 5). As optimal reactor size for the CSTR a 9 m<sup>3</sup> fermenter, for PFR five fermenters connected in series with each 1.8 m3, and for 2sR 9 m3 were chosen.

Due to the higher energy content, the additional fermentation of black water should be considered for the ICU-concept. For black water usage, the separation of grey and black water is needed. This required a two-pipe system and separation or vacuum toilets (e.g., Jenfelder Au, [50]). The implementation is technically demanding, but allows the reuse of greywater, which would further reduce water consumption.

#### 3.1.2. Dynamic Behavior of the Anaerobic Digester

For the first simulation a constant supply of biowaste and black water with the chance that fluctuations occur was assumed. To assess the dynamic behavior of the system, after initial conditions (i.), a shift in the feeding rate from (ii.) constant 33.6 kg/d to (iii.) an increase of biomass of 18% for four months (iii.) to (iv.) a feeding rate of 30.7 kg/d for only five days a week was considered (Figure 6). The simulation of all three fermenter types (2sR, CSTR and PFR) shows smooth transitions for the different changes in the feeding strategy indicating an overall stable process behavior. This suggests that the ICU concept could be easily integrated into buildings. However, systems for handling fluctuations in gas production such as gas storage tanks or gas torches should be considered in case of technical problems (data not shown). Furthermore, it is known that ICU systems are prone to long periods of biomass overloading [45] and require a relatively long period for the start-up (Figure 6).

**Figure 6.** Dynamic behavior of the ICU model. Feed fluctuations: Energy production from biowaste (green) and additional black water (BW) (brown) for 2sR, CSTR and PFR fermenters. (**i.**): Initiation phase of the model. (**ii.**) Feeding rate of 33.6 kg/d (**iii.**) Feeding rate of 41.1kg/d (**iv.**) Feeding rate of 30.7 kg/d for five days a week.
