**1. Introduction**

One of the most demanding challenges for the future is progressive global warming caused by excessive carbon dioxide (CO2) emissions and other greenhouse gases. To stop global warming, our society must reduce the CO2 emission and make our entire lifestyle CO2-neutral. While many concepts for sustainable electrical energy production already exist, CO2-neutral agriculture and biomass circulation concepts are lacking. Additionally,

**Citation:** Meinusch, N.; Kramer, S.; Körner, O.; Wiese, J.; Seick, I.; Beblek, A.; Berges, R.; Illenberger, B.; Illenberger, M.; Uebbing, J.; et al. Integrated Cycles for Urban Biomass as a Strategy to Promote a CO2-Neutral Society—A Feasibility Study. *Sustainability* **2021**, *13*, 9505. https://doi.org/10.3390/ su13179505

Academic Editor: Carlos Morón Fernández

Received: 22 July 2021 Accepted: 20 August 2021 Published: 24 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

since half of the world population now lives in cities, these concepts have to be also applicable to urban areas. For example, in some urban districts (e.g., the Jenfelder Au in Hamburg, Germany [1]) black water is used on-site to produce heat and electricity by anaerobic digestion (AD). Furthermore, roof-top gardens enable the production of food in the cities [2]. Fuldauer et al. [3] demonstrated that it is even feasible to connect a small-scale anaerobic digestion plant (ADP) with a hydroponic or algae cultivation system to close the biomass cycle. The benefits of closed urban biomass cycles are an efficient utilization of the resources and the avoidance of transport [4].

Although there is plenty of research about single methods for biomass recovery (e.g., biogas plants) [5–8], fertilizer production from bio waste [9], or urbane farming [10,11], there is a lack of concepts for closed biomass circles and systematic feasibility evaluations.

Therefore, this study investigates the potentials and limitations of a concept for urban biomass circulation regarding energy and food production, carbon dioxide equivalent (CO2 eq) savings, costs, and social–cultural aspects in Germany. The concept, called Integrated Cycles for Urban Biomass (ICU), demands an in-house ADP to degrade biowaste from residential buildings to biogas and digestate. The biogas generated is converted on-site to heat and electricity through a combined heat and power plant (CHP). The remaining fermenter liquid is upgraded by a soiling®-process (EP3684909A1 [12]) and a nitrification process to refined fertilizer. Finally, the liquid fertilizer is used to produce fruits, vegetables, and ornamental plants using either in-house integrated hydroponic systems, soil-based agriculture or roof-top gardens. Finally, the residents can consume the food while the accruing plant residues are fed into the ADP to close the biomass cycle again (Figure 1). Whereas soil-based agriculture is more robust, the use of hydroponic systems for the production of vegetables enables faster growth, higher product quality and needs less space [13].

**Figure 1.** Process chart of ICU concept: Integrated Cycles for Urban Biomass (IUC) processes. ❍: Input of biomass through the building complex into the process; outputs like solid fraction and digestate. " **... .**": additional blackwater for biogas production.

A key challenge to close the biomass cycle between AD and agriculture is transforming the digestate into fertilizer. Digestates contain high amounts of ammonium (NH4). However, while NH4 can be used as a nitrogen (N) source by plants, high NH4 contents potentially increase N-losses by emission and can inhibit plant growth, especially in hydroponics. Therefore, NH4 has to be oxidized via nitrite (NO2) to nitrate (NO3) (e.g., by the soiling®-process). In particular, for hydroponic-based crop production, fertilizer quality is of high importance [14] as, among others, its buffer capacity is very low (compared to soil). For hydroponics, under optimal conditions, synthetic or inorganic-based fertilizers are commonly applied [15–17]. With the adjustment of the correct dilution ratio and nutrient concentrations of the organic fertilizers, similar or even higher yields compared to a commercial nutrient solution are possible [18,19]. Finally, based on a Life Cycle Assessment (LCA), the reduction of CO2-eq can be calculated [20].

#### **2. Materials and Methods**

To assess the ICU concept regarding energy and food production, the conversion of biowaste to heat and electricity using AD (Section 2.1) and agriculture (Sections 2.2 and 2.3) were simulated. CO2-eq reductions as an indicator for the global warming potential of parts of the ICU process were evaluated by LCA (Section 2.4). In addition, the costs for the implementation of the ICU concept in real buildings were estimated (Section 2.5). Finally, social–cultural aspects of the implementation were reviewed for Germany (Section 3.5.5).

#### *2.1. Modelling In-House Biowaste Degradation for Energy Production*

Biogas production was used as a key process to model biowaste conversion to heat and electricity. The entire process was simulated using pre-implemented building blocks from the software SIMBA#Biogas (version 4.3) [21] (Figure 2) assuming a building with 100 residents.

**Figure 2.** Simba model for in-house degradation of biowaste to heat and electricity. 1. Biomass input of biogenic waste and blackwater with 70 ◦C pre-treatment. 2. The Converter block determines the biomass composition. 3. The Fermenter block determines the biogas and digestate output. 4. The Gas analysis determines the biogas composition. 5. The CHP block converts biogas into energy with a calorific value of 10.5 kJ/kg 6. The Separation block determines thick (solid) and thin (liquid) fractions.

The "Biowaste block" (input) assumes homogenization at 70 ◦C to ensure the necessary sanitization conditions. Therefore, the biomass will be pretreated in a homogenization tank for 2–3 days before fermentation [22]. On average, residents are supposed to produce 33.6 kg of biogenic waste and 720 L of blackwater per day. Since the amount of biowaste fluctuates over the year, an increase of 22% for four months and a feeding of 30.7 kg for five days a week was applied to monitor the dynamic behavior ([23], Supplementary File 1). The "Converter block" takes into account the biomass composition based on literature values (Supplementary File 1, [23,24]). The "Biowaste Fermenter block" represents the conversion of kitchen and hydroponic biowaste into biogas and digestate. The simulations

use the Anaerobic Digestion Model Number 1 da (ADM1da), which is an extension of the Anaerobic Digestion Model Number 1 (ADM1) [25]. The ADM1da comprises 32 differential and algebraic equations. They represent all relevant steps of the biomass degradation and physicochemical process parameters. Operation of the ADP at 55 ◦C was assumed. The "Gas analysis block" defines the biogas composition. The "CHP block" is used to determine the methane (CH4) yield into electrical energy and heat. Here, an electrical efficiency of 38% and a thermal efficiency of 45% was used [26,27]. The electrical efficiency increases with the purity of the AD product gas [26]. The "Separation block" is used to split the digestate into a thin (liquid) and a thick (solid) fraction. The liquid effluent is further processed and nitrified by the soiling®-process into refined digestate. Soiling®-nutrient recycling fertilizer is composed of mineral N plus macro- and micronutrients. Note: only the N amount can be taken into account with Simba; for further calculations, the macro- and micronutrients were neglected.

For the "Blackwater block" (input) a second fermenter is considered as there is currently no approval for a fertilizer containing anthropogenic raw material in Germany according to the so-called Fertilizer Ordinance [28]. Therefore, blackwater fermentation is only considered for energy production but not for fertilizer production. Furthermore, the "Blackwater Fermenter block" is modeled by three different reactor block configurations to identify the most efficient one. The first scenario considers biogas production inside a continuously stirred-tank reactor (CSTR). The second scenario takes into account five CSTR blocks connected in series to simulate a plug flow bioreactor (PFR). The third scenario describes a two-stage reactor (2sR). Here, a small CSTR is used for hydrolysis and fermentation of biowaste, whereas acidogenesis and methanogenesis occur in a bigger second CSTR. The specific parameter settings for all scenarios are in the Supplementary File 2.

The ADP considered in this work was assumed to produce biogas with a calorific value of 10.5 kJ/kg Thus, neither combined cycle power plants (CCPP) nor CHP can be applied on-site because of their lower efficiency. In practice, many operators of small-scale biogas plants favor a satellite CHP over on-site power production. A satellite CHP is supplied with biogas from multiple small-scale biogas plants via a local micro gas grid [27]. The assumption is that a large number of small-scale biogas plants are nearby; an option that could also be applied here.

Suitable for the on-site power production of small-scale biogas plants are fuel cells, microturbines (MT), and engines (igniting beam engine, gas engine). Fuel cells can reach high electrical efficiencies and run quietly. However, they are comparatively large, expensive and their operation requires a high gas purity, which would make additional biogas upgrading necessary. Therefore, the use of a fuel cell was not considered in this study. MT, on the other hand, can operate with a wide range of CH4 concentrations (30–100%) [27]. MTs reach an electrical efficiency of 25–33% with a thermal efficiency of ~49% [29], and operate silently and environmentally compatible [30]. MTs are commonly applied on a 30–550 kW scale [27,31]. Commercial 1 kW scale turbines are under development. However, at the current state, small-scale implementations are inefficient (11% electrical efficiency) at costs of around 6000 € per turbine [30,32]. Alternatively, engines (igniting beam engine, gas engine) can reach a higher electrical efficiency than MTs of 30–40% and a thermal efficiency of ~47%. Compared to MTs, engines are louder, produce noxious side products and require more maintenance. In particular igniting beam engines, which require the addition of pilot oil for combustion, produce noxious side products and soot, which inhibits the efficient use of excess heat [29]. The preferable alternative are gas engines, which operate without pilot oil, but require CH4 concentrations above 45% [29].

In summary, gas engines were considered the most attractive option for on-site production of electrical energy from AD product gas in this study. However, if other small-scale biogas plants were available, a satellite CHP could be the more efficient alternative.

### *2.2. Crop Production Systems*

The amount of ammonia nitrogen (NH4-N), in the liquid effluent and the nitrification rate of the soiling process (European patent, EP3684909) was used to calculate the amounts of plant available N (NH4-N and nitrate nitrogen (NO3-N)) of one year. For cultivation planning (system sizing), the ratio of fresh biomass production to available N was considered. As a model, crop lettuce (*Lactuca sativa* ssp.) was used. A fresh matter N content of 0.18% was assumed (Feller et al., 2019) with a fixed dry matter fraction of 0.048%.

Four possible methods were considered for lettuce cultivation: Scenario 1 and 2 are open-air plant cultivation systems with raised beds or vertical hydroponics, respectively. The residents drive these scenarios on the roof-top with a cultivation period from April to October (vegetation period of Berlin). Both scenarios are complex as they involve the participation of community members (that are outside of the scope of the present simulations). Scenario 3 and 4 are protected cultivations with hydroponic greenhouses or plant factories, respectively. Both have to be operated year-round by trained staff and can be located on the roof-top or in the basement of buildings. Here, a pure bio-technical assessment using deterministic explanatory simulation models was applied.

A numerical simulator for controlled environments and greenhouses was used that is a further development of earlier published greenhouse simulators [33,34]. The simulator was programmed using MATLAB (MathWorks Inc., Natick, MA, USA). It was connected to a replica of commercially available climate controllers, including a setpoint generator that calculated climate setpoints for heating, ventilation, light and CO2 concentration. The simulator was fitted to a standard Venlo-type greenhouse structure or a vertical farming hydroponics-controlled environment (scenario 3 and 4). The simulator's crop-basis is a photosynthesis-driven growth model with microclimate predictions for water and nutrient uptake according to the Penman–Monteith equation [35]. Nutrient uptake was calculated assuming that the diluted nutrients in the irrigation system are optimally taken up by the crop. As such, a perfect pH, electrical conductivity (EC), and a root environment with optimal nutrient solution composition with an optimal availability of all nutrients were assumed. In accordance with Goddek and Körner (2019) [36], all element-specific chemical, biological or physical resistances were set to zero.

For technical layout, supplementary lighting was applied with LED lamps installed either under the roof above the crop with an installed capacity of 80 W m−<sup>2</sup> power and an output of 192 μmol m−<sup>2</sup> s−<sup>1</sup> or at an installed capacity of 110 W m−<sup>2</sup> power and an output of 264 μmol m−<sup>2</sup> s−<sup>1</sup> in scenario 3 or 4, respectively. The light was controlled dynamically with setpoints generated using a daily light integral (DLI) of either 12 mol m−<sup>2</sup> d−<sup>1</sup> or 20 mol m−<sup>2</sup> d−<sup>1</sup> for greenhouse or vertical farming, respectively [37]. In both scenarios, CO2 in the air was set to 700 μmol mol−<sup>1</sup> and supplied according to the demand (max. at 15 g m2 h−1) during lightening when greenhouse vents were closed and at all times in the vertical farming scenario. In the greenhouse scenarios, heat exchange for cooling was calculated with passive roof ventilation while active cooling and dehumidification were used in the vertical farming-controlled environment scenarios (active cooler based on ANSI/AHRI standards 1200). Dehumidification was implemented with a commercially available dehumidification unit of the type ventilated latent heat energy converter. Further model parameters are summarized in supplementary file 3.

The simulator calculated macro- and microclimate in a time-step of 5 min, integrated hourly using controlled actuators (e.g., heating, ventilation, cooling, CO2, light) that were re-adjusted as described by Körner and Van Straten (2008) [34]. The simulations' output included hourly biological and physical variables related to lettuce production, such as microclimate conditions, photosynthesis, yield, and resource consumption (electrical power, heating energy, water, CO2). Input to the simulation program included, among others, physical location (latitude (LAT), longitude (LON)), humidity set point (%), set points for heating and ventilation (◦C), crop planting density (plants m−2) and temperaturesum related harvest time. Input climate data were hourly data sets for Berlin (Germany, LAT 52.5N, LON 13.4) from 2009 to 2018 (Meteoblue; [38]). Calculations were performed for all scenarios for single years of each of the 10-year horizons.

Simulations were performed targeting nutrient and water uptake, yield, and energy demand for heat and lighting for either a greenhouse with a size of 70 m2, or for a vertical farm with a four-layer system (17.5 m2 each, in a room of 30 m2 area and 2.50 m height). As commercially viable climate control in small greenhouses is challenging to maintain, a 500 m2 greenhouse as minimum commercial size was modeled in addition. All simulations were done for year-round production of hydroponically grown lettuce with a fixed planting density of 36 plants m−<sup>2</sup> as in commercial practice (e.g., Brechner et al., 2013 [39]).
