Life cycle assessment was employed to assess the environmental sustainability of the different synergistic options. The functional unit used for the environmental assessment is the annual production of plants from the hypothetical vertical farm (roughly 185 000 plants per year with an edible portion weight of ~5 500 kg) available to consumers. It is assumed that the farm will produce a mix of leafy greens and herbs. Furthermore, to show the impacts related to the plants, the environmental impacts per kg edible portion of the plants are also highlighted. The study is limited to the production and final availability of the product to consumers. As such, the study was conducted using a cradle-to-gate perspective, including all upstream processes in the cultivation, energy use, transportation of materials, and distribution of the plants to regional supermarkets. No waste handling for the final product after consumption was modeled as it was considered outside the scope of the assessment. This includes the pots and growing media in the potted plants. However, waste handling from any waste generated at the farm was included in the assessment, including biowaste, plastics, and water. A depiction of the system boundaries is available in Figure 1.

For the LCA of symbiotic exchanges, this article employs a system expansion methodology (i.e., semi-consequential approach). This is done to allow for the depiction of the benefits of integration with the building and surrounding businesses and motivated by the scope of the study, to explore the implications of symbiotic exchanges with the host building and residual flows from regional systems to improve the environmental performance of the vertical farm, see, e.g., guidelines in Martin et al. (2015) and Brandao et al. (2017). As such, the vertical farm is of primary importance, and the benefits provided to surrounding systems (other than the building with which it is integrated) through the use of residual materials, etc., are excluded. Any intermediate processes for the synergies, e.g., for the use of new materials, are included in the assessment as outlined in the section below for different synergies.

The LCA for the different scenarios was conducted in the OpenLCA software v 1.10.3 (Greendelta, 2018). For this study, the International Reference Life Cycle Data System (ILCD, 2011) midpoint life cycle impact assessment (LCIA) method was employed as it is widely regarded as a consistent and robust method for exploring the implications of products and services in the European market (JRC, 2010). The impact categories included in the main text are Global warming potential (measured in kg CO₂-eq) and denoted (GHG), Water resource depletion (WRD) measured in m³ water, Acidification Potential (AP) measured in molc H+ eq, Freshwater Eutrophication (FE) measured in kg P-eq, and Mineral, fossil, and renewable resource depletion (RD)measured in kg Sb-eq. These particular impact categories were chosen as they provide a review of the regional, global, and resource implications of food systems and their importance in the changes studied, i.e., altering the energy demand and sources, fertilizers, material inputs, and waste handling methods. These are also of particular interest for our comparisons, and similar impact categories are commonly used in a number of studies on agricultural systems and urban farming applications. However, the ILCD method contains 16 indicators, and results from the study for those indicators not highlighted in the main text are provided in Supplementary Material for further information.

All life cycle inventory (LCI) data were obtained from the LCI database Ecoinvent v. 3.6 (Ecoinvent, 2019). Further details are also provided in the subsequent sections for all scenarios assessed. See further details in Supplementary Material for corresponding specific datasets, references for all processes and materials used, in addition to further descriptions on modeling choices.

From the Reference scenario vertical farm, prospective synergies with surrounding businesses and the host building are assessed. The different synergies are illustrated in Figure 2 and further described in the subsequent sections. These are outlined and assessed as separate synergies, while an additional scenario with several synergies integrated is also evaluated.

From the results, it was found that there can be sensitivity to certain datasets and assumptions employed in the assessment. For these datasets and assumptions, the Sensitivity Analysis section below analyzes the influence of these datasets and assumptions. A description of the comparisons is described in these sections, and further details can also be found in Supplementary Material.

As illustrated in Figure 3, for the reference scenario, in nearly all environmental impact categories, energy contributes to the largest share of the environmental impacts. This is primarily a result of the electricity use for the LEDs and the HVAC system employed, contributing to over 50% of the GHG emissions and acidification and freshwater eutrophication impacts. Furthermore, the majority of the resource and water depletion impacts, over 98%, were due to electricity demand.

Roughly 20% of the GHG and acidification impacts are due to the pots and growing media. Transportation, for both inputs and of the products to market, had only a small contribution to the environmental impacts. Together, transportation of inputs and final products to the market accounted for <8% of all GHG emissions in addition to acidification and eutrophication impacts. For freshwater eutrophication, GHG emissions, and acidification, packaging also has a minor contribution to these impact categories, accounting for roughly 8% in all aforementioned impact categories, which was primarily a result of the polyethylene packaging.

The results illustrate that all scenarios, except those with PV arrays, were found to have lower GHG emissions. GHG emissions are reduced largely through integration with residual streams from the local brewery. In this scenario, these two options, i.e., the use of BSG as a growing media and using the CO₂, are illustrated in Figure 4 to reduce the impacts from the Pots and Media and CO₂ processes. The scenario including the use of biofertilizer had a slight reduction in impacts, again primarily due to the reduction of conventional fertilizers required.

As illustrated in Figure 4, the scenarios with the largest reduction in environmental impacts are those of building energy system integration and synergies with a local brewery. For the integrated building energy scenarios, i.e., Bldng Int. HP-WWTP and Bldng Int. DH-W2E, despite no large changes to the overall farm-based emissions, the credits for the use of heat extracted from the LEDs were illustrated to have large potential benefits. For the latter scenario, replacing district heating from waste-to-energy sources had the largest benefit for the system. Finally, when combining the most beneficial scenarios, i.e., in the Opt. Integ, the overall GHG emissions are further reduced.

The scenarios with PV arrays were illustrated to have higher overall GHG emissions compared to all other scenarios. This was partly due to an increase in infrastructure emissions for installing the PVs for the vertical farm, see Figure 4 above. Upon further analysis, this was also found to be due to higher GHG emissions per kWh from photovoltaics compared to the chosen electricity source. For example, while the chosen LCI dataset for electricity from the Swedish electricity mix had GHG emissions of roughly 45 g CO₂-eq per kWh, the resulting GHG emissions from PV sourced electricity was roughly 76 g CO₂-eq per kWh.

However, despite lowering GHG emissions, as Table 2 illustrates, there may be trade-offs with the symbiotic scenarios, where other impact categories assessed may increase. As an example, the local market scenario was illustrated to have lower GHG emissions and acidification impacts compared to the Reference scenario, due in part to a reduction in transportation. However, the resource and water resource depletion impacts were slightly higher, owing to the added infrastructure required for the local pick-up depot, e.g., metals and electronics. The biofertilizer scenario showed reductions in GHG emissions and acidification impacts from reducing conventional fertilizer application, though no notable changes in resource depletion or water resource depletion were illustrated.

While the scenarios with PVs were found to have increased GHG emissions and acidification emissions, in comparison to the Reference scenario, they were found to have lower resource and water resource depletion. This was primarily a result of less demand for electricity from the grid despite added infrastructure for the PVs.

The building integration scenarios had lower environmental impacts in all categories. The scenario Building Integ-W2E was illustrated to substantially reduce GHG emissions, acidification, and resource depletion, owing primarily to the credits provided from the avoidance of district heating from waste incineration. However, the water resource depletion was only reduced slightly considering the low impact of the avoided heat for water resource depletion.

As illustrated in Figure 4, there are considerable reductions possible when the synergies are integrated. Integration of the three scenarios with the largest reductions in environmental impacts, i.e., the BrewInteg, Building Integ-W2E, and Local Market scenarios, was also assessed to analyze the development of more than one synergy and create a symbiotic network between the building, farm, and regional market and businesses.

The previous results highlight the annual impacts. However, as illustrated in Figure 5, the GHG emissions per kg of edible mass from the plants are largely influenced by the synergies. The results show how the carbon footprint associated with the product can be reduced through more circular development, with carbon footprint reductions of over 50% possible.

The results of this study indicate that symbiotic development, more specifically urban symbiosis, through synergies with the host building and surrounding businesses, can lead to large environmental performance improvements. The results are consistent with the findings from several previous studies that also show that synergies with regional systems can improve the environmental performance of vertical farms by moving away from a linear approach, see, e.g., Martin et al. (2019) and Arcas-Pilz et al. (2022). However, this greatly depends on the selected synergies. While several synergies have a more considerable beneficial contribution, others have less influence on the environmental performance. Furthermore, while many of the environmental impact categories are improved, there are some trade-offs between the environmental impact categories.

As the results highlight, synergies to improve the energy demand of the host building through the use of residual heat from the vertical farm showed the most promise in reducing GHG emissions. These results also concur with earlier studies that find environmental performance benefits for urban farms through symbiotic development with energy systems (Sanjuan-Delmás et al., 2018; Sanyé-Mengual et al., 2018; Gentry, 2019). Additionally, several further studies have also suggested similar potential, despite not assessing the environmental performance, see, e.g., Thomaier et al. (2014), Eigenbrod and Gruda (2015), and Chance et al. (2018). However, for some synergies, resource depletion and water resource depletion impacts may also increase due to increased infrastructure, as shown in the Local market and BrewInteg scenario. Such concerns can be an essential aspect of the sustainability of symbiotic networks (Martin, 2020).

Developing a local market for the products also reduced the environmental impacts. In the local market scenario, the reduced transportation and logistics required for product deliveries largely reduced these farm-to-gate emissions from transport. This does partially concur with the discourse used to promote vertical urban farms (Specht et al., 2014). However, this is only applicable for more localized markets for the products with little to no distribution logistics. A number of previous studies have shown that the transportation of foods has a relatively minor impact on the overall impacts of foods (Edwards-Jones et al., 2008; Coley et al., 2009), and specifically vertical farms (Martin and Molin, 2019). Nonetheless, for vertical farms, the close proximity to consumers can also influence the choice of crops grown, which can be selected for specific attributes such as flavor and taste, which is not always possible in conventional varieties found in retail (Bogomolova et al., 2018; Harada and Whitlow, 2020).

While the study only addresses the environmental performance, the benefits also apply to other sustainability aspects, including economic performance and socio-economic development. Urban symbiosis can provide many opportunities for a number of socio-economic benefits, which may be hard to quantify. This can include shared learning, leading to human and social capital development (Velenturf and Jensen, 2016; Chance et al., 2018). For example, a previous study by Chance et al. (2018) studied the material flows and social implications of an urban symbiotic network for hydroponic systems. Furthermore, Marchi et al. (2018) also explore the possibility of symbiotic networks to improve horticulture, suggesting potential economic performance benefits. Nonetheless, few assessments of the environmental performance of the role of urban symbiosis for improving the sustainability of vertical farming systems are available in the literature, pointing to the novel results provided in this study.

As the results suggest, developing circular systems also benefits the environmental performance of the products. The impacts may be halved with more circular development compared to a reference scenario. Similar results for the potential improvements have been outlined in previous studies, see, e.g., Martin et al. (2015) and Martin (2020). However, few other studies have assessed the implications on the environmental impacts of products from vertical farms, and there are few assessments to compare with. In a recent study, however, Orsini et al. (2020) found that greenhouse production methods have reported carbon footprints from 0.2 to 3.2 kg CO₂-eq per kg edible product, while vertical farms may have between 10 and 20 kg CO₂-eq per kg edible for salad. Previous studies of vertical farms from a Swedish context have also shown environmental impacts from roughly 2.6–5.3 kg CO₂-eq per kg edible portion for mixed herbs and leafy greens from vertical farms (Martin et al., 2019; Barge, 2020). As such, the findings of this study are in line with these figures, and symbiotic development has been shown to reduce the impacts, where the reference scenario has a carbon footprint of 3.2 kg CO₂-eq per kg edible while the best case for symbiotic development halves these emissions to roughly 1.5 kg CO₂-eq per kg edible.

While the results of the photovoltaic scenarios are slightly higher than the reference scenario, such results are not unexpected in Sweden, where the energy system has low impacts. Similar findings have been outlined for different regions in Sweden, where the mitigation potential of solar microgrids had higher emissions than the regional electricity system; see e.g., Papageorgiou et al. (2020). This is also due in part to the datasets employed. Ecoinvent data available for photovoltaic energy may be outdated, although the GHG emissions of photovoltaic electricity are similar to those outlined in the aforementioned study. Despite this, and despite the minor contribution the photovoltaics may have to supply the entire demand of the system, it is important also to recognize that this would require a much larger surface. However, this may not negate the emissions associated with the photovoltaic production compared to other sources from regional electricity, i.e., the Swedish electricity mix. Recent studies have also outlined the potential of photovoltaics, suggesting it can cover the energy demand of vertical farms depending on the configuration and sizing of the growing area, see, e.g., Rehman (2021), although further research into the use of photovoltaics in urban environments is needed.

The results point to utilizing biofertilizers as an alternative to conventional fertilizers. Despite only minor reductions in the overall impacts of the vertical farm, the use of biofertilizers largely reduced impacts from the fertilizer use. While this study makes assumptions on their replacement, based primarily on comparing macronutrients, biofertilizer from anaerobic digestion has been tested by several of the regional vertical farms in the Stockholm region. Additionally, in several recent studies, biofertilizers from anaerobic digestion have also been shown to be viable alternatives with added infrastructure for nitrification for these vertical farms, see, e.g., Bergstrand et al. (2020) and Lind et al. (2021). The use of biofertilizers in vertical farms in urban environments can create new markets for regional biogas plants, as the economic viability and use of biofertilizers have been identified as a bottleneck in biogas production systems, see, e.g., Olsson and Fallde (2014) and Martin (2015).

This study outlines several synergies with brewing industry residual products. Replacing conventional soil containing a combination of coir and peat with BSG as a growing media was illustrated to reduce the environmental impacts. These results concur with those suggested in Barrett et al. (2016), who highlight that utilizing industrial by-products can improve the sustainability of soilless plant cultivation systems, and also outlined in an exploratory study by Martin et al. (2019). Furthermore, a number of studies highlight the viability of BSG and other fibrous residual products for different horticultural applications with minimal effect on the plant production (Chong, 2005; Verhagen and Boon, 2008; Christoulaki et al., 2014; Barrett et al., 2016; Chrysargyris et al., 2018). Nonetheless, the use of BSG in this study is conducted only to explore its potential and the viability has not been tested experimentally. Another promising synergy from the brewing industry is the use of carbon dioxide from fermentation for carbon enrichment. This was also illustrated to reduce environmental impacts. However, carbon dioxide extraction from the brewing industry is rare in Sweden. Despite this, ethanol producers in Sweden currently capture CO₂. Carbon capture and its use for carbon enrichment for greenhouses have been conducted in other areas on industrial scales, see, e.g., Short et al. (2014).

While this study has highlighted the possibilities for the use of residual products from the brewing industry, these are not currently explored to their fullest potential. This is especially true for small-scale breweries, which have grown extensively in recent years (SBA, 2020), and may not have handling processes to deliver by-products and wastes in an efficient manner. Many large-scale brewers currently dispose of their by-products, e.g., BSG, for low-value applications, i.e., to produce animal feed or even fuel for boilers (San Martin et al., 2020). Furthermore, there are a large number of urban breweries in Sweden, which lack efficient handling methods for their by-products (Stockholm Brewing, 2021), which ultimately end up in municipal waste streams. Many of these lack efficient handling methods for their by-products, increasing the need to manage resources more sustainably. The scale of the exchange in this study, i.e., the share of BSG used annually for the vertical farm, is relatively minor. As such, effects from changing markets for BSG are considered negligible. Such flows have also been outlined in a number of studies to be ripe for valorization (Santos et al., 2003; Enweremadu et al., 2008; Gmoser et al., 2020; San Martin et al., 2020), although added processing to ensure they do not contain pathogens is critical (Mycoterra Farm, 2014) and further viability and validity of its use as a growing medium is needed. The results of the sensitivity analysis show that this would not significantly change the environmental impacts of the system as the growing media does not largely influence the overall environmental impacts.

Furthermore, while this study does not assess the economic potential of these residual streams, their use in new areas may provide new revenue streams. Still, it may also require added infrastructure and processes to upgrade these to a marketable product. As such, this may incur added costs for the vertical farms, which often struggle with economic viability (Shao et al., 2016) and may impede their development and realization to progress the industry toward synergies as outlined in this study.

The results point to energy demand, primarily electricity, for the different processes as a major influencing input. This stems largely from electricity demand from LED lighting and the HVAC system. These results concur with a number of previous studies, suggesting that energy consumption for artificially maintained climatic and light regimes contribute significantly to the environmental impacts of such systems; see, e.g., Chance et al. (2018), Romeo et al. (2018), and Martin and Molin (2019). This was also apparent in the sensitivity analysis, where the choice of electricity heavily influenced the results.

The results indicate that benefits derived from using waste heat for space heating were substantial. There are two reasons for this. First, virtually all electricity used is converted to heat (through LED waste heat, reflection, and plant evapotranspiration), which is eventually extracted through the AC unit and absorbed by the refrigerant. The majority of this heat can be used for space heating due to adequate temperature levels. As the COP of the AC unit was considered to be between 3 and 4, this leads to a higher quantity in kilowatt terms of heat theoretically available for building integration (189 630 kWh) compared with the total electricity input (130 240 kWh for LEDs and 44 640 kWh for the AC). In practice, this value is not achieved as space heating requirements are seasonal and heat transfer losses occur (106 720 kWh of heat effectively replaced). Second, in the negative impact scenarios, renewable electricity indirectly replaces heat from fossil fuels, which means the inputs are much less carbon-intensive than the replaced products. These indicative values need to be carefully considered. In this study, roughly 15% heat losses were assumed. However, this needs to be validated in practice as the heat loss through walls might be higher depending upon the building envelope. On the other hand, the replacement of hot water heating was not considered as additional heat pumps would be required to lift temperatures to adequate levels (roughly 60°C). This would increase the symbiosis potential as waste heat in summer could be partially utilized. In locations where electricity is generated by burning fossil fuels, the GHG emissions for energy would be proportionally higher (compared with ~50 g CO₂/kWh assumed in this study). Similar assertions have also been highlighted for the use of heat from the vertical farm by Barge (2020). Similar findings from energy-integrated rooftops greenhouses have also been found to save energy and heating requirements for buildings (Bass and Baskaran, 2001; Sanjuan-Delmás et al., 2018; Muñoz Liesa et al., 2020).

While the study addresses building energy integration, further synergies have also been outlined in the literature to integrate vertical farms in buildings. In a recent study, Shao et al. (2021) also highlight the potential of employing vertical farms to reduce CO₂ concentration and reduce ventilation energy consumption in commercial buildings. Furthermore, additional studies have also highlighted the use of optical fibers in buildings to reduce the energy consumption for LEDs; see e.g., Asiabanpour et al. (2018). Thus, further research could be focused on the potential of further integration with buildings to reduce energy and resource consumption for the farm and buildings.

Once again, the results of this study point to the benefits of including vertical farming in buildings and the potential of urban symbiosis to link with local energy systems and surrounding businesses. Such information can provide evidence to support the integration of urban agriculture and vertical farms in existing buildings and in new developments. Integrating vertical farms with their buildings can be used to progress toward decarbonizing buildings, which is needed to reduce the environmental implications of the building sector; see stipulations in Castleton et al. (2010) and Mata et al. (2022). This study also highlights the potential of utilizing residual spaces in urban areas and buildings. While a number of studies have also reviewed the use of vacant lots, few study the use of indoor vacant spaces; see, e.g., Grewal and Grewal (2012). As such, the benefits of employing residual spaces for producing food through vertical farms can add to the rental returns for real estate owners, improve the quality of the spaces, and provide new business opportunities for food production and markets (Martin and Bustamante, 2021; Shao et al., 2021), providing insights to residential building owners on the benefits of employing residual space for urban food provisioning. As the results of this study show, these systems can be multi-functional by providing energy savings for the building, offering locally produced food, utilizing and valorizing residual streams, and increasing awareness of food production in urban areas. Similar assertions have been highlighted in previous studies (Ferreira et al., 2018; Rufí-Salís et al., 2020; Arcas-Pilz et al., 2022). Such development should be encouraged in urban areas to promote urban agriculture, food systems and promote circular economy development in urban areas (Specht et al., 2014; Petit-Boix and Leipold, 2018; Arcas-Pilz et al., 2022).