Bacterial Protein for Food and Feed Generated Via Renewable Energy

Global Food Security 22 (2019) 25–32

Contents lists available at ScienceDirect

Global Food Security

journal homepage: www.elsevier.com/locate/gfs

Bacterial protein for food and feed generated via renewable energy and T direct air capture of CO2: Can it reduce land and water use? ∗ Jani Sillmana, , Lauri Nygrena, Helena Kahiluotoa, Vesa Ruuskanena, Anu Tamminenb, Cyril Bajamundib, Marja Nappab, Mikko Wuokkob, Tuomo Lindha, Pasi Vainikkac, Juha-Pekka Pitkänenc, Jero Aholaa a LUT School of Energy Systems, Lappeenranta University of Technology, 53850, Lappeenranta, Finland b VTT Technical Research Centre of Finland, P.O. Box 1000, FI-02044, VTT, Finland c Solar Foods Ltd, 00320, Helsinki, Finland

ARTICLE INFO ABSTRACT

Keywords: The global food demand is projected to signiﬁcantly increase. To maintain global food security in the future, Carbon dioxide protein production needs to become more eﬃcient regarding the use of limited land and water resources. Direct air capture Protein-rich biomass can be produced via direct air capture of CO2 with the help of H2-oxidizing bacteria and Hydrogen-oxidizing bacteria renewable electricity in a closed, climate-independent system. This quantitative literature review conservatively Microbial protein (MP) estimated the direct land and water use of bacterial protein production relying on secondary data for the Microbial biomass components of the technology and for the reference protein sources. A several times higher potential protein Bio-electrochemical system Land use yield per land area can be achieved by this technology with approximately one-tenth of the water use compared Carbon capture and utilization to that required for soybean production. Environmental sustainability

1. Introduction Rosegrant et al., 2009). Consequently, there is a need for novel approaches to realise environmentally sustainable, climate-resilient pro- The growing human population and the projected decline of agri- tein production that is less dependent on land and water and has the cultural yields due to climate change challenge global food security ability to adapt to climate change (FAO, 2017; Campbell et al., 2016). (Godfray et al., 2010; Porter et al., 2014). To provide the increasing Proteins are one of the main components in food, and their avail- population with a sufficient amount of protein, approximately 110% ability, quality and environmental impacts are good indicators by more crop protein will have to be produced in 2050 than in 2005, ac- which to measure food security (Coles et al., 2016). The lack of access cording to the estimate of Tilman et al. (2011).), while the availability to high-quality protein due to unequal distribution in the world causes of additional arable land and water for agricultural use is limited (FAO, malnutrition, especially in developing countries. For example, millions 2017). Currently, agriculture (excluding fibre, rubber and tobacco) of children subsist on poor sources of essential amino acids (Semba, accounts for 43% of the World’s ice-free and desert-free land area 2016). In the future, the access to high-quality protein will become even (Poore and Nemecek, 2018) and 70% of freshwater use (AQUASTAT, more challenging due to the increase in protein demand along with 2016). There are also competing claims for land and freshwater use population growth and increase in wealth under climate change. To from urbanization, the forest industry, biofuel production and carbon produce high-quality proteins to maintain food security with minimal stock conservation (e.g., Gibbs et al., 2007) via forests. In addition, environmental impact, microbial proteins (MPs), especially bacteria- agriculture, forestry and other land uses cause approximately one based MPs, are seen as promising alternatives compared to traditional quarter of anthropogenic greenhouse gas emissions and most nutrient animal- and plant-based proteins (Litchfield, 1989; Matassa et al., emissions (FAO, 2017; Smith et al., 2014). The production of animal 2016). For example, in animal feed, MPs from gram-negative soil bac- proteins accounts for the major part of the use of these resources, even terium Curpiavidus necator have been found to be comparable with though their share of the global protein consumption is approximately those from basic protein sources, such as soymeal and fishmeal, and one third only (Campbell et al., 2017; Rojas-Downing et al., 2017; include all essential amino acids required for growth (Volova and

∗ Corresponding author. E-mail address: jani.sillman@lut.ﬁ (J. Sillman). https://doi.org/10.1016/j.gfs.2019.09.007 Received 6 February 2019; Received in revised form 21 September 2019; Accepted 24 September 2019 2211-9124/ © 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). J. Sillman, et al. Global Food Security 22 (2019) 25–32

Barashkov, 2010; Schøyen, 2007). direct air capture (DAC), bioreactors with an in situ water electrolyser There are several beneﬁcial characteristics that bacterial proteins and separation of the bacterial biomass from the cultivation medium. have over other protein sources. Bacteria-based MPs have high protein Separating CO2 from ambient air to produce MP by binding CO2 to the contents ranging from 50% to 83% of the dry biomass, and the bacteria biomass without increasing atmospheric CO2 makes the application have high growth rates relative to other protein sources (Anupama and less-dependent on location compared solutions with point sources of

Ravindra, 2000; Jørgensen, 2011; Litchfield, 1989; Matassa et al., 2016; CO2 (e.g., Matassa et al., 2015; Pikaar et al., 2018; Yu, 2014). The Srividya et al., 2014; Upadhaya et al., 2016.). The bacterial conversion greatest potential of the approach is to decouple protein production efficiency from substrates to protein-rich biomass is high compared to from the use of agricultural land and minimal water use, in comparison that of other MP sources, such as yeasts and fungi, and to that of tra- with traditional protein sources, whereas the climate impact crucially ditional protein sources. MP production is usually based on inexpensive depends on the used energy source. Here we have assumed a renewable carbon substrates, including waste, which are available in large quan- energy source, which increases the land use, and thus provide a con- tities (Matassa et al., 2015, 2016; Upadhaya et al., 2016; Øverland, servative estimate of the potential to reduce land and water use. 2012; Steinkraus, 1986). In addition, the production process for MPs The aim of this study is to quantify the potential reduction in land can be designed as a closed system highly independent of seasonal and water use from bacterial MP production via the use of direct air changes, making efficient nutrient use without runoff to the sur- capture of CO2, a water electrolyser and renewables in comparison with rounding environment possible, and the system does not require her- soybean production and other MPs available on markets having studied bicides or pesticides (Israelidis, 1987; Pikaar et al., 2017). MP pro- requirement of land and water. Soybeans are a widely used plant-based duction applications can have higher solar-to-biomass efficiencies than protein source with a high protein content and high yields, making those of crops, but efficiencies among different applications vary soybean protein a well-suited plant-based protein for comparison (Day, greatly. Regarding the biomass of crops that can be harvested annually, 2013; Mekonnen and Hoekstra, 2012; FAOSTAT, 2018). The land and the yearly average solar-to-biomass efficiency does not typically exceed freshwater use were assessed based on the energy and raw material 1%. During the growing season, the energy use efficiency may reach requirements of the system components reported in the literature. 3.5% for C3 plants, such as most staple cereals and vegetables, and 4.3% for C4 plants, such as maize (corn), sorghum and sugarcane, and even the theoretical efficiency is limited to 6% (Blankenship et al., 2. Materials and methods 2011; Zhu et al., 2008). Using the cyanobacterium Arthrospira plan- tensis, an efficiency of 2% has been reached (Tredici and Zittelli, 1998), 2.1. System boundary and methods while significantly higher solar-to-biomass efficiencies of up to 10% have been achieved using H2-oxidizing bacteria (Liu et al., 2016). The studied system consists of the determined energy and material The possibility of producing MP using H2-oxidizing bacteria has consumption of the components of technologies of MP production. The long been known (Schlegel, 1965). H2-oxidizing bacteria are able to performance data from the best electricity-to-biomass efficiency of a utilize H2 as an energy source, with the help of CO2 and O2, for mi- lab-scale bioreactor found in literature was used for energy consump- crobial growth (Yu, 2014). H2 and O2 can be produced by electrolysing tion calculations. For other processes, the electricity consumptions of water inside a bioreactor (Schlegel, 1965), and CO2 can be separated commercially available products and estimates from literature were from the atmospheric air by adsorbents (Leung et al., 2014). Although used. The energy requirements for producing nutrients for microbial

CO2 separation and water electrolysis are energy-intensive processes, growth were neglected. The theoretical energy and material require- there exists a growing interest in opportunities for MP production, ments of the application were determined by calculating the stoichio- because renewable energy sources can be used. (e.g., Matassa et al., metry of the substance requirements for biomass growth and by com- 2016; Liu et al., 2016). bining experimental data from system components found in the The growing capacity of wind and solar power increases utility literature. The production process was assumed to be a closed system. fluctuation, leading to occasional oversupplying of power. In particular, Thermal energy was assumed to be taken from waste heat sources. The the availability of solar energy is increasing due to the rapid decline in technologies selected consisted of direct electrolysis in the bioreactor, its production costs. Thus, flexible applications capable of using elec- post-processing to dry the biomass, and DAC to provide the source of trical loads for demand response are needed in the future (Lehner, CO2, and the bacterial species used was C. necator. The simplistic 2014; Park et al., 2016). The electricity oversupply can be used to scheme of the application can be seen in Fig. 1. produce MP from H2-oxidizing bacteria. C. necator survives night cycles The conservative land and water requirements of MP production when the required substances cannot be provided to the bioreactor were compared to the requirements of soybean and FeedKind® pro- through renewables and continues to grow again during daytime (Liu duction using a quantitative literature review. FeedKind® is a bacterial et al., 2016). Food-insecure regions with high amounts of non-arable MP used for feed purposes in markets (Cumberlege et al., 2016). The land area could be used for producing solar or wind energy, which land use comparison between MP and soybean production is based on could be directed to MP production upon oversupply or when required. direct land use values in the U.S, which is one of the major soybean For instance, the International Renewable Energy Agency (2014) has producers in the world. The land use was estimated according to the estimated for Africa annual theoretical energy production potentials of direct land occupation of protein production per year similarly to, e.g., 660 PWh and 460 PWh for photovoltaics (PVs) and wind energies, re- Nijdam et al. (2012) and Nguyen et al. (2010). The production facilities spectively (Hermann et al., 2014). The energy production potentials for MP have only a minor impact on land use, and they can be located exclude the areas used for agriculture and exceed the global annual on non-arable land or even under the ground; thus, their impact are electricity demand of 25 570 TWh in 2017 (IEA, 2018). excluded from the analysis. MP production is energy-intensive and uses Despite the growing interest in MP production, the environmental renewables as energy sources. The land area required by energy pro- sustainability of MP production has been quantitatively evaluated only duction and the related infrastructure was included in the calculation. in a few studies (e.g., Cumberlege et al., 2016; Knudsen et al., 2016; The land occupation of energy production required for MP production Pikaar et al., 2018). In a comparison of different MP production tech- was calculated by multiplying the energy consumption per produced nologies, production via water electrolysis requires the least land and biomass and land occupation per energy generation and time: causes the lowest greenhouse gas emissions (Pikaar et al., 2018). However, to our knowledge, no studies have evaluated the land and water needs of MP production by H2-oxidizing bacteria, taking into account the energy requirements of state-of-the-art technology using

26 J. Sillman, et al. Global Food Security 22 (2019) 25–32

Fig. 1. The scheme of MP production by the selected technology. U refers to voltage and I refers to electric current.

Land occupationkWh explosion (C&EN, 2016). Therefore, controllable gas solubility is es- Land useMP = Energy generationkWh × time sential. Problems with the mass transfer of gases and safety in the cultivation of H -oxidizing bacteria can be avoided, at least to some ⎛ Energy consumptionMP ⎞ 2 × Σ ⎜⎟ extent, by performing the electrolysis directly in the bioreactor. The ⎝ Massbiomass ⎠ (1) application in the studied case employs internal water electrolysis, and

For the water requirement of soybean production, values including thus, only CO2 is supplied from outside the bioreactor. green and blue water requirements for soybean production was used for comparison. Green water measures the volume of rainwater consumed, 2.2.2. Direct air capture of CO2 and blue water measures the ground and surface water consumed CO2 is required as the carbon source for the growth of bacteria. CO2 during the growth period (Mekonnen and Hoekstra, 2014). The water supplied to the reactor can be taken from many sources. Flue gases from consumption of MP was calculated based on the stoichiometry of the combustion process (Leung et al., 2014) and by-products of fermenta- production process: tion or air are a few examples of sources of CO2. The studied case application uses DAC, which is an emerging technology for the collection ⎛Water consumptionMP ⎞ Water consumption Σ ⎜ ⎟ MP = of CO2 directly from the atmosphere. Ambient air passes through either ⎝ Massbiomass ⎠ (2) a solid or a liquid medium in which CO2 molecules are retained. When energy is introduced into the medium, concentrated CO2 is released, 2.2. The description of the case technology enabling it to be collected, stored, or used. Absorption refers to the case in which a liquid medium is used in the capture process, and adsorption 2.2.1. Hydrogen-oxidizing bacteria cultivation in bioreactors is defined as the process in which a solid medium is used in the capture Microorganisms can be cultivated in bioreactors, where they can be process (APS, 2011; Bajamundi et al., 2019; Jones, 2011; Jones, 2015; grown on a wide variety of substrates. H2-oxidizing bacteria, also called Zeman, 2014.). Most DAC systems are based on an adsorption/deso- Knallgas bacteria, are able to use H2 as an electron donor and O2 as an rption process (e.g., Carbon Engineering, 2016; Climeworks, 2016; fi electron acceptor to fixCO2 to build up their biomass. In addition, some Global Thermostat, 2016; In nitree, 2016; The Center for Negative nutrients are required to increase the biomass in the cultivation Carbon Emissions, 2016). The scheme of an adsorption-based direct air medium. There exist different species of H2-oxidizing bacteria, such as capture device is illustrated in Fig. 2. C. necator (also called Ralstonia eutropha and Alcaligenes eutrophus), In principle, a pure CO2 source could be omitted, and atmospheric

Rhodococcus opacus, and Hydrogenobacter thermophiles.H2-oxidizing air could be used as the carbon source for bacteria. According to Liu bacteria are mostly aerobic and facultative chemolithoautotrophs using et al. (2016), the electricity-to-biomass efficiency decreased from 54% Calving cycle for carbon fixation. They can derive their energy through chemosynthesis or grow heterotrophically using organic substrates (Aragno, 1998; Kuenen, 1999.; Lee, 2015.). Microorganisms can be cultivated in batch or continuous bioreactors. There are several different types of bioreactors, such as stirred- tank, bubble column, trickle-bed, membrane-based, and U-loop reactors (Munasinghe and Khanal, 2010; Petersen et al., 2016). In addition to the various technological approaches for splitting water using electricity (Section 2.2.3), it is also possible to produce H2 and O2 externally or internally. External water electrolysis uses a separate unit process outside the bioreactor, and internal electrolysis is performed directly in the cultivation medium in the bioreactor. Supplying the required gases from outside the reactor is a more conventional method of growing H2-oxidizing bacteria, but the challenge of this approach is the low mass transfer of H2 and O2 to the aqueous solution (Yu, 2014). In addition, H2 and O2 can produce flammable mixtures in the reactor headspace, which may be ignited, for instance, by measurement instruments in the bioreactor and cause an Fig. 2. Adsorption-based direct air capture device (Fasihi et al., 2019).

27 J. Sillman, et al. Global Food Security 22 (2019) 25–32

Fig. 3. Approximated main material and energy flows required to produce 1 kg of bacterial biomass. to 20% when atmospheric air was supplied to the headspace of a reactor polyhydroxyalkanoates can also be separated from the protein fraction instead of 100% CO2, even though the partial pressure of CO2 was re- using solvent extraction. To estimate the energy needed for post-produced by 2500 times. cessing, centrifugation and evaporation were chosen as example processes. 2.2.3. Electrolysis of water Currently, alkaline water electrolysis is the most mature and widely 2.3. Material and energy requirements of the case technology exploited water electrolysis technology for water splitting (Bertuccioli et al., 2014). However, proton exchange membrane (PEM) technology The amount of required input substances can be approximated by is currently challenging alkaline technology by producing higher cur- the stoichiometry of the bacteria. Liu et al. (2016) reported the fol- rent densities, thereby offering a more compact design and higher ef- lowing stoichiometry for C. necator biomass formation when the bac- ficiency (Carmo et al., 2013). Efficiencies based on the higher heating teria were cultivated using direct electrolysis in the bioreactor: value of H2 for the present PEM electrolysers have been reported to be 65–82%, whereas the alkaline electrolyser efficiencies have been re- CO2 ++ 0.24 NH3 0.525 H2 O → CH1.77 O 0.49 N 0.24 + 1.02 O2 ported to be significantly lower, at 59–70% (Zeng and Zhang, 2010). According to this stoichiometry, 0.38 kg of H2O, 1.76 kg of CO2, and Solid oxide electrolysers are still in the R&D stage (Ursúa, 2012). 0.16 kg of NH3 are required to produce 1 kg of biomass. In addition, Direct electrolysis in a bioreactor was invented in the 1960s 1.31 kg of O2 is produced per kilogram of biomass. Usable protein (Schlegel, 1965), but owing to its high voltage requirement, its con- content of 50–65% of dry biomass are typical estimates for C. Nectator ffi ffi version e ciency was poor. An electricity-to-biomass e ciency of 54% (Anupama and Ravindra, 2000; Chee et al., 2019; Volova and has been achieved by applying direct electrolysis in a bioreactor, which Barashkov, 2010). Assuming that the protein content of the biomass is ffi would approximately correspond to a 10% solar-to-biomass e ciency, 60%, the water consumption is approximately 0.6 L kg -1. The ac- ffi protein assuming 18% e ciency for photovoltaics (Liu et al., 2016). Similar tual water consumption is higher because there is always some residual ffi kind of electricity-to-biomass e ciency has been achieved by using moisture after solid–liquid separation that is then removed via drying. external water electrolysis (Yu, 2014). The challenge in direct electro- In addition to H ,O, and CO , bacteria also require a mineral salt fi 2 2 2 lysis is to nd the balance between a suitable environment for bacteria medium containing essential nutrients, such as nitrogen, phosphorus, ffi in the cultivation medium, a reasonable e ciency of electrolysis, and a and sulphur (Rittman and McCarty, 2001). ffi su cient H2 formation rate (Liu et al., 2016). Therefore, the design of The estimated material and energy flows required to produce 1 kg of electrode structures, anode and cathode materials and the optimization bacterial biomass are illustrated in Fig. 3. The energy input for pro- ffi of growth conditions play key roles in improving the conversion e - ducing bacterial biomass consists of the energy required for the elec- ciency. In addition, when using internal water electrolysis to grow trolysis of water, CO2 capture and post-processing of the biomass. An bacterial biomass for food or feed purposes, it must be ensured that no electricity-to-biomass efficiency of 54% for C. necator bacteria has been toxic components are dissolved from the electrodes or from salts used in achieved using direct electrolysis, assuming the free energy gain from the cultivation medium. A typical electrode material in bioelec- -1 CO2 to biomass to be ΔrG° = 479 kJ mol (Liu et al., 2016), which trochemical systems has been rare and expensive platinum, but new corresponds to an electric energy consumption of the electrolysis of materials have also been found, such as cobalt phosphate CoPi for the 9.86 kWh per 1 kg of produced biomass. The energy consumption of anode and NiMoZn, stainless steel (Liu et al., 2016; Torella et al., 2015). CO2 capture depends on the technology applied. Climeworks an- nounced that their CO2 capture device has a thermal energy demand of -1 2.2.4. Post-processing of bacterial biomass 1.8–2.5 kWh kg CO2 and an electric energy demand of 0.35–0.45 kWh -1 Separation of the bacterial biomass from the cultivation medium is kg CO2 (Climeworks, 2016). According to the stoichiometry, the challenging, as the bacterial cells are small (1–2 μm), their density is average energy consumption of CO2 capture would thus be 0.71 kWh close to that of water (1.003 kg l-1), and the cell dry weight densities at electric energy and 3.78 kWh thermal energy per 1 kg of produced the end of cultivation are relatively low (10–30 g l-1)(Lee, 2015). The biomass. At the end of the cultivation process, the biomass concentra- separation is typically carried out by centrifugation (Litchfield, 1989), tion in the medium can be assumed to be 2.5% (Lee, 2015). Davis et al. but microfiltration via a membrane can also be used (Frenander and (2016) estimated an average value of 1.35 kWh m-3 for centrifugation Jönsson, 1996). After solid–liquid separation, other post-treatment to obtain a biomass concentration of 20%. Therefore, to harvest 1 kg of processes, such as thermal disruption, acid treatment, flash drying, and biomass via centrifugation, 40 kg of reactor solution must be treated, -1 grinding, can be implemented depending on the end-product properties which leads to 0.05 kWh kgbiomass . If evaporation is used to achieve a desired (Brocklebank, 1987). Various cell components such as lipids or biomass concentration of 90%, 3.89 kg of water has to be evaporated.

28 J. Sillman, et al. Global Food Security 22 (2019) 25–32

With the heat requirement for water vaporization of 2260 kJ kg-1 and Table 1 the estimated thermal efficiency of 84%, 2.91 kWh of heat is needed for Water and land use of the main components in the case technology. evaporation per 1 kg of biomass. As the electricity consumption for DAC Bioreactor Post-processes Total water evaporation is 87 kWh m-3, 0.34 kWh is needed per 1 kg of bio- 2 -1 mass. Land use [m kgprot ] MPSolar 0.01–0.018 0.16–0.23 0.0065–0.0091 0.18–0.26 MP 0–0.0026 0–0.033 0–0.0013 0–0.04 2.4. Land use of renewable energy Wind -1 Water use [L kgprot ] The direct land area impact of renewables consists of land directly Input - 65.82 65.18 - occupied by solar arrays and wind turbine pads, access roads, substa- Output - −65.18 −65 - Net use 0.63 0.18 0.82 tions, service buildings and other energy infrastructure physically oc- - cupying land area. According to the national renewable energy laboratory, the generation-weighted direct land area impact of solar PV Table 2 – 2 -1 -1 varies within the range of 0.010 0.014 m kWh y in U.S. In the case Water and land use of soybean and MPs with other production characteristics. of wind energies, the capacity-based direct land area impact is within Sources: the range of 0–0.002 m2 kWh-1 y-1 in U.S, when the average capacity Land and water requirement Soybean MP MP MP factor of 0.367 is used (Denholm et al., 2009; Ong et al., 2013.). Solar Wind Methane -1 Water consumption [L kgprot ] 2.67–6.67 0.82 0.82 0.5–1.45 2 -1 -1 2.5. Land and water use of soybean production Land use [m kgprot y ] 6.40–15.86 0.18–0.26 0–0.04 0 Other characteristics [a] [a] [a] To evaluate the yields of soybean production in U.S, the yields of the Sterile production environment No Yes Yes Yes Pesticides and herbicides use Yes No[b-c] No[b-c] No[b-c] last ten years were taken from National Agricultural Statistic Service of Fertilizer run-offs Yes No[b-c] No[b-c] No[b-c] United States Department of Agriculture. The top maximum and Controllable production process No Yes[a-c] Yes[a-c] Yes[a-c] minimum values were not used for the evaluation as they were counted Seasonal dependence Yes No[b-c] No[b-c] No[b-c] as exceptions. The range of the yields were 2.4–5.95 m2 kg-1y-1 in the Need for arable land Yes No No No years 2008–2018 in U.S (USDA, 2019). The soybean protein content a Lee (2015). varies between 35% and 40% (Day, 2013). If the average protein b Israelidis (1987). content is used, the yield of soybean protein varies between c Pikaar et al.(2017). 6.40–15.86 m2 kg-1y-1. In the case of water use, the range of green-blue – -1 water use of soybeans is approximately 2.67 6.67 L kgprotein (Konar of soybean production and have similar kind of impacts than MP via et al., 2013; Mekonnen and Hoekstra, 2014), when using the average methane production (Table 2). The direct land occupation of bacterial protein content. protein have ten times lower land area requirement per a year than soybean production in the case of the highest land occupation value of 2.6. Land and water use of the another bacterial MP solar energy generation. The results show that land occupation of MP production is minimal, especially, when wind energy is relied upon. ® Cumberlege et al. (2016) evaluated FeedKind commercial products The MP production consumes approximately one tenth of the water ® in terms of water use and land occupation. FeedKind is produced from compared to soybean production. The characteristics of MP via me- bacterium capable of methanotrophy, thus methane is required for the thane and the studied MP production are assumed to be the same as bacterial growth. The land occupation values are in range of both processes can be designed to be closed systems. – 2 -1 0 0.052 m kgprotein and the water consumption in the range of The efficiency of electricity-to-biomass of 54% reported in this – -1 10 29 L kgprotein depending on how the product is post-processed. The paper is high compared with other values found in the literature (e.g., high water consumption is due to energy used and vegetable oil in the Torella et al., 2015). For instance, if the electricity requirement of the fi nalized product. The land occupation included only the land re- electrolysis would increase by 10% per produced amount of biomass, quirement of vegetable oil production. For impacts of MP via methane the increase in the demand for electricity would be approximately to be comparable with the studied MP, only the impacts of protein 0.99 kWh, which would decrease productivity, land use per produced production were included in the evaluation. When the impacts of en- protein, by approximately 8%. The protein content of the final MP ergy and vegetable oil are omitted, the results are approximately product is also affecting the productivity. When calculating land use by – -1 0.5 1.45 L kgprotein for water consumption and zero for land occupa- using 50% and 65% protein contents of MP, the productivity decreases tion. by 16% or increases by 10%. Other process sensitivities to the total electricity demand are minor compared with that of electrolysis 3. Results (Table 1).

The sum of the electrical energy consumed during the electrolysis and the average value of the range given for CO2 capture together with 4. Discussion and conclusions the estimated requirement for biomass drying would be 10.96 kWh -1 kgbiomass . The consumed energy per kilogram of produced protein 4.1. Land and water use of MP production and other characteristics would be 18.26 kWh (Fig. 3). The bioreactor within-situ electrolyser uses most of the resources in the production process of bacterial protein The main ﬁnding of this study was that MP protein production using

(Table 1). The land use of main components of the case technology CO2 capture from air and renewables needs many times less land area originates in the calculation by multiplying the range of direct land and water than does soybean production. Although, the studied case occupation of electrical power generation (Section 2.4) with electricity technology uses energy intensive DAC devise to provide CO2 for the demand of the main components in the technology (Equation (1); process, the results of this study are similar kinds with those of studies Section 2.1). The water consumption of MP production originates in estimating the land and water use of other MP production methods calculation according to Equation (2) (Section 2.1). (e.g., Pikaar et al., 2018; Cumberlege et al., 2016). The electricity-to- The land use and water consumption of bacterial MP via studied biomass eﬃciency of electrolysis majorly aﬀected the energy demand technology are clearly lower than the land use and water consumption and, thus, should play a key role in the application design.

29 J. Sillman, et al. Global Food Security 22 (2019) 25–32

Pikaar et al. (2018) simulated avoided land use, nitrogen flows and cell walls of bacteria are indigestible for non-ruminant animals. The cell greenhouse gas emissions in agriculture, if bacterial MPs from different walls can be destroyed and the nucleic acid concentration can be re- pathways replaces traditional protein sources. The results showed sig- duced by mechanical, chemical, or enzymatic means (Lee, 2015; nificant reduction potentials. The pathways were MP via water elec- Nasseri et al., 2011). Bacterial biomass has already been commercia- trolysis, sugarcane-to-MP, cellulose gasification-to-MP and biomethane- lized as animal feed, but the nutritional and toxicological testing re- to-MP. MP via water electrolysis had the most significant reduction quired for its approval for food purposes may take several years and potential from the above mentioned MPs. The MP via water electrolysis cost millions of euros (Lee, 2015; Dominique et al., 2016). The use of uses similar hydrogen oxidizing bacterium species to produce biomass bacteria as food seems also to be unacceptable to many people, even as the studied case technology, but the production technology was though, for instance, lactic acid bacteria play a key role in many fer- different, e.g., bioreactor, external water electrolysis, and point sources mented foods (Caplice and Fitzgerald, 1999). Some products, such as of CO2 were different. By using DAC instead of point sources to provide yoghurts, are even marketed based on the presence of lactic acid bac- CO2 opens new possibilities for mitigation, adaptation and location teria. Examples of successful commercialization of unconventional mi- purposes for MP production as CO2 is taken from air and the production crobial protein sources, such as the mycoprotein Quorn, exist, in- facility is not bound to specific locations. dicating that there are commercial possibilities for microbial food MP production via the proposed method can be a competitive al- products. ternative to conventional protein production also because of other beneficial characteristics and potential environmental advantages re- 4.3. Reliability of the analysis and future research needs lated to land and water use. Because the application can be designed as a closed system, climate change has no impact on the growing condi- Currently, soybeans are the most important source of plan-based tions and it is possible to recycle nutrients until they are completely protein for feed purposes, thus, we used only them as a plant-based used, resulting in high nutrient utilization efficiency without run-offs reference protein. Other commonly used alternatives for fodder are (Lee, 2015). Another beneficial aspect is that MP production does not mainly side streams from other products (e.g., sunflower and oilseed require pesticides and herbicides, the use of which might cause biodi- rape), the sources are mainly used for crop rotation, specific sources are versity losses. In addition, it is possible to use wastewaters as a nutrient not used so widely for feed than soybeans are (e.g., legumes) or the sources, which would reduce the fresh water and nutrient consumption sources have low protein contents (e.g., maize and barley). In addition, of the application. However, the safety and process sterility of waste- the content and quality of the studied protein is comparable to soymeal. water use has to be ensured (Matassa et al., 2015). Owing to the The study was based on values given in the literature and by tech- abundance and availability of the required raw materials to cultivate nology manufacturers. The energy requirements of the unit processes MP this method of protein production is highly independent of place excluded from the study have only a negligible effect on the total en- and climate and, therefore could be exploited in food production even ergy consumption. Similarly, the material flows of production process where there is no arable land. Due to the rapid growth rate of microbes, and the impact of the MP production facilities excluded from the ana- the method can be fast implemented when food or feed availability lysis on direct land area requirement, are minimal. The potential con- becomes challenging. servative land and water use estimation of the state-of-the-art technology has many disadvantages compared to soybeans. The nutrient use 4.2. Usability of MP as food and feed is more efficient than that of soybeans due to closed production system, the land use of MP production includes the land use of renewables and Until now, the main use of bacterial protein has been a feed for the production does not require pesticides or herbicides. In addition, animals, an application that has been widely studied and implemented the land occupation of energy production depends on the location, used in an industrial scale, e.g., Methylophilus methylotrophus using me- technology and energy sources available, which can have a major effect thanol as an energy source (Litchfield, 1989). Furthermore, animals are on the requirements of land area. For example, a significant amount of typically inefficient in converting plant-based proteins into animal PV solar power plants can be installed on the roofs of buildings, and proteins (Steinkraus, 1986), and thus, a large amount of plant-based wind farms can be located offshore resulting in a minimal direct land proteins are consumed as feed resources. The provision of MP as feed area requirement. would save plant-based proteins, such as soybean protein, for human The proposed protein production method is highly energy-intensive consumption or land and water for other uses. However, as Pikaar et al. compared with conventional production practices in agriculture (2018) showed in their simulations, MPs can replace only a certain (Woods et al., 2010). Approximately 90% of the energy demand of the amount of nutrition value apart from protein in feed, but with other application comes from the electrolysis process, and thus, the energy sustainable practices, MPs could be a part of the solution in transition requirement is highly dependent on the electrolysis efficiency and the towards sustainable food systems. mass transfer of H2 and O2. The used electricity-to-biomass efficiency Despite the superior growth rate, low resource needs and high was based on a lab-scale bioreactor; thus, scaling up the process is es- protein content, in general, bacterial biomass still has some dis- sential when larger production volumes are preferred. Although it has advantages which limit its wider use. These limiting factors are related been shown that scale-up of designs for producing H2-oxidizing bac- to the downstream processes after cultivation, regulation, and safety teria-based biomass is not a problem (Reed et al., 2015), piloting a challenges. For instance, it must be ensured that the selected bacteria facility with a larger production capacity is essential to validate the are not pathogenic to humans or animals and do not contain toxic results of this study. compounds. Another challenge is to provide a sterile production en- With current electricity prices, only the cost of energy required to vironment because contaminating microorganisms typically grow well produce MP using the application is higher than the price of soybeans in the culture medium (Lee, 2015; Litchfield, 1979, 1989; Ritala et al., (USDA, 2018), even if the capital and other operational costs are not 2017). taken into account. In the future, the situation may change, as elec- MP for example from Methylophilus methylotrophus or C. necator have trobioreactors could be used to provide demand-response services for amino acid profiles and a nutritive value comparable with fishmeal and the grid, electricity prices could go down, or the price of soybeans could soybean meal but have high nucleic acid concentrations. The high nu- go up. The energy-related costs can also be reduced by using different cleic acid content of the bacterial biomass is not a problem for rumi- technologies than the ones presented in this method. Thus, techno- nants, and monogastric animals tolerate it (Litchfield et al., 1979; economic research is preferred before commercialising the proposed Srividya et al., 2014). For humans, the consumption of more than 2 g/d application. of nucleic acids may lead to kidney stone formation and gout, and the H2-oxidizing bacteria survive night cycles when no electricity is

30 J. Sillman, et al. Global Food Security 22 (2019) 25–32 used to provide substances for biomass growth and continue their (6), 459–479. https://doi.org/10.1016/S0734-9750(00)00045-8. growth during day cycles when renewable electricity is available. Thus, Bajamundi, C.J.E., Koponen, J., Ruuskanen, V., Elfving, J., Kosonen, A., Kauppinen, J., Ahola, J., 2019. Capturing CO2 from air: technical performance and process control the production of bacterial biomass could be balanced according to the improvement. J.CO2 .Util. 30, 232–239. https://doi.org/10.1016/j.jcou.2019.02. varying production and load of the grid, which is known as a demand 002. response, and can lead to reduced electricity costs and/or incentive Bertuccioli, L., Chan, A., Hart, D., Lehner, F., Madden, B., Standen, E., 2014. Study On Development of Water Electrolysis in the EU; Final Report in Fuel Cells and Hydrogen payments (Albadi and El-Saadany, 2008). However, further research is Joint Undertaking (FCH JU). (Brussels). needed to conclude the suitability of the application for use as a de- Blankenship, R.E., Tiede, D.M., Barber, J., Brudvig, G.W., Fleming, G., et al., 2011. mand-response tool. Especially, the time required to achieve the op- Comparing photosynthetic and photovoltaic efficiencies and recognizing the poten- – timum productivity of MP is a topic of interest. tial for improvement. Science 332 (6031), 805 809. https://doi.org/10.1126/ science.1200165. Because the application binds atmospheric CO2, investigating the Brocklebank, M.P., 1987. Large scale separation and isolation of proteins. In: King, R.D., greenhouse gas emissions of the entire production process would also Cheetham, P.S.J. (Eds.), Food Biotechnology – 1. Elsevier Applied Science Publishing – be of interest. To study the different environmental impacts over the LTD, London, pp. 139 192. ’ Campbell, B.M., Vermeulen, S.J., Aggarwall, P.K., Cornell-Dollof, C., Girvetz, E., product s lifetime, a life-cycle approach can be used (Finkbeiner, 2011). Loboguerrero, A.M., Ramirez-Villegas, J., Rosenstock, T., Sebastian, L., Thornton, A pilot-scale application can help in estimating the actual energy and P.K., Wollenberg, E., 2016. Reducing risks to food security from climate change. – material flows of the production processes and in estimating reasonable Global food security 11, 34 43. https://doi.org/10.1016/j.gfs.2016.06.002. Campbell, B.M., Beare, D.J., Bennett, E.M., Hall-Spencer, J.M., Ingram, J.S.I., Jaramillo, production scales. A cradle-to-grave impact assessment takes into ac- F., Ortiz, R., Ramankutty, N., Sayer, J.A., Shindell, D., 2017. Agriculture production count the supply chains for product distribution, materials for pro- as a major driver of the Earth system exceeding planetary boundaries. Ecol. Soc. 22 duction facilities and other material flows. Another research need (4). https://doi.org/10.5751/ES-09595-220408. Caplice, E., Fitzgerald, G.F., 1999. Food fermentations: role of microorganisms in food would be to estimate the available amount of renewable energy for MP production and preservation. Int. J. Food Microbiol. 50 (1-2), 131–149. https://doi. production and to which extent MPs could replace plant-based feeds org/10.1016/S0168-1605(99)00082-3. while considering, for example, the entire nutrient value and post- Carbon Engineering, 2016. Industrial-scale Capture of CO2 from Ambient Air Website. http://carbonengineering.com/, Accessed date: 1 August 2018. processing requirements of the alternative feed protein sources. Carmo, M., Fritz, D.L., Mergel, J., Stolten, D., 2013. A comprehensive review on PEM water electrolysis. Int. J. Hydrogen Energy 38 (12), 4901–4934. https://doi.org/10. 4.4. Conclusions 1016/j.ijhydene.2013.01.151. Center for Negative Carbon Emissions, 2016. https://engineering.asu.edu/cnce/, Accessed date: 1 August 2018. The land and water requirements of H2-oxidizing bacteria-based MP Chee, J.Y., Lakshmanan, M., Jeepery, I.F., Hairudin, N.H.M., Sudesh, K., 2019. The po- via the carbon capture and renewable energy had not been previously tential application of cupriavidus necator as polyhydroxy-alkonates producer and fi studied. We have shown that this technology exhibits a minor direct single cell protein: a review on scienti c, cultural and religious perspective. Appl.Food.Biotechnol. 6 (1). https://doi.org/10.22037/fb.v%vi%i.22234. land occupation and water consumption compared to traditional pro- Climeworks, 2016. Facts & Figures Website. www.climeworks.com/facts-figures.html, tein sources such as soybeans. Currently, MPs are not widely used Accessed date: 1 August 2018. protein sources; thus, there is much progress that can be made by using Coles, G.D., Wratten, S.D., Porter, J.R., 2016. Food and nutritional security requires adequate protein as well as energy, delivered from whole-year crop production. PeerJ the proposed MP production technology to improve food security and 4 (e2100). https://doi.org/10.7717/peerj.2100. sustainability of food or feed chains. Therefore, the technology should Cumberlege, T., Blenkinsopp, T., Clark, J., 2016. Assessment of environmental impact of be considered as a promising tool for sustainable agricultural devel- FeedKind protein carbon trust. https://www.carbontrust.com/media/672719/ calysta-feedkind.pdf. opment. However, before the commercialization of the studied tech- Davis, R., Kinchin, C., Markham, J., Tan, E.C.D., Laurens, L.M.L., Sexton, D., Knorr, D., nology, ensuring the safety as food and piloting with larger production Schoen, P., Lukas, J., 2016. Process Design and Economics for the Production of Algal capacities including a techno-economic and a more comprehensive Biomass: Algal Biomass Production in Open Pond Systems and Processing through Dewatering for Downstream Conversion. National Renewable Energy Laboratory, environmental assessment is essential. Denver. Day, L., 2013. Proteins from land plants – potential resource for human nutrition and food Declaration of competing interest security. Trends Food Sci. Technol. 32 (1), 25–42. https://doi.org/10.1016/j.tifs. 2013.05.005. Denholm, P., Hand, M., Jackson, M., Ong, S., 2009. Land-Use Requirements of Modern Authors declares that there are no conflicts of interest. Wind Power Plants in the United States. U.S. National Renewable Energy Laboratory, Denver. Acknowledgements Dominique, T., Bresson, J.-L., Burlingame, B., Dean, T., Fairweather-Tait, S., Heinonen, M., Hirsch-Ernst, K., et al., 2016. Guidance on the preparation and presentation of an application for authorization of a novel food in the context of Regulation (EU) 2015/ The authors would like to thank the Academy of Finland for pro- 2283. EFSA Journal 14 (11). https://doi.org/10.2903/j.efsa.2016.4594. viding funding for the study Microbial Oil and Proteins from Air by FAO, 2017. The Future of Food and Agriculture – Trends and Challenges. FAO, Rome. FAOSTAT, 2018. Crops Website. www.fao.org/faostat/en/#data/QC, Accessed date: 15 Electricity-Driven Microbes; funding decision 295883 for VTT and August 2018.

295866 for LUT. Fasihi, M., Efimova, O., Breyer, C., 2019. Techno-economic assesment of CO2 direct air capture plants. J. Clean. Prod. 224, 957–980. https://doi.org/10.1016/j.jclepro. 2019.03.086. References Finkbeiner, M., 2011. Towards Life Cycle Sustainability Management. Springer, Netherlands. fl fi Aragno, M., 1998. The aerobic, hydrogen-oxidizing (knallgas) bacteria. In: Burlage, R.S., Frenander, U., Jönsson, A.S., 1996. Cell harvesting by cross- ow micro ltration using a – Atlas, R., Stahl, D., Geesey, G., Sayler, G. (Eds.), Techniques in Microbial Ecology. shear-enhanced module. Biotechnol. Bioeng. 52 (3), 397 403. https://doi.org/10. Oxford University Press, New York, pp. 3917–3933. 1002/(SICI)1097-0290(19961105)52:3<397::AID-BIT5>3.0.CO;2-F. Gibbs, H.K., Brown, S., Niles, J.O., Foley, J., 2007. Monitoring and estimating tropical APS, 2011. Direct air capture of CO2 with chemicals: a technology assessment for the APS panel on public affairs. In: Hotinski, R. (Ed.), APS Physics. AQUASTAT. Water uses carbon forest stocks: making REDD a reality. Environ. Res. Lett. 2. https://doi.org/ Website. www.fao.org/nr/water/aquastat/water_use/index.stm, Accessed date: 1 10.1088/1748-9326/2/4/045023. August 2018. Global Thermostat, 2016. A Carbon Negative Solution Website. http://globalthermostat. AQUASTAT, 2016. Water uses Website. , Accessed date: 1 August 2018. com/, Accessed date: 1 August 2018. C&EN, 2016. Spark from Pressure Gauge Caused University of Hawaii Explosion, Fire Godfray, H.C.J., Beddington, J.R., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F., Pretty, Department Says'. Chemical & Engineering News Website. http://cen.acs.org/ J., Robinson, S., Thomas, S.M., Toulmin, C., 2010. Food security: the challenge of – articles/94/web/2016/04/Spark-pressure-gauge-caused-University.html, Accessed feeding 9 billion people. Science 327, 812 818. https://doi.org/10.1126/science. date: 1 August 2018. 1185383. Kuenen, G., 1999. Oxidation of inorganic compounds by chemolithotrophs. In: Lengeler, Hermann, S., Miketa, A., Fichaux, N., 2014. Estimating the Renewable Energy Potential in J.W. (Ed.), Biology of the Prokaryotes, pp. 187–233. Africa, IRENA-KTH Working Paper. International Renewable Energy Agency, Abu Albadi, M.H., El-Saadany, E.F., 2008. A summary of demand response in electricity Dhabi. markets. Electr. Power Syst. Res. 7 (11), 1989–1996. https://doi.org/10.1016/j.epsr. IEA, 2018. Global Energy & CO2 Status Report 2017. International Energy Agency. fi fi 2008.04.002. In nitree, 2016. Technology website. www.in nitreellc.com/technology/, Accessed date: Anupama, Ravindra, P., 2000. Value-added food: single cell protein. Biotechnol. Adv. 18 1 August 2018.

31 J. Sillman, et al. Global Food Security 22 (2019) 25–32

Israelidis, C.J., 1987. Nutrition – single cell protein, twenty years later. In: First B.I.O: Ebi, K.L., Estrada, Y.O., Genova, R.C., Girma, B., Kissel, E.S., Levy, A.N., MacCracken, International Conference 1987. S., Mastrandrea, P.R., White, L.L. (Eds.), Climate Change 2014: Impacts, Adaptation,

Jones, C.W., 2011. CO2 capture from dilute gases as a component of modern global and Vulnerability. Part A: Global And Sectoral Aspects. Contribution Of Working carbon management. Annu. Rev.Chem.Biomol. Eng.2011 2, 31–52. https://doi.org/ Group II to the Fifth Assessment Report Of the Intergovernmental Panel On Climate 10.1146/annurev-chembioeng-061010-114252. Change. Cambridge University Press, Cambridge, United Kingdom and New York,

Jones, C.W., 2015. Amine-based solid sorbents for CO2 separations – from sorbent design NY, USA, pp. 485–533. to novel separation processes. In: International Conference on Carbon Dioxide Reed, J., Geller, J., McDaniel, R., Kiverdi, )., 2015. CO2 Conversion By Knallgas Utilization XIII, (Singapore). Microorganisms. California Energy Commission Publication number: CEC-500-2017- Jørgensen, J.B., 2011. Exploiting microorganisms for animal feed production. In: The 7th 005. Malaysia International Agro-Bio Business Conference: what Next Agriculture II, Ritala, A., Häkkinen, S.T., Toivari, M., Wiebe, G., M.G., 2017. Single cell protein—state- (Malaysia). of-the-art, industrial landscape and patents 2001–2016. Front. Microbiol. 8. https:// Knudsen, M.T., Dorca-Preda, T., Hermansen, J.E., 2016. Life cycle assessment of alter- doi.org/10.3389/fmicb.2017.02009. native protein sources: constraints and potentials. In: Andersen, K., Tybirk, K. (Eds.), Rittman, B.E., McCarty, P.L., 2001. Environmental Biotechnology: Principles and Nordic Alternative Protein Potentials: Mapping of Regional Bioeconomy Applications. McGraw-Hill, New York, U.S.A. Opportunities. Nordic Council of Ministers, Denmark, pp. 101–118. Rojas-Downing, M.M., Nejadhashemi, A.P., Harrigan, T., Woznicki, 2017. Climate change Konar, M., Hussein, Z., Hanasaki, N., Mauzerall, D.L., 2013. Virtual water trade flows and and livestock: impacts, adaptation, and mitigation. Climate Risk Management 16, savings under climate change. Hydrol. Earth Syst. Sci. Discuss. 10 (1), 67–101. 145–163. https://doi.org/10.1016/j.crm.2017.02.001. https://doi.org/10.5194/hessd-10-67-2013. Rosegrant, M.W., Fernandez, M., Sinha, A., 2009. Looking into the future for agriculture Lee, B.H., 2015. Fundamentals of Food Biotechnology, second ed.. John Wiley & Sons, and AKST. In: McIntyre, B.D., Herren, H.R., Wakhungu, J., Watson, R.T. (Eds.), Chichester, UK. International Assessment of Agricultural Knowledge. Science and Technology for Lehner, M., Ticler, R., Steinmüller, H., Koppe, M., 2014. Power-to-Gas: Technology and Development (IAASTD), Agriculture at a crossroads, Island Press, Washington, DC, Business Models. Springer International Publishing, Lonkon, UK. pp. 307–376 2009. Leung, D.Y.C., Caramanna, G., Maroto-Valer, M.M., 2014. An overview of current status Schlegel, H.G., Lafferty, R., 1965. Growth of ‘knallgas’ bacteria (hydrogenomonas) using of carbon dioxide capture and storage technologies. Renew. Sustain. Energy Rev. 39, direct electrolysis of the culture medium. Nature 205, 308–309. https://doi.org/10. 426–443. https://doi.org/10.1016/j.rser.2014.07.093. 1038/205308b0. Litchfield, J., 1989. Single-cell proteins. In: Marx, J.L. (Ed.), A Revolution in Schøyen, H.F., 2007. Bacterial Protein Produced from Natural Gas as a Protein Source in Biotechnology. Press Syndicate of the University of Cambridge, New York, pp. 71–81. Feed for Monogastric Animal. Dh.D. Dissertation. Norwegian University of life sci- Litchfield, J., 1979. Production of single-cell protein for use in food or feed. In: Peppler, ences (Ås). H.J., Perlman, D. (Eds.), Microbial Technology Volume 1: Microbial Processes. Semba, R.D., 2016. The rise and fall of protein malnutrition in global health. Ann. Nutr. Academic Press, New York, pp. 93–155. Metab. 69 (2), 79–88. https://doi.org/10.1159/000449175. Liu, C., Colón, B.C., Ziesack, M., Silver, P.A., Nocera, D.G., 2016. Water splitting-bio- Smith, P., Bustamante, M., Ahammad, H., Clark, H., Dong, H., Elsiddig, E.A., Haberl, H.,

synthetic system with CO2 reduction efficiencies exceeding photosynthesis. Science Harper, R., House, J., Jafari, M., Masera, O., Mbow, C., Ravindranath, N.H., Rice, 352 (6290), 1210–1213. https://doi.org/10.1126/science.aaf5039. C.W., Robledo Abad, C., Romanovskaya, A., Sperling, F., Tubiello, F., 2014. Matassa, S., Boon, N., Verstraete, W., 2015. Resource recovery from used water: the Agriculture, forestry and other land use (AFOLU). In: Edenhofer, O., Pichs-Madruga, manufacturing abilities of hydrogen-oxidizing bacteria. Water Res. 68 (1), 467–478. R., Sokona, Y., Farahani, E., Kadner, S., Seyboth, K., Adler, A., Baum, I., Brunner, S., https://doi.org/10.1016/j.watres.2014.10.028. Eickemeier, P., Kriemann, B., Savolainen, J., Schlömer, S., von Stechow, C., Zwickel, Matassa, S., Boon, N., Pikaar, I., Verstraete, W., 2016. Microbial protein: future sustain- T., Minx, J.C. (Eds.), Climate Change 2014: Mitigation Of Climate Change. able food supply route with low environmental footprint. Microb. Biotechnol. 9 (5), Contribution Of Working Group III to the Fifth Assessment Report Of the 568–575. https://doi.org/10.1111/1751-7915.12369. Intergovernmental Panel On Climate Change. Cambridge University Press, Mekonnen, M.M., Hoekstra, Y., 2012. A global assessment of the water footprint of farm Cambridge, United Kingdom and New York, NY, USA. animal products. Ecosystems 15 (3), 401–415. https://doi.org/10.1007/s10021-011- Srividya, A.R.,, Vishnuvarthan, V.J.,, Murugappan, M.,, Dahake, P.G., 2014. Signle cell 9517-8. protein- a review. Int. J. Pharm. Res. Scholars 2 (4) 2277–7873. Mekonnen, M.M., Hoekstra, Y., 2014. Water footprint benchmarks for crop production: a Steinkraus, K.H., 1986. Microbial biomass protein grown on edible substrates: the in- first global assessment. Ecol. Indicat. 46, 214–223. https://doi.org/10.1016/j. digenous fermented foods. In: Moo-Young, M., Gregory, K.F. (Eds.), Microbial ecolind.2014.06.013. Biomass Proteins. Elsevier Applied Science, Essex, pp. 33–43. Munasinghe, P.C., Khanal, K.K., 2010. Biomass-derived syngas fermentation into biofuels: Tilman, D., Balzer, C., Hill, J., Befort, B.L., 2011. Global food demand and the sustainable opportunities and challenges. Bioresour. Technol. 101 (13), 5013–5022. https://doi. intensification of agriculture. Proc. Natl. Acad. Sci. 108 (50), 20260–20264. https:// org/10.1016/j.biortech.2009.12.098. doi.org/10.1073/pnas.1116437108. Nasseri, A.T., Rasoul-Amini, S., Morowvat, M.H., Ghasemi, Y., 2011. Single cell protein: Torella, P.J., Gagliardi, C.J., Chen, J.S., et al., 2015. Efficient solar-to-fuels production production and process. Am. J. Food Technol. 6 (2), 103–116. https://doi.org/10. from a hybrid microbial-water-splitting catalyst system. Proc. Natl. Acad. Sci. U. S. A 3923/ajft.2011.103.116. 112 (8), 2337–2342. https://doi.org/10.1073/pnas.1424872112. Nguyen, T.L.T., Hermansen, J.E., Mogensen, L., 2010. Environmental consequences of Tredici, M.R., Zittelli, G.C., 1998. Efficiency of sunlight utilization: tubular versus flat different beef production systems in EU. J. Clean. Prod. 18 (8), 756–766. https://doi. photobioreactors. Biotechnol. Bioeng. 57 (2), 187–197. AID-BIT7 > 3.0.CO;2-J. org/10.1016/j.jclepro.2009.12.023. https://doi.org/10.1002/(SICI)1097-0290(19980120)57:2<187. Nijdam, D., Rood, T., Westhoek, H., 2012. The price of protein: review of land use and Upadhaya, S., Tiwari, S., Arora, N.K., Singh, D.P., 2016. Microbial protein: a valuable carbon footprints from life cycle assessments of animal food products and their component for future food security. In: Singh, J.K., Singh, D.P. (Eds.), Microbes and substitutes. Food Policy 37 (6), 760–770. https://doi.org/10.1016/j.foodpol.2012. Environmental Management. Studium Press, New Delhi, pp. 259–278. 08.002. Ursúa, A., Gandía, L.M., Sanchis, P., 2012. Hydrogen production from water electrolysis: Ong, S., Campbell, C., Denholm, P., Margolis, R., Heath, G., 2013. Land-Use Requirements current status and future trends. Proc. IEEE 100 (2), 410–426. https://doi.org/10. for Solar Power Plants in the United States. U.S. National Renewable Energy 1109/JPROC.2011.2156750. Laboratory, Denver. USDA, 2018. USDA Soybean Baseline, 2010-19. United States Department of Agriculturae Øverland, M., 2012. Potential of microbial ingredients as protein sources for farmed Economic Research Service Website. www.ers.usda.gov/topics/crops/soybeans-oil- animals and fish. In: International Symposium on European Protein Position, 2012. crops/market-outlook/usda-soybean-baseline-2010-19/, Accessed date: 20 August Park, C., Bigler, F., Korba, P., 2016. Power-to-Gas consept for integration of increaced 2018. photovoltaic generation into distribution. Eneregy Procedia 99, 411–417. https://doi. USDA, 2019. Quick stats. United states department of agriculture: national agricultural org/10.1016/j.egypro.2016.10.131. statistical service. https://quickstats.nass.usda.gov/, Accessed date: 23 August 2019. Petersen, L.A.H., Villadsen, J., Jørgensen, J.B., Gernaey, K.V., 2016. Mixing and mass Volova, T.G., Barashkov, V.A., 2010. Characteristics of proteins synthesized by hydrogen- transfer in a pilot scale U-loop bioreactor. Biotechnol. Bioeng. 114 (2), 344–354. oxidizing microorganisms. Appl. Biochem. Microbiol. 46 (6), 574–579 PMID https://doi.org/10.1002/bit.26084. 21261071. Pikaar, I., Matassa, S., Rabaey, K., Bodirsky, B.L., Popp, A., Herrero, M., Verstraete, W., Woods, J., Williams, A., Hughes, J.K., Black, M., Murphy, R., 2010. Energy and the food 2017. Microbes and the next nitrogen revolution. Environ. Sci. Technol. 51 (13), system. Philos. Trans. R. Soc. Biol. Sci. 365 (1554), 2991–3006. https://doi.org/10. 7297–7303. https://doi.org/10.1021/acs.est.7b00916. 1098/rstb.2010.0172. Pikaar, I., Matassa, S., Bodirsky, B.L., Weindl, I., Humpenöder, F., Rabaey, K., Boon, N., Yu, J., 2014. Bio-based products from solar energy and carbon dioxide. Trends Bruschi, M., Yuan, Z., van Zanten, H., Herrero, M., Verstraete, W., Popp, A., 2018. Biotechnol. 32 (1), 5–10. https://doi.org/10.1016/j.tibtech.2013.11.001.

Decoupling livestock from land use through industrial feed production pathways. Zeman, F., 2014. Reducing the cost of Ca-based direct air capture of CO2. Environ. Sci. Environ. Sci. Technol. 52 (13), 7351–7359. https://doi.org/10.1021/acs.est. Technol. 48 (19), 11730–11735. https://doi.org/10.1021/es502887y. 8b00216. Zeng, K., Zhang, D., 2010. Recent progress in alkaline water electrolysis for hydrogen Poore, J., Nemecek, T., 2018. Reducing food's environmental impacts through producers production and applications. Prog. Energy Combust. Sci. 36 (3), 307–326. https:// and consumers. Science 360 (6392), 987–992. https://doi.org/10.1126/science. doi.org/10.1016/j.pecs.2009.11.002. aaq0216. Zhu, X., Long, S.P., Ort, D.R., 2008. What is the maximum eﬃciency with which pho- Porter, J.R., Xie, L., Challinor, A.J., Cochrane, K., Howden, S.M., Iqbal, M.M., Lobell, D.B., tosynthesis can convert solar energy into biomass? Curr. Opin. Biotechnol. 19 (2), Travasso, M.I., 2014. Food security and food production systems. In: Field, C.B., 153–159. https://doi.org/10.1016/j.copbio.2008.02.004. Barros, V.R., Dokken, D.J., Mach, K.J., Mastrandrea, M.D., Bilir, T.E., Chatterjee, M.,