Environmental Life Cycle Assessment of a Novel Cultivated Meat Burger Patty in the United States

Abstract

The meat industry has a substantial negative impact on the environment. As a result, this industry is in a period of change to alternative meat to mitigate the environmental issues caused by conventional meat production. Cultivated meat is highlighted as an alternative to conventional meat-based diets. SCiFi Foods has developed such a novel cultivated meat burger as a potential successor to the currently available burgers. Based on the process information provided by SCiFi Foods, this work performed a life cycle analysis on the novel cultivated meat burger and compared it with alternatives. The life cycle impacts of the novel burger were evaluated using four indicators: greenhouse gas emissions (CML-IA); energy use (cumulative energy demand); land use (ReCiPe midpoint); and water use (ReCiPe midpoint). The study found that the cultivated meat burger generated 87% less greenhouse gas emissions, required 39% less energy, had 90% less influence on land use, and 96% less water use than the comparable beef patty. The effects of uncertainty in the data, sensitivity to major assumptions, and the effect of the manufacturing plant location were analyzed. The studied burger was also found to have a life cycle environmental impact that is comparable with plant-based commercialized burgers that are currently available.

1. Introduction

The agriculture and food industry contributes more than 25% of all global greenhouse gas emissions, inflicts considerable negative impacts on freshwater, and uses about half of the arable land on earth [1,2,3,4]. This raises the following questions: (1) How can the food industry develop innovations toward sustainable development? (2) What could replace our current food industry?

Within the food production sector, the meat industry plays an outsized role in pollution and is the largest consumer of fossil fuels, land, and water resources [5]. Resource use, pollution, and the use of farm animals contribute to these severe consequences. Specifically, the livestock production sector accounts for 18% of anthropogenic greenhouse gas emissions to the atmosphere worldwide, since this sector contributes 37% of anthropogenic methane [5]. A total of 30% of the land surface is used for livestock production, and 33 and 26% of arable land are used to grow livestock feed crops and grazing, respectively. Additionally, livestock feed crop production highly impacts the freshwater quality due to nitrogen and phosphorus emissions, pesticide contamination of water, and acid rain from ammonia emissions [6,7,8]. Thus, the livestock and the meat industries contribute significantly to environmental damage. The livestock and meat industries contribute significantly to the current biodiversity loss crisis by converting land use and greenhouse gas emissions: over three-quarters of all deforested lands are used for livestock pasture and feed crop production [9,10]. These environmental damages including reduced biodiversity and increased greenhouse gas emissions destroy the link between all organisms on Earth, and ultimately, they threaten the sustainability of the food industry [11,12].

So far, most of the existing technologies and industries have focused on economic indicators and productivity, which has caused tremendous environmental damage due to insufficient consideration of the environmental and social aspects [13,14]. To reduce the chance of burden shifting along a value chain, the life cycle analysis (LCA) approach was developed in the 1970s and has been extended to become even more comprehensive [15,16]. The food industries (e.g., Coca-Cola Company and Nestle) have applied the LCA approach to assess and develop food products and packaging [17,18]. Some studies [19,20] have discussed that dietary shifts could powerfully contribute to increasing the sustainability of food systems, along with reducing meat consumption.

However, it is predicted that the increase in meat consumption will continue due to simultaneous increases in the population and per capita meat consumption with greater prosperity [21]. Thus, various substitutes for meat are being studied. Plant-based meat made from soy and pea protein has been discussed and utilized for several decades as an alternative. Furthermore, plant-based proteins represented by tofu, tempeh, and seitan, textured vegetable proteins, have been invented and used as vegan versions of meat-based dishes [22]. Additionally, the EU granted insect-based products such as protein bars, gummies, and insect flour for human consumption in May 2021. To mimic the more similar meat specifications (e.g., texture, flavor, and nutrition) compared to plant-based meat and insect-based protein, the cultivated meat produced by in vitro cell cultures of animal cells using tissue engineering is considered as one of the promising alternatives. Listrat et al. [23] analyzed the components of skeletal muscle and suggested research directions for the production of cultivated meat to achieve production efficiency and improve the meat/flesh quality of conventional products (e.g., cattle, pigs, and poultry).

Many studies have performed a LCA of meat analogues for replacing conventional meat [24,25,26,27]. Mattick et al. [24] performed a life cycle analysis of in vitro biomass cultivation for cultivated meat. The produced cultivated meat had 25, 4, and 6% of global warming potential, eutrophication potential, and land use than that of beef, respectively. However, the industrial energy use for cultivated meat had a 35% higher value than beef. Smetana et al. [25] performed a life cycle analysis of proteins made from Hermetia illucens. In their study, fresh insect biomass was evaluated that was almost twice as sustainable as chicken meat. They evaluated that insect-based products and vitro biomass could contributed to the sustainable food industry as protein sources. Their target product was a different form from conventional meat; thus, more studies are needed on the food and nutritional aspects (taste and texture) for commercialization. Since 2016, the Impossible (is produced by Impossible Foods Inc., Redwood City, CA, USA) [26] and Beyond (is produced by Beyond Meat Inc., El Segundo, CA, USA) [27] burgers have been launched commercially. However, there is still a lack of texture or a nutritionally perfect imitation of conventional meat, so they are continuously being improved [28]. Impossible and Beyond Burgers are currently being commercialized, but their life cycle environmental impact has not been well discussed in their reports [26,27] and studies [29]. To produce an alternative that is more similar to conventionally-produced meat, one approach is to replace the plant-based material from animal-source nutrients including serum. SCiFi Foods, California, USA has developed a novel cultivated meat burger by combining the benefits of cultivated bovine muscle and fat and plant-based material (e.g., soy protein isolate and coconut oil), which contributed to a more similar meat analog. In collaboration with SCiFi Foods, this work aims to assess the life cycle impacts (LCIs) of their novel blended cultivated meat burger with a breakdown of the contribution from various activities in the product life cycle using LCA methodology.

The rest of this paper is organized as follows. Section 2 introduces the research methods including system boundaries, functional unit, and impact assessment. Notably, the data collection procedure and replacement strategies are described to overcome the lack of the life cycle inventory for the novel product, which can contribute to follow-up studies. The LCI for the novel burger is investigated and analyzed by comparing the U.S. beef patty in Section 3. We discuss the potential improvement in GHG emissions for the novel burger and competitiveness with alternative patties, which showed that the selection of a factory site has a significant effect on the GHG emissions. In addition, the impact oof the technological innovations is discussed. The work is summarized and concluded in the final section.

2. Methods and Materials

2.1. Goal and Scope

This section describes the goal and scope of the study (research methodology, system boundaries, functional unit, and impact assessment) to evaluate the life cycle impacts of the developed burger, produced using plant-based materials and cultivated cells including fat and muscle cells developed by SCiFi Foods.

2.1.1. System Boundaries

The life cycle of products consists of five stages, from raw material extraction to end-of-life through material processing, product manufacturing, and product use [30]. Depending on the research goal, several types of scopes (cradle-to-grave, cradle-to-gate, and gate-to-gate) can be considered: cradle-to-grave includes impacts from raw material extraction to disposal and reuse; cradle-to-gate accounts for the effects from raw material to production except for the end-of-life; and gate-to-gate is only from material processing to the product manufacturing. Notably, this study aimed to compare the LCI of a novel burger with the conventional burger and to identify the significant contributor or sector. Additionally, the novel burger’s raw materials differ from conventional meat. However, the gate-to-grave part of all burgers was virtually identical. Therefore, we chose the cradle-to-gate system boundary since it allows for a fair comparison between the selected burgers. The study evaluated the life cycle of the novel burger from the raw material extraction for fat and muscle cell cultivation to burger production using cultivated meat, which includes the cultivated fat and muscle cell from SCiFi Foods, as shown in Figure 1. In this study, we ignored the distribution and cold storage steps because they are common to competing burgers. Thus, the system boundary in this study was from cradle-to-gate.

2.1.2. Functional Unit

The functional unit is critical to ensure a reliable and fair comparison of alternatives and to analyze the ecological effects on a common basis [31]. The functional unit (FU) of this study was 1 kg of novel burger, which includes cultivated meat cultivated at SCiFi Foods, plant-based materials, etc. We compared the life cycle impacts of 1 kg of patties including beef, Beyond burger, and Impossible Burger, which have been commercialized in the market.

2.1.3. Impact Assessment

This study considered four impact categories: GHG emissions, energy demand, land use, and water use. These indicators are widely used and accepted for assessing the environmental impacts of livestock and agricultural products [32]. According to the life cycle impact assessment (LCIA) methods, the available impact categories are different: The CML-IA baseline is composed of global warming, human toxicity, and freshwater aquatic ecotoxicity; ReCiPe midpoint includes climate change, ionizing radiation, and freshwater eutrophication [33]. The target product in this study was a novel burger that includes various compounds such as cultivated fat and muscle cells and burger ingredients. Thus, we considered all impacts of renewable and non-renewable sources to cover all energy use for the life cycle (e.g., energy production from wind and solar, and fertilizer). Life cycle impacts in each category were converted to a standard unit: GHG emissions are represented in units of kg of CO_2-eq; energy demand in MJ; land use in m²a_eq; water use in m³_eq. These conversions require characterization factors such as 28 kg CO_2-eq being equivalent to 1 kg CH₄. According to the impact categories, various LCIA methods and normalization and weighting sets are utilized such as CML, ReCiPe, and IPCC. To compare the environmental impacts of novel burger with similar products, this study selected the suitable impact categories: the CML-IA baseline with World 2000 allocation and cumulative energy demand methods were used to assess the GHG emissions and energy use, respectively. The ReCiPe 2016 midpoint (E) method with World (2010) (E) allocation was used to evaluate the land and water use.

2.2. Comparison Studies

Many studies have performed the LCA of meat analogues to replace conventional meats (e.g., beef and chicken) and address the meat industry’s environmental impacts [24,26,27,34,35]. This study reviewed three papers and two reports and summarized the characteristics of their studies.

Table 1. Comparative analysis results of the life cycle assessment method in similar studies.

	Tuomisto and Mattos (2011) [34]	Mattick et al. (2015) [24]	Smetana et al. (2015) [35]	Beyond Meat (2017) [27]	Impossible Foods (2019) [26]	This Study
Functional Unit	1 ton of cultivated meat	1 kg of Chinese hamster ovary biomass	1 kg of protein-concentrated meal	4 oz of Beyond burger	4 oz of Impossible burger	1 kg of novel burger
System boundaries	Cradle-to-gate	Cradle-to-gate	Cradle-to-gate	Cradle-to-distribution	Cradle-to-gate	Cradle-to-gate
Life cycle inventory	European Life Cycle (ELCD) V2.0 and calculation	US LCI, Ecoinvent V2.0, European Life Cycle (ELCD) V2.0	Agri-footprint and Ecoinvent V3.1	Ecoinvent V3.0 and Agrifootprint V3.0	Ecoinvent V3.3, World Food Lifecycle Database V3.1	Ecoinvent V3.8 and Agribalyse V3.0.1
LCIA method
GHG emission	ISO14000	CML 2001	Impact 2002+	IPCC 2007	Impact 2002+	CML-IA
Energy use	ISO14000	Cumulative energy demand	Impact 2002+	Cumulative energy demand	-	Cumulative energy demand
Land use	ISO14000	Ecological footprint	Impact 2002+	Calculation	Impact 2002+	ReCiPe 2016 midpoint (E)
Water use	ISO14000	-	Impact 2002+	Calculation	Impact 2002+	ReCiPe 2016 midpoint (E)

shows the comparative analysis results including the system boundary and the LCIA method. In each study, 1 kg of the product was the functional unit, and a cradle-to-gate LCA was performed from raw materials to product manufacture. Different life cycle inventories and LCIA methods were adopted. Since this study aimed to analyze the life cycle impacts of SCiFi Foods’ novel burger before commercialization, we did not compare other types of products such as Chinese hamster ovary and protein-concentrated meal.

Two companies, Beyond Meat and Impossible Foods, have developed plant-based burgers that are supposed to be similar to conventional beef in terms of appearance, taste, and nutrition. These are available for purchase at the supermarket. Their life cycle studies show many differences such as raw materials, system boundaries, and life cycle inventory. Beyond Meat’s system boundary is cradle-to-distribution, which includes ingredients and raw material supply, processing and packaging operations, cold chain, distribution, and the disposal of packaging. Impossible Foods performed cradle-to-gate LCA from the raw material supply to packaging. The raw materials for the Beyond and Impossible burger production differ: The Impossible Burger [26] includes leghemoglobin, coconut oil, and sunflower oil; Beyond Burger [27] is made of pea protein, canola oil, and coconut oil. Their functional unit is a 4 oz burger patty. They also utilized different life cycle inventories, as shown in Table 1. Two LCA reports analyzed that both had lower impacts than beef in all life cycle indicators. It is not possible to perform a direct comparative analysis with novel burger since they did not publish detailed information and different assumptions (e.g., system boundary and functional unit). This study performed a simple comparative analysis by adjusting the system boundary and functional unit.

2.3. Life Cycle Inventory Analysis

2.3.1. Data Collection Procedure

SCiFi Foods, the developer of the novel cultivated meat production process, provided qualitative and quantitative information (e.g., compound types and amounts, process energy use, and equipment specification) to assess the LCI of the novel burger. The detailed information is the property of the companies and will not be recognizably published to guard the confidentiality of the data. All information including inputs and energy matched the appropriate processes in the selected database for inclusion in the inventory and entered into the OpenLCA software.

As discussed in Section 2.2, the inventory data of compounds vary depending on the database, which significantly affects the life cycle impacts. Before incorporating the compounds in the OpenLCA software, the commercial database for life cycle analysis was investigated.

Table 2. Material data availability in the database for fat and muscle cells.

	USCLI	Ecoinvent V3.8	ELCD V3.0	Agri-Footprint V5.0	Agribalyse V3.0.1
Compounds (62)	6	18	2	2	20
Salts (16)	5	12	1	1	12
Amino acids (21)	0	3	0	0	6
Vitamins (11)	0	0	0	0	0
Lipids (2)	0	0	0	0	1
Others (12)	1	3	1	1	2
Utilities (Heat & Elec.)	O	O	O	O	O

shows the life cycle inventory availability according to the database for fat and muscle cell cultivation. As shown in Table 2, the Agribalyse database included the most compounds. We selected Agribalyse V3.0.1 and Ecoinvent V3.8 to maintain consistency of the life cycle impacts for compounds.

2.3.2. Fat and Muscle Cell Production

The compounds can be categorized into six categories: salts, amino acids, vitamins, and lipids for incorporating the fat and muscle cell data in OpenLCA. We established the replacement strategy to release the lack of data availability in the database.

• Salts: Eighteen compounds (e.g., calcium chloride, ferrous sulfate, sodium chloride, and potassium chloride) were classified as salts. Among the 18 compounds, 12 compounds were included in the database. Six compounds (i.e., ferric nitrate, magnesium chloride, zinc sulfate, sodium pyruvate, hypoxanthine sodium, thymidine) were replaced with other compounds. The following compounds have been produced in a similar process or have a similar supply chain with the replaced compounds [24,36,37,38]:

  -

  Ferric nitrate (Fe(NO₃)₃·9H₂O): Iron sulfate.

  -

  Magnesium chloride (MgCl₂): Magnesium sulfate.

  -

  Zinc sulfate (ZnSO₄·H₂O): Manganese sulfate.

  -

  Sodium pyruvate (NaC₃H₃O₃): Sodium phosphate.

  -

  Hypoxanthine sodium (NaC₅H₃N₄O): Sodium phosphate.

  -

  Thymidine (C₁₀H₁₄N₂O₅): Glucose.

• Amino acids: Among the 21 compounds in the group of amino acids, only six compounds (i.e., glutamine, glycine, lysine HCl, threonine, tryptophan, and valine) are included in the database. We replaced other compounds using lysine and threonine since these compounds are produced using similar processes with lysine and threonine as follows [24,36,39,40]:

  -

  Lysine (C₆H₁₄N₂O₂·HCl): Arginine HCl, cysteine.HCl·H₂O, histidine.HCl·H₂O, methionine, phenylalanine, tyrosine.2Na·2H₂O.

  -

  Threonine (C₄H₉NO₃): Alanine, asparagine, aspartic acid, cysteine, glutamic acid, isoleucine, leucine, proline, and serine.

• Vitamins: There were no life cycle data for 11 vitamins in the database. Thus, the compounds were replaced by glutamine and lysine since the biotechnological production processes (i.e., fermentation and microbial/enzymatic transformation) are constructed for the commercial production of vitamins and related compounds, and the vitamin production process is similar to the other fermentation processes [41,42]:

  -

  Lysine (C₆H₁₄N₂O₂·HCl): Biotin, folic acid, riboflavin, and vitamin B12.

  -

  Glutamine (C₅H₁₀N₂O₃): D-calcium pantothenate, choline chloride, i-inositol, pyridoxal.HCl, pyridoxine.HCl, and thiamine.HCl.

• Lipids: We assumed that two compounds were synthesized in the process, from soybean oil to fatty acid. Using the process input and output data, we incorporated two compounds in the OpenLCA [24].
• Others: Twelve compounds (e.g., D-glucose, FBS, and water) were categorized as others in this study. Among these, four compounds (i.e., D-glucose, water, sodium selenite, and ascorbic acid) were included in the databases (i.e., Agribalyse and USLCI). One compound, HEPES, requires a negligible amount for fat and muscle cell cultivation; thus, it did not make the replacing strategy. As discussed in the vitamin categories, hormones (i.e., insulin, transferrin, FGF-2, and AlbuMAX Ⅱ) and putrecine.2HCl were replaced using lysine [41].
• Utilities: A heating source (i.e., steam) is required for fat and muscle cell cultivation in a high-temperature short-time sterilizer [43]. The consumed heat in the process was converted to the amount of natural gas from the Artemys food. The required amount of natural gas was incorporated into OpenLCA. Electricity, which is utilized for mixing and cell cultivation, was supplied from the energy grid. The life cycle data for electricity were based on the U.S. average data in the Ecoinvent database.

2.3.3. Novel Burger Production

As the next step, the input and output data for the novel burger were incorporated into OpenLCA.

Table 3. Information of the content of the burger ingredients provided by SCiFi Foods.

Substances	[wt%]
Water	50.5
Cultivated meat	16.9
Soy protein isolate	10.0
Coconut oil	7.0
Mushrooms	5.3
Sunflower oil	5.0
Oat	2.0
Methylcellulose	1.5
Potato starch	1.5
Salt	0.2
Citric acid	0.2

shows the content of the ingredients for the novel burger. Similar to the cell cultivation process, some life cycle inventory data were not included in the selected databases.

This study established the replacement strategy for methylcellulose, soy protein isolate, and cultivated meat. Carboxymethyl cellulose was substituted for methylcellulose since it has been used for similar objectives [44]. There are no data on the life cycle of soy protein isolate (SPI). The life cycle impacts of SPI showed a wide range (5.3 [45] to 20.22 [46] kg of CO_2-eq per kg of SPI) from consequential modeling (i.e., estimating from the energy use for soy protein isolate). Additionally, soybean has 0.38 kg of CO_2-eq per kg of soybean in the GHG emissions. In this study, we substituted soy protein isolate with soybean, and the required amount of soybean was revised using the GHG emissions difference using two GHG emission values (i.e., 0.38 and 5.3 kg of CO_2-eq for 1 kg of soybean and soy protein isolate, respectively). The fat and muscle cells were produced, then mixed to produce the cultivated meat at the mixing ratio of 50 and 50% of the fat and muscle cells, respectively [47].

3. Results and Discussion

3.1. Life Cycle Impacts

In this analysis, we compared the LCI of SCiFi Foods’ cultivated meat burger to the U.S. beef patty. The LCI of the novel burger for GHG emissions, energy demand, land use, and water use are shown in Figure 2. The novel burger generated 87% less greenhouse gas emissions, required 39% less energy, had 90% less influence on land use, and 96% less water use than the U.S. beef patty. CM cell cultivation accounted for the most significant portion of GHG emissions (63%) and energy demand (69%), and the burger ingredients showed the most part in land use (99%), and water use (86%). In contrast, the CM-ingredient and cleaning had a comparatively negligible impact on the LCIA results. These drastic reductions in LCI are similar to those in the reviewed studies. The novel burger showed different LCIA results than other alternative burgers and cultivated meat: GHG emissions of the Beyond and Impossible burgers were 3.3 and 3.4 kg of CO_2-eq/FU, respectively; the energy demand of Beyond burger was 50.1 MJ/FU (Impossible burger did not analyze the energy demand); the land use for the Beyond and Impossible burgers were 3.3 and 2.5 kg of M²_a-eq/FU, respectively; the water use for the Beyond and Impossible burgers were 0.01 and 0.11 kg of M³_eq/FU, respectively.

Detailed analysis is restricted due to several reasons: (1) A difference in the database and assessment method; (2) Input (e.g., compounds and energy demand) and output (produced amount and byproducts) for the cultivated meat production, as shown in Table 1; and (3) Detailed data were not provided in their reports [26,27].

In the GHG emissions, the most considerable GHG emissions were from CM-cell cultivation since that process is electricity-intensive. The second largest contributor to GHG emissions was burger ingredients utilized for the novel burger production using cultivated cells. Among the burger ingredients, SPI accounted for 72% of their life cycle impact since it contributed the largest share (16.85% of total) in the novel burger, except for water. The portion of CM-cell cultivation increased from 63% (GHG emissions) to 69% (energy use), since some parts of CM-cell cultivation are supplied from renewable sources (e.g., biomass, water, wind, and solar) for electricity production. The relative contribution of methylcellulose, oat, and sunflower to cumulative energy were increased in the burger ingredients compared to GHG emissions due to renewable energy use (e.g., geothermal energy and solar energy) for cultivation. Compared to the previous results (i.e., GHG emissions and cumulative energy demand), CM-cell cultivation did not critically impact land use. The SPI (8.57 m² a crop-eq/kg of novel burger) showed the most considerable contribution to land use. Additionally, the impact of coconut oil on the burger ingredients decreased more than that in the effects of the energy use of SPI due to two reasons: (1) A 141% higher content (16.9% and 7.0% of SPI and coconut oil in the novel burger, respectively) in the burger; and (2) a higher unit land use (50.9 m²a crop-eq and 6.6 m²a crop-eq per 1 kg of SPI and coconut oil, respectively.). In the water use indicator, the fraction of consumptive water use in the life cycle of burger ingredients (86%) was higher than that of the CM-cell cultivation (10%). However, the portion of coconut oil was the highest in the burger ingredients compared to other indicators: the unit water use for 1 kg of SPI was 0.12 m³; on the other hand, coconut oil used more water (0.96 m³/kg of coconut oil).

3.2. Uncertainty Analysis

The LCIA results depend on the quality and adequacy of life cycle data that are used for the study. This section describes the uncertainty assessments for two products, a beef patty and the novel cultivated meat burger. In the uncertainty assessment, the range of uncertainty in estimating the flows of compounds and energy in the systems and the uncertainty in the emissions of pollutants were considered. Monte Carlo simulation is widely used for uncertainty assessment. It varies the values of the input data based on knowledge about their uncertainty [48]. The database contains many processes, and a general LCA evaluates environmental impacts using the average value of various process data. The environmental impacts were calculated using many assumptions for each process. As more processes are included in the database, the results in the uncertainty analysis showed a higher uncertainty and a larger range of values. For all stochastic variables of the LCI, the Monte Carlo simulation randomized all parameters according to their respective range of uncertainty. This approach was implemented in common LCA software including OpenLCA. The Monte Carlo method requires many simulations to obtain representative results over 1000 runs; commonly, 1000 or 10,000 runs are used in the LCIA [49,50]. This study applied 1000 iterations to ensure stable variance for the novel burger. The uncertainty assessment results for beef were also extracted from other studies. The results were used to select the manipulated values and scopes in further analysis (i.e., case study and sensitivity analysis).

Figure 3 shows the uncertainty results based on the Monte Carlo simulations for the beef and novel burger in accordance with the GHG emissions, energy demand, land use, and water use. The boxes spanned ±1 standard deviation around the means, which were as follows: GHG emissions: 3.5 and 30.6 kg of CO_2-eq; energy demand: 46.8 and 78.6 MJ; land use: 9.0 and 62.0 m²a_eq; water use: 0.0 and 0.9 m³_eq per kg of products for the novel burger and beef, respectively. The ends of the whiskers represent the 5th and 95th percentiles. Specifically, the water use reached 0.0 m³_eq due to double accounting of the input and the use of natural rainfall in some processes: the output from the wastewater treatment plant was connected with the input from the process, which produces wastewater [51,52], causing the negative water use in the LCIA results. To compare the overall precision of two products, the coefficient of variation was utilized in this study, and the coefficient of variations (COV) were calculated using the following equation [53]:

$$\mathrm{COV}\hspace{0.17em}=\hspace{0.17em}\frac{\mathrm{Standard}\hspace{0.17em}\mathrm{deviation}}{\mathrm{Mean}}\hspace{0.17em}[\%]$$

As shown in

Table 4. The coefficient of variations for the novel burger and beef patty.

	GHG Emissions	Energy Demand	Water Use	Land Use
Novel burger	6.8%	4.6%	14.9%	130.0%
Beef burger	10.3%	12.2%	40.0%	29.2%

, the beef burger showed higher COV except for land use, which means that the beef had higher uncertainties than the novel burger. Through this comparison, it can be interpreted that the estimation and evaluation of the LCA of the novel burger were more precise than that of a beef burger in real-life.

The COV ranged from 10% (GHG emissions) and 40% (water use) for the beef burger. In contrast, the novel burger showed from 5% (energy demand) to 130% (land use) of COV. Generally, a value of less than 10% for COV implies that the LCA results are reliable [54]. The results in the GHG emissions and energy demand for the novel burger can be evaluated as reliable. On the other hand, land use showed a comparatively higher COV. Thus, the included data in the Ecoinvent database for land use of the compounds and results had a higher uncertainty. Nevertheless, the value fluctuation range of the novel burger was lower than that of the beef burger, which can be interpreted as being more eco-friendly than a traditional burger.

3.3. Spatial Analysis: Effect of Facility Location

Similar studies [24,26,34] have analyzed that energy consumption critically impacts the life cycle indicators. However, their research solves the problem by using a high content of renewable energy sources or energy mix from 2010. Additionally, the GHG emissions for electricity production are highly volatile, according to the energy sources (e.g., crude oil, natural gas, and solar). Recently, renewable energy, which has low GHG emissions, is being utilized to mitigate global warming. In particular, renewable energy sources are being rapidly adopted for electricity production and are transparently documented on the website [55]. In this section, we performed a sensitivity analysis to show that the facility location analysis significantly affects the GHG emissions and requirements for the production of sustainable alternative burgers.

Parameters and assumptions. The GHG emissions according to the eGRID are very different: NEWE (NPCC New England/Eastern Power Grid) and HIOA(HICC Oahu/Hawaii power Grid) have 0.106 kg of CO_2-eq/kWh and 0.750 kg of CO_2-eq/kWh, respectively [55]. There is a big gap in GHG emissions according to the eGRID subregions. There are 27 eGRID subregions in the U.S. This study selected two subregions: regions with the highest and lowest GHG emissions, among the west, midwest, north, and south regions. According to the year and subregion, the GHG emissions were different, as shown in Figure 4. For example, the reductions of GHG emissions in the NYUP (NPCC Upstate NY/Eastern Power Grid) reached 134% from 2010 to 2020. In some sub-regions (e.g., NYLI (NPCC Long Island/Eastern Power Grid) and U.S. average), GHG emissions in 2020 were higher than in 2018. This study used the newest GHG emission values (in 2020) in each eGRID subregion.

In Section 3, we carried out the LCIA of the novel burger by using the value of 507 kg of CO_2-eq/1 MWh for electricity. However, the GHG emissions in 2020 were at least 105.9 kg of CO_2-eq/MWh (in NYUP). As the portion of fossil fuels including coal and oil was higher, the GHG emissions were more elevated. This analysis shows the implications on the LCIA of the eGRID region in which the cultivated meat plant is built. Since the novel burger is ready to be commercialized, the GHG emissions of the currently commercialized products including the Impossible burger and Beyond burger were compared.

Results and discussion. Using the differences in GHG emissions according to eGRID, we conducted a the case study for eight sub eGRIDs and the U.S. average, as shown in Figure 4. We identified that the GHG emissions for 1 kg of novel burger are 3.84 kg of CO_2-eq (using 2019 data). The GHG emissions in the case study were from 1.93 to 4.09 kg of CO_2-eq/kg of the novel burger since the unit GHG emissions per electricity were different according to the eGRID subregions (Figure 5).

Additionally, the ranges were from 3.1 to 4.0 kg of CO_2-eq/kg for the Impossible burger [26]. This range was calculated through uncertainty analysis, but they most likely did not consider the effect of regional variation in the electricity mix. The GHG emissions were 3.31 kg of CO_2-eq/kg for Beyond Burger [27]. The GHG emissions for Beyond burger was lower than the prior results for the novel burger, and the novel burger was in the range of the Impossible burger. Significantly, the three eGRID subregions (e.g., CAMX, NYUP, and SRVC) showed lower GHG emissions. We found that for the GHG emissions from the novel burger to be less than the Impossible and Beyond burgers, the unit GHG emissions per unit of electricity had to be 477.2 kg of CO_2-eq/MWh for the electricity supply in the CM-cell cultivation step. If the novel burger production facility is installed in 18 eGRID regions except for nine out of 27 eGRIDs in the United States (based on the 2020 year), it is possible to produce alternative meat with less GHG emissions than the commercialized burger patty. This is under the assumption that the source of electricity for the competing burgers does not change.

3.4. Sensitivity Analysis

In this study, some data on cultivated meat production were based on many assumptions, thus, having high uncertainty and possibility for enhancement in the LCIs. We analyzed the effect of the technical data used in this study on LCI for the novel burger. This study can further investigate whether the improvement in the environment and sustainability of the novel burger is increased with technological progress.

Parameters and assumptions. The sensitivity analysis of the significant technical parameters and assumptions allows for decision-makers (e.g., policymakers or stakeholders) to acquire the preliminary assessment results before commercialization for the novel burger and identify the hotspots and bottlenecks. In Section 3, the CM-cell cultivation and burger ingredients had the highest impact among the LCIA results: CM-cell cultivation accounted for the most considerable portion of GHG emissions and energy demand; soy protein isolate and coconut oil highly impacted land use and water use, respectively.

In the techno-economic assessment performed by SCiFi Foods, the motor’s average output was assumed to be 60% (bioreactor motor) and 70% (compressor), and this compressor output can be increased or decreased due to various improvements such as technical innovations [25]. Additionally, agricultural technology development has increased the cultivation yields per area for soybean and coconut [56,57]. In the database (Agribalyse) used for this study, the required areas were 3.61 and 6.61 m² for 1 kg of soybean and coconut, respectively. We assumed ±25% changes in the yields of the technical data, as shown in

Table 5. Assumptions about the technical yields for the sensitivity analysis of the LCI.

	−25%	Base Line	+25%
Motor yield [%]	45/53	60/70	75/88
Soybean cultivation yield [kg/m2]	0.045	0.059	0.074
Coconut cultivation yield [kg/m2]	0.107	0.143	0.179

. We assumed and adjusted that all inputs (e.g., energy and water) for each technical data changed along with the yield variations.

Results and discussion. According to the selected technical data, the sensitivity analysis results are summarized in Figure 6. The baseline corresponded to the LCIA results for the novel burger, as discussed in Section 3. As the yields increase, the LCIs of the novel burger increase, which means that technological innovations can contribute to enhancing the environmental impact. The GHG emissions are very sensitive to the motor yield (±16% changes), since GHG emissions are significantly impacted by the electricity to utilize the motor. Furthermore, the motor yield showed a higher impact on the energy demand than the GHG emissions. In contrast, the effects of the burger ingredients were decreased compared to the GHG emissions since CM-cell cultivation accounted for a larger part of the energy demand. However, the land-use variation by the motor efficiency was negligible. Overall, as the portion of each compound or process to the LCI was high, the variation in the LCI was also higher. It can be interpreted that the assumed data and parameters to construct the LCI of the compounds or processes significantly impact the uncertainties of the LCI, and there is a high probability of updating the LCI.

4. Conclusions

The meat industry significantly impacts the environmental problems; cultivated meat has been evaluated as one of the most promising alternatives to mitigate the climate and environmental issues caused by the current meat industry. The study aimed to assess the environmental impacts of a novel burger that aims to combine the best of plant- and animal-based ingredients. The input and output data for LCA were supplied from the pilot plant of SCiFi Foods. This study performed various analyses to identify the bottleneck in the environmental perspectives and tentative variations of the LCA. The major findings in this study are as follows:

• Like other alternative meat and burgers, SCiFi Foods’ novel burger is environmentally better than traditional beef. GHG emissions, land use, and water use were less than 15% of the beef patty. Moreover, the energy use of the novel burger was 44% lower than that of the beef patty.
• Through the uncertainty analysis, the LCIA results of the novel burger were comparatively more reliable than the traditional meat for assessing the real-life impact by comparing the COV. Using the sensitivity analysis and case study, the LCIA results can be enhanced for the novel burger and can be further improved through technological innovation.

This study demonstrated that the novel burger could make significant environmental improvements over traditional beef. In addition, it can be competitive with Impossible and Beyond burgers from an environmental perspective since the novel burger had similar LCIA results to the Impossible and Beyond burgers. Using the LCIA results for alternative meat, it was identified that shifting from livestock to cultivated meat can contribute to the sustainability of the food industry. Additionally, this study provides the milestones for additional related studies including updating the life cycle inventory to mitigate the uncertainty of the inventory data and the analysis of meat analogs before commercial production. Stakeholders and investors can utilize these results to select the factory location and establish the strategy for replacing the ingredients to improve the environmental impacts. Based on this study, future research can be extended to different system boundaries by considering the packaging and distribution process.