Data for: Environmental Life Cycle Assessment of the SpiralG Biorefinery

SUPPLEMENTARY DOCUMENT 1 Description of the SpiralG biorefinery model Last updated on January 2025. School of Biosystems and Food Engineering, University College Dublin, Ireland BiOrbic, Bioeconomy SFI Research Center, University College Dublin, Ireland 1 2 Léa Braud , Kevin McDonnell , Fionnuala Murphy 1 1,2 *1,2 Spiralgorithm This work was supported by the Circular Bio-based Industries Joint Undertaking (CBE-JU) [Grant No. 792257]. The CBE-JU receives support from the European’s Union’s Horizon Europe research and innovation programme and the Bio-Based Industries Consortium. The funders had no role in study design as well as data collection and analysis. SUPPLEMENTARY DOCUMENT 1 Description of the SpiralG biorefinery model. Braud L.1,2, McDonnell, K.1, and Murphy, F.1,2 1School of Biosystems and Food Engineering, University College Dublin, Ireland 2BiOrbic, Bioeconomy SFI Research Center, University College Dublin, Ireland Content: This document describes the methods used to collect and aggregate the pilot scale data of the SpiralG biorefinery which serve as a basis for the environmental life cycle assessment (LCA) studies. A general overview of the biorefinery is given before detailing the approaches used to convert the raw data into aggregated data. Then, the model constructed from the resulting datasets is presented with a focus on two influencing parameters: Spirulina biomass productivity and phycocyanin extraction efficiency. This model is used to analyse the aggregated data and perform a mass balance of the biorefinery. Finally, the framework coded in the Python programming language using Brightway, used to perform the LCA calculations and visualise the results, is described. Contents 1 General description of the Spirulina biorefinery 3 1.1 Development of three marketable products . . . . . . . . . . . . . . . . . . . . . . 3 1.2 Construction of the biorefinery system diagram . . . . . . . . . . . . . . . . . . . 4 2 Data collection and aggregation 4 2.1 Collectionoftherawdata............................... 4 2.2 Aggregation of the raw data at process level . . . . . . . . . . . . . . . . . . . . . 6 3 Spirulina biorefinery model 8 3.1 Technologicalperiod .................................. 8 3.2 Spirulina biomass productivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 3.2.1 Definition of the Spirulina biomass productivity parameter . . . . . . . . 10 3.2.2 Calculation of the initial values . . . . . . . . . . . . . . . . . . . . . . . . 10 3.2.3 Variations of Spirulina biomass productivity . . . . . . . . . . . . . . . . . 11 3.3 Phycocyanin extraction efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 4 Environmental life cycle assessment 15 4.1 Foreground and background databases . . . . . . . . . . . . . . . . . . . . . . . . 15 4.2 LCA calculations using Brightway . . . . . . . . . . . . . . . . . . . . . . . . . . 15 4.2.1 Initialisation .................................. 15 2 SD2: Description of the SpiralG biorefinery model 4.2.2 Databaseimport ................................ 17 4.2.3 LCAcalculations................................ 17 4.3 Visualisation of the LCA results . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 1 General description of the Spirulina biorefinery 1.1 Development of three marketable products The SpiralG project gathers three industrial partners located in France and Italy which are each responsible for a portion of the Spirulina value chain. Livegreen1is a company established in Arborea, a city of approximately 4,000 inhabitants located in the province of Oristano in Sardinia (Italy). The economy of Arborea is based on the local agricultural activities revolving around the Cooperativa Assegnatari Associati Arborea2founded in 1956. Livegreen cultivates Spirulina in open raceway ponds (ORPs) and commercialises “pure” biomass in the form of tablets and spaghettini as well as transformed products inspired by traditional Sardinian cuisine (e.g. Fregula Sarda, Carasau Guttiau bread). This work was limited to the assessment of the environmental impacts associated with the production of spaghettini. Indeed, only this form of biomass was used in the phycocyanin extraction process. The meteorological conditions in Arborea (e.g. light incidence, temperature) are favourable for the cultivation of Spirulina 90% of the year. The biomass is harvested daily, dewatered, shaped into spaghettini, and dried. These pre-processing steps allow the reduction of the water content of Spirulina biomass from 99% to 5% and facilitate its transportation for further processing. The dry spaghettini are shipped to Greensea3, a company located in the city of Mèze, in the South of France. Phycocyanin is extracted using water as a solvent and approximately 50% of pure pigment is recovered in the blue extract after several separation (e.g. centrifugation) and purification (e.g. ultrafiltration) steps. This process co-produces two main biomass fractions with added value. The co-product obtained from the first separation process, called co-product A (CPA), is transported to Algaia4, a company located in Saint-Lô (France). After an acid hydrolysis to break down the proteins and two purification steps, the product enriched in amino acids is concentrated. Due to its high content in nitrogen, the CPA concentrate (CPAc) has promising applications in the agricultural industry. The second co-product, called co-product B (CPB), is obtained from the ultrafiltration of the blue extract and is treated at Greensea directly. CPB is concentrated and the product obtained, which contains small colourless proteins, can be used in the cosmetic industry. The three extracts produced in the Spirulina biorefinery, i.e. the blue extract, CPAc, and CPB concentrate (CPBc), are sold as ingredients to formulators and incorporated into food, cosmetics, and agricultural products (see Table 1). Due to confidentiality, the exact biochemical composition of each fraction is not detailed. 1https://livegreen.bio/ 2https://www.arborea1956.com. 3https://en.greensea-all.com/ 4https://www.algaia.com/ 3 SD2: Description of the SpiralG biorefinery model Table 1: Description of the Spirulina biorefinery products and their potential applications. Biomass fraction General composition Potential application(s) Blue extract 50% of pure phycocyanin Natural food colouring & ingredient for cosmetics products CPAc Enriched in amino acids Ingredient for agricultural & feed products CPBc Enriched in colourless proteins Ingredient for food & cosmetics products 1.2 Construction of the biorefinery system diagram In order to facilitate the assessment of the environmental impacts of the Spirulina biorefinery, three subsystems were defined (i.e. each corresponding to one industrial partners of the SpiralG project): Spirulina cultivation and biomass pre-processing in Arborea (Italy) (subsystem 1 or S1), phycocyanin extraction in Mèze (France) (subsystem 2 or S2), and co-product A treatment in Saint-Lô (France) (subsystem 3 or S3) (see Fig. 1). In the articles, the name of the companies are not mentioned and each section of the biorefinery is referred to as S1, S2, and S3. Each of these subsystems is divided into processes, also called “activities” representing the smallest unit for which data were collected on-site. The activities “S1.A0.Building” and “S1.A0.Operation”, initially included in S1, were transferred to the subsystems “infrastructures” and “operation”, respectively. Similarly. “S1.A8.Transport” and “S2.A8.Transport” were both transferred to the subsystem “transport”. 2 Data collection and aggregation 2.1 Collection of the raw data Six data collection campaigns were conducted at the three pilot scale facilities from 2019 to 2022 (see Table 2). The raw data collected on-site (i.e. primary data) correspond to the amounts of electricity, water, nutrients, chemicals, and materials used in the three subsystems. Data were obtained from direct measurements since no data logger or sensors allowed the retrieval of data for most of the equipment. Only the centrifuge and ultrafiltration machines in S2 had data loggers to measure the volumes of water and chemicals used during the cleaning programs. In S1, Spirulina biomass was harvested daily. The processes from “S1.A2.Filtration” to “S1.6.Packaging” were repeated, allowing the collection of daily datasets. The average values obtained over the week were used as inventory data in the LCA studies. Regarding S2 and S3, the processes were sequentially performed over the week. No data duplicates were measured and the datasets are based on unique values. First written by hand in a notebook, the data were then digitalised and presented in the form of Excel files containing the original data, calculations, and text (e.g. description of the data collection procedure for each item, description of the data obtained). The data for S1, S2, and S3 were collected independently since the respective facilities are located in different regions of Italy and France. Regarding S1, the data were collected in July 2019 and July 2022, during the week 30 and 28, respectively. The climatic conditions were similar during the two data collection periods. This is of particular importance since the temperature of the air, light incidence, and cloudiness affect Spirulina growth in ORPs. In contrast, the data for S2 and S3 were collected in 2021 and 2022 at different periods. Since the phycocyanin extraction and co-product treatment are performed indoor, the meteorological conditions do not significantly affect the processes. 4 SD2: Description of the SpiralG biorefinery model S1.A1.Cultivation S1.A2.Filtration S1.A3.Dewatering S1.A4.Shaping S1.A5.Drying S1.A6a.Packaging S1.A7.Freezing Spirulina broth Spirulina slurry S1.A6b.Packaging Spirulina paste Spirulina wet spagh. Spirulina paste, packaged Spirulina paste, packaged, frozen Dry line Wet line Spirulina dry spaghettini, packaged S1.A0.Building S1.A0.Operation Subsystem 1 Infrastructures Operation Spirulina dry spagh. S2.A1.Maceration S2.A2.Centrifugation S2.A3.Filtration S2.A4.Ultrafiltration1 S2.A5.Ultrafiltration2 S2.A7a.Packaging Spirulina mix Filtrate Retentate Blue extract Supernatant Blue extract, packaged CPA, unprocessed CPB, unprocessed S2.A6.Concentration S2.A7b.Packaging CPB concentrate CPB concentrate, packaged Subsystem 2 S3.A1.Extraction S3.A2.Diafiltration S3.A3.Ultrafiltration S1.A4.Concentration S1.A5.Stabilisation S1.A6.Packaging Hydrolysate Permeate Retentate Permeate Concentrate Concentrate, stabilised Retentate CPA concentrate, packaged, stabilised Subsystem 3 S1.A8.TransportS12 Transport S2.A8.TransportS23 T T Legend: Transport T Process or activity Product flow (France) (France) (Italy) Figure 1: Simplified system diagram of the Spirulina biorefinery. The abbreviation “spagh.” refers to “spaghettini”. 5 SD2: Description of the SpiralG biorefinery model In the case of S2, the temperature in the pilot hall influences the need to cool down the blue extract after extraction. In fact, the high temperature in the pilot hall in summer results in a larger consumption of energy to cool down the blue extract. Table 2: Description of the data collection campaigns conducted in the frame of the SpiralG project. No. Period Description 1 July 2019 Data collection at the Spirulina cultivation and biomass pre-processing facility in Arborea, Italy. The plant was inaugurated in April 2019 and it was therefore its first year of operation. 2 January 2021 Data collection at the phycocyanin extraction facility in Mèze, France. The pilot scale trial was performed using 50 kg of dry Spirulina spaghettini. The second ultrafiltration step was interrupted and the concentration of the CPB not conducted. 3 February 2021 Data collection at the CPA treatment facility in Saint-Lô, France. The pilot scale trial corresponded to the first based on acid hydrolysis to extract the proteins from CPA. The concentration of the permeate from ultrafiltration was interrupted. 4 April 2022 Data collection at the CPA treatment facility in Saint-Lô, France. Not all the permeate from diafiltration was ultrafiltered. The evaporation of the retentate from ultrafiltration was interrupted. 5 July 2022 Data collection at the Spirulina cultivation and biomass pre-processing facility in Arborea, Italy. No measurements of electricity consumption were performed. 6 August 2022 Data collection at the phycocyanin extraction facility in Mèze, France. The pilot scale trial was performed using 12.5 kg of dry Spirulina spaghettini. The CPB was concentrated using an evaporator. The two successive ultrafiltration steps worked. 2.2 Aggregation of the raw data at process level The raw data were aggregated at process level to overcome the issue of confidentiality. The aggregation consists of summing the individual measurements for each exchange in an activity. For instance, the total volume of water used in “S1.A3.Dewatering” was measured by collecting the water exiting the water presses into containers which were weighed at the end of each water press cycle (i.e. each batch). The volume of water used to clean the water presses at the end of the process (e.g. steel structure, cotton bags) was measured using a water meter installed on the hose used for cleaning. The details of each volume measured was noted in the raw dataset. During the aggregation, the total volume used per day was calculated by summing the individual values of the raw datasets to obtain a unique volume of water. The same procedure was applied to electricity, chemicals, materials etc. In addition, the aggregation procedure allowed the identification of data gaps. Although the data were collected on-site, certain measurements could not be performed due to a malfunctioning or lack of equipment. For instance, only one water meter was available to measure volume of Spirulina broth harvested in S1. The data collected for two ORPs were used as proxy for the four other ORPs. Mass and water balances were conducted at process level to verify the dataset and fill in the data gaps were necessary. The use of Lavoisier’s law of mass conservation allowed to calculate missing values. This principle states that for any closed system to transfer of matter and energy, the mass of the system remains constant over time. The quantity of mass 6 SD2: Description of the SpiralG biorefinery model within the system cannot change and therefore nothing can be added or removed. Several processes were simultaneous, i.e. the biomass output of a process was directly pumped into the next machine and could therefore not be measured. For instance, the hydrolysate obtained from “S3.A1.Extraction” was directly pumped into the tank of the ultrafiltration machine. As a result, the amount of hydrolysate produced was calculated from the mass balance using the principle of conservation (biomass input = biomass output). Ultimately, no losses were accounted for in the process. In addition, certain measurements led to unbalanced processes (e.g. the amount of biomass measured at the end of a process was larger than at the start). In this case, the amounts were adjusted to balance the process by using simple cross multiplications. The mass and water balances of each activity of the three subsystems are detailed in another document. S1 - Spirulina cultivation - Livegreen Confidential dataset - S1.A1.Cultivation Léa Braud - WP5 - SpiralG project 12 Livegreen, July 2022 Data collected for the activity/process "S1.A1.Cultivation" Amount of nutrients added to the ORPs Table 1: Calculation of the amount of nutrients added to the ORPs weekly Nutrient (kg) ORP1 ORP2 ORP3 ORP4 ORP5 ORP6 Tot./week Av./day 105.77 0.00 0.00 0.00 37.82 0.00 143.60 20.51 20.00 17.00 19.00 35.00 38.00 35.00 164.00 23.43 1.84 0.00 0.00 0.00 0.00 0.00 1.84 0.26 0.65 0.40 0.46 0.22 0.31 0.40 2.44 0.35 Chelated Iron (6%) (Fe-EDDHA) 0.16 0.22 0.24 0.12 0.17 0.21 1.12 0.16 Table 2: Estimation of the amount of plastic packaging used (nutrient salts bags) Amount Unit Total amount of sodium bicarbonate 143.60 kg Amount of NaHCO3 per bag 25 kg Number of bags used 5.7 - Number of bags discarded 50.0031 kg Total amount of PE-LD (4) 0.449 kg General description: The dataset of the activity "S1.A1.Cultivation" contains information regarding the amount nutrients supplied to the ORPs to meet Spirulina growth requirements, electricity and water as weel as the volume of broth harvested (exiting the ORPs). The cells coloured in dark blue indicate the vaues used to perform the LCA study (see Table I). The dataset was aggregated at process level i.e. only the sum of the data obtained for each exchange are shown. These data were used in the LCA study to be disclosed to the public (e.g. research article, SpiralG deliverable). Data collection procedure: The nutrients (or nutirent salts) are added once a week to the ORPs, usually on Saturdays. During the data collection, the nutrients were added on Friday 15/07/2022. The amounts were calculated from the analysis of the ORP culture medium conducted on the same day. The amount of magnesium sulfate and chelated iron is based on the equivalent dry biomass produced over]the five previous working days. Sodium bicarbonate (NaHCO3) Potassium nitrate (KNO3) Ammonium phosphate (NH4H2PO4) Magnesium sulfate (MgSO4) Data analysis: The amount of magnesium sulfate and chelated iron added to the ORPs corresponded to 20 g / kg dry biomass and 17 g / kg biomass, resp. A daily average of nutrients was calculated to fit with the rest of the data (i.e. daily production of wet or dry biomass). It was therefore assumed that the nutrients were supplied in small quantities everyday rather than all on the Saturday. The total amount of each nutrient added to the six ORPs was divided per seven since the algae grows continuously i.e. over a full week and not a "working week" of 6 days. The average per day therefore corresponds to the total values presented in Table 1, divided per seven. These values were used in the inventory data of the LCA study. Regarding the plastic wastes related to the packging of the nutrients, none was considered. In 2019, only the plastic packaging of sodium bicarbonate was considered since it was used in large quantities. The amount of empty plastic bags discarded was estimated from the data collected in 2019. Data collection procedure: In 2019, the amount of plastic bag discarded after the supply of sodium bicarbonate to the ORPs was measured. A total value of 2.452 kg of plastic for the supply of 785 kg of sodium bicarbonate was obtained. Amount of plastic wasted / kg NaHCO3 Figure 2: Extract of a data collection sheet used to communicate the results with the industrial partners. The data for S1 were collected for each ORP, every day. The raw data were replaced by blank spaces due to their confidentiality. Only the aggregated data i.e. per day were shown. 7 SD2: Description of the SpiralG biorefinery model 3 Spirulina biorefinery model The Spirulina biorefinery model was developed in the Python programming language using the data collected on-site and described in Section 1. The first objective of the model is to facilitate the comparison of the data collected during the different data collection campaigns. The datasets consisting of data collected in 2019 and 2021 correspond to the period 1 while the most up-to-date data, i.e. collected in 2022, correspond to the period 2. The comparison of the datasets of periods 1 and 2 allows to evaluate the effects of the technological improvements made across the value chain on the consumption of energy, water, materials, and chemicals (and to a larger extent on the environmental sustainability of the biorefinery). The second objective of the model is to evaluate the effects of changes in the Spirulina biomass productivity and phycocyanin content on the mass balance of the biorefinery (i.e. proportion of blue extract, CPAc, and CPBc) and the consumption of the energy, water, materials, and chemicals (and to a larger extent on the environmental sustainability of the biorefinery). The model is divided into the three subsystems described above. Each process (or activity) is computed with a set of functions which typically take as input the amount of biomass to process and the technological period to return the amount of energy, water, materials, and chemicals used in the process plus the waste generated. Therefore, the outputs of the model correspond to datasets for each subsystem with data aggregated at process level. The datasets can be directly used to perform the LCA study of the biorefinery. The parameters of the model were defined based on the research objectives and classified into four categories. The operational parameters include the technological period, number of working days, type of the line (e.g. wet, dry), and material lifespan. In addition, the technical parameters consist of the harvesting efficiency, phycocyanin extraction efficiency, number of ORPs, number of vibrating filters (VFs), and volume of ORP. Finally, the biological parameters include the Spirulina biomass productivity and phycocyanin content. The last parameters correspond to the transportation distances between S1, S2, and S3. In total, 14 parameters were defined among which 6 were used in the environmental LCA studies (see Table 3). The other parameters were considered as constant. The number of working days was set to 330, the total cultivated area to 705 m2, the number of ORPs to 6, the number of VFs to 3. In addition, the volume of the ORPs was considered as constant (i.e. the volume of water used to fill in the ORPs compensates the losses per evaporation). 3.1 Technological period The technological period is the main parameter to all activities except for the ones that occurred in period 2 only (e.g. “S2.A5. Ultrafiltration 2”, “S2.A6.Concentration”). This parameter was used to calculate distinctive datasets for periods 1 or 2 based on the data collected for each activity. For instance, the technological period influences the amount of Spirulina paste obtained after dewatering (see Listing 1). 1def BiomassDewatering ( tech_period , slurry_DW ): 2biomass_balance_dict_sc1 = {} 3biomass_balance_dict = {} 4if tech_period == ’1’: 5## data collected 6slurry_sc1 = 257.71 #amount slurry [kg] 7DM_slurry_sc1 = 14.75 # dry matter content slurry [%] 8slurry_DW_sc1 = slurry_sc1 * DM_slurry_sc1 /100 # amount slurry [ kg DW - eq ] 9paste_sc1 = 91.45 # amount paste [ kg ] 8 SD2: Description of the SpiralG biorefinery model Table 3: Description of the parameters defined in the Spirulina biorefinery model. The underlined parameters correspond to the ones used in the LCA studies. The rest were considered as constant. Parameter Description Unit Values Technological period Period during which the data were collected. Period 1 corresponds to 2019 (S1) and 2021 (S2,S3). Period 2 corresponds to 2022 (S1, S2, S3). - ‘1’ or ‘2’ Productivity Daily areal dry spaghettini productivity g/m2/day [0; 12] Phycocyanin content Proportion of phycocyanin in Spirulina biomass % [5; 15] Working days Number of days during which Spirulina is harvested and the biomass processed day [330; 365] Harvesting efficiency Efficiency of the filtration step i.e. ratio between the amount of biomass recovered in the slurry and the amount of broth filtered % [70; 100] Line Dry or wet line for Spirulina biomass preprocessing. - ‘dry’ or ‘wet’ Distance by car Distance between the Spirulina cultivation facility and the local industrial harbour km [10; 50] Distance by ship Distance between the industrial harbours in Italy and France km [500; 800] Distance by truck Distance between the industrial harbour in France and the phycocyanin extraction facility km [50; 200] Distance by refrigerated truck Distance between the phycocyanin and the coproduct treatment facility in France km [500; 900] Number of ORPs Number of open raceway ponds in the Spirulina cultivation facility unit 6 Number of VFS Number of Vibrating Filtering Systems in the Spirulina biomass processing facility unit 3 Volume of ORP Volume of one open raceway pond m3[150; 250] Material lifespan Lifespan of the materials used in the construction of the greenhouse, cultivation facility, and biomass processing facility. year [5; 20] 10 DM_paste_sc1 = 33.07 #dry matter content of paste [%] 11 paste_DW_sc1 = paste_sc1 * DM_paste_sc1 /100 12 biomass_balance_dict_sc1[’slurry’]={’wet_mass ’: slurry_sc1 , ’DM_content ’ : DM_slurry_sc1 , ’dry_mass ’ : slurry_DW_sc1 } 13 biomass_balance_dict_sc1[’paste ’]={’wet_mass ’:paste_sc1 , ’DM_content ’: DM_paste_sc1 , ’dry_mass ’:paste_DW_sc1} 14 ## data modelled 15 paste_DW = slurry_DW * paste_DW_sc1 / slurry_DW_sc1 16 biomass_balance_dict[’slurry’] = slurry_DW 17 biomass_balance_dict[’paste ’] = paste_DW 18 biomass_balance_dict[’losses’] = slurry_DW - paste_DW 19 if tech_period == ’2’: 20 ## data collected 21 slurry_sc1 = 198.99 #amount slurry [kg] 22 DM_slurry_sc1 = 10.39 # dry matter content of slurry [%] 23 slurry_DW_sc1 = slurry_sc1 * DM_slurry_sc1 /100 # amount slurry [ kg DW - eq ] 24 paste_sc1 = 77.38 # amount paste [ kg ] 25 DM_paste_sc1 = 23.26 #dry matter content of paste [%] 26 paste_DW_sc1 = paste_sc1 * DM_paste_sc1 /100 27 biomass_balance_dict_sc1[’slurry’]={’wet_mass ’: slurry_sc1 , ’DM_content ’ : DM_slurry_sc1 , ’dry_mass ’ : slurry_DW_sc1 } 28 biomass_balance_dict_sc1[’paste ’]={’wet_mass ’:paste_sc1 , ’DM_content ’: DM_paste_sc1 , ’dry_mass ’:paste_DW_sc1} 29 ## data modelled 9 SD2: Description of the SpiralG biorefinery model Figure 4: Structure of the LCA framework based on Brightway. 1input : 2analyses: 3CA_subsystems : 4relative_stacked_bar_plot: True 5bar_chart_one_ic: False 6CA_processes: 7relative_stacked_bar_plot: False 8pie_chart_stacked_bar_plot: False 9stacked_bar_plot_comparison: False 10 CA_input_category: 11 stacked_bar_plot: False 12 heatmap : False 13 CA_processes: 14 relative_stacked_bar_plot: 15 databases : 16 - db_infrastructures 17 - db_operation 18 - db_S1 19 - db_S2 20 - db_S3 21 - db_transport 16 SD2: Description of the SpiralG biorefinery model 22 lcia_methods: 23 - !! python / tuple [ ReCiPe Midpoint (H) V1 .13 , fossil depletion , FDP ] 24 - !! python / tuple [ ReCiPe Midpoint (H) V1 .13 , climate change , GWP100 ] 25 - !! python / tuple [ ReCiPe Midpoint (H) V1 .13 , metal depletion , MDP ] 26 - !! python / tuple [ ReCiPe Midpoint (H) V1 .13 , water depletion , WDP ] Listing 2: Extract of the recipe YAML file containing information to run the LCA algorithm. 4.2.2 Database import Once a project is selected, the background and foreground databases are imported following the recipe. First, the background databases biosphere3 and ecoinvent 3.6 cut-off are imported using the Ecospold2Importer. The database biosphere3 contains elementary flows for which the names have been normalised to eocinvent 3. The function bw2setup() imports biosphere3, LCIA methods, and additional metadata required to import other databases. The datasets are differentiated using unique identifiers which include the name of the database and a code such as a number, UUID, and name. For instance, the code ("biosphere", "f66d00944691...") is a valid identifier. LCIA methods corresponds to tuples with the general name of the method (e.g. ReCiPe Midpoint (H) V1. 13), the name of the impact category (e.g. climate change), and its abbreviation (e.g. GWP100). The foreground databases are imported in the bw2package format and are linked to the ecoinvent 3.6 cut-off database. The linking between the datasets obtained from the biorefinery model described in Section 3 is performed using an Excel file containing the information regarding the ecoinvent 3.6 cut-off datasets used (e.g. name of the exchange, location, unit). 4.2.3 LCA calculations The LCA calculations are performed according to the type of contribution analysis selected in the recipe. The environmental impacts can be calculated at three different levels: subsystems, processes, and exchanges. The basic structure for LCA calculation was used at each of the three levels. Subsystem level: The LCA scores are calculated for each subsystem to analyse the contribution of infrastructures, operation, S1, S2, S3, and transport to the overall environmental impacts of the biorefinery. Two types of graphs can be generated. The relative stacked bar plot shows the contribution of each subsystem to the overall environmental impacts of the biorefinery for as many impact categories as indicated in the recipe. The second graph consists of a bar plot that shows the impacts of different foreground database scenarios for a specific impact category. Process level: The LCA scores are calculated for each process (or activity) to evaluate the contribution of each activity to the environmental impacts of a specific subsystem (e.g. S1, S2, S3). Two types of graphs can be generated. The relative stacked bar plot shows the contribution of each activity (e.g. “S1.A1.Cultivation”, “S2.A1.Maceration”) to the environmental impacts of a specific subsystem for as many impact categories as indicated in the recipe. The second graph consists of a stacked bar plot that shows the impacts of different scenarios of a same subsystem (e.g. S1 in 2019 versus S1 in 2022) for a specific impact category. 17 SD2: Description of the SpiralG biorefinery model Exchange level: The LCA scores are calculated per category of exchanges across the subsystems (i.e. for the entire biorefinery). The categories include chemicals, energy, transport, equipment, construction, nutrient, packaging, water, and waste (see Table 7). The graph generated corresponds to a heat map which shows the relative contribution of each category of exchange to the impact categories selected. Table 7: Categories of inputs used to perform the contribution analysis per exchange. Exchange categories Exchanges included Chemicals Sodium hydroxide, hydrogen peroxide, phosphoric acid, sulfuric acid, potassium hydroxide, nitric acid, sodium hypochlorite Energy Electricity (French mix), electricity (Italian mix), natural gas, heat from anaerobic digestion Materials Galvanised steel, polypropylene pipes, sand, propylene pipes, polyvinylchloride cover, polypropylene random copolymer (PPR) pipes, polyethylene film, mosquito net made of polyethylene, polycarbonate walls and ceiling, concrete, EPS bricks, ceramic floor tiles, wall pumice bricks, wall concrete bricks, insulated panels made of polyurethane, polyisocyanurate, rock wool, solar shading net made of high density polyethylene, PEX pipes (HDPE), cellulose filters, nylon, food grade packaging (polyethylene). Nutrients Sodium bicarbonate (NaHCO3), carbon dioxide (CO2), chelated Iron (6%) (Fe EDDHA), TKPP (K4P2O7), potassium sulfate (K2SO4), magnesium sulfate (MgSO4), ammonium phosphate (NH4H2PO4), potassium nitrate (KNO3), ammonium phosphate (NH4H2PO4) Transport Transport by truck, transport by refrigerated truck, transport by ship, transport by refrigerated ship,transport by car Wastes Wastewater, waste polyethylene, plastic waste, paperboard waste, general waste Water Ground water, tap water, ultrapure water 4.3 Visualisation of the LCA results Specific attention was given to the visualisation of the LCA results. According to the level at which the contribution analysis was performed (e.g. subsystem, process, exchange), several types of graphs can be plotted (see Fig. 5). 18 SD2: Description of the SpiralG biorefinery model Dataset from the biorefinery model in the bw2package format Analyse the contribution of each subsystem to the overall environmental impacts of the biorefinery Analyse the contribution of each process to the overall environmental impacts of the biorefinery Perform the LCA calculations at subsystem level Perform the LCA calculations at process level Perform the LCA calculations at exchange level Analyse the contribution of each category of exchange to the overall environmental impacts of the biorefinery Figure 5: Three different types of visualisation according to the contribution analysis performed 19 SD2: Description of the SpiralG biorefinery model References [1] Luca Attene et al. “Efficient Nitrogen Recovery from Agro-Energy Effluents for Cyanobacteria Cultivation (Spirulina)”. In: Sustainability 15.1 (2023), p. 675. [2] Chris Mutel. “Brightway: an open source framework for life cycle assessment”. In: Journal of Open Source Software 2.12 (2017), p. 236. 20