https://orcid.org/0000-0003-0236-5970
Abstract
Wastewater treatment plants (WWTP) have extensive energy processes that undermine their economic and environmental performance. In this context, the integration of wastewater treatment with other biochemical processes such as co-digestion of sludge with organic wastes, and production of value-added products at their downstream processes will shift conventional WWTPs into biorefinery platforms with better sustainability performance. The sustainability of such a biorefinery platform has been investigated herein using an economic and life cycle assessment approach. This WWTP-based biorefinery treats wastewater from Copenhagen municipality, co-digests the source-sorted organic fraction of municipal solid waste and sludge, and upgrades biogas into biomethane using a hydrogen-assisted upgrading method. Apart from bioenergy, this biorefinery also produces microbial protein (MP) using recovered nutrients from WWTP's reject water. The net environmental savings achieved in two damage categories, i.e., −1.07 × 10−2 species.yr/FU in ecosystem quality and −1.68 × 106 USD/FU in resource scarcity damage categories along with high potential windows for the further environmental profile improvements make this biorefinery platform so encouraging. Despite being promising in terms of environmental performance, the high capital expenditure and low gross profit have undermined the economic performance of the proposed biorefinery. Technological improvements, process optimization, and encouraging incentives/subsidies are still needed to make this platform economically feasible.
HIGHLIGHTS
A biorefinery for microbial protein and bioenergy production has been developed.
Wastewater treatment and anaerobic digestion of sludge incorporated in biorefinery.
Microbial production had lower environmental profile than soybean meal.
High CAPEX undermined the economic feasibility of MP production.
ABBREVIATIONS
- LCA
life cycle assessment
- LCI
life cycle inventory
- CLCA
consequential life cycle assessment
- FU
functional unit
- WWTP
wastewater treatment plant
- RW
reject water
- NH4-N
ammonium nitrogen
- TAN
total ammonium nitrogen
- MOB
methane-oxidizing bacteria
- AD
anaerobic digestion
- CSTR
continuously stirred tank reactor
- HRT
hydraulic retention time
- MP
microbial protein
- dAMS
diluted ammonium medium salt
- N
nitrogen
- P
phosphorus
- BES
bio-electrochemical system
- MFC
microbial fuel cells
- CH4
methane
- CO2
carbon dioxide
- N2O
nitrous oxide
- O2
oxygen
- DALY
disability-adjusted life year
- USD
United States dollar
- LHV
low heat value
- NPV
net present value
- ROI
return on investment
- M€
million €
- IRR
internal rate of return
- CAPEX
capital expenditure
- OPEX
operational expenditure
- FOPEX
fixed operational expenditure
- VOPEX
variable operational expenditure
- GWP
global warming potentials
INTRODUCTION
Extensive energy demand (i.e., electricity and heat) is an inevitable challenge that wastewater treatment plants (WWTPs) are facing (Tian et al. 2020). Energy from external sources is supplied to meet WWTPs’ demand, which brings about serious challenges among them undermining environmental and economic performance (Gu et al. 2017). Most of the large WWTPs worldwide are now supplemented with anaerobic digesters to supply their internal energy demand by biogas production via anaerobic digestion (AD) of sewage sludge. However, AD cannot always compensate for the high energy demand in WWTPs. In some countries such as Denmark, utility companies are front runners and have worked intensely to decrease energy consumption optimize overall processes. Their attempts have resulted in a positive energy balance (150% or more) (WATER VISION DENMARK 2021; Energy & Resources 2022). However, to produce even more energy and contribute to the future circular bioeconomy, the AD of other organic wastes from external sources is in focus. Despite the great potential of integrating the wastewater treatment process and AD to enhance the energy flow within WWTPs, the low biodegradability of sludge results in lower energy recovery compared with agro-industrial biogas plants fed with biowaste with high biogas yield. For instance, the organic fraction of municipal solid waste (OFMSW) has a biochemical methane yield of 474 mLCH4/gVS (Hanum et al. 2019; Zhang & Li 2019) while that of sludge is 376 mLCH4/gVS (Khoshnevisan et al. 2018a; Zha et al. 2020). In this context, the anaerobic co-digestion of sludge and OFMSW is an attractive option to deal with the above-mentioned problems (Zha et al. 2020). The co-digestion of sludge and OFMSW has shown synergistic effects leading to increased biomethane potential of the feedstocks (Zha et al. 2020). It is worth mentioning that there are considerable debates in the literature on full-scale anaerobic co-digestion, the relative importance of performance conditions, cost-effectiveness, and technical requirements (Cavinato et al. 2013; Wehner et al. 2021). Regarding energy recovery, environmental, and financial benefits, full-scale anaerobic co-digestion had very encouraging results (Nghiem et al. 2017), even more than doubling energy recovery (Koch et al. 2016). Most of the published literature reached similar results when other wastes were co-digested with sludge in a full-scale application (Park et al. 2011; Di Maria et al. 2015; Wong et al. 2018).
While the co-digestion of sludge and OFMSW increases the energy flows within WWTPs, the reject water (RW), i.e., the liquid fraction of digestate also contains considerable amounts of nutrients, more specifically nitrogen (N) and phosphorous (P), which can be further valorized into value-added products (Tsapekos et al. 2021). Overall, the nutrients in the RW are supplied from two sources: wastewater and biopulp. The source of N and P in wastewater is greywater (i.e., any water that is consumed at home, including washing, bathing, and kitchen, except toilet water), urine, and faeces (Vinnerås et al. 2006; Morales Pereira 2014). Also, biopulp is mainly composed of source-separated organic household waste, green waste, superstore waste, and leftover food from restaurants and cafeterias (Alvarado-Morales et al. 2021), which make up the rest of the nutrients in the RW. The RW after the dewatering process reportedly contains a high amount of recoverable ammonium nitrogen, i.e., ∼15–20% of the input flow (Horan 1994; Gustavsson et al. 2010), the removal of which plays a key role in making WWTPs environmentally friendly (Guo et al. 2010). The co-digestion of sludge with OFMSW increases the nutrient content of RW compared to that of conventional WWTPs as OFMSW is rich in nutrients (Khoshnevisan et al. 2018b). Therefore, the valorization of nutrient-rich RW into value-added products can contribute to circular bioeconomy and play the same role as a non-waste-derived primary source without environmental damage. In this way we can apply the end-of-waste status to RW and cease to be waste if it meets the end-of-waste criteria in accordance with the Waste Framework Directive (Delgado et al. 2009). Achieving end-of-waste criteria for waste materials can provide potential environmental and economic benefits (Zorpas 2016).
The valorization of N-rich effluents into microbial protein (MP), i.e., a protein-rich biomass that can be used as supplementary animal feed, has provoked considerable attention (Matassa et al. 2016b; Tsapekos et al. 2019; Khoshnevisan et al. 2020a). The growth of the world's population and increased living standards have raised concerns over food insecurity and an imbalanced diet. To meet the high demand for stable food sources, proteinaceous food from vegetal and animal sources has expanded rapidly over the past decades, leading to significant environmental damages (Herrero et al. 2016; Khoshnevisan et al. 2021). The environmental degradation caused by feed production has undermined the sustainability of animal protein sources (Du et al. 2018; Uwizeye et al. 2020). In this context, the production of land-independent MP, i.e., microorganisms which are rich in protein and amino acid (i.e., 10% to 80% based on culture media, growth conditions, and microorganism type (Jones et al. 2020)), can substitute conventional proteinaceous feed with high environmental impacts such as soybean meal and fishmeal (Matassa et al. 2016a; Tsapekos et al. 2019).
Methane-oxidizing bacteria (MOB), a bacterial-based MP, have shown great potential to assimilate ammonium-nitrogen (NH4-N) using methane as their energy/carbon source (Strong et al. 2015; Khoshnevisan et al. 2019). Initial attempts have led to the commercialization of 1st generation MOB-based MP such as UniProtein and FeedKind, which are used as a protein feed supplement for livestock and aquaculture. Unibio® and Calysta® companies use synthetic nitrogen and natural gas as their N and C sources, respectively, which do not seem to be long-term sustainable solutions (Sillman et al. 2020; Verbeeck et al. 2020). This challenge has directed investigators towards the production of 2nd generation MP by using waste effluents as N source and biomethane as C source (Matassa et al. 2015a, 2015b). Previous studies have demonstrated that the upcycling of N-rich waste effluent to high-quality MP would decrease the induced environmental impacts of conventional proteinaceous feed such as fishmeal and soybean meal (Khoshnevisan et al. 2020b; Matassa et al. 2020). Furthermore, biogas or biomethane (biologically upgraded biogas) have been shown as promising alternatives to natural gas for the production of MOB-based MP (Khoshnevisan et al. 2019; Acosta et al. 2020).
The previous studies have provided valuable insights on the possibility of valorizing different types of waste streams into MP such as municipal solid waste (Khoshnevisan et al. 2020b), pig manure (Acosta et al. 2020; Verbeeck et al. 2021), urine (Khoshnevisan et al. 2020a), household waste (Tsapekos et al. 2020), and sewage sludge (Zha et al. 2021). The literature review showed that the potential of WWTP-based biorefinery platforms for MP and bioenergy production by the treatment of municipal wastewater and OFMSW has not been scrutinized yet. A recent study was conducted by Marami et al. (2022) to investigate various pathways for nutrients extraction from wastewater and deploying methane in different forms (i.e., biogas and upgraded biomethane) for MP production. They found that the lack of sufficient nutrients in wastewater would hinder the sustainability of integrated WWTPs and MP production facilities. Accordingly, the main objective of this study was to investigate the environmental and economic performance of a biorefinery platform that produces MP and bioenergy. This platform is established on a WWTP to valorize N and carbon flows from wastewater treatment and anaerobic co-digestion of sludge as well as OFMSW. To do so, a real WWTP in Copenhagen, Denmark, was selected as our case study to model N, C, and energy flows within the WWTP and implement life cycle assessment (LCA), economic assessment (net present value (NPV), gross profit, return on investment (ROI), and internal rate of return (IRR)) to evaluate the sustainability of the proposed biorefinery.
MATERIALS AND METHOD
LCA
In the following sub-sections, the detailed description of how LCA is to be implemented will be presented in terms of goal and scope of the study, life cycle inventory (LCI), life cycle impact assessment (LCIA), and interpretation of results. Guidelines and recommendations made by ISO 14040 (Finkbeiner et al. 2006) and International Reference Life Cycle Data System (ILCD) handbook (JRC 2010) were followed to correctly implement this LCA study.
Goal and scope definition
The main objective of this study was to assess the environmental impacts of a biorefinery platform established on a WWTP aimed at valorizing N and carbon flows into an animal grade proteinaceous feed supplement (known as microbial protein) and bioenergy. To do so, Avedøre municipal WWTP, which is one of the WWTP facilities managed by BIOFOS company (https://biofos.dk/), located in the south of Copenhagen (56.2639° N, 9.5018° E), Denmark, was selected as our case study (plant description is shown in SI-1). Hence, the geographical scope of this study was Denmark and the temporal scope was 2030. The production of 2nd generation MP using N- and carbon-rich waste streams relies on various technologies some of which are still at pilot scale demonstrations and have not been commercialized yet. Hence, scaled-up MP production and the penetration of the final product into the animal feed market as a substitution for conventional protein feeds such as soybean meal need a minimum time frame of 10 years. Because the mass-production of MP and its use as supplementary animal feed may affect the current animal feed market and will have consequences outside the system boundary of this study (it will be later discussed in section 2.1.2.6), consequential LCA was applied herein to include the above-mentioned impacts within the system boundary opted for this study. The multifunctionality problem was dealt with system expansion with substitution as recommended by ISO guidelines (Finkbeiner et al. 2006).
A cradle to gate system boundary was used covering all the processes within the WWTP, nutrient recovery, and MP production facility as well as electrolytic hydrogen production using off-peak surplus electricity needed for biological biogas upgrading. The system boundary of this study along with the mass and energy balances is shown in Figure 1 (see details in section 2.1.2). Because this biorefinery is proposed on a real WWTP as our case study, the inlet raw sewage stream to the Avedøre WWTP in 2030, i.e., 27.3 million m3/year, was used as the functional unit (FU). However, a second FU, i.e., 1 metric ton of finished pure protein, was also opted later to allow comparing the environmental impacts of 2nd generation MP with other proteinaceous feed sources.
Process flow diagram, mass flow, and energy flow through the WWTP. Marginal products/ energies and final products are shown with dashed boxes and yellow boxes, respectively. Numbers 1, 2, and 3 represent the soybean-palm oil-barley loops (see details in subsection 2.1.2.6).
Process flow diagram, mass flow, and energy flow through the WWTP. Marginal products/ energies and final products are shown with dashed boxes and yellow boxes, respectively. Numbers 1, 2, and 3 represent the soybean-palm oil-barley loops (see details in subsection 2.1.2.6).
LCI
The data used for LCA modeling were collected from various sources: (1) data needed to perform material and energy flows within WWTP were obtained from Avedøre WWTP as well as HCS company supplying biopulp (i.e., source sorted OFMSW pretreated by biopulp technology, more details in the supplementary file ‘SI-2’); (2) data related to MP production and biological biogas upgrading were collected from authors’ earlier studies conducted in this context; (3) background data related to the production of auxiliaries, energy carriers, transportation, etc. were obtained from ecoinvent database (version V3.4); and finally (4) other required information that could not be gathered from the above-mentioned sources were obtained through literature review, technical reports, and commercialized protein suppliers. A summary of all inventory data is shown in Table 1 and detailed information is provided in the following sections.
Overview of life cycle inventory data
| Unit process | Parameter | Amount | Unit | Ref. | Description |
|---|---|---|---|---|---|
| WWTP inventory data | |||||
| Preliminary sedimentation | Power consumption | 741 | MWhel | Provided by WWTP | Energy intensity for pumping: 0.027 kWh/m3 WW |
| N2O diffusion | 139 | kg | Sun et al. (2015); Bao et al. (2016); Hwang et al. (2016) | 0.002–0.005 g N2O /m3 WW | |
| CH4 diffusion | 919 | kg | Liu et al. (2014); Masuda et al. (2015) | 0.034 g CH4/m3 WW | |
| Aeration tank | Power consumption | 3,972 | MWhel | Provided by WWTP | 0.145 kWh/m3 WW |
| N2O diffusion | 164 | t | Hwang et al. (2016); Nguyen et al. (2019) | 1.5 g N2O /m3 WW | |
| CH4 diffusiona | 39 | t | Liu et al. (2014); Bao et al. (2016); Nguyen et al. (2019) | 1.425 g CH4/m3 WW | |
| CO2 diffusiona | 9,609 | t | Bao et al. (2016); Nguyen et al. (2019) | 350.82 g CO2/m3 WW | |
| Clarifying tank | Power consumption | 1,522 | MWhel | Provided by WWTP | Based on energy balance 2016 |
| N2O diffusion | 37 | t | Kyung et al. (2015) | 0.027 kg N2O/kg TN | |
| Pumping station | Power consumption | 759 | MWhel | Provided by WWTP | 0.028 kWh/m3 WW |
| N2 diffusion | 94 | t | – | Emission to water based on mass balance 2016. | |
| Final sedimentation | Power consumption | 740 | MWhel | Provided by WWTP | Based on energy balance 2016. |
| N2O diffusion | 21 | kg | Sun et al. (2015); Bao et al. (2016); Hwang et al. (2016) | 0.002–0.005 g N2O /m3 WW | |
| CH4 diffusion | 10 | kg | Liu et al. (2014); Masuda et al. (2015) | 0.034 g CH4/m3 WW | |
| Anaerobic digestion | Power consumption | 2,366 | MWhel | Pöschl et al. (2010) | 3% of produced biogas |
| Heat consumed | 18,303 | MWhth | Based on energy balance 2016. | 39% of MWh produced CH4 | |
| CH4 diffusiona | 91 | t | Naroznova et al. (2016); Khoshnevisan et al. (2018b) | 2.7% of the total CH4 produced | |
| Water used | 33,805 | m3 | Khoshnevisan et al. (2020b) | Dilution of biopulp with a ratio of 1: 1 | |
| Sludge dewatering | Power consumption | 1,959 | MWhel | Calculated | Energy intensity for pumping and centrifuge: 0.027 and 3.55 kWh/m3 |
| Drying-combustion | Power consumption | 4,720 | MWhel | Provided by WWTP | Energy intensity for drying and combustion: 23.59 and 86.82 kWh/m3 |
| Heat consumed | 13,485 | MWhth | Provided by WWTP | Was supplied internally | |
| Biological biogas upgrading inventory data | |||||
| Biological biogas upgrading | Inlet biogas | 7.89 | Mm3 | – | – |
| Power consumption | 237 | MWhel | 10% consumed electricity in digester | ||
| Electrolysis | Water used | 11,348 | m3 | Kuckshinrichs et al. (2017) | 10.11 kg water/kgH2 |
| Power consumption | 62,859 | MWhel | Ivy (2004) | 56 kWhel/kgH2 | |
| MP production inventory data | |||||
| BES-P recovery | Na2HPO4 consumed | 0.43 | t | Xavier et al. (2014); Almatouq & Babatunde (2018) | Based on the stoichiometric relationships |
| Mg(OH)2 consumed | 74 | t | Xavier et al. (2014); Almatouq & Babatunde (2018) | Based on the stoichiometric relationships | |
| BES-TAN recovery | Na2HPO4 consumed | 0.43 | t | – | Calculated |
| Fermentation process | Power consumption | 495 | MWhel | Calculated | Air blower: 0.025 kWh/m3 |
| Chemical consumed | Table 3 | t | Khoshnevisan et al. (2018b), Unibio® | – | |
| Biomethane consumed | 1.97 | Mm3 | Provided by DTU | CH4/TAN: 9.4 | |
| Purification process | Power consumption | 419 | MWhel | Calculated | GNLW554EP-VFD: 0.83 kWh/m3 |
| Protein substitution | Soybean meal | 1,620 | t | Calculated | Soybean meal loop (Section 2.1.2.6) |
| Palm oil | −1,092 | t | Calculated | Soybean meal loop (Section 2.1.2.6) | |
| Barley grain | 16 | t | Calculated | Soybean meal loop (Section 2.1.2.6) | |
| Heat substitution | Biomass-wood chips/wood pellet | 7,617 | MWhth | Calculated | – |
| Oxygen substitution | Oxygen | 8,868 | t | Calculated | Based on stoichiometric relationships |
| Biomethane substitution | Natural gas | 5.91 | Mm3 | Calculated | – |
| Biopulp production inventory data | |||||
| Biopulp unit | Power consumption | 851 | MWhel | Khoshnevisan et al. (2018b) | 26.6 kWh/t biopulp |
| Water used | 12,846 | m3 | Khoshnevisan et al. (2018b) | 400 kg water/t biopulp | |
| Transportation inventory data | |||||
| Transportation | Chemical materials | 2,589 | tkm/year | Calculated | Distance: 5 km, it was supposed to be shipped from the markets of central Copenhagen. |
| Drying-combustion unit | 49,240 | tkm/year | Calculated | Distance: 20 km, it was supposed to be shipped from Lynetten plantb. | |
| Biopulp | 468,876 | tkm/year | Calculated | Distance: 15 km, it was supposed to be shipped from HCS company. | |
| Unit process | Parameter | Amount | Unit | Ref. | Description |
|---|---|---|---|---|---|
| WWTP inventory data | |||||
| Preliminary sedimentation | Power consumption | 741 | MWhel | Provided by WWTP | Energy intensity for pumping: 0.027 kWh/m3 WW |
| N2O diffusion | 139 | kg | Sun et al. (2015); Bao et al. (2016); Hwang et al. (2016) | 0.002–0.005 g N2O /m3 WW | |
| CH4 diffusion | 919 | kg | Liu et al. (2014); Masuda et al. (2015) | 0.034 g CH4/m3 WW | |
| Aeration tank | Power consumption | 3,972 | MWhel | Provided by WWTP | 0.145 kWh/m3 WW |
| N2O diffusion | 164 | t | Hwang et al. (2016); Nguyen et al. (2019) | 1.5 g N2O /m3 WW | |
| CH4 diffusiona | 39 | t | Liu et al. (2014); Bao et al. (2016); Nguyen et al. (2019) | 1.425 g CH4/m3 WW | |
| CO2 diffusiona | 9,609 | t | Bao et al. (2016); Nguyen et al. (2019) | 350.82 g CO2/m3 WW | |
| Clarifying tank | Power consumption | 1,522 | MWhel | Provided by WWTP | Based on energy balance 2016 |
| N2O diffusion | 37 | t | Kyung et al. (2015) | 0.027 kg N2O/kg TN | |
| Pumping station | Power consumption | 759 | MWhel | Provided by WWTP | 0.028 kWh/m3 WW |
| N2 diffusion | 94 | t | – | Emission to water based on mass balance 2016. | |
| Final sedimentation | Power consumption | 740 | MWhel | Provided by WWTP | Based on energy balance 2016. |
| N2O diffusion | 21 | kg | Sun et al. (2015); Bao et al. (2016); Hwang et al. (2016) | 0.002–0.005 g N2O /m3 WW | |
| CH4 diffusion | 10 | kg | Liu et al. (2014); Masuda et al. (2015) | 0.034 g CH4/m3 WW | |
| Anaerobic digestion | Power consumption | 2,366 | MWhel | Pöschl et al. (2010) | 3% of produced biogas |
| Heat consumed | 18,303 | MWhth | Based on energy balance 2016. | 39% of MWh produced CH4 | |
| CH4 diffusiona | 91 | t | Naroznova et al. (2016); Khoshnevisan et al. (2018b) | 2.7% of the total CH4 produced | |
| Water used | 33,805 | m3 | Khoshnevisan et al. (2020b) | Dilution of biopulp with a ratio of 1: 1 | |
| Sludge dewatering | Power consumption | 1,959 | MWhel | Calculated | Energy intensity for pumping and centrifuge: 0.027 and 3.55 kWh/m3 |
| Drying-combustion | Power consumption | 4,720 | MWhel | Provided by WWTP | Energy intensity for drying and combustion: 23.59 and 86.82 kWh/m3 |
| Heat consumed | 13,485 | MWhth | Provided by WWTP | Was supplied internally | |
| Biological biogas upgrading inventory data | |||||
| Biological biogas upgrading | Inlet biogas | 7.89 | Mm3 | – | – |
| Power consumption | 237 | MWhel | 10% consumed electricity in digester | ||
| Electrolysis | Water used | 11,348 | m3 | Kuckshinrichs et al. (2017) | 10.11 kg water/kgH2 |
| Power consumption | 62,859 | MWhel | Ivy (2004) | 56 kWhel/kgH2 | |
| MP production inventory data | |||||
| BES-P recovery | Na2HPO4 consumed | 0.43 | t | Xavier et al. (2014); Almatouq & Babatunde (2018) | Based on the stoichiometric relationships |
| Mg(OH)2 consumed | 74 | t | Xavier et al. (2014); Almatouq & Babatunde (2018) | Based on the stoichiometric relationships | |
| BES-TAN recovery | Na2HPO4 consumed | 0.43 | t | – | Calculated |
| Fermentation process | Power consumption | 495 | MWhel | Calculated | Air blower: 0.025 kWh/m3 |
| Chemical consumed | Table 3 | t | Khoshnevisan et al. (2018b), Unibio® | – | |
| Biomethane consumed | 1.97 | Mm3 | Provided by DTU | CH4/TAN: 9.4 | |
| Purification process | Power consumption | 419 | MWhel | Calculated | GNLW554EP-VFD: 0.83 kWh/m3 |
| Protein substitution | Soybean meal | 1,620 | t | Calculated | Soybean meal loop (Section 2.1.2.6) |
| Palm oil | −1,092 | t | Calculated | Soybean meal loop (Section 2.1.2.6) | |
| Barley grain | 16 | t | Calculated | Soybean meal loop (Section 2.1.2.6) | |
| Heat substitution | Biomass-wood chips/wood pellet | 7,617 | MWhth | Calculated | – |
| Oxygen substitution | Oxygen | 8,868 | t | Calculated | Based on stoichiometric relationships |
| Biomethane substitution | Natural gas | 5.91 | Mm3 | Calculated | – |
| Biopulp production inventory data | |||||
| Biopulp unit | Power consumption | 851 | MWhel | Khoshnevisan et al. (2018b) | 26.6 kWh/t biopulp |
| Water used | 12,846 | m3 | Khoshnevisan et al. (2018b) | 400 kg water/t biopulp | |
| Transportation inventory data | |||||
| Transportation | Chemical materials | 2,589 | tkm/year | Calculated | Distance: 5 km, it was supposed to be shipped from the markets of central Copenhagen. |
| Drying-combustion unit | 49,240 | tkm/year | Calculated | Distance: 20 km, it was supposed to be shipped from Lynetten plantb. | |
| Biopulp | 468,876 | tkm/year | Calculated | Distance: 15 km, it was supposed to be shipped from HCS company. | |
Mm3, Million m3.
aBiogenic.
bDenmark (55.696̊ N, 12.637̊ E).
Substrate
The RW from Avedøre WWTP (i.e., liquid fraction from dewatering process of digested sludge, Figure 1) and biopulp collected from HCS Company were considered as the primary N and P sources for MP production in this study. The water and electricity demands for the biopulping process, producing a biopulp with a TS of 12–16%, reportedly are 400 kg water/t biopulp and 26.6 kWhel/t biopulp, respectively (Khoshnevisan et al. 2018b).
The WWTP model was created based on the annual average data and the projected data for 2030. The comprehensive characteristics of biopulp have been already detailed in the earlier study of Khoshnevisan et al. (2018b) and the key components are shown in Table 2. The amount of biopulp that is co-digested with sludge in AD reactors was determined based on the optimal volumetric ratio of sludge/biopulp, i.e., 60:40 (VS), which creates the synergistic effect and maximizes the biomethane potential (Zha et al. 2020). Previous studies have reported a methane yield of 428 mLCH4/gVS, for mono-digestion of biopulp under continuous mode (Khoshnevisan et al. 2020b). The data from our case study demonstrated an average methane yield of 265 mLCH4/gVSsludge under continuous mode. Zha et al. (2020) achieved a methane yield of up to 430 mLCH4/gVS under the co-digestion of biopulp and sewage sludge at the above-mentioned volumetric ratio. The characteristics of various streams within Avedøre WWTP are shown in Table 2.
Characteristics of the feedstock used in this study
| Characteristics | Raw sewage | Inlet sludge to AD | Reject water | Biopulp | Unit |
|---|---|---|---|---|---|
| Flow | 27,295,120 | 325,875 | 380,997 | 33,805 | m3/year |
| COD | 16,205 | 10,485 | 247 | 8,381 | t/year |
| BOD | 6,441 | – | 267 | – | t/year |
| SS | 8,250 | 8,817 | 149 | – | t/year |
| TN | 1,300 | 315 | 38 | 158 | t/year |
| TP | 193 | 218 | 380,997 | 17 | t/year |
| Characteristics | Raw sewage | Inlet sludge to AD | Reject water | Biopulp | Unit |
|---|---|---|---|---|---|
| Flow | 27,295,120 | 325,875 | 380,997 | 33,805 | m3/year |
| COD | 16,205 | 10,485 | 247 | 8,381 | t/year |
| BOD | 6,441 | – | 267 | – | t/year |
| SS | 8,250 | 8,817 | 149 | – | t/year |
| TN | 1,300 | 315 | 38 | 158 | t/year |
| TP | 193 | 218 | 380,997 | 17 | t/year |
Nutrient recovery
, phosphate
, and magnesium (Mg2+) for struvite settlement. In this regard, pumping magnesium hydroxide solution i.e., Mg(OH2), to the cathode compartment as a source of Mg2+ would lead to a 95% P precipitation efficiency (Xavier et al. 2014; Almatouq & Babatunde 2018). It is worth mentioning that in both NH4-N and P recovery, Na2HPO4 was used for pH adjustment in the cathode chamber. It should be pointed out that the nutrients needed for the growth of active microorganisms in the anodic chamber were supplied by the trace elements available in biopulp.
Biological upgrading of biogas
The number of studies on methanation strategies for biogas upgrading has been increasing. Elyasi et al. (2021) studied the sustainability of the biological biogas upgrading using off-peak surplus electricity in Denmark. Their findings confirmed that from economic and environmental points of view, biological upgrading of biogas can be a good competitor to conventional water scrubbing technology. During the high penetration of renewable electricity into the Danish power system, MOB-based MP production integrated with biological biogas upgrading as a downstream process has been introduced as a sustainable pathway (Khoshnevisan et al. 2022). Accordingly, the use of biological biogas upgrading as a sub-category of power-to-gas technology is considered in this study.
The implementation of biological biogas upgrading improves the efficiency of WWTPs by converting the CO2 content of biogas to biomethane by the addition of hydrogen (i.e., 
). Taking the thermodynamic relationships and mass balance into account, 0.18 kg of hydrogen is required to upgrade 1 kg CO2 to biomethane (Mehmeti et al. 2018). To produce 1 kg of hydrogen, 10.11 kg H2O water is consumed (Kuckshinrichs et al. 2017). Considering the stoichiometric relationships, the amount of water consumed by the electrolysis unit and subsequently the amounts of H2 and O2 generated were estimated. Therefore, the amount of electricity consumption was calculated at 56 kWh/kg H2 (Ivy 2004). Accordingly, in the biological biogas upgrading facility, standing at upstream of the MP production facility, 1,122 t/year hydrogen was required to react with 6,175 t/year CO2 to generate 5,626 t/year biomethane. Furthermore, 62,859 MWh of electricity was annually consumed by electrolyzers to break down water into H2 and O2, which was supplied by excess wind power. About 25% of the total upgraded biomethane was consumed for MP fermentation, i.e., 1,974 million m3/year. The remaining biomethane (i.e., 5,913 million m3/year), after supplying the requirements of the fermentation unit, was assumed to be injected into the natural gas grid.
It should be pointed out that excess wind power is assumed to be utilized by electrolyzers to produce hydrogen for biological biogas upgrading. Surplus power refers to the amount of electricity produced in excess and has no market demand leading to wind turbines curtailment and zero/negative electricity price in the Nord Pool electricity market (Elyasi et al. 2021). The curtailment of wind turbines increases the environmental impacts of wind electricity production (caused by the manufacturing of fixed and moving parts of wind turbines) within their life cycle. On the contrary, avoiding curtailment by using surplus electricity for other purposes would decrease the impact per kWh electricity that is produced within the life cycle of wind turbines. On the other hand the surplus electricity can be further valorized into high-added-value products or stored in the existing gas pipelines infrastructures.
MP fermentation
Mixed methanotrophic culture is used for the fermentation of the MP. The highest concentration of extracted TAN without any dilution was considered for the preparation of the culture medium, i.e., 297 mg/l. Moreover, industrial data has shown a water circulation rate of 80% in the 1st generation MP fermentation process (https://www.unibio.dk/end-product/protein/), which decreases the freshwater requirement. The carbon/energy source for methane oxidizing bacteria was supplied from biologically upgraded biogas as shown promising results for MP production (Khoshnevisan et al. 2019; Tsapekosa et al. 2019). The earlier studies showed that a CH4/TAN ratio of 9.4 would maximize MP yield, hence, a total of 1,408 t CH4/year is consumed within the fermentation process.
The pretreated sludge and biopulp were digested in a continuously stirred tank reactor (CSTR). The fermentation of MP took place at ambient temperature; hence, this stage did not have any heat demand. The power demand of the fermenters was estimated based on the amount of air/oxygen required to be pumped in the culture medium (i.e., the energy demand of pumps). The oxygen and methane ratio of 2:1 would provide an optimal growth condition for MP production (Khoshnevisan et al. 2019; Tsapekos et al. 2019). Accordingly, an air blower with a capacity of 43.5–134 m3/min and electricity consumption of 0.025 kWh/m3 (model: GM130 L/DN300) was considered to supply air/oxygen to the fermenters.
The output flow from BES lacks micro/macro nutrients required for the balanced growth of the methanotrophic culture. Diluted ammonium medium salt (dAMS) is the standard medium used as the growth medium for the cultivation of methanotrophs. As N and P are supplied from the RW, the original protocol is altered and the modified-dAMS is used for the fermentation of MP. The modified-dAMS were prepared according to the recommendations of previous studies (Khoshnevisan et al. 2019). This protocol has been modified based on the authors’ experiences and the instructions made by experts from Unibio®. Table 3 provides an overview of the chemicals needed for the preparation of modified-dAMS. As the cultivation medium has been patented by Unibio®, detailed information is not allowed to be shown herein.
Modified-dAMS instruction for MP production
| Chemicals | Amount (t/year) |
|---|---|
| NH4-N from RW | 149.82 |
| Potassium and P source | 417.1 |
| Magnesium source | 12.1 |
| Calcium source | 128.8 |
| Iron source | 5.5 |
| Trace metals | 0.39 |
| Chemicals | Amount (t/year) |
|---|---|
| NH4-N from RW | 149.82 |
| Potassium and P source | 417.1 |
| Magnesium source | 12.1 |
| Calcium source | 128.8 |
| Iron source | 5.5 |
| Trace metals | 0.39 |
System expansion and marginal technologies
Consequential LCA is characterized by using marginal data/technology instead of average technology. Weidema et al. (1999) have recommended a 5-step procedure to identify marginal technologies, which was employed herein. As the temporal scope of this study is 2030, offshore wind electricity was used as marginal technology. The choice of offshore wind electricity as marginal technology was in line with projections made by Denmark's Energy and Climate Outlook for 2030 (Danish Energy Agency 2019). Regarding the marginal source of heat, the renewable energy sources in central heating installations have the highest proportion of Denmark's heat supply and are expected to grow 80% by 2030 (Danish Energy Agency 2019). Biomass, with a 10% increase in annual consumption, is projected to be the best alternative to the most common district heating sources in Denmark, such as natural gas, heat pumps, and coal (Danish Energy Agency 2019). Accordingly, biomass-wood chip and biomass-wood pellet were chosen as marginal qualified sources for the centralized heating system. Regarding the biomethane produced, natural gas was considered to be substituted, which is in line with the Danish energy outlook as shown in the recent studies (Elyasi et al. 2021; Faragò et al. 2021). Considering conventional protein sources for animal feeds, the most preferred proteinaceous feeds that are imported to Denmark are rapeseed, sunflower, and soybean meal. Among them, soybean meal accounts for up to 64% of the overall imported protein-based feeds (Gylling & Hermansen 2018; Index Mundi data 2020); mostly from Germany, Argentina, and Brazil (OEC 2018). The statistical analysis demonstrated that the demand for soybean/soybean meal in Denmark will rise in the next decade (KNOEMA 2016; Index Mundi data 2020); hence, soybean meal was chosen as marginal technology herein.
Soybean meal loop
As mentioned earlier, it is expected that 2nd generation MP substitutes for soybean meal in the feed market. However, this substation will have important consequences on the soybean oil market as a decrease in soybean meal production is accompanied by a drop in soybean production, which should be compensated by palm oil production (Khoshnevisan et al. 2020b). Accordingly, any changes in soybean demand will have subsequent changes in demand/supply of other products, i.e., soybean oil, palm oil, palm kernel meal, and barley. For a detailed description please read Dalgaard et al. (2007). So, a continuous chain of demands/supplies (loop) will be created, due to the replacement of soybean meal by MP and consequently soybean oil by palm oil (Schmidt & Weidema 2008), which must be addressed within the framework of this study. The accumulation of protein and energy of each product was considered, as a major criterion, to determine the amounts of marginal feed substituted. The soybean meal loop has been shown in Figure 1 using a ‘loop’ sign, which implies the quantity of avoided soybean meal and spring barley as well as the stimulated palm oil (Dalgaard et al. 2007).
LCIA
The impact assessment method, ReCiPe 2016 v1.1 (hierarchies perspective) (Huijbregts et al. 2017) at the endpoint and midpoint levels were used to investigate the environmental impacts of the developed biorefinery. This impact assessment method has 18 midpoint impact categories (i.e., global warming, ionizing radiation, water consumption, land use, ozone formation (terrestrial ecosystems), ozone formation (human health), stratospheric ozone depletion, human non-carcinogenic toxicity, human carcinogenic toxicity, fine particulate matter formation, freshwater eutrophication, freshwater ecotoxicity, marine ecotoxicity, terrestrial ecotoxicity, mineral and fossil resource scarcity, terrestrial acidification, and marine eutrophication) and three endpoint areas of protection (i.e., human health, resource availability, and ecosystem quality).
Economic assessment
An overview of all the steps, included in the economic assessment of the proposed biorefinery is demonstrated in Figure 2. The sum of capital expenditure (CAPEX) and operating expenditure (OPEX) has been made up of the total cost of the project. In this respect, operating expenditure OPEX can be classified into fixed OPEX and variable OPEX, both of which are included in our economic assessment.
Economic assessment steps in current work.
The CAPEX of biological biogas upgrading facility and its associated technologies including electrolyzers, hydrogen storage, and biogas upgrading was taken from the recent study of Elyasi et al. (2021). FOPEX covers the running expenditure of employees, insurance, tax, maintenances, and depreciation. VOPEX is also attributed to the expenses of feedstock materials, power, chemicals, and water consumed.
The electricity used in the electrolyzer facility for hydrogen production is supplied from off-peak surplus wind electricity. According to Denmark's Energy Outlook (Danish Energy Agency 2019), Denmark will become an electricity exporter by 2030 due to the expansion of offshore wind turbines, which is accompanied by surplus wind electricity production (See more details in Elyasi et al. 2021). Accordingly, the price of surplus electricity was assumed to be zero for the economic assessment. However, the impact of various prices of surplus electricity on the economic performance of MP production was later investigated by the sensitivity analysis. In this context, the effect of increasing the surplus electricity price up to 50% of the conventional electricity price in Denmark was used for the sensitivity analysis. It should be pointed out that the Danish Energy Agency has projected that the electricity prices will gradually decrease, reaching 44.72 €/MWh in 2030 (Danish Energy Agency 2019).
The transportation costs were estimated based on the distance between the market and biorefinery and the amount of fuel consumed for such transports. The chemicals used in the cultivation medium are assumed to be transported from Copenhagen to the Avedøre plant with an average distance of 5 km. The biopulp is shipped from a biopulping facility standing at a distance of 15 km. According to Hagos & Ahlgren (2020), diesel will be still the major transportation fuel for heavy-duty vehicles in 2030. Accordingly, the price of diesel fuel in 2030 was predicted based on annual data published in three conditions: 1.89, 1.93, and 1.19 €/L for the base, pessimistic, andoptimistic conditions, respectively (Statista Data Platform) (see the supplementary file ‘SI-4’).
The price of MP was considered to be equal to its counterpart, i.e., soybean meal, in 2030 to make it competitive with conventional animal feed ingredients. Through the analysis of historical data between 2000 and 2020 (IndexMundi data portal), the sale prices of soybean meal for 2030 under various scenarios were projected; 502, 541, and 323 €/t for the base, optimistic, and pessimistic conditions,respectively. Oxygen is the by-product of hydrogen in the electrolyzer facility, which should be accounted for in the economic assessment. The price of electricity in the reference year was assumed to be 0.082 €/kg O2 (Chen et al. 2018) with a 6% compound annual growth rate (CAGR) from 2020 to 2030 (Elyasi et al. 2021). Accordingly, themarket price for oxygen was calculated to be 0.17 €/kg O2 in 2030. The economic performance of the MP production facility was evaluated using gross profit, internal rate of return (IRR), net present value (NPV), and return on investment (ROI) (more details in the supplementary file ‘SI-6’).
RESULT AND DISCUSSION
Environmental performance of developed biorefinery
The developed biorefinery showed promising results in two damage categories of ecosystem and resources scarcity where net negative impacts were attained, i.e., −1.07 × 10−2 species.yr/FU, and −1.68 × 106 USD/FU, respectively. However, further improvements are still needed in the damage category of human health where a net impact of 2.37 DALY/FU was achieved. The results achieved herein demonstrated that a biorefinery platform developed on a municipal WWTP for the processing of WW and OFMS can be encouraging for closing the gaps in animal protein feed and supplying bioenergy. To identify the key hotspots in each damage category and potential possibilities for further improvements, the results of each damage category are discussed in detail separately in the following sub-sections.
Human health damage category
The contribution of various processes to the damage category of human health is shown in Figure 3. In this damage category, the unit process of anaerobic digestion had the highest contribution with a share of 41%. This impact attributes to methane leakage and the background emissions of heat and electricity production. Following the anaerobic digestion unit process, electrolysis and fermentation processes imposed the highest impacts on this damage category with a share of 27 and 15%, respectively. The impact from the electrolysis unit is caused by the consumption of electricity. Although wind electricity, which is known as an environmentally friendly and renewable energy source, is used for water electrolysis, it imposed a high environmental burden on the human health damage category. This attributes to the fact that the manufacture of wind turbines (both fixed and moving parts), as well as the maintenance operations, necessitate the use of materials such as concretes, steels, and glass-fibers (Ghenai 2012), which create such negative impacts within their life cycle from extraction to use phase. Delpierre (2019) already reported that wind electricity contributed most to the environmental impacts of electrolytic hydrogen production. In another study conducted by Elyasi et al. (2021), the same conclusion was also reached, demonstrating that wind electricity production contributes to the damage category of human health when surplus wind electricity is used for the biological upgrading of biogas.
Contribution of all unit processes to damage category of human health resulting from integrating WWTP (with sludge and biopulp co-digestion) and MP production.
Contribution of all unit processes to damage category of human health resulting from integrating WWTP (with sludge and biopulp co-digestion) and MP production.
The impacts from the fermentation process attributed to the electricity consumption (i.e., 1.4% of total imposed impacts) as well as the consumption of chemicals used for culture medium preparation. The background emissions from the production of such chemicals as well as their transportation are all shown under the fermentation process. As discussed earlier, the outflows of BESs do not contain any supplementary nutrients, needed for the growth of methanotrophic bacteria, except NH4-N and P. Thereby, a balanced enriched nutrient culture medium is provided by the addition of various chemicals to prepare modified-dAMS. The detailed analysis demonstrated that P/potassium sources, used as the buffering agents, were the major contributors to the environmental impact of the fermentation process, with over 70% of total induced impacts. As biopulp is not rich in P (2.70 g /kg TS biopulp), the co-digestion of a P rich waste stream such as fruit-vegetable waste, livestock manure, and agricultural waste with sludge and biopulp and the subsequent extraction of struvite P can decrease demand for P sources and hence will improve the environmental profile of fermentation process. Moreover, optimizing the fermentation process and recycling the culture medium to ensure the complete consumption of nutrients can further improve the environmental performance in this damage category. Other processes such as aeration tank, also contributed to the human health damage category but to a letter extent, i.e., 1.54 DALY/FU. These impacts are rooted in on-site emissions from biological wastewater treatment.
As shown in Figure 3, the avoided impacts caused by the substitution of soybean meal led to the highest saving in the human health damage category. The substitution of soybean meal has a net impact of −7.22 DALY/FU, contributing to 46% of total net savings. Following that, the substitution of biomethane for natural gas and oxygen as a co-product of the electrolysis process, led to the highest savings in this damage category with a net impact of −5.30 and −2.87 DALY/FU, respectively. Although considerable amounts of environmental impacts in human health damage category could be avoided by the substituted products, it was not enough to reach a net negative impact in this damage category. Optimizing the TAN extraction process and enhancing the recovery efficiency of BES (above 61%) would be one of the possibilities to improve the environmental performance of the biorefinery. The higher TAN recovery, the higher MP production, and hence the higher substitution of conventional proteinaceous feed. Moreover, improving N-to-protein conversion rate and the optimization of MP fermentation will also bring about more savings in this damage category. Furthermore, it is highly suggested to investigate the possibilities of storage and use of the electricity which is generated during NH4-N and P recovery in BES. The use of BES-driven electricity can contribute to the better environmental and economic performance of MP-based biorefinery platforms.
Ecosystem quality damage category
On the contrary to the damage category of human health, a net saving of −0.011 species.yr/FU was attained in ecosystem quality (Figure 4). Contribution analysis demonstrated that the anaerobic digestion and fermentation process had the greatest impacts, i.e., 0.047 and 0.016 species.yr/FU, contributing to 60 and 20% of total net impacts, respectively. The detailed contributional analysis demonstrated that heat consumption was the major contributor to the environmental impact of anaerobic digestion, i.e., about 80% of the total impact. As discussed in the previous sub-section, the impacts associated with the fermentation process originate from the consumption of chemicals and electricity demand. Faragò et al. (2021) reported that the utilization of electricity and chemicals for P recovery from sludge ash contributed to the impact category of water eutrophication and subsequently to the ecosystem quality damage category. Following AD and fermentation, the electrolysis process imposed the highest impact on this damage category with a share of 9% (i.e., 0.007 species.yr/FU).
Contribution of all unit processes to damage category of ecosystem quality resulting from integrating WWTP (with sludge and biopulp co-digestion) and MP production.
Contribution of all unit processes to damage category of ecosystem quality resulting from integrating WWTP (with sludge and biopulp co-digestion) and MP production.
Resource scarcity damage category
Similar to the ecosystem quality damage category, a net saving of −1.68 × 106 USD/FU was also attained in the resource scarcity damage category. In other words, the saving achieved by the avoided products under this damage category was 7.18 times as much as the impacts imposed by the unit processes. According to the results shown in Figure 5, electrolysis and fermentation unit processes dominated this damage category with a share of 32 and 28%, respectively. The impact from electrolysis unit process attributes to the consumption of materials used for the manufacture of fixed and moving parts of offshore wind turbines. The impact from the fermentation process attributes to the background emissions induced by the production of chemicals and minerals needed for the preparation of modified-dAMS. As can be observed in Figure 5, the substitution of natural gas (the marginal source of biomethane production in 2030 in agreement with Elyasi et al. (2021) and Faragò et al. (2021)) caused the highest saving in the resource scarcity damage category, accounting for 93% of the total saving. The co-digestion of biopulp and sludge results in higher methane yield (Zha et al. 2020). Moreover, the biological biogas upgrading increases the total biomethane production of the facility. The whole biorefinery's heat and electricity demand are from biomass (wood chips and pellet) and surplus wind power, respectively, while a high portion of the biomethane (approximately 75%) can be injected into the gas grid and substitute the marginal biomethane sources. Accordingly, the substitution of natural gas had higher net savings in resource scarcity than soybean meal substitution.
Contribution of all unit processes to damage category of resources resulting from integrating WWTP (with sludge and biopulp co-digestion) and MP production.
Contribution of all unit processes to damage category of resources resulting from integrating WWTP (with sludge and biopulp co-digestion) and MP production.
Environmental cost of MP production
The environmental impact of 1 tonne pure protein from 2nd generation MP was calculated at 0.007 DALY in human health, 6.75 × 10−5 species.yr in ecosystem quality, −711 kg CO2 in climate change, and −2,555 USD in resources damage categories. Thecommercialized methanotrophs-based MPs such as FeedKind® use natural gas and chemical nitrogen as their energy and nitrogen sources, respectively. According to Cumberlege et al. (2011), even with 100% use of biogas and renewables, FeedKind® MP, could not outperform the conventional protein sources such as soybean meal. The average global soybean meal and fishmeal production are considered herein to have an insight into the environmental profile of 2nd generation MP compared to conventional protein sources (www.ecoinvent.org). The production of 1 tonne of protein from soybean meal and fishmeal has a net impact of 0.009 and 0.011 DALY in human health, 7.06 × 103 and 3.00 × 103 kg CO2 in climate change, 9.84 × 10−5 and 1.51 × 10−5 species.yr in ecosystem quality, and 1.17 × 102 and 9.47 × 101 USD in resources damage categories, respectively. As can be seen, the production of MP from waste effluents is a promising approach to produce more environmentally friendly proteinaceous feed in the future.
Sensitivity analysis for LCA
Sensitivity analysis was performed to investigate to what extent the environmental profile of the WWTP-based biorefinery will change with respect to the assumptions made herein and the improvements which can be implemented. As discussed earlier, the use of chemicals in modified-dAMS solution as supplementary nutrients plays a key role in the overall environmental performance of the biorefinery. Moreover, improving the protein profile of MP and increasing its total protein content (by enhancing conversion rate from N to protein) can substitute more soybean meal as a competing product, hence, will help to improve the environmental performance. The results showed that (Table 4) the environmental profile of the biorefinery had the highest sensitivity to improving the conversion rate of N to protein. According to the lab experiments, 1 g of NH4-N could be valorized into 4.36 g of protein. Theoretically, proteins contain 16% N (Shuuluka et al. 2013; Krul 2019). Hence, improving the protein profile of the produced biomass can increase the protein yield from a given amount of nitrogen. The enhancement of NH4-N to protein conversion rate from 4.36 to 5.25 and optimization of the fermentation process and recycling of culture medium could to some extent decrease the impacts in the human health damage category. However, such a reduction in the environmental profile was not sufficient to achieve a net saving in this damage category. Considering both strategies, i.e., a 20% reduction in consumption of dAMS and enhancing pure protein content (Table 4; Sens. All), could decrease the net impact by 86% reaching 0.34 DALY/FU in human health.
Sensitivity analysis over the reduction of dAMS solution by 5% (Sens. 5%), 10% (Sens. 10%), 20% (Sens. 20%); improve protein yield of NH4-N (Sens. Pure protein); both a 20% reduction in dAMS and improve pure protein content (Sens. All)
| Parameter | Human health | Ecosystem | Resources | |||
|---|---|---|---|---|---|---|
| Net saving/impact (DALY) | Improve (%) | Net saving/impact (Species.yr) | Improve (%) | Net saving/impact (USD) | Improve (%) | |
| Base | 2.366 | – | −1.07 × 10−2 | – | −1.679 × 106 | – |
| Sen.5% | 2.204 | 6.85 | −1.16 × 10−2 | 7.64 | −1.683 × 106 | 0.26 |
| Sen.10% | 2.072 | 12.44 | −1.23 × 10−2 | 15.00 | −1.687 × 106 | 0.48 |
| Sen.20% | 1.803 | 23.78 | −1.39 × 10−2 | 29.77 | −1.695 × 106 | 0.94 |
| Sen. Pure protein | 0.899 | 62.01 | −2.53 × 10−2 | 135.55 | −1.686 × 106 | 0.41 |
| Sens. All | 0.336 | 85.78 | −2.85 × 10−2 | 165.32 | −1.702 × 105 | 1.35 |
| Parameter | Human health | Ecosystem | Resources | |||
|---|---|---|---|---|---|---|
| Net saving/impact (DALY) | Improve (%) | Net saving/impact (Species.yr) | Improve (%) | Net saving/impact (USD) | Improve (%) | |
| Base | 2.366 | – | −1.07 × 10−2 | – | −1.679 × 106 | – |
| Sen.5% | 2.204 | 6.85 | −1.16 × 10−2 | 7.64 | −1.683 × 106 | 0.26 |
| Sen.10% | 2.072 | 12.44 | −1.23 × 10−2 | 15.00 | −1.687 × 106 | 0.48 |
| Sen.20% | 1.803 | 23.78 | −1.39 × 10−2 | 29.77 | −1.695 × 106 | 0.94 |
| Sen. Pure protein | 0.899 | 62.01 | −2.53 × 10−2 | 135.55 | −1.686 × 106 | 0.41 |
| Sens. All | 0.336 | 85.78 | −2.85 × 10−2 | 165.32 | −1.702 × 105 | 1.35 |
USD, United States dollar; DALY, Disability-adjusted life year.
The supply of plant's electricity demand from excess wind power and the upgrading of all biogas into biomethane for further injection to gas grids instead of combusting in CHP have higher environmental benefits (Faragò et al. 2021). However, the results achieved herein demonstrated that supplying the plant's heat demand from the grid would result in significant environmental burdens. As mentioned earlier the marginal heat sources are wood chips and wood pellets in 2030 whose combustion has a wide range of air emissions. Having considered to supply plant's heat demand (i.e., 19,784 MWhth/year) through the combustion of part of biogas in a boiler instead of supplying from the grid, sensitivity analysis was performed over the source of heat production. By assuming the lower heat value (LHV) of biomethane (36 MJ/m3 CH4 or 10 kWh/m3 CH4 (Panpong et al. 2014; Firdaus et al. 2017)) and 95% projected efficiency of modern boilers in 2030 (Elyasi et al. 2021), approximately 44% of the biogas (i.e., 3,471 million m3 per year biogas) shall be combusted in the boiler to meet plant's heat requirement (see boiler's emissions inventory in Table 5). It is worth bearing in mind that CO2 and CH4 emissions from biogas combustion were considered biogenic due to their non-fossil origin. The remaining biogas (i.e., 4,417 million m3/year), after supplying the requirements of the MP facility, was upgraded then and injected into the natural gas grid to substitute natural gas. Supplying plant's heat demand from biogas combustion in the boiler could improve the environmental profile of the biorefinery by 79 and 73% in human health and ecosystem quality damage categories, respectively (Figure 6). On the contrary, lower saving was attained in the resource scarcity damage category. The detailed contributional analysis demonstrated (see the supplementary file ‘SI-7’) that the unit process of fermentation dominated in ecosystem quality and resources damage categories with a share of 45 and 46%, respectively.
Emission inventory for the boiler unit
| Emission | Abbreviation | Emission to | Amount (kg/year) |
|---|---|---|---|
| Sulfur dioxide | SO2 | Air | 1.87 × 103 |
| Nitric oxides | NOX | Air | 2.10 × 103 |
| Non-methane volatile organic compounds | NMVOC | Air | 1.50 × 102 |
| Methane | CH4 | Air | 7.50 × 101a |
| Carbon monoxide | CO | Air | 2.70 × 103 |
| Carbon dioxide | CO2 | Air | 6.31 × 106a |
| Nitrous oxide | N2O | Air | 7.50 × 100 |
| Total suspended particles | TSP | Air | 1.12 × 102 |
| Particulate matter (aerodynamic diameters<=10 μm) | PM10 | Air | 1.12 × 102 |
| Particulate matter (aerodynamic diameters<=2.5 μm) | PM2.5 | Air | 1.12 × 102 |
| Black carbon | BC | Air | 3.71 × 100 |
| Chromium | Cr | Air | 1.35 × 10−2 |
| Copper | Cu | Air | 2.32 × 10−2 |
| Mercury | Hg | Air | 9.00 × 10−3 |
| Zinc | Zn | Air | 2.96 × 10−1 |
| Arsenic | As | Air | 3.00 × 10−3 |
| Cadmium | Cd | Air | 1.50 × 10−4 |
| Nickel | Ni | Air | 1.72 × 10−2 |
| Lead | Pb | Air | 3.75 × 10−4 |
| Selenium | Se | Air | 1.57 × 10−2 |
| Emission | Abbreviation | Emission to | Amount (kg/year) |
|---|---|---|---|
| Sulfur dioxide | SO2 | Air | 1.87 × 103 |
| Nitric oxides | NOX | Air | 2.10 × 103 |
| Non-methane volatile organic compounds | NMVOC | Air | 1.50 × 102 |
| Methane | CH4 | Air | 7.50 × 101a |
| Carbon monoxide | CO | Air | 2.70 × 103 |
| Carbon dioxide | CO2 | Air | 6.31 × 106a |
| Nitrous oxide | N2O | Air | 7.50 × 100 |
| Total suspended particles | TSP | Air | 1.12 × 102 |
| Particulate matter (aerodynamic diameters<=10 μm) | PM10 | Air | 1.12 × 102 |
| Particulate matter (aerodynamic diameters<=2.5 μm) | PM2.5 | Air | 1.12 × 102 |
| Black carbon | BC | Air | 3.71 × 100 |
| Chromium | Cr | Air | 1.35 × 10−2 |
| Copper | Cu | Air | 2.32 × 10−2 |
| Mercury | Hg | Air | 9.00 × 10−3 |
| Zinc | Zn | Air | 2.96 × 10−1 |
| Arsenic | As | Air | 3.00 × 10−3 |
| Cadmium | Cd | Air | 1.50 × 10−4 |
| Nickel | Ni | Air | 1.72 × 10−2 |
| Lead | Pb | Air | 3.75 × 10−4 |
| Selenium | Se | Air | 1.57 × 10−2 |
aBiogenic.
Sensitivity analysis over the supply of heat demand by a boiler in ecosystem quality, human health, and resource scarcity damage categories.
Sensitivity analysis over the supply of heat demand by a boiler in ecosystem quality, human health, and resource scarcity damage categories.
Economic assessment
The following sub-sections will comprehensively discuss the economic performance of MP production in terms of expenditures, incomes, and subsequent economic indices.
Expenditures
As demonstrated in Table 6, equipment and more specifically fermenters had the highest contribution to the total CAPEX. The expenditure on fermenters is straightly appertained to providing sufficient hydraulic retention time (HRT) for MP fermentation and consequently the volume of the reactors (Skadborg 2018). Considering a 72-hours HRT to secure high MP yield and N to protein efficiency (Tsapekos et al. 2019) and an input flow rate of 1,384 m3/day (i.e., extracted N-rich flow), a total fermenters’ volume of 4,152 m3 is needed. Among all the facilities used for biological biogas upgrading technology, the water electrolysis facility had the largest contribution to CAPEX. The results achieved herein are in line with those of Elyasi et al. (2021), who proved that, although learning technology and increasing the volume of the water electrolysis system will significantly reduce the investment cost in 2030 (Zauner et al. 2020), it will still have the largest contribution to the CAPEX of biological biogas upgrading technology.
Total CAPEX of a microbial protein production facility and its complementary electrolyzer facility standing at the downstream of a WWTP treating wastewater and organic fraction of municipal solid waste
| Item | Cost (M€) | Portion of total CAPEX (%) | Ref. | Description |
|---|---|---|---|---|
| Equipment expenses | 6.4 | 27.96 | – | Table SI-1 |
| Installation part | 2.1 | 9.32 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 33% of equipment capital |
| Process pipeline | 1.3 | 5.59 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 20% of equipment capital |
| Controls and instrumentation | 0.86 | 3.73 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 13% of equipment capital |
| Electrical systems | 0.86 | 3.73 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 13% of equipment capital |
| Buildings | 2.1 | 9.32 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 33% of equipment capital |
| Yard enhancements | 0.43 | 1.86 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 7% of equipment capital |
| Assisting facilities | 1.1 | 4.66 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 17% of equipment capital |
| Engineering and technical | 1.3 | 5.59 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 20% of equipment capital |
| Construction expenses | 1.7 | 7.46 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 27% of equipment capital |
| Legal expenses | 0.43 | 1.86 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 7% of equipment capital |
| Contractor's payment | 0.64 | 2.80 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 10% of equipment capital |
| Contingency | 2.1 | 9.32 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 33% of equipment capital |
| Electrolysis system | 0.73 | 3.18 | Zauner et al. (2020) | 500 €/kWel |
| Hydrogen storage tank | 0.06 | 0.25 | Gorre et al. (2020) | 33 €/m3 H2 |
| Biological upgrading | 0.43 | 1.88 | Thema et al. (2019); Zauner et al. (2019) | 300 €/kWel |
| Install, design, and plan | 0.34 | 1.49 | Leonzio (2017) | 28% of sum CAPEX |
| Total | 23.1 | 100 | – | – |
| Item | Cost (M€) | Portion of total CAPEX (%) | Ref. | Description |
|---|---|---|---|---|
| Equipment expenses | 6.4 | 27.96 | – | Table SI-1 |
| Installation part | 2.1 | 9.32 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 33% of equipment capital |
| Process pipeline | 1.3 | 5.59 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 20% of equipment capital |
| Controls and instrumentation | 0.86 | 3.73 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 13% of equipment capital |
| Electrical systems | 0.86 | 3.73 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 13% of equipment capital |
| Buildings | 2.1 | 9.32 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 33% of equipment capital |
| Yard enhancements | 0.43 | 1.86 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 7% of equipment capital |
| Assisting facilities | 1.1 | 4.66 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 17% of equipment capital |
| Engineering and technical | 1.3 | 5.59 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 20% of equipment capital |
| Construction expenses | 1.7 | 7.46 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 27% of equipment capital |
| Legal expenses | 0.43 | 1.86 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 7% of equipment capital |
| Contractor's payment | 0.64 | 2.80 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 10% of equipment capital |
| Contingency | 2.1 | 9.32 | Peters et al. (2003); Acién et al. (2012); Kwan et al. (2015) | 33% of equipment capital |
| Electrolysis system | 0.73 | 3.18 | Zauner et al. (2020) | 500 €/kWel |
| Hydrogen storage tank | 0.06 | 0.25 | Gorre et al. (2020) | 33 €/m3 H2 |
| Biological upgrading | 0.43 | 1.88 | Thema et al. (2019); Zauner et al. (2019) | 300 €/kWel |
| Install, design, and plan | 0.34 | 1.49 | Leonzio (2017) | 28% of sum CAPEX |
| Total | 23.1 | 100 | – | – |
M€, million euro; Construction expenses, construction of equipment and tools, safety, permits, etc.; Yard enhancements, site expansion and landscaped; Contingency, uncertainty in evaluation; Assisting facilities, non-process equipment, facilities, and packaging and distribution.
VOPEX and FOPEX of the current process are profoundly calculated and shown in Table 7.In this regard, chemicals used in BES (for P precipitation and pH adjustment) and fermentation process (as supplementary nutrients for methanotrophs), had the highest contribution to OPEX with a share of 35%. The annual cost of chemicals adds up to 0.41 million €/year, which significantly affects the economic performance of MP production. On the contrary, OPEX of biological biogas upgrading made up a low proportion of total OPEX (i.e., 3.5%). This can be attributed to the fact that only 26.35% of the upgraded biogas is consumed within the MP facility (either in form of methane for MP production or in form of electricity), hence, the OPEX of biological biogas upgrading is allocated between the MP facility and wastewater treatment plant.
Total OPEX of MP production with reject water
| Item | Amount | Unit | Cost (€/year) | Description | Refs. |
|---|---|---|---|---|---|
| Chemicals substance | |||||
| Potassium and phosphora | 417.2 | ton | 340,840* | Avg. unit price: 602 €/t | Taken from various sources such as chemical reports, peer-reviewed papers, and reputable supplier companies. |
| Magnesium sourcea | 86.1 | ton | 29,712* | Avg. unit price: 210 €/t | |
| Calcium source | 128.8 | ton | 32,661* | Avg. unit price: 252 €/t | |
| Iron source | 5.53 | ton | 1,737* | Avg. unit price: 231 €/t | |
| Trace metals | 0.39 | ton | 730* | Avg. unit price: 1,868 €/t | |
| Sum | 405,680 | ||||
| Portion of total OPEX (%) | 35.2 | ||||
| Employees costs | |||||
| Staff | 4 | Person | 306,950 | For a person: 26.28 €/hour | Estimated |
| Supervision and control | – | – | 12,278 | 0.04 staff | LaTurner et al. (2020) |
| Sum | 319,228 | ||||
| Portion of total OPEX (%) | 27.7 | ||||
| Insurance, tax, and depreciation cost | |||||
| Depreciation | – | – | 257,929 | According to Equation (4) | Calculated |
| Property tax | – | – | 2,579 | 0.01 of depreciation | LaTurner et al. (2020) |
| Insurance | – | – | 1,548 | 0.06 of depreciation | LaTurner et al. (2020) |
| Sum | 262,056 | ||||
| Portion of total OPEX (%) | 22.7 | ||||
| Utility costs | |||||
| Electricity | 851 | MWh | 0 | 0 | Estimated |
| Biopulp | 32,115 | ton | 111,756 | Unit price: 17 €/t | Cimpan & Wenzel (2016) |
| Sum | 125,507* | ||||
| Portion of total OPEX (%) | 10.9 | ||||
| Biological biogas upgrading costs | |||||
| Electrolyzer service | – | – | 14,689 | 2% of CAPEX | Zauner et al. (2019) |
| Hydrogen storage | – | – | 577 | 1% of CAPEX | Zauner et al. (2019) |
| Biological upgrading | – | – | 21,694 | 5% of CAPEX | Zauner et al. (2019) |
| Electricity | 53,017 | MWh | 0 | 0 | Estimated |
| Water | 11,348 | m3 | 14,299 | Unit price:1.26 €/m3 | Elyasi et al. (2021) |
| Sum | 39,886 | ||||
| Portion of total OPEX (%) | 3.5 | ||||
| Item | Amount | Unit | Cost (€/year) | Description | Refs. |
|---|---|---|---|---|---|
| Chemicals substance | |||||
| Potassium and phosphora | 417.2 | ton | 340,840* | Avg. unit price: 602 €/t | Taken from various sources such as chemical reports, peer-reviewed papers, and reputable supplier companies. |
| Magnesium sourcea | 86.1 | ton | 29,712* | Avg. unit price: 210 €/t | |
| Calcium source | 128.8 | ton | 32,661* | Avg. unit price: 252 €/t | |
| Iron source | 5.53 | ton | 1,737* | Avg. unit price: 231 €/t | |
| Trace metals | 0.39 | ton | 730* | Avg. unit price: 1,868 €/t | |
| Sum | 405,680 | ||||
| Portion of total OPEX (%) | 35.2 | ||||
| Employees costs | |||||
| Staff | 4 | Person | 306,950 | For a person: 26.28 €/hour | Estimated |
| Supervision and control | – | – | 12,278 | 0.04 staff | LaTurner et al. (2020) |
| Sum | 319,228 | ||||
| Portion of total OPEX (%) | 27.7 | ||||
| Insurance, tax, and depreciation cost | |||||
| Depreciation | – | – | 257,929 | According to Equation (4) | Calculated |
| Property tax | – | – | 2,579 | 0.01 of depreciation | LaTurner et al. (2020) |
| Insurance | – | – | 1,548 | 0.06 of depreciation | LaTurner et al. (2020) |
| Sum | 262,056 | ||||
| Portion of total OPEX (%) | 22.7 | ||||
| Utility costs | |||||
| Electricity | 851 | MWh | 0 | 0 | Estimated |
| Biopulp | 32,115 | ton | 111,756 | Unit price: 17 €/t | Cimpan & Wenzel (2016) |
| Sum | 125,507* | ||||
| Portion of total OPEX (%) | 10.9 | ||||
| Biological biogas upgrading costs | |||||
| Electrolyzer service | – | – | 14,689 | 2% of CAPEX | Zauner et al. (2019) |
| Hydrogen storage | – | – | 577 | 1% of CAPEX | Zauner et al. (2019) |
| Biological upgrading | – | – | 21,694 | 5% of CAPEX | Zauner et al. (2019) |
| Electricity | 53,017 | MWh | 0 | 0 | Estimated |
| Water | 11,348 | m3 | 14,299 | Unit price:1.26 €/m3 | Elyasi et al. (2021) |
| Sum | 39,886 | ||||
| Portion of total OPEX (%) | 3.5 | ||||
★Including the cost of transportation.
aUsed in BES and culture medium.
Economic indices
Having considered optimistic, base, and pessimistic scenarios, the economic indices are calculated and shown in Table 8. As can be seen, the gross profit for this process was estimated at 0.29 M€ and 0.22 M€ under optimistic and base scenarios, respectively. It is noteworthy that the economic indices under the pessimistic scenario were not reported as they were found to be unprofitable. Although the results of gross profit were slightly promising but the overall economic performance of MP production was not encouraging in terms of NPV and IRR. The high CAPEX associated with 2nd generation MP from waste streams has undermined its economic performance. For a project to be economically viable, IRR should be equivalent to the discount rate (i.e., 8%) and NPV should be positive. The detailed analysis demonstrated that CAPEX associated with fermenters (i.e., cost of fermenters) had the largest share. As discussed earlier, considering a 72-hours HRT to secure high MP yield and N to protein efficiency increased the total reactors’ volume needed for the MP production plant, and subsequently raised the total CAPEX. The future advancement in the technology of fermenters and the development of U-Loop fermenters can play a significant role in decreasing the CAPEX of MP production in the future. Having compared with CSTRs used for the MP production in this study, U-Loop fermenters can perform better in terms of production rate while shortening the HRT and decreasing the required reactor size. In this regard, the economic performance of MP under the assumption of replacing CSTRs with U-Loop fermenters has been investigated under sensitivity analysis.
Revenue, expenditure, and economic profitability parameters under pessimistic, base, and optimistic scenarios
| Item | Scenarios | Unit | ||
|---|---|---|---|---|
| Optimistic | Base | pessimistic | ||
| Revenue from products sales | ||||
| MP | 0.88 | 0.81 | 0.52 | M€/year |
| Oxygen | 0.31 | 0.31 | 0.31 | M€/year |
| Sum | 1.18 | 1.12 | 0.83 | M€/year |
| Expenditure of 1 ton of pure protein production | ||||
| CAPEX | 35,293 | 35,293 | 35,293 | €/t pure protein |
| FOPEX | 946 | 946 | 946 | €/t pure protein |
| VOPEX | 817 | 829 | 830 | €/t pure protein |
| Economic profitability indices | ||||
| Gross profit* | 0.29 | 0.22 | – | M€ |
| ROI | 5.14 | 4.87 | – | % |
| NPV | −20.18 | −20.88 | – | M€ |
| IRR | −11.03 | −12.74 | – | % |
| Item | Scenarios | Unit | ||
|---|---|---|---|---|
| Optimistic | Base | pessimistic | ||
| Revenue from products sales | ||||
| MP | 0.88 | 0.81 | 0.52 | M€/year |
| Oxygen | 0.31 | 0.31 | 0.31 | M€/year |
| Sum | 1.18 | 1.12 | 0.83 | M€/year |
| Expenditure of 1 ton of pure protein production | ||||
| CAPEX | 35,293 | 35,293 | 35,293 | €/t pure protein |
| FOPEX | 946 | 946 | 946 | €/t pure protein |
| VOPEX | 817 | 829 | 830 | €/t pure protein |
| Economic profitability indices | ||||
| Gross profit* | 0.29 | 0.22 | – | M€ |
| ROI | 5.14 | 4.87 | – | % |
| NPV | −20.18 | −20.88 | – | M€ |
| IRR | −11.03 | −12.74 | – | % |
M€, million €.
*Gross profit does not take the property tax and depreciation into account; FOPEX, fixed operation cost; CAPEX, capital expenditure; VOPEX, variable operation cost; IRR; internal rate of return; ROI, return on investment; NPV, net present value.
Sensitivity analysis for economic assessment
Sensitivity analysis was performed by considering the potential economic changes due to the following modifications: (1) replacing conventional fermenters with U-loop reactors, (2) reducing the consumption of Modified-dAMS solution by 5, 10, and 20% through process optimization and proper recycling of culture medium, (3) improving pure protein yield on NH4-N, and (4) considering of surplus electricity price. Table 9 summarises the expected changes in economic indices.
Economic sensitivity analysis under an optimistic scenario
| Parameters | Economic profitability indices | |||||||
|---|---|---|---|---|---|---|---|---|
| Gross profit* (M€) | Improve (%) | ROI (%) | Improve (%) | NPV (M€) | Improve (%) | IRR (%) | Improve (%) | |
| Sen. Reactor | 0.294 | 0.34 | 13.13 | 7.99 | −6.14 | 69.59 | −4.38 | 6.65 |
| Sen. 5% | 0.35 | 20.69 | 5.14 | 0 | −19.57 | 3.02 | −9.80 | 1.23 |
| Sen. 10% | 0.37 | 27.59 | 5.14 | 0 | −19.39 | 3.91 | −9.46 | 1.57 |
| Sen. 20% | 0.41 | 41.38 | 5.14 | 0 | −19.01 | 5.80 | −8.81 | 2.22 |
| Sen. Pure protein | 0.47 | 62.07 | 5.91 | 0.77 | −18.43 | 8.67 | −7.88 | 3.15 |
| Sens. electricity price | −0.93 | −420.68 | 5.14 | 0 | – | – | – | – |
| Sens. All | 0.59 | 101.33 | 15.10 | 9.96 | −3.23 | 84.01 | 1.91 | 12.94 |
| Parameters | Economic profitability indices | |||||||
|---|---|---|---|---|---|---|---|---|
| Gross profit* (M€) | Improve (%) | ROI (%) | Improve (%) | NPV (M€) | Improve (%) | IRR (%) | Improve (%) | |
| Sen. Reactor | 0.294 | 0.34 | 13.13 | 7.99 | −6.14 | 69.59 | −4.38 | 6.65 |
| Sen. 5% | 0.35 | 20.69 | 5.14 | 0 | −19.57 | 3.02 | −9.80 | 1.23 |
| Sen. 10% | 0.37 | 27.59 | 5.14 | 0 | −19.39 | 3.91 | −9.46 | 1.57 |
| Sen. 20% | 0.41 | 41.38 | 5.14 | 0 | −19.01 | 5.80 | −8.81 | 2.22 |
| Sen. Pure protein | 0.47 | 62.07 | 5.91 | 0.77 | −18.43 | 8.67 | −7.88 | 3.15 |
| Sens. electricity price | −0.93 | −420.68 | 5.14 | 0 | – | – | – | – |
| Sens. All | 0.59 | 101.33 | 15.10 | 9.96 | −3.23 | 84.01 | 1.91 | 12.94 |
*Gross profit does not take the property tax and depreciation into account; M€, million €; ROI, return on investment; NPV, net present value; IRR; internal rate of return.
Having replaced CSTRs with U-loop fermenters, the required HRT is expected to decrease much less than 24 hours (Drejer et al. 2017; Petersen et al. 2019), so that the size of the reactors decreases without affecting the biomass production rate. Although, the amount of much less than 24 hours (up to 5 hours) has been reported by unpublished reports from companies working in this context, 24-hours HRT was opted for sensitivity analysis based on journal peer review. Accordingly, IRR, NPV, and ROI can be improved by 6.65, 69.59, and 7.99%, respectively, though changes in gross profit were negligible, i.e., 0.34%. Hence, the cost-efficiency of this MP production pathway was highly dependent on reactor volume and the HRT required to optimal production rate.
Although increasing the NH4-N to protein conversion ratio up to 5.25 is needed to have a better environmental profile for 2nd generation MP and can raise gross profit by 62%, it was not sufficient to make MP production profitable. As shown in Table 9, by considering all possible changes together (Table 9; Sens. All), the economic performance of 2nd generation MP production can be significantly improved, although it will not be still economically profitable. Among all the economic indices calculated herein, ROI demonstrated the least sensitivity, i.e., 9.96%. This is because ROI is independent of the raw materials and utility cost (i.e., OPEX). Therefore, reducing the consumption of modified-dAMS and considering surplus electricity price did not affect the ROI. In respect to MP production for substituting soy-meal and fishmeal, LaTurner et al. (2020)reported that the market price of soy-meal undermines the economic feasibility of MP production. Therefore, granting subsidies and incentive schemes are needed to promote the production of 2nd generation MP.
CONCLUSION
In the current study, the environmental and economic feasibility of a biorefinery platform established on a WWTP was assessed. This platform has targeted to valorising N and carbon flows from wastewater treatment and anaerobic co-digestion of sludge and OFMSW into microbial protein and bioenergy. The environmental profile of such a biorefinery was found promising in ecosystem and resources scarcity damage categories where net negatives of −1.07 × 10−2 species.yr/FU, and −1.68 × 106 USD/FU, respectively, were attained. Further system modifications and improvements are still needed to gain net savings in the damage category of human health. Under the baseline scenario a net impact of 2.37 DALY/FU was achieved in the human health damage category. Impacts from anaerobic digestion, electrolysis unit process, and background emissions from the production of supplementary chemicals that are used for the fermentation of MP had considerable impacts on theenvironmental profile. However, having compared 1 ton of pure protein from MP against conventional proteinaceous feed sources such as soybean and fishmeal, better perspectives are imagined for the future of MP. In terms of economic performance, although gross profit was slightly promising, but the overall economic performance of MP production was not encouraging. The high CAPEX has undermined the economic feasibility of MP production. More specifically, conventional and low-efficient fermenters increase investment costs. Sensitivity analysis indicated that improving pure protein yield on ammonium nitrogen and replacing conventional fermenters with U-loop reactors can to some extent increase the gross profit and decrease the investment costs, respectively, although still not enough to make the process economically profitable.
ACKNOWLEDGEMENT
We acknowledge support from the FUBAF project financed by the Danish EPA – MUDP (J.nr. Mst-11700508).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
%20cropped.png?versionId=5945)