Evaluation of Environmental Sustainability of Biorefinery and Incineration with Energy Recovery Based on Life Cycle Assessment ^†

Abstract

Based on Life Cycle Assessment (LCA) and ISO standards, we compared the global environmental sustainability (ES) of two technologies that process the organic fraction of municipal solid waste (OFMSW) in Mexico. The first technology was a biorefinery (BRF) known as HMEZSNN-BRF (abbreviation for Hydrogen-Methane-Extraction-Enzyme-Saccharification/Nanoproduction Biorefinery); it produces the gas biofuels hydrogen (H) and methane (M), organic acids (E), enzymes (Z), saccharified liquors (S), and bionanobioparticles (BNBPs) in a nanoproduction stage (NN). The second technology was incineration with energy recovery (IER). An LCA was performed with a functional unit (FU) of 1000 kg of OFMSW. The BRF generates 166.4 kWh/FU (600 MJ) of net electricity, along with bioproducts such as volatile organic acids (38 kg), industrial enzyme solution (1087 kg), and BNBPs (40 kg). The IER only produces 393 net kWh/FU electricity and 5653 MJ/FU heat. The characterization potential environmental impacts (PEIs) were assessed using SimaPro software, and normalized PEIs (NPEIs) were calculated accordingly. We defined a new variable alpha and the indices σ-τ plane for quantifying the ES. The higher the alpha, the lower the ES. Alpha was the sum of the eighteen NPEIs aligned with the ISO standards. The contributions to PEI and NPEI were also analyzed. Four NPEIs were the highest in both technologies, i.e., freshwater and marine ecotoxicities and human non-carcinogenic and carcinogenic toxicities. For the three first categories, the NPEI values corresponding to IER were much higher than those of the BRF (58.6 and 8.7 person*year/FU freshwater toxicity; 93.5 and 13.6 marine ecotoxicity; 12.1 and 1.8 human non-carcinogenic toxicity; 13.7 and 13.9 human carcinogenic toxicity, for IER and the BRF, respectively). The total α values were 179.1 and 40.7 (person*yr)/FU for IER and the BRF, respectively. Thus, the ES of IER was four times lower than that of the BRF. Values of σ = 0.592 and τ = −0.368 were found; the point defined by these coordinates in the σ-τ plane was located in Quadrant IV. This result confirmed that the BRF in this work is more environmentally sustainable (with restrictions) than the IER in Mexico for the treatment of the OFMSW.

1. Introduction

The treatment and disposal of solid waste remains among the most significant environmental and economic challenges in developing and developed countries. In 2018, it was estimated that 2.01 × 10¹² t of waste is generated annually worldwide, of which ~44% is yard and food waste, followed by dry recyclables (cardboard and paper, 17%), plastics (12%), and glass and metals (5 and 4%, respectively). This waste releases ~1.6 × 10¹² t of greenhouse gas (GHG) emissions into the atmosphere [1]. These impacts are most severe in low- and middle-income countries [1,2,3]. In Mexico, it is estimated that approximately 120,000 t of municipal solid waste (MSW) was generated daily in 2019 [4]. The Secretary of the Environment and Natural Resources (known by its acronym in Spanish, SEMARNAT) [4] has reported that 46% of the waste is organic from food, crops, and wood. This waste is disposed of in landfills (LFs) and open dumps. The National Institute of Ecology and Climate Change (known by its acronym in Spanish, INECC) [5] estimates emissions of 21,921 Gg of CO e from the final disposal of MSW. Worldwide, 37% of MSW is disposed in LFs and 33% in open dumps, 11% is incinerated, and 19% is recycled and composted [1]. Middle-income countries (or developed countries; DCs) have the highest LF percentage, at 54%, followed by high-income countries at 39%. In high-income countries, 22% of MSW is incinerated and 35% is either composted or recycled [1].

LF is a relatively low-cost and widely used technology for MSW disposal [6], used in high-income and upper-middle-income countries. In contrast, open dumping predominates in low-income countries (93%) [1]. The adverse environmental impacts of LFs are of concern and are related to GHG emissions in poorly designed sites or those with no biogas recovery, as well as the contamination of aquifers and subsurface soils via leachate [6,7,8]. The scarcity of available land is a practical limitation to implementing LFs (Supplementary Materials, Annex S13, Section S13.2). For a more detailed examination of the importance, trends, advantages, and disadvantages of LFs, BRF, and IER technologies, please refer to Supplementary Materials Annex S13.

In DC waste incineration represents a crucial component of the solid waste management system [9]; it is regarded as a competitive technology for simultaneously reducing the waste volume and allowing energy recovery, which can reduce GHGs, including CH₄ emissions from buried waste [10,11,12]. Incineration is reported to process 18% of MSW across Europe and Central Asia, 12% in North America, and 4.5% in Latin America and the Caribbean [1]. Escamilla-García et al. [13] and the SEMARNAT-GIZ German Society for International Cooperation (GIZ) study [14] have demonstrated that the incineration of waste can be an efficient option for electricity generation in Mexico. However, implementing incineration technology gives rise to several concerns, primarily related to its adverse potential environmental impacts [15,16,17].

According to Duan et al. [18], Khoshnezisan et al. [19], and Ladakis et al. [20], among other experts [21,22], an alternative that is currently attracting attention for the treatment of MSW is the BRF approach (Supplementary Materials Annex S13, Section S13.4). BRF technology employs core biotechnological processes and waste biomass as feedstock, ensuring compliance with the principle of environmental sustainability. Escamilla-Alvarado et al. (2017) [23] and Fava et al. [24] state that this principle is achieved by ensuring that inputs intended for human consumption are not compromised, among other features [25,26]. The BRFs are facilities that employ an integrated, efficient, and flexible conversion of organic waste to various valuable products and energy. Depending on the integration and type of BRF, it has been proposed that biogas, other bioenergies, and value-added products (VAPs) such as microbial protein, bio-succinic acid, and lactic acid can be produced [19,27,28], along with the extraction of oils and fats, the extraction of proteins, and the production of enzymes, among other uses [27,28]. In 2022, the International Energy Agency (IEA) published a report indicating that there were 1312 operational BRFs worldwide [29,30]. As indicated by the U.S. Energy Information Administration [30], over 40 BFRs were identified in the USA, Finland, Germany, France, the Netherlands, Sweden, and China; 1 BRF was constructed in Mexico.

Mexico is undergoing a transition in solid waste management, whereby the replacement of LFs [14,31] with more intensive and potentially more sustainable waste treatment methods is proposed and supported [13,32]. These findings are encouraged by several sources [23,33,34] (Supplementary Materials Annex S13, Section S13.5). The use of more intensive technologies such as IER and BRFs, which process the OFMSW and provide energy/bioenergy (along with VAPs in the second case), could prove to be an avenue worth exploring. For example, numerous researchers and institutions in Mexico and elsewhere have advocated for a paradigm shift that supports the use of BRFs for solid waste management [23,34]. This could include changes in the Mexican regulatory framework to facilitate or boost the implementation of waste-based BRFs in Mexico. Similarly, other experts [35] have also concluded that the design and approval of new regulations and technologies will be necessary to contribute to sustainable development in Mexico. Concurrently, as more stringent environmental regulations related to waste LFs have been issued in several DCs and underdeveloped countries (UCs) [36,37,38], this change has drawn the attention of some decision-makers and government officers to the opportunity to replace LFs with potentially more cost-effective energy-from-waste (EfW) technologies [10,39]. IER is among the EfW technologies that could provide a more intensive and sustainable solution to waste management [32]; in Mexico, it is supported by SEMARNAT [4,31] and other specialists [13,14], while other DCs [10,40] have also favored IER.

In summary, IER and BRF can be seen as intensive alternatives to replace LF in Mexico and other developing countries seeking improved OFMSW management, with the recovery of valuable energy and resources, along with possible reductions in the environmental impacts of LF and open dumps. Therefore, comparative studies on the environmental sustainability of IER and BRF for OFMSW management in large cities are needed to inform the selection of such technologies (Supplementary Materials Annex S13, Section S13.6).

Unfortunately, the information in the open literature supporting a side-by-side evaluation of the overall environmental sustainability of IER and an original model of BRF, such as our BRF for the treatment of urban organic waste, is scarce (Annex S12; Annex S13 Section S13.6 in the Supplementary Materials; readers are encouraged to thoroughly read Annex S12 because these paragraphs are just an excerpt). There is a knowledge gap that remains to be filled. Furthermore, side-by-side comparisons of the environmental sustainability of IER and BRF using indicators aligned with ISO standards and LCA are still scarce in the open literature. The most frequent reasons are described below:

(i) According to the developed literature review, none of the identified related studies has applied the methodology used in our work to an innovative and integrated MSW biorefinery such as the GBAER-type BRF or compared it to a non-biologically driven intensive technology such as IER, to the best of our knowledge. On the one hand, most of the BRFs of the few that have appeared in comparative studies are simple setups (Table S12.1, Annex S12, Supplementary Materials), such as a “waste refinery” [41] that consists of just enzymatic hydrolysis followed by either anaerobic digestion or incineration [42] or plain ethanol production followed by incineration, or only a larvae-stabilizing process followed by composting [42], or just a sorting plant followed by anaerobic digestion and composting [43], which can hardly be called biorefineries because they only exhibit one or two co-bioproducts at most and two to the three main processes [24]. On the other hand, our BRF includes two bioenergies and three co-bioproducts based on a relatively complex network of biological, physical, and phys–chem processes.

(ii) In some cases, there is no evidence that the comparison uses the same FU or at least a harmonized FU for both technologies [43]. Thus, these could lead to debatable results and an unfair comparison.

(iii) In other cases, the comparison is made with the use of different indicators (such as Eco-Indicator 99, now superseded, Tonini and Astrup (2012) [41]) or emergy indicators [43] whose results are not compared or translated to equivalent results of ISO-LCA-based indicators.

(iv) In other works, the results of technology comparison only use a reduced set of PEIs or NPEIs (from one to seven, representing only 0.06 to 38% of the eighteen contributing PEIs).

(v) Furthermore, there is no evidence of using an adequate single score in most references, based on those impacts, which could summarize the overall ES trends ([42,44,45,46]). Exceptions are the work by Tonini and Astrup [41], who used the Eco-Indicator 99 milliPoints, and Clasen et al. [43], who reported results with the ERI (emergy) indicator. Unfortunately, they did not use other indicators (midpoint or endpoint ones based on ISO-LCA and the ReCiPe methodology) for further insight and internal comparison with the reported milliPoints or emjoules.

(vi) As a consequence of points (iii), (iv), and (v), in most articles, even those with higher ranking, a deep discussion of either the PEIs or NPEIs, as well as their critical contributions and recommended improvements to the ES of the examined technologies, is missing.

A key tool for the further determination of ES is the LCA methodology, which is an iterative technique used to evaluate a wide and comprehensive variety of environmental impacts through a set of protocols and accepted methodologies [47,48]. LCA also evaluates the consumption of resources during each stage of production, use, and final disposal [49,50]. LCA can be used to compare waste management technologies [51,52,53] to understand which are the most effective, and to minimize environmental impacts [45,54]. To date, the use of LCA results has been limited to the analysis of a few characterization PEIs, and no application of LCA for building overall metrics of ES has been reported.

The Environmental Biotechnology and Renewable Energy Group (GBAER) in Mexico has developed a BRF called “HMEZS-NN”, the core processes of which are based on the fermentation of MSW (Supplementary Materials Annex S13 Section S13.8). The BRF integrates additional processes into the HMZS BRF [23,55,56], such as a step for the extraction of organic acids from the fermented solids prior to the production of industrial enzymes, as well as a step for the production of BNBPs using saccharified liquors [23,55,56]. An intrinsic advantage of this type of BRF (now known as HMEZS NN [23,26]) is its ability to comply with the fundamental principle of sustainability and four auxiliary principles that contribute to its achievement [57]. These sustainability principles are also supported at least partially by the findings of [58,59,60]. However, the quantitative, overall environmental sustainability of BRFs remains to be demonstrated.

The development of waste-based biorefineries would not only help to shift the linear economies toward circular economies but also contribute to improving public health and the environment [43,61], reducing waste generation, as well as producing bioenergy and bioproducts that are incorporated into the productive cycles of modern societies, and decreasing society’s reliance on raw material imports. Sadeleer et al. [62] observed that boosting the circular economy through BRFs can minimize the waste stream sent to landfills by promoting the continual circulation of waste-based products through redesign, reuse, recycling, and energy recovery. Therefore, assessment and scientific validation of the environmental sustainability of a BRF, as well as the ES of alternative intensive technologies of MSW treatment and management (such as IER), is of paramount importance to avoid misleading results and poor decision-making.

Following the probable shift toward replacing LFs with more intensive and potentially sustainable technologies (such as IER and BRF) in Mexico, along with the lack of information on side-by-side comparisons of the overall ES of such technologies, as documented above, we aimed to evaluate and compare the overall ES of these two intensive urban waste treatment technologies (the BRF HMEZS-NN and the IER), in order to select the most ES option for treating MSW in a large city such as Mexico City. The evaluation of ES was based on the results of eighteen environmental impacts according to the LCA methodology. As a subsidiary objective, a new environmental sustainability index (α) aligned with LCA and ISO-related standards (14040 and 14044) [47,48], which uses all of the normalized environmental impacts and a novel methodology (σ-τ), was developed to determine and compare the overall ES of both technologies.

2. Materials and Methods

2.1. Experimental Design

The core of the experimental design was the side-by-side comparison of the environmental impacts of the BRF and IER, using methods and metrics based on or aligned with LCA and the ISO standards of GBAER HMEZS-NN BRF and IER. The goal and scope definition, life cycle inventory (LCI), LCA methodology, and interpretation of results are presented in the following subsections. The methodology of our LCA follows the model implemented by other authors who have conducted LCA in the context of evaluating the potential environmental impacts of environmental technologies [63,64,65], all of whom followed the ISO 14040 standard [47,48]. The experimental design is complemented with specific methods in the following subsections. Among them, we highlight the index μ for direct comparison of a given characterization PEI between the two technologies, the calculation of normalized potential environmental impacts, and the development of the new sustainability indices α, σ, and τ used to determine the environmental sustainability of both technologies.

2.1.1. Goal, Scope, System Description, and Limitations in the LCA of the BRF and IER

The objective of this work was to evaluate the environmental sustainability of two organic waste-processing technologies, BRF HMEZS-NN and IER, using LCA to determine their potential environmental impacts. The functional unit (FU) and reference flow consisted of the processing of 1000 kg of OFMSW. The composition was 60% paper and cardboard and 40% food waste, while the moisture content was 35%. The scope of this study began with the reception of the waste and continued until the final disposal. Each system has its specificities, which are described in the following sections.

Biorefinery HMEZS-NN

Our lab-scale biorefinery, also known as the GBAER type or HMEZS-NN BRF (where GBAER stands for the acronym in Spanish of the Environmental Biotechnology and Renewable Energies Group), produces bioenergy and bioproducts via six stages: (H) hydrogen production, (M) methane generation, (E) extraction of a concentrate of organic acids and solvents, (Z) production of industrial enzymes, (S) production of saccharified liquors, and (NN) production of bionanobioparticles (BNBPs) (Figure 1 and

Table 1. Biorefinery HMEZS-NN: Description of stages.

Stage Name and (Notation)	Description
Production of biohydrogen (H)	OFMSW conditioning to provide humidity (35%) and alkalinity. Use of domestic wastewater or recirculated internal effluents. Hydrogen production by dark fermentation (hydrogen biogas, BG-H) and fermented wastes (F a). Hydrogen purification at 99% v/v. Electric power generation using purified H2 in hydrogen fuel cells.
Production of methane (M) 40% of FW	CH4 production (methane biogas, BG-M). CH4 purification to 96% v/v.Electricity and heat generation (using previously purified CH4) in a combined-cycle heat and electricity cogeneration plant.
Extraction of organic acids and solvents	VOA b and low-molecular-weight solvents (acetone, butanol) were extracted from the FW flow (60%).
Enzyme production (Z)	Production of industrial enzymes from 40% of the FW current.
Hydrolysates or saccharified liquors(S) 20% of FW	Production of saccharified liquors from acid hydrolysis of extracted FW. Neutralization of saccharified liquors. Detoxification of saccharified liquors with activated carbon.
Production of nanobioparticles (NN)	Production of BNBPs in methanogenic bioreactors (colonize and nanodecorate the bioparticles) using saccharified liquors as a substrate and a solution of iron chloride as a Fe(III) source. Methanogenic biogas purification. Cogeneration of electric power and heat.

^a FW (fermented waste); ^b VOA (volatile organic acids).

). Table 1 describes each step in detail [55]. As a first step, the OFMWS is received at the BRF gate. The processes of collection, recyclables’ separation (also known as sorting), and transport to the plant are not considered by other experts in their published works on LCA and sustainability analysis of MSW (Ebrahimian et al. [66]; Ebrahimian et al. [67]; Bozorgirad et al. [68]; Ma et al. [42], among others). Note that there is an implicit separation of recyclables upstream of the BRF, which could contribute to the circularity and to potential reduction in environmental impacts of the overall waste management approach. However, the recycling process is not part of the current BRF.

The OFMWS is conditioned with municipal wastewater (or internal wastewater recirculated for conditioning) to add moisture and phosphate in order to stabilize the alkalinity and pH in the downstream bioreactors. Municipal wastewater is considered to be available at no cost or with minimal material or transportation costs. In addition, internal recirculation of effluent is considered to minimize the discharge of wastewater from the BRF to the environment or outside treatment, as well as offering an opportunity to save on the addition of phosphate material, since effluents recycled from methanogenic bioreactors present in the BRF and from Stage Z are sources of alkalinity. The waste (output) currents correspond to solids from the saccharification process, thickened anaerobic (methanogenic) digestates, and re-fermented solids from Stage Z, sent to a sanitary landfill.

Allocation was performed according to either co-product mass or economic value in each stage of the BRF. For the H-stage, the allocation was performed according to mass, which was 20% to biogas production and 80% to fermented wastes. In the M-stage, the allocation was economic, with 100% methane. For the E-stage, 100% was allocated to the concentrate of volatile organic acid solvents of low molecular weight. For the Z-stage, the allocation was economical, with 90% for enzyme concentrate and 10% for referred solids. In the S-stage, the allocation was 100% for the saccharified liquors. Finally, for the NN stage, the allocation was economic, with 80% for bionanobioparticles and 20% for methane.

Incineration with Energy Recovery

The design of the IER system was based on the “proxy” process for IER in the Ecoinvent 3.5 database, entitled “Biowaste {CH}| treatment of municipal incineration with fly ash extraction|Cut-off, U” [69]. It includes a municipal solid waste incinerator with ash extraction recommended for mixed biowaste (garden, food/kitchen waste). It is included in the SimaPro v 9.0.0.35, Analyst/Professional software [70]. The incinerator consists of an average Swiss MSW incinerator from 2010 (grate incinerator), with an electrostatic precipitator for fly ash (ESP) and a wet flue gas scrubber [69] (Figure 2).

The LCA activities start with the reception of waste at the factory gate and its delivery to the waste reception area at the incinerator site (transport to the incinerator is omitted). The waste is burned in the incinerator [69]. The activities end with short-term, waste-specific emissions to water (mainly from leachate) and long-term emissions from LF to groundwater. The plant is fed with OFMSW at 35% humidity. The waste enters the combustion chamber through grates. Heat recovery takes place in a boiler, followed by a turbogenerator for power generation (Figure 2). The flue gas treatment consists of an electrostatic precipitator (ESP) for particulate removal, a wet flue gas scrubber, a non-selective catalytic reduction (NSCR) system with an efficiency of 25%, and a high-and low-dust selective catalytic reduction (SCR) system (Figure 2) with efficiencies of 42.77 and 32.68%, respectively [70]. Finally, the filter ash treatment is carried out by the acid leaching of ash (FLUWA) system and operates with an efficiency of 46.22% [69], which comes from the proxy selected from Ecoinvent. The FLUWA process (Flugaschenwäsche, meaning “fly ash washing” in German) was developed to treat fly ash generated from MSW incinerators and is extensively used in Swiss waste-to-energy plants. The FLUWA process aims to extract and recover heavy metals from fly ash, reducing its environmental impact and enabling safer disposal or reuse of the treated residue. The process involves washing the fly ash, separating it, and eventually recycling heavy cation streams [74,75,76].

Zucha et al. [77] reported an average recovery of 43% for Zn, Pb, Cd, and Cu using FLUWA in the incineration process in Switzerland, which is similar to the values selected for our IER. Weibel et al. [76] published a study on optimizing the FLUWA process at the MSW incineration plant of Linth in Switzerland by determining ideal process parameters for optimal metal recovery. The authors showed that heavy metal recovery could be improved up to 63.5%.

2.1.2. Life Cycle Inventory

The life cycle inventory (LCI) of the BRF was obtained using the reported data on the mass and energy balances of the different processes, along with experimental information obtained by the GBAER [78,79,80] and published in master’s theses, doctoral theses, and international articles [23,55,81]. The LCI of this technology is presented in Annex S1 of the Supplementary Materials. The LCI of the IER was obtained from information in the SimaPro proxy for biowaste incineration from the Ecoinvent database, as mentioned above, and following the LCA methodology (ISO 14044 and 14040) [47,48].

The composition of biogas produced in the hydrogenogenic bioreactor was 30% v/v hydrogen, 69% CO₂, and H₂S traces based on our experiments published in journals indexed in the JCR Web of Science. Our values are similar to those reported by Martínez-Fraile et al. [82] and Moussa et al. [83]. Martínez-Fraile et al. [82] reported an H₂ content within a range of 24.3–44.7% from the dark fermentation of organic wastes, while Moussa et al. [83] observed an H₂ composition of 28.45% from dark co-fermentation of organic/food waste, sludge, and biomass.

Regarding the biogas production in the methanogenic bioreactor, a methane content of 65%, CO₂ of 35%, and H₂S traces were taken from our published experiments. The M biogas composition profile was similar to the results of Tonini and Astrup [41], who focused on the treatment comparison of sorted MSW processed in an IER or a waste refinery. The composition of our M biogas was also in the order of the CH₄ ranges found by Frankiewicz [84] and Mes et al. [85], who reported CH₄ contents within 50–70% and 55–70%, respectively, with the remainder CO₂. For the electricity production efficiency in IER from municipal waste, 20% was assumed, coinciding with Niessink [86], who reported 20% efficiency for electricity production from Dutch waste incinerators, and Buekens [72], who compiled thermal (electric) efficiency into electricity of 16–24% for refuse incinerators in the Netherlands. Tonini and Astrup [41] performed an LCA of a municipal solid waste incinerator with an electricity efficiency of 20% in a Danish incinerator. Fruergaard and Astrup [87] also stated an electricity efficiency value of 19.5% for most Danish waste incinerators. For IER, 70% heat recovery efficiency was taken, within the range presented by Tchobanoglous and Kreith 70–75% [73].

Concerning the emissions assigned to our IER, the steps shown in Figure S19.1, Annex S19, Supplementary Materials, were followed. First, the proxy entitled “Biowaste {CH}|treatment of municipal incineration with fly ash extraction|Cut-off, U” was selected from the Ecoinvent 3.5 database. Next, the FU was harmonized by the total mass of the two functional units (the proxy and our BRF) and by the two dry matter contents, using the factors δ and ϕ (Section 2.1.4). Emissions for the harmonized FU were estimated by multiplying the emissions in the proxy by both δ and ϕ. Subsequently, the harmonization of the proxy emissions for the harmonized FU was carried out considering the maximum permissible levels (MPV) of emissions in the regulations on emissions to the atmosphere by incinerators in Mexico and Switzerland (NOM-098-SEMARNAT-2002 [88]; European Energy Agency, 2022 [89]). Then, the ratios of the MPV of each reported item were determined, with the Mexican MPV in the numerator and the MPV of the Swiss regulation in the denominator. Based on these ratio values, an average quotient was obtained (factor). Next, the corresponding emissions of the harmonized proxy were multiplied by this factor to account for Mexican regulations’ permissibility and poorer maintenance, among other issues. The ratio average factor might change (decrease) in the future, but the Mexican Government has not announced any intention of updating the emissions regulations to the atmosphere.

The energy flows in the HMEZS-NN BRF were estimated by a combination of (i) our experimental data published in articles in the open literature, master’s and doctoral theses of the GBAER of Cinvestav (Escamilla-Alvarado et al. (2017) [23], Escamilla-Alvarado et al. (2013) [81], and seven more articles by Romero-Cedillo et al., Bretón-Deval et al., and Valdez-Vazquez, among others), (ii) accepted procedures of sizing and design in chemical ([90,91] McCabe et al. [92]) and biochemical engineering [93,94], (iii) the use of proxies (life cycle inventories in the Ecoinvent database) close to the BRF process in question, particularly when the process or stage was a very complex one [95] [the energy flows in the proxy were summed in each category, heat input and electricity input, and the results were expressed on our FU basis by the procedure that takes into account the FU of the proxy and our FU, similar to Section 2.1.4 of this manuscript (using factors ϕ and δ, Equations (3) and (4)], and (iv) a calculation technique that we called the “ping-pong strategy” in several cases.

The latter was used for selected processes in several stages of the biorefinery that require the estimation of heat losses from equipment to the environment, along with the electricity used for pumping and mixing. Calculations performed for continuously operated process units were assumed to be in a steady state. All equipment was assumed to be fitted with appropriate thermal insulation, 1 to 2 inches thick, and with correct maintenance. The interested reader is referred to “Annex S15. Calculations of energy flows of the biorefinery and incineration plant” in the Supplementary Materials, where a detailed description of the “ping-pong” strategy can be found, along with the equipment (such as ambient heat loss of the equipment per FU) or, on volumetric indicators (e.g., electricity per FU for mixing, based on the power intensity factor β), the application of the strategy for the calculation of energy flows of the H₂-producing bioreactor in Stage H, several equations used for diverse energy flows, a compilation Table S15.1 of energy flows in the biorefinery, and more information on energy flows in the incineration plant.

The energy balances and energy performance evaluation of the IER were based on the energy generated in the system, along with the latent heat, sensible heat, heat losses, and the net heat delivered to the boiler from the combustion process [72,73,96]. The LCI of the IER is presented in Annex S2 of the Supplementary Materials, and the calculation procedures are described in Annex S3. Technological harmonization of the Mexican conditions was performed. The first modification to the LCI was to adjust the FU. The second modification was to correct the factors that determine the allowable emission limits for Mexico and the EU (European Emissions Commission) [71,97]. The adequacy procedure rational can be consulted in Annex S19. The technological adequacy factors used here can be found in Annex S4 of the Supplementary Materials. Data quality analysis was performed for the BRF (Annex S5, Table S5.1) and IER (Annex S5, Table S5.2).

2.1.3. Environmental Impact Assessment

The LCA was conducted according to the steps described in ISO 14040 and ISO 14044 (International Organization for Standardization) [47,48]. SimaPro version 9.0.0.35, Analyst Professional software was used [70]. The PEIs, also known as characterization potential environmental impacts, were first obtained. The method selected to evaluate the LCI was ReCiPe 2016 hierarchical intermediate point v.1.03, which evaluates eighteen categories of PEI. The emissions were classified according to the eighteen PEIs analyzed and characterized according to their respective equivalent units. The PEIs analyzed can be reviewed in Annex S6 of the Supplementary Materials of this article.

The included PEIs were global warming (GW), ionizing radiation (IR), scarcity of fossil resources (SOD), ozone formation terrestrial ecosystems (OFTEs), ozone formation, human health (OFH), deterioration of the ozone layer, formation of fine particles (FPs), carcinogenic human toxicity (HCT), non-carcinogenic human toxicity (HNCT), terrestrial acidification (TA), ecotoxicity to freshwater (FWEc), marine ecotoxicity (MEc), land use (LU), scarcity of mineral resources (SMR), and water consumption (WC).

Environmental impact normalization was used to compare the results and thus determine their significance. According to the EDIP 1997 methodology, normalization allows for the determination of the relative proportions of the indicators of each category with respect to the reference values (current values) [98,99,100,101] as well as the comparison the magnitudes of the eighteen environmental impacts between the LCAs of different processes or technologies (the latter provided that the FU is equivalent in all cases) [70]. Reference information can refer to a specific community, person, or other system over time.

Normalization was global or worldwide, the only options available in the version of SimaPro that we used. These results are presented as normalized potential environmental impacts (NPEIs). In accordance with this concept, the NPEIs have the same units (person*yr/FU). The definition and derivation of the units of the NPEIs are included in Annex S7 of the Supplementary Materials. The important conclusion is that the values of NPEI_j (j = 1, 2, …, 18) can be summed, subtracted, and divided because the eighteen NPEI_j have the same simple units. This feature allows them to be combined in the search for more general overall ES indices or metrics.

2.1.4. Interpretation of Results

This section presents the results, formulation of conclusions, and recommendations of this study. The interpretation considers data verification, assumptions, system specifications, the methodology used, calculation models, the analysis of process contributions to impact categories, and sensitivity and uncertainty analyses [47,95]. Following the steps described below, we compared PEIs with reported BRFs and biowaste incineration cases. To determine the most sustainable technology for OFMSW treatment, we used the ES indices α, σ, and τ, as described in Section 2.2.

First, we reviewed the reported cases of BRFs and IER for biomass or similar wastes, and the PEIs of the BRF-GBAER and our IER were compared accordingly. The composition of the feedstock, the bioenergy and bioproducts produced, and the environmental performance in terms of GWP, eutrophication, acidification, and human toxicity were considered (if reported). Selected cases of biowaste BRFs or similar biowaste IERs were selected from the literature. The PEI_js were harmonized to the FU and reference flow in the BRFs for each technology according to Equations (1)–(5), as described below.

The harmonization procedure can be summarized as follows: First, the selected PEI of a BRF from a case in the literature (subscript a) was compared with the corresponding PEI of our BRF (reference or subscript b). The environmental impact must be expressed on the same basis to make the comparison objective. Therefore, the PEIs were harmonized as described below. The PEI_j obtained from a BRF from the literature is symbolized as PEI_j,a, where the subscript a reflects the membership of this BRF (BRF_a).

This BRF_a will have its own FU_a and feed waste reference flow m_FUa, with the latter having its own dry matter content TS_a (TS, total solids). Our BRF is the reference or base BRF and uses the subscript b (BRF_b); it has its own FU_b and feed waste reference flow m_FU,_b, and this feed has its own dry matter content TS_b.

The procedure converts the PEI_j,a into a modified impact PEI_j,a,mod, which is expressed using the conditions of the BRF_b (i.e., using FU_b and the corresponding dry matter). The harmonization procedure is divided into two phases: first, a correction for the different reference flows, represented by the factor ϕ, and second, another correction for the different dry matter contents in the reference flows, represented by the factor δ.

$${PEI}_{j,a,mod}={PEI}_{j,a}\ast \varnothing \ast \delta$$

(1)

It is worth clarifying that PEIs of all types and each of our two BRFs have specific units. Typically, these units are physical (P), chemical (C), biological (B), or toxicological (T) units, referred to as the corresponding FU. For example, the non-carcinogenic human toxicity PEI has units of kg 1,4-DCB e per FU. Thus, the units for a PEI_j,a are represented by the symbol φ, as shown in Equation (2):

$${\phi (PEI}_{j.a})={\displaystyle \frac{Units Phys-chem-bio-tox}{{FU}_b}}={\displaystyle \frac{PCBT}{{FU}_b}}$$

(2)

The ϕ factor is given by Equation (3):

$$\varphi =\left(m_{FU,b}\right)\left({\displaystyle \frac{{kg}_{feed,b}}{{FU}_b}}\right)\ast {\displaystyle \frac{1}{\left(m_{FU,a}\right)\ast \left(\frac{{kg}_{feed,a}}{FU,a}\right)}}$$

(3)

The δ correction is shown in Equation (4):

$$\delta ={\displaystyle \frac{{TS}_b}{{TS}_a}}\left({\displaystyle \frac{\frac{kg dry matter, b}{ kg feed,b}}{\frac{kg dry matter,a}{ kg feed a}}}\right)$$

(4)

Substituting the expressions for ϕ and δ into Equation (1), we obtain Equation (5) below:

$${PEI}_{j,a,mod}\left({\displaystyle \frac{ PCBT}{{FU}_b}}\right)={PEI}_{j,a}\left({\displaystyle \frac{ PCBT}{{FU}_a}}\right)\ast \left[{\displaystyle \frac{m_{FU.b}}{m_{FU,a}}}\right]\left({\displaystyle \frac{\frac{ kg feed,b}{FU,b}}{\frac{kg feed,a}{FU,a}}}\right)\ast {\displaystyle \frac{{TS}_b}{{TS}_a}}\left({\displaystyle \frac{\frac{kg dry matter,b}{kg feed, b}}{\frac{kg dry matter,a}{ kg feed, a}}}\right)$$

(5)

Now, ${PEI}_{j,a,mod}$ is comparable with the corresponding PEI_j,b.

2.2. Comparison of Environmental Impacts and Sustainability of the Biorefinery and Incineration in Our Work

This subsection deals with specific and meaningful indices and methods for comparing the environmental impacts and sustainability of both technologies:

(i) The index μ for direct comparison of a given characterization PEI between the two technologies. (ii) The calculation of NPEIs, which are the cornerstone of the new sustainability indices. (iii) The development of the new sustainability indices α, σ, and τ, which were used to determine the environmental sustainability of both technologies. These new indices are based on the NPEIs calculated in (ii). Selected information on detailed definitions and calculation procedures has been included in Annexes 7 (for the calculation of the NPEIs) and 8 (for the calculation and application of the α, σ, and τ indices).

Also, we have included a discussion and comparison of our indicators with selected ES indicators, such as those based on exergy analysis, Eco-Indicator 99, and End-Point Recipe 2016 in Annex S17 of the Supplementary Materials. Readers are encouraged to read the complete Annex because these paragraphs are only the tip of the iceberg for reviewing and discussing ES indicators. We conclude that the GBAER indicators, based on midpoint NPEIs ReCiPe 2016 and posterior own definitions and calculations (indicators α, σ, and τ), can be helpful and easy to apply at this stage of our research for the rapid determination and characterization of the environmental sustainability of environmental technologies or processes, and likely of industrial technologies or processes as well. Our indicators will characterize the comparison of ES with the concepts of dominance and restrictions, a feature not included in other indicators to the best of our knowledge.

The comparative analysis of the IER and the BRF followed the steps previously presented in Section 2.1.3, comparing selected (i.e., the most important) PEIs and NPEIs: GW, FWEc, MEc, HNC, and HCT. In order to compare the values of a given category of potential environmental impact, PEI_j, between two processes or technologies (i.e., between BRF and IER, for the same FU), we propose the variable μ_j below (Equation (6)), which indicates which technology is more polluting in this “jth” category of impact. The index μ_j is dimensionless and expressed as a percentage. The variable ∆μ_j (%) gives the percentage increase in the maximum impact with respect to the minimum impact, as defined by Equation (7) below.

$${\mathsf{\mu }}_{\mathrm{j}}={\displaystyle \frac{\max ({PEI}_{j,IER}, { PEI}_{j,BRF})}{\min ({PEI}_{j,IER}, { PEI}_{j,BRF}) }}\ast 100; j=1, 2, 3, ...., 18$$

(6)

$$\mathsf{\Delta }{\mathsf{\mu }}_{\mathrm{j}} (\%)={\mathsf{\mu }}_j-100$$

(7)

where max indicates the maximum value of the two values between parentheses, min indicates the minimum value of the two values between parentheses, PEI_j,IER is the value of the potential environmental impact category _j for the IER, and PEI_j,BRF is the value of the potential environmental impact category _j for the BRF.

However, even if all values of μ_j are calculated, it is impossible to obtain expressions for the total ES and to compare the two technologies based on the latter, because the PEI units are heterogeneous and cannot be mixed. Therefore, it is impossible to process them, i.e., to sum and divide the μ_j. Consequently, it is difficult or impossible to define an overall index that determines the most environmentally sustainable technology based on a combination of μ_j. Thus, normalization according to the ISO standards was followed, allowing us to compare different environmental impacts by relating the environmental impact values to a standard reference. The normalized impacts are abbreviated as NPEI_j, and their units are (person*yr)/FU, where j is the subscript representing the j-th impact. The determination of NPEIs in our work is based on the eighteen characterization PEIs determined by LCA (in units of characterization impact per FU) and the unit loads (amount of a given impact/[person*year]). These unit loads are the inverse of the corresponding normalization factors F_n in Recipe Reports 2016; see Tables S7.1 and S7.2, Annex S7, Supplementary Materials. This definition is essential but is almost hidden as a footnote in the Simapro Manual [70]. We use F_n for the normalization value and U for the unit environmental load instead of N, Equation (S7.2), Annex S7.

The normalization factors were also confirmed by dividing each NPEI reported by SimaPro by the corresponding characterization PEI reported by SimaPro software version 9.0.0.35, Analyst Professional, Table S7.2. Annex S7 in Supplementary Materials. The normalization factors from both methods coincided. Since each PEI has the units (phys–chem–biol–tox units/FU) and the normalization factors have units (person*year/phys–chem–biol–tox units), the units of the NPEI are (person*year/FU); the phys–chem–biol–tox units are canceled. This is very clear and important. The units of NPEIs show that NPEIs are not dimensionless. Also, the units of NPEIs stress that a key feature is the FU, and a comparison between NPEIs of different processes or technologies can be made if and only if the functional unit is the same or harmonized. One significant property of NPEIs is that they all have the same units for the same FU, and then the NPEIs can be added, subtracted, and divided.

So, the sum of the eighteen normalized potential impacts is the normalized index α of a technology or process (Equation (S7.8) in Annex S7, Supplementary Materials). Therefore, if α is higher, then environmental sustainability is lower. Thus, the sum of the NPEIs makes it possible to compare the environmental impacts of different technologies. Based on this concept, the GBAER Environmental Sustainability Indices were also generated and are presented in the following paragraphs. The GBAER developed the Environmental Sustainability Indices σ and τ to compare different technologies and determine their environmental sustainability (Equations (S8.2) and (S8.3)) described in Annex S8 of the Supplementary Materials). These indices allow for the analysis of the results obtained by the SimaPro software and the positioning of the results from comparing the technologies on the σ-τ plane (Figure S8.1 in Annex S8). The α index allows us to obtain an overall or global result of the environmental sustainability of a technology and helps to determine which technology is more environmentally sustainable among a set of technologies with the same FU.

However, the α index does not allow us to distinguish whether environmental sustainability is dominant in all or most of the impact categories analyzed, or whether there are categories in which the evaluated technology has more significant individual environmental impacts (restrictions) even though the overall sustainability is favorable. The σ-τ plane allows us to distinguish the most environmentally sustainable technology, and it also indicates whether ES is dominant or with restrictions. The σ-τ plane is divided into four quadrants (Figure S8.1 in Annex S8). Quadrants I and III indicate the dominant environmental sustainability of the assessed and reference technologies, respectively. The point σ-τ is located in Quadrant I when the evaluated technology is more environmentally sustainable. If the reference technology is more environmentally sustainable in a dominant way, the point σ-τ is in Quadrant III. Quadrants II and IV contain intermediate cases whose better ES entails restrictions.

3. Results and Discussion

This section presents the following results: (i) The energy performance of the HMEZS-NN BRF and IER is described and compared in terms of gross electricity, net electricity, and net thermal load (heat) produced. (ii) Then, the environmental performance results for the BRF are described in terms of the PEIs (characterization and normalized impacts) found by the LCA. Both the magnitudes and the contributions of the process stages and critical processes and materials to the environmental impacts are highlighted and commented on. (iii) The environmental performance of the IER is described similarly to (ii). (iv) Finally, we present the determination and comparison of the environmental sustainability of both technologies based on novel indices developed by our Group and aligned with the LCA and ISO standards for LCA. As a preamble, the NPEIs previously determined in (ii) and (iii) are compared.

3.1. Energy Performance of the Biorefinery and Incineration in This Work

The summarized energy performances of our BRF and IER are presented in

Table 2. Energy performance of the biorefinery HMEZS-NN per functional unit in this work.

Stage	Input		Output
	Heat (MJ)	Electricity (kWh)	Heat (MJ)	Electricity (kWh)
Conditioning		8.88
H	392.60	3.47	0.00	0.00
M	70.42	2.35	0.00	0.00
Biogas H: Purification fuel cells	0.00	22.93	0.00	40.25
Biogas M: Purification		36.58	799.82	407.31
E	97.58	0.57
Z	707.96	79.78
S (acid)	36.69	44.63
NN Purification	0.00	3.84	70.04	42.80
Subtotal	1305.25	203.03	869.86	490.36
	Heat (MJ)		Electricity
			(kWh)	(MJ)
Total	435.39		287.33	1034.39
Total energy (Electricity–Heat)	599.00 MJ		166.39 kWh

and

Table 3. Energy performance per functional unit of incineration with energy recovery in this work.

Energy Amount	MJ	kWh
Heat of combustion	9981.1	2773
Latent heat	1797.3	499
Gross available heat	8183.8	2273
Heat losses (5%) from the combustion chamber	409.2	114
Heat losses in the incinerated waste (ashes)	13.7	4
Boiler heat losses (5% of energy inside the boiler 77,609.8 MJ)	388.0	108
Heat loss from flue gases (538 °C and 100% excess air)	6.5	2
Other heat losses (heat loss in piping and fittings, cracked insulation, steam leaks, heating makeup water, etc.)	300.0	83
Net available heat	7066.4	1963
Electricity produced at 20% efficiency	1413.3	393

, respectively. The BRF is self-sufficient in terms of total energy, with a net value of approximately 600 MJ (167 kWh) from a mixture of high-voltage and low-voltage electricity. This represents the electricity obtained from the cogeneration of biological methane and that obtained from bio-hydrogen-powered fuel cells (Table 2). The BRF is 100% self-sufficient in terms of electrical energy, but this only provides 60% of the heat required by the BRF. This 40% heat deficit can be satisfied with part of the significant excess of electrical energy. This leaves a net energy production of 167 kWh (599 MJ), as shown in Table 2.

Although the BRF is intended to be energy self-sufficient, during some periods of its operation, such as the startup period when the methanogenic reactor’s performance is not well-established, energy requirements from the technosphere could arise. External heat or electricity inputs from conventional energy services could meet these needs. For clarification, in the SimaPro LCI, the electricity consumed by the processes was replaced by electricity produced by the BRF. Approximately half of the required (consumed) thermal loads in the BRF were replaced by heat generated by the BRF. The BRF also produces a menu of bioproducts. On the other hand, the IER generates only two energy products: 5653 MJ of heat and 393 kWh of electricity per FU, i.e., heat for building or industrial heating, and net electricity, respectively (

Table 4. Characterization of potential environmental impacts, identification of significant contributions by process stages and materials in the biorefinery in this work (HMEZS-NN), normalized potential environmental impacts, and process stages’ contributions.

Impact Category	Units	Stage H a	Stage M	Stage E	Stage Z	Stage S	Stage NN	Total Impact
Global warming	PEI contribution (%)	13.44	16.95	3.59	18.07	10.04	37.92	100.00
	PEI (kg CO2e/FU)	126.64	159.73	33.79	170.34	94.63	357.39	942.52
	NPEI (person*yr/FU)	0.02	0.02	0.004	0.02	0.01	0.05	0.12
Stratospheric ozone depletion	PEI contribution (%)	40.99	39.73	0.19	8.04	3.79	7.26	100.00
	PEI (kg CFC11 e/FU)	9.5 × 10−4	9.2 × 10−4	4.4 × 10−4	1.9 × 10−4	8.8 × 10−5	1.7 × 10−4	2.3 × 10−3
	NPEI (person*yr/FU)	0.02	0.02	0.00	0.003	0.001	0.003	0.04
Ionizing radiation	PEI contribution (%)	8.39	2.83	3.12	19.53	34.97	31.16	100.00
	PEI (kBq Co-60 e/FU)	1.75	0.59	0.65	4.08	7.30	6.51	20.88
	NPEI (person*yr/FU)	0.004	0.001	0.001	0.008	0.02	0.01	0.04
Ozone formation, human health	PEI contribution (%)	6.62	4.99	44.20	10.61	6.34	27.23	100.00
	PEI (kg NOx e/FU)	0.18	0.14	1.21	0.29	0.17	0.75	2.74
	NPEI (person*yr/FU)	0.009	0.007	0.06	0.01	0.008	0.04	0.13
Fine particulate matter formation	PEI contribution (%)	23.06	36.90	1.36	8.33	6.22	24.14	100.00
	PEI (kg PM2.5 e/FU)	0.77	1.22	0.05	0.28	0.21	0.80	3.32
	NPEI (person*yr/FU)	0.03	0.05	0.002	0.01	0.008	0.03	0.13
Ozone formation, terrestrial ecosystems	PEI contribution (%)	5.31	3.97	55.01	9.01	5.06	21.65	100.00
	PEI (kg NOx e/FU)	0.19	0.14	1.91	0.31	0.18	0.75	3.48
	NPEI (person*yr/FU)	0.01	0.008	0.11	0.02	0.01	0.04	0.20
Terrestrial acidification	PEI contribution (%)	28.32	50.81	0.57	4.19	2.22	13.89	100.00
	PEI (kg SO2 e/FU)	5.40	9.68	0.11	0.80	0.42	2.65	19.05
	NPEI (person*yr/FU)	0.13	0.24	0.003	0.02	0.01	0.07	0.47
Freshwater eutrophication	PEI contribution (%)	9.10	2.58	2.11	19.67	12.95	53.60	100.00
	PEI (kg P e/FU)	0.03	0.007	0.006	0.05	0.04	0.15	0.27
	NPEI (person*yr/FU)	0.04	0.01	0.01	0.08	0.06	0.23	0.42
Marine eutrophication	PEI contribution (%)	3.02	0.93	0.49	74.25	4.90	16.42	100.00
	PEI (kg N e/FU)	2.0 × 10−3	6.3 × 10−4	3.3 × 10−4	0.05	0.003	0.01	0.07
	NPEI (person*yr/FU)	4.4 × 10−4	1.4 × 10−4	7.1 × 10−5	0.01	0.001	0.002	0.02
Terrestrial ecotoxicity	PEI contribution (%)	8.23	2.78	32.71	29.87	7.94	18.47	100.00
	PEI (kg 1,4-DCB e/FU)	68.50	23.10	272.23	248.57	66.07	153.69	832.16
	0.066	0.02	0.26	0.24	0.06	0.15	0.80	0.07
Freshwater ecotoxicity	PEI contribution (%)	8.06	2.23	3.38	34.69	9.87	41.77	100.00
	PEI (kg 1,4-DCB e/FU)	0.87	0.24	0.36	3.73	1.06	4.49	10.74
	NPEI (person*yr/FU)	0.71	0.20	0.30	3.01	0.86	3.66	8.76
Marine ecotoxicity	PEI contribution (%)	8.73	2.44	2.11	31.39	10.65	44.68	100.00
	PEI (kg 1,4-DCB e/FU)	1.22	0.34	0.30	4.41	1.50	6.27	14.04
	NPEI (person*yr/FU)	1.18	0.33	0.29	4.27	1.45	6.08	13.60
Human carcinogenic toxicity	PEI contribution (%)	3.74	1.15	1.54	59.95	6.50	27.12	100.00
	PEI (kg 1,4-DCB e/FU)	1.45	0.44	0.60	23.14	2.51	10.47	38.60
	NPEI (person*yr/FU)	0.52	0.16	0.22	8.35	0.91	3.78	13.93
Human non-carcinogenic toxicity	PEI contribution (%)	10.44	2.96	2.22	23.66	11.23	49.50	100.00
	PEI (kg 1,4-DCB e/FU)	28.30	8.02	6.03	64.15	30.44	134.22	271.14
	NPEI (person*yr/FU)	0.19	0.05	0.04	0.43	0.20	0.90	1.82
Land use	PEI contribution (%)	5.76	0.74	0.27	87.25	1.07	4.91	100.00
	PEI (m2a crop e/FU)	2.45	0.32	0.11	37.17	0.45	2.09	42.60
	NPEI (person*yr/FU)	3.9 × 10−4	5.1 × 10−5	1.8 × 10−5	6.0 × 10−3	7.4 × 10−5	3.4 × 10−4	6.9 × 10−3
Scarcity of mineral resources	PEI contribution (%)	0.93	0.05	0.22	97.22	0.63	0.96	100.00
	PEI (kg Cu e/FU)	1.7 × 10−2	8.9 × 10−4	3.9 × 10−3	1.73	0.01	0.02	1.78
	NPEI (person*yr/FU)	1.4 × 10−7	7.5 × 10−9	3.2 × 10−8	1.4 × 10−5	9.3 × 10−8	1.4 × 10−7	1.5 × 10−5
Fossil resource scarcity	PEI contribution (%)	8.92	2.59	8.33	25.53	11.85	42.78	100.00
	PEI (kg oil e/FU)	18.11	5.27	16.91	51.83	24.05	86.85	203.01
	NPEI (person*yr/FU)	0.01	0.01	0.02	0.05	0.03	0.09	0.21
Water consumption	PEI contribution (%)	22.39	8.25	10.55	41.78	16.64	0.40	100.00
	PEI (m3/FU)	2.05	0.76	0.97	3.83	1.52	0.04	9.16
	NPEI (person*yr/FU)	0.01	0.003	0.004	0.01	0.01	1.4 × 10−4	0.03

References: [47,48,70]. ^a Includes conditioning stage. Notes: The conditioning stage was included in Stage H. NPEI was calculated using normalization factors in Ecoinvent. Normalization W ReCiPe 2016 Midpont (H) V1.03 World (2010) H.

). The net thermal efficiency of electricity generation is ~20%, which is typically low due to moisture in the waste and losses. Most of the energy produced by the IER is in the form of heat.

The electricity production in our IER was higher than that reported by Lou et al. [98], who obtained 210 kWh/t of waste. This result could be due to differences in feedstock quality (i.e., lower moisture content of OFMSW in our case) and higher net thermal efficiency. Regarding the external energy consumption in our IER, the primary fuel used was natural gas, of which 186 MJ is required (Annex S2, Table S2.1) and is used in the FGT (the DeNOx system) as auxiliary fuel during the startup of the incinerator after maintenance episodes (Ecoinvent Database 3) [69].

The net electricity generation in the IER was ~2.4 times higher than in the BRF (393 and 166 kWh per FU, respectively). In contrast, the net heat production of the IER was much higher than that of the BRF (5653 and 0 net MJ per FU, respectively). Despite the lower energy production of the BRF, the latter could be advantageous from other points of view, such as the production of a significant surplus of electricity and VAP generation. In fact, the IER produces only two energies and no VAPs. In the contrast, the BRF allows the production of two bioenergy sources, several bioproducts, and VAPs (concentrates of organic acids, industrial enzymes, and BNBPs). In addition, the BRF is self-sufficient in electricity production, with a surplus of electricity for sale or export.

3.2. Environmental Performance of the Biorefinery

This subsection describes the results of the environmental performance of the BRF in terms of eighteen PEIs (characterization and normalization) found by the LCA; it also identifies and discusses the contributions of the BRF’s process stages, key processes, and materials to the environmental impacts. Finally, we compare the environmental results of our BRF with those of reported cases of BRFs found in the open literature, in order to identify similarities and differences in performance and discuss their likely reasons. The main results of the PEIs for the BRF are presented in Table 4, and in more detail in Annex S10 of the Supplementary Materials, where key issues such as total impacts (eighteen) and characterization for each environmental impact category are presented, along with other information to be used in the following subsection (normalization factors and the NPEIs).

3.2.1. Environmental Impact Description and Contributions

The Z and NN stages contributed significantly to eleven environmental impact categories, while the NN influenced six of the eleven impacts. Stage Z influenced five PEIs, with the highest contributions in MEc with 74.2% and HCT with 59.9%. The NN stage showed the most significant contributions to GW, FWEu, FWEc, MEc, and HMCT with 38, 53, 41.7, 44.6, and 49.5%, respectively. The potential GW emissions from the NN stage were estimated at 357 kg of CO₂e, and the most significant contribution corresponded to the use of activated carbon in this stage for the production of BNBPs. On the one hand, the production of activated carbon (Table S10.8, Annex S10 in Supplementary Materials) is energy-intensive, with an input of 1.8 kWh per kg of activated carbon (Table S9.4, Annex S9, in Supplementary Materials), which leads to the release of large amounts of GHGs, mainly CO₂ (5.3 kg of CO₂ e per kg of activated carbon). On the other hand, the use of activated carbon and the associated treatment of coal mining waste is responsible for the highest emissions in the categories HNCT, GW, and FWEu, as well as eutrophication and ecotoxicity to fresh and seawater (Table 4 and Tables S9.6 and S9.7 of Annex S9 in the Supplementary Materials).

The NN-stage emissions of As (V) and Zn (II) to water likely contribute to PEIs FWEc and MEC. The emissions to water per kg of “treated coal mining spoil” were found to be 6.25 × 10⁻⁷ and 1.90 × 10⁻⁵ kg of As and Zn, respectively (Annex S9, Table S9.6). Contributions from stage Z predominated in the PEIs HCT, MEc, land use, and scarcity of mineral resources. The category of greatest concern was HCT, with the most significant contribution being 60% and 23,139 kg of 1,4-DBC e (Table 4). The SimaPro analysis revealed that this contribution is related to the treatment of H₃PO₄ waste (Tables S9.2 and S9.3, Annex S9 in the Supplementary Materials), which is a residue from the production and use of Na₃PO₄ used in the formulation of the culture medium for the enzymes in this stage (Table S9.1, Annex S9 in the Supplementary Materials).

The results of the PEIs were normalized, converted to NPEIs, and summarized as shown in Table 4. The normalization procedure was as described in the Methodology (see Section 2.2) and in Annex S7 (Table S7.1) in the Supplementary Materials. For the BRF, the highest NPEIs were HCT and MEc with 13.93 and 13.60 (person*yr)/FU, respectively, followed by HNCT with 1.82 (person*yr)/FU. FWEc and FWEu were 8.75 and 0.42 (person*yr)/FU, respectively. Regarding the emissions to the atmosphere, 0.118 (person*yr)/FU was obtained for GW—a relatively low value. The NPEIs showed the same tendency as the PEIs, where the NN and Z stage predominated and had the most remarkable effects on HCT, MEc, and FWEc. The NPEI in stage Z was 19.46 (person*yr)/FU, whereas in the NN stage, it was 16.22, i.e., 42 and 35%, respectively. Stage Z had the most outstanding contribution to the HCT with 60% (Table 4), representing the highest NPEI, with 13.93 (person*yr)/FU. The latter is due to the use of phosphates in the formulation of the culture medium in this stage, which is related to the environmental impact due to the residues of the H₃PO₄ treatment (Annex S9, Tables S9.1–S9.3 in the Supplementary Materials). Stage Z also contributes to the MEc category with 34.69% (Table 4). This contribution is related to a variety of coal mining processes.

With respect to the GW category, the NN stage has a predominant contribution of 38% (Table 4). The most prominent contribution is associated with the use of activated carbon of fossil origin (Table S10.8, Annex S10 in the Supplementary Materials), with a similar discussion presented in this section. Given the importance of the NN in the highest NPEI and GW, one can speculate that the use of activated carbon in the BRF merits further study.

The LCA analysis strongly suggests that using activated carbon can become a negative burden in terms of the environmental impact of the BRF, especially for the NN stage. As the LCA is iterative, it is helpful to consider replacing activated carbon of fossil origin with activated carbon of plant origin. It would be worthwhile to conduct such a study—if a benefit is found, the impacts of key BRF stages, processes, and products on key NPEIs could be reduced and a modified BRF would be more sustainable.

Regarding the role of phosphate waste from phosphate use in the Z stage and its influence on key NPEIs, one recommendation is to determine whether the amount of Na₃PO₄ added to the Z medium exceeds the nutrient P requirement. If so, the Na₃PO₄ is likely fulfilling a pH control role rather than a nutrient role. Therefore, another P-free buffer system could replace some of the total Na₃PO₄. This knowledge would allow for the quantification of the new (lower) amount of Na₃PO₄ to be used in the LCI. A second LCA should then be performed to determine if the environmental impacts associated with the reduced use of Na₃PO₄ have decreased.

3.2.2. Comparison with Reported Cases of Biorefineries

This section presents four reported cases of BRFs from biomass or similar wastes, and the BRF-GBAER PEIs have been compared where possible.

Table 5. Comparison with selected Life Cycle Assessment studies on organic waste-based biorefineries—harmonized functional unit.

Feed Waste Composition	Impact Assessment (PEI) Categories					Remarks/Bioenergy, Bioproducts	Impact Assessment Method and Software	Ref.
	Functional Unit/Harmonization Factor	Global Warming (kg CO2e)	Acidification (kg SO2e)	Freshwater Eutrophication (kg Pe)	Human Toxicity (Cancer and Non-Cancer kg 1,4-DCBe)
OFMSW: paper, kitchen waste 65.7%, garden waste 26.7%, other 7.6% Total solids 39.7% VS 27.5%	OFMSW 145,000 t/yr Harmonization factor 7.43 × 10−6	−3.75 × 106 a,*	2.39 × 105 a/ 1.78 b	4.78 × 105 a/ 3.55 b	2.16 × 107 a/ 160.58 b	biomethane, fertilizers and PHAs (polyhydroxyalkanoates)	Midpoint CML-IA baseline V3.02/EU25	Rossi et al. [102]
Biowaste Sewage sludge 25% TS of the total VS VS 27.5%	1 kg produced polymer polyhydroxyalkanoates (PHA) Harmonization factor 52.46	10 (kg CO2 e /kg polymer) a 524.66 b	c	c	d	e Depletion of fossil 80 (MJ eq) Caso Trento Polihidroxyalkanoates	Environmental footprint 3.0	Bassi et al. f [104]
Raw food waste	1 kg raw food waste Harmonization factor 1324	5.5 kg CO2 e a 7282 kg CO2 e b	c	c	c	bioethanol, biomethane, and oil	SimaPro 8.5.2.0 IMPACT 2002+	Soleymani Angili et al. [105]
Food waste: Coked rice 47%, boiled vegetables 16%; vegetables peel 18%; spoiled vegetables and fruits 6%; eggs and meet 13%	1 kg of Bio-H2 production Harmonization factor 637.20	−1.203 a/−766.61 b	3.90 × 10−3 a/2.485 b	2.58 × 10−5 a/0.0162 b	c	Bio-H2 VOA Acuatic ecotoxicity 5.8392 (kg TEG water)	Impact 2002 Endpoint method	Sarkar et al. [103]
Food waste 60% Paper 40%	1 t OFMSW 35% humidity	942.52	19.05	0.274	309.74	130 kWh	SimaPro	This work

VOA (volatile fatty acid), mPt (milli Point), ^a value reported by the author, ^b value calculated with the harmonization factor, ^c no data available, ^d freshwater ecotoxicity, human non-carcinogenic toxicity, human toxicity potential (HTP), human carcinogenic toxicity. Human non-carcinogenic toxicity and freshwater ecotoxicity showed the highest impact values; however, they were not included in the calculated single indicator because the authors considered that they had considerable uncertainty and were not robust enough, ^e 1 kg of PHA produced from urban organic waste in five geographical regions was modeled; six alternatives with varying food waste and sewage, and eight framework scenarios, ^f this work does not meet the main feature of a BRF, because it produces just one bioproduct or bioenergy (recall that a BRF is typically a multi-process and multi-product plant). Nevertheless, the authors call the process a biorefinery. * It is not clear whether the global warming PEI suffers from double accounting (double subtraction), so we have refrained from making the comparison (cases of Rossi et al. and Sarkar et al. [102,103]).

shows the composition of the raw material, the bioenergy and bioproducts generated, and the environmental performance in GW, eutrophication, acidification, and human toxicity (whenever reported) for each case. The PEIj were harmonized to the FU and reference flow in the BRFs according to Equations (1)–(5) presented in the methodology at the end of Section 2.1.4. All cases analyzed in Table 5 show the characterization PEI results for the GW category, but not necessarily for the rest of the impact categories. Furthermore, in some cases, it is unclear whether the GW PEI suffers from double accounting (double subtraction), so we have refrained from making the comparison (cases of Rossi et al. and Sarkar et al. [102,103]).

Regarding the production of bioenergy and bioproducts, the BRF that corresponds to Rossi et al. [102] produces biomethane, fertilizer, and polyhydroxyalkanoates (PHA). In contrast, the BRF reported by Bassi et al. [104] produced only one product (PHA). Therefore, it is questionable whether we should call this process a BRF. Concerning the latter, Fava et al. [24] of the European Federation of Biotechnology stated the following:

“Biorefineries could be described as integrated bio-based industries that use a variety of technologies to produce products such as chemicals, biofuels, food and feed ingredients, biomaterials, fibers, and heat and power, with the aim of maximizing added value along the three pillars of sustainability (environment, economy, and society)”.

This means that a BRF is, among other things, a multi-process and multi-product plant. For their part, Soleymani Angili et al. [105] chose to produce biomethane, bioethanol, oils, and fats, while Sarkar et al. [103] reported on the production of biohydrogen and VOAs. The PEI of the BRF HMEZS-NN corresponding to GW was 942.5 kg CO₂ e/FU_b. Bassi et al. [104] found a value of the same order (slightly lower) of 524.66 kg CO₂ e/FU_b.

It is assumed that the average heat consumption was 363 MJ/t of food waste and that this thermal load was 100%, covered by the internal production of biogas, without the aid of burning fossil energy from the technosphere, which could emit more CO₂e. Global warming (the only category reported by Bassi et al. [104]) is determined by direct (or avoided) atmospheric emissions of fossil CH₄ and CO₂. Although our case shows a GW impact of the same order of magnitude as that reported by Bassi et al. [104], the difference in values may be due to several factors. As mentioned above, our BRF produces H₂, CH₄, a concentrate of organic acids and low-molecular-weight solvents, enzymes, and BNBPs. Our BRF is complex and multi-stage, which can result in significant energy consumption, requiring energy supply for six process stages. Although heat recovery is achieved and the energy generated in stages H and M is reused, more heat is needed to supply the thermal load required in the last two stages, NN and Z. As mentioned in Section 3.1, the excess electricity from the BRF could be used to meet the thermal load requirements. If we take into account the energy savings due to the internal use of excess electricity in the BRF, then our GW PEI could easily show negative numbers. This would reflect a favorable outcome for our BRF in terms of GW impact. The NN stage also contributes primarily to GW due to processes related to the activated carbon used to form the BNBPs. The M stage also contributes to GW.

Soleymani Angili et al. [105] reported 7281 kg CO₂ e for the GW category, mainly due to electricity consumption and the production of nutrients/chemicals. Their BRF requires electricity supply for seven process steps. The electricity sources are mostly fossil, such as gas, lignite, and oil. In addition, several of the stages of their BRF require high energy consumption, such as grinding and drying in the process feed (Figure 1 in Soleymani Angili et al. [105]), and drying of extracted solids and alcoholic fermentation, which require electricity for their operation. The authors do not propose to use the generated CH₄ in the plant, which further promotes the release of CO₂ emissions from fossil fuels. In contrast, our BRF implements the in-plant use of CH₄ and H₂ to produce electrical and thermal loads used in the BRF itself, thereby reducing “imports” of fossil energy from the technosphere.

Rossi et al. [102] analyzed a BRF integrated with anaerobic digestion and composting that processes OFMSW to produce bioplastics (PHAs). They considered avoiding environmental impacts resulting from substituting bioenergy and bio-products (e.g., biomethane, compost, and PHA production). The approximate recovery is 25% by volume of biomethane, which is used as thermal energy within the plant as a fuel for steam generation. Compost was also considered as a substitute for fertilizer and polyhydroxyalkanoate (PHA) as a substitute for polyethylene terephthalate (PET).

Regarding HCT, the value reported in our BRF was about 309.74 kg 1,4-DCB e (Table 5, the sum of HCT and HNCT 38.60 + 271.14 kg 1,4-DCB e correspondingly) higher than those reported in the other cases. However, the results are still in the same order of magnitude (Rossi et al. [102] with about 160 kg of 1,4-DCB e). In our BRF, the NN stage dominates and contributes to the HNCT with 49.5% and 134.21 kg of 1,4-DCBe (Table 4). This contribution is derived from the emissions to water per kg of treated material from the activated carbon production process and waste treatment from this process (“coal mining tailings”). Stages M and Z also contribute in our case. The toxicity PEI 160.5 kg 1,4-DCB e observed by Rossi et al. [102] was mainly related to the extraction of PHA. MEc dominates the environmental load due to the large amount of supernatant to be treated. In our case, the Z and NN stages showed the most significant contributions, corresponding to the production of enzymes and BNBPs. Biorefinery complexity could have influenced differences between MEc values in the model BRF analyzed. On the one hand, Rossi et al.’s model [102] only had to quantify the PEI of three stages: anaerobic digestion, PHA extraction, and compost (three stages). In contrast, in our case, six stages are analyzed, which could explain our higher input of materials and energy requirements that could result in higher contributions to MEc.

The acidification category is only reported in two cases, Sarkar et al. [103] and Rossi et al. [102] (Table 5). The value obtained in our work was on the lower side of the range (0.5 kg SO₂ e), but the three results are in the same order of magnitude (2.456 kg SO₂ e by Sarkar et al. [106]; 1.007 kg SO₂ e by Rossi et al. [102]) (Table 5). Acidification was related to air emissions of sulfur oxides (SO₂), CO₂, and nitrogen (as NO_x and NH₃) expressed in kg SO₂ e [100]. In our work, CO₂ emissions are significant, as seen above, but direct emissions of kg SO₂ and nitrogen are relatively low. In the eutrophication category, our work showed a PEI of 0.274 kg PO₄ e, in the middle of the range of the three reported values (0.0162 kg PO₄ eq in Sarkar et al. [103] and 2.01 kg PO₄ e in Rossi et al. [102]). In this category, the differences between neighboring values are about one order of magnitude (Table 5). Rossi et al. [102] suggested that their relatively high eutrophication PEI could be due to environmental loads associated with digestate dewatering streams from their anaerobic digestion process.

3.3. Environmental Performance of Incineration with Energy Recovery

In this subsection, we present and discuss the environmental performance results for the IER in terms of 18 PEIs (characterization and normalized) found by LCA. Furthermore, we identify and discuss the contributions of the main products (heat and electricity) and critical processes and materials to the environmental impacts. Finally, the subsection compares the environmental results of our IER with those of reported cases of solid waste incinerators found in the open literature to identify agreements and disagreements in performance and discuss their likely reasons.

3.3.1. Description of Environmental Impacts and Contributions

Table 6. Contributions to potential environmental impacts in incineration with energy recovery by processes and materials.

Impact Category	Units	Total PEI	Process or Subsystem with the Highest Contribution (PCBT a units/FU)	Subtotal Due to Process (PCBT a Units/FU)	Process Contribution (%)	Heat (NPEI)	Electricity (NPEI)	Total NPEI
Global warming	kg CO2e	119.58	Electricity	98.04	81.99	0.007	0.014	0.021
Stratospheric ozone depletion	kg CFC-11e	0.004	Electricity	0.004	94.74	0.028	0.061	0.089
Ionizing radiation	kBq Co-60 e	0.87	Tailing, from uranium milling {GLO}| treatment | Cut-off, U	0.78	89.58	0.001	0.002	0.003
Ozone formation, terrestrial ecosystems	kg NOX e	1.35	Electricity	1.26	93.69	0.034	0.073	0.107
Ozone formation, human health	kg NOX e	1.35	Electricity	1.26	93.59	0.029	0.063	0.092
Marine ecotoxicity	kg 1,4-DCB e	68.74	Electricity	65.38	95.11	29.94	63.62	93.56
Human carcinogenic toxicity	kg 1,4-DCB e	27.17	Electricity	24.86	91.49	4.41	9.37	13.78
Terrestrial ecotoxicity	kg 1,4-DCB e	175.67	Electricity	138.38	78.77	0.076	0.162	0.238
Freshwater ecotoxicity	kg 1,4-DCB e	51.19	Electricity	48.70	95.14	18.75	39.85	58.6
Mineral resource scarcity	kg Cu e	0.03	Clay {RoW}| clay pit operation | Cut-off, U	0.01	25.96	1.22 × 10−7	2.59 × 10−7	3.8 × 10−7
Terrestrial acidification	kg SO2 e	0.59	Electricity	0.53	90.03	0.006	0.014	0.02
Fossil resource scarcity	kg oil e	3.84	Natural gas, high pressure {NO}| petroleum and gas production, offshore | Cut-off, U	0.55	14.19	0.002	0.004	0.006
Freshwater eutrophication	kg P e	0.20	Electricity	0.19	93.87	0.138	0.294	0.432
Human non-carcinogenic toxicity	kg 1,4-DCB e	1288	Electricity	1226.6	95.24	3.88	8.25	12.13
Land use	m2 × yr crop e	0.09	Wood chips, wet, measured as dry mass {SE}|hard-wood forestry, birch, sustainable forest management| Cut-off, U	0.01	12.51	6.45 × 10−6	1.37 × 10−5	2.0 × 10−5
Marine eutrophication	kg N e	0.08	Electricity	0.07	94.74	0.007	0.016	0.023
Fine particulate matter formation	kg of particles 2.5 µm	0.21	Electricity	0.19	89.49	0.004	0.008	0.012

^a In the physical–chemical–biological-toxicity units that correspond to each PEI.

shows the main contributions of materials and processes to the environmental impacts of IER. Energy production is the most significant contributor to PEI due to combustion and its emissions. These processes have a significant influence on ecotoxicity and toxicity impacts. The Supplementary Materials (Annex S11) provides the characterization of the other environmental impacts. The PEIs of the IER technology for processing 1 t of OFMSW and related emissions to the atmosphere were estimated to be 119.58 kg and 0.089 kg CFC11e for GW and SOD, respectively. The HCT and HNCT were 27.17 and 1288 kg 1.4-DCB e, respectively. Regarding the emissions to water, the PEIs were 0.20 kg Pe in FWEu, 0.076 kg Ne in MEu, and 51.19 and 68.74 kg 1.4-DCBe in FWEc and MEc, respectively.

Electricity production is the process that contributes the most to thirteen of the PEIs due to the combustion of OFMSW and auxiliary fossil fuels. It has a significant impact on the ecotoxicity and toxicity categories (Table 6). The most critical emission for this category, as reported by Maresca et al. [106], comes from the FGT subsystem. Furthermore, in our IER, emissions of 2.93 × 10⁻³, 0.137, and 0.212 kg were found for As, Cu, and Zn, respectively (Table S2.1, Annex S2 in Supplementary Materials). According to Hauschild et al. [107], toxic effects on freshwater ecosystems are determined by emissions of metals such as zinc (II) and copper (II), which are associated with heat generation. Freshwater ecotoxicity is also affected by FGD, probably due to the heavy metal content and extreme pH of the liquid streams in the subsystem. As a result, ash management is critical [73,108,109].

The electricity generation process also contributes 48.7 kg of 1,4-DCB e to the FWEc (Table 6). Emissions of 2.93 × 10⁻³, 0.137, 0.212, 0.0433, and 0.0426 kg/FU to the water were found for As, Cu, Zn, Ni, and Br, respectively (Table S2.1, Annex S2 in the Supplementary Materials). As mentioned above, these emissions are toxic to freshwater ecosystems. In our case, we found that the FGT process affects the MEc with 65.38 kg of 1,4-DCB e (Table 6), mainly due to the emission of heavy metals to water, with values of 4.77 × 10⁻³, 0.0433 and 0.212 kg/FU, for Cr, Ni, and Zn, respectively (Table S2.2, Annex S2 in Supplementary Materials). The HCT shows a total of 24.86 kg of 1,4-DCB e (Table 6) probably due to emissions of dioxins (6.18 × 10⁻¹⁰ kg), Cr(VI) (4.76 × 10⁻³ kg), As 1.65 × 10⁻² kg, and hydrogen chloride (1.85 × 10⁻² kg/FU) to the air (Table S2.1, Annex S2 in the Supplementary Materials), which according to Cole-Hunter et al. [109], have a significant impact and are highly relevant for this category. Incinerator installation equivalent to 4.64 × 10⁻⁷ p (piece), the heat emitted by the discharge of slag (33 kg) and fine ash (3.2 kg), and calcium oxide (lime) used in the cleaning of combustion gasses (the product of the heat emitted by the reaction of lime with water) (Table S2.1, Annex S2 in the Supplementary Materials) also appear to be significant contributors.

The emissions of our IER are consistent with the results of the analysis carried out by Turconi et al. [110]. They showed that direct emissions to the atmosphere and energy recovery/substitution were the main potential impacts on human toxicity. The latter was related to emissions of Hg, As, and Cr. Their study focused on IER with a capacity of 230,000 t of MSW/yr in an incinerator equipped with a semi-dry and wet emission control system.

Concerning the NPEIs, similar comments apply to those made earlier in this section regarding the similar trends, characteristics, and underlying causes for the IER characterization of PEIs. The IER NPEIs for processing 1 t of OFMSW were as follows: for MEc and MEu, 93.56 and 0.023 (person*yr)/FU, respectively; for FWEc and FWEu, 58.60 and 0.432 (person*yr)/FU, respectively. The HCT and HNTC NPEIs were 13.78 and 12.13 (person*yr)/FU, respectively. For emissions to the atmosphere, GW was 0.021 (person*yr)/FU. MEc showed the highest normalized impact, with 94 (person*yr)/FU (Table 6, sum of 63.61 + 29.94). This feature could be related to Hg and Pb emissions (1.45 × 10⁻⁴ kg of Hg and 0.106 kg of Pb/FU, Table S2.1, Annex S2 in the Supplementary Materials). Hauschild et al. and Cole-Hunter et al. and [107,109] consider such emissions to be toxic to human beings.

The thermal processes of the IER contribute to one-third of the GW NPEI. Waste heat of 2614 MJ, including heat associated with the use of natural gas as an auxiliary source of 186 MJ, can be associated with this effect. The emissions that affect the ecotoxicity and toxicity NPEIs (as mentioned above for the characterization PEIs) are related to waste incineration processes and FGT [107,109,111]. Overall, the GW NPEI does not show a high value (0.021 (person*yr)/FU); nevertheless, we decided to analyze it in some detail due to its importance and impact on the fate of humanity and the planet in the present and near future). In this regard, it can be seen in Table S2.1 (Annex S2 of the Supplementary Materials) that the process that contributes most to GHGs is electricity production; it is primary emissions are CO₂ and CO, with 1190 and 0.657 kg, respectively.

Finally, it is worth mentioning that the FGT process, in general, significantly reduces the PEIs and NPEIs associated with particulate matter (up to 88%), POI (76–90%), and TA (83–90%). However, the FGT contributes to fourteen other impacts that would not have occurred in the absence of combustion gas treatment [111]. In particular, the influence of the reported advantages of dry FGT systems over wet treatment systems must be considered in order to reduce the adverse effects of FGT to some extent. Therefore, the effects of any changes introduced in the FGT subsystem to improve the ES of the IER need to be verified by performing new LCAs.

3.3.2. Comparison with Reported Incineration Cases

A review of IER cases that treated municipal solid waste or organic waste was conducted (

Table 7. Comparison with selected examples of Life Cycle Analysis studies of MSW incineration plants. Basis: harmonized functional unit.

FeedWaste Composition	Functional Unit/Harmonization Factor	Impact Assessment Categories				Electricity Output	Remarks	Software and Method	Ref.
		Global Warming	Terrestrial Acidification	Fresh-Water Eutrophication	Human Toxicity (Cancer and Non-Cancer)
Paper 40% OSW 60%	1 t of average MSW Harmonization factor 1.650	119.6 a/ 197.310 b (kg CO2 e)	0.590 a/0.973 b (kg SO2 e)	0.119 a /0.328 b (kg PO4 e)	27.17 a/44.824 b (kg 1,4-DCB e)	180 (kWh)		SimaPro 8.5.2. IILC 2011	Di Maria and Micale [112]
Paper 15.1% Food waste 40.5%Wood 14.1% Plastic 13.7% Others 16.6% Water content 45.23%	kg of wet waste Harmonization factor1.187	525 a/626.18 b (kg CO2 e)	0.540 a/0.641 b (kg SO2 e)	0.080 a/0.095 b (kg PO4 e)	6.070 a/7.204 b (kg1,4-DCB e)	188 kWh/t MSW	cd FAETP 1.25 × 10−2 (kg DCB e) POCP 2.83 × 10−2 (kg Ethene-e) TETP 0.51 (kgDCB e)	Gabi v.8.7, CML2001	Liu et al. [113]
See in Footnotes e	e	0.93 a kg CO2 e/kWh		e	e	264.13 kWh/t	N/A	SimaPro. 8.2. CML	Song et al. [114]
Food waste 35.9%Paper 20.7% Plastics 15.9% Water content 40.40%	kg of MSW Harmonization factor 1.091	273.0 a/297.735 b (kg CO2 e)	2.37 a/2.585 b(kg SO2 e)	0.350 a/0.382 b (kg PO4 e)	e	e			Chaya and Gheewala [115]
OSW 70.6% Plastics 12.8% Paper 7.3% Glass 3% Others 6.3%	kg of MSW Harmonization factor 1.159	314.1 a/363.93 b (kg CO2 e)	0.172 a/0.199 b (kg SO2e)	0.090 a/0.104 b (kg PO4e)	2.396 a/2.776 b	210–310 kWh/t MSW	NE (nutrient enrichment) 0.001 PE/t MSW	EASE WASTE Diesel	Lou et al. [98]
Organic 42.4% Paper 30.9% Plastic 9.4% Glass 6.8% Wood 4.0% Others 6.6%	Incineration of 1 Mgww (t of wet waste) Harmonization factor 1.142	42 mPE a/365.4 g (kg CO2 e)	12 mPE a/1.488 g	12 mPE a/3.576 g (kg NO3)	50 mPE a/2950 kg C2H4e	h	1 mPE = 10−3 PE,	SimaPro	Turconi et al. [110]
SFR: Paper and cardboard 30% Wood 30% Plastics 37% Textiles 3%	kg of SFR/organic waste for energy purposes Harmonization factor 0.691	15 mPE a/t 130.5 (kg CO2 e) b	Acidification (AC) 20 mPE/t a	Nutrient enrichment (NE), 20 mPE/t a	Reports savings in this category (negative value)	Biogas pro-duction 108 Nm3/t of waste (ww)	f Net calorific value (TS) 19.4 MJ/kg Net calorific value (ww) 16.5 MJ/kg	EASE WASTE EDIP97	Fruergaard and Astrup [87]
Food waste 60% Paper 40% Humidity 35%	1 t OFMSW	119 kg CO2 e	2.48 kgSO2 e	0.281 kg PO4e	27 kg1,4-DCB e	393 kWh	N/A	SimaPro	This work

Notes: SFR (solid fuels recovered); the average person’s yearly impact yields the following relationship: 10% of the latter is equivalent to 100 mPE. ^a Value reported by the author. ^b Value calculated with the harmonization factor. ^c GWP is mainly contributed by transportation; the IER was not considered because the IPCC [98] indicates that the carbon dioxide produced by burning short-lived biogenic materials should be considered with a zero value. Also, it could be due to considering the energy generated in the incineration plants as having an “avoided” impact of fossil fuels equivalent [116]. The PEI GWP of the overall IER is more significant than that of transportation if the GWP (i.e., 525 kg CO₂ e) of the emissions released by incineration is considered. ^d Most environmental impact assessments (EIAs) only consider GW, eutrophication, acidification, human toxicity potential, and EP. EIA generally does not consider abiotic depletion potential or equivalent, freshwater ecotoxicity potential, terrestrial ecotoxicity potential, or photochemical ozone creation potential. ^e MSW composition varies greatly, especially in terms of food waste content. In this study, determining accurate quantities of materials that make up the MSW was complicated. ^f No data available. ^g Calculated with normalization factors EDIP 1997 [101]. ^h Includes collection and transportation, consumption of resources and energy for handling and treating the waste, emissions to soil, water, and air, upstream and downstream processes.

). The impacts of selected IERs from the literature were harmonized based on the FUs and total solids of their feedstocks. Their harmonized PEIs were then compared with those of our IER.

The harmonization procedure was similar to that reported for the harmonization of BRF results (see Section 2.1.4 and Section 3.2.2). The PEI was harmonized to 650 kg dry OFMSW (whenever possible), taking into account the reference flow of OFMSW and water content (1000 kg OFMSW and 35% humidity). The cases analyzed from the literature do not necessarily report all environmental impact categories as reported in our IER study (eighteen environmental impacts). Therefore, the PEIs chosen for comparison were GW (kg CO₂ e), TA (kg SO₂ e), FWEu (kg PO₄ e), and human toxicity (1–4 kg DCB e).

It is worth noting that some cases, such as Liu et al. [113], Turconi et al., and Fruergaard and Astrup [88,110], reported negative PEIs. This result is probably due to their consideration that the International Plant Protection Convention (IPCC) guidelines [117] indicate that the carbon dioxide produced by the combustion of short-lived biogenic materials should be considered with a zero value. It could also be due to the fact that the energy produced in the incinerators is considered as “avoided” fossil fuel equivalent impacts. Therefore, in our analysis, the PEIs were selected for comparison based on whether the reported impact was positive and did not minimize or omit the environmental impacts generated. For example, Fruergaard and Astrup and Turconi et al. [87,110] considered savings in electricity generation. They argue that the use of coal in electricity generation was avoided and thus converted to a negative PEI.

The GW for our IER was 119 kg CO₂ e (Table 7), and the emissions with the most significant contributions were CO₂ and CO, with 118 and 0.657 kg/FU, respectively (Table S2.1 Annex S2 in Supplementary Materials). In addition, our IER requires natural gas to generate 186 MJ of heat, which is needed in the FGT process (Table S2.1, Annex S2 in Supplementary Materials), and our analysis does not include waste collection/transport to the IER.

Di Maria and Micale [112] reported 197.3 kg CO₂ e/FU_b, which is higher than our case (119 kgCO₂ e) but in the same order of magnitude (Table 7). The difference in value may be due to differences in FGT. Our IER was equipped with an electrostatic precipitator and a wet scrubber and used selective catalytic reduction for NO_x removal, while Di Maria and Micale’s incineration had dust removal followed by dry gas scrubbing [112]. They analyzed the incineration of 36 t/d of solid organic waste (including collection, treatment, and disposal), which required 2.98 L of fossil fuel/t of waste. Electricity generation was considered according to the climatic conditions and the plant’s location (310 kWh/t, net electrical efficiency 22%). In contrast, the fossil energy consumption was 14 MJ/t. Chaya and Gheewala [115] determined a GW of 297.73 kg CO₂e/FU_b; however, the composition of their MSW varied enormously and the PEIs could not be harmonized. Liu et al. [113] reported that the GW of their incinerator was 626.18 kg CO₂ e/FU (Table 7). This value, although higher, is in the same order of magnitude as our case (120 kg CO₂ e) and the case reported by Lou et al. [98] of 314.1 kg CO₂ e/FU_b (Table 7). The difference could be due to the consideration of waste transportation to the IER plant (for example, they reported diesel consumption of 0.263 and 2 kg/t MSW for transportation and incineration, respectively) in the case of Liu et al. [113].

Such diesel consumption is directly related to the kg of CO₂ eq emitted. In contrast, the transportation of the waste to the incineration plant was not considered in our IER. Turconi et al. [110] estimated 365 kg CO₂ e/FU_b for the GW category, which is more than double the value reported in our work (120 kg CO₂ e) (Table 7). However, it is still in the same order of magnitude as our case and the others analyzed. It is worth noting that the values reported by Turconi et al. [110] include transportation and waste collection, unlike ours. Thus, the order of Turconi et al.’s GW value is consistent. For their part, Fruergaard and Astrup [87] reported 130.5 kg CO₂ e/FU_b, resulting in a GW very close to that observed in our IER (Table 7).

The IER affects FWEu, which may be due to emissions of phosphates and nitrates. These can adversely affect groundwater quality through leaching and ammonia air emissions. In this sense, NO_x is identified by experts as a significant emission resulting from the incineration process [107]. All cases analyzed for the FWEu impact were of the same order of magnitude (Table 7). Our IER had a PEI of 0.281 kg PO₄ e/FU (Table 7) and generated 1.85 kg NO_x/FU (Table S6.1, Annex S6, in Supplementary Materials). Liu et al. [113] found the lowest value for this category with 0.095 kg PO₄ e/FU_b and 0.50 kg NOx/FU. For their part, Lou et al. [98] reported 0.104 kg PO₄ e/FU_b and 1.09–1.34 kg NOx/FU for the FWEu category, and Di Maria and Micale [112] observed 0.328 kg PO₄e/FU_b and 0.265 kg NO_x/FU. Chaya and Gheewala [115] determined 0.382 kg PO₄ e/FU_b (they did not show the NO_x produced). The phosphate and NO_x emissions ranges were in the order of 10⁻¹ for most of the incinerators included in Table 7.

The impact of TA is mainly related to air emissions of nitrogen, such as NO_x and NH₃, and sulfur as SO₂ [107]. In addition, acid gases such as HF and HCl are emissions from MSW incineration and contribute significantly to the acidification category [115]. An inspection of Table 7 shows that TA varied from 0.2 to 2.6 kg SO₂/FU. Given the variety of incinerators and wastes in Table 7, this can be considered a relatively narrow range of variation. Our IER showed a TA of 2.48 kg SO₂/FU, with emissions of 5.85 × 10⁻³ kg NH₃/FU, 1.8 kg NO_X/FU, 0.01 kg SO₂/FU, 1.09 × 10⁻² kg HF/FU, and 1.85 × 10⁻² kg HCl/FU. Chaya and Gheewala [115] published a value of 2.585 kg SO₂/FU_b, which is similar to the PEI determined for our IER (Table 7). Lou et al. [116] reported a TA of 0.199 kg SO_2/FU_b, with emissions of 1.34 kg NO_x/t, 0.44 kg SO₂/FU, 0.060 kg HF/FU, and 0.18 kg HCl/FU. The results of the analysis performed by Di Maria and Micale [112] were 0.973 kg SO₂ e/FU_b for TA and emissions of 1.17 kg NO_x/FU, 0.0333 kg SO₂/FU, 0.01877 kg NH₃/FU, and 0.013 kg HCl/FU. Liu et al. [113] also showed values in the same order of magnitude with a total TA of 0.641 kg SO_2/FU_b and emissions of 0.50 kg NO_X, 0.12 kg SO_2/FU_b, 0.14 kg HCL/FU_b, and 4.59 × 10⁻³ kg HF/FU_b (Table 7). Regarding human toxicity (HT), the value determined in our work was 27 kg 1,4-DCB e/FU. Our value of this PEI was in the middle of the reported range. Liu et al. [113] reported 7.204 kg 1,4-DCB e/FU_b, while Lou et al. [98] determined the lowest value with 2.78 kg 1,4-DCB e/FU_b. On the other hand, the incinerator of Di Maria et al. [112] showed the highest value of HT 44.824 kg 1,4-DCB e/FU_b (Table 7). Turconi et al. [110] described the case of IER Aarhus, which processes 230,000 Mg of wet waste/yr. Unfortunately, we could not compare the HTs because the authors expressed the HT effect in terms of ethylene, which made it difficult to convert their values to a DCB e basis. Similar difficulties were encountered with the HT results of Fruergaard and Astrup [87].

3.4. Comparison of the Global Environmental Sustainability of the Biorefinery and Incineration in This Work

3.4.1. Comparison of Environmental Sustainability

This key subsection consists of two parts. The first compares selected environmental impacts of the IER and BRF GBAER, where we will present the trends of selected PEIs between the two technologies, understanding that the trends of NPEIs and the possible underlying causes and processes are similar. The second part is dedicated to the analysis of the global sustainability of the technologies in terms of novel indices based on the NPEIs from the LCA. These indices are α, σ, τ, and the auxiliary plane σ-τ, and they are consistent with the ISO standards for LCA (see Section 2.2 in the Materials and Methods). The original concepts of “dominant sustainability” and “more environmentally sustainable with restrictions” are also proposed, applied, and discussed.

Figure 3a,b show the characterization PEIs of the BRF and IER standardized to 100%, respectively. The GW PEIs were 119.5 and 942.52 (kg CO₂ e) for the IER and BRF, respectively. The corresponding μ_j and the percentage difference ∆μ_j were 788 and 688%, respectively (see Materials and Methods Section 2.2), indicating that the impact of the BRF was much more significant. This difference may be due to the use of fossil energy in the embedded processes (embedded proxies) of the BRF-LCI, among other factors.

It is worth recalling that the complexity of the biorefinery (six-step flow sheet and production of two bioenergies plus up to four bioproducts) is likely to result in higher material use and higher fossil energy consumption in the LCIs of the embedded processes (proxies). The associated use of fossil energy may, in turn, increase the CO₂ contributions, boosting the GW impact of the BRF. In this regard, the dominant stages for GW contributions in the BRF are M, Z, and NN, with 17, 18, and 38% contributions, respectively (Figure 3 and Table 4). The proxies included (embedded) in the LCI of such stages of our BRF are likely to consume fossil energy from the technosphere. This, in turn, could present a source of CO₂ and other GHG emissions. On the other hand, the IER only requires auxiliary fossil fuel in the startup and restart (intermittent) and the FGT subsystem (continuous), i.e., 186.94 MJ/FU. This discussion may explain why, contrary to expectations, the GW PEI of the IER is only a fraction of that of the BRF. It may be helpful to consider replacing the fossil energy used for heating in the embedded proxies of the BRF with solar energy, which would likely reduce their contributions to the GW PEI of the BRF. This integration may be feasible where heating requirements are associated with low–medium temperatures, such as mesophilic heating processes at 35–37 °C in the M and Z stages, and the NN of the BRF.

The reported PEIs for FWEc were 51.19 and 10.74 1,4-DCBe/FU for the IER and BRF, respectively (Figure 3a,b). The corresponding μ_j and the percentage difference ∆μ_j were 476 and 376%, respectively (see Section 2.2 Equations (6) and (7)), indicating that the impact of IER was more significant than that of the BRF by a factor of ~5. This feature could be related to the emissions from direct electricity generation in IER, which contributes 48.7 kg of 1,4-DCB e (Table S11.4 Annex S11 in the Supplementary Materials). Maresca et al. [106] reported several major emissions to water from IER in this category. We found 2.93 × 10⁻³, 0.137, 0.212, 0.0433, and 0.0426 kg/FU for As, Cu, Zn, Ni, and Br, respectively, in our incinerator (Table S2.1, Annex S2 in the Supplementary Materials). Regarding the BRF, the NN stage had the highest contribution to the FWEc with 4488 kg 1,4-DCB e/FU (Figure 3a, Table 4), which was related to activated carbon production processes and the treatment of coal mining waste (Table S10.4, Annex S10 in the Supplementary Materials).

The PEIs for HNCT were 271 and 1288 kg1,4-DCB e/FU for BRF and IER, respectively. The corresponding μ_j and the percentage difference ∆μ_j were 475 and 375%, respectively (see Methodology Section S2.2), suggesting that the impact of IER was more significant than that of BRF. The IER emissions associated with this category are likely to be As (V) and Zn (II) emitted into freshwater [107]). In this regard, as mentioned above for our IER, emissions of 6.25 × 10⁻⁷ and 1.90 × 10⁻⁵ kg of As (V) and Zn (II), respectively, were reported to contribute to HNCT (Table S2.1, Annex S2 in the Supplementary Materials).

Regarding the discussion and comparison of NPEIs, we should first recall that the procedure for evaluating normalized environmental impacts can be found in the Methodology section (Section S2.1.3 and Annex S7 in the Supplementary Materials). In short, the relationships between PEIs and NPEIs are based on the unit impact loads per person and year reported by the SimaPro database as normalization factors that are the inverse of the unit loads [70]. In general, the comparison of the most significant NPEIs between the BRF and IER follows the trends highlighted and discussed for the corresponding characterization PEIs. The highest NPEIs corresponding to the IER technology were predominantly mainly in two categories: MEc and FWEc, with approximately 94 and 58 (person*yr)/FU, respectively, followed by HNCT (12.14) and FWEu (0.432) (

Table 8. Normalized environmental impacts of incineration and biorefinery in this work in units (person*yr)/FU) and α indices for each technology, which helps define their environmental sustainability.

Impact Category	Incineration with Energy Recovery	Biorefinery
Global warming	0.021	0.118
Stratospheric ozone depletion	0.089	0.039
Ionizing radiation	0.003	0.043
Ozone formation, human health	0.092	0.133
Fine particulate matter formation	0.012	0.130
Ozone formation, terrestrial ecosystems	0.107	0.196
Terrestrial acidification	0.020	0.465
Freshwater eutrophication	0.432	0.422
Marine eutrophication	0.023	0.015
Terrestrial ecotoxicity	0.238	0.803
Freshwater ecotoxicity	58.601	8.756
Marine ecotoxicity	93.557	13.604
Human carcinogenic toxicity	13.777	13.933
Human non-carcinogenic toxicity	12.139	1.819
Land use	2.02 × 10−5	0.007
Mineral resource scarcity	3.82 × 10−7	1.49 × 10−5
Fossil resource scarcity	0.006	0.207
Water consumption	<0.001	0.034
Index α (sum of the eighteen NPEIs)	179.114	40.725

, Figure 4).

As mentioned in Section 3.3.1, the normalized impacts related to the IER, with the most significant contribution, are associated with the electricity generation processes. In turn, as stated above, these processes are also related to waste incineration and FGT, and they mainly affect the NPEIs of ecotoxicity and toxicity. The primary emissions in this category are detailed in Section 3.3.1, and some measures can be advanced to reduce their environmental impact. For the BRF, its NPEIs for ecotoxicity (marine and freshwater) and HCNT are 14, 9, and 1.8 (person*yr)/FU, respectively, which are almost seven times lower than the corresponding NPEIs for IER (Figure 5; Table 8). The NPEI for HCT is essentially the same for BRF and IER, at ~14 (person*yr)/FU (Table 8, Figure 5).

To evaluate global environmental sustainability, we first used the GBAER index α (see Annex S8 in the Supplementary Materials) to determine the most environmentally sustainable technology, where α is the sum of the NPEIs, and higher values of α indicate lower ES. According to Equation (S8.1), as shown in Annex S8, we obtained a value of α_IER = 179.11. The BRF HMSZ-NN was α_BRF = 40.72 (Table 8). When α is larger, ES is lower, so the BRF is more sustainable (environmentally) than IER. Overall, the BRF is almost four times more sustainable (environmentally) than the IER for the treatment of Mexico City’s OFMSW.

Second, we calculated the σ-τ indices shown in Annex S8 (Equations (S8.2) and (S8.3)); their values were σ = 0.643 and τ = −0.268, representing the coordinates of a point P that expresses the sustainability comparison between the BRF and IER (Figure S8.2, Annex S8 in the Supplementary Materials). Point P is located in Quadrant IV of the σ-τ plane (Figure 5 and Figure S8.2 in Annex S8 of the Supplementary Materials). This result confirms that the HMSZ-NN BRF is more environmentally sustainable than the IER, with restrictions. These restrictions mean that a subset of the least significant NPEIs are lower for the least sustainable option. However, the large values of the highest NPEIs of the least sustainable option determine its loss of ES.

As an example of how the concept of “restrictions” works, an inspection of Table 8 shows that the IER has lower values of NPEIs in up to twelve impact categories, although they are the least significant in terms of their contribution to the α index (GW, IR, TA, FP, OFH, OFTE, TE, HCT, FRS, LU, WC, and MRS). In contrast, the BRF shows lower NPEI values in six impact categories (HNCT, FWEc, MEc, FWEu, MEu, and SOD); at least three of these six are almost seven times lower for the BRF (e.g., the first three categories from the previous list) (Figure 4, Table 8). Therefore, the profile of NPEIs, in this case, is such that the BRF is more environmentally sustainable overall than the IER, even though the latter is more sustainable than the BRF in a larger subset of less significant impact categories. This profile is acknowledged with the word “restrictions”, and negative values of the calculated index τ accompanied the positive values of σ. The statement is that BRF is more environmentally sustainable than the IER, with restrictions.

Another typical profile is possible. One technology is more environmentally sustainable overall than the other (i.e., the reference technology), and the former is more sustainable in a larger subset of impact categories than the latter one (lower NPEIs). This profile is acknowledged with the word “dominant” or “dominance” and shows positive values of the calculated index τ and positive values of σ. This indicates that the first technology (i.e., a new technology) is more environmentally sustainable in a dominant way than the reference technology. In colloquial terms, the first technology would be significantly more environmentally sustainable than the second by a landslide. Two additional profiles can occur when the reference or second technology is more environmentally sustainable than the first. In both cases, the σ values would be negative. However, the sustainability statement would have restrictions whenever τ was positive and would be dominant when τ was negative. Therefore, the new σ-τ plane or procedure is more informative and constitutes a step beyond the more Spartan (albeit novel and valuable) α criterion for characterizing the overall ES of two technologies.

3.4.2. Sensitivity Tests

Sensitivity of Biorefinery Environmental Sustainability to Changes in the Type of Activated Carbon Used in the NN Stage of the Biorefinery

Our previous analysis above of the environmental impacts, as well as the contributions of stages, processes, and materials on PEI and NPEIs of the BRF, identified that the use of activated carbon of fossil origin (FAC) in the NN stage of the BRF could be associated to high values of selected impacts of ecotoxicity and human health in that stage, as well as contributions to other NPEIs. The NN stage was critical for HNCT, GW, FWEu, FWEc, and MEc. The NN stage contributed 49.5% to the HNCT category, probably related to using activated carbon of fossil origin. The latter is the core support of the bionanobioparticles (BNBPs) produced in the NN stage, and its impacts are associated with treating mining wastes of fossil coal. Therefore, using activated carbon of fossil origin is a burden for the environmental impact of BRF, especially for the NN stage.

Furthermore, we suggested above in our article and previous works [117] that it would be worth determining the effect of a change in material, that is, a replacement of FAC by another adsorbent such as activated carbon of vegetal origin (VAC), on the NN stage contribution to the environmental sustainability of the BRF. As the LCA is iterative, it is helpful to consider replacing activated carbon of fossil origin with activated carbon of plant origin. It would be worthwhile conducting such a study—if a benefit is found, the impacts of key BRF stages, processes, and products on key NPEIs could be reduced, and a modified BRF would be more sustainable. Consequently, looking for an alternative to activated carbon of fossil origin (FAC), such as activated carbon of plant origin (VAC) or any other suitable adsorbent material with a lower environmental impact, is advisable.

Therefore, a sensitivity analysis was performed for BRF using either FAC or VAC. The LCI for the FAC was obtained from the proxy “Activated carbon, granular {RoW}| activated carbon production |Cut-off, U”. In contrast, the LCI for the VAC was reported by Arena et al. [118] (Table 4 of this reference, scenario 2) and is exhibited as Table S18.1, Annex S18, Supplementary Material document. The readers are advised to thoroughly read Annex S18 since the sub-section is only a summary of the Annex with fewer tables and figures and a shorter discussion of results. The characterization of PEIs of both biorefineries, the one with fossil-origin activated carbon (BRF-FAC) and the second one with vegetal-origin activated carbon (BRF-VAC), is exhibited in Figure S18.1, Annex S18. The main trends are the following:

(i) Seventeen PEIs of BRF-VAC were lower than those of BRF-FAC. The highest percentage decreases were found for the PEIs: FWEu 48.31%, HNCT 44.93%, MEc 40.50%, FRS 38.00%, and FWEc 37.91%. There was one tie: the stratospheric O₃ depletion.

(ii) All the relative contributions (in percentage) of the stage NN to PEIs of BRF-VAC were lower than those of BRF-FAC.

(iii) As a consequence, the absolute contributions (in units of PEI) of the stage NN to the BRF-VAC impacts were lower than those absolute contributions to impacts of BRF-FAC

The analysis of the NPEIs of both biorefineries confirmed the findings for the characterization PEIs in Figure S18.1, Annex S18. Figure 6 depicts the eighteen NPEIs of both the BRF with fossil-origin activated carbon and vegetal-based activated carbon. Seventeen environmental impacts decreased their NPEIs for BRF VAC. The most outstanding decreases corresponded to high-value ecotoxicity-related NPEIs and human health-related impacts. The immediate consequence of this pattern was that the indicator α for the BRF-VAC was significantly lower than that of BRF-FAC. Their values are shown in Table S18.2, Annex S18. The α indicator decreased from 40.73 to 27.041 (person*yr/FU) (−40.13%) when replacing FAC with VAC in the BRF, which represented an absolute net decrease of 13.42 (person*yr/FU) (Table S18.2, Annex S18). The values of the other two indicators were σ = 0.178 and τ = 1.091, defined as Point B. Point B, which compares both BRFs in the σ-τ plane, lies on Quadrant 1 in Figure 5. According to the discussion in Annex S8 and Section S3.4.1 above, this means that the BRF that used VAC was more environmentally sustainable than the BRF that used FAC with dominance (Figure 5). Please also see Annex S8 in the Supplementary Materials for calculations of the environmental sustainability indices α, σ, and τ based on normalized environmental impacts.

The ES comparison between IER and the BRF-VAC is summarized in Table S18.3 and Figure S18.4, in Annex S18. The values of indicator α were 179.11 and 27.041 (person*yr/FU) for the IER and the BRF-VAC, respectively (a difference of 73.77%), which represented an absolute, net decrease of 152.069 (person*yr/FU) (Table S18.3, Annex S18). The restrictions on the ES for the comparison IER/BRF-VAC (point B) were lower than those for the comparison IER/BRF-FAC because the τ value for the first case increased. This favors the BRF-VAC from the ES evaluation/comparison with IER standpoint.

In conclusion, replacing fossil-origin-activated carbon with VAC in the BRF significantly improved the BRF’s environmental sustainability and increased its environmental sustainability superiority concerning IER. The technologies order in descending ES is BRF-VAC > BRF-FAC > IER.

Sensitivity of Indicators and Environmental Sustainability to Weighting

In the following, the results of the sensitivity of weighting approaches and error analysis of the indicators α and α_w (Equations (S16.1), (S16.2), (S16.9), and (S16.10), Section S16.1, Annex S16, Supplementary Materials) used in this work will be briefly discussed; the readers are encouraged to thoroughly read Annex S16 because this subsection is only a rapid summary of the detailed equations’ derivation, a Lemma and its proof, and four tables with results; the whole discussion presented in the Annex, along with final concluding remarks on the GBAER indicators applied in this work. When assigning high weights to the GW and WC scenarios (w_p = 4), there was no increase in the α_w indicator, even for the highest weight factors. The result was “counter-intuitive” since an increase in weights on the GW and WC NPEIs by 300% led to a decrease in the indicator α_w from 179.12 to 112.02 (37.46%) and 40.724 to 24.966 (36.24%) for IER and BRF, respectively (Table S16.1, Section S16.2.1, Annex S16, Supplementary Material). Interestingly, the percentage decrease was very close for both technologies. Thus, the ratios α_IERα_BRF for equal weighting and the weighted indicators were very similar, 4.398 and 4.314, respectively. The corresponding values of σ indicators followed a similar trend. Therefore, the pattern of the ES for both technologies based on those ratios suggests that the BRF is nearly four times more environmentally sustainable than IER, irrespective of the weighting. It seems that the decrease in the contribution of the other NPEIs that were assigned the complementary weight factor w_r = 0.625 (particularly the large NPEIs of ecotoxicity and human health) was more significant than the meager increases in GW and WC impacts; this explains the reduction in α_w for both technologies. Section S16.2.3.1, Annex S16 also includes the Lemma S16.1, which predicts these trends qualitatively for several weighting scenarios; when using the indicator, this Lemma can be used to check the results of numerical calculations.

More tests were run, where the NPEIs related to ecosystem effects were assigned a w_p = 2, while the remaining ones were multiplied by the complementary w_r = 036354 (Table S16.1 and Section S16.2, Annex S16, Supplementary Materials). The α_w of both technologies increased, at 76.12% for IER and 33.84% for BRF. As a result, the ratio α_w,IER/α_w,BRF was 5.787, nearly 31.52% higher than the baseline ratio of 4.39 (EW baseline). The new ratio suggests that the BRF is almost six times more environmentally sustainable than the IER. In summary, the seven NPEIs related to ecosystem toxicity and damage contribute decisively to the indicators α and α_w. Using high-weighting factors for these impacts increases the environmental sustainability gap between IER and BRF, favoring the BRF.

Two more tests were performed to determine the sensitivity of high weighting the impacts associated with human health effects (w_p = 2; and another with w_p = 2.5) (Table S16.1 and Section S16.2.3, Annex S16, Supplementary Materials). The other NPEIs were multiplied by the complementary weigh w_r = 0.5 and 0.25, respectively (Table S16.1). The w_p/w_r ratio was 4 for the first combination. The changes on α_w were different for each technology. The IER α_w decreased by −28.13%, whereas the value of BRF slightly increased by 9.29% (Table S16.1, Annex S16, Supplementary Materials). As a result, the ratio α_w,IER/α_w,BRF was 2.892, nearly 34.27% lower than the baseline ratio of 4.39 (EW baseline; Table S16.1, Annex S16, Supplementary Material). There seems to be a “convergence” trend between α_w,IER, and α_w,BRF; however, this convergence was not enough to reverse the superiority of BRF over IER in terms of environmental sustainability. The new ratio suggests that the BRF is nearly three times more environmentally sustainable than the IER. The second combination gave similar trends to the first one. The indicator σ exhibited trends similar to those of the ratio α_w,IER/α_w,BRF (Table S16.1). Lemma S16.1 was developed; it was able to qualitatively predict the distinct responses of the α_w for IER and BRF in these scenarios (Table S16.3).

In summary, the ES of BRF was better than IER’s for the three weighing scenarios, pointing to the indicators’ α_w and σ consistency. For GW and WC high weighting, the relationship of ES was nearly uniform (ca. 4). For high weighting of ecotoxicity-related impacts, the relationship of ES of BRF to IER increased compared to the baseline scenario. For high weighting of human health-related impacts the relationship of ES of BRF to IER decreased, but still, the ES of BRF was superior.

The error analysis revealed that the absolute error of the indicators α and α_w increased with the magnitude of the errors of the NPEIs, as expected (Tables S16.1 and S16.4, Section S16.3 of Annex S16, Supplementary Materials). That increase was not rampant. It led to relatively “controlled” (reasonably narrow) confidence intervals (CIs) of 95% for the indicators. CIs for indicators of IER and BRF in each examined scenario did not overlap. Moreover, even for the most conservative scenario, the Lower Limit_IER >> Upper Limit_BRF. This suggests that the indicator for IER is significantly higher than the value for BRF, either for baseline or weighted indicators examined in this work. Thus, from this point of view, the BRF is, in general, more environmentally sustainable than IER and likely with statistical significance.

It is important to discuss the relative economic accessibility of the BFR vs. IER. However, a complete economic analysis and evaluation would necessitate a complex and long piece of work [119,120], which would double the size of the present article. Thus, an economic analysis is beyond the scope of this article. Nevertheless, we conducted a preliminary assessment of the specific investment costs based on the open literature and Internet information. For more details, please consult Annex S14, “Economic assessment of specific investment costs of the biorefinery and the incineration plant” in the Supplementary Material document. Here, the definition of the specific investment costs (γ) is given by Equation (8), as shown below [121], as the ratio of total investment cost (TIC) in USD to Γ the plant capacity (t/yr) in terms of the waste or biomass:

$$\gamma ={\displaystyle \frac{total investmet costs }{\Gamma }}\ast F_{correction} \left({\displaystyle \frac{USD}{(t/yr)}}\right)$$

(8)

where F_correction is the correction factor to standardize to USD January 2025, given that the cost data were usually reported in different years.

Annex S14 (in the Supplementary Materials) provides compilation tables of biorefineries (γ) (Tables S14.1 and S14.2) and the trends of (γ) in terms of incinerator capacity and biorefinery complexity (Figure S14.1), along with further discussion. The comparison indicates that the specific investment cost of the incineration plant will most likely be higher than that of the BRF, at 484 and 680 USD/t in 2025, respectively, or 40.49% higher in relative terms. These preliminary results are encouraging and confirm that the economic analysis of these technologies should be continued and expanded soon.

4. Conclusions

Our results indicate that the BRF GBAER, a biorefinery based on the fermentation of OFMSW, is more sustainable (environmentally) than the IER for the treatment of such waste in Mexico. The alpha values were 179.1 and 40.7 (person*yr)/FU for the IER and BRF, respectively, indicating that the IER technology is four times more polluting than our BRF. Regarding the new indices σ and τ, values of σ = 0.5923 and τ = −0.368 were found. The point defined by these coordinates in the σ-τ plane was located in Quadrant IV. This result also confirms that the BRF considered in this work is more environmentally sustainable (with restrictions) than the IER for treating OFMSW in México.

Our results point to the implementation of waste-based BRFs to replace landfills in major Mexico cities, and possibly in major cities of other DCs and UCs. Our model BRF is significantly more environmentally sustainable than the other target option IER and can provide electricity and bioproducts that benefit the sustainable development of modern societies. In terms of energy, the BRF produces 166.4 net kWh/FU (600 MJ) of electricity, along with several bioproducts such as VOAs (38 kg), an industrial enzyme solution (1087 kg), hydrolysates for bioproduct fermentations (515 kg), BNBPs (40 kg), and a soil amender. In contrast, the IER produces only energy, 393 net kWh/FU of electricity, and 5653 MJ/FU of heat. Summarily, from an energy point of view, the IER is superior to the BRF; however, the latter is self-sufficient and has a modest electricity surplus for sale or export. In addition, the BRF generates a menu of bioproducts, some of which qualify as VAPs, as mentioned above.

The highest NPEIs were marine and freshwater ecotoxicity, followed by human non-carcinogenic and carcinogenic toxicities. The first three were significantly higher for IER than for the BRF. The waste products most likely to contribute to these categories in IER were Cr(VI) emitted into freshwater and heavy metals (Hg and Pb) contained in combustion gasses. In contrast, As (V) and Zn (II) emissions contribute to the HNCT category for humans. With respect to the BRF, the most prominent contributors to the NPEIs were the Z and NN stages, related to the production of enzymes and the generation of BNBPs, respectively, with particularly high contributions to the effects of human carcinogenic toxicity. Stage Z contributed greatly to the HCT (up to 60% of this PEI), derived from the use of Na₃PO₄, which in turn is associated with a non-explicit process derived from the treatment of waste related to the production of H₃PO₄, contributing 60% to this impact category.

The NN stage was critical for HNCT, GW, FWEu, FWEc, and Mec, contributing 49.5% to the HNCT category, which was probably related to the use of activated carbon of fossil origin. The latter is used to produce BNBPs and is associated with the treatment of mining waste and fossil coal. Therefore, the use of activated carbon of fossil origin again becomes a burden in terms of the environmental impact of the BRF, especially for the NN stage. Consequently, it is advisable to look for an alternative to activated carbon of fossil origin, such as activated carbon of plant origin or any other suitable adsorbent material with a lower environmental impact.

In short, our side-by-side comparison of the ES of IER and a BRF demonstrates that a Mexican model of a BRF is more environmentally sustainable than IER for the treatment and management of MSW. Our research also included the development and application of innovative indices based on NPEIs, in line with ISO standards and LCA, to calculate and compare ES. These indices confirmed the conclusion of the environmental impact assessment. The indices and the methodology show promise for application to other environmental technologies (and to industrial processes/technologies) for the selection of more sustainable waste management and treatment options. A wide variety of potential users could likely benefit from their use, including planners and policymakers, governmental decision-makers, manufacturers/suppliers of equipment, environmental consultants, practitioners, academicians, and communities interested in waste management.