Application of Multi-Criteria Decision-Making Tools for Assessing Biogas Plants: A Case Study in Reykjavik, Iceland

: The European Union is planning a new program to achieve climate neutrality by 2050. In this context, the Icelandic government plans to ban new registrations of fossil fuel cars after 2030 as one of the strategies to make Iceland a carbon-neutral country by 2040. Upgraded biogas can be directly used in vehicles with CNG engines, reducing CO 2 emissions by 80%. In this paper, several alternatives of biogas plants, simulated in previous research, were evaluated by considering techno-economic and environmental criteria through the application of multi-criteria decision-making tools. Twelve alternatives were analyzed using the Deﬁnite 3.1 software. A weighted summation algorithm, which transforms all criteria into the same scale by multiplying them by weights and then summing them to obtain the results, was used in the analysis. The multi-criteria analysis of the twelve proposed alternatives included eleven criteria (three technical, ﬁve economic, and three environmental) whose weights were changed in a total of eleven scenarios. From a global perspective, when all criteria were considered (9.1% weight) the best alternative with a score of 0.58 was the single-stage biogas plant working with municipal solid waste. Sensitivity and uncertainty analyses also demonstrated that the multi-criteria results obtained were robust and reliable.


Introduction
The EU has announced an ambitious plan to reduce GHG emissions to 80-95% below 1990 levels by 2050 [1]. In Europe, so far, much of the combined total GHG emissions reduction has come from the power and heating and cooling sectors, achieving a reduction of 23% from 1990-2014, while the transport sector emissions increased by 20.1% in the same period [2]. In the meantime, recent studies indicate that the energy demand will grow by up to 50% until 2050 and, therefore, the security of the energy supply is a crucial challenge [3].
There are several alternatives for reducing GHG emissions, among them, the EU strategy proposes intensifying the use of biomass. Processing biomass into bio-based and renewable products allows for a decrease in the consumption of non-renewable resources and boosts the circular economy [4,5]. Anaerobic digestion (AD) of organic compounds to produce biogas is a promising alternative for biomass utilization. AD is a biological process where organic matter is biodegraded under anaerobic conditions, leading to the production of biogas along with a digestate [6]. The digestate is a black-in-color by-product that can be divided into solid (SD) and liquid phases (LD). The SD can be transformed into energy through incineration, pyrolysis, gasification, or hydrothermal carbonization (HTC); composted, or used to fertilize agricultural crops. The LD contains high concentrations of nitrogen, phosphorous, and potassium and must be treated before its discharge into  [23]. Reproduced with permission from Elena Dosal, Doctoral thesis "Towards and improved framework for construction and demolition waste management (C&DW) using decision support tools", 2015.
In this case, the method selected (step 4 of Figure 1) was weighted summation (WS) This method can be used to address problems that involve a finite and discrete set of alternatives that must be evaluated based on conflicting objectives [25]. For any given objective, one or more different attributes or criteria are used to measure the performance in relation to the objective. Impacts of all alternative options for all criteria are presented in the impact matrix. Such criteria are usually measured on different scales and therefore cannot be compared with each other directly.
The process to be followed to carry out WS can be further detailed as follows: (1) definition of the alternatives that will be compared against each other; (2) selection and definition of criteria identifying the most relevant indicators for the decision; (3) assessment of scores for each alternative by assigning values to each indicator for all the alternatives; (4) standardization of the scores to make the criteria comparable with each other (5) weighting of criteria to assign priorities to them; (6) ranking of the alternatives. A tota score for each alternative is calculated by multiplying the standardized scores with their appropriate weight, followed by summing the weighted scores of all criteria.

Anaerobic Digestion Plants
A total of twelve alternatives (biogas plants) were considered and evaluated in the MCA as shown in Figure 2. The biogas plants included three different inlet flows (300 t/d 320 t/d, and 323 t/d) and three kinds of feedstocks (MSW, FW, and LCB), operating in  [23]. Reproduced with permission from Elena Dosal, Doctoral thesis "Towards and improved framework for construction and demolition waste management (C&DW) using decision support tools", 2015.
In this case, the method selected (step 4 of Figure 1) was weighted summation (WS). This method can be used to address problems that involve a finite and discrete set of alternatives that must be evaluated based on conflicting objectives [25]. For any given objective, one or more different attributes or criteria are used to measure the performance in relation to the objective. Impacts of all alternative options for all criteria are presented in the impact matrix. Such criteria are usually measured on different scales and therefore cannot be compared with each other directly.
The process to be followed to carry out WS can be further detailed as follows: (1) definition of the alternatives that will be compared against each other; (2) selection and definition of criteria identifying the most relevant indicators for the decision; (3) assessment of scores for each alternative by assigning values to each indicator for all the alternatives; (4) standardization of the scores to make the criteria comparable with each other; (5) weighting of criteria to assign priorities to them; (6) ranking of the alternatives. A total score for each alternative is calculated by multiplying the standardized scores with their appropriate weight, followed by summing the weighted scores of all criteria.

Anaerobic Digestion Plants
A total of twelve alternatives (biogas plants) were considered and evaluated in the MCA as shown in Figure 2. The biogas plants included three different inlet flows (300 t/d, 320 t/d, and 323 t/d) and three kinds of feedstocks (MSW, FW, and LCB), operating in single or in co-digestion. The chemical composition of the three materials was previously described and can be consulted in the literature [18].
The single-stage AD model approach includes a total of 7 reactions based on the Nduse and Oladiran model [26]; the two-stage AD model approach considers 45 reactions and is based on the Rajendran model [27]. In this case, hydrolysis occurs in the stoichiometric reactor, whereas acidogenesis, acetogenesis, and methanogenesis occur in the continuously stirred tank reactor. All details of the AP developed models, assumptions, and considerations were fully described in a previous publication [18]. A block scheme of the alternatives (A1 to A12) is summarized in Figure 2.

Criteria Definition
To build a comprehensive evaluation dimension system, the bioenergy sector focuses on environmental, technical, economic, and social factors. Consequently, the evaluation dimensions of the present paper mainly consisted of eleven criteria divided into three categories, respectively, technical, economic, and environmental, as shown in Table 1.
The eleven criteria included in the impact matrix to perform the MCA were quantitative. The WS method requires criteria to be comparable amongst each other. Therefore, all criteria were standardized by dividing each one of them by the inlet mass flow of feedstock. Among the eleven criteria, the three technical ones were of the type "benefit", which means that the higher the score of this effect is, the better are the alternative results. The rest of the criteria (five economic and three environmental) were of the type "cost" or in other words, the higher the score of this effect is the worse are the alternative results. Plant sizes and kinds of feedstocks were taken from the SORPA landfill. In a period of four years (from 2016 to 2019) the landfill collected on average 110,000 t/y of MSW, 8000 t/y of FW, and 1000 t/y of LCB. Consequently, the inlet flows of co-digestion alternatives were set at 323 t/d (all residues), 320 t/d (MSW and LCB), and 3 t/d (MSW and FW). Two Aspen Plus model approaches (one or two anaerobic digestion steps in series) were evaluated. The AP model approaches work under thermophilic conditions of 55 • C. The single-stage AD model approach includes a total of 7 reactions based on the Nduse and Oladiran model [26]; the two-stage AD model approach considers 45 reactions and is based on the Rajendran model [27]. In this case, hydrolysis occurs in the stoichiometric reactor, whereas acidogenesis, acetogenesis, and methanogenesis occur in the continuously stirred tank reactor. All details of the AP developed models, assumptions, and considerations were fully described in a previous publication [18]. A block scheme of the alternatives (A1 to A12) is summarized in Figure 2.

Criteria Definition
To build a comprehensive evaluation dimension system, the bioenergy sector focuses on environmental, technical, economic, and social factors. Consequently, the evaluation dimensions of the present paper mainly consisted of eleven criteria divided into three categories, respectively, technical, economic, and environmental, as shown in Table 1. The eleven criteria included in the impact matrix to perform the MCA were quantitative. The WS method requires criteria to be comparable amongst each other. Therefore, all criteria were standardized by dividing each one of them by the inlet mass flow of feedstock. Among the eleven criteria, the three technical ones were of the type "benefit", which means that the higher the score of this effect is, the better are the alternative results. The rest of the criteria (five economic and three environmental) were of the type "cost" or in other words, the higher the score of this effect is the worse are the alternative results.
Regarding the technical criteria, the CH 4 :CO 2 ratio gives an idea of the selectivity of the AD process and is calculated as shown in Equation (1). Biogas and methane yields are among the most important indicators of these kinds of plants. They were calculated as shown in Equations (2) and (3).
are the volumetric flow rate of methane and carbon dioxide in the biogas outlet stream, respectively.
and . m f are yield of biogas, density of biogas, outlet mass flow of biogas, and inlet mass flow of feedstock, respectively.
where yield CH 4 , ρ CH 4 , . m CH 4 , and . m f are yield of methane, density of methane, outlet mass flow of methane, and inlet mass flow of feedstock, respectively.
Regarding the environmental criteria, carbon emissions measured in kg of CO 2 eq. per kg of inlet feedstock were determined through the Aspen Energy Analyzer (AEA). The digestate generated is the result of the bottom stream, named 'sludge' in the flowsheet previously published by the co-authors [18], and is obtained as an outlet stream of the flash unit, which is placed after the anaerobic digestion units. The criterion digestate contaminants (Equation (4)) considers the sum of the mass fractions of carbon dioxide, methane, ammonia, acetic acid, propionic acid, benzene, and the phenol forming part of the digestate matrix ('sludge' stream of the AP flowsheet). These components were selected based on their effects on atmospheric acidification, stratospheric ozone depletion, photochemical ozone (smog) formation, human health (carcinogenic) effects, aquatic oxygen demand, or ecotoxicity to aquatic life [28]. In addition, CO 2 and CH 4 were considered because they are GHG.
are the mass fractions of the previously aforementioned components considered in the digestate.
Regarding the economic criteria, all of them were determined through the Aspen Plus Economic Analyzer tool (APEA). A brief description of the fixed costs (capital, equipment, and installed costs) and variable costs (operating and utility costs) considered in the MCA is given [29]: • Capital costs include all cost-based technical project details, including design, equipment, civil, structural, piping, mechanical, steel, instrumental, electrical, insulation, paint, labor, and management; • Operating costs include the total raw materials costs, utility costs, operating labor costs, maintenance costs, operating charges, and plant overhead; • Utility costs are the costs incurred by using electricity, water, heating, or waste disposal; • Equipment costs include costs of vessels, pipelines and in general all the unit operations used in the plant; • Installed costs are the total costs of labor and materials of the facility.

Formulated Scenarios
Once the criteria were selected, the next step in the MCA-WS methodology is to define the importance (weight) of each criterion and decide the scenarios that will be considered for evaluating the biogas plant alternatives. Such scenarios are represented in Table 2.  Table 2 summarizes the scenarios (SCE) contemplated, together with the weights assigned to each criterion and the purpose pursued in each scenario. The developed scenarios had the aim of analyzing the biogas plant alternatives not only from a global perspective, which corresponds to the SCE.11, but also from different perspectives depending on the stakeholders' viewpoints. Thus, looking at Table 2, there is a total of four scenarios (SCE.1 to SCE.4) considering only technical criteria, three scenarios (SCE.5 to SCE.7) considering environmental criteria, three scenarios (SCE.8 to SCE.10) considering economic criteria, and the last scenario (SCE.11) combines all criteria with the same weight (9.1%).

Robustness Analysis
Weights of the criteria considered in the formulated scenarios of Section 2.4 together with the scoring values of the alternatives could contain some uncertainties. An important issue is to know how the final ranking of the alternatives is sensitive to the changes of some input parameters of the decision model. Therefore, uncertainty and sensitivity analyses were conducted. The uncertainty of input data was set to 10% and 25%. Sensitivity analysis of each criterion to determine how its weight affected the final ranking was also studied by varying weights from 0 to 100%. In addition, perspectives of the ranking alternatives at 20%, 40%, 60%, 80%, and 100% weights of three selected criteria (biogas yield, CO 2 eq., and capital costs) were also checked.

Results of the Multi-Criteria Analysis by the Weighted Summation Method (MCA-WS)
The first step to starting the MCA-WS is to build the impact matrix with all the quantitative effects. Table 3 includes each of the twelve biogas plant alternatives in the columns and the eleven quantitative criteria in the rows. The impact matrix values were introduced into the Definite software, and each of the formulated scenarios (SCE.1 to SCE.11) will be discussed. Figure 3 shows all the MCA-WS results for a given scenario. The scores are represented in bars graphs and all the alternatives are plotted in descending order. Weighting criteria distribution is represented in the circular graphs.
The technical scenarios (SCE.1 to SCE.4) plotted in Figure 3a-d show similar ranking scores. In general, two-stage AD plants showed better results except for SCE.2 (100% weight-to-criterion biogas yield). In this case the highest production of biogas occurred working in the co-digestion mode in a two-stage model, followed by in the biogas plant alternatives working in single-stage conditions. From the rest of the technical scenarios (SCE.1, SCE.3, and SCE.4) the two-stage models in mono-digestion and co-digestion modes become the best choices due to their higher CH 4 presence in the biogas. The highest biogas amount under single-stage conditions was obtained by using MSW (score of 0.95) as can be seen in SCE.2 (Figure 3b) because the MSW used in the AP models contains higher amounts of carbohydrates than do FW and LCB feedstocks. In addition, the C/N ratio of MWS fits better with the 20-30 C/N ratio requirements than do those of LCB and FW [30,31]. Regarding the methane yield (SCE.3 in Figure 3c), LCB is the feedstock that obtained the highest score (1.00). Among all the feedstocks checked, LCB is the one with the most carbohydrates, and according to the bibliography, carbohydrate concentrations higher than 8.3% and proteins and lipids lower than 5.0% and 5.6%, respectively, could be an effective way for maintaining higher methane production and shorter digestion retention [32]. In line with this, the results of SCE.1, where the CH 4 :CO 2 biogas ratio was considered, had the best score when the biogas plants worked with FW in a two-stage model, with a maximum score of 1.00 (SCE.1 in Figure 3a). The co-digestion alternatives also obtained good scores in all the formulated technical scenarios. Such behavior can be explained because co-digestion offers an improvement of the balance of nutrients and the C/N ratio, alleviation of inhibitory effects, and enhancement of methane production kinetics [33,34].  MCA results of the economic scenarios SCE.8, SCE.9, and SCE.10 are plotted in Figure 3h-j. SCE.8 was formulated with the purpose of minimizing fixed costs and SCE.9 to minimize variable costs, while the last scenario, SCE.10, represented a compromise solution where all the economic criteria were considered with 20% weights. In all cases analyzed, the best scores were obtained with single-stage AD models. In the last scenario SCE.11 represented in Figure 3k, where all criteria were evaluated with 9.1% weights, the best biogas plant alternative was MSW-1S with a total score of 0.58.  MCA results of the three environmental scenarios SCE.5, SCE.6, and SCE.7 are represented in Figure 3e-g, respectively. When the CO 2 .eq is considered with a weighing of 100% (SCE.5 in Figure 3e), all two-stage AD model approaches resulted in better scores than those of the single-stage AD model approaches. This behavior can be explained because the direct CO 2 emissions in the biogas are lower in the two-stage models than they are in the single-stage models, which is in accordance with the CH 4 :CO 2 ratios analyzed in the technical scenario SCE.1. When the two digestate criteria were evaluated with the same weight of 50% in SCE.6 (Figure 3f), the trend was the opposite of the trend observed in SCE.5. In this case, the single-stage AD model approaches achieved the best scores in the range of 0.61 and 0.56 due to all the side-reactions considered in the two-stage AD model approaches, which translated into higher mass fractions of toxic compounds in the digestate. Finally, when all the environmental criteria were evaluated with the same weights of 33.3% in SCE.7 (Figure 3g), the highest score of 0.53 was obtained in two biogas plant alternatives: MSW-2S and co-digestion of MSW and FW-2S.
MCA results of the economic scenarios SCE.8, SCE.9, and SCE.10 are plotted in Figure 3h-j. SCE.8 was formulated with the purpose of minimizing fixed costs and SCE.9 to minimize variable costs, while the last scenario, SCE.10, represented a compromise solution where all the economic criteria were considered with 20% weights. In all cases analyzed, the best scores were obtained with single-stage AD models. In the last scenario SCE.11 represented in Figure 3k, where all criteria were evaluated with 9.1% weights, the best biogas plant alternative was MSW-1S with a total score of 0.58.

Ranking MCA Results of Selected Criteria
The next step was to show the perspectives of certain criteria. In this case, biogas yield (technical criterion), CO 2 eq. (environmental criterion), and capital and operating costs (economic criteria) were selected.
Biogas yield was chosen because all the literature consulted agrees that it is a key parameter at any AD biogas plant [35][36][37][38]. CO 2 eq. was chosen because a key point of this research was the reduction of GHG according to the Paris Agreement, and no other biogas plant alternative can fulfill this objective better.
Finally, fixed capital and variable operating costs were also studied in this section because they include the biggest expenses of the biogas plants.
Looking at the ranking scores of the biogas yield (Figure 4a), the single-stage AD plant operating with MSW seems the most adequate alternative. This trend becomes higher as the weight of this criterion increases. The scores at 20%, 40%, 60%, 80%, and 100% were 0.62, 0.70, 0.78, 0.86, and 0.94, respectively. Even though scores constantly increase, the co-digestion mode alternatives become the best choice, reaching a maximum score of 1.00 in the two-stage plant working with all wastes (MSW, FW, and LCB) at the same time. This behavior in co-digestion is in line with the literature that demonstrates how co-digestion is an effective way to enhance the digestion process for better biogas quantity and quality [30].
The best results in capital and operating costs (Figure 4c) were obtained in the singlestage AD model approaches either in mono-digestion or in co-digestion modes. In none of the scenarios did the two-stage alternatives have better results than those of the single-stage biogas plants.

Robustness Analysis
Sensitivity analysis (Figures 5 and 6) assesses the influence of the weights assigned to each criterion, while uncertainty analysis (Figure 7) assesses the effect of uncertainties in the criteria scores.   In general, alternatives that use MSW in a single stage are the best biogas plants fo most of the technical and economic criteria.
Results of the uncertainty analysis are plotted in Figure 7. The robustness of th MCA-WS ranking results was demonstrated by changing the impact matrix values to 10% (Figure 7a) and 25% (Figure 7b).
Biogas plants (alternatives) are represented on the y-axis, whereas the ranking pos tion is represented on the x-axis. The size of the circles is proportional to the probabilit that each biogas plant alternative occupies a certain position in the rank order. The large sized circles on the main diagonal indicate that, despite the scores deviating from the as signed values of up to 10%, the ranking hardly varied. This behavior changes in Figure 7 when the criteria values change by 25%. In this case, the biogas plant alternatives locate in the second, third, eighth, ninth, and tenth ranking position might vary based on th smaller circle size in comparison with the circle size at 10% uncertainty (Figure 7a). Th A key method to determine the robustness of the findings of an MCA study is to determine how sensitive the results are to changes in the weighting factors [39]. In the previous sections, MCA-WS results from eleven scenarios, with different weights and combinations of the criteria, were discussed. Next, the effect of the weight of each criterion over the ranking score was analyzed. It is of great importance to know how the final ranking of the alternatives is sensitive to the changes of some input parameters of the decision model [22]. For this reason, the evolution of the ranking order (y-axis) with the weight distribution (x-axis) is represented in Figures 5 and 6. The sensitivity analysis of the technical and environmental criteria is displayed in Figure 5; the sensitivity analysis of the economic criteria is shown in Figure 6.
sponding to MSW-2S, FW-1S, LCB-1S, all wastes co-digestion 2-S, and LCB-2S, respectively, might vary. This behavior can be detected based on the smaller circle size in comparison with that of the 10% uncertainty analysis of Figure 7a. Nevertheless, the probability of changing the ranking position is still low since medium size circles prevail over small size circles.

Conclusions
The complexity, heterogeneity, and seasonality of the organic residues collected in the metropolitan area of Reykjavik and the significant amount of technical, economic, and environmental parameters resulting from the biogas plant simulated in Aspen Plus makes it very difficult to establish the optimum solution considering all the stakeholders' viewpoints. Therefore, an adequate decision-making tool able to solve for these concerns was required. In this sense, MCA-WS methodology was applied to evaluate the twelve biogas plant alternatives within eleven different scenarios by using a total of eleven criteria (three technical, three environmental, and five economic criteria). In Figures 5 and 6, the y-axis represents in ascending order the ranking position of the biogas plants (alternatives). The x-axis is the distribution of weights given to a specific criterion. Each line represents a biogas plant alternative. The best alternative is the one located in the highest position (i.e., in Figure 5a, the red line that corresponds to the MSW-1S alternative is the best solution when the weight of biogas yield reaches between 0 and 70%).
Looking at the technical criteria, in general, a single-stage biogas plant working with MSW (red line) seems to be the best alternative at all the weight ranges (from 0 to 100%) except for the methane yield ( Figure 5c) and the CH 4 :CO 2 ratio (Figure 5e). This trend is in line with the biogas yield ranking perspective studied in the previous section (Figure 4a). Regarding the methane yield, when this criterion weights more than 30%, the best alternative is the two-stage plant working with LCB, reaching the maximum score of 1.00 when the importance of the methane yield achieves 100%. In the case of the CH 4 :CO 2 ratio at the beginning, when the weight of this criterion is in the range of 0 to 45%, the scores of MSW-1S, MSW-2S, and FW-1S are quite similar and these three are the best alternatives. However, when this indicator gains importance from 45 to 100% weight, FW-2S becomes the best alternative.
Regarding the environmental criteria, LCB-2S is the worst biogas plant alternative when analyzing the effect of the digestate generation ( Figure 5e) and the digestate contamination (Figure 5f). This trend is the same in both criteria at all the range of weights tested in the two-stage biogas plant alternatives, due to the amount of degradation compounds and reactions considered in the two-stage model approach in comparison with those of the single-stage model. To make these alternatives more sustainable, valorization of the digestate is required. A different behavior was observed in the CO 2 eq. sensitivity analysis (Figure 5f). In this case, the best choice was the FW-2S plant followed by LCB-2S. The same ranking was achieved when looking at the CO 2 eq. weight given in Section 3.3 and represented in Figure 4b. Looking at the impact matrix values, the two-stage model approach has the lowest CO 2 eq. emissions. This behavior is also explained by the CH 4 :CO 2 ratio because the CO 2 present in the biogas outlet streams at the two-stage biogas plants is significantly lower than that in single-stage biogas plants, whereas the CO 2 present in the digestate is higher in the two-stage biogas plants. This can be explained again with the scheme of reactions considered in the two-stage plants.
The last sensitivity analysis was carried out for the economic criteria. Looking at Figure 6, when analyzing equipment costs (Figure 6b) and utility costs (Figure 6d), the best scores at all weights were obtained by plants working with MSW. As was expected, the worst ranking positions when studying capital ( Figure 6c) and operational (Figure 6a) costs were the two-stage plants due to the second reactor and all the costs derived from this issue.
In general, alternatives that use MSW in a single stage are the best biogas plants for most of the technical and economic criteria.
Results of the uncertainty analysis are plotted in Figure 7. The robustness of the MCA-WS ranking results was demonstrated by changing the impact matrix values to 10% ( Figure 7a) and 25% (Figure 7b).
Biogas plants (alternatives) are represented on the y-axis, whereas the ranking position is represented on the x-axis. The size of the circles is proportional to the probability that each biogas plant alternative occupies a certain position in the rank order. The large-sized circles on the main diagonal indicate that, despite the scores deviating from the assigned values of up to 10%, the ranking hardly varied. This behavior changes in Figure 7b when the criteria values change by 25%. In this case, the biogas plant alternatives located in the second, third, eighth, ninth, and tenth ranking position might vary based on the smaller circle size in comparison with the circle size at 10% uncertainty (Figure 7a). The highest score in both analyses was the single-stage biogas plant working with MSW (first green circle) and the worst ranking position belonged to the two-stage biogas plant working in co-digestion mode with MSW and LCB (last red circle). Looking at the 25 % uncertainty analysis for all the criteria (Figure 7b), ranking positions 2, 3, 8, 9, and 10 corresponding to MSW-2S, FW-1S, LCB-1S, all wastes co-digestion 2-S, and LCB-2S, respectively, might vary. This behavior can be detected based on the smaller circle size in comparison with that of the 10% uncertainty analysis of Figure 7a. Nevertheless, the probability of changing the ranking position is still low since medium size circles prevail over small size circles.

Conclusions
The complexity, heterogeneity, and seasonality of the organic residues collected in the metropolitan area of Reykjavik and the significant amount of technical, economic, and environmental parameters resulting from the biogas plant simulated in Aspen Plus makes it very difficult to establish the optimum solution considering all the stakeholders' viewpoints. Therefore, an adequate decision-making tool able to solve for these concerns was required. In this sense, MCA-WS methodology was applied to evaluate the twelve biogas plant alternatives within eleven different scenarios by using a total of eleven criteria (three technical, three environmental, and five economic criteria).
The best solution from a technical viewpoint was the anaerobic digestion biogas plant working at co-digestion in a two-stage model approach, reaching a maximum score of 0.88 when the three technical criteria were considered in SCE. 4.
From an environmental perspective (SCE.5 to SCE.7) there was a dichotomy: (i) when the most important criterion was the CO 2 eq., the two-stage model approaches obtained the best scores due to its lower amount of direct CO 2 emissions in the biogas; (ii) when the two digestate criteria were evaluated, the three single-stage co-digestion plants alternatives obtained the best scores (three of them with the same score of 0.61) because the reaction schemes did not contemplate the by-products presented in the digestate.
From an economic perspective, all the evaluated scenarios (SCE.8 to SCE.10) presented the best results in the single-stage model approach. In general, best scores were obtained in the biogas plant using FW in a single stage, reaching scores of 0.65 (SCE.8) and 0.48 (SCE.10).
Finally, a compromise solution was obtained in the last scenario (SCE.11) where all criteria were analyzed under the same weights of 9.1%. In this case, as occurs in the technical scenarios, the best choice was the biogas plant fed with MSW in a single-stage model anaerobic digestion approach.