MATTI – a multi-criteria decision analysis framework for assessing wastewater treatment technologies

AFP

Aerated Facultative Pond

AHP

Analytic Hierarchy Process

AP + FP

Anaerobic Pond & Facultative Pond

AP + FP + MP

Anaerobic Pond & Facultative Pond & Maturation Pond

AS-C

Activated Sludge – Carbon removal

AS-EA

Activated Sludge – Extended Aeration

ASN

Activated Sludge – Nitrogen removal

BOD

Biochemical oxygen demand

CAPEX

Capital expenditure ($.inhabitant ⁻¹.year⁻¹)

CMAP + SP

Complete Mix Aerated Pond & Sedimentation Pond

FP (P)

Facultative Pond – Primary

MDCA

Multi-Criteria Decision Analysis

MATTI

Most Appropriate Treatment Technology Index

O&M

Simplicity in operation and maintenance (being 1 for complex – 5 for simple)

OPEX

Operational expenditure ($. inhabitant⁻¹.year⁻¹)

PC

Power consumption (kWh. inhabitant ⁻¹.year⁻¹)

RBC

Rotating Biological Contactors

SAR

Surface Application Rate (m³.m⁻².day⁻¹)

SD

Sludge to be disposed (L. inhabitant ⁻¹.year⁻¹)

SDG

Sustainable Development Goals

TF-HR(P)

Trickling Filter – High Rate (Plastic media)

TF-HR(R)

Trickling Filter – High Rate (Rock media)

TF-LR

Trickling Filter – Low Rate

UASB

Upflow Anaerobic Sludge Blanket

UASB + AFP

Upflow Anaerobic Sludge Blanket & Aerated Facultative Pond

UASB + AS-C

Upflow Anaerobic Sludge Blanket & Activated Sludge – Conventional

UASB + BAF

Upflow Anaerobic Sludge Blanket & Biological Aerated Filter

UASB + CMAP + SP

Upflow Anaerobic Sludge Blanket & Complete Mix Aerated Pond & Sedimentation Pond

UASB + TF-HR(P)

Upflow Anaerobic Sludge Blanket & Trickling Filter – High Rate (Plastic media)

UASB + TF-HR(R)

Upflow Anaerobic Sludge Blanket & Trickling Filter – High Rate (Rock media)

UASB + PP

Upflow Anaerobic Sludge Blanket & Polishing Pond

The exponential population growth and the unplanned urbanization process in recent decades have intensified environmental and public health problems worldwide. According to Delanka-Pedige et al. (2020), currently, almost half of humanity already lives in urbanized centers, and by 2050, this number is expected to increase to two thirds. This scenario, therefore, shows a marked increase in sewage production in cities, which must be properly treated before being released into the environment. However, it is estimated that only 20% of the sewage produced worldwide is treated before being discharged (Rodriguez et al. 2020). In Brazil, according to ANA (2020), only 43% of the population has its sewage collected and treated and 12% adopt individual processes, such as septic tanks. In this context, it is considered that 55% of the Brazilian population has an adequate sanitary wastewater solution. Of the remainder, 18% have collected but untreated sewage, and 27% have neither collection nor treatment. Currently, the new legal framework for sanitation in Brazil, Law 14.026/2020 (Brazil 2020), is seen as a possible driver for the universalization of sanitation systems in the country.

The discharge of raw wastewater into water bodies causes, in addition to other serious problems of pollution, contamination and eutrophication of water bodies, and the proliferation of infectious diseases. According to Sims & Kasprzyk-Hordern (2020), these problems are considered one of the most critical threats to global public health today. Delanka-Pedige et al. (2020) state that the basic infrastructure to support society, such as clean water, food, energy, and sanitation services, is already facing serious supply issues, especially in regions with low and medium economic development. In this context, SDG #6 stands out, with target 6.2 that aims to ensure that everyone safely uses sanitation services managed and operated with quality, regardless of region and accessible technologies. According to Findikakis et al. (2020), to achieve target 6.2, it is important to use solutions that represent the best option in different parts of the world, as well as innovative solutions, with the active participation of engineering proposing appropriate solutions, with creative and circular economy concepts in sanitation management.

However, the lack of adequate conception of wastewater treatment systems leads to projects inappropriate to the local reality, with the consequent low performance of the units. This whole scenario, which takes into consideration the low rates of sewage treatment and the low levels of operational quality of the implemented systems, aggravates the problems of raw water quality for general uses such as domestic supply and irrigation, besides emphasizing the water shortage at different levels and in several regions of the planet, already so present today (Lima et al. 2020). Studies at the watershed level reiterate that the implementation of wastewater treatment systems is key to river rehabilitation (Bringer et al. 2018; Costa et al. 2021).

The process of defining technologies for wastewater treatment is complex due to the different alternatives of treatment technologies available and the different types of desirable characteristics (Castillo et al. 2016). In this sense, Multiple Criteria (or MultiCriteria) Decision Analysis (MDCA) allows the use of assigned weights to prioritize needs, according to user preference, to improve the decision-making process amid the complexity of choosing among alternatives (Achillas et al. 2013; Marttunen et al. 2017). Many applications of MDCA models are known for wastewater treatment plant (WWTP) selection, but are limited to select best alternatives for specific scenarios or situations, such as to treat industrial effluents (Castillo et al. 2017); to treat municipal wastewater for small communities (Molinos-Senante et al. 2015); and to optimize water quality in the watershed, under water management principles (Bringer et al. 2018).

Given that the available tools have limited scope that limits their applicability, the objective of this study was to develop a ready-to-use decision support methodology. The methodology applies an Analytic Hierarchy Process (AHP) based solution (Wiecek et al. 2016), to select the most appropriate technologies for wastewater treatment applicable to any desired scenario and available technology. The methodology proposed in this paper is expressed by the Most Appropriate Treatment Technology Index (MATTI), which encompasses the physical characteristics and performance of each technology, adjusted for the local needs of the decision-maker. The index is based on the fundamentals of the multi-criteria analysis and was applied to the Brazilian reality for validation of the proposed tool.

Currently, there are many technologies available for wastewater treatment in Brazil and worldwide (Goffi et al. 2018). However, their different characteristics end up subsidizing long discussions about the choice of the most appropriate technology for local conditions. In this sense, in the first stage of MATTI's development, a survey was carried out on the main technologies available and accessible for the treatment of effluents. Afterwards, the technologies that meet the standards for effluent discharge, established in legal regulations, were selected.

Different characteristics involve the distinct wastewater treatment technologies: area demand, quality performance, mechanization rates, electric power consumption, implementation costs (CAPEX) and operation and maintenance costs (OPEX), and operational complexity, among others. Thus, in the context of the present work, in this step, the main characteristics absorbed in the development of MATTI were indicated.

The performance evaluation of wastewater treatment units is based on removal efficiency values of certain water quality parameters. In this step, the most relevant parameters to be considered for the performance evaluation of wastewater treatment plants were initially determined. It is emphasized that the water quality parameters must be compatible with its destination. If the purpose is to be discharged into water bodies, the applicable legislation must be observed; if the objective is to reuse the effluent, the parameters must be defined according to the quality requirements for the desired purpose. From this, only the technologies that meet the defined water quality parameters were selected.

The definition of the weights for the selected parameters was performed based on the priorities and needs of each location. Since some analyzed characteristics are less desired as their numerical values grow, i.e., implementation costs, all values were subtracted from 1 (1-Normalized). In this way, the negative characteristics become of lesser value.

Depending on local characteristics, each region has specific needs and should adopt the criteria that best suit its reality. Thus, the decision-maker must identify these demands to allocate the weights appropriately. A scale between 1 and 5 is recommended, where 1 is not very relevant and 5 is very relevant. To avoid excessive prioritization, a strategy to guarantee the adequate allocation of weights is to define a maximum of points to be distributed, for example, three times the number of analyzed parameters. Otherwise, the weighting may not reflect the real need of the population served, and the tool loses its meaning.

For a better identification and understanding of the results, after determining the final index of each technology, they are classified according to the following categories and their respective colors:

• > 0.75–1.0: ;

• > 0.5–0.75: ;

• > 0.25–0.5: ;

• 0.0–0.25: .

To select the most appropriate technology to meet the standards for discharging effluents into Brazilian water bodies, the MATTI methodology was applied to assist in decision-making by managers. In this context, we adopted the following aspects for each stage of the work:

• Step 1: The main technologies, or a combination of them, adopted in the 3,668 WWTP in operation in Brazil were evaluated (ANA 2020);

• Step 2: The following variables were defined, according to Trianni Negri & Cagno (2021) to characterize the technologies: surface application rate (SAR) (m³.m⁻².day⁻¹); power consumption (PC) (kWh. inhabitant⁻¹.year⁻¹); sludge to be disposed of (SD) (L. inhabitant⁻¹.year⁻¹); simplicity in operation and maintenance (O&M) (1 complex – 5 simple); capital expenditure (CAPEX) ($.inhabitant⁻¹.year⁻¹); and operational expenditure (OPEX) ($.inhabitant⁻¹.year⁻¹).

• Step 3: The National Environmental Council Resolution No. 430/2011 (Brazil 2011) was adopted, which addresses the effluent discharge standards in the Brazilian territory, as a criterion for evaluating the performance of the WWTPs. Thus, only the typical BOD removal efficiency (%) was indicated, since nutrient standards and fecal contamination indicators are not established in the Brazilian regulatory document.

• Step 4: Weights were assigned in the application of the MATTI methodology in two hypothetical and representative scenarios in the context of Brazilian cities.

The ranges of values assigned to the variables were obtained in the Brazilian technical-scientific literature (Von Sperling 2014; Chernicharo 2016; Jordão & Pessôa 2017) and in the Brazilian Standard NBR 12.209:2011, for sanitary hydraulic designing wastewater treatment plants (ABNT 2011). The final values assigned to each variable were the mean values of the data ranges obtained in the literature. The exchange rate considered for the CAPEX and OPEX was 5.05 Brazilian Reais for 1 U.S. Dollar from Central Bank of Brazil (2021).

The surface application rates were determined after the reactors' design, by following the standards and parameters established in Brazilian Standard (ABNT 2011) and Von Sperling (2017), for ponds. Then, applying the flow-to-area ratio, the SAR were determined with the units (m³.m⁻².d⁻¹). Table 1 presents the SAR equations for each technology, applying the design standards and derived from the flow-to-area ratio.

Table 1

Summary of achieved surface application rates for each technology determined from the flow-to-area ratio

Technology .	Surface Application Rate (m³.m−2.d−1 or m.d−1) .
Upflow anaerobic sludge blanket
Activated sludge (carbon removal, nitrogen removal and extended aeration)
Rotating biological contactors
Trickling filter (high rate and low rate), Biological aerated filter and anaerobic pond
Facultative pond
Aerated facultative pond, complete mix aerated pond, sedimentation pond, maturation and polishing pond

Technology .	Surface Application Rate (m³.m−2.d−1 or m.d−1) .
Upflow anaerobic sludge blanket
Activated sludge (carbon removal, nitrogen removal and extended aeration)
Rotating biological contactors
Trickling filter (high rate and low rate), Biological aerated filter and anaerobic pond
Facultative pond
Aerated facultative pond, complete mix aerated pond, sedimentation pond, maturation and polishing pond

where: D – Depth (m); – Hourly coefficient adjustment to average daily flow; HRT – Hydraulic retention time (d); – Influent BOD concentration (g BOD.m⁻³); – Food-to-microorganisms ratio (g BOD.g VSS⁻¹); MLVSS – Mixed liquor volatile suspended solid (g VSS.m⁻³); SAR – Surface application rate (m³.m-².d⁻¹); VOL – Volumetric organic load (g BOD.m⁻³.d⁻¹); SLR – Surface loading rate (g BOD.m⁻².d⁻¹).

Obs: The hourly coefficient adjustment to average daily flow (⁠⁠) is only applied to the UASB reactor according to Chernicharo (2016) design procedure. The other technologies are designed according to the average daily flow, as per the Brazilian Standards (ABNT 2011). The adopted value for was 1.5 (ABNT 2011; Von Sperling 2014).

View Large

Table 2 shows, for all the wastewater treatment technologies defined in the study, both the variables for characterization of the units and the typical BOD removal efficiency for performance evaluation.

For a better understanding of the application of the MATTI methodology, two representative scenarios were simulated, among the 5570 Brazilian municipalities, with opposite characteristics and conditions, requiring the application of distinct weights (Step 4) according to local priorities and needs. Based on the Brazilian context, a country with great cultural, socioeconomic, and environmental (e.g., physical and geographical, among others) differences, each scenario presented the following characteristics:

• Scenario 1: Considering that about one-third of the Brazilian population resides in metropolises (IBGE 2018), it was considered a compatible scenario for a municipality with area restrictions for the construction of the WWTP and imposition of more restrictive standards for effluent discharge, but with a high capacity for tariff collection and available resources for construction, maintenance, and operation of sanitation services.

• Scenario 2: Considering that 87.5% of the Brazilian municipalities are small (less than 50 thousand inhabitants) (IBGE 2021), scenario 2 was simulated considering a municipality with a large land availability for the WWTP construction and less restrictive local standards for effluent discharge, but with low capacity to collect tariffs and limited resources for construction, maintenance and operation of sanitation services.

It is noteworthy that other scenarios can be simulated using the MATTI calculator, developed by the authors, and made available in the supplementary material. This calculator was implemented as an Excel spreadsheet (not fully compatible, yet functional, in LibreOffice and Google Sheets) and is customizable, from varying weights, adding analyzed variables, to adding technologies.

According to Goffi et al. (2018), one of the most efficient ways to select the best technologies to be applied in the design of a WWTP is from conducting in-depth studies on determining the best criteria and variables involved in the decision-making process by managers. Still, according to the same authors, due to the low investments in sanitation in developing countries, the economic factor (CAPEX and OPEX) should have a significant weight in decision-making.

Some related studies highlight the importance of selecting some criteria, even in developed countries, to help managers be informed to select the most appropriate technologies for specific local characteristics. Some of the recommended features to be considered in wastewater treatment technologies' selection are: technical (efficiencies, simplicity of operation), socio-environmental (area demand, energy demand, and discharge limits), and economic feasibility (construction, maintenance, and operation costs, and cost-effectiveness) (Goffi et al. 2018; Ling et al. 2021; Salamirad et al. 2021).

The application of the five steps of the MATTI methodology, including the determination of the SAR and other variables (BOD reduction efficiency; operational complexity; power consumption; sludge to be disposed of; implementation cost – CAPEX; and operation and maintenance costs – OPEX) for each selected technology, can be seen in the results presented for scenarios 1 and 2, in Tables 3 and 4, respectively. Further details, such as the results of the normalized values, can be found in the supplementary information.

Table 3

Application of MATTI methodology and normalization/weight assignment – scenario 1

MATTI: highly recommended in blue (>0,75–1,0); adequate in green (>0,5–0,75); poor adequate in yellow (>0,25–0,5); inadequate in red (0,0–0,25). The acronyms are the same as presented in Table 2.

View Large

Table 4

Application of MATTI methodology and normalization/weight assignment – scenario 2

MATTI: highly recommended in blue (>0,75–1,0); adequate in green (>0,5–0,75); poor adequate in yellow (>0,25–0,5); inadequate in red (0,0–0,25). The acronyms are the same as presented in Table 2.

View Large

Most of the wastewater treatment technologies among the best appropriate to meet the needs and priorities in scenario 1 involve the UASB reactor. The flowchart composed of UASB + trickling filter of high rate with plastic media support was the most recommended, reaching a final score of 0.77. However, the configuration composed of UASB reactor + submerged aerated biofilter stood out as suitable with a MATTI value of 0.68, which is considered the second most suitable option. The technology known as submerged aerated biofilter was also considered the most suitable for some locations whose performance requirement is high, with the possibility of nitrification, and low space demand (Castillo et al. 2016; Bringer et al. 2018).

In Brazil, among the technological alternatives for sewage treatment, UASB reactors stand out for their low cost of implementation, small space requirements, simple operation, and potential for energy recovery (Rodrigues Mesquita et al. 2021). Such characteristics confirm the data from a recent survey conducted in Brazil, which highlights more than 650 reactors in operation, at full scale (ANA 2020). Of these, approximately 90% are installed in the south and southeast regions, with a capacity to serve nearly 23 million inhabitants and an operational flow rate of approximately 43 m³·s⁻¹ (Chernicharo et al. 2018). According to Rodrigues Mesquita et al. (2021) about 67% of the WWTPs with UASB reactors either have no post-treatment, or present post-treatment technologies with low-energy demand, such as biological filters or stabilization ponds. These data corroborate the results presented in Table 3, which classifies UASB reactors and flowcharts composed of them as appropriate for scenario 1. It should be noted that UASB reactors can also be considered the most widespread anaerobic process used to treat domestic wastewater in the world (Chernicharo et al. 2015).

On the other hand, although the UASB reactor presents characteristics such as high organic load capacity, low-energy cost compared to aerated systems, and the possibility of using the biogas produced as an alternative energy source in the WWTP, they have a limited effect on nutrients (nitrogen and phosphorus), pathogenic organisms, and solids (Mainardis et al. 2020). Moreover, these reactors require post-treatment of the effluent to meet the legislative standards required in Brazil (Ribeiro et al. 2020). According to ANA (2020), there is significant variability of post-treatment systems for UASB reactors in Brazilian WWTPs. However, 42% of these are represented by activated sludge systems, including their variants, and aerobic filters.

From the application of the MATTI methodology and the results of Table 3, it was possible to verify that the treatment technology that was most often classified as adequate or highly recommended was the UASB reactor, with seven indications, from the individual reactors to combinations with another technology. As an option of post-treatment technology (secondary process), the high rate trickling filters (with plastic or rock support medium) is indicated as adequate or highly recommended, corroborated by Bressani-Ribeiro et al. (2018) findings. The single trickling filter technologies differ from the ones associated with UASB reactors because the first assumes primary clarifier (PC) as the primary treatment unit, which removes a considerable lesser amount of BOD (60% for UASB and 30% for PC). Since the trickling filter is designed according to the organic load, as the lower the influent BOD, the smaller the reactor shall be, which is reflected on the higher SAR for the combined UASB + TF.

Santos et al. (2021) indicate an inconsistency between the hydraulic-sanitary design parameters established by Brazilian Standard (ABNT 2011) for high rate trickling filters with plastic support medium. This Standard indicates that the TF-HR (plastic) has a height limit of 12 m, which could, potentially, achieve a SAR over 120 m³.m⁻².day⁻¹. Nevertheless, this SAR is theoretical since the Brazilian Standard limits the SAR to 75 m³·m⁻²·day⁻¹, the value adopted by this study, as presented in Table 2. Moreover, Santos et al. (2021) discuss that plastic media provides advantages such as promoting slender structures, occupying less area and higher specific surface area. Regardless, since this has not yet been experimentally demonstrated and is not allowed by the Brazilian Standard, the theoretical potential of plastic media calculated is not fully achievable.

In scenario 2, the primary facultative ponds technology was the most recommended, achieving the index of 0.87 (highly recommended). However, the individual UASB technology and the sequence of stabilization ponds (anaerobic ponds + facultative ponds) were also rated as highly recommended, achieving the index of 0.85 and 0.86, respectively.

In addition to the UASB reactors, facultative ponds were also indicated by Bringer Reis & Mendonça (2018) and Molinos-Senante et al. (2015), as systems that require simplified construction, operation and maintenance, with virtual zero energy expenditure, low sludge production, and for small communities. Such technologies individually or in the composition of other flowcharts, besides presenting better classifications, according to data and information from the scientific literature, also stood out in the parameters of energy consumption, simplicity of maintenance and operation, CAPEX, and OPEX.

Wastewater treatment systems have been improved both to achieve adequate traditional performances (Goffi et al. 2018) and to meet the new demands of society regarding the removal of compounds with the potential to cause toxic effects on exposed biota. However, understanding the capability of the different treatment technologies in bringing down their concentration to non-toxic levels is a challenging task for many of decision-makers and an area that needs significant attention (Parida et al. 2021). The greater amount of available wastewater treatment technological alternatives has been making the process of selecting the most suitable technologies that at the same time meet local needs, an even more complex task for managers (Achillas et al. 2013; Rabello et al. 2019; Parida et al. 2021). Thus, the MATTI methodology helps managers make decisions that are technical, democratic, and based on the demands of society, as recommended by Achillas et al. (2013).

Activated sludge systems and their variants, when evaluated from MATTI values, were classified as inadequate for scenario 2, being the ones that presented the lowest values, especially in energy consumption (high demand for energy) and simplicity of maintenance and operation (high costs and difficulties inherent to the process). Economic analysis is fundamental for any type of investment decision, especially in sanitation projects, where direct financial return is not always guaranteed, and future gains appear indirectly, which makes this decision even more complex (Goffi et al. 2018). According to user-friendly, Kalbar et al. (2012), the conventional activated sludge system is generally more suitable for urban settings with little availability of areas for construction of the WWTP and with a higher demand on the quality of the final effluent, imposed by the regulatory document or by the quality of the receiving water body.

In scenario 2, the UASB system was also highlighted. In this context, it was classified as suitable or highly recommended six times, highlighting its participation as an individual unit or in the composition of some flowchart, especially with trickling filters and their variants. The main advantages presented by the trickling filters and its variants have a simpler design, easy construction, little space occupied, low-energy requirements, and low operation and maintenance costs (Bressani-Ribeiro et al. 2018; Bezirgiannidis et al. 2019). Hence, Santos et al. (2021) point out that, when compared to low rate filters, high rate trickling filters have lower area requirements, since in these cases, it is possible to apply a greater hydraulic load per unit area (SAR concept). However, there is a slight reduction in performance due to the effluent recirculation requirement that helps to equalize the flow rate and influent load, and remove additional organic matter; which led to an increase in electricity consumption, OPEX and operational complexity.

For scenario 2, the options involving stabilization ponds also stand out, with the verification of eight adequate combinations, with high scores (0.56–0.87). The factor of large space availability and the criteria of low-energy consumption and simplicity of maintenance and operation were determining factors for such high scores. The technologies indicated as most suitable for scenario 2 were also recommended, according to the economic evaluation carried out by Goffi et al. (2018). Furthermore, according to these authors, the wastewater treatment technologies that presented the lowest costs (following the economic criteria determined in the study) were trickling filters, UASB reactors, anaerobic ponds, and facultative ponds.

Each WWTP project is unique and requires an appropriate solution that should preferably be evaluated in multicriteria analysis. This action tends to ensure the solution's suitability to the legal requirements, local needs and priorities. Therefore, it is necessary to consider, besides the simplistic expected compounds removal performance, other variables simultaneously (simplicity of operation; area demand; energy demand; discharge limits; and the construction, maintenance, and operation costs) to assist decision-making. An adequate and careful conception represents the most important aspect to achieve the expected goals towards implantation and operation of a wastewater treatment plant: environmental and public health protection.

The two hypothetical scenarios outcomes were compared to other methodologies developed for specific situations, which validated MATTI's methodology. The calculator was developed to simulate different scenarios, allowing for sensibility analysis by assigning different weights and design criteria, depending on the desired outcomes. The final 0–1 index is classified into four categories and represented by a color pattern aimed to assist in the process of selecting individual technologies or completing optimal flowcharts for each scenario, considering different factors simultaneously.

Thus, MATTI, through its user-friendly interface, presented as an accessible and comprehensive tool capable of assisting managers in decision-making processes, especially in developing countries, such as Brazil, where wastewater treatment technologies alternatives are not often scrutinized. Therefore, this multicriteria framework helps to reduce the subjectivity in wastewater treatment facilities selection and eases the process to boost sanitation development towards its universalization.

This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brazil (CAPES) - Finance Code 001 and National Council for Scientific and Technological Development – Brazil (CNPq).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.