Framework for a multi-criteria decision support system for water permit allocation in Brazil

Estrutura para um sistema de suporte à decisão multi-critério para a alocação de permissões de água no Brasil

ABSTRACT

This study presents an operational framework for a decision support system (DSS) aligned with Brazilian water resource legislation, designed to manage the allocation of surface water resources (WRA) through water permit concessions, aimed at resolving conflicts between competing users seeking withdrawals and effluent dilution. The DSS integrates a one-dimensional model for assessing water availability with a multi-criteria decision analysis module, employing the Analytic Hierarchy Process combined with fuzzy logic for criteria weighting. When simulated water availability falls below sustainable limits for a new water use permit, the system triggers a processing scheme to propose reallocation scenarios for withdrawals and effluents among new and existing upstream users. The DSS then provides ensembles of optimal solutions to ensure that previously defined minimum water availability targets are met in the stream. A case study in the Cuiabá River basin validates the functionality of the modular system. Given the decentralized and participatory nature of Brazilian water resource legislation, the WRA-DSS framework is flexible and prepared for the inclusion of new decision criteria or water protection goals, as well as the integration of more advanced flow and water quality models.

Keywords:
Water resource allocation; Brazil; MCDA; AHP; Fuzzy logic

RESUMO

Este estudo apresenta uma estrutura operacional para um sistema de suporte à decisão (DSS) alinhado à legislação brasileira de recursos hídricos, projetado para gerenciar a alocação de recursos hídricos superficiais (WRA) por meio de concessões de permissões de uso de água, com o objetivo de resolver conflitos entre usuários competidores que buscam retiradas e diluição de efluentes. O DSS integra um modelo unidimensional para avaliar a disponibilidade de água com um módulo de análise de decisão multi-critério, empregando o Processo de Hierarquia Analítica combinado com lógica fuzzy para o ponderamento dos critérios. Quando a disponibilidade de água simulada fica abaixo dos limites sustentáveis para uma nova permissão de uso de água, o sistema aciona um esquema de processamento para propor cenários de realocação de retiradas e efluentes entre usuários novos e existentes rio acima. O DSS então fornece conjuntos de soluções ótimas para garantir que as metas mínimas de disponibilidade de água previamente definidas sejam atendidas no curso d'água. Um estudo de caso na bacia do Rio Cuiabá valida a funcionalidade do sistema modular. Dada a natureza descentralizada e participativa da legislação brasileira de recursos hídricos, a estrutura WRA-DSS é flexível e preparada para a inclusão de novos critérios de decisão ou metas de proteção hídrica, assim como para a integração de modelos mais avançados de fluxo e qualidade da água.

Palavras-chave:
Alocação de recursos hídricos; Brasil; MCDA; AHP; Lógica fuzzy

INTRODUCTION

Population growth and ongoing economic development are driving a steadily increasing demand for water, particularly in developing countries (Paquin & Cosgrove, 2016). This growing demand is often exacerbated by reduced availability due to climate change-induced droughts, increased pollution, and the degradation of aquatic ecosystems (Wiek & Larson, 2012). To address these challenges, legislators worldwide are working to prevent supply shortages for multiple uses and to ensure the ecological functioning of freshwater ecosystems through the regulation and control of water resource allocation.

Despite having the largest amount of freshwater in the world (Getirana et al., 2021), Brazil faces significant disparities in water availability, water scarcity, and health issues related to inadequate supply and deteriorating water quality across its diverse hydroclimate regions (Andrade et al., 2020; Chiquito Gesualdo et al., 2021).

Inspired by the French model of water resource management, Brazilian federal and state legislations established since 1997 are characterized by a decentralized and participatory approach (Cánepa et al., 2017). Policy instruments include the issuance of water permits and concessions known as “Outorga”, which are based on estimates of low flow availability and the loss of freshwater resources due to the river's dilution capacity. The legislation typically adopts criteria such as a minimum ecological discharge in river reaches, which must be maintained throughout the year, and a maximum concentration of Biochemical Oxygen Demand (BOD₅) under low flow conditions. These upper limits are determined by a classification system designed to sustain water quality for the most demanding use of a river reach (“Enquadramento”). For instance, a class 2 river must maintain BOD₅ concentrations no higher than 5 mg/l, making it suitable for human consumption after conventional drinking water treatment.

When the Water Availability (WA) of a stream has been exceeded by existing water concessions, it is common practice for water resource authorities, endorsed by regulatory legislation, to deny any additional water permits. However, this approach may violate fundamental principles of the Brazilian National Water Resources Policy, as defined in the National Water Act of 1997 (Law 9.433). These principles include: (i) prioritizing water use to meet basic human needs, especially during critical periods, and (ii) ensuring that water management encompasses and encourages multiple uses. For example, in a rural watershed where the majority of available water resources have historically been allocated for irrigation, a downstream urban sanitation agency might be prevented from withdrawing water to meet the increasing demand for human consumption or from diluting effluents from a wastewater treatment plant.

Decision-making (DM) in water resource management (WRM) often involves balancing multiple, sometimes conflicting, objectives (Davijani et al., 2016). Consequently, DM processes in WRM are complex and can yield intangible outcomes (Gough & Ward, 1996; Hajkowicz & Higgins, 2008), which may reduce transparency, particularly for non-technicians, and limit the ability to audit decisions. Theoretical-methodological frameworks for WRM in administered, market-based, or combined Decision Support Systems (DSS) have been developed, and their efficiencies compared (Zhao et al., 2013). This study focuses on developing an administered water permit concession system, where governmental agencies control and regulate water allocation according to established laws. However, the proposed tool should also be applicable for strategic planning, offering long-term perspectives for sustainable basin management.

Complex DM problems can be effectively addressed using Multiple Criteria Decision Analysis (MCDA), a framework for evaluating alternative decision options against multiple, hierarchical criteria (Hajkowicz & Higgins, 2008). MCDA has been successfully applied in WRM (Calizaya et al., 2010; Zolghadr-Asli et al., 2021) and, more specifically, in addressing water allocation challenges (Yan et al., 2017; Golfam et al., 2019; Gorgoglione et al., 2019). According to a review by Zolghadr-Asli et al. (2021), the Analytic Hierarchy Process (AHP) is the most frequently used algorithm in WRM DSS and performs particularly well when different conflicting interest groups must be considered in DM or when decision-makers need to learn from the application of the MCDA tool.

Here, we propose a modular computational implementation of a Decision Support System (DSS) for water permit concessions, called Outorga-Web. Although it has not yet been used operationally, its implementation would require amendments to existing regulatory decrees to prevent legal disputes. Nevertheless, the framework is fully functional and, in our view, can support the implementation of Water Act principles, including water use prioritization, multiple water uses, and the decentralization of water resource management. To address water allocation challenges while minimizing negative impacts on surface water volume and quality, the framework employs a fuzzy logic inference engine combined with an Analytical Hierarchy Process-based (FAHP) multi-criteria approach. Additionally, it facilitates the creation of negotiation scenarios to help limit necessary restrictions on already allocated permits for existing users.

SYSTEM CONCEPTION

Overview

To streamline operational DM in WRA, Da Hora (2001) proposed combining withdrawal (direct use) and effluent dilution (indirect use) into a single demand metric. This metric is expressed as the discharge retrieved from the affected stream reach, a method commonly employed by federal and state environmental agencies for water resource allocation. The criterion for water rights concession is then based on WA in a stream, both before and after the inclusion of a new user.

When a potential user requests a water permit for withdrawal and/or effluent dilution from a stream reach, the first system component estimates WA along the river reaches. This estimation is based on the initial low flow discharge and biochemical oxygen demand (BOD₅) concentrations, taking into account pre-existing withdrawals and effluent inflows (Figure 1). WA at a reach is thus determined as the combination of direct and indirect water uses.

Figure 1
Decision-making Pipeline in the Outorga-Web Application for Water Permit Concession.

If a WA deficit is detected after the preliminary inclusion of a new water user, a DSS component is triggered. The DSS aims to identify a set of possible solutions that limit the withdrawals and effluents of the new and existing upstream water users to meet a predefined minimum WA at the reach. The WRA-DSS, called OutorgaWeb, was developed using CASE GENEXUS software, version 9.0 by Artech Corp. From this tool, a JAVA code was generated, which interacts with a PostgreSQL, version 8.3 database and can be executed independently on any platform in web environments.

Water Availability modelling tool

Following the conceptual framework of Da Hora (2001), a simple one-dimensional model is used for water availability (WA) simulations. WA is estimated by flow balancing, which uses previous estimates of statistical low flow discharges for each reach and applies the one-dimensional prediction method for the auto-depuration of pollutants based on the Streeter-Phelps equations. BOD₅ loads from inflows are transformed into an indirect withdrawal, required for the dilution of pollutants in the stream.

After selecting the first reach and resetting the attributes storing the accumulated direct and indirect water uses along the stream network, the maximum available discharge for water use is calculated (Figure 2). In Brazil, this calculation is governed by the classification of water bodies according to designated uses (“Enquadramento”). For example, a reach targeted for Class 2 (suitable for human consumption after conventional treatment) must not have BOD₅ concentrations higher than 5 mg/L.

Figure 2
Multi-Step Processes in the Water Allocation Modeling Module of the Outorga-Web Application.

BOD₅ loads are then converted into the required discharge for dilution, based on the following principle: sewage affects downstream water quality and consumes a portion of the runoff volume necessary for diluting the effluent. Consequently, stream discharge is needed for both direct (withdrawals) and indirect (effluent dilution) water uses. The residual WA discharge at a river reach segment i at time t, expressed as v_d(i,t), is given by the difference between the maximum WA discharge for water permit licensing (v⁺(i,t)), - typically a percentage of a statistical low flow index - and the sum of all direct (withdrawals) and indirect uses (effluents) upstream, including upstream effluent discharges (Equation 1). Accordingly, v_d(i,t) is the threshold value, which must not be exceeded. In other words, (v⁺(i,t)) must never be lower than the sum of water uses.

${\text{v}}_{\text{d}}\left(\text{i}\text{,t}\right){\text{=v}}^{\text{+}}\left(\text{i}\text{,t}\right)\text{-}{\displaystyle \sum }_{\text{j}\in {\text{ε}}_{\text{M}}}\text{(u}\left(\text{j}\right)\text{.v(j}{\text{,t)+v}}_{\text{q}}\text{(j}\text{,t))+}{\displaystyle \sum }_{\text{j}\in {\text{ε}}_{\text{M}}}\left(\text{1-u}\left(\text{j}\right)\right)\text{.v}\left(\text{j}\text{,t}\right)$ (1)

$\forall \text{εM e}\forall \text{t}$

with:

u(j): coefficient of water usage at a upstream river reach segment j, varying between 0 and 1.

v(j,t): withdrawn/effluent discharge at a upstream river reach segment j at time t

v_q(j,t): discharge consumed for dilution at a upstream river reach segment j at time t

At river reaches with no withdrawal or effluent, the concentration of BOD₅ is estimated as a function of its decay rate k and the time t needed for a water volume to pass through a reach.

For execution, Outorga-Web requires a spatial representation of streams and their network topology as input. Each reach has a unique identifier and attributes to define confluences. Additionally, there are fields to input model entries, including statistical low flow discharge and initial values of water quality parameters. In our case, we used the Q₍₉₅₎ (95% percentile from the flow duration curve) and BOD₅ concentrations, which serve as references for water concession permits in Mato Grosso state.

Multicriteria decision support system module

Premises

The DSS module of Outorga-Web is based on the following premises:

1. The outright rejection of new water concessions after a stream reaches its sustainable limits of WA is not a viable solution, given the increasing conflicts over water use.

2. DM should aim to find solutions that do not favor excessive or low-priority water uses, which may exist in a basin simply because they were the first to receive permits. Instead, we propose that existing users with disproportionate volumes of direct and/or indirect withdrawals should be involved in a negotiation process to allow the inclusion of new water users, especially if these new users have higher priority and adopt practices that minimize impacts on water resources. We consider this approach appropriate, as water permits in Brazil are generally limited to 5 or 10 years, with the exception of those granted to hydropower plants, which have a duration of up to 30 years.

The DSS module has been initially configured to present water resource authorities with a list of up to 10 groupings of water users located upstream of a water user requesting a new concession. Each grouping consists of a maximum of five members. The DSS ranks these groupings according to parameters (initially up to three) that can be selected by the operator. The operator can then decide which grouping is meant to reduce withdrawals and/or dilution and determine the percentage by which each member should contribute to accommodate the new concession.

Implementation

The implemented DSS module is a fuzzy logic inference engine for Multiple Criteria Decision Analysis (MCDA). It is based on the fuzzy decision-making methodologies of Bellman & Zadeh (1970) and its extensions by Yager (1977), which allow for the weighting of criteria when they are not of equal importance to the decision maker. It also incorporates the Analytical Hierarchy Process (AHP) (Saaty, 2008), a method for paired comparisons of criteria for the identification of best alternatives for conflict solving. This integration was first proposed by O’Hagan (1988) and is referred to here as the Fuzzy Analytic Hierarchy Process (FAHP).

In FAHP, solving a complex problem involves four major steps. First, the problem is decomposed into a hierarchy of sub-problems at different levels, where each sub-problem can be evaluated independently. Layer elements may consist of criteria, sub-criteria, constraints, or other relevant factors (Kepaptsoglou et al., 2013).

In our case, four predefined constraints and three criteria were established. Constraints determine whether an existing water user is considered for a reduction in direct or indirect water use and were defined as follows:

1. The user is upstream of the reach in the stream network where the major water availability (WA) deficit occurs.

2. The user does not have a permit for human consumption.

3. The user's water permit is responsible for at least 5% of the WA deficit detected in the WA model.

4. The user has not reduced their water consumption by more than 20%.

The three criteria available to prioritize conflict resolution are:

1. Percentage of reduction in withdrawals – Either a higher reduction applied to a small number of users or lower reductions distributed across a larger number of upstream users, thereby increasing the number of users involved in renegotiation.

2. Remaining duration of existing concessions – Prioritization based on the time left before current water permits expire.

3. Group size – User groups are formed for decision-making, with their concessions subject to reduction. Smaller groups result in higher individual BOD₅ reductions, whereas larger groups lead to lower individual effluent reductions.

A pairwise comparison is then conducted between layer elements, from which a decision matrix X of fuzzy numbers is formulated using a Fuzzy Membership Function—commonly, as in our case, a Triangular one (TFM). Here, X consists of Triangular Fuzzy Numbers (TFN), denoted as x_ij= {l_ij, m_ij, u_ij}, with l_ij, m_ij, u_ij being lower, modal, and upper values (Equation 2).

$\text{X=}\left\{{\text{x}}_{\text{ij}}\right\}\text{=}\left[\begin{array}{cccc}\text{1}& \left({\text{l}}_{\text{12}}{\text{,m}}_{\text{12}}{\text{,u}}_{\text{12}}\right)& \text{…}& \left({\text{l}}_{\text{1n}}{\text{,m}}_{\text{1n}}{\text{,u}}_{\text{1n}}\right)\\ \left({\text{u}}_{\text{12}}^{\text{-1}}{\text{,m}}_{\text{12}}^{\text{-1}}{\text{,l}}_{\text{12}}^{\text{-1}}\right)& \text{1}& \text{…}& \left({\text{l}}_{\text{2n}}{\text{,m}}_{\text{2n}}{\text{,u}}_{\text{2n}}\right)\\ \text{…}& \text{…}& \text{…}& \text{…}\\ \left({\text{u}}_{\text{1n}}^{\text{-1}}{\text{,m}}_{\text{1n}}^{\text{-1}}{\text{,l}}_{\text{1n}}^{\text{-1}}\right)& \left({\text{u}}_{\text{2n}}^{\text{-1}}{\text{,m}}_{\text{2n}}^{\text{-1}}{\text{,l}}_{\text{2n}}^{\text{-1}}\right)& \text{…}& \text{1}\end{array}\right]$ (2)

Fuzzy ranks r and weights w are calculated using the Equations 3 and 4.

${\text{r}}_{\text{i}}\text{=}{\left({\displaystyle \coprod }_{\text{j=1}}^{\text{n}}{\text{x}}_{\text{ij}}\right)}^{\frac{\text{1}}{\text{n}}}$ (3)

${\text{w}}_{\text{i}}{\text{=r}}_{\text{i}}{\left({\displaystyle \sum }_{\text{i=1}}^{\text{n}}{\text{r}}_{\text{i}}\right)}^{\text{-1}}$ (4)

At last, crisp weights are obtained through the defuzzification of weights. Details of the mathematical formalization for the definition (Bellman & Zadeh 1970) and weighting of fuzzy sets can be found in Yager (1977) and O’Hagan & O’Hagan (1993). FAHP is detailed in .

In Outorga-Web, the Decision Support System (DSS) operator—typically a technician at the environmental agency or a member of the water committee responsible for water concessions, ideally supported by domain specialists—conducts pairwise comparisons between criteria. Each comparison reflects the relative importance of the criteria and is assigned a discrete numerical value (Figure 3). In this case, a nine-level reference scale was used for discretization.

Figure 3
Multi-Step Processes in the FAHP Decision Support Module of the Outorga-Web Application.

After resetting temporary tables, all upstream users subject to potential renegotiation are identified. Groups of up to five members are then formed and classified by the FAHP algorithm. Finally, these groups are visualized for the system operator to select a group and, if necessary, adjust the percentage reductions for individual group members.

CASE STUDY

Study area and input data preparation

The WRM tool was set up for the upper and mid Cuiabá River basin, which covers an area of approximately 28,300 km² and is located between 54°45’ and 56°55’ west longitude and 14°15’ and 16°10’ south latitude, in the southern part of the Brazilian state of Mato Grosso, within the Cerrado savannah biome. A pilot study for system validation was conducted in the watershed of the Pari stream (730 km²), a right-bank tributary of the Cuiabá River (Figure 4). The tropical semi-humid climate of the Cuiabá River basin is characterized by mean annual temperatures ranging from 22 to 25 °C and annual precipitation between 900 and 1,800 mm. The basin includes the urban agglomeration of Cuiabá and Várzea Grande, with a population of approximately 800,000 inhabitants in 2010. Wastewater treatment in this area is estimated to serve less than 50% of the total population. The main industrial water users include beverage production, slaughterhouses, and milk processing facilities. Despite their obligation for conventional wastewater treatment, discontinuous measurements reported BOD₅ concentrations higher than 1000 mg/l at river disposal. Fish farming is also an important economic activity in the region, with pond outflows reported to have BOD₅ concentrations exceeding 12 mg/L.

Figure 4
Cuiabá River Basin and Pari Stream Watershed Featured in the Pilot Case Study.

To exemplify and functionally validate the Outorga-Web system, we simulated discharge and diffuse BOD₅ loading using highly simplified modeling approaches.

Basin parametrization was conducted using the 1-arc resolution SRTM 4 Digital Elevation Model (DEM). A total of 338 sub-basins and stream reaches were generated using spatial processing tools available through the simple NGFlow precipitation-runoff model (Ye et al., 1996). NGFlow was also used to simulate a 12-year daily discharge time series (1997-2008) from 12 precipitation stations (Agência Nacional de Águas e Saneamento Básico, 2017). This series was then used to regionalize the Q95 low flow values for the sub-basins using the Regio tool developed by Zeilhofer & Lima (2006). It should be noted that the short duration of the underlying series may limit the robustness of Q95 estimates (Salinas et al., 2013). To ensure compatibility with simulations at the watershed scale, each stream reach was subdivided into 1 km segments using standard ArcMap GIS, 10.5 (ESRI) procedures. Average diffuse BOD₅ loading was estimated from land use and cover (LUC) data obtained from World Wide Fund for Nature (2009) using the simple PLOAD model, which is part of the Better Assessment Science Integrating Point and Nonpoint Sources (BASINS) tool (Kinerson et al., 2009).

For setup, we further included estimates of withdrawals and sewage disposal from: i) Licensed water users according to the State Environmental Agency (SEMA). ii) A spatial cadaster of fish farms obtained from SEMA and iii) The domestic sanitation sector, estimated from population density and average sewage treatment rates for each IBGE census sector (Zeilhofer et al., 2010).

Further details on data processing pipelines for sub-basin and hydrographic network parametrization, discharge modeling, estimation of diffuse BOD₅ loading, and preparation of Outorga-Web inputs can be found in Oliveira (2009). Despite the use of simplistic modeling approaches and uncertainties in input data, comparisons with recently available discharge and water quality data from 29 rural and urban monitoring stations (Universidade Federal de Mato Grosso, 2024) for 2010, 2015 and 2023/24, indicate that discharge and DBO estimates from Outorga-Web have a realistic magnitude.

The DS module of Outorga-Web has been primarily validated in low-order streams within the Precambrian peripheral depression zone of Cuiabá, an area characterized by shallow soils and streams with low base flow ratios. In this region, competition between water users and inflows of untreated domestic wastewater are common. For the following validation run in the Pari stream sub-basin, a hypothetical large food-processing facility was included as a new user in the database, with an average withdrawal of 0.020 m³/s and an effluent discharge of 0.015 m³/s, containing a BOD₅ concentration of 110 mg/l. The river reach is classified as Class 2 under Brazilian legislation, meaning the BOD₅ concentration of the stream must not exceed 5 mg/l under low flow conditions.

Through the initial interface, the system operator can visualize stream reach attributes (“Visualizar trecho”), initiate the inclusion of a new water user for an Outorga-Web simulation run (“Simular inclusão”), or modify system configurations (“Configurações”) (Figure 5).

Figure 5
Main Interface of the Outorga-Web Application.

In this example, the criteria preferences have been set to higher values throughout: a greater percentage reduction in withdrawals from a smaller number of users (Figure 6, orange arrow), a longer remaining concession duration (Figure 6, red arrow), and a higher individual BOD₅ reduction (Figure 6, dark red arrow).

Figure 6
Interface of the Outorga-Web Application for Configuring and Pairwise Comparing the Relative Importance of Criteria.

In the second block of the interface, the operator can define the prevalence between pairs of criteria and assign relative weights. In the example shown, the first comparison weighs the average withdrawal discharge (“vazão média de retirada”) against the average remaining time span of concessions within the user group (“tempo de contrato médio dos grupos”) (Figure 6, double-ended blue arrow). In this case, priority is given to discharge, which is rated as “strongly preferable” (“fortemente preferível”) compared to the remaining concession time span.

For the configuration shown in Figure 6, Equation 5 presents the resulting FAHP comparison matrix and the eigenvector, which represents the weight of each criterion:

$\begin{array}{c}\text{W}\\ \text{T}\\ \text{D}\end{array}\left[\begin{array}{ccc}\text{1}& \text{5}& \text{3}\\ \text{1/5}& \text{1}& \text{8}\\ \text{1/3}& \text{1/8}& \text{1}\end{array}\right]\begin{array}{c}\text{W}\\ \text{T}\\ \text{D}\end{array}\left[\begin{array}{c}\text{0}\text{.57}\\ \text{0}\text{.32}\\ \text{0}\text{.11}\end{array}\right]$ (5)

In this example, withdrawal (W) has the highest weight (0.57), followed by remaining time span (T) (0.32), and BOD₅ load (D) (0.11).

The facility requests a water withdrawal of 0.02 m³/s and a sewage discharge of 0.015 m³/s with a BOD₅ concentration of 110 mg/l at Segment 54 of the Pari stream (Figure 7).

Figure 7
Attributes of Stream Reach Segments After Inclusion of a New User at Segment 54.

According to the low flow discharge and permitted dilution capacity, the discharge deficit in the reach segment is approximately 0.032 m³/s. The generated decision support model identifies six possible user groups that could successfully resolve the conflict (Figure 8).

Figure 8
Conflict Resolution Alternatives Generated by the FAHP Module for Renegotiation of Water Permits to Ensure Licensing of the New User.

The possible solutions are listed in descending order based on the defined criteria. If the system operator selects option 3, the new stream conditions after the inclusion are simulated as shown in Figure 9.

Figure 9
Water Availability (WA) in Pari Stream Reach Segments Upstream and Downstream of the Newly Included User and Reduction of Preexisting User Ensemble.

The “% Outorga” field indicates how much of the available capacity would be utilized after the facility is implemented, in this case, more than 99% at reach segment 54, where withdrawal and sewage discharge will occur.

DISCUSSION

Despite being considered controversial and having shortcomings in terms of effectiveness (Miranda & Reynard, 2020; Marques, 2022), the concession of water permits remains, and will continue to be, a key instrument for establishing sustainable water resource management in Brazil, largely due to its deep integration into the country’s water policies.

However, in practice, the technical procedures for granting permits - such as those used in the state of Mato Grosso - are often static. In our view, these procedures do not fully align with the key principles of the legislation, which include: i) the constitutional right of equal access to water for all; ii) the hierarchical prioritization of water uses, where human consumption and livestock watering take precedence over industrial uses or irrigation; and iii) the decentralization of water use concessions and management decisions, which are intended to be tailored to the local or regional environmental and socioeconomic conditions of each watershed (Cánepa et al., 2017).

If the permissible water use limits have already been reached by upstream users, a new, potentially small or priority water user—even one proposing maximum preservation efforts—may be rejected, while previously approved, large, or low-priority upstream users can continue their consumption without penalty or reduction. This occurs simply because the latter applied for their permits earlier. Water permits are typically granted for limited periods, commonly five or ten years. As a result, if a new downstream user is rejected, the upstream low-priority user may still renew their water permit upon expiration.

We suppose that this situation could lead to legal uncertainties or, at the very least, political pressure on water management authorities tasked with law enforcement, particularly due to socioeconomic factors. Therefore, we propose that a reserve of water allocations be safeguarded, especially in regions where increased demand is projected. Ideally, already approved licenses should be subject to renegotiation based on technical, flexible criteria. These criteria could include the priority of existing upstream users, the remaining duration of their current licenses, and their impact on water availability upstream.

Given this context, Outorga-Web was developed as an administered water permit concession system under the premise that concession permits can be renegotiated under the guidance of water agencies, which are currently predominantly governmental but should increasingly transition to basin committees. We acknowledge that this shift is not trivial in practice and will require additional administrative resources (Marques, 2022). However, we believe that this transition is fundamental to upholding the principle of participatory water management as outlined in Brazilian water law. The presented validation run indicates that the DSS functions effectively and produces outputs consistent with the real discharge and water quality conditions, as well as user conflicts in the pilot watershed.

In its implementation, we adopted the use of Fuzzy rules for criteria definition to facilitate decision-making (Ma et al., 2020). Particularly in data-poor regions, establishing well-founded crisp criteria can be challenging. Following the Brazilian water law's prerogative for decentralized water management, the Outorga-Web source code allows for the modification and quick inclusion of additional criteria in the concession process. In the future, criteria should ideally be modifiable through adaptive user interfaces (Miraz et al., 2021), a task that would require soft coding efforts in the core module to improve scalability (Zhai et al., 2020). Furthermore, Outorga-Web was implemented using a commercial development tool, and efforts should be made to clean up the current Java code and database interfacing to create platform-independent code for component reusability and further development with open-source software.

Thanks to its simple water availability modeling approach, Outorga-Web is computationally efficient, enabling simulations in near real-time, even when the DSS requires thousands of simulations to identify optimized scenarios. Nevertheless, the framework should include interfaces to state-of-the-art discharge and water quality models in the future for more robust water availability (WA) estimations (Candido et al., 2022), a feature that is supported by the modular architecture of Outorga-Web.

CONCLUSION

In this study, we present an administrative decision support system (DSS) for water resource allocation, specifically designed for the concession of water permits in Brazil. Compared to existing computational systems currently in use by water agencies, Outorga-Web introduces novel functionality that, in our view, better aligns with the core principles of Brazilian water policies. Specifically, the system offers:

1. The integration of a water availability model with a DSS to assist technical staff in managing water permits in situations involving conflicting water demands.

2. An interactive web-based tool that provides auditable decision support, enabling the balancing of multiple conflicting objectives and enhancing transparency, particularly for non-technical stakeholders.

3. A modular DSS architecture with a database interface, designed to accommodate new decision-making criteria and interface with advanced models for improved water availability estimates.

ACKNOWLEDGEMENTS

Both authors have received research support from the Interdisciplinary Center for Environmental Sanitation Studies (“Núcleo Interdisciplinar de Estudos em Saneamento Ambiental – NIESA”) of the Federal University of Mato Grosso (UFMT).

REFERENCES

• Agência Nacional de Águas e Saneamento Básico – ANA. (2017). Hidroweb. Brasília. Retrieved in 2024, November 30, from https://www.snirh.gov.br/hidroweb
  » https://www.snirh.gov.br/hidroweb
• Andrade, E. P., Araújo Nunes, A. B., Alves, K. F., Ugaya, C. M. L., Alencar, M. C., Lima Santos, T., Silva Barros, V., Pastor, A. V., & Figueirêdo, M. C. B. (2020). Water scarcity in Brazil: Part 1 - Regionalization of the AWARE model characterization factors. The International Journal of Life Cycle Assessment, 25(12), 2342-2358. http://doi.org/10.1007/s11367-019-01643-5
  » http://doi.org/10.1007/s11367-019-01643-5
• Bellman, R. E., & Zadeh, L. A. (1970). Decision-making in a fuzzy environment. Management Science, 17(4), 41-64. http://doi.org/10.1287/mnsc.17.4.B141
  » http://doi.org/10.1287/mnsc.17.4.B141
• Calizaya, A., Meixner, O., Bengtsson, L., & Berndtsson, R. (2010). Multi-criteria decision analysis (MCDA) for Integrated Water Resources Management (IWRM) in the Lake Poopo Basin, Bolivia. Water Resources Management, 24, 2267-2289.
• Candido, L. A., Coêlho, G. A. G., de Moraes, M. M. G. A., & Florêncio, L. (2022). Review of decision support systems and allocation models for integrated water resources management focusing on joint water quantity-quality. Journal of Water Resources Planning and Management, 148, 03121001.
• Cánepa, E. M., Pereira, J. S., & Lanna, A. E. L. (2017). Water and Economy. In C. M. Bicudo, J. G. Tundisi, & M. C. B. Scheuenstuhl (Eds.), Waters of Brazil (pp. 25-37). New York: Springer. http://doi.org/10.1007/978-3-319-41372-3_3
  » http://doi.org/10.1007/978-3-319-41372-3_3
• Chiquito Gesualdo, G., Sone, J. S., Galvao, C. D. O., Martins, E. S., Montenegro, S. M. G. L., Tomasella, J., & Mendiondo, E. M. (2021). Unveiling water security in Brazil: current challenges and future perspectives. Hydrological Sciences Journal, 66, 759-768.
• Da Hora, A. F. (2001). Metodologia para a outorga do uso de recursos hídricos com ênfase no caso de aproveitamentos hidrelétricos (PhD thesis). Universidade Federal do Rio de Janeiro, Rio de Janeiro.
• Davijani, M. H., Banihabib, M. E., Anvar, A. N., & Hashemi, S. R. (2016). Multi-objective optimization model for the allocation of water resources in arid regions based on the maximization of socioeconomic efficiency. Water Resources Management, 30, 927-946.
• Getirana, A., Libonati, R., & Cataldi, M. (2021). Brazil is in water crisis - it needs a drought plan. Nature, 600(7888), 218-220. PMid:34880440. http://doi.org/10.1038/d41586-021-03625-w
  » http://doi.org/10.1038/d41586-021-03625-w
• Golfam, P., Ashofteh, P. S., Rajaee, T., & Chu, X. (2019). Prioritization of water allocation for adaptation to climate change using multi-criteria decision making (MCDM). Water Resources Management, 33, 3401-3416.
• Gorgoglione, A., Crisci, M., Kayser, R. H., Chreties, C., & Collischonn, W. (2019). A new scenario-based framework for conflict resolution in water allocation in transboundary watersheds. Water, 11, 1174.
• Gough, J. D., & Ward, J. C. (1996). Environmental decision making and lake management. Journal of Environmental Management, 48(1), 1-15. http://doi.org/10.1006/jema.1996.0063
  » http://doi.org/10.1006/jema.1996.0063
• Hajkowicz, S., & Higgins, A. (2008). A comparison of multiple criteria analysis techniques for water resource management. European Journal of Operational Research, 184, 255-265.
• Kepaptsoglou, K., Karlaftis, M. G., & Gkountis, J. (2013). A fuzzy AHP model for assessing the condition of metro stations. KSCE Journal of Civil Engineering, 17, 1109-1116.
• Kinerson, R. S., Kittle, J. L., & Duda, P. B. (2009). BASINS: better assessment science integrating point and nonpoint sources. In A. Marcomini, G. W. Suter II, & A. Critto (Eds.), Decision support systems for risk-based management of contaminated sites (pp. 1-24). New York: Springer.
• Ma, X., Wang, H., Yu, L., Li, Y., Fan, Y., Zhang, J., & Zhang, J. (2020). Multi-preference based interval fuzzy-credibility optimization for planning the management of multiple water resources with multiple water-receiving cities under uncertainty. Journal of Hydrology, 591, 125259.
• Marques, G. F. (2022). Water allocation in Brazil: main strategies, learning and challenges. In J. Rouillard, C. Babbitt, E. Challies, & J. D. Rinaudo (Eds.), Water resources allocation and agriculture: transitioning from open to regulated access (pp. 159-171). London: IWA Publishing. http://doi.org/10.2166/9781789062786_0159
  » http://doi.org/10.2166/9781789062786_0159
• Miranda, G. M., & Reynard, E. (2020). Integrated water resources management in federations: the examples of Brazil and Switzerland. Water, 12(7), 1914. http://doi.org/10.3390/w12071914
  » http://doi.org/10.3390/w12071914
• Miraz, M. H., Ali, M., & Excell, P. S. (2021). Adaptive user interfaces and universal usability through plasticity of user interface design. Computer Science Review, 40, 100363. http://doi.org/10.1016/j.cosrev.2021.100363
  » http://doi.org/10.1016/j.cosrev.2021.100363
• O’Hagan, M. (1988). Aggregating template or rule antecedents in real-time expert systems with fuzzy set logic. In Proceedings of the Twenty-Second Asilomar Conference on Signals, Systems and Computers (pp. 681-689). New Jersey: IEEE. http://doi.org/10.1109/ACSSC.1988.754637.
• O’Hagan, N. K., & O’Hagan, M. (1993). Decision-making with a fuzzy logic inference engine. In Proceedings of the SPIE 2061, Applications of Fuzzy Logic Technology (pp. 320-330). Washington, D.C.: SPIE. http://doi.org/10.1117/12.165038.
• Oliveira, C. U. R. D. (2009). Sistema de suporte a decisão baseado em lógica fuzzy para outorga de recursos hídricos (Master thesis). Universidade Federal de Mato Grosso, Cuiabá.
• Paquin, M., & Cosgrove, C. (2016). The United Nations world water development report 2016: water and jobs. Retrieved in 2024, November 30, from https://unesdoc.unesco.org/ark:/48223/pf0000243938
  » https://unesdoc.unesco.org/ark:/48223/pf0000243938
• Saaty, T. L. (2008). Decision making with the analytic hierarchy process. International Journal of Services Sciences, 1(1), 83-98. http://doi.org/10.1504/IJSSCI.2008.017590
  » http://doi.org/10.1504/IJSSCI.2008.017590
• Salinas, J. L., Laaha, G., Rogger, M., Parajka, J., Viglione, A., Sivapalan, M., & Blöschl, G. (2013). Comparative assessment of predictions in ungauged basins. Part 2: flood and low flow studies. Hydrology and Earth System Sciences, 17(7), 2637-2652. http://doi.org/10.5194/hess-17-2637-2013
  » http://doi.org/10.5194/hess-17-2637-2013
• Universidade Federal de Mato Grosso – UFMT. Núcleo Interdisciplinar de Estudos em Saneamento Ambiental – NIESA. (2024). Monitoramento de vazão e qualidade de água na bacia hidrográfica do Rio Cuiabá, MT. Cuiabá.
• Wiek, A., & Larson, K. L. (2012). Water, people, and sustainability - a systems framework for analyzing and assessing water governance regimes. Water Resources Management, 26(11), 3153-3171. http://doi.org/10.1007/s11269-012-0065-6
  » http://doi.org/10.1007/s11269-012-0065-6
• World Wide Fund for Nature – WWF. (2009). Monitoramento das alterações da cobertura vegetal e uso do Solo na Bacia do Alto Paraguai - porção brasileira - período de análise: 2002 a 2008. Brasília. Retrieved in 2018, April 11, from https://www.wwf.org.br/?25181/Monitoramento-das-alteracoes-da-cobertura-vegetal-e-uso-do-solo-na-Bacia-do-Alto-Paraguai
  » https://www.wwf.org.br/?25181/Monitoramento-das-alteracoes-da-cobertura-vegetal-e-uso-do-solo-na-Bacia-do-Alto-Paraguai
• Yager, R. R. (1977). Multiple objective decision-making using fuzzy sets. International Journal of Man-Machine Studies, 9(4), 375-382. http://doi.org/10.1016/S0020-7373(77)80008-4
  » http://doi.org/10.1016/S0020-7373(77)80008-4
• Yan, D., Ludwig, F., Huang, H. Q., & Werners, S. E. (2017). Many-objective robust decision making for water allocation under climate change. The Science of the Total Environment, 607, 294-303. PMid:28692899. http://doi.org/10.1016/j.scitotenv.2017.06.265
  » http://doi.org/10.1016/j.scitotenv.2017.06.265
• Ye, Z., Maidment, D. R., & McKinney, D. C. (1996). Map-based surface and subsurface flow simulation models: an object oriented and GIS approach. Austin: University of Texas. Retrieved in 2024, November 30, from https://repositories.lib.utexas.edu/server/api/core/bitstreams/6753eec1-fb06-4a33-8d09-54163f239072/content
  » https://repositories.lib.utexas.edu/server/api/core/bitstreams/6753eec1-fb06-4a33-8d09-54163f239072/content
• Zeilhofer, P., & Lima, R. M. (2006). Regio: uma aplicação SIG para subsidiar a regionalização de vazões. Caminhos de Geografia, 7(17), 54-61. http://doi.org/10.14393/RCG71715271
  » http://doi.org/10.14393/RCG71715271
• Zeilhofer, P., Lima, E. B. N. R., & Lima, G. A. R. (2010). Land use effects on water quality in the urban agglomeration of Cuiabá and Várzea Grande, Mato Grosso State, central Brazil. Urban Water Journal, 7(3), 173-186. http://doi.org/10.1080/1573062X.2010.484496
  » http://doi.org/10.1080/1573062X.2010.484496
• Zhai, Z., Martínez, J. F., Beltran, V., & Martínez, N. L. (2020). Decision support systems for agriculture 4.0: survey and challenges. Computers and Electronics in Agriculture, 170, 105256. http://doi.org/10.1016/j.compag.2020.105256
  » http://doi.org/10.1016/j.compag.2020.105256
• Zhao, J., Cai, X., & Wang, Z. (2013). Comparing administered and market-based water allocation systems through a consistent agent-based modeling framework. Journal of Environmental Management, 123, 120-130. PMid:23597927. http://doi.org/10.1016/j.jenvman.2013.03.005
  » http://doi.org/10.1016/j.jenvman.2013.03.005
• Zolghadr-Asli, B., Bozorg-Haddad, O., Enayati, M., & Chu, X. (2021). A review of 20-year applications of multi-attribute decision-making in environmental and water resources planning and management. Environment, Development and Sustainability, 23(10), 14379-14404. http://doi.org/10.1007/s10668-021-01278-3
  » http://doi.org/10.1007/s10668-021-01278-3

Edited by

• Editor-in-Chief:
  Adilson Pinheiro

• Associated Editor:
  Iran Eduardo Lima Neto

Publication Dates

• Publication in this collection
  23 May 2025
• Date of issue
  2025

History

• Received
  30 Nov 2024
• Reviewed
  10 Feb 2025
• Accepted
  19 Feb 2025