A multi-criteria multidimensional model for optimal selection of rural water supply systems

One of the top priorities considered by local managers in different regions is to provide residents with access to adequate water. Surface waters (rivers, streams, and lakes) have been the most accessible source of domestic water in most urban and rural communities in developing countries (Edokpayi et al. 2017). The combination of rapid population growth, inappropriate spatial population distribution, and mismanagement, coupled with a thirst for development, particularly in developing countries, has led to increased pressure on inland water resources, exacerbating the water crisis (Madani 2014; Malistani et al. 2022; Nijhawan et al. 2022). Researchers argue that water demand management can help address this crisis (Muller 2019), but economic, social, cultural, administrative, and technological challenges make the water supply debate critical (McDonnell 2008). A significant challenge facing human societies in the 21st century is the decreasing quantity and quality of drinking water resources (Liu et al. 2020b).

Water resources in most regions are facing great challenges due to increasing demand, decreasing water supply, and catastrophic distribution. The sustainability of water resource systems is influenced by a variety of socioeconomic and environmental factors, so the interplay of different parameters should be studied when evaluating water supply alternatives. Many parameters and their relationships in water supply and management systems are associated with high uncertainties and risks due to the lack of information and the complexity of water systems (Ghandehari et al. 2020).

Severe water scarcity affected around 4 billion people worldwide for at least one month in 2020 (Warner & Diaz 2020). Meeting the growing demand for agriculture, drinking water, and industry is increasingly challenging, and it is expected that water demand will rise by 20–30% over the next three decades, particularly in the Middle East, North Africa, and the countries around the Persian Gulf, which consume more water per capita than the world average (Waha et al. 2017; Boretti & Rosa 2019). Rural settlements often pose a challenge to expanding existing water services or creating new ones due to their remote locations. Thus, policymakers must focus on these areas to achieve Sustainable Development Goal 6, which aims to provide safe drinking water to everyone by 2030 (Machado et al. 2019; Hegarty et al. 2021).

In recent years, there has been an increased interest in designing and implementing rural water supply (RWS) projects to make rural areas more efficient and sustainable. However, alternative water supply options in rural areas are highly complex due to their special economic, social, and technological challenges, making their future uncertain. Failure to consider expert opinions may increase the risk of project failure during ongoing operations.

Although rural areas are crucial to achieving sustainable development in the MENA region, they are facing significant water-related challenges, according to several studies (Bekturganov et al. 2016; Fragaszy et al. 2020). In many developing countries' rural areas, the scarcity of water resources does not seem to be the primary problem. Instead, they lack the appropriate technology to provide, purify, and distribute water based on local and regional capacities, fail to adequately utilize available national or international financial resources, and do not meet these requirements (Nouri-Sepehr 2007).

As a member of the MENA region, Iran has witnessed a significant decrease of over 63% in its rural population in the past century due to the migration of villagers to urban areas in search of better prospects. One of the primary reasons for this migration is the lack of safe water resources in rural areas, which has led to severe water scarcity and forced people to move to urban centers in search of better prospects (Jafari et al. 2016). The lack of sustainable water supply systems is one of the reasons for the declining share of the rural population in the total population (Amini & Hesami 2017).

The multi-criteria decision-making (MCDM) technique is considered one of the most suitable approaches for addressing multi-objective (or multi-criteria) water allocation issues (Gebre et al. 2021). A variety of multi-criteria computer-aided decision-making tools are available to support decision-makers. Many MCDM methods have been used from different perspectives by researchers in water resource management (Minatour et al. 2014; Arasteh & Farjami 2021; Noori et al. 2021; Zamanigandomani et al. 2021). For example, Ebad Ardestani et al. (2020) presented an MCDM framework for evaluating agricultural water supply systems, with a focus on promoting cooperation among all stakeholders in the system.

Also, Li et al. (2022) introduced a method for fuzzy mathematical programming that can be used to reconfigure water allocation. Their approach mitigates the risks associated with irrational water allocation, reduces system risks, and alleviates decision-making risks and uncertainties that are often associated with the allocation of water resources. Mohapatra et al. (2023) developed an analytical model using trapezium fuzzy numbers for decision-making problems in an Internet of Things-based water distribution network. The model was integrated with an MCDM-based location routing protocol to select an optimal route between the utility center and the consumer, taking into account multiple criteria. Recently, Mosetlhe et al. (2023) used TOPSIS and VIKOR methods for selecting the optimal microgrid configuration for rural water pumping applications. However, there is little research on the use of these methods to select optimal alternatives for rural water supply, or simply RWS, systems. Sikder & Salehin (2015) used the MCDM methodology to evaluate suitable alternative water supplies for RWS systems in coastal areas of Bangladesh.

The main aim of this study is to explore the potential of using MCDM methods to identify the optimal water supply methods for rural areas in Iran. Several widely used MCDM techniques, including AHP, Fuzzy-AHP, and TOPSIS, are employed to identify the most crucial criteria and sub-criteria for selecting the most suitable alternatives for RWS. The use of both AHP and Fuzzy-AHP methods can provide a cross-check on the results obtained from each method, ensuring their consistency and identifying any inconsistencies or errors in the decision-making process. Additionally, using both methods can help to capture both the objective and subjective aspects of the decision-making process. Ultimately, this study aims to offer valuable insights to policymakers, water resource managers, and other stakeholders involved in RWS planning and decision-making in Iran.

This study examines the importance and prioritization of effective parameters when implementing RWS projects in areas with less than 10,000 populations. In the first step, all possible alternatives for RWS were explored. Five alternatives have been proposed for water supply in rural areas, based on the opinions of national managers and technical experts in the field of hydraulic engineering. These alternatives include temporary storage dams and reservoirs, diversion of water from rivers, groundwater extraction (springs and wells) and pipe transport, groundwater extraction (springs and wells) and tanker transport, and branching off the pipeline in nearby towns or villages.

In the second step, after discussing expert perspectives, four criteria including environmental, social-security, economic, and technical-management, were chosen for their comprehensive coverage of the key aspects that should be considered when evaluating RWS projects. Firstly, the environmental criterion evaluates the impact of RWS projects on the natural environment, including water sources, biodiversity, and the ecological balance. It is important to ensure that RWS projects do not harm the environment and have a sustainable impact in the long run. Secondly, the social-security criterion assesses the social impacts of the RWS projects. These impacts are critical to ensuring that RWS projects serve the needs of the local communities without adding extra conflicts. Thirdly, the economic criterion evaluates the financial sustainability of the RWS projects, including their cost-effectiveness and financial viability. Finally, the technical-management criterion assesses the technical feasibility of the RWS projects and the effectiveness of their management systems. This includes factors such as the reliability of the water supply, the maintenance of the infrastructure, and the capacity of the local institutions to manage the RWS projects effectively.

The sub-criteria selected for each criterion were chosen after a series of fine-tuning, based on their relevance to the specific context of the study, which aimed to find the optimal RWS systems. The environment criteria included sub-criteria such as adverse environmental impacts on humans and ecosystems other than humans, as well as harmony with climate. These were chosen because RWS systems can have significant impacts on the environment, and it is essential to consider these when selecting the optimal system. The economic criteria included sub-criteria such as project implementation costs, ease of project financing, and the possibility of private sector financial participation. These were included because they can significantly affect the feasibility and sustainability of the project, as well as the potential for private sector investment. The technical-management criteria included sub-criteria such as ease of supplying manpower, machinery and equipment, and materials, as well as ease of execution and project management. These were relevant because they can impact the efficiency and effectiveness of project implementation, as well as the overall cost and quality of the water supply system. Finally, the social-security criteria included sub-criteria such as the possibility of identification and risk-taking by subversive agents, resolving conflicts with beneficiaries and stakeholders, and affectability by natural disasters. These were included because they can impact the social and security aspects of the project, as well as the potential for community acceptance and participation.

In the third step, questionnaires were developed according to the above criteria and sub-criteria and made available to relevant project experts and managers. They were asked to use numbers between 1 and 9 to demonstrate the importance of criteria, sub-criteria, and alternatives from a professional perspective. Numbers 1–9 are used for pairwise comparisons of the criteria, sub-criteria, and alternatives according to Saaty (1989). Expert Choice software was used to determine weighting factors for criteria, sub-criteria, and RWS alternatives. Finally, a ranking of criteria, sub-criteria, and alternatives was performed using relationships governing this method.

Iran is one of the largest and most populous countries in the MENA region. Therefore, studying Iran can provide insights into the challenges and opportunities of implementing an optimal system in a complex and diverse context. Also, Iran has a long history of energy production and consumption, and the country is currently facing significant challenges in balancing its energy needs with environmental concerns. This makes Iran an interesting case study for exploring how an optimal system can help address such challenges. More than 22 million people live in approximately 36,000 rural areas in Iran. Approximately 92% of this rural population has access to safe drinking water, of which 67% is supplied by water supply networks and 25% by water tankers (Maleksaeidi et al. 2018; Ansari et al. 2019).

Iran's rural population is unevenly distributed, with half the population living in the hills, 27% in the mountains, 13% in the plains, and 10% along the riverbanks. This kind of population distribution indicates that water supply management in Iranian villages needs to be seriously formulated with necessary and appropriate mechanisms according to the country's conditions (Imandel 1999; Nouri-Sepehr 2007). The water provided to the rural areas is utilized for different usages in addition to drinking. A study found that garden size, greenhouse size, and number of garden waterings per month were related to water use in rural households in the Ramjad region of Fars province in southern Iran (Keshavarzi et al. 2006).

As shown in Figure 1, AHP, Fuzzy-AHP, and AHP-TOPSIS methods were used in this study. Each of these three calculation methods involves four basic steps (corresponding to the steps presented in Figure 1) leading to the determination of the weights of the criteria and sub-criteria. These steps are briefly described, with additional explanations and associated formulas and calculations. Buckley's method was utilized to determine the weight coefficient values (Buckley 1985). In the fifth step, the results of the fourth step are sorted and the ranking of each sub-criterion is determined. Finally, depending on various factors such as data fuzzification, the results are analyzed and evaluated.

This study utilized questionnaires, which are a commonly employed method for data collection. A team of national experts including university professors and governmental decision-makers was established to conduct the investigation, and since there was no pre-existing questionnaire available to evaluate the most appropriate RWS system for Iran, a panel of experts was assembled to generate questions based on the four mentioned criteria. Questionnaires were used to establish the priorities for water supply projects. Following this, a group of experienced hydraulic engineers evaluated the content of the questionnaire for both qualitative and quantitative validity. To execute the weighting process, two distinct questionnaires were employed. The first type of questionnaire was used to conduct pairwise comparisons of criteria, while the second type was dedicated to ranking the available options for each criterion. Both questionnaires are available as supplementary documents, and they employed the Saaty (1999) scale as a basis for their scales.

The analytical hierarchy process (AHP) is a structure for solving problems that need to be solved analytically but have a layered form (Saaty 1980, 1999). The AHP method recognizes that although there are many criteria or choices, the magnitude or importance of each criterion may not be the same as the others. The AHP method is a technique that must pay attention to both weights and objective function optimization. It is a powerful tool for pairwise comparison of problem alternatives and criteria based on the opinion of relevant experts, their weighting, and scoring (Ghorbani & Hamidifar 2022; Ghorbani et al. 2022b). Checks and balances thus ensure that reasonable solutions are found when the relative importance of the criteria in the process is determined by weighting. That is why AHP has become one of the most popular decision-making techniques in various sciences today.

To obtain fuzzy weight, the mean values of the fuzzy geometry must be multiplied by the sum of the reversed matrix's values. Finally, the weighted values of the problems criteria, or the arithmetic average of the new upper, middle, and lower limits, should be normalized and then ranked.

The results of Equation (7) need to be ranked. Any value that is close to 1 denotes the higher rank of the criterion, sub-criteria, or RWS option.

The AHP method is used to present weighted criteria and sub-criteria as well as inconsistency coefficients and ranking of the problem's sub-criteria in Table 1. In this table, it can be seen that the inconsistency coefficients of the sub-criteria are all less than 0.1. Inconsistency coefficients are a measure of the degree of agreement or disagreement among responses to different parts of the questionnaire. A low inconsistency coefficient (less than 0.1) indicates that there is a high degree of agreement among respondents, and therefore greater confidence in the validity of the questionnaire (Saaty 1977). These results suggest that the completed questionnaire is a robust and effective tool for measuring the relevant criteria and sub-criteria. Furthermore, based on the AHP method, the environmental and economic criteria are ranked higher than other criteria, while the technical-management and social-security criteria are ranked in the last ranks. Among the sub-criteria, the most important ones are the adverse environmental effects on humans, the costs of implementing the project, and the ability to identify and take risks by subversive agents. Meanwhile, ease of implementation, project management, and project financing are in the last rank when choosing the type of RWS project. Similarly, the weights are calculated in a fuzzy environment and the ranking criteria and sub-criteria are then performed based on the relationships governing the Fuzzy-AHP method mentioned in the methodology section. Table 2 shows the final results, including the weighting coefficient, the ranking of the problem's criteria and sub-criteria, and the overall ranking of the sub-criteria using the Fuzzy-AHP method.

It can be seen that in the Fuzzy-AHP method, although the prioritization of criteria is similar to AHP, the prioritization of sub-criteria is quite different. Therefore, the highest priority is given to the sub-criteria of the negative environmental impact on people, the possibility of identifying harmful substances and the possibility of risk-taking, as well as the costs of project implementation, but the sub-criteria ease of implementation, exposure to natural disasters and project management were less important when choosing the type of RWS project. The initial data was then subjected to a third method, called the AHP-TOPSIS method, to compare the previous results. AHP-TOPSIS has more precise functions and it seems that with the mathematical logic behind it, the prediction of the results is more tangible. The calculations for this method were performed based on the methods section and the results are summarized in Tables 3 and 4.

Table 3 shows all the steps of the AHP-TOPSIS procedure, including the formation of the primary matrices, the normalized matrices, the normalized weight matrices, and finally the ranking of the problem criteria. Table 4 shows the final results, which include the weighting factors and rankings of the problem criteria and sub-criteria as well as the overall ranking of the sub-criteria of the AHP-TOPSIS method. Based on the results in the table, it can be seen that environmental and economic criteria are still more important according to the TOPSIS method. However, the social-security criterion is superior to the technical-management criterion. In addition, the adverse environmental effects on humans, the possibility of identification and risk-taking by subversive agents, and project implementation costs sub-criteria were given the highest priority in choosing the type of water supply project in rural areas. Meanwhile, the project management, negative environmental impacts on ecosystems other than humans, and harmony with the climate criteria are placed among the last priorities in determining the type of water supply project.

Figure 2 shows the results of criteria prioritization in different procedures. Based on this figure, the three methods AHP, Fuzzy-AHP, and AHP-TOPSIS show the great importance of environmental criteria when selecting the type of RWS projects. Subsequently, the economic criterion also receives the second priority, based on the three methods used. Although the AHP and Fuzzy-AHP methods indicate greater importance of the technical-management criterion compared to the social-security criterion, in the AHP-TOPSIS method, the ranks of these two criteria are reversed and the social-security criterion is more important.

Table 5 shows the weight and final ranking of the alternatives chosen in the studied problem using the AHP, Fuzzy-AHP, and AHP-TOPSIS methods. As can be seen from this table, the rankings of the Fuzzy-AHP methods are sorted according to the proposed model, while the rankings of the AHP-TOPSIS methods show a shift in the priority of water supply projects.

The Fuzzy-AHP method suggests that storage dams have the highest priority, followed by diverting water from the river, among the five available alternatives. However, the AHP-TOPSIS method indicates that diverting water from the river is the top priority, with constructing storage dams ranked second. Both methods agree that branching off the pipeline in the nearby town or village is the least favorable option. It should be noted that some other alternatives, for example, underground dams, qanats, and dug wells, have been used in some areas for RWS (Wagner & Lanoix 1959; Hanson & Nilsson 1986; Nasiri & Mafakheri 2015). As these options were not recommended by the expert in initial screening due to the high risk of pollution in shallow groundwater, they are not listed as alternatives for RWS in this paper.

Based on each method's advantages and disadvantages, different results can be obtained when selecting the best option for RWS projects. For instance, AHP's ease of use, adaptability, and ability to investigate inconsistencies are its main advantages. Additionally, users can input simple data by completing the pairwise comparison form. However, a significant number of pairwise comparisons must be filled because using AHP causes the decision problem to be divided into several subsystems. On the other hand, ideal negative and positive solutions are realized in the case of TOPSIS. Additionally, the TOPSIS framework supports the use of various measurement units. However, TOPSIS is conclusive in its original and standard form and does not account for uncertainty in the calculations of final weighting (Jozaghi et al. 2018). A summary of the advantages and disadvantages of different methods is provided in Table 6 based on the literature (Oguztimur 2011; Zarghami & Ehsani 2011; Velasquez & Hester 2013, Liu et al. 2020a, 2020b).

In the present study, it was aimed to prioritize the most important criteria and sub-criteria in the implementation of RWS projects and eliminate other criteria or sub-criteria that have a negligible impact. Reducing the cost and time required to implement an RWS project will boost the project productivity and stakeholder satisfaction. The criteria were selected as appropriate and effective for the purpose of the research and were weighed by experts. The value of each criterion was first determined and then a multi-criteria decision matrix was created. When choosing an RWS system, the AHP was used to rank the importance of various criteria. A questionnaire was prepared and designed at this point with the opinions of experts from consulting engineering firms and local industry contractors who were familiar with the RWS system. Integrated AHP approaches support better resolution of real decision problems when the number of alternatives is limited. The fuzzy framework and TOPSIS techniques were then integrated into the AHP approach to dissect different RWS choices and select the most ideal choice. By comparing the sub-criteria, it was also found that the adverse environmental effects on humans, the possibility of detection and exposure to subversive factors, the cost of project implementation, and the mere provision of human resources are of greater importance. In general, the results indicate the great importance of environmental and economic criteria when selecting the type of RWS projects. Additionally, technical-management and social-security criteria play a secondary role. These findings suggest that underestimating technical-management and social-security requirements may have contributed to the failure of some RWS projects.

On the other hand, the project management, the negative environmental impacts on ecosystems other than humans, and harmony with the climate come last in the choice of the type of RWS project. Furthermore, diverting water from the river and temporary storage dams and reservoirs were found to be the top priorities. Besides, branching off the pipeline in the nearby town or village pipeline is chosen as the last priority among the alternatives. In conclusion, although the focus of this paper was on finding suitable alternatives and key criteria for RWS in Iran, it should be noted that the criteria and alternatives considered are not limited to Iran. Therefore, the methodology used in this study could be useful in solving RWS problems in other countries, especially in the Middle East and North Africa regions. For future research plans, various methodological possibilities can be combined by the proposed model in order to increase the accuracy of the obtained results and reduce the evaluation risk. For example, the partial risk map method (PRISM) proposed by Bognár et al. (2022) and Bognár & Benedek (2022) can be used for risk identification during pairwise comparisons in the AHP method, supporting the analysis of best and worst ideal solutions in the TOPSIS method. Also, the use of machine learning algorithms or artificial intelligence models could enhance the accuracy and precision of the decision-making process. Additionally, further investigation could be done to explore the potential of incorporating stakeholder perspectives and preferences in the decision-making process to increase stakeholder participation and satisfaction. Finally, it could be useful to conduct a longitudinal study to assess the long-term effectiveness of RWS projects implemented based on the prioritized criteria and sub-criteria identified in this study.

All relevant data are available from an online repository or repositories (Hamidifar 2023).

The authors declare there is no conflict.