Sustainability framework for water infrastructure development in Nigeria: a modelling approach

The pursuit of safe and sustainable water supplies is a critical global challenge, affecting billions of people. Over 2.1 billion people worldwide lack access to safely managed drinking water, and 785 million lack basic water services, primarily in rural areas (Bain et al. 2014; UNICEF/WHO 2019). This highlights the urgency of achieving Sustainable Development Goal (SDG) 6.1, which calls for universal and equitable access to safe and affordable drinking water. Globally, the sustainability of water resources remains a pervasive challenge, with innovative methodologies emerging across various regions to enhance the predictability and sustainability of water infrastructure.

For instance, in Taiyuan, China, a region facing severe water shortages, researchers have developed a model using principal component analysis and a backpropagation neural network to predict water demand, outperforming traditional models (Wu et al. 2021). Similarly, in another study, a multifaceted approach to assess water supply sustainability was proposed, employing a multivariate water supply index to combine various sources of supply (Khoshoei et al. 2023). This method provides a new basis for evaluating water sustainability in regions dependent on multiple water sources.

Further refining prediction methodologies, another study from China used an optimised grey and Markov model to improve the accuracy of domestic water consumption predictions, addressing issues like overfitting which plague traditional grey prediction methods (Wang et al. 2021). Meanwhile, Song et al. (2024) introduced a risk analysis for water resource carrying capacity in the central plains region of China, utilising an improved entropy weighting method combined with gray correlation analysis. This comprehensive approach considers economic, social, and ecological environments, offering a methodological guide for sustainable water resource management.

In Africa, particularly Nigeria, the challenge of achieving sustainable water infrastructure is notably critical. Approximately 46% of water infrastructure is non-functional, significantly impeding progress towards SDG 6.1 (Andres et al. 2018; Adeniran et al. 2021). Failures in water infrastructure arise from a complex interplay of factors, including technical limitations, social dynamics, financial constraints, institutional weaknesses, environmental pressures, and political realities (Adeoti et al. 2023). These challenges vary greatly depending on the location; developed countries typically experience minimal failure rates, whereas developing nations like Nigeria face widespread issues (UN Water 2022). Within Nigeria itself, stark disparities exist between urban and rural settings, exemplified by the Isale-Oja community in Ibadan, one of this study's focus locations (Adeoti et al. 2024). This underscores the critical need to assess local conditions influencing water infrastructure sustainability comprehensively.

This research responds to this need by pioneering an innovative methodology to predict infrastructure sustainability in the planning phase of project implementation, utilising the concept of ‘Location Viability Rating’ (LVR), a metric specifically designed to measure the unique circumstances of each community that impact the long-term success of water infrastructure projects.

Predicting the sustainability of water projects requires a precise definition of sustainability. The concept of sustainability, particularly in the context of development, often faces challenges due to its broad and varied interpretations (Dovers & Handmer 1992; Barbosa et al. 2014). This study specifically focus on ‘infrastructure sustainability,’ defined as the ability of infrastructure to support and contribute to long-term economic and environmental well-being (Olanipekun et al. 2014; Song & Wu 2021).

Sustainable infrastructure is conceptualised to be durable, resilient to environmental changes, and capable of enhancing community well-being. It must remain functional throughout its designated lifespan, addressing challenges such as security issues, vandalism, wear and tear, and the impacts of climate change. The integration of innovative technologies for efficiency, alongside strong governance and institutional support, is crucial. Consistent funding is vital to maintain operational infrastructure, providing the necessary financial resources, and human capital (Oyegoke et al. 2012; Khan et al. 2018; Adeoti et al. 2023).

A holistic approach is essential for addressing water infrastructure sustainability. While Adeoti et al. (2023) propose a Water Infrastructure Sustainability Framework encompassing all project stages, a notable gap exists in Nigeria, as evidenced by a systematic literature review by Koppa et al. (2023). Their analysis found only six frameworks related to sustainable infrastructure in Africa, with none specifically targeting water infrastructure, highlighting a significant shortfall in tailored frameworks for regions like Nigeria (Oraegbune & Ugwu 2020).

Rigourous pre-implementation assessments are a hallmark of infrastructure projects in developed nations, with stringent standards exemplified in Europe (Niekerk & Voogd 1999). However, Diaz-Sarachaga et al. (2016) highlight a critical shortcoming when applying these frameworks to the developing countries like Nigeria. Addressing this disparity, our study introduces an innovative framework designed for the real-world conditions of developing countries. This framework, built on a foundation of global insights, local data, and integration of multiple theoretical frameworks, developed the Viability Rating (VR) determined through a Likert scale survey assessing community readiness for sustainable water projects, based on 265 sustainability issues identified by Adeoti et al. (2023). The Sustainability Rating (SR) is another pivotal component, which assesses project sustainability by incorporating financial factors and leveraging VR to enhance prediction accuracy. This methodology enables iterative refinement of SR predictions through a conceptually integrated model that evolves with real-life project data.

In essence, our study represents a practical and strategic framework and tool for planning and executing water projects, significantly advancing the objectives of SDG 6.1 and contributing to the global water sustainability efforts. The innovative methodology used in this study has global significance as it is uniquely designed to predict infrastructure sustainability in the context of Nigeria, which is applicable to similar regions across Africa and the world, enhancing sustainable infrastructure implementation.

This study introduces a novel, transdisciplinary framework for assessing water infrastructure sustainability in Nigeria. The framework integrates insights from ecology, behavioural science, socio-technical studies, and complex systems to address the multifaceted challenges of water infrastructure in Nigeria. As advocated by Robinson (2008) and Wickson et al. (2006), this transdisciplinary approach is inherently problem-focused and collaborative, making it well-suited to tackling the social, cultural, environmental, institutional, and economic aspects of water infrastructure sustainability (Boyer et al. 2015). The study employs iterative and participatory methods, integrating insights from multiple disciplines, stakeholder feedback, and local observations. This approach ensures a balance of academic rigour and practical application, exemplifying how a transdisciplinary method can bridge theoretical knowledge with real-world relevance.

Bronfenbrenner's Ecological Systems Theory (1979), as applied to environmental contexts by Crawford (2020), forms the foundation for our community selection process. This theory recognises that the community well-being is shaped by complex interactions within environmental systems. We employed a standardised tool to identify communities facing severe water scarcity, considering the interplay of ecological, social, and economic factors in these locations. This approach ensures that the chosen communities are not only grappling with immediate water challenges but also serve as exemplars of how ecological systems influence the long-term success of water infrastructure projects.

The questionnaire design is based on the Theory of Planned Behaviour (TPB) (Ajzen 1991), informed by Adeoti et al. (2023)'s systematic review which identified 265 sustainability challenges. Adapting TPB principles, we use a Likert-scale format to assess community attitudes towards sustainable water infrastructure. This approach goes beyond simply predicting behaviour; by understanding community perspectives, we can develop a community VR for sustainable water infrastructure projects.

Complex Adaptive Systems Theory (Chan 2001; Holden 2005) informs the development of our VR and SR. This approach acknowledges the dynamic interplay of sustainability factors within communities. The VR and SR act as indicators for predicting the potential success of sustainable water infrastructure projects. Both theoretical foundations and real-world data from water projects inform the VR and SR. This adaptability ensures that our model remains responsive to the unique circumstances of each community, enabling a comprehensive evaluation of water infrastructure sustainability across diverse contexts.

This study investigates five communities within Oyo State, Nigeria, selected from the 1,696 communities assessed by Adeoti et al. (2024). Data from Adeoti et al. (2024) were thoroughly reviewed to identify the areas with most pressing water challenges. These communities, identified through precise GPS coordinates and enhanced by local insights, were chosen to represent a range of urban and rural water scarcity challenges. While conceptually, three communities would have sufficed to test the predictive capacity of our model, five were ultimately selected to broaden the scope of data collection and enhance the robustness of our sustainability assessments.

• Aponmode Community (Lat: 7.5276, Long: 3.9142) in Moniya, Ibadan, faces suburban water scarcity with residents relying on an unprotected well as their primary water source, reflecting issues with ageing infrastructure and inadequate water supply systems.

• Isale Oja in Ijaye Orile (Lat: 7.6301, Long: 3.8462) is a stark example of rural water poverty, where the community depends on an unhygienic and unprotected well, highlighting the absence of reliable water sources.

• Omilabu Village (Lat: 7.5465, Long: 3.9456) deals with extreme remote area water challenges, with the main water source being an unhygienic stream, indicative of the acute difficulties in accessing safe drinking water.

• Alabata Village (Lat: 7.5874, Long: 3.8688), located along Iseyin Road in Ibadan, primarily relies on rainwater and unprotected wells, illustrating the peripheral urban area's struggle with inconsistent water supply.

• Ajila/Irepodun in Idi Ayunre (Lat: 7.1611, Long: 3.9214) represents semi-rural areas with no improved water sources, depending on rain and an unhygienic stream, which are often overlooked in water infrastructure planning.

To ensure representative data collection across the five chosen communities, we employed Cochran's Formula and considered factors like population, household size, and house counts. It is important to note that interviews were conducted at the household level, where a household represents a family unit residing in a single dwelling. Even though a single house might contain multiple households, interviewing one household provides valuable insights for that entire house.

While the calculated sample size for Aponmode Community was 61 households, we surveyed more residents to strengthen the generalisability of our findings (Vasileiou et al. 2018).

In Alabata Community, the required sample from 500 households was at least 87. Omilabu Community, with 122 households, necessitated a sample of 37 households. For Isale Oja Ijaye, with 422 households, 80 were needed in the sample. Lastly, Ajila/Irepodun Idi Ayunre, with 367 households, required a sample of 56. Totally, 321 households required across all 5 communities.

The comprehensive formula used for determining the sample space in each community is as follows:

Here:

• b = Average number of households per house

• ICC = Intra-cluster correlation coefficient (0.1 assumed)

Detailed sample size calculation for Aponmode Community is provided in the Supplementary Information (SI). The same methodological framework was applied uniformly to the remaining four communities. This ensures transparency in our approach and clarity in our household interview selection process.

The Predictive Iterative Sustainability Model (PISM) is a dynamic framework and practical tool designed to predict the viability and sustainability of water infrastructure within communities. PISM employs a two-step rating process. First, the VR is calculated using a conceptual formula applied to weighted survey responses. This formula generates a nuanced index reflecting each location's potential for sustainable water infrastructure projects. The model then progresses to the SR, which further refines the viability assessment by incorporating financial considerations such as predicted income, operational, and maintenance costs. This ensures the sustainability score aligns with a project's long-term financial feasibility.

PISM is designed to be progressively enhanced through iterative refinements based on real-life project data. While the specifics of these enhancements fall outside the scope of this study, the conceptual formula guiding this iterative process is meticulously detailed in the SI material and subsequent sections of this methodology. This iterative process is crucial for perfecting the model's predictive capabilities, aiming to progressively improve its accuracy until it reaches a point of convergence.

The study utilised a survey using a Likert scale to gather data on community sustainability factors. Respondents rated their answers on a scale of 1 (least desirable) to 5 (most desirable). The detailed questionnaire can be found in Table S1 of the supplementary information (SI). Responses were tallied under respective factors; each was assigned a weight reflecting its sustainability impact. These weighted responses were aggregated to calculate the LVR for each community, providing a quantitative assessment of their sustainability potential.

This iterative refinement process is core to model improvement. Recognising the data requirements for VR assessment, PISM incorporates a method to substitute VR with the average of actual SRs (aSRs). This average becomes the new predictive coefficient, improving SR prediction practicality and accuracy by leveraging real-world project data. Figure 1 illustrates this transition, while the detailed conditions and mathematical proofs supporting it are elaborated in the Supplementary Information (SI). The SI provides a comprehensive breakdown of the PISM methodology, including the nuanced calculations and logical flow summarised in Figure 1.

The iterative process aims for convergence, where the gap between predicted and actual SRs falls within an acceptable error margin, given sufficient real-world project data. This process, from initial prediction to convergence, establishes a foundation for predicting sustainable water infrastructure.

Even in its current form, before achieving full convergence, the PISM demonstrates its effectiveness as a predictive tool for sustainability. The SR predictions for the five communities studied were meticulously informed by extensive surveys, reflecting actual conditions with accuracy. Therefore, the PISM, in its current state, represents a significant and influential framework with the potential to considerably enhance water infrastructure sustainability.

Our study's methodology, leveraging a transdisciplinary approach, integrated insights from both industry experts and academic researchers to ensure the validity of our survey instruments and data collection protocols. We standardised data collection across all study locations to minimise biases, using Likert scale questionnaires to transform subjective qualitative perceptions into consistent, quantitative data across surveyed households.

To handle anomalous or erroneous data, we initiated a rigourous verification process during data entry and subsequent analysis phases. Prior to data collection, all data collectors underwent comprehensive training to ensure uniformity in methods, procedures, questions, and techniques across all sites. This standardisation was crucial in minimising data variability and potential biases. Utilising Likert heatmaps, we meticulously reviewed each questionnaire response to identify potential anomalies, enabling us to make necessary corrections or exclude erroneous data to uphold the integrity and accuracy of our findings.

These methodological enhancements significantly improved the reliability of our results, providing a robust framework for assessing the sustainability of water infrastructure in the studied communities.

The VRs obtained from our Likert-scale questionnaire reflect comprehensive assessments across the five studied communities. The questionnaire, summarised in Table S1 of the Supplementary Material, includes a sample of 10 questions specifically assessing community water poverty, one of the 12 sustainability factors covered by the full set of 70 questions. To enhance data robustness and mitigate non-response bias, we collected responses from 380 households (88 in Aponmode, 96 in Alabata, 45 in Omilabu, 86 in Isale-Oja, and 65 in Ajila), exceeding the methodology's required sample size of 321.

Utilising the VR formula outlined in the Methodology section, we computed the aggregated Community VR following Likert survey analysis. This process transitioned the qualitative community perspectives into a quantifiable VR, encompassing various sustainability factors.

The VRs in Table 1 range from 63.95 to 67.91%, indicating a similar baseline viability for water projects across these communities despite their water needs. This similarity highlights that VR alone cannot predict sustainability outcomes. Instead, the VR serves as a springboard, influencing a model that predicts the SR. The community with the highest VR is expected to have the best predicted SR, suggesting a higher likelihood of successful sustainable infrastructure projects.

Table 1

Community VR score

View Large

By pinpointing communities with the highest VRs, we can prioritise investments in locations where water projects are most likely to be viable and sustainable in the long term. VRs inform investment decisions by highlighting communities with conditions that are favourable for successful and sustainable water projects.

The SR for our studied communities is derived using a formula that balances the predicted income with the predicted expenses, modulated by the VR. The computation process is as follows:

Predicted income is estimated through a combination of average water demand per household (AWDH) and the estimated amount households are willing to pay per liter of water (EAHWP), factoring in the volume of water supply capacity (VWSC) relative to AWDH and the proportion of households willing to pay (PHWP).

These are the total expected costs of operation (⁠⁠), maintenance (⁠⁠), and repair (⁠⁠) combined.

In instances where the water supply capacity per household (VWSC/AWDH) exceeds the total number of households (EHC), the formula adjusts to use EHC in place of VWSC/AWDH.

The formula considers the total water project supply capacity per household in relation to the community's population to account for variations in water demand, income generation, and financial sustainability. This ensures that the predicted SR reflects the specific dynamics of each community.

The explanations for each variable, and the detailed development of the SR prediction formula, are provided in the Supplementary Information (SI) material for further reference. Table 2 presents the determined SR for each of the five studied communites.

Table 2

Predicted water infrastructure SR in the surveyed communities

View Large

Our analysis of Predicted SRs in Table 2 reveals a standout case: Isale-Oja with an SR of 424%. This translates to a predicted financial ratio of income being 4.24 times greater than expenses, indicating a strong potential for Isale-Oja to sustain water infrastructure through healthy income generation relative to projected expenses.

An interesting contrast emerges when comparing VRs and SRs. Despite a lower VR of 63.95% for Ajila community compared to Omilabu 64.6%, Ajila exhibits a higher SR of 211%. This seemingly counterintuitive finding can be explained by water project supply capacity in relation to population density. Omilabu's higher TWPSCH compared to its EHC suggests a lower population density. This translates to lower water demand and potentially lower income from water sales compared to Ajila, despite a slightly higher VR. Therefore, while VR suggests Ajila has a slightly lower baseline suitability for the project, the SR suggests its potential for financial sustainability is stronger due to a more favourable population-to-water supply ratio. Ajila's higher SR highlights the crucial role of population density and water demand in financial sustainability. This case underscores the need to consider community-specific factors beyond just baseline viability for accurate forecasting of water project success.

Isale-Oja's exceptional SR of 424% (predicted income-to-expense ratio of 4.24:1) reinforces the link between economic surplus and sustainable water infrastructure. The variation in SRs across communities with similar VRs further emphasises this point. This highlights the significant influence of economic realities beyond baseline viability on the long-term financial sustainability of water projects.

The PISM enhances the accuracy of SR predictions by leveraging the VR as a starting point to account for various sustainability factors within the model. This strategic use of predicted SRs in the pre-construction phase empowers decision-makers to identify communities that not only require but are also economically prepared for sustainable water infrastructure investments. This methodology is crucial in addressing the persistent challenges in Nigeria's water sector, where over $3 billion has been lost to failed projects, hindering water access for millions of Nigerians (Otun et al. 2011; Andres et al. 2018; Adeniran et al. 2021). The strength of the PISM lies in its iterative design, incorporating real-world data to refine projections and enhance decision-making. This process not only supports the targeted allocation of resources but also aligns with global sustainability goals, ensuring the long-term success and effectiveness of water infrastructure developments.

The iterative refinement of the PISM is designed to ensure that the predicted SR converges with the actual SR within an acceptable margin of error, represented by ⁠. This convergence criterion, a quantitative measure, is crucial for model validation using real-life water project data for calibration. Achieving convergence suggests the model's predictive capability is reliable enough to forecast the success of water infrastructure projects before construction begins.

The model's core design hinges on integrating empirical data into its prediction mechanism. Through iterative recalibration driven by the ratio of actual to predicted SRs: , the model is continuously refined. This adjustment process ensures the model remains aligned with real-world outcomes, thereby enhancing the precision of its forecasts.

In this study, five communities utilised VR to comprehensively assess critical factors for sustainable water infrastructure projects. VR effectively facilitates the prediction of a project's SR during the planning and pre-construction phase, aiding in the selection of viable locations. However, VR assessments can be resource-intensive, prompting the need for a more efficient approach. We propose replacing VR with the average SR of completed projects. This substitution leverages real-world data, potentially improving model accuracy while simplifying its application in water infrastructure planning.

The methodology section details an alternative method, iteratively refined for convergence, to enhance PISM's usability for water agencies. This method replaces VR with the average SR of completed projects. Since VR acts as a multiplier in SR prediction, this substitution could effectively adjust predictions and improve PISM's utility for strategic water infrastructure decisions. Convergence between predicted and actual SRs strengthens PISM's predictive power and underscores its value in water infrastructure execution. Continuous refinement and integration of real-world data will further enhance its forecasting capabilities, solidifying PISM as a crucial tool for achieving sustainable water infrastructure development.

Refining the accuracy of PISM in Nigeria and similar contexts requires real-world implementation. This, in turn, necessitates post-construction data collection to assess infrastructure performance, a process that demands substantial financial backing. Securing this funding hinges on collaboration with government agencies, philanthropies, and international organisations. The primary challenge lies in overcoming potential financial constraints and obtaining long-term commitment from these key partners.

This study introduces the PISM, a novel framework designed to address critical gaps in sustainable water infrastructure development for Nigeria. PISM integrates VR and SR into a cyclical evaluation process, offering a valuable tool for assessing project sustainability during the early planning stages. Field data collected through surveys conducted across five communities established VR scores ranging from 63.95 to 67.91%, creating a robust benchmark for evaluating community readiness.

The wide range of SR scores (179–424%) highlights PISM's adaptability to diverse economic and operational realities, identifying projects with high sustainability potential. However, PISM's current strength lies in its predictive capabilities, which require real-world application for validation. Future research should focus on implementing PISM in actual projects to verify its predictions and refine its methods based on observed outcomes. Streamlining data collection and expanding its applicability to various contexts are crucial for enhancing PISM's practicality and scalability.

This study highlights the urgent need for collaboration among policymakers, institutions, and philanthropic organisations. By working together, they can effectively adopt and adapt innovative frameworks like PISM. Such collaboration is critical to addressing water poverty and infrastructure failures in Nigeria and other developing regions. Refining and implementing PISM in real-world projects are essential steps towards achieving SDG 6.1: ensuring universal access to safe and affordable drinking water.

The PISM's potential for global application offers a bright future for sustainable water infrastructure in developing regions. This research bridges the gap between academic discourse and practical solutions, empowering policymakers and practitioners in sustainable development. Integrating PISM into water projects and continuously refining it can turn the vision of universal access to safe water into reality.

The support of Fairaction International Pty Ltd, The University of Technology Sydney, Australian Government Research Training Program Scholarship, Clean Water for All Pty Ltd and Essential Need Projects Ltd is gratefully acknowledged.

This study was conducted in strict adherence to the ethical guidelines and standards set forth by the University of Technology Sydney, with ethics approval granted under UTS HREC REF NO. ETH23-7980. This approval underscores our commitment to responsible research practices, particularly when collecting data involving human participants.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.