1 Introduction

Oil spills are a major global environmental problem that have a negative impact on ecosystems and public health, especially in areas where transportation and hydrocarbon extraction are common. While the effects of oil spills on bodies of surface water, such rivers and oceans, have received a significant attention, there is still much to learn about the potential for crude oil to contaminate groundwater resources.

Numerous studies have emphasized the risks that oil and hydrocarbon spills pose to the quality of groundwater. The possibility of contamination from unconventional gas extraction is highlighted (Jackson, 2014; Jackson et al., 2013 and Lefebvre, 2017), mainly signposting the hazards of surface pollutant releases and well integrity failures. Raynauld et al. (2016) delve more into the mechanisms of pollutant release and how they affect groundwater. A more comprehensive viewpoint is offered by Kurwadkar (2017) and Kurwadkar et al. (2020), who addresses the effects of both organic and inorganic contaminants on groundwater quality as well as the necessity of taking preventative action to avoid contamination. When taken as a whole, these findings highlight the critical need for more investigation and strict laws to safeguard groundwater supplies in areas with high levels of oil exploration and production.

The Niger Delta region (NDR), renowned for its substantial oil reserves, serves as a vivid example of the complexities inherent in resource-rich regions. Over the years, this area has borne witness to extensive hydrocarbon contamination as a consequence of persistent oil exploration, transportation accidents, and unintentional spillages (Akinwumiju et al., 2020; Giadom and Tse, 2014; Sam et al., 2017b; Umar et al., 2019). Set within the delta unconsolidated alluvium is a bountiful aquifer system, a critical lifeline for the region’s inhabitants. A majority of the local population depends on these groundwater reserves, primarily sourced through shallow boreholes and dug wells (UNEP, 2011). However, analyses reveal elevated concentrations of hydrocarbons and heavy metals in these water sources, directly linking them to nearby oil spillages (Ozoemenam et al., 2018). Oil leaks can originate from various facilities and pipelines, including aging, deteriorating, and abandoned infrastructure. Additionally, accidents during the transportation and artisanal processing of stolen oil under harsh conditions can lead to spills of varying magnitudes (Lindén and Pålsson, 2013; Sam et al., 2017b). While the immediate and recognizable health hazards of inhaling hydrocarbon vapours are acknowledged, there exists another concealed peril, of associated heavy metals. Groundwater, serving as the primary drinking water source in the Niger Delta, is suspected to be a significant vector for hydrocarbon ingestion. The dissolved hydrocarbons present within groundwater pose an imperceptible threat, as they neither alter taste nor emit any odour (Johannsen et al., 1995), yet, their carcinogenic attributes pose a significant health risk, potentially exposing communities to harmful substances with every sip (Abdel-Shafy and Mansour, 2016; Barbosa Jr et al., 2023; Shukla et al., 2022). The need for a comprehensive risk assessment is underscored by the immediate and long-term health risks posed by groundwater contamination. Nevertheless, progress toward effective risk assessment in the Niger Delta has been hampered by a persistent lack of comprehensive data. In cases of limited information, the source-pathway-receptor (S-P-R) model stands out as a promising tool (EA., 1999, 2000a, 2000b, 2003; Smith et al., 2005). This model, at its foundation, seeks to elucidate the relationships between a contamination’s origin (the source), its transmission medium (the pathway), and those at risk (the receptors). Such a model offers a panoramic view of contamination risks (Belle et al., 2023; Mbae et al., 2023; Sun et al., 2023) and incorporating geospatial analysis to map the distribution and extent of contamination (Umar et al., 2019). In the context of the Niger Delta, the sources are predominantly hydrocarbon spills, pathways include soil and groundwater systems through which contaminants migrate, and receptors are the ecological and human systems that are adversely affected (Giadom and Tse, 2014).

In recent years, there has been notable progress in the quantitative aspects of hydrological studies in West Africa as a whole and within the Niger Delta region. This includes, for example, studies by Getirana et al. (2021), who developed a fully coupled land surface and flood model to improve the simulation of wetland dynamics in the Inner Niger Delta, and Ndehedehe et al. (2016), who used GRACE satellite data to analyse changes in terrestrial water storage across West Africa. These advancements have been fueled by improved data collection methods, enhanced modeling techniques, and a growing understanding of the complex hydrological dynamics in the area. However, despite these studies, the issue of groundwater contamination remains a pressing concern in the region. A variety of techniques, each with a specific focus and application, have been employed to evaluate the risk of groundwater contamination. Burkart et al. (1999) used statistical, overlay and index, and process-based modeling techniques to evaluate vulnerability to agrichemical contamination in the Midwest US, while Zhao et al. (2018) Zhao developed a straightforward risk index method that integrates aquifer vulnerability, pollution sources, and groundwater function zoning. In order to evaluate the danger of pollution in the Beijing Plain, China, Wang et al. (2012) combined groundwater value, a modified DRASTIC model, and hazard quantification. To assess groundwater quality and identify the sources of contamination in the Lalin River basin in Northeastern China, Chen et al. (2019) used principal component analysis, modified grey relational analysis, absolute principal component score-multiple linear regression, and positive matrix factorization receptor modeling. Although these traditional risk assessment models exhibit robustness, they often struggle in data-deprived environments, such as the Niger Delta; and also demonstrate the necessity of a thorough method for evaluating the risk of groundwater contamination that takes into account both inherent susceptibility and particular causes of contamination.

The Source-Pathway-Receptor (S-P-R) model (Belle et al., 2023; EA., 1999), characterized by its simple and flexible conceptual design, offers a promising solution by providing a comprehensive risk profile. It envisions risk as the cumulative product of source potency, pathway transmissivity, and receptor susceptibility. While individual open-source datasets may possess limitations, their collective integration holds the potential to effectively address and close existing knowledge gaps. To create a thorough and reliable map that illustrates the latent dangers related to groundwater contamination by hydrocarbons, this research attempts to investigate the integration of heterogeneous datasets. Therefore, to improve the comprehension of possible contamination hazards, the study aims to create an integrated framework using open-source data.

2 The Niger Delta Region (NDR) – a Brief Overview

The NDR comprises of nine states in the southern part of Nigeria and is located on the east of the Gulf of Guinea. It is an easy to drill terrain, with considerable groundwater potential and hydrocarbon-prolific region, with several interconnected pipelines which result in environmentally degrading incidents such as oil spills (Fig. 1). Just like many deltas, the course of the waterbodies like channels and creeks are constantly altered due to consistent erosion and deposition. Artisanal farming and fishing are the principal occupations in most communities.

The Niger Delta formed due to the accumulation of fluvial and marine sediments since the upper Cretaceous period and consists of three main formations, namely topmost Benin, middle Agbada and bottom Akata Formations. The Benin Formation is the uppermost layer geologic layer comprising significantly of variably thick multiple layers of sand, discontinuous clay, peat/lignite and conglomerate, thus resulting localized aquifers or hydraulically interconnected aquifers (Akinwumiju et al., 2020; Ojoboh et al., 2020; Raimi et al., 2021a; Raimi et al., 2021b). The aquifers are generally shallow, ranging from close to the surface to 10 m below the ground surface, and fluctuate seasonally. In recent times, groundwater has become crucial to the region as most of its residents depend on it as a main source of potable water. Residents access these aquifers via indiscriminate and randomly sited hand-dug wells with diameters of around 60 cm and/or shallow boreholes of about 10–30 m deep, either manually or with pumps.

The majority of the settlements in the area are small, dispersed hamlets (Sam et al., 2017b), with nearly 31 million people residing in this region within 13, 231 rural and 98 urban settlements respectively (Wekpe et al., 2022), with a heavy dependence on the land and natural resources for their livelihoods, especially agriculture, which includes farming and fishing (Sam et al. 2017a, b). However, the ongoing environmental degradation and depletion of economic resources in the NDR births far-reaching socio-economic challenges (including unemployment, conflict, poverty, insecurity and migration) as many peasant farmers and fishermen are grossly impacted by the loss of farmlands, vegetation, wildlife, biodiversity and various ecosystems (Agbonifo, 2022; Sam et al., 2017a; Wekpe et al., 2022).

Fig. 1
figure 1

Integrated view of the Niger Delta region, highlighting the spatial relationship between population density and oil-related infrastructure. The varying shades of colour in the background denote population density categories, which are based on local adaptations of global standards (see Sect. 3.3 for details)

Full size image

3 Data and Method

The method utilized the Source- Pathway-Receptor (S-P-R) linkage, involving multiple criteria analysis for assessing the environmental risk (Abdullah et al., 2022; Danumah et al., 2016; Sam et al., 2017b; Sun et al., 2023). The conceptualization derived from examining the interplay between contamination sources, their migration pathways, and the impacted receptors, providing a structured approach to risk assessment.

Given the absence of readily available datasets encompassing sources, pathways, or receptors, proxies have been adopted to characterize source potency, pathway transmissibility, and receptor susceptibility. Utilizing a multistage decision-making process involving literature review and scientific logic, fourteen parameters were identified and classified according to the S-P-R model (Table 1). These parameters were selected based on various previous research and publications that engender risk mapping (Akinwumiju et al., 2020; Lyu and Yin, 2023; Mastouri et al., 2022; Sam et al., 2017b; Zheng et al., 2021), groundwater potential (Benjmel et al., 2020, 2022; Phong et al., 2021; Regmi et al., 2022), groundwater contamination (Chen et al., 2019; Chinye-Ikejiunor et al., 2021; Eisenhauer et al., 2016; El-Alfy, 2024; Giadom and Tse, 2014; Zhao et al., 2018), and flood mapping (Allafta and Opp, 2021; Dung et al., 2022; Lyu et al., 2023; Pham et al., 2021; Siam et al., 2022) as all these have significant relationships with the focus of study. Relative weightings were assigned after normalized to ensure weights sum to 100% or 1 and resampled into 30 m resolution., allowing for easy comparison and aggregation of factors (supplementary material Figure S.1 for more details). It attempts to quantify the relevance and contributions to a specific parameter, when compared with other contributory factors (Tonidandel et al., 2009). These weightings were based on expert judgments and comparisons with existing literature, thus representing the level of significance deemed appropriate for each parameter in the current investigation. While the subjective nature of this approach could serve as a limitation, as is common with multi-criteria decision analyses, it nevertheless provides a valuable opportunity to investigate the relative variation of the risk. Normalised parameters were assigned weights ranging from 0.1 to 1.0, indicating their relative importance. The inverse (x− 1) of the thematic layers were utilized to depict the inverse relationship between such criteria (x) and the parameters being analysed. However, in such cases, the initial actual values were inversed, then normalized, as depicted above.

The fourteen criteria were grouped into three main classes (Fig. 2), viz.: Source Potency (Sp), Pathway Transmissivity (Pt) and Receptor Susceptibility (Rs). Sp describes all the factors that could impact on the source properties, while Pt refers to all the criteria that could influence the rate at which the oils could penetrate and infiltrate into the subsurface, especially into the groundwater system of the area; and Rs describes the human susceptibility to contaminated groundwater and the resulting impact of such on the health of the human population (Figure S.1 – S.3).

The final scores for each component (Sp, Pt, and Rs) were computed and integrated to generate the final overall regional risk map.

Fig. 2
figure 2

The S-P-R framework for groundwater contamination risk assessment of the Niger Delta in Nigeria. The source potency (Sp) is characterised by five parameters, the pathway transmissivity (Pt) by eight parameters and the Receptor Susceptibility by population density. The risk area is defined by the intersection of all three main parameters while three main interactions exist with individual pairs, including the source potency-pathway transmissivity (Sp-Pt), source potency-receptor susceptibility (Sp-Rs) and Receptor Susceptibility-Pathway Transmissivity (Rs-Pt). These indicate potential future risk areas

Full size image
Table 1 Assessment of parameters influencing the S-P-R Risk Model. This table categorizes the identified criteria into three primary classes: Source Potency (Sp), Pathway Transmissivity (Pt), and Receptor Susceptibility (Rs), elucidating their respective impacts within the risk assessment framework
Full size table

3.1 Source Potency Sp

Source potency is conceptualized as a function of five key parameters: the volume (Vs) and type of spill (Ts), along with distances to spill sites (Ss), pipelines (Pp), and oil wells (Ow) (Eqs. 1–2). This equation reflects the integration of both physical and spatial attributes of spills, offering a comprehensive view of the risk they pose to the environment.

$$Sp=f (Vs, Ts, Ss, Pp, Ow)$$
(1)
$$Sp=0.2Vs+0.2Ts+0.3{Ss}^{-1}+0.15{Pp}^{-1}+0.15{Ow}^{-1}$$
(2)

Where \({Ss}^{-1}=Normalised \left[\right({Actual\,values\,of\,Ss)}^{-1}]\)

$${Pp}^{-1}=Normalised \left[\right({Actual\,values\,of\,Pp)}^{-1}]$$
$${Ow}^{-1}=Normalised \left[\right({Actual\,values\,of\,Ow)}^{-1}]$$

This multifactorial approach determines source potency based on the understanding that each parameter uniquely contributes to the environmental impact of a spill (Table 1). The weightings signify authors’ opinions on significance of each factor, after considering and comparing those used by previous researchers in related risk mapping. The volume of a spill directly correlates with the risk of groundwater contamination, with larger volumes posing greater threats. Similarly, the type of spill is crucial, as crude oil significantly increases the risk of contamination compared to more volatile refined oils. Wekpe et al. (2024) and Wekpe and Idisi (2024) emphasize the impact of quantities and frequencies of spills in clusters aiding oilspill hotspot regions and stress on vegetation while El-Alfy (2024) & Fei-Baffoe et al. (2024) insinuate that some fuels (crude oils like total petroleum hydrocarbons TPHs & polycyclic aromatic hydrocarbons PAHs) have more residence time and therefore higher probability of contaminating groundwater than volatiles. The spatial parameters, including proximity to spill sites, pipelines, and oil wells, are critical. Closer proximity typically translates to a higher risk of contamination, emphasizing the importance of spatial dynamics in environmental risk assessment. Sam et al. (2017b) emphasise the importance of proximities to various facilities such as spill locations and pipelines in likelihood of exposures during risk-based assessments.

3.2 Pathway Transmissivity Pt

The migration potential of contaminants, or pathway transmissivity, is determined by eight key factors: percentage of sandiness (Sa), water table (Wt), elevation (El), slope (Sl), river proximity (Rp), drainage density (Dd), precipitation/rainfall (Pr) and land us/land cover (Lc) (Eqs. 3–4). These criteria can aid or hinder the recharge of groundwater along with contaminants. Since infiltration, groundwater movement, and flooding events are considerably related to groundwater contamination (Table 1), these factors were integrated together as the degree of transmissivity of the pathway materials depends significantly on them and their interactions. In addition to extracting weighting and scoring choices from experts and practitioners like geologists, hydrologists, environmental managers, policy makers, environmentalists, and scientists studying plants and animals, expert opinions were also gathered from sources of existing literature.

$$Pt=f (Sa,Wt, El,Sl, Rp,Dd, Pr, Lc)$$
(3)
$$\begin{array}{l} Pt=0.2Sa+0.1{El}^{-1}+0.1{Sl}^{-1}+0.1Rp\\+0.1Dd+0.1Pr+0.1Lc+0.2{Wt}^{-1}\end{array}$$
(4)

Where \({El}^{-1}=Normalised \left[\right({Actual\,values\,of\,El)}^{-1}]\)

$${Sl}^{-1}=Normalised \left[\right({Actual\,values\,of\,Sl)}^{-1}]$$
$${Wt}^{-1}=Normalised \left[\right({Actual\,values\,of\,Wt)}^{-1}]$$

3.3 Receptor Susceptibility Rs

Receptor susceptibility to environmental contaminants can be influenced by a range of factors, including age, sex, socio-economic status, and health condition, as highlighted by (Sam et al., 2017b). However, due to the limitations of available data, we focused on population density as the sole parameter. A higher density implies a greater total impact (Eqs. 5–7), acknowledging that human density variations directly affect contamination risk. Thus, Rs, defined as a function of population density, allows for the assessment of human vulnerability based on proximity and density relative to contamination sources. Expert information and knowledge taken from numerous published studies were investigated in order to assign suitable weights to the various evaluation criteria.

$$Rs=f (Pd)$$
(5)
$$Rs={\sum }_{ }^{ } Pd$$
(6)

Where:

$${\sum }_{ }^{ } Pd=0{Pd}_{nv}+0.25{Pd}_{v}+0.75{Pd}_{pu}+{Pd}_{u}$$
(7)

This approach, while simplified, offers valuable insights into the generalized potential impact of environmental contaminants on human health within the constraints of the current data. A range of population density (Pd) is utilised, where natural vegetations with no human habitats (with\({Pd}_{nv}\)) is believed to be of zero susceptibility; villages (\({Pd}_{v};\)0 < Pd < 100) is believed to be about 25% more susceptible; peri-urban regions (\({Pd}_{pu}\); 100 < Pd < 200) is weighted to be three times more susceptible than the villages; and urban regions (\({Pd}_{u}\); Pd > 200) are assumed to be of peak susceptibility to impact of hydrocarbon contamination in groundwater.

The categorisation of population density, and thus receptor susceptibility, is informed by global standards, albeit adjusted to reflect the local context of Nigeria. For this study, the population density standards applied are less stringent than those of more densely populated and expansive nations such as India and the USA. In India, urban areas are defined by densities upwards of 396 persons per km² (Long et al., 2001), while in the USA, urban densities are considered to be 620 persons per km² or greater (McDonnell, 1990). These benchmarks are recalibrated for Nigeria, taking into account the country’s less extensive land area and population distribution (Ofem, 2012). In this case, that of Australia as stated in Ofem (2012) was adapted and recalibrated to fit Nigeria situation such that population densities of 0, < 100, 100–250 and > 250 were classified as natural reserves, rural, peri-urban and urban areas respectively.

3.4 Overall Risk and Classifications

The final scores for each component (Sp, Pt, and Rs) were computed and integrated to generate the final overall risk score (R in Eq. 8). This method adheres to the principle that risk (R) is a product of hazard, exposure, and vulnerability, ensuring a comprehensive assessment of environmental risks.

$$R=Sp\times Pt\times Rs$$
(8)

This methodological framework offers a detailed and reproducible approach for assessing groundwater contamination risk in the Niger Delta. It enables future researchers to apply similar analyses in other contexts. Normalization of values ensures that the sum of the weights equals 100% or 1, facilitating easy comparison and aggregation of factors, and keeps the overall risk score between 0 and 1. Additionally, this approach delineates interactions/overlaps between Sp, Pt, and Rs, identifying areas where risks could escalate under changing conditions.

Determining cutoff values to classify the Niger Delta into low, moderate, and high groundwater risk zones was a nuanced process. The overall risk scores, while quantitative, represented relative rather than absolute risks. Initially, the Jenks natural breaks method (Chen et al., 2013) was employed for zoning, optimizing the arrangement of overall risk scores into categories that minimized variance within and maximized variance between classes. However, given the relative nature of risk scores, refining these classifications to reflect the complex realities of the Niger Delta required further steps.

To ensure the relevance and accuracy of our risk zones, we aligned our initial statistical classification with documented high-risk areas identified in prior literature, including significant sources such as the UNEP report on Ogoniland (UNEP, 2011) and studies by Giadom and Tse (2014) and Akinwumiju et al. (2020). This literature-based validation and adjustment process was crucial for accurately mapping areas affected by hydrocarbon pollution and contamination.

The final risk map, informed by both the Jenks natural breaks and literature insights, provides a comprehensive view of the Niger Delta’s vulnerability to groundwater contamination. It not only corroborates known areas of environmental risk but also reveals previously unidentified zones of concern. This approach ensures that our methodological framework, while detailed and reproducible, is also grounded in the contextual realities of the region, offering a valuable tool for future research and intervention planning.

4 Results

4.1 Source Potency

Spatial variation in source potency – derived from the cumulative impact of spill volume and type, in conjunction with proximity to spill sites, pipelines, and oil wells – indicates that over 50% of the landscape is under threat from hydrocarbon contamination (Fig. 3a).

In terms of geographical extent, 48% of the Niger Delta Region (21,000 km²) is demarcated as low source potency. These areas predominantly encompass the parts of Ondo and Delta states, extend into the northern reaches of Edo, the southern regions of Imo, the southwestern quadrant of Rivers, and span across the major portions of Bayelsa and the entirety of Akwa Ibom state. The moderate source potency zones cover 39% of the region (17,000 km²), while high source potency areas, which are intimately associated with extensive pipeline networks and historical spill sites, constitute 13% (6,000 km²). These high-risk zones are primarily concentrated in significant parts of Delta and Rivers states, with some impacted areas within Bayelsa.

Fig. 3
figure 3

Visual assessment of hydrocarbon risk factors in the Niger Delta Region. On the left: (a) Source Potency: Over 50% of the region shows moderate to high contamination risk. (b) Pathway Transmissivity: Two-thirds of the region have moderate to high permeability. (c) Receptor Susceptibility: Over 50% of the region, especially in urban areas with high population densities. On the right (with areas in red denoting the most significant interactions) (d) Source Potency and Pathway Transmissivity (Sp-Pt) Interactions: Highlighting areas where contaminants could reach groundwater. (e) Receptor Susceptibility and Pathway Transmissivity (Rs-Pt) Interactions: Indicating potential  future risk areas, particularly in the event of new oil spills. (f) Source Potency and Receptor Susceptibility (Sp-Rs) Interactions: Highlight areas with significant non-groundwater exposure risks. Areas show significant interactions despite low pathway transmissivity

Full size image

4.2 Pathways Transmissivity

The spatial analysis, which synthesises the interplay of eight morphological parameters — sand content, elevation, slope, proximity to rivers, drainage density, rainfall, land cover, and groundwater level — has yielded a nuanced map of pathways transmissivity depicting the fluid and contaminant migration potential of the regions (Fig. 3b). This map is a demonstration to the complexity of subsurface, delineating the variance in permeability and, consequently, the potential for oil spill hazard progression.

The directional spread of transmissivity begins in the northwestern sector with Edo state, where the prevalence of red indicates high pathway transmissivity. As one traverses southwards and eastwards, a transition is observed, with Delta, Rivers and Imo states displaying a mix of moderate and high transmissivity zones. Further southeast, the states of Abia and Akwa Ibom also show significant areas with moderate to high transmissivity, aligning with the broader pattern observed across the region. Notably, approximately 21% of the region (8,995 km2), predominantly these states, demonstrates a high propensity for pollutant percolation, with subsurface materials conducive to rapid fluid movement. Conversely, the lower transmissivity zones, indicated by green (33%, 14,584 km2), are more pronounced in the coastal and southwestern areas, particularly within Bayelsa, and parts of Rivers, and Delta states. These areas, constituting about a third of the region, exhibit subsurface characteristics that are less conducive to the mobility of contaminants. The lower transmissivity areas coincide with regions that exhibit a denser clay composition, reduced drainage density, and flatter geomorphology, which collectively impede the migration of contaminants.

4.3 Receptor Susceptibility

Receptor susceptibility in the Niger Delta Region directly stems from dependency on groundwater by the population, a vital resource imperiled by the widespread hydrocarbon contamination. The population density map (Fig. 1) provides a proxy measure for this susceptibility, with more densely populated areas reflecting a heightened potential for human exposure to environmental hazards. The analysis of receptor susceptibility was refined through a grading system where population density was proportionately weighted, asserting that higher densities amplify the susceptibility index (Fig. 3c). The map reveals that around 22% (9,900 km2) of the region falls within moderate to high receptor susceptibility categories, typically associated with urban or peri-urban areas. These higher vulnerabilities are especially prevalent in the central locales of Edo and Delta states, as well as in the northeastern sections of Rivers and Bayelsa, extending into significant portions of Akwa Ibom and Imo. In stark contrast, areas such as Ondo and peripheral zones of Edo present lower population densities, with vast expanses of nature reserves and rural villages.

4.4 Interaction between Risk Components

The data analysis further explored the interplay between two risk components at a time, across all three potential combinations: Source Potency and Pathway Transmissivity, Source Potency and Receptor Susceptibility, and Pathway Transmissivity and Receptor Susceptibility (Fig. 3, d - f). These dual interactions not only shed light on the contribution of the third component to the total risk but also highlight the spatial significance of that component.

The intersection between Source Potency and Pathway Transmissivity (Fig. 3d) reveals spatial variability in the merged score. Regions where both source potency and pathways are transmissive—indicating a higher potential for groundwater contamination—are shown. Moderate to high merged scores (36%; 15,000 km² to 10%; 4,000 km²) were observed across the Niger Delta, particularly in significant areas of Delta, Imo, and Rivers states, as well as the northern and southern regions of Abia. Conversely, the northern parts of Edo, southern Ondo and Imo, western Delta, and most parts of Bayelsa and Akwa Ibom states show low interactions, accounting for approximately 54% of the area (23,000 km²).

The overlap of Receptor Susceptibility and Pathway Transmissivity highlights regions with a high capacity for infiltration within populated areas. Approximately 10% (4,500 km²) of the NDR is characterised as moderate, and 11% (5,000 km²) as high for these interaction scores. These span substantial portions of Akwa Ibom and Abia, the northern parts of Imo, central Delta, western Edo, northeastern Bayelsa, and the north and northeast of Rivers state (Fig. 3e). In areas without oil spill events, other contaminants such as surface chemicals, and even septic-tank can serve as alternative sources of groundwater contamination.

In the case of Source Potency and Receptor Susceptibility, higher merge scores are mostly confined to Delta and Rivers states (Fig. 3f). Around 17% of the Niger Delta Region corresponds to moderate (6,000 km²) to high (1,500 km²) interaction zones. These include the northern reaches of Imo and Abia, central and northwest Delta, central, northern, and southeast Rivers, and northwest Akwa Ibom. These areas potentially represent high human health risks, not through groundwater contamination but through direct exposure via inhalation, dermal contact, and other pathways.

4.5 Overall Risk Map

The synthesis of the multicriteria analysis with the source-pathway-receptor (S-P-R) framework culminated in the generation of an overall risk map for the Niger Delta Region (Fig. 4). This map delineates the spatial distribution of groundwater risk zones (GWRZ) based on a composite index derived from the interplay of fourteen parameters encompassing source potency, pathway transmissivity, and receptor susceptibility.

The risk index—a product of the aforementioned factors—was computed for each pixel within the study area, yielding a dimensionless map. This map was subsequently categorised into regions representing different levels of GWRZ, reflecting the cumulative hydrocarbon contamination risk each area faces. The map reveals that a portion of the region, encompassing 13% (5,000 km²), exhibits moderate GWRZ, while a smaller yet critical 3% (1,400 km²) is characterised by high GWRZ. These high-risk zones are prominently located in central and northern Delta, northern Imo, northwest Akwa Ibom, north and south Abia, eastern and northern parts of Rivers, northeast Bayelsa, and northwest Edo, with scattered patches observed throughout these states.

In stark contrast, most of the region, approximately 84% (35,000 km²), is identified as low GWRZ. This predominantly encompasses significant expanses of Ondo, Edo, Delta, Bayelsa, Akwa Ibom, and Imo states. The red regions on the map signify areas of absolute intersection among the risk components, indicating the highest susceptibility to groundwater contamination risks.

Fig. 4
figure 4

Overall risk map generated from integrating the multicriteria analysis under the source-pathway-receptor (S-P-R) framework. Here, regions in red depict areas of high-risk levels which signify likely intersection of the fourteen parameters contributing to the source potency, pathway transmissivity and receptor susceptibility

Full size image

5 Discussions

5.1 The Source-Pathway-Receptor Model for Regional Risk Assessment

The employment of the source-pathway-receptor (S-P-R) framework (EA., 1999) in a regional context, as demonstrated in this investigation, marks a novel approach to evaluating groundwater contamination risk due to oil spills within the Niger Delta. This study’s integration of source potency, pathway transmissivity, and receptor susceptibility (Sp, Pt & Rs) transcends traditional analyses by illuminating the interactive dynamics of contamination risk. Our findings elucidate that risk is not solely contingent upon the presence of contamination sources but hinges on the interplay among all three S-P-R components.

The utilisation of the S-P-R model has delineated distinct risk zones, with a pronounced emphasis on areas exhibiting absolute parameter interactions (Fig. 4). These regions of heightened risk, notably within the Delta and Rivers states, correlate with historical spill incidents and extensive petroleum infrastructure, necessitating prioritised remediation efforts (Akinwumiju et al., 2020; Sam et al., 2017b; Umar et al., 2019; UNEP, 2011; Wekpe et al., 2022). In contrast, areas characterised by source-pathway (S-P) connections may not present immediate risks but signal potential future concerns, contingent upon an increase in receptor activity. This is also consistent with research by Wekpe et al. (2024), Mba et al. (2019) and Ugwu et al. (2021) who discovered that the majority of oil spill interdiction happens at pipeline locations used to carry refined goods and crude oil from oil fields to export terminals. Furthermore, our research acknowledges the possibility of hydrocarbon exposure through alternative pathways in regions with source-receptor (S-R) linkages (Fig. 3f), corroborating with prior documentation of oil pollution and associated inhalation, and other exposure risks (Anyanwu et al., 2023; UNEP, 2011). Pathway-receptor (P-R) associations suggest vulnerability to a spectrum of contaminants beyond hydrocarbons, a scenario evidenced in segments of the delta where non-petroleum-based pollutants, such as wastewater and pesticides, have been identified (Abdel-Shafy and Mansour, 2016; Olalekanand Dodeye, 2020; Olalekan et al., 2021; Olalekan et al., 2018; Olalekan and Emman, 2020; Ozoemenam et al., 2018).

While the S-P-R model is a staple in contaminated land management within the UK, primarily within small-site applications (Sam et al. 2017a), its adaptation to the scale of this study represents an unique extension. Historical precedents in NDR such as the investigations of oil spill events by Akinwumiju et al. (2020) and the geospatial distribution analyses of Umar et al. (2019) have not employed the S-P-R framework to assess hydrocarbon spill risks on groundwater quality. Our investigation reveals that risk zones emerge not merely from the presence of a contamination source but from the intricate interactions of all three S-P-R elements. The innovative employment of open-source data proxies and multifaceted datasets further enhances the model’s applicability to large-scale environmental assessments.

5.2 Appropriateness of S-P-R Approach and Overall risk

The selection of pertinent factors in environmental risk assessments, such as those for flood mapping, often lacks uniform consensus (Tehrany et al., 2014). Nevertheless, the variables integrated within this study for assessing groundwater contamination risk from oil spills have been widely adopted in related fields, including flood and groundwater potential mapping, underscoring their relevance (Benjmel et al., 2020, 2022; Dar et al., 2021; Mukherjee and Singh, 2020; Parsian et al., 2021; Phong et al., 2021; Regmi et al., 2022; Tehrany et al., 2015). These issues share commonalities with oil spill disaster mapping, where the source-pathway-receptor (S-P-R) framework is particularly apt. Key factors such as slope, proximity to rivers and spill sites, drainage density, land use, and elevation were chosen due to their established correlation with both geology and groundwater vulnerability (Fig. 2; Table 1).

The derived overall risk map illustrates the spatial distribution of risk across the delta, pinpointing high-risk zones predominantly in the central regions, in contrast to the low-risk areas in the northern and southern areas (Fig. 4). The employment of the S-P-R approach within a multistage, multicriteria decision analysis has proven substantial for delineating areas prone to contamination.

Despite the strength of the S-P-R model, inherent uncertainties persist, stemming from raw data quality, data processing, criteria selection, and threshold determination (Lyu et al., 2020a, 2020b). To fortify the findings and mitigate uncertainties (Zheng et al., 2021, 2022) recommend augmenting the S-P-R approach with additional techniques to yield more comprehensive outcomes, especially in future remediation endeavours. Validation and sensitivity analysis are also advocated by Mukherjee and Singh (2020) to address potential subjectivity and biases, particularly those introduced through expert judgment influencing parameter weightings.

Correlating our risk assessment results with existing groundwater quality reports, areas classified as moderate to high risk coincide with locations reporting hazardous concentrations of hydrocarbons, including benzene, PAHs, and TPHs ((Alinnor et al., 2014; Amangabara and Njoku, 2012; Egirani and Chidi, 2023; Egirani and Alaowei, 2024; Egirani and Dagogo-Jack, 2023; Lindén and Pålsson, 2013; Lubi and Akinluyi, 2022; Nwankwoala and Mzaga, 2017; Raimi et al., 2022; UNEP, 2011). Notably, Akinwumiju et al. (2020) discerned a relatively low incidence of oil spills across the Niger Delta Region, potentially attributable to the integration of limited datasets and a predominantly statistical approach. While their research did not focus on groundwater risk, they identified Rivers and Bayelsa states as areas of concern for spill incidents, corroborating with our findings of heightened source potency and thus elevated risk. Significantly, our study demonstrates the presence of both pathway transmissivity (Pt) and receptor susceptibility (Rs) in these same regions, which were identified as exhibiting moderate to high risk of spill incidents, with corrosion highlighted as a principal causal factor. This convergence of all three S-P-R components establishes a comprehensive linkage, underscoring the risk these areas face.

Another study by Umar et al. (2019) highlighted hotspots in Rivers, Bayelsa, and Delta states through spatial analysis of previous spill events but did not extend their scope to groundwater risk implications. In contrast, our study expands upon this foundational work by applying the source-pathway-receptor (S-P-R) model, which incorporates a holistic view of how oil spills can potentially impact groundwater quality. While Umar et al. (2019) provided valuable spatial insights into spill occurrences, our approach integrates these insights within the S-P-R framework to assess the interconnectedness of spill sources, environmental pathways, and vulnerable receptors. This comprehensive methodology not only reaffirms the relevance of the identified hotspots but also enhances our understanding of their potential environmental and public health impacts.

5.3 Implications for Groundwater Sustainability and Risk Management

In this research, various open source datasets were utilised such as historical spill data from Nigerian Oil spill response and Detection Agency (NOSDRA), elevation and other related data from the USGS website, population data from the landscan website and static water level from published articles (Table 1). Each dataset on its own tend to be poor and insufficient in providing a wholesome and accurate picture of the specificities of the NDR and would otherwise be unreliable. This is especially true for the historical data on oil spills as there exists a disparity between datasets from the Nigerian agency, licensed oil companies and indigenes’ reports/perspectives (Sam et al., 2017b; Wekpe et al., 2022). In other words, a significant uncertainties revolve around some of the datasets, however, output of integration of these dataset can provide real insights with improved reliability (Dalla Valle and Kenett, 2018).

Due to insecurity in NDR and lack of a systematic, accessible and frequently updated database, direct measurements of oil spill risk assessments may not be possible in the Niger Delta. This necessitates the formulation of efficient, cost-effective frameworks towards mitigating the risk of oil spills impacting its groundwater quality (Bruckberger et al., 2020; Ivshina et al., 2015). The current study was able to identify and present the most important conditioning parameters for groundwater contamination from oil spill events, as well as the most oil spill vulnerable areas (Fig. 4). This indicates the need for further investigations, especially with relation to groundwater quality should be paramount in southern and central parts of the delta, within states such as Delta, Abia, Akwa Ibom, Imo and Rivers.

The weights assigned to each parameter in our risk assessment were derived through a combination of expert judgment and comparison with existing literature. This approach ensured that the weights reflected both the significance of each parameter in the context of the Niger Delta and the findings from previous studies in similar settings. Despite the robustness of multi-criteria decision analyses (MCDA), we acknowledge the inherent subjectivity and potential limitations in our weighting process.

To evaluate the impact of these weights and address the limitations of MCDA, we conducted a simple multi-parametric sensitivity analysis. This analysis involved the elimination of each parameter to observe changes in the risk assessment outcomes. The results, detailed in Supplementary Figures S.4, S.5 and S.6 and Supplementary Table S.1, indicate that most parameters are insensitive, except for Euclidean distance to spill sites and pipelines for source potency, and elevation, slope and river proximity for pathway transmissivity. For the overall risk mapping, significant sensitivity was observed only for Euclidean distance to pipelines and population densities. These findings suggest that certain parameters play a critical role in determining groundwater contamination risk. The immobility of pipelines poses a challenge for risk management, indicating that while the risk of oil spills impacting groundwater quality cannot be entirely mitigated, focus should be directed towards effective remediation and consistent monitoring of affected regions. Our sensitivity analysis underscores the necessity for targeted interventions and continuous assessment to safeguard groundwater resources in the region.

This risk mapping research offers a preliminary, large-scale assessment and insights; therefore, researchers can use these results as a foundation for further investigations and studies. The findings can serve as a valuable resource for understanding the specific risks and vulnerabilities in each area; and are highly replicable in other oil producing deltaic environments. It can also aid governmental agencies in making informed decisions and allocating resources such as remedial activities, monitoring, early warning systems, emergency response protocols and appropriate location-based legislation (Adeniran et al. 2023; Bingari 2023; Bingari et al. 2023; Sam et al. 2017b). It could also help to signpost areas to engage in vital community-level awareness and advocate for improved protective measures; while considering alternative potable water sources such as deeper borehole which is likely to be in separated from the shallow contaminated flow system (Hoque et al. 2017). Effective collaboration and proactive actions between all stakeholders, including government entities, researchers and individuals will considerably reduce the potential harm and preserve the health and well-being of both the communities and the ecosystems.

6 Conclusions

In this study, the Source-Pathway-Receptor (S-P-R) model, augmented by a multilevel, multicriteria analysis, has been employed to assess the risk of groundwater contamination due to oil spills in the Niger Delta Region (NDR), Nigeria. Through the innovative integration of fourteen thematic data layers, this research delineates the NDR into low, moderate, and high groundwater risk zones, uncovering areas previously unrecognized as vulnerable to contamination. Our findings reveal the critical interplay between source potency, pathway transmissivity, and receptor susceptibility, offering new insights into the spatial dynamics of environmental risk within the delta. This underscores the necessity for further research into the actual groundwater quality within the Niger Delta Region, particularly advocating for large-scale screening of groundwater quality to identify contaminated areas and prevent usage. For instance, it was revealed that about 52% of the NDR ranges from moderate to high source potency, while 80% of the region exhibits moderate to high pathway transmissivities and about 78% of the NDR suggest low receptor susceptibility levels. The final risk map suggests that about 16% of the region falls within the moderate to high groundwater risk zones and mainly encompasses the central and eastern parts of the region.

High-risk zones prone to oil spills were pinpointed, providing essential information for the monitoring, remediation, and management of groundwater resources in the Niger Delta and other oil-rich regions. To ensure the reliability of our findings, the study’s risk maps were validated against previously published works, addressing potential limitations of the individual datasets. The study identifies moderate to high-risk zones, especially in Rivers, Bayelsa, Delta states, and low-risk areas in Ondo and significant portions of Edo, Bayelsa, and western Imo states.

By applying the S-P-R model to a complex and ecologically sensitive region, this study not only advances the field but also sets a precedent for future large-scale environmental risk assessments using open-source data. In line with UN Sustainable Development Goal (SDG) #6, it underscores the necessity of targeted strategies to ensure clean water provision in high-risk areas.