Understanding Complex Hydraulic Heterogeneities in Crystalline Basement Aquifers Used as Drinking Water Sources

Abstract

Managing groundwater flow in crystalline basement aquifers (CBAs) remains challenging due to their dependence on secondary permeability fields characterized by high spatial variability. This study combines pumping and tracer tests to estimate the hydraulic properties and connectivity in four bedrock wells within a CBA in Southwestern Nigeria. The pumping tests caused drawdowns up to 4.13 m and 12.60 m in observation and pumping wells, with significant drawdowns only in three of four wells, revealing poor connection with the fourth well. The time-drawdown plots confirm double porosity effects suggesting fracture and matrix flow and release of water from a fractured dyke. Fracture and matrix hydraulic conductivities exceeded 7.9 × 10⁻⁷ m/s and 1.00 × 10⁻¹⁰ m/s, while the aquifer yield ranged from 0.08 to 0.34%. Groundwater flow velocity and dispersivity of 5.80 × 10⁻⁴ m/s and 2.60 m were estimated from the tracer test, while a Peclet number of 3.25 suggests dominant advective flow. Calculated sustainable yield shows that each well could provide water for up to 1600 people under controlled low pumping at 0.50 l/s with higher rates possible using larger diameter wells. These results confirm high variability in groundwater flow within CBAs, justifying the need to characterize them effectively.

1. Introduction

Crystalline basement aquifers (CBAs) are significant sources of drinking water in most parts of sub-Saharan Africa, including Southwestern Nigeria [1] despite being characterized by low hydraulic conductivity and specific yield [2,3]. Sustainable management of these critical water resources requires understanding the aquifer yield and the distribution of their hydraulic properties and connectivity [4,5]. The hydraulic properties, including transmissivity, hydraulic conductivity, storativity, flow velocity, and dispersivity, are essential for informing groundwater models used to predict flow and transport within these aquifers [6,7]. Limited studies have, however, investigated the hydraulic properties distribution and connectivity within CBA in Southwestern Nigeria, which are inherently heterogeneous [8]. This study presents a field assessment of hydraulic properties variation, connectivity, and sustainable yield of a CBA at the Ibadan Hydrogeophysics Research Site in Southwestern Nigeria using multiple pumping and tracer tests.

CBAs are formed by intense weathering and fracturing of impermeable igneous and metamorphic rocks, such as granite and gneiss, creating regolith of varying thickness, which overlies crystalline rocks with discontinuities, such as fractures, joints, and faults [9,10,11]. Generally, the geometry of CBAs consists of a topsoil layer, a weathered overburden termed saprolite, a weathered and fractured crystalline rock termed saprock, and a fresh crystalline rock serving as an aquitard [2]. CBAs exhibit complex groundwater flow characteristics due to their low primary porosity and dependence on secondary porosity created by fractures, joints, faults, and weathering [12].

The intensity and connectivity of fractures determine the hydraulic properties, such as hydraulic conductivity controlling flow and transport within these aquifers [13,14]. In intensely fractured zones, groundwater flow is enhanced due to increased hydraulic conductivity, whereas in areas with sparse or poorly connected fractures, groundwater flow is significantly restricted [15,16]. Also, groundwater storage and transmission within CBAs are strongly affected by the degree of weathering within the crystalline basement rocks [17]. The weathered zone or regolith consisting of weathered crystalline rock and clay-rich material stores considerable amounts of water [11,18]. However, the high clay content reduces its hydraulic conductivity, restricting groundwater movement [2]. The transition zone between the weathered layer and the fractured bedrock allows water storage which recharges the fractured basement [19,20,21]. The complex geology, weathering, and fracture distribution within CBAs make their flow and transport properties vary over several orders of magnitude [14].

The high variability in hydraulic conductivity and other flow and transport parameters poses significant challenges in groundwater assessment and management within crystalline basements [2,22]. Unlike sedimentary aquifers, where flow properties are relatively less variable, CBAs exhibit localized high-permeability pathways interspersed within low-permeability rocks [14,23]. This heterogeneity complicates the prediction of groundwater flow and contaminant transport, requiring advanced hydrogeological investigations and modeling approaches. Quantifying these variations necessitates the use of geophysical surveys [24,25], hydraulic testing, and numerical modeling techniques, all of which are constrained by inherent uncertainty [4,6,26,27].

Hydraulic testing, including pumping and tracer tests, has been widely used to estimate the variation in aquifer flow and transport properties [28,29,30,31]. Pumping tests are used to estimate aquifer transmissivity, hydraulic conductivity, and storativity by analyzing the drawdown response to pumping in wells over time [32,33,34]. These tests typically involve withdrawing water at a constant or stepwise varying rate from a well while monitoring the resulting water level changes in both the pumping and observation wells [31]. The interpretation of pumping test data relies on analytical and numerical models that assume aquifer homogeneity, isotropy, and confined or unconfined flow regimes [35,36,37]. For fractured crystalline basement aquifers, drawdown curves often deviate from the idealized Theis-type response, reflecting the dual-porosity nature of these systems, where groundwater flow occurs both within the fracture network and the surrounding rock matrix [38,39]. The presence of fractures can lead to early rapid drawdown followed by a delayed response due to matrix diffusion and storage effects [40,41]. Additionally, time-drawdown analysis can reveal whether the aquifer behaves as a single or double porosity system, with fracture-dominated flow exhibiting a characteristic non-linear decline in drawdown over time [42,43].

Besides the classical constant-rate pumping tests, step-drawdown tests, in which the pumping rate is increased in discrete steps, are also used to assess the well efficiency and to identify turbulent flow effects near the wellbore [44,45]. The interpretation of step-drawdown data allows for distinguishing between laminar and turbulent flow components, which is crucial for understanding the impact of well losses on measured hydraulic properties [46]. After a pumping test is completed, the recovery of the well can also be monitored with time by recording the rate at which the water level recovers to its original static level. Recovery data provide an alternative means of estimating transmissivity and storativity while minimizing potential errors associated with well losses and boundary effects [47]. The interpretation of recovery tests is typically based on the Theis recovery method, which assumes that drawdown and recovery curves mirror each other when plotted semi-logarithmically [34,47]. However, in fractured aquifers, recovery trends may deviate from conventional porous media behavior due to delayed water release from fractures and the surrounding matrix [39,40,48,49]. Recovery tests are particularly useful in cases where drawdown data are ambiguous or affected by external influences, such as recharge or interference from nearby wells [49].

Tracer tests also provide complementary information on groundwater flow velocity, dispersivity, and connectivity between wells. These tests involve the injection of a tracer (e.g., sodium chloride, or fluorescein) into an injection well and monitoring its breakthrough at downstream observation or extraction wells [1,50,51]. The time taken for the tracer to appear, its peak concentration, and its tailing behavior provide valuable insights into the dominant transport mechanisms within the aquifer. The interpretation of tracer test results depends on the nature of groundwater flow within the aquifer. In porous media, solute transport is typically governed, depending on groundwater velocity, by advection and hydrodynamic dispersion [52]. However, in fractured crystalline aquifers, solute movement is often dominated by advection through high-permeability fractures and diffusion into the surrounding low-permeability matrix [53,54]. The Peclet number (Pe), which describes the relative importance of advection versus dispersion, helps characterize the transport regime within the aquifer. A high Pe value suggests that advection dominates, whereas a low Pe value indicates significant diffusive mixing [55]. The Pe can particularly be used to assess pronounced skewness of non-Fickian transport in fractured aquifers [56,57]. Tracer breakthrough curves (BTCs) can exhibit pronounced tailing due to matrix diffusion, where solutes temporarily migrate into low-permeability zones before re-entering the main flow path [58,59,60]. The shape of BTCs also reveals the degree of fracture connectivity; well-connected fractures produce sharp breakthrough peaks, while poorly connected systems exhibit delayed and dispersed arrival times [61].

Despite the established importance of pumping and tracer tests in groundwater characterization, a few studies have applied both techniques to systematically evaluate the variation in hydraulic properties and connectivity of crystalline basement aquifers in Southwestern Nigeria. Where pumping and tracer tests were reported, the tests were conducted separately at different locations limiting their integrated use for estimating the flow and transport properties of aquifers in Southwestern Nigeria. This study aims to estimate key hydraulic parameters (e.g., hydraulic conductivity and storativity) using constant-rate and step-drawdown pumping tests. Fracture connectivity and groundwater flow velocity are assessed using pumping and a salt tracer test. The transport behavior of solutes within the crystalline basement aquifer is evaluated using the tracer breakthrough curves. The hydraulic test results are then compared with existing conceptual models of fractured basement aquifer flow to improve the understanding of groundwater behavior in similar geological settings. By integrating pumping and tracer test data, this study provides a comparative assessment of groundwater flow and transport processes in a crystalline basement aquifer. Compared to earlier studies, the combined use of multiple pumping and a tracer test in the same aquifer system provide estimates of both flow and transport properties of the aquifer and an opportunity to assess connectivity within the aquifer. Knowledge gained from this assessment would be important for sustainable water resource management. Results of this study will inform hydrogeological modeling and support decision-making for groundwater development in fractured rock environments.

2. Study Site

This study was conducted at the Ibadan Hydrogeophysics Research Site (IHRS) located within the University of Ibadan campus in Oyo State, Nigeria. The site serves as a field research laboratory for advancing the understanding of hydrological processes within crystalline basement aquifers in parts of the country with a similar lithology and structure and as a guide for studies in regions where the crystalline basement aquifer has a significantly different architecture, e.g., a thicker weathered zone and a deeper active fractured zone. The study area is part of Southwestern Nigeria’s Precambrian basement complex, which is dominated by augen gneiss, granite gneiss, banded gneiss, quartzites, and migmatites with the intrusion of pegmatite and dolerite dyke, covered by weathered superficial sediments [62,63,64,65]. Major faults trending northwest and southeast have been mapped in the area [66], with multiple fractures oriented in the northeast and southwest serving as major sources of groundwater supply [67]. The study area covers approximately 15,000 m² in size, within coordinates 7°28′24″ to 7°28′26″ Easting and 3°53′52″ to 3°53′54″ Northing and elevation ranging from 205 m to 212 m above mean sea level (Figure 1). The site is bounded to the north by a perennial stream that flows NW-SE, to the east by a road, and to the south and west by a cultivated farmland [4,68].

The site was established in 2019 with four bedrock groundwater wells drilled down to 30 m. Lithology logs, prior geological observations, and electrical resistivity tomography profiles of the site indicate that the hydrogeological units consist of an overburden made up of clayey sand and clay, as shown in the lithological section of the wells in Figure 2. The clay-rich top layer is underlain by the weathered basement rocks, the fractured rocks, and lastly, the fresh basement rocks [68,69].

3. Materials and Methods

3.1. Constant-Rate Pumping Test

Constant rate pumping tests have been conducted in several aquifer characterization projects that give a step-by-step guide to conducting such tests [70,71,72,73]. Although theoretically similar, pumping tests differ from interference tests commonly used in reservoir engineering to access hydraulic interaction between multiple wells [74]. While pumping can be conducted in single or multiple wells in an interference test, the pumping test involves pumping in a single well with drawdown monitored in the pumping and nearby observation wells [75]. For this study, we conducted constant-rate pumping tests at all four wells at the site by first pumping at well A (Figure 1) at a constant rate of 0.90 L/s while simultaneously monitoring groundwater level changes (drawdown) in the pumping well (well A) and the other wells which served as observation wells (wells B, C, and D). We then repeated the test by pumping wells B, C, and D, respectively, while measuring drawdown in the pumping and observation wells in each case. A submersible pump installed at a depth of ~25 m was used for pumping, while the observation and pumping wells were equipped with automated CTD Diver data loggers (VanEssen Instruments, Tucker, GA, USA) to take temperature, electrical conductivity, and water level measurements every 30 s. A flow meter was installed to monitor the extraction rate every 10 min, and the mean flow rate of 0.90 l/s was used for data processing. Manual water level measurement was also performed using a battery-powered water level meter, beginning with a thirty-second interval for the first hour and progressing to a one-minute interval for the next hour, two minutes for the next hour, and finally, every five minutes until the exercise was completed. Two cycles of pumping tests were conducted for all four wells, with the first lasting for 9 h and the second lasting for 12 h, giving a total of 8 constant rate pumping tests. During each test, the extracted water was discharged far away from the research site via a network of pipes to avoid reabsorption into the pumped aquifer. To ensure that the aquifer had fully recovered before carrying out successive tests, we allowed a break of at least 72 h before successive pumping exercise. Two days before conducting the pumping test, all four wells were also developed to ensure a good connection to the aquifer.

Time series of hydraulic head measurements during each pumping test were pressure-compensated to account for the effects of changing atmospheric pressure during the test using barometric pressure measured simultaneously for the same period using a Van Essen CTD Diver baro-logger. Changes in the hydraulic head with respect to the starting hydraulic head were used to estimate drawdown. Linear, semi-log, and log-log plots of drawdown with time were used to assess the aquifer’s water level response to pumping. Analyzing the pumping test drawdown versus time data required fitting an appropriate analytical solution to the data to estimate the aquifer hydraulic properties. To choose an appropriate model for the data, we first compared the measured drawdown data to established hydrogeologic signatures represented in the site [33,72]. We also plotted a logarithmic derivative of drawdown as a function of time and compared that with standard derivative plots [73] to assess the hydrogeological and boundary conditions at the site. Drawdowns in the pumping and monitoring wells were matched with type curves of analytical models while derivate analysis techniques developed by Bourdet et al. [76] were used to identify characteristic patterns representing different flow regimes in the aquifer which help to identify the aquifer flow regime needed for constructing a hydrodynamic conceptual model. Analysis of the derivative and diagnostic plots were used to identify boundary conditions and flow regimes, such as the well-bore storage effect, fracture dewatering effect, double-porosity, and radial flow diagnosis [73].

We fitted semi-log and log-log plots of drawdown versus time with an analytical solution by Baker [40] for a generalized radial flow (GRF) in the fractured aquifer implemented in AQTESOLV software version v4.5 (Hydrosolve, Inc., Reston, VA, USA). The model’s best fit was used to estimate the fracture and matrix hydraulic conductivities, the fracture and matrix specific storage, and the effective dimension number (n-dimension) of the fractured aquifer [72]. With the aquifer conceptualized as an unconfined aquifer, we expect the specific yield to be much greater than the specific storage. Hence, we calculated the specific yield following Remson and Lang’s [76] approach by comparing the volume of dewatered material in the cone of depression with the total volume of water discharged during the pumping exercise assuming a porous aquifer framework. The specific yield is given as follows:

$$SpecificYield={\displaystyle \frac{Qt}{7.48V}}$$

(1)

where,

$$LogV= Log {\displaystyle \frac{ Q{r}^2}{4T}}+{\displaystyle \frac{5.45Ts}{Q}}$$

(2)

T is transmissivity (product of hydraulic conductivity, K, and the aquifer thickness, b), r is the horizontal distance from the axis of the pumped well to a point on the cone of depression, s is the drawdown at distance r, Q is discharge, and V is the volume of dewatered material.

Based on the groundwater discharge during the pumping tests in each of the wells, we computed the sustainable yield using the flow characteristics method (FC-Method) presented by van Tonder et al. [73]. The sustainable yield describes the discharge rate that will not cause the water level in the borehole to drop below a prescribed limit (usually the position of a major water strike), to avoid the borehole drying up [77,78,79].

3.2. Step Drawdown Tests

During this study, we also conducted two step-drawdown tests in wells C and D (Figure 1) to assess the well performance, including well loss, well efficiency, well bore skin factor, and effective well radius in the fractured aquifer system. Pumping at each well started with an extraction rate of 0.35 L/s for the first three hours, which increased to 0.65 L/s between 3 to 6 h of extraction, and lastly to 0.88 L/s for the final six hours following recommendations by Kruseman and De Ridder [33]. A gauge valve was installed at the extraction well to control the flow rate, which was monitored using a flow meter. Data loggers in the pumping and monitoring wells with similar settings as used for the constant rate pumping tests were used for the step drawdown tests.

Similar compensation and processing steps were followed as that of the constant-rate pumping data processing. The data was further analyzed using the Hantush and Jacob [80] model implemented in AQTESOLV software v4.5. Hantush and Jacob [80] constructed a rigorous mathematical model of the transient flow of water at different constant pumping rates in a well in an infinitely large homogeneous, isotropic, and leaky-constrained aquifer.

3.3. Recovery Tests

The aquifer’s recovery (i.e., the increase in water level) after pumping was monitored for all the pumping test scenarios. Since the recovery data were less noisy than pumping data due to fewer external perturbations during data acquisition, it helps to better understand aquifer connectivity and transport rate, both of which are critical in understanding aquifer sustainability [71]. Following each pumping test, the aquifer’s recovery was monitored for 2 to 14 days using the data loggers used for the pumping tests. Our goal was to achieve full recovery and ensure that the aquifer returned to its stable ambient condition before successive pumping tests. The hydraulic head recovery (termed residual drawdown) versus time data were plotted to show the water return relationship in all the wells. The recovery data was corrected for barometric influence and analyzed using the Agawal method, which is a modified Theis method implemented in the AQTESOLV software v4.5 [81].

3.4. Tracer Test

We conducted a convergent forced gradient flow field tracer test with continuous water extraction at well A and tracer injection at well C (Figure 1) to assess the connection between the wells and determine the groundwater flow velocity, average porosity, and dispersivity. The tracer solution was prepared in an inflatable pool by mixing 40 kg of table salt (NaCl) with 3240 L (3.24 m³) of water with a baseline electrical conductivity of 0.52 mS/cm. The baseline value observed at this site is consistent with values measured around the Ibadan metropolis of Southwestern Nigeria. After mixing, the electrical conductivity of the tracer was 28 mS/cm. Prior to the tracer injection, we extracted water from well A (extraction well) at 0.90 L/s for two hours using a submersible pump to create a quasi-steady state flow condition. The tracer was then injected into well C at a rate of 0.50 L/s over two hours while continuing pumping at the extraction well to maintain an undisturbed flow field. A 0.50 inch diameter hose was used to lower the tracer to a depth of 15 m while a submersible pump was installed in the injection well at a depth of 25 m to pump water from the lower section of the injection well and was re-injected at the top (at 5 m) to ensure continuous mixing of the saline tracer in the injected well. After the tracer injection, we continued pumping water at the same rate from the extraction well for 8 h and 25 min to maintain the flow field. The total duration of the tracer test (pre-injection, injection, and post-injection phases) was 12 h and 25 min.

Automated loggers were installed in all four wells at the site to monitor water level, temperature, and electrical conductivity. To monitor the tracer breakthrough, we used a Hannahn HI9811-51 conductivity meter (Hanna^® Instrument, Woonsocket, RI, USA) to continuously measure the electrical conductivity of water discharged from the extraction well. The time-series tracer data were analyzed to obtain the tracer’s first and peak arrival times, estimate the recovered tracer mass, and assess the shape of the tracer breakthrough curve. The tracer test data were further analyzed with BRGM (French Geological Survey)’s TRAC software using the radial convergent transport model to estimate the flow velocity and dispersivity [82]. An analytical solution of the classical advection-dispersion equation was fitted to the normalized concentration–time curve to estimate the tracer transport properties.

4. Results

4.1. Constant Rate Pumping Tests

Pumping from the CBA at the study site at 0.90 L/s for 9 and 12 h resulted in drawdowns ranging from 4.94 to 12.47 m and 5.43 to 12.58 m (

Table 1. Drawdown observation across all wells during the 9 h and 12 h pumping tests. Depth values are with reference to the ground surface.

Well Name	Status	Distance from Pumped Well (m)	9 h Drawdown (m)	12 h Drawdown (m)
A	Pumping	0.00	4.94	5.43
B	Observation	11.40	2.48	2.86
C	Observation	8.30	3.63	4.13
D	Observation	8.50	1.63	2.03
A	Observation	11.40	2.73	2.96
B	Pumping	0.00	12.47	12.58
C	Observation	6.85	2.68	3.01
D	Observation	15.70	1.73	2.08
A	Observation	8.30	3.52	4.12
B	Observation	6.85	2.55	3.08
C	Pumping	0.00	7.22	9.00
D	Observation	9.20	1.28	2.31
A	Observation	8.50	1.77	2.31
B	Observation	15.70	1.74	2.19
C	Observation	9.20	1.86	2.34
D	Pumping	0.00	6.18	7.30

) in the pumping wells, with the minimum drawdown observed while pumping from well A and the maximum observed while pumping from well B. Corresponding drawdowns in the observation wells ranged from 1.28 to 3.63 m during the 9 h pumping test and 2.03 to 4.13 m during the 12 h pumping test (Table 1). The minimum drawdown in the observation wells was observed in well D while pumping at well C, while the maximum drawdown was observed in well C while pumping in well A. Direct linear, semi-log and log-log plots of hydraulic head measurements with time are presented in Figure S1 in the supplementary file. Linear, semi-log, and log-log plots of drawdowns with time for the 12-h pumping tests (Figure 3, Row 1) show a sharp increase in drawdown in wells A, B, and C while pumping in well A. In contrast, well D showed a gently rising drawdown curve. At late times, the drawdown curve of wells A, B, and C tend towards a steady state while well D shows a slight continual increase visible on the linear and the log-log drawdown–time plots (Figure 3). This behavior is consistent during pumping in wells B and C as well. However, pumping in well D produced the smallest drawdowns (2.19 to 2.34 m) with the drawdown similar in all the observation wells (A, B, and C) and all the observations showing a gradual continuous decrease in drawdown with no steady state achieved.

The log-log and semi-log plots of drawdown versus time were compared with type curves for different hydrogeological structures presented by Kruseman and De Ridder [33]. The log-log plot of drawdown versus time for observation well A while pumping at D, for example, showed a slope of 0.50 at the early stage and ~0.25 at the late stage (Figure 4), similar to parts C in Figure 4 presented by Kruseman and De Ridder [33]. Similarly, the semi-log plot showed a gentle slope at an early pumping time and a steep slope at a late pumping time. This corresponds to drawdown response while pumping from a permeable dyke in an otherwise impermeable aquifer (Figure 4).

We also compared the log-log plot of the derivative of drawdown with time for wells B and D (Figure 5) with the analytical type-curve provided by van Tonder et al. [73]. The derivative plot of well B at early times follows a steep slope typical of well bore storage effects, while that of well D showed no well bore storage effect. However, the derivative drawdown plot of well D at about 400 s after the start of pumping showed a depression that corresponded to a double porosity effect. At late pumping times (i.e., ~6000 to 12,000 s after the start of pumping), well B showed successive depressions in the derivative of drawdown, which corresponds to a fracture dewatering effect in the analytical type-curve presented by van Tonder et al. [73]. Further derivative plots of drawdown with time while pumping at well B (Figure S2), pumping at well C (Figure S3), pumping at well D (Figure S4) and pumping at well A (Figure S5) are provided in the supplementary file.

Lastly, the fitted analytical solution by Baker [40] for the fractured aquifer to the log-log and semi-log plots of drawdown and its derivative with time produced a good fit for all pumping test data with slight deviations, mostly at the early times within the first 400 s of pumping (Figure 6). The fracture and matrix hydraulic conductivity estimated from the model’s best fit varied over two orders of magnitude and ranged from 1.45 × 10⁻⁵ to 7.85 × 10⁻⁷ m/s and 1.00 to 7.69 × 10⁻¹⁰ m/s (

Table 2. Estimated aquifer parameters in both 9 and 12 h pumping tests.

Wells	Hydraulic Parameters	Pumping at Well A (9 h)	Pumping at Well D (9 h)	Pumping at Well C (9 h)	Pumping at Well B (9 h)	Pumping at Well B (12 h)	Pumping at Well C (12 h)	Pumping at Well D (12 h)	Pumping at Well A (12 h)
WELL A	Fracture Hydraulic Conductivity (m/s)	6.57 × 10−6	1.23 × 10−6	2.43 × 10−6	2.97 × 10−6	2.99 × 10−6	3.22 × 10−6	6.41 × 10−7	1.00 × 10−5
	Matrix Hydraulic Conductivity (m/s)	1.09 × 10−6	1.92 × 10−9	2.43 × 10−10	3.54 × 10−10	3.40 × 10−10	2.61 × 10−10	1.00 × 10−10	1.00 × 10−10
	Fracture Specific Storage (m−1)	2.46 × 10−4	2.30 × 10−6	2.77 × 10−7	3.03 × 10−7	3.15 × 10−7	2.14 × 10−7	1.96 × 10−6	1.08 × 10−4
	Matrix Specific Storage (m−1)	5.48 × 10−2	9.75 × 10−4	7.33 × 10−5	6.88 × 10−5	6.53 × 10−5	2.42 × 10−5	4.85 × 10−3	7.02 × 10−5
	Transmissivity (m2/s)	1.01 × 10−4	1.89 × 10−5	3.71 × 10−5	4.54 × 10−5	4.56 × 10−5	4.92 × 10−5	9.79 × 10−6	1.54 × 10−4
	Flow Dimension (n)	2.06	2.22	1.97	2.00	2.00	2.00	2.00	1.78
WELL B	Fracture Hydraulic Conductivity (m/s)	3.82 × 10−6	3.64 × 10−6	2.64 × 10−6	3.53 × 10−6	3.54 × 10−6	3.17 × 10−6	7.85 × 10−7	1.66 × 10−5
	Matrix Hydraulic Conductivity (m/s)	9.19 × 10−10	4.29 × 10−8	6.95 × 10−10	7.69 × 10−10	1.00 × 10−10	7.63 × 10−10	1.00 × 10−10	4.25 × 10−2
	Fracture-Specific Storage	5.78 × 10−7	5.07 × 10−5	2.06 × 10−5	4.60 × 10−6	1.09 × 10−6	1.89 × 10−5	6.35 × 10−7	4.51 × 10−5
	Matrix Specific Storage	9.19 × 10−5	1.16 × 10−4	2.23 × 10−4	1.00 × 10−10	1.00 × 10−10	1.20 × 10−4	5.40 × 10−4	1.16 × 10−10
	Transmissivity (m2/s)	1.06 × 10−4	1.01 × 10−4	7.30 × 10−5	9.75 × 10−5	9.77 × 10−5	8.70 × 10−5	2.17 × 10−5	2.53 × 10−4
	Flow Dimension (n)	2.35	2.00	2.00	2.00	2.00	2.00	2.00	1.71
WELL C	Fracture Hydraulic Conductivity (m/s)	7.19 × 10−6	1.91 × 10−6	2.76 × 10−6	3.71 × 10−6	1.82 × 10−6	2.49 × 10−6	1.45 × 10−5	2.66 × 10−6
	Matrix Hydraulic Conductivity (m/s)	1.87 × 10−10	8.06 × 10−9	7.67 × 10−9	1.43 × 10−9	1.45 × 10−10	1.00 × 10−10	6.14 × 10−2	1.00 × 10−10
	Fracture-Specific Storage	1.16 × 10−5	1.29 × 10−5	1.90 × 10−3	2.24 × 10−5	1.70 × 10−5	2.46 × 10−3	1.95 × 10−4	3.02 × 10−6
	Matrix Specific Storage	1.30 × 10−5	4.94 × 10−4	9.31 × 10−4	7.84 × 10−5	5.65 × 10−4	1.00 × 10−10	1.00 × 10−10	1.43 × 10−4
	Transmissivity (m2/s)	1.99 × 10−4	5.30 × 10−5	7.66 × 10−5	1.03 × 10−4	5.05 × 10−5	6.91 × 10−5	4.03 × 10−4	4.06 × 10−6
	Flow Dimension (n)	2.00	2.00	2.00	2.05	2.00	2.00	0.93
WELL D	Fracture Hydraulic Conductivity (m/s)	2.63 × 10−7	5.11 × 10−7	5.59 × 10−8	5.50 × 10−6	1.36 × 10−6	4.36 × 10−7	4.80 × 10−7	1.64 × 10−5
	Matrix Hydraulic Conductivity (m/s)	3.52 × 10−10	1.00 × 10−00	1.71 × 10−10	6.70 × 10−5	1.00 × 10−10	1.13 × 10−10	4.04 × 10−1	1.42 × 10−5
	Fracture-Specific Storage	1.27 × 10−5	1.25 × 10−00	1.90 × 10−6	1.00 × 10−3	1.12 × 10−6	1.86 × 10−6	1.10 × 10−0	2.40 × 10−3
	Matrix Specific Storage	1.02 × 10−2	1.00 × 10−0	3.34 × 10−3	1.00 × 10−3	5.44 × 10−4	2.48 × 10−3	9.54 × 10−1	2.83 × 10−8
	Transmissivity (m2/s)	7.28 × 10−6	1.41 × 10−5	1.54 × 10−6	1.52 × 10−4	3.75 × 10−5	1.21 × 10−5	1.33 × 10−5	1.64 × 10−5
	Flow Dimension (n)	2.68	2.00	2.00	2.00	2.00	2.00	2.00

). The average value of fracture hydraulic conductivity for wells A, B, C, and D are 3.76, 4.71, 4.63, and 3.12 (× 10⁻⁶) m/s, while their average matrix hydraulic conductivity is 1.37 × 10⁻⁷, 5.31 × 10⁻³, 7.68 × 10⁻³, and 1.75 × 10⁻¹ m/s (Table 2).

The Baker model [40] used to estimate the hydraulic conductivity was also used to estimate the specific storage of both the fractures and matrix (Table 2). The specific storage of the fractures was not surprisingly small with values ranging from 2.14 × 10⁻⁷ to 2.46 × 10⁻³ m⁻¹ while those of the matrix range from 1.00 × 10⁻¹⁰ to 9.31 × 10⁻⁴ m⁻¹ (Table 2). We also obtained the effective dimension number of the fractured aquifer (n-dimension) with n ranging from 0.90 to 2.60 with most of the pumping tests giving values around 2 (Table 2). The range of n values show variations in the aquifer structure typical of fractured and rocks dual porosity systems (n = 1–2 or <1), homogeneous and isotropic systems (where n = ~2), and connected fractures (n = 2–3+) [40]. This further highlights the high level of heterogeneity and complex structure typical of crystalline basement aquifers. The aquifer specific yields (

Table 3. Specific yield at various distances of the monitoring well to the extraction during the 12 h pumping exercise.

Pumping Well	Extraction Well	A	B	C	D
A	Specific Yield (%)		0.16%	0.22%	0.34%
	Distance from pumping well		11.40 m	8.30 m	8.50m
	Volume of material dewatered (m3)		3426.80 m3	2427.60 m3	1575.90 m3
	Volume of water drained (L)		38.84 L	53.41 L	53.58 L
B	Specific Yield (%)	0.03%		0.08%	0.05%
	Distance from pumping well	11.40 m		6.85 m	15.70 m
	Volume of material dewatered (m3)	16,358.75 m3		6295.01 m3	9494.89 m3
	Volume of water drained (L)	50.71 L		50.36 L	47.48 L
C	Specific Yield (%)	0.04%	0.17%		0.20%
	Distance from pumping well	8.30 m	6.85 m		9.20 m
	Volume of material dewatered (m3)	12,448.40 m3	3075.70 m3		2595.83 m3
	Volume of water drained (L)	49.79 L	52.29 L		51.93 L
D	Specific Yield (%)	0.28%	0.08%	0.24%
	Distance from pumping well	8.50 m	15.70 m	9.20 m
	Volume of material dewatered (m3)	1906.70 m3	6372.93 m3	2247.88 m3
	Volume of water drained (L)	53.38 L	50.98 L	53.94 L

) were highly variable, as typical of fractured aquifers with values ranging from 0.03 to 0.34 [-], while the sustainable yields (

Table 4. Estimated sustainable yield of each well using FC-Method.

Well	Sustainable Yield	Number of Persons That Can Be Supplied
A	1166.40 m3 at 0.45 L/s	1555 People
B	414.72 m3 at 0.16 L/s	553 People
C	803.52 m3 at 0.31 L/s	1071 People
D	984.96 m3 at 0.38 L/s	1313 People

) ranged from 415 m³ when pumping at 0.00016 m³/s to 1.17 m³ when pumping at 0.00045 m³/s.

4.2. Step-Drawdown Pumping Test

The step drawdown test while pumping at well C produced drawdowns ranging from 0.40 to 2.30 m with pumping at a rate of 0.35 L/s, 0.90 to 5.10 m at the second-rate step of 0. 65 L/s, and 1.80 to 8.05 m at the last pumping rate of 0.90 L/s (Figure 7). Large drawdowns (>2.80 m) were observed in the pumping well (well C) and observation wells A and B. Well D, however, showed very low drawdown (<1.80 m), with a nearly linear increasing drawdown that hardly reflects the three steps changes in the pumping rate. In contrast, the step drawdown test at well D only showed large drawdowns in the pumping well (~1.20 to 6.05 m), reflecting the stepwise changes in the pumping rate while low drawdown (<2.10 m), with an approximately linear increase that hardly reflects the step changes in pumping rates were observed in wells A, B, and C (Figure 7).

The semi-log and log-log plots of the step drawdown test with pumping at well C fitted with the Hantush–Jacob analytical solution [80] show a good fit with mismatch at the earlier pumping time (Figure 8). Transmissivity and storage coefficients estimated from the model fit range from 7.05 × 10⁻⁵ to 1.17 × 10⁻⁴ m/s and 0.01 to 0.12 [-], respectively while the non-linear well efficiency factor ranged from 1.50 to 2.00, consistent with the range of 1.50 to 3.50 suggested by Rorabaugh [83]. The well efficiency exceeded 92%.

4.3. Recovery Test

Recovery into the aquifer began at the end of the extraction phase, albeit the duration varies across each well, indicating the effect of heterogeneity in the aquifer system. As with the pumping data, the residual drawdowns (recovery) for the wells are shown in linear, semi-log, and log-log forms during the twelve-hour pumping exercises for pumping at wells A and D (Figure 9). The hydraulic head recovery pattern is comparable to the pumping induced drawdown, with an initial high rate of recovery that peaks early and then declines with recovery length. During the first 600 s of recovery, there was a constant recovery rate, followed by a rapid decline until recovery stopped.

4.4. Tracer Test

The tracer test result shows a first arrival time after 27 min at the extraction well (well A) 8.30 m away from the injection well (Well C) with the electrical conductivity changing from the background value of 0.55 mS/cm to 0.58 mS/cm at an extraction rate of 0.90 L/s (Figure 10). The tracer concentration continued to increase until peak arrival time after 123 min with electrical conductivity at a peak concentration of 2.66 mS/m. The measured concentration at the observation well decreased thereafter with continued decline in measured electrical conductivity showing a tailing behavior. We stopped monitoring after 610 min (at 7:00 PM West African Time) with the EC dropping to 1.58 mS/cm.

A fit of the tracer concentration versus time plot with an analytical solution of the advection-dispersion equation produced a good fit for the early to peak tracer breakthrough times but showed deviation at late times when tailing was observed (Figure 10). Mean flow velocity, longitudinal dispersivity, and porosity were estimated at 5.80 × 10⁻⁴ m/s, 2.60 m, and 0.01, respectively. About 95% of the injected salt tracer mass was recovered during the monitoring duration.

5. Discussion

5.1. Aquifer Hydraulic Properties

The hydraulic properties of CBAs are primarily controlled by the presence of secondary porosity and permeability, which arises from fracturing and weathering processes [84]. Unlike sedimentary aquifers, which may show well to poorly sorted pore spaces, basement aquifers exhibit significant heterogeneity due to the variability in fracture density, aperture, and connectivity. This makes significant variation in estimated hydraulic conductivity in CBAs highly typical. The estimated hydraulic conductivity values (Table 2) from this study indicate a highly variable permeability structure, with higher values (~1.45 × 10⁻⁵ m/s) likely associated with well-developed fracture networks and lower values (~7.85 × 10⁻⁷ m/s) corresponding to less fractured zones. These findings align with theoretical models that suggest fluid flow in crystalline aquifers is predominantly fracture-controlled, with primary porosity playing a negligible role [14,85].

Comparing these results to previous studies, the observed hydraulic conductivity variations further emphasize the dependence of groundwater movement on the spatial distribution of interconnected fractures, similar to findings by Taylor and Howard [86] in crystalline basement aquifers in South Africa. However, this study contradicts reports by Acheampong and Hess [87] that suggested relatively uniform hydraulic conductivity in some weathered basement terrains. The variation in estimated hydraulic conductivity over two orders of magnitude (7.85 × 10⁻⁷ m/s–1.45 × 10⁻⁵ m/s) within an area < 400 m² has an implication for the exploration and management of the aquifer system in the region. This shows that groundwater flow is highly non-uniform with high volumetric flux in an area with high hydraulic conductivity. Groundwater in this area would follow discontinuous and anisotropic flow paths. This implies that within a small area, groundwater availability and flow would vary significantly. Such non-uniform flow field could limit pumping and remediation in cases of contamination with recharge and potential pollutant migration being highly non-uniform. Localized high hydraulic conductivity areas would yield a significant amount of water, and hence, are suitable for productive boreholes while adjacent areas with low hydraulic conductivity may lead to boreholes that dry out over time. This is consistent with dry boreholes observed within the Ibadan metropolis in Nigeria. Hence, we recommend site-specific geophysical and hydrogeological characterization prior to drilling groundwater wells as boreholes drilled a few meters apart could produce significantly different results.

Also, the upper (1.45 × 10⁻⁵ m/s) and lower (7.85 × 10⁻⁷ m/s) bounds of estimated hydraulic conductivity at the site could sustain groundwater wells with moderate and marginal to unproductive wells. This further justifies the need for detailed site-specific investigations and performance cannot be generalized from one location to surrounding areas. We recommend using these results to guide extraction rates from these aquifers to avoid over-exploitation in areas with low hydraulic conductivity. The storativity values observed in this study also suggest a relatively limited storage capacity, reinforcing the need for careful water balance assessments to ensure sustainable groundwater use.

5.2. Aquifer Connectivity

The extent of hydraulic connectivity within crystalline basement aquifers is largely governed by the degree of fracture interconnectivity and the presence of weathered zones [14]. The hydraulic gradient analysis at the studied site suggests differential groundwater movement, likely influenced by variations in fracture permeability and structural controls, such as faults and joint sets. The results indicate that some areas (e.g., around wells A, B, and C) exhibit high hydraulic connectivity, facilitating relatively rapid groundwater movement, whereas others (e.g., around well D) show restricted flow due to isolated or poorly connected fractures. These findings are consistent with conceptual models that describe basement aquifers as compartmentalized systems where water movement is constrained by discontinuities in the fracture network [4,14].

The results of the tracer test provided additional insights into the connectivity within the aquifers at IHRS. The sodium chloride (NaCl) tracer injected at well C (40 kg total mass) was observed at Well A (8.30 m from the injection well) with the first tracer arrival time recorded in about 27 min after injection, and the groundwater electrical conductivity (a proxy for salinity) increasing from an initial 0.55 mS/cm to a peak of 2.66 mS/cm at 2 h post-injection before steadily declining. This rapid tracer breakthrough confirms a well-connected flow path between wells C and A. The retrieved mass of 38 kg (95% recovery) highlights a minimal loss to matrix diffusion or dead-end fractures. Calculated parameters, such as flow velocity, longitudinal dispersivity, and kinematic porosity of 5.80 × 10⁻⁴ m/s, 2.60 m and 0.01, further support these observations, while the Peclet number (Pe) of 3.25 indicates an advective flow with mechanical dispersion dominating molecular diffusion, reinforcing the fracture-dominated flow regime [88,89].

These findings align with previous studies, such as Lapworth et al. [90], which documented preferential flow pathways in crystalline basement aquifers but contrast with results from Wright [11], who suggested more homogeneous fracture connectivity in similar settings. The implications of these findings are critical for groundwater management, as localized high connectivity zones may allow for efficient water extraction, while poorly connected regions may require alternative strategies, such as clustering groundwater extraction wells or artificial recharge interventions.

5.3. Well Yield and Sustainable Aquifer Use

Well yield in crystalline basement aquifers is primarily a function of fracture aperture, density, and interconnectivity, making it highly variable across different locations. The results of well yield estimates demonstrate that high-yielding wells are situated in zones with enhanced fracture permeability, while lower-yielding wells correspond to sparsely fractured or unfractured bedrock. These findings align with the work of Lachassagne et al. [2] and Neuman [14], which identified high productivity wells in regions of dense fracture networks. This study reveals pronounced spatial variability in well yield, highlighting the importance of localized hydrogeological assessments.

Over-extraction from productive fracture zones can lead to rapid drawdowns and localized dewatering, emphasizing the need for a controlled pumping strategy. Given the relatively low storage capacity of basement aquifers, excessive withdrawal without adequate recharge can cause long-term declines in groundwater levels, leading to potential aquifer depletion. Results of calculated sustainable yield in this study (Table 4) show that wells within the crystalline basement aquifers in Southwestern Nigeria, where this study was conducted, could sustainably provide water for up to 1,550 people but with controlled low pumping rates of 0.45 L/s. The pumping rate could be increased by drilling larger diameter wells.

The results of this study show that crystalline basement aquifers have low to medium yield. This makes them susceptible to drying up particularly with over pumping or during extended dry periods when recharge is limited. To manage these aquifers sustainably, a managed aquifer recharge program, including the utilization of surface water infiltration and artificial recharge techniques, is recommended. Furthermore, future climate variability could impact recharge dynamics by altering precipitation patterns and surface water-groundwater interactions, necessitating adaptive management strategies [8,84]. The findings highlight the critical role of hydrogeological characterization in optimizing well placement and ensuring the long-term viability of groundwater resources in crystalline basement settings.

6. Conclusions

This study provided a detailed characterization of the hydraulic properties, connectivity, and sustainable yield of a crystalline basement aquifer in Southwestern Nigeria, offering valuable insights for groundwater resource management for the over 55 million people in the region who depend on it. The results confirmed that groundwater flow in the crystalline basement aquifers is fracture-dominated, with permeability and storativity strongly influenced by the distribution and connectivity of fractures. Areas with high fracture intensity show a corresponding increase in groundwater storage and flow. The variability in hydraulic conductivity and transmissivity highlights the inherent heterogeneity of crystalline basement aquifers, which directly impacts water well productivity and groundwater availability.

A combination of pumping and tracer tests was used in this study to assess groundwater flow and potential solute transport within the crystalline basement aquifer. The aquifer connectivity analysis using measured drawdown from the pumping tests and solute breakthrough from the tracer test revealed a complex interplay of structural controls that influence groundwater movement, with some areas exhibiting well-integrated flow networks while others remain hydraulically isolated. This study confirms previous findings on the importance of structural influences in basement aquifers while highlighting the need for improved characterization techniques to resolve aquifer complexity. Sustainable aquifer use in crystalline basement settings, such as in Southwestern Nigeria, would require a balanced approach that incorporates groundwater recharge, controlled extraction, and continuous monitoring to prevent resource depletion.

The implications of these findings extend to regional water resource management, particularly in Southwestern Nigeria, where better-informed drilling practices and sustainable groundwater abstraction strategies can be implemented. Future research should focus on integrating long-term hydrogeological data with numerical modeling to improve predictions of groundwater behavior under changing climatic and anthropogenic influences. The results of this study serve as a foundation for developing targeted management strategies that can enhance the resilience and sustainable use of crystalline basement aquifer as sources of water supply, particularly in Southwestern Nigeria.