Hydrologically, Cotonou (experiences) is subjected to a sub-equatorial climate with four seasons and an average annual rainfall of approximately 1300 mm [1]. Situated between the Atlantic Ocean and Lake Nokoué, the city is subject to hydrological exchanges that facilitate the recharge of shallow sandy formations through precipitation and seasonal flooding.

Hydrogeologically, the site is underlain by the unconfined Quaternary aquifer of the coastal sandbar. Piezometric monitoring conducted in March 2025 revealed a stable water table at an average depth of 2.04 m. A Lefranc test performed at a depth of 2 to 4 m measured a permeability of K = 2.429 × 10⁻⁵ m/s, characteristic of permeable fine-to-medium sand [2]. The aquifer, recharged by precipitation and connected to both Lake Nokoué and the Atlantic Ocean, exhibits seasonal fluctuations ranging from 0.5 to 2.5 m. Its shallow depth keeps the surface sandy layers saturated, thereby increasing their susceptibility to liquefaction. However, the presence of potentially liquefiable soil is not sufficient to define a significant risk, as this also depends on the intensity of seismic loading and soil’s resistance [3].

The site is located within the Beninese coastal sedimentary basin [4], which extends along the Gulf of Guinea and is the result of continuous sedimentation from the Late Cretaceous to the present day. According to the 1:200,000 scale geological map of Benin [5], the surface geology of the Cotonou area consists primarily of clayey sands and sandy clays of deltaic and lagoonal origin, overlain by recent Quaternary deposits forming the coastal barrier; these deposits compressed between marine and aeolian sands reworked by coastal dynamics. Core samples taken at the site confirm this structure: the observed lithology is dominated by clean sand at the surface, followed by silty sands, sandy silts, and finally clayey to highly clayey sands with increasing depth, down to 25 m. This sequence, typical of West African coastal barriers, reflects a gradual transition from an open marine depositional environment to appease, lagoonal conditions, and explains the alternation between loose sandy layers and finer, denser layers observed.

Liquefaction is a geotechnical phenomenon in which saturated, loose, non-cohesive soil temporarily loses its shear strength under cyclic loading due to an increase in pore water pressure. Its assessment relies on the simplified method developed by Seed and Idriss [6] and updated by the NCEER/NSF [7], which compares cyclic stress (CSR) with the soil’s cyclic resistance (CRR). Initially based on the SPT, this approach was extended to the CPT and subsequently to the Ménard pressuremeter test, thanks to the work of [8] and [9] introduced the Liquefaction Potential Index (LPI), which incorporates the entire geotechnical profile. Although these methods have been successfully applied in Beyrouth [10] and Tunisia, they remain poorly documented in West Africa, where pressuremeter tests often constitute the primary source of geotechnical data.

West Africa is a region of low to moderate seismicity, lacking any major active faults [11]. The most significant documented earthquake remains the one that struck Accra, Ghana, on December 22, 1939, with a magnitude of 6.4 [12]. In Benin, historical earthquakes are rare and generally have magnitudes below 5, justifying its classification as a seismic zone 2 area (a_gR = 0.06 g). Consequently, the low regional seismic hazard is the primary reason for the “no-liquefaction” result obtained.

The geotechnical investigation was carried out within the SImAU building footprint at the Cotonou Financial City site. Figure 2 presents the location of the main investigation points, including the pressuremeter and cored boreholes, providing an overview of their spatial distribution within the study area.

Figure 2. Aerial view of the SImAU site showing the location of the geotechnical investigation points.

The geotechnical campaign utilized GPS for positioning boreholes, a piezometer for monitoring the water table, a Ménard pressuremeter for the mechanical characterization of soils, and a core drilling rig for lithological assessment and sample collection. The Ménard pressuremeter enables the determination of the limit pressure and the pressuremeter modulus parameters essential for evaluating liquefaction potential. The investigation comprised two Ménard pressuremeter boreholes (SP3 and SP4) and two cored boreholes (SC3 and SC4), all drilled to 25 m depth. Pressuremeter tests were performed at 1.00 m intervals throughout each borehole, yielding 25 test levels per sounding. The boreholes are located within the SImAU building footprint at the Cotonou Financial City site. The type of ressurometer used is presented in Figure 3.

Figure 3. Schematic diagram of the Ménard pressuremeter test and equipment.

The core drilling system that is shown in Figure 4 provides information on the lithological succession of the formations penetrated, as well as representative samples for soil identification and laboratory testing. The combined use of these two pieces of equipment ensures a comprehensive geotechnical characterization of the site by linking the mechanical properties of the ground to its geological nature.

3.2.1. Determination of the Seismic Coefficient

NF EN 1998-5 [13] governs the assessment of liquefaction. The conditions for exemption are not met due to the presence of clean sands with a low fines content (<2%), loose layers, and a groundwater table located at a depth of 2.04 m. An assessment of liquefaction potential is therefore required for the site under study. The seismic coefficient is determined using the method specified in NF EN 1998-5, in accordance with the following equation:

With r = 1 (saturated, cohesionless soils)

The design magnitude adopted (Mw = 5.1) corresponds to the maximum

Figure 4. Diagram of the core sampling principle.

historical earthquake recorded in Benin at Nikki on August 3, 2014. This choice provides an initial assessment consistent with available data but does not replace a probabilistic study of local seismic hazard. The following input values were used: total and effective vertical stresses ( σ v , σ ′ v ) computed from measured soil unit weights (γ_h = 17.14 - 19.00 kN/m³) and groundwater table depth hw = 2.04 m; depth-reduction factor rd after Idriss and Boulanger (2004); reference ground acceleration a_gR = 0.06 g (Poggi et al. [14], seismic Zone 2, Benin); soil amplification factor S = 1.8 (site class D, Eurocode 8); importance factor γ_I = 1.0 (ordinary structure, class II); yielding ag·S = γ_I × a_gR × S = 0.108 g.

where P a = 100 kPa.

Le CRR (7.5) is given by:

The magnitude correction is introduced by the factor:

We have: MSF = 1.8 pour Mw = 5.1

The factor of safety is then calculated using the relationship:

This method allows for the direct use of pressuremeter results without requiring a standard penetration test performed on-site.

3.2.2. PMT-SPT Correlation

The PMT-SPT correlation method, proposed by [15], is based on the indirect estimation of Standard Penetration Test (SPT) resistance from pressuremeter parameters. In the absence of SPT test results for the site under study, this approach allows the pressuremeter modulus to be converted into an equivalent standard penetration index (N₆₀) using the following empirical relationship:

The overload correction from [16]

For N 1 ( 60 ) < 30 the CRR (7.5) is calculated using the following formula:

When N 1 ( 60 ) ≥ 30 , the soil is directly classified as non-liquefiable within the framework of the method in question. The factor of safety is then obtained by:

This approach provides an indirect solution for utilizing pressuremeter data in the absence of SPT tests. However, it must be interpreted with caution, as the relationship between pressuremeter parameters and the SPT index can depend on the soil type, its density, and the geological context.

3.2.3. Determination of the Liquefaction Potential Index

The liquefaction potential of the profile is summarized using the Liquefaction Potential Index (LPI). This index integrates the contribution of each layer weighted by its proximity to the surface over a reference depth of 20 m:

where

The function F(z) represents the degree of susceptibility to liquefaction, while W(z) assigns greater weight to horizons close to the surface.

4.1.1. Factors of Safety

Table 1 shows that both methods yield consistent conclusions. All layers exhibit a factor of safety greater than 1. Layer C5 is the most critical (FS = 2.80), whereas layers C2 and C4 fall outside the liquefiable range.

The value of the factor of safety per layer for each method is shown in Table 2.

Table 2. Factor of safety values per layer according to each method.

It should be noted that both methods yield consistent results. All layers exhibit a factor of safety greater than 1, indicating no risk of liquefaction. Layer 5 is the most unfavorable.

All the layers exhibit FS > 1 for both methods. Consequently, the severity function F(z) = max(1 − FS, 0) equals zero for every layer throughout the investigated profile. The depth-weighted, profile-wide Liquefaction Potential Index (LPI), integrated from the surface to z = 20 m, is therefore LPI = 0, regardless of the integration bound selected (10 m or 20 m), since no layer contributes to the summation. An LPI of zero places the site in the lowest risk class (“very low to nil”) of the Iwasaki et al. [17] classification.

4.1.2. Determination of the Critical Magnitude

Table 3 presents an isolated parametric sensitivity analysis for the most vulnerable

Figure 5. Position of the layers relative to the liquefaction limit curve.

layer (C5), obtained by varying Mw while keeping ag·S = 0.108 g constant. This is a methodological exercise only: the seismic coefficient is not updated consistently with each magnitude value, as no regional ground-motion prediction equation is available for Benin. The sole purpose is to identify the parametric threshold at which FS = 1 under the fixed design acceleration. These Mw values do not represent physically consistent earthquake scenarios and must not be interpreted as design-risk references.

The liquefaction zones of the soil tested are shown in Figure 5.

Table 3. Parametric sensitivity of FS (layer C5, most vulnerable) to varying Mw at fixed ag·S = 0.108 g isolated methodological exercise, not a physically consistent hazard scenario.

Under the fixed ag·S = 0.108 g condition, layer C5 remains above the liquefaction threshold (FS > 1) for all magnitudes up to Mw ≈ 9.1. Beyond this parametric threshold, magnitudes above Mw = 9.1 have never been recorded on Earth and are entirely outside the seismotectonic context of the West African passive margin (historical maximum: Mw = 6.4 in Ghana, 1939). These results are presented solely to characterize the mathematical behavior of the method and carry no design significance.

The results obtained differ from those reported for the Beirut coastline, where safety factors close to or below 1 were calculated for several sand layers due to higher regional seismic hazard. Conversely, the study by [18], conducted in a Tunisian coastal sedimentary basin subject to moderate seismicity comparable to that of Benin, yields consistent conclusions, with safety factors exceeding 2 and a low to very low liquefaction risk. This comparison demonstrates that the level of regional seismic hazard rather than soil type accounts for the differences observed between the studied contexts. From a practical standpoint, this study represents one of the first documented applications of pressuremeter-based liquefaction assessment methods in Benin, where SPT and CPT tests are rarely available. The convergence of the two methods used, along with their consistency with results from comparable coastal basins, confirms the suitability of this approach for the West African context and provides a reproducible methodological framework for the seismic design of foundations along the Beninese coast.

However, three main limitations should be noted: the absence of SPT and CPT tests necessitates the use of correlations not calibrated for the coastal sands of the Gulf of Guinea; the selected magnitude is based on the maximum historical value rather than a probabilistic study of local seismic hazard; and finally, seasonal groundwater fluctuations and potential sand disturbance remain factors requiring monitoring. A supplementary campaign of CPTu and SPT tests would help confirm the results, particularly for Layer 5, which was identified as the most vulnerable. Although the risk of liquefaction is ruled out, foundations should be anchored in the most resistant layers (Layer 2 or 4), while local seismic microzoning and piezometric monitoring would usefully complement this assessment.