Susceptibility to Liquefaction of Iron Ore Tailings in Upstream Dams Considering Drainage Conditions Based on Seismic Piezocone Tests

Abstract

One of the critical challenges facing the mining sector is related to the prevention and mitigation of catastrophic incidents associated with its tailing dams. As mining tailings are very heterogeneous and field characterization is expensive and complex, geotechnical properties of these materials are largely unknown. The seismic cone penetration test (SCPTu) provides a field approach to estimate a large array of geotechnical information, including the liquefaction potential of tailing dams. Yet, the exploration of strain softening behaviors in geomaterials under undrained loading, utilizing the state parameter ($\psi$) inferred from SCPTu tests initially applied to soft soils, has been often used for mining tailings. This study is concerned with the implementation of a tailing classification system which uses the ratio between the small strain shear modulus and the cone tip resistance (${\mathrm{G}}_0/{\mathrm{q}}_{\mathrm{t}}$). A series of laboratory tests was executed, and three different methodologies were adopted to assess the effects of (partial) drainage conditions based on 531.26 m of SCPTu measurements conducted at three different upstream iron ore tailing dams in Brazil. Furthermore, the ${\mathrm{G}}_0/{\mathrm{q}}_{\mathrm{t}}$ ratio is integrated with $\psi$ to assess the liquefaction tendencies of the investigated materials. The findings reveal the heterogeneous nature of the tailings, wherein indications of partial drainage are discernible across numerous records. Liquefaction analyses demonstrate that the tailings exhibit a contractive behavior in over 94% of the SCPTu data, confirming their susceptibility to flow liquefaction. Our findings are relevant for site characterization within iron ore tailing dams and other mining sites with similar geotechnical attributes.

1. Introduction

The mining sector holds paramount relevance and makes a substantial economic contribution in Brazil, positioning the country as the second-largest exporter of iron ore in 2022 [1]. The state of Minas Gerais (southeastern Brazil) hosts the most extensive iron ore deposits nationwide. Furthermore, this state is home to 37% of all of Brazil’s dams, and among them, 33% were constructed with upstream methods after their initial embankment [2].

Recently, the world has been left impressed by two catastrophic incidents involving upstream dams in Brazil. These events, known as the Fundão disaster [3,4] and the Brumadinho disaster [5], occurred within the state of Minas Gerais. Unfortunately, this state continues to deal with the precarious conditions of 56 structures at risk [6]. The studies by Morgenstern et al. [7] and Robertson et al. [8] indicated that the failures of these two tailing dams were triggered by flow liquefaction. Recent literature addressing tailing dam liquefaction further supports this perspective, emphasizing the large risks posed to both human safety and the environment [9,10,11]. Consequently, conducting thorough geotechnical investigations on tailing dams assumes critical importance.

Laboratory tests can be employed for the characterization of tailing dams [12,13]. However, tailing properties obtained from laboratory tests conducted on limited soil samples are typically unsuitable for effectively characterizing the in situ conditions at field scales. Moreover, laboratory tests are time-consuming, while much effort is required to collect undisturbed samples [14,15]. As a result, most research relies on field experiments such as the seismic cone penetration test (SCPTu), which offers a combination of cone penetration test (CPT) data and shear wave velocity (V_s) profiles [16], enabling one to analyze the behavior of several soil types and evaluate their physical and mechanical properties [17]. This approach offers the advantage of enabling the estimation of multiple geotechnical parameters through measurements of key variables, including the cone tip resistance (${\mathrm{q}}_{\mathrm{t}}$), lateral friction (${\mathrm{f}}_{\mathrm{s}}$), pore pressure (u₂), and V_s. For instance, a useful geotechnical parameter derived from SCPTu tests is the small strain shear modulus (${\mathrm{G}}_0$). The obtained V_s results facilitate geotechnical classifications and establish several valuable correlations [18,19]. Furthermore, results from SCPTu tests can be used to assess liquefaction susceptibility in soils and tailings [20,21,22].

In contrast to conventional soils, mining tailings are of anthropogenic origin, exhibiting diverse characteristics influenced by mineral processing and by the tailing disposal method. This variety of factors contributes to the inherent heterogeneity of the material. Consequently, the interpretation of in situ tests, for instance, can be challenging due to the potential variation in drainage conditions throughout the depth of a containment structure. Classical soil mechanics offers established equations for deriving parameters of materials exhibiting either fully drained or undrained behavior under load [23]. Materials referred to as “transitional” possess intermediate permeability, necessitating ongoing investigations and monitoring [24]. This allows the enhancement of field test outcomes through the adjustment of the derived parameters. Several studies indicate that mining tailings are silty materials and present an intermediate permeability (10⁻⁵ m/s < k < 10⁻⁸ m/s) in which standard cone penetration tests (v = 20 mm/s) are affected by partial drainage during penetration. This may induce errors in the prediction of soil parameters [25,26,27].

In a study by Nierwinski, Schnaid and Odebrecht [28,29,30], a classification methodology using the G_o/q_t ratio was introduced for granular materials subject to liquefaction, along with drainage assessment. The authors highlighted that the combination of cone tip resistance (q_t), a large strain measurement, and the small strain shear modulus (G₀) addresses the limitations of relying solely on cone resistance and other traditional parameters. By incorporating measurements sensitive to the small strain properties of the soil, this approach results in a more comprehensive understanding of the soil’s geotechnical behavior, leading to improved predictions and assessments in engineering practice. Their primary aim was to systematically classify various types of soils while investigating the vulnerability to liquefaction. This involves the correction of parameters, including tip resistance (${\mathrm{q}}_{\mathrm{t}}$), which are influenced by drainage conditions. Yet, these studies should also encompass a thorough liquefaction potential acknowledgment and a comprehensive drainage analysis within tailing dams. To fill this research gap and given the catastrophic incidents involving tailing dams and their negative impact on the environment and communities due to flow liquefaction failures, this study aims to assess the susceptibility of iron ore tailings to such flow-induced liquefaction. To evaluate the liquefaction susceptibility of iron ore tailings, we estimated the state parameter from SCPTu tests. This estimation required assessing drainage conditions and correcting test results obtained under partially drained conditions. Consequently, the distinctive attributes of each site and prevailing drainage conditions were considered. The dataset under examination was originated from SCPTu tests carried out on three distinct upstream tailings dams situated in the Iron Quadrangle region [31], state of Minas Gerais, Brazil.

2. Background: Classification, Drainage Analysis and State Parameter Assessment

In this section, we provide a concise introduction and clarification on the following subjects:

• The classification methodology proposed by Nierwinski, Schnaid and Odebrecht [28,29,30] subsequently served as the basis for classifying the tailings examined in this study.

• Three distinct methodologies were employed to evaluate drainage conditions during the implementation of SCPTu tests.

• The evaluation of liquefaction susceptibility utilizing the state parameter approach ($\psi$) was derived from SCPTu tests.

2.1. Classification Methodology

Field data acquired from SCPTu tests were analyzed using the classification methodology introduced by Nierwinski, Schnaid and Odebrecht [28,29,30]. This method establishes a correlation between the ratio of the small strain shear modulus and the adjusted tip resistance (${\mathrm{G}}_0/{\mathrm{q}}_{\mathrm{t}}$) and the normalized tip resistance (${\mathrm{Q}}_{\mathrm{t}\mathrm{n}}$) (see Figure 1). This approach introduces a new classification perspective anchored in ${\mathrm{G}}_0$, thereby offering a different methodology for evaluating the potential for liquefaction susceptibility through the utilization of the state parameter ($\psi$). The normalized tip resistance can be computed as follows:

(1)${\mathrm{Q}}_{\mathrm{t}\mathrm{n}}=\left(\frac{{\mathrm{q}}_{\mathrm{t}}-{\sigma }_{\mathrm{v}}}{{\mathrm{p}}_{\mathrm{a}}}\right)\times {\left(\frac{{\mathrm{p}}_{\mathrm{a}}}{{\sigma \mathrm{\prime }}_{\mathrm{v}}}\right)}^{\mathrm{n}}$

where σ_v and σ′_v are the total and effective vertical stress, respectively, n is assumed to be 0.5 for granular soils, and p_a is the atmospheric pressure.

As illustrated in Figure 1, the classification method presents a distinct differentiation between plastic and non-plastic soils. The shaded region, indicated by Roman numerals, corresponds to plastic soils, whereas the unshaded region, marked with Arabic numerals, is associated with non-plastic soils. Additionally, there is a reduction in particle size from right to left and from top to bottom.

2.2. Drainage Analysis

The classification introduced by Nierwinski, Schnaid and Odebrecht [28,29,30] further encompasses a differentiation in terms of drainage conditions. The method delineates a zone of partial drainage conditions by assuming 10 < ${\mathrm{Q}}_{\mathrm{t}\mathrm{n}}$ < 50, which designates ${\mathrm{Q}}_{\mathrm{t}\mathrm{n}}$ < 10 as an undrained zone and identifies ${\mathrm{Q}}_{\mathrm{t}\mathrm{n}}$ > 50 as a fully drained zone.

Additionally, other parameters, like the pore pressure ratio (B_q) and the normalized penetration velocity (V), are employed to assess drainage conditions.

Regarding the pore pressure ratio (B_q), Jefferies and Been [32] defined a threshold of 0 to establish full drainage, thereby classifying B_q < 0 as indicative of full drainage conditions. Meanwhile, Schnaid, Lehane and Fahey [25] determined in their study that B_q > 0.5 means a non-drained state for bauxite tailings. Consequently, the range between these two benchmarks could suggest a scenario of partial drainage conditions. B_q can be estimated as follows [33]:

(2)${\mathrm{B}}_{\mathrm{q}}=\frac{\left(\mathrm{u}-{\mathrm{u}}_0\right)}{({\mathrm{q}}_{\mathrm{t}}-{\sigma }_{\mathrm{v}\mathrm{o}})}$

where the $\mathrm{u}$ variable represents the excess of pore pressure, and ${\mathrm{u}}_0$ is the hydrostatic water pressure.

Regarding the dimensionless normalized penetration velocity parameter (V), a fully drained condition is characterized when V falls within the range of less than 0.01 to 0.03, while an undrained penetration state is indicated when V exceeds 30 to 100 [34]. The calculation of V can be achieved using Equation (3):

(3)$\mathrm{V}=\frac{\mathrm{v}\times \mathrm{d}}{{\mathrm{c}}_{\mathrm{v}}}$

where $\mathrm{v}$ is the cone penetration velocity, $\mathrm{d}$ represents the cone diameter, and c_v is the vertical consolidation coefficient that can be obtained from dissipation tests.

In this study, the three parameters (B_q, Q_tn and V) obtained after SCPTu results were investigated in terms of drainage conditions. Even though B_q and Q_tn can be calculated for each row of SCPTu results, the parameter V can only be calculated when dissipation tests are performed.

To address the effects of partial drainage conditions, Nierwinski, Schnaid and Odebrecth [29,30] proposed a methodology to correct q_t20 (standard cone tip resistance) obtained in CPTu tests. The authors proposed the following equation to assess the effects of partial drainage conditions:

(4)$\frac{{\mathrm{q}}_{\mathrm{t}\mathrm{D}}}{{\mathrm{q}}_{\mathrm{t}20}}=1+\left(1+\frac{{\mathrm{q}}_{\mathrm{t}\mathrm{D}}}{{\mathrm{q}}_{\mathrm{t}\mathrm{U}\mathrm{D}}}\right)\times {\left({\mathrm{B}}_{\mathrm{q}}\right)}^{0.5}$

where ${\mathrm{q}}_{\mathrm{t}\mathrm{D}}$ is the drained cone tip resistance and ${\mathrm{q}}_{\mathrm{t}\mathrm{U}\mathrm{D}}$ represents the undrained cone tip resistance. The ratio ${\mathrm{q}}_{\mathrm{t}\mathrm{D}}$/${\mathrm{q}}_{\mathrm{t}\mathrm{U}\mathrm{D}}$ assumes a fixed value based on the material’s friction angle obtained from triaxial tests (Figure 2).

2.3. Liquefaction Susceptibility Evaluation

To assess liquefaction susceptibility, the state parameter ($\psi$) methodology introduced by Plewes, Davies and Jeffereis [36] offers a suitable approach. In this context, a value of $\psi$ > −0.05 signifies a contractive behavior during shear, consequently suggesting the potential for flow-induced liquefaction, as described by Equation (5):

(5)$\psi =\alpha {\left(\frac{\mathrm{p}\mathrm{\prime }}{{\mathrm{p}}_{\mathrm{a}}}\right)}^{\beta }+\chi \ln \left(\frac{{\mathrm{G}}_0}{{\mathrm{q}}_{\mathrm{t}}}\right)$

where p′ is the mean effective stress acting on the material and p_a is the atmospheric pressure. Also, α = −0.520, β = −0.07 and χ = 0.180 are (dimensionless) mean values obtained from tests in calibration chambers [37]. Equation (5) was first introduced to be used for granular soils, and therefore, materials with full drainage behavior.

3. Materials and Methods

3.1. Iron Ore Tailing

The iron ore tailings used in this study were sourced from three distinct upstream dams situated within the Iron Quadrangle region of Minas Gerais state. Throughout this paper, we designated these structures as Structure 1, Structure 2, and Structure 3 for clarity. Figure 3 provides an overview of the studied region, highlighting the presence of the tested dams within the Itabira and Nova Lima geological groups, which yielded a diverse range of results. The materials were extracted using sampling techniques that ensured the inclusion of both undisturbed and disturbed samples. Figure 3 also illustrates the diverse iron ore mining sites within each supergroup formation of the Iron Quadrangle. Notably, this figure highlights geosites within banded iron formations, which hold substantial geological significance as areas amenable to study, observation, and appreciation. Furthermore, the figure also encompasses several historical cities situated within the Iron Quadrangle area. The geographic overview (right) of the field site in Minas Gerais, Brazil is depicted with a red cross, whereas the catastrophic dam failures of Brumadinho (west) and Fundão (east) are denoted by red diamonds.

3.2. Field and Laboratory Program

A comprehensive series of both field and laboratory tests was executed in this study. The testing program was composed of the following data: (1) 531.26 m of SCPTu tests, (2) 317 seismic tests, and (3) characterization and triaxial tests on 16 samples collected from trenches and using a Shelby sampler.

For the survey campaign of Structure 1, 2 SCPTu tests, 5 seismic tests, and 6 samples were collected. As for Structure 2, 7 SCPTu tests were carried out along with 168 seismic tests, and 7 samples were collected. Structure 3 underwent 11 SCPTu tests, 144 seismic tests, and sampling for 3 tailing samples. The seismic tests were executed at intervals of either 1 or 2 m.

To complement the field tests, laboratory investigations were conducted to gain further insights into the characteristics of the iron ore tailings studied. Across all structures, a series of laboratory tests was performed, encompassing gradation analysis, void ratio determination, Atterberg limits assessment, specific gravity measurement, and triaxial tests. More detailed information regarding these tests are provided in Table 1.

Field and laboratory data were collected by well-trained professionals from the mining sector. The best preparation and arrangement were adhered to ensuring the optimal setup and calibration of field equipment. This included precise calibration of the cones, saturation of their porous elements, and the adjustment of the standard CPTu velocity (2 mm/s) throughout the entire testing process. Similar attention to detail was devoted to the laboratory tests, encompassing thorough sample treatment and careful preparation. Triaxial tests were performed using undisturbed samples collected directly from the tailing dams. These tests facilitated the determination of cohesion and friction angle, with the latter parameter subsequently employed in conjunction with the Senneset, Sandven and Janbu [35] abacus (as shown in Figure 2) for further analysis and interpretation of the drainage conditions (Section 2.2).

3.3. Data Analysis

During SCPTu tests, tip resistance (q_t), lateral friction (f_s) and pore pressure (u₂) were continuously measured. Pore pressure ratio (B_q) (Equation (2)) and material classification index (I_c) (Equation (6)) were also calculated after piezocone test results. The latter parameter can be calculated as follows [39]:

(6)${\mathrm{I}}_{\mathrm{c}}=\sqrt{{\left(3.47-\mathrm{l}\mathrm{o}\mathrm{g}{\mathrm{Q}}_{\mathrm{t}}\right)}^2+{\left(\mathrm{l}\mathrm{o}\mathrm{g}{\mathrm{F}}_{\mathrm{r}}+1.22\right)}^2}$

where ${\mathrm{Q}}_{\mathrm{t}}$ is the normalized tip resistance, and ${\mathrm{F}}_{\mathrm{r}}$ is the normalized friction ratio. Figure 4 illustrates depth-dependent profiles of q_t, f_s, u₂, B_q, and I_c in a standard SCPTu result for Structure 3.

Seismic tests were performed at specific depths in all SCPTu tests. The small strain shear modulus (G₀) was obtained considering the shear wave velocities (V_s) recorded during field tests and the specific gravity (ρ) of the tailings obtained from the laboratory tests executed, as shown in Equation (7), proposed by Hardin and Black [40]:

(7)${\mathrm{G}}_0=\rho \times {{\mathrm{V}}_{\mathrm{s}}}^2$

Dissipation tests were additionally conducted on specific boreholes. The termination criterion for these tests was set at a minimum of 50% dissipation from the initially recorded pore pressure. Consequently, the equivalent time (t₅₀) was employed to calculate the horizontal consolidation coefficient (${\mathrm{c}}_{\mathrm{h}}$), as per Equation (8), originally proposed by Houlsby and Teh [41]:

(8)${\mathrm{c}}_{\mathrm{h}}=\frac{\mathrm{T}\times {\mathrm{r}}^2\times {{\mathrm{I}}_{\mathrm{r}}}^{0.5}}{{\mathrm{t}}_{50}}$

where T is the time factor, r is the piezocone radius, and I_r is the stiffness index. During these tests, data for ${\mathrm{c}}_{\mathrm{h}}$ were recorded, and subsequently, ${\mathrm{c}}_{\mathrm{v}}$ was computed using Equation (9):

(9)${\mathrm{c}}_{\mathrm{v}}=\frac{{\mathrm{c}}_{\mathrm{h}}}{\raisebox{1ex}{${\mathrm{K}}_{\mathrm{h}}$}\!\left/ \!\raisebox{-1ex}{${\mathrm{k}}_{\mathrm{v}}$}\right.}$

where the ratio ${(\mathrm{K}}_{\mathrm{h}}/{\mathrm{k}}_{\mathrm{v}}$) represents the anisotropy between vertical and horizontal permeabilities. The ratio ${(\mathrm{K}}_{\mathrm{h}}/{\mathrm{k}}_{\mathrm{v}}$) is set at a value of 2, following the approach by Vick [42].

Parameters such as G₀ (Equation (7)) and c_v (Equation (9)) were exclusively calculated at specific depths corresponding to the locations of seismic or dissipations tests. Consequently, associated parameters, such as Ψ and V, were likewise computed at the same specific depths.

4. Results and Discussion

In this section, field and laboratory tests are presented and discussed. Subsequently, we resorted to the methodology proposed by Nierwinski, Schnaid and Odebrecht [28,29,30] for the classification of iron ore tailings. The assessment of drainage conditions was conducted through the examination of pore pressure ratio (B_q), Q_tn, and normalized penetration velocity (V), allowing the identification of partial drainage. Finally, the analysis of liquefaction susceptibility was undertaken, utilizing parameters derived directly from field and laboratory data, along with corrections applied to q_t.

4.1. Field and Laboratory Test Results

Laboratory characterization tests, as shown in Figure 5 and summarized in Table 1, revealed that most samples consisted predominantly of sands or silts as their primary and secondary components. The average D₅₀ fell within the order of magnitude of 10⁻² mm, a typical value associated with silts and fine sands. Of all the samples, 66% displayed no plasticity limit, while among the remaining 33%, 80% exhibited a plasticity index ranging between 7 and 15, signifying an intermediate level of plasticity.

To provide a more detailed description of the materials under investigation, please refer to Table 1. This table presents a detailed summary of the laboratory test results, organized according to the structures studied. Specific gravity (SG) values ranged between 2.82 and 4.77, with most samples displaying no noticeable plasticity index. Consoli et al. [43] classified iron ore tailings from the Iron Quadrangle as predominantly silty sand, with a specific gravity of solids of 2.92. Similarly, Mmbando, Fourie and Reid [44] concurred, characterizing most of the Iron Quadrangle iron ore tailings as primarily silt or sand, and their specific gravity was 3.71. In contrast, Duarte et al. [45] identified the iron ore tailings from the Fundão Dam as predominantly clayey silt. Additionally, Siqueira et al. [46], in their investigation of Brumadinho Dam tailings, determined them to be primarily composed of sands and silts, with specific gravity values ranging from 2.43 to 4.86. The liquid limit (LL) and plastic limit (PL) of the samples were established, revealing that Structure 1 exhibits plastic behavior, whereas the other structures, owing to their larger particle sizes, exhibit minimal or negligible plasticity.

The material classification index (I_c) calculated from field tests (Equation (6)), illustrated in Figure 4, varies from 2.05 to 4, indicating a very heterogeneous material, with layers of sandy, silty, clayey, and even organic material, according to Robertson [39]. Apart from the laboratory tests that characterize the collected samples, SCPTu tests can only infer of the materials behavior, since no samples are extracted. The pore pressure ratio (B_q) (Equation (2)), also shown in Figure 5, varies from 0 to 1, with approximately all readings comprehended in the interval of 0 and 0.5, which could already indicate a partial drainage condition.

Triaxial tests yielded cohesion ($c$) intercept results ranging from 0 to approximately 50 kPa, with a mean value of 10.71 kPa and a median of 5 kPa. Additionally, the friction angle ($\varphi$) results spanned from 21° to 37°, with a mean value of 32.09° and a median of 36°. Figure 6 illustrates a representative result, displaying the Mohr circle along with the obtained Mohr–Coulomb parameters. The observed range of variation in the Mohr–Coulomb parameter results is consistent with those reported in the literature for copper and iron tailings [47]. Morgenstern et al. [7] investigated the Fundão Dam, reporting friction angles ranging from 28° to 36°, classifying the tailings as cohesionless sands. In a separate study, Becker, Cavalcanti and Marques [48] examined the tailings from the Germano Dam (Iron Quadrangle region), finding friction angle values that varied from 28° to 37°. Albuquerque Filho [49] conducted research on tailings from four different dams in the Iron Quadrangle region and documented friction angle results spanning from 26° to 45° for the material, along with observations of low or negligible cohesion.

4.2. Classification Methodology

The results of all SCPTu tests were compiled and are graphically represented in Figure 7, following the classification methodology proposed by Nierwinski, Schnaid and Odebrecht [28,29,30]. The plot illustrates that iron ore tailings exhibit a heterogeneous nature, as the data points are widely distributed on both the horizontal and vertical axes. This distribution spans both plastic and non-plastic behavior areas, ranging from non-plastic sensitive soils to silts and sands. However, it should be noted that approximately 97% of the data fall within the non-plastic region, indicating a predominantly granular behavior in the tailings. Consequently, the assessment of the susceptibility to liquefaction is performed for the entire material studied.

While the correlation derived from SCPTu tests (I_c) suggests a heterogeneous composition of the tailings, with layers containing clays and organic materials, the classification methodology illustrated above categorizes the tailings as a non-plastic material, with many tests falling within the sands and silts categories (regions 3 and 4). This classification is corroborated by the particle size distribution curves (Figure 5a–c), which confirms a sandy/silty granulometry and is further supported by additional laboratory tests such as Atterberg limits and triaxial tests.

4.3. Drainage Analysis

Drainage conditions were analyzed using three different parameters: B_q, Q_tn and V. The histograms shown in Figure 8 emphasize the count of recorded parameters that indicate partial drainage, highlighted within the shaded grey region. According to the B_q analysis, 40% of the readings suggested partial drainage conditions, based on the Q_tn analysis, 62%, and as per the V analysis, 37%. As such, it is evident that the three methodologies yielded divergent results from each other.

To apply the correction to q_t using the methodology proposed by Nierwinski, Schnaid and Odebrecth [29,30], triaxial tests were conducted on undisturbed samples collected from the tailing dams. The analysis revealed a mean friction angle of 35.23° for Structure 1, 22.25° for Structure 2, and 36.55° for Structure 3. Consequently, the q_tD/q_tUD ratios adopted for each structure were 8.50, 4.69, and 9.02, respectively, using the upper limit of the approach outlined by Senneset, Sandven and Janbu [35], as previously illustrated in Figure 2.

Figure 9 shows the tip resistance (q_t) after applying the correction outlined in Equation (4), juxtaposed with the uncorrected parameter for the same borehole shown in Figure 4. Due to the limited number of dissipation tests conducted, the analysis involving the V parameter presented the smallest number of q_t corrections. Furthermore, the classification index (I_c) was recalibrated using the corrected q_t values. Figure 10 provides a comparative view of I_c (Equation (6)) before and after the q_t adjustment for the same SCPTu test illustrated in Figure 4. The adjusted I_c changes noticeably to the left, signifying a more granular behavior compared to the initial classification. This realignment agrees with the methodology proposed by Nierwinski, Schnaid and Odebrecht [28,29,30] and is consistent with the results of laboratory tests.

Based on the original data from the SCPTu test presented in Figure 4, it is noteworthy that at depths where excess pore pressure (u₂) was recorded, the corrections resulted in higher q_t values, corresponding to a drained condition. It is also observed that at depths below 30 m, the original results indicated peaks in the generation of excess pore pressure and B_q values. Consequently, the corrected q_t values exhibited peaks, distinguishing themselves from the uncorrected test results. The increase in q_t values with corrections directly affected I_c values, altering the soil behavior classification. Higher values of excess pore pressure led to more pronounced changes, particularly at depths below 30 m, where peaks in the generation of excess pore pressure and B_q were observed.

4.4. Liquefaction Susceptibility

For the non-plastic materials shown in Figure 7, the state parameter was calculated and graphed against G₀/q_t (estimation without correction in Figure 11). Notably, around 94% of the tailings exhibited Ψ values greater than −0.05, indicating a contractive behavior during shear. This observation suggests a potential susceptibility to flow liquefaction [36]. A similar analysis was conducted following the correction of q_t for partially drained conditions. Given the notable disparities between the B_q, Q_tn and V methods, the assessment of liquefaction susceptibility accounted for the q_t correction specific to each test identified as partially drained using any of the three methods. Upon correcting q_t, the SCPTu estimations with corrections showed a slight upward adjustment, aligning more effectively within the stress limits, defined by Schnaid and Yu [37] as in situ stresses generally experienced by soils (estimation with correction in Figure 11). However, in the context of the analysis of state parameters and susceptibility to liquefaction, the material still exhibited a significant risk of experiencing flow liquefaction. In fact, even when accounting for the adjustment of partially drained conditions, the analysis based on B_q revealed that 96% of the tests still indicated susceptibility to flow liquefaction (Ψ > −0.05) [36], while the Q_tn analysis identified susceptibility in 98%, and the V analysis identified susceptibility in 99% of the cases.

5. Conclusions

The Brazilian Iron Quadrangle region is home to numerous upstream tailing dams, making it necessary to assess the potential for liquefaction in these structures, a critical aspect for their management. This research evaluated the liquefaction susceptibility based on the estimation of Ψ values [36] in upstream Brazilian iron ore tailing dams, considering drainage analysis based on SCPTu tests. Field and laboratory tests were conducted in three upstream dams located within the Iron Quadrangle region of Brazil. The testing program included 531.26 m of SCPTu tests, 317 seismic tests, and characterization and triaxial tests performed on samples obtained from trenches and using a Shelby sampler. Our methodology can be delineated into three main components: First, we employed the classification methodology proposed by Nierwinski, Schnaid and Odebrecht [28,29,30] to distinctly differentiate between plastic and non-plastic tailings while simultaneously assessing drainage conditions. Second, three methodologies were employed to assess drainage conditions during the execution of SCPTu tests, involving the utilization of normalized tip resistance (Q_tn), pore pressure ratio (B_q), and the normalized penetration velocity parameter (V). This analysis was complemented by adjusting the standard cone tip resistance obtained in CPTu tests (q_t) to consider the influence of partial drainage conditions. Lastly, we examined liquefaction susceptibility employing the state parameter approach (Ψ) [36] derived from SCPTu tests.

Our findings reveal the high susceptibility of the iron ore tailings from the Iron Quadrangle region to flow liquefaction. The state parameter approach [36] revealed that 94% of the samples exhibited vulnerability to liquefaction (e.g., presented Ψ > −0.05) [36]. Even after accounting for the correction of partially drained conditions, the analysis based on B_q indicated that 96% of the tests still showed susceptibility to flow liquefaction, while the Q_tn analysis identified 98%, and the V analysis identified susceptibility of 99%. The results also indicate a significant level of heterogeneity in the studied tailings. Various parameters, such as specific gravity, material index (I_c), and the proposed classification method, exhibited considerable variations. Additionally, the three distinct approaches used to assess drainage conditions demonstrated some degree of variability among each other, emphasizing the complexity of defining geotechnical analyses in terms of total and effective stress. This study also serves to validate the methodologies developed for classifying and assessing liquefaction susceptibility in different soil types, with a particular focus on expanding the knowledge and database on iron ore tailings.

By employing the state parameter (Ψ) derived from SCPTu tests, our research demonstrates a methodology for assessing the liquefaction potential of iron ore tailings, considering drainage effects. This approach allows for a more precise characterization of tailing materials, which is crucial for designing safer and more reliable tailing storage facilities. Our findings offer a comprehensive framework that can be applied to similar mining sites, aiding in the prevention and mitigation of catastrophic incidents associated with tailing dam failures. This study not only advances the field characterization of tailings but also contributes to the development of more effective design strategies and risk management practices in the mining industry.

Despite our efforts to characterize iron ore tailings in upstream dams through extensive laboratory and SCPTu tests, future research could benefit from the collection of undisturbed samples and triaxial tests to directly obtain the state parameter. Additionally, laboratory tests, such as those involving bender elements, could be conducted to refine and calibrate the results obtained from the field tests conducted in this study. From a parallel perspective, evaluating other parameters such as residual shear strength and their rates of variation can provide valuable insights for a better understanding of the behavior of mining tailings.

Author Contributions

Conceptualization, G.C.L.R.d.C. and H.P.N.; methodology, G.C.L.R.d.C. and H.P.N.; software, G.C.L.R.d.C.; validation, G.J.C.G. and H.P.N.; formal analysis, G.C.L.R.d.C., G.J.C.G. and H.P.N.; investigation, G.C.L.R.d.C.; resources, G.C.L.R.d.C., G.J.C.G. and H.P.N.; data curation, G.C.L.R.d.C.; writing—original draft preparation, G.C.L.R.d.C.; writing—review and editing, G.J.C.G. and H.P.N.; visualization, G.C.L.R.d.C.; supervision, G.J.C.G. and H.P.N.; project administration, G.J.C.G.; funding acquisition, G.J.C.G. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Data is contained within the article.

Acknowledgments

The authors wish to thank the Federal University of Ouro Preto (UFOP) and, especially, Chammas Engenharia for providing the data.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Table

Figure 1. Classification methodology adapted from Nierwinski, Schnaid and Odebrecht [28,29,30], clearly distinguishing between plastic and non-plastic soils while simultaneously evaluating drainage conditions.

Figure 2. Correlation between the qtD/qtUD ratio and the friction angle ([Forumla omitted. See PDF.]) obtained from triaxial tests. The friction angle value can be used to approximate the qtD/qtUD ratio, consequently facilitating the correction of qtD through Equation (4). The shaded region in gray highlights a range of variations documented in the literature. The listed references can be found in Senneset, Sandven and Janbu [35].

Figure 3. Study site: Simplified geological map of the Iron Quadrangle (Minas Gerais, Brazil), highlighting mining sites and historical cities. This map was produced based on research by Cavalcanti et al. [38].

Figure 4. Typical SPCTu results from Structure 3 showing qt, fs, u2, Bq and Ic plotted against depth. The layers from the Ic graphic stand for: 1—Gravelly sand, 2—Sands: clean to silty, 3—Silty sands to sandy silts, 4—Clayey silt to silty clay, 5—Clays, and 6—Organic soils.

Figure 5. Particle size distribution curves for the three structures: (a) Structure 1, (b) Structure 2, and (c) Structure 3 under study, revealing that the predominant material composition consists of a combination of sands and silts.

Figure 6. Mohr circles and the failure envelope derived from a representative triaxial test result. The plot includes the Mohr–Coulomb derived parameters: cohesion (0) and friction angle (37°).

Figure 7. Results of the classification method proposed by Nierwinski, Schnaid and Odebrecht [28,29,30] applied to the investigated tailings (structures). Ninety-seven percent of the data are situated within the non-plastic zone, implying a prevailing granular characteristic within the tailings.

Figure 8. Results of the drainage conditions analysis using (a) Bq, (b) Qtn and (c) V parameters. The histogram highlights, within the shaded grey region, the count of records obtained from SCPTu tests indicating partial drainage conditions according to three distinct methodologies.

Figure 9. SCPTu results are displayed with qt plotted against standard velocity, followed by the application of drainage corrections based on (a) Bq, (b) Qtn and (c) V.

Figure 10. Comparison of Ic with qt on standard velocity and subsequent application of drainage correction based on (a) Bq, (b) Qtn and (c) V.

Figure 11. Estimation of state parameters for the iron ore tailings before and after qt correction according to Bq, Qtn and V. P limit values are boundaries defined by in situ stresses generally experienced by soils [37].

Table 1

Summary of laboratory tests executed and their results for each structure studied.

Structure	Sample Type	Clay (%)	Silt (%)	Sand (%)	D50 (mm)	LL (%)	LP (%)	SG	Void Ratio	Triaxial Type	Cohesion (kPa)	Friction Angle (°)
1	Reconstituted	14.96	34.14	7.05	0.01	37.24	14.05	2.84	1.03	CIDSat	42.48	33.46
1	Disturbed	53.98	16.04	19.28	0.001	38.34	11.52	3.04	*	*	*	*
1	Disturbed	57.08	19.29	19.82	0.001	59.75	26.94	2.94	*	*	*	*
1	Disturbed	65.92	23.36	10.71	-	43.44	14.61	3.15	*	*	*	*
1	Reconstituted	19.15	60.82	17.6	0.01	54.34	23.57	2.87	1.01	CIUSat	20	37
1	Disturbed	15.25	48.26	11.97	0.01	28.34	13.98	2.82	*	*	*	*
2	Reconstituted	8.81	57.1	34.1	0.01	/	/	3.54	0.854	CIUSat	0	23.33
2	Reconstituted	4.96	28.79	52.31	0.1	/	/	3.5	0.827	CIUSat	7.5	21.17
2	Undisturbed (Shelby)	4	53	38	0.1	/	/	4.66	*	*	*	*
2	Undisturbed (Shelby)	3	83	14	0.01	/	/	4.77	*	*	*	*
2	Undisturbed (Shelby)	3	81	16	0.01	/	/	4.7	*	*	*	*
2	Undisturbed (Shelby)	4	60	37	0.1	/	/	4.37	*	*	*	*
2	Undisturbed (Shelby)	5	76	19	0.01	/	/	4.47	*	*	*	*
3	Reconstituted	1.91	19.5	78.59	0.1	/	/	3.3	0.67	CIUSat	0	37
3	Reconstituted	3.11	24.04	72.85	0.1	/	/	3.38	0.653	CIUSat	0	36
3	Reconstituted	2.81	19.18	78.02	0.1	/	/	3.36	0.696	CIUSat	5	36.65

* No tests were conducted; / Results nonexistent.

References

1. USGS—U.S. Geological Survey, Mineral Commodity Summaries. Iron Ore, January 2023. Available online: https://pubs.usgs.gov/periodicals/mcs2023/mcs2023-iron-ore.pdf (accessed on 15 August 2023).

2. Brazilian Government. Brazil Stands out in the Export of Iron Ore. 2022; Available online: https://www.gov.br/resolveuid/532757dd80974e2ab16ce09293524942 (accessed on 28 July 2023).

3. Carmo, F.F.; Kamino, L.H.Y.; Tobias, R., Jr.; Campos, I.C.; Carmo, F.F.; Silvino, G.; Castro, K.J.S.X.; Mauro, M.L.; Rodrigues, N.U.A.; Miranda, M.P.S. et al. Fundão tailings dam failures: The environment tragedy of the largest technological disaster of Brazilian mining in global context. Perspect. Ecol. Conserv.; 2017; 15, pp. 145-151. [DOI: https://dx.doi.org/10.1016/j.pecon.2017.06.002]

4. Quaresma, V.S.; Aguiar, V.M.C.; Bastos, A.C.; Oliveira, K.S.; Vieira, F.V.; Sá, F.; Baptista Neto, J.A. The impact of trace metals in marine sediments after a tailing dam failure: The Fundão dam case (Brazil). Environ. Earth Sci.; 2021; 80, 571. [DOI: https://dx.doi.org/10.1007/s12665-021-09817-x]

5. Cambridge, M.; Shaw, D. Preliminary reflections on the failure of the Brumadinho tailings dam in January 2019. Dams Reserv.; 2019; 29, pp. 113-123. [DOI: https://dx.doi.org/10.1680/jdare.19.00004]

6. National Mining Agency. Mining Dams System of Safety Management. 2023; Available online: https://app.anm.gov.br/SIGBM/Publico/GerenciarPublico (accessed on 7 August 2023).

7. Morgenstern, N.R.; Vick, S.G.; Viotti, S.G.; Watts, B.D. Report on the Immediate Causes of the Failure of the Fundão Dam. 2016; Available online: https://www.resolutionmineeis.us/sites/default/files/references/fundao-2016.pdf (accessed on 4 August 2023).

8. Robertson, P.K.; de Melo, L.; Williams, D.J.; Wilson, G.W. Report of the Expert Panel on the Technical Causes of the Failure of Feijão Dam I Expert Panel Technical Report. Available online: http://www.b1technicalinvestigation.com/ (accessed on 4 August 2023).

9. Vergaray, L.; Macedo, J.; Arnold, C. Static and cyclic liquefaction of copper mine tailings. J. Geotech. Geoenviron. Eng.; 2023; 149, 04023021. [DOI: https://dx.doi.org/10.1061/JGGEFK.GTENG-10661]

10. Reid, D.; Urbina, F.; Tiwari, B.; Fanni, R.; Smith, K.; Fourie, A. Effect of saturation confining pressure on accessible density and shear behaviour of sandy silt tailings. Géotechnique Lett.; 2023; 13, pp. 113-117. [DOI: https://dx.doi.org/10.1680/jgele.23.00001]

11. Wang, Y.; Vo, T.; Russell, A.R. Modelling unsaturated silty tailings and the conditions required for static liquefaction. Géotechnique; 2023; ahead of print [DOI: https://dx.doi.org/10.1680/jgeot.22.00074]

12. Ozcan, N.T.; Ulusay, R.; Isik, N.S. A study on geotechnical characterization and stability of downstream slope of a tailings dam to improve its storage capacity (Turkey). Environ. Earth Sci.; 2013; 69, pp. 1871-1890. [DOI: https://dx.doi.org/10.1007/s12665-012-2016-1]

13. Fanni, R.; Reid, D.; Fourie, A. On reliability of inferring liquefied shear strengths from simple shear testing. Soils Found.; 2022; 62, 101151. [DOI: https://dx.doi.org/10.1016/j.sandf.2022.101151]

14. Macek, M.; Smolar, J.; Petkovsek, A. The reliability of CPTu and DMT for the mechanical characterisation of soft tailings. Bull. Eng. Geol. Environ.; 2018; 78, pp. 2237-2252. [DOI: https://dx.doi.org/10.1007/s10064-018-1299-1]

15. Rocha, B.P.; Silveira, I.A.; Rodrigues, R.A.; Lodi, P.C.; Giacheti, H.L. Identifying Collapsible Soils from Seismic Cone (SCPT): A Qualitative Approach. Buildings; 2023; 13, 830. [DOI: https://dx.doi.org/10.3390/buildings13030830]

16. Campanella, R.G.; Robertson, P.K. A seismic cone penetrometer to measure engineering properties of soil. SEG Technical Program Expanded Abstracts; Society of Exploration Geophysicists: Houston, TX, USA, 1984; pp. 138-141. [DOI: https://dx.doi.org/10.1190/1.1894361]

17. Kawa, M.; Baginska, I.; Wyjadlowski, M. Reliability analysis of sheet pile wall in spatially variable soil including CPTu test results. Arch. Civ. Mech. Eng.; 2019; 19, pp. 598-613. [DOI: https://dx.doi.org/10.1016/j.acme.2018.10.007]

18. Wang, H.; Wu, S.; Qi, X.; Chu, J. Site characterization of reclaimed lands based on seismic cone penetration test. Eng. Geol.; 2021; 280, 105953. [DOI: https://dx.doi.org/10.1016/j.enggeo.2020.105953]

19. Long, M. Practical use of shear wave velocity measurements from SCPTU in clays. Cone Penetration Testing 2022; 1st ed. Gottardi, G.; Tonni, L. CRC Press: London, UK, 2022; pp. 28-52.

20. Cordeiro, D.; da Fonseca, A.V.; Ferreira, C.; Molina-Gómez, F.; Rodrigues, C. Liquefaction assessment through SCPTu and DMT tests: Aveiro case study. Proceedings of the 6th International Conference on Geotechnical and Geophysical Site Characterization (ISC2020); Budapest, Hungary, 7–11 September 2020; [DOI: https://dx.doi.org/10.53243/ISC2020-199]

21. Bala, A.; Hannich, D. Liquefaction potential analysis in bucharest city as a result of the ground shaking during strong Vrancea earthquakes. Athens J. Technol. Eng.; 2021; 8, pp. 113-138. [DOI: https://dx.doi.org/10.30958/ajte.8-2-1]

22. Zafarani, A.R.; Ewane, M.S.; Zamani, A.H.; Gomes, P.; Painchaud, J.F. In-situ geotechnical investigations for evaluating static liquefaction potential of mine tailings. Proceedings of the GeoCalgary 2022; Calgary, AB, Canada, 2–5 October 2022.

23. Lambe, T.W.; Whitman, R.V. Soil Mechanics; Wiley: New York, NY, USA, 1969.

24. Schnaid, F. The Ninth James, K. Mitchell Lecture: On The Geomechanics and Geocharacterization of Tailings. Proceedings of the 6th International Conference on Geotechnical and Geophysical Site Characterization (ISC2020); Budapest, Hungary, 7–11 September 2020; [DOI: https://dx.doi.org/10.53243/ISC2020-501]

25. Schnaid, F.; Lehane, B.M.; Fahey, M. In situ test characterization of unusual soils. Keynote lecture. Proceedings of the 2nd International Geotechnical and Geophysical Site Characterization (ICS2004); Porto, Portugal, 19–22 September.

26. DeJong, J.T.; Randolph, M.F. Influence of partial consolidation during cone penetration on estimated soil behavior type and pore pressure dissipation measurements. J. Geotech. Geoenviron. Eng.; 2012; 138, pp. 777-788. [DOI: https://dx.doi.org/10.1061/(ASCE)GT.1943-5606.0000646]

27. Dienstmann, G.; Schnaid, F.; Maghous, S.; DeJong, J. Piezocone Penetration Rate Effects in Transient Gold Tailings. J. Geotech. Geoenviron. Eng.; 2018; 144, 04017116. [DOI: https://dx.doi.org/10.1061/(ASCE)GT.1943-5606.0001822]

28. Schnaid, F.; Nierwinski, H.P.; Odebrecht, E. Classification and state-parameter assessment of granular soils using the seismic cone. J. Geotech. Geoenviron. Eng.; 2020; 146, 06020009. [DOI: https://dx.doi.org/10.1061/(ASCE)GT.1943-5606.0002306]

29. Nierwinski, H.; Schnaid, F.; Odebrecht, E. In-situ state parameter assessment of non-plastic silty soils using the seismic cone. Proceedings of the 6th International Conference on Geotechnical and Geophysical Site Characterization (ISC2020); Budapest, Hungary, 7–11 September 2020; [DOI: https://dx.doi.org/10.53243/ISC2020-304]

30. Nierwinski, H.P.; Schnaid, F.; Odebrecht, E. Evaluation of Flow Liquefaction Susceptibility in Non-Plastic Silty Soils Using the Seismic Cone. Mining; 2024; 4, pp. 21-36. [DOI: https://dx.doi.org/10.3390/mining4010003]

31. Vicq, R.; Matschullat, J.; Leite, M.G.P.; Nalini, H.A., Jr.; Mendonça, F.P.C. Iron Quadrangle stream sediments, Brazil: Geochemical maps and reference values. Environ. Earth Sci.; 2015; 74, pp. 4407-4417. [DOI: https://dx.doi.org/10.1007/s12665-015-4508-2]

32. Jefferies, M.; Been, K. Soil Liquefaction: A Critical State Approach; 2nd ed. CRC Press: Boca Raton, FL, USA, 2015; [DOI: https://dx.doi.org/10.1201/b19114]

33. Robertson, P.K. Soil classification using the cone penetration test. Can. Geotech. J.; 1990; 27, pp. 151-158. [DOI: https://dx.doi.org/10.1139/t90-014]

34. Randolph, M. Characterisation of soft sediments for offshore applications. Proceedings of the 2nd International Geotechnical and Geophysical Site Characterization (ICS2004); Porto, Portugal, 19–22 September 2004.

35. Senneset, K.; Sandven, R.; Janbu, N. Evaluation of soil parameters from piezocone tests. Transp. Res. Rec.; 1989; 1235.

36. Plewes, H.D.; Davies, M.P.; Jefferies, M.G. CPT based screening procedure for evaluating liquefaction susceptibility. Proceedings of the 45th Canadian Geotechnical Conference; Toronto, ON, Canada, 26–28 October 1992.

37. Schnaid, F.; Yu, H.S. Interpretation of the seismic cone test in granular soils. Géotechnique; 2007; 57, pp. 265-272. [DOI: https://dx.doi.org/10.1680/geot.2007.57.3.265]

38. Cavalcanti, J.A.D.; da Silva, M.S.; Schobbenhaus, C.; Atencio, D.; de Lima, H.M. Geoconservation of geological and mining heritage related to banded iron formation of Itabira Group, Quadrilátero Ferrífero, Minas Gerais, Brazil: A challenging issue. Int. J. Geoheritage Parks; 2023; 11, pp. 118-148. [DOI: https://dx.doi.org/10.1016/j.ijgeop.2022.12.002]

39. Robertson, P.K. Soil behaviour type from the CPT: An update. Proceedings of the 2nd International Symposium on Cone Penetration Testing (CPT’10); Huntington Beach, CA, USA, 8–12 May 2010.

40. Hardin, B.O.; Black, W.L. Sand stiffness under various triaxial stresses. J. Soil Mech. Found. Div.; 1966; 92, pp. 27-42. [DOI: https://dx.doi.org/10.1061/JSFEAQ.0000865]

41. Houlsby, G.T.; Teh, C.I. Analysis of the piezocone in clay. Proceedings of the 1st International Symposium on Penetration Testing (ISOPT-1); Orlando, FL, USA, 20–24 March 1988.

42. Vick, S.G. Planning, Design and Analysis of Tailing Dams; 1st ed. Wiley: New York, NY, USA, 1983.

43. Consoli, N.C.; Vogt, J.C.; Silva, J.P.S.; Chaves, H.M.; Scheuermann Filho, H.C.; Moreira, E.B.; Lotero, A. Behaviour of Compacted Filtered Iron Ore Tailings–Portland Cement Blends: New Brazilian Trend for Tailings Disposal by Stacking. Appl. Sci.; 2022; 12, 836. [DOI: https://dx.doi.org/10.3390/app12020836]

44. Mmbando, E.; Fourie, A.; Reid, D. Mechanics of an Iron Ore Tailings Exhibiting Transitional Behaviour. Geotech. Geol. Eng.; 2023; 41, pp. 2211-2220. [DOI: https://dx.doi.org/10.1007/s10706-023-02379-8]

45. Duarte, E.B.; Neves, M.A.; de Oliveira, F.B.; Martins, M.E.; de Oliveira, C.H.R.; Burak, D.L.; D’Azeredo Orlando, M.T.; Rangel, C.V.G.T. Trace metals in Rio Doce sediments before and after the collapse of the Fundão iron ore tailing dam. Chemosphere; 2021; 262, 127879. [DOI: https://dx.doi.org/10.1016/j.chemosphere.2020.127879] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32781334]

46. Siqueira, D.; Cesar, R.; Lourenço, R.; Salomão, A.; Marques, M.; Polivanov, H.; Teixeira, M.; Vezzone, M.; Santos, D.; Koifman, G. et al. Terrestrial and aquatic ecotoxicity of iron ore tailings after the failure of VALE S.A mining dam in Brumadinho (Brazil). J. Geochem. Explor.; 2022; 235, 106954. [DOI: https://dx.doi.org/10.1016/j.gexplo.2022.106954]

47. Hu, L.; Wu, H.; Zhang, L.; Zhang, P.; Wen, Q. Geotechnical properties of mine tailings. J. Mater. Civ. Eng.; 2017; 29, 04016220. [DOI: https://dx.doi.org/10.1061/(ASCE)MT.1943-5533.0001736]

48. Becker, L.D.B.; Cavalcanti, M.C.R.; Marques, A.A.M. Statistical Analysis of the Effective Friction Angle of Sand Tailings from Germano Dam. Infrastructures; 2023; 8, 61. [DOI: https://dx.doi.org/10.3390/infrastructures8030061]

49. Albuquerque Filho, L.H. Iron Ore Tailing Dams Geotechnical Behavior Evaluation through Piezocone Tests. Master’s Thesis; Federal University of Ouro Preto: Ouro Preto, Brazil, 2004; (In Portuguese)