Study on blind zone identification and genetic mechanism of leaching process of ion-adsorption rare earth ore

1 Introduction

Ion-adsorption rare earth elements are valuable mineral resources that are abundant in China but scarce worldwide (Li et al., 2017; Navarro, 2015; Xu et al., 2023). Although their proven reserves account for only approximately 1.4% of global rare-earth resources, they supply more than 90% of the global demand for heavy rare earth elements. In 1991, the State Council of the People’s Republic of China classified ion-adsorption rare earths as protected minerals subject to controlled extraction and designated them as strategic resources (Lin, 1991). The Rare Earth Industry Development Plan (2016–2020) further emphasized the critical role of rare earth elements in upgrading traditional industries, fostering emerging sectors, and supporting national defense and advanced technology development (Luo et al., 2022).

At present, ion-adsorption rare earth resources are primarily distributed across seven provinces in southern China, with particularly abundant deposits in southern Jiangxi. To balance resource recovery efficiency, production costs, and surface environmental protection, in-situ leaching has been widely adopted as the predominant extraction method (Luo et al., 2018). This technique involves injecting a leaching solution directly into the ore body without disturbing surface vegetation or excavating topsoil and ore. Through ion-exchange reactions, cations in the leaching solution replace rare earth ions adsorbed on the surfaces of clay minerals. As a result, a pregnant leach solution is generated, from which rare earth elements are subsequently recovered. Owing to its advantages in minimizing surface disturbance and reducing mining costs, in-situ leaching is well suited for large-scale application (Hu et al., 2017; Nie et al., 2020). Despite these advantages, the widespread application of in-situ leaching presents several challenges. Long-term intensive mining and outdated extraction technologies have resulted in low rare-earth recovery rates and excessive consumption of leaching agents (Fan, 2022; Zhang, 2009). One of the primary causes of these inefficiencies is the formation of so-called leaching blind zones within the ore body. These zones arise from the combined effects of unfavorable geological conditions and suboptimal engineering parameters. In such areas, leaching solutions cannot effectively penetrate, and ion-exchange reactions remain incomplete (Wang et al., 2016; Zheng et al., 2020). The presence of leaching blind zones impede the complete recovery of residual rare earth resources. It also necessitates repeated reinjection of leaching solutions to enhance ion exchange. As a consequence, reagent consumption and overall production costs are substantially increased (Wang et al., 2016).

Two main types of leaching blind zones have been identified. The first type occurs between injection wells, where areas between adjacent permeation ellipsoids cannot be effectively wetted, resulting in incomplete leaching (Wang et al., 2016). The second type results from internal heterogeneity of the ore body, including anisotropy in the ore–soil structure and nonuniform seepage behavior, which also promotes the formation of leaching blind zones (Zheng et al., 2020). The formation of leaching blind zones is mainly attributed to two factors. First, weathered crust–type rare-earth deposits contain both macropores and micropores. The proportion of micropores that hold immobile water is higher than that in soils with similar texture. According to soil science theory, pore-water velocity is proportional to the fourth power of pore diameter. Consequently, water flows substantially faster through larger pores (Sun, 2021). As a result, preferential flow paths develop during in-situ leaching. This leads to nonuniform distribution of the leaching solution and incomplete leaching, ultimately forming blind zones. Second, inappropriate design of injection well spacing, injection intensity, and injection sequence during in-situ leaching can promote the leaching solution to form fixed flow channels. These channels prevent the solution from reaching rare-earth minerals outside the dominant pathways, thereby creating leaching blind zones (Lu et al., 1997; Wang et al., 2016; Zheng et al., 2020). The presence of leaching blind zones leads to the loss of rare-earth resources, reduces recovery efficiency, and increase production costs. For many years, this problem has remained unresolved in ion-adsorption rare-earth mining, primarily because the spatial locations and distributions of blind zones are difficult to identify accurately. Accurate identification of these zones and characterization of their spatial distribution and occurrence through appropriate monitoring techniques would enable targeted recovery measures to be implemented, thereby facilitating the extraction of residual rare earth resources and reducing resource waste.

Research on leaching blind zones in ion-adsorption rare-earth mines remains limited. Lu et al. (1997) identified the problem of leaching blind zones, explained their formation mechanisms, and proposed effective preventive measures. Zheng et al. (2020) conducted borehole sampling in abandoned rare-earth mines. They found that blind zones and areas affected by channelized flow retained high rare-earth contents from the surface to a depth of 5 m. Based on these findings, they recommended the secondary recovery of residual rare earth resources from these zones. Using numerical simulations, Zhang (2023b) investigated the effects of pressure head, pore size, and matric potential on blind-zone formation. A predictive model for blind-zone area was subsequently proposed. Zhang (2024) applied industrial computed tomography (CT) scanning and three-dimensional reconstruction to quantify flow velocity within blind zones, reporting a value of 8.2 × 10⁻⁷ m/s. Optimization of the injection conditions significantly reduced the blind-zone area Through simulations of two-phase water–air displacement in soil pores, Xiao et al. (2024) identified ore-body saturation as a key factor contributing to blind-zone formation. Zhang et al. (2023a) developed a theoretical wetting-front model based on high-density resistivity method tests of rare-earth ores. This model was used to estimate wetting-front height and provided guidance for optimizing injection and recovery parameters. Yin et al. (2018) investigated the permeability characteristics of ore–soil materials with different pore structures to elucidate the formation of extensive blind or stagnant zones caused by poor seepage conditions. They further developed a permeability-coefficient model to support field operations. Guo et al. (2018) examined dead corners and blind zones formed by overly sparse injection-well networks. They proposed a characteristic-value model for wetted bodies, providing practical guidance for well-network design. Zhao et al. (2018), analyzed in-situ leaching processes and seepage mechanisms, concluded that the primary cause of leaching blind zones lies in the injection stage. Shi (2023) investigated both blind-zone formation and slope instability associated with inappropriate injection-parameter design. An in-situ leaching seepage model was developed and applied to guide field experiments. Overall, previous studies have primarily focused on reducing blind zones by elucidating seepage mechanisms, optimizing injection parameters, and improving leaching processes. However, reports on the direct detection and identification of blind zones in mines remain scarce.

High-density resistivity method surveying is a modern geophysical exploration technique that employs array-based electrode configurations and high-precision data acquisition systems. It enables non-destructive, efficient, and accurate characterization of subsurface structures. Owing to its capability for multi-parameter data acquisition, this method has been widely applied in engineering investigations, including slope stability monitoring and the detection of underground cavities and groundwater (Bao et al., 2020; Jia et al., 2022; Kang et al., 2023; Yang et al., 2024). Olawale et al. (2020) employed high-density resistivity method to investigate soil contamination at a waste disposal site in Nigeria and successfully delineated the contamination plume in the central area of a cassava wastewater and waste disposal site. Jia et al. (2022) applied high-density resistivity method to identify collapse zones in mined-out areas and distinguished them from the surrounding strata based on their relatively low resistivity. Bongkoch et al. (2021) used electrical geophysical techniques to evaluate the potential for refuse-derived fuel recovery from an open dumpsite in Thailand. Their results revealed a strong correlation between electrical resistivity and plastic content, demonstrating the potential of this approach for identifying fuel-rich zones within waste deposits. Mohamed et al. (2022) investigated an abandoned mine in South Dakota, USA, and successfully used high-density resistivity method to distinguish dry tunnels from water-filled tunnels. Jabrane et al. (2023) applied high-density resistivity method to characterize the two- and three-dimensional geometries of sediment-filled sinkholes associated with interconnected fractures and joints in a karst region of southeastern Spain. Similarly, Jendruś et al. (2023) successfully employed high-density resistivity method to locate loosened zones beneath a coalfield, which were characterized by relatively low resistivity. In addition to these applications, several studies have investigated the application of high-density resistivity method to ion-adsorption rare earth mines. Chen et al. (2022) applied two-dimensional and three-dimensional high-density resistivity surveys to investigate deep seepage pathways in rare-earth mines. They found that three-dimensional surveys provide more comprehensive subsurface information. Min et al. (2022) used high-density resistivity method to investigate deep zones within a rare-earth mining area and successfully delineated the spatial distribution of deep fracture structures. Zhong et al. (2023) employed high-density resistivity surveying for stratigraphic delineation in an ion-adsorption rare-earth deposit and reported an exploration error of approximately 1.75 m. Chen et al. (2023) applied high-density electrical methods to production-stage exploration in rare-earth mines, providing technical support for refined exploration. Huang et al. (2024) used high-density resistivity methods to monitor leaching-solution flow paths in rare-earth mines, enabling visualization of the solution recovery process. Li and Zhu, 2024 reviewed the applications and future prospects of ground-penetrating radar and high-density resistivity methods in rare earth mining. They highlighted that emerging geophysical techniques provide important support for resolving ore-body structures and optimizing injection-well network parameters. Zhong et al. (2023) combined high-density resistivity surveys with borehole data to accurately delineate the distribution of bedrock and aquifers. Overall, previous studies have primarily focused on the application of high-density resistivity methods to characterize stratigraphic structures and leaching-solution flow paths. However, limited attention has been given to detecting blind zones formed during the leaching process. Therefore, this study applies high-density resistivity methods to identify leaching blind zones formed during in-situ leaching of ion-adsorption rare-earth deposits and to elucidate their formation mechanisms.

This study investigates an ion-type rare-earth mining block in southern China. High-density resistivity imaging and soil moisture sensors are employed to monitor apparent resistivity and volumetric water content in situ and in real time along a representative ore-body profile. By integrating the monitoring results with the injection history, precipitation data, groundwater level variations, and relevant geological information, the spatial distribution of leaching blind zones within the rare-earth ore body were analyzed. The findings provide technical guidance for optimizing injection parameters and improving resource recovery efficiency.

2 Engineering overview

2.1 Geological and hydrogeological characteristics of the study area

The experimental site is located at the intersection of the Nanxiong–Zhoutian first-order structural belt in the eastern segment of the east–west–trending Nanling fold system and the Guangchang–Dingnan north–northeast–trending structural belt of the Neocathaysian system. It is situated in the central part of the Zhaibei granite body. The structural setting of the study area is relatively simple and is characterized mainly by minor faults and fractures. The stratigraphy is uniform. Quaternary deposits occur only in valley areas, and no other strata are exposed. Igneous rocks are widely exposed and consist entirely of acidic granite. These granites form the principal parent rock of the weathered-crust ion-adsorption rare-earth deposit, and a well-developed weathering crust is present. In the vertical profile, the weathering crust exhibits clear zonation in lithology, structure, and material composition.

From top to bottom, the weathering crust can be divided into topsoil, a completely weathered layer, a partially weathered layer, and bedrock (Li, Y., 2014). As shown in Figure 1, the geological structures of rare-earth deposits indicate that ion-type rare-earth ores are primarily hosted within the completely weathered layer. Therefore, this study focuses on the completely weathered layer and upper portion of the semi-weathered layer (Li et al., 2026).

The study area is in a subtropical, monsoon-influenced region of southeastern China. The long-term mean annual air temperature is 19.5 °C, and the frost-free period lasts for 302 days. The mean annual precipitation is 1560.2 mm, with an average annual maximum of 2072.0 mm and an average annual minimum of 1099.0 mm. The maximum daily precipitation reached 335 mm was recorded on 5 May 2010. The mean annual evaporation is 1636.8 mm, with a maximum of 1850.0 mm and a minimum of 1417.9 mm. According to precipitation statistics from the Jiangxi Provincial Meteorological Bureau (2024) show that rainfall in August 2024 was approximately 2.3 times that in September 2024.

Surface water in the study area primarily consists of seasonal valley streams. Stream discharge typically ranges from 0.052 to 0.66 L/s and is strongly influenced by precipitation. The discharge is high during the rainy season, decreases sharply in the dry season, and may occasionally cease. The surface water and groundwater in the overlying aquifers are hydraulically connected and mutually exchanged water. Overall, groundwater discharges primarily into surface water. It emerges as springs in low-lying areas at the foot of the hills and recharges nearby surface channels and small streams through short-duration runoff. The groundwater recharge within the study area is mainly dominated by rainfall infiltration. This recharge process is controlled by seasonal conditions, lithology, and the degree of rock fracturing. Rainfall infiltration and downward percolation occur more rapidly in areas with high permeability and well-developed fractures, which leads to higher groundwater levels or increased groundwater discharge during wet periods.

2.2 Principle of in situ leaching

The experimental block was treated using an in-situ leaching process. During this process, rare-earth cations adsorbed on the surfaces of clay and other minerals are displaced by magnesium ions (Mg²⁺) from a magnesium sulfate leaching solution through ion-exchange reactions. Consequently, magnesium ions are retained within the ore body, while rare-earth ions are released into the leachate. The leachate along the bedrock surface toward the liquid collecting ditch at the foot of the hill to recover rare earth elements (Wang et al., 2024a).

2.3 Injection parameters of the test ore block

The experimental block has an average topsoil thickness of approximately 2 m. Injection wells were arranged in a 2 m × 2 m grid, with depths ranging from 4 to 5 m. 2% magnesium sulfate solution was used as the leaching agent. The monitoring period extended from August 6 to 2 November 2024. Leaching began on August 7, with three rows of injection valves at the slope crest were opened. On August 8, the number of active valve rows increased to five. By August 9, seven rows of injection valves were in operation. Leaching was suspended from August 13 to September 2 and resumed on September 3. On September 11, the injection intensity was reduced. Leaching was temporarily suspended from September 15 to 16 and again from September 19 to 22. On September 23, all injection valves on the gentle-slope side were opened, while those on the steep-slope side were completely closed. On October 7, injection of the clarified solution commenced. All injection activities were terminated on November 2. The injection parameters for the experimental block are summarized in Table 1.

TABLE 1. Injection parameters for the experimental block.

2.4 Physical properties of ore soil

To characterize the properties of the soil within the experimental block, moisture sensors were installed in advance along the electrical survey lines. Soil samples were collected at monitoring points using hand-auger drilling. The field testing setup is shown in Figure 2. Soil samples were tested for boundary moisture content and particle size analysis in accordance with the Geotechnical Engineering Test Method and Criterion (GB/T 50123-2019, 2019). Particle-size analysis was performed using a combined sieving method and laser particle size analyser (HL2020-B). The basic physical properties and particle-size grading of the ore-bearing soils are summarized in Tables 2 and 3, respectively. Following the saturated hydraulic-conductivity test method described by Wang (2023b) for rare-earth ores, the saturated hydraulic conductivity at the monitored points within the experimental block was determined as (Ks = 0.0488 cm/min). In addition, van Genuchten curve-fitting parameters previously obtained for the ore-bearing soils from the same ion-adsorption rare earth ore were (α = 0.0086) and (n = 1.5587) (Wang et al., 2024b).

TABLE 2. Basic physical properties of the soil samples.

TABLE 3. Particle size gradation of the test samples.

3 Field testing methods

3.1 Testing principle and data processing

This study employed a DEM-M1 multiparameter direct-current resistivity monitoring system to investigate variations in apparent resistivity during the leaching process of an experimental block in an ion-adsorption rare-earth ore. High-density resistivity surveying is a geophysical exploration technique based on resistivity measurements. Its fundamental principle involves densely arranged electrode arrays that measure the spatial distribution of apparent resistivity in subsurface materials. These measurements are then used to infer subsurface geological structures (Bi et al., 2020). The apparent resistivity is calculated using the following Equation 1 (Xiao et al., 2024):

ρ s = K Δ U M N I (1)

In this equation, ρ_s denotes the apparent resistivity (Ω·m) and represents the overall electrical conductivity of the subsurface medium. K is the geometric factor determined by the electrode configuration. ΔU_MN represents the potential difference between the M and N electrodes (V). I denotes the electric current injected into the ground (A).

During in-situ leaching, moisture content within the ore body exhibits substantial spatial heterogeneity. This heterogeneity leads to corresponding variations in resistivity at different locations. Leaching blind zones correspond to regions of ore–soil that are not penetrated by the leaching solution and therefore remain at their natural moisture content. As a result, their apparent resistivity differs markedly from that of saturated ore–soil. High-density resistivity surveys exploit these electrical contrasts to characterize the subsurface resistivity distribution. This information is then used to infer the spatial distribution of leaching blind zones. In this study, electrical resistivity data were acquired using a Wenner–α array. Using the Wenner–α depth calculation coefficient of 0.173 provided by Loke (2004), the theoretical investigation depth for the 180 m survey line was calculated to be 31.14 m.

Raw monitoring data were exported from the online electrical monitoring platform and imported into the RES2DINV software for initial inversion. During the inversion, the initial damping factor was set to 0.15, with a minimum damping factor of 0.02. The resistivity values were constrained to a range of 1–300000 Ω m. The inversion results were then converted into Surfer-compatible files for further processing and visualization. The processed data were re-plotted in Surfer to generate high-resolution vector images. These images provide detailed inverted resistivity profiles along the survey lines throughout the leaching process.

Topographic correction was performed using the RES2DINV high-density two-dimensional inversion software. First, the observational data exported from the geophysical monitoring platform in.dat format with the corresponding topographic elevation values. The updated dataset was then imported into RES2DINV, where a least-squares inversion scheme with a deformed finite-element mesh was used to perform topographic modeling correction. During forward modeling, RES2DINV incorporates the true undulating topography as boundary conditions in the forward modeling stage to compute the electric-field distribution under the specified terrain. A smoothness constraint is applied within a least-squares iterative framework during the inversion. The inversion fits the measured apparent resistivity and the topography-induced electric-field distortion simultaneously. This approach suppresses spurious terrain effects-associated anomalies and produces a reliable subsurface resistivity model. The inversion results were exported as Surfer standard input files. Surfer was then used to redraw high-resolution vector maps and generate high-resolution inverted resistivity cross-sections along the geophysical survey line throughout the leaching process.

3.2 Survey line layout

The survey-line layout, number of survey lines, and electrode spacing were determined based on the ore-layer thickness and the required investigation depth. As shown in Figure 3, a single survey line was deployed perpendicular to the ridge. The line had a total length of 180 m and consisted of two cable sections. Each cable section was equipped with 30 copper electrodes, spaced at 3 m intervals.

The monitoring period extended from August 6 to 2 November 2024. The monitoring data were interpreted in conjunction with field observations, and the corresponding inversion results are presented in Figure 4.

3.3 Layout of soil moisture sensors

Three moisture sensors were installed along the survey line. The sensors were labeled 6#–1 m, 7#–10 m, and 9#–1 m. The label 6#–1 m indicates that the sensor was installed in monitoring borehole No. Six at a depth of 1 m below the ground surface. Sensor 7#–10 m was installed 10 m below the ground surface near the crest of the experimental block. The remaining two sensors were installed 1 m below the ground surface near the slope toe. The surface locations and profile distribution of the sensors are shown in Figure 5. The slash-filled region in Figure 5 is defined as the spatial distribution of ore bodies delineated along the high-density resistivity method survey line profile.

4 Experimental results

4.1 Evolution of moisture content

Figure 6 presents the moisture-monitoring results recorded by the three sensors during the leaching process in the experimental block. These data allow evaluation of both the extent and timing of leaching solution infiltration at each sensor location.

Before leaching, the initial volumetric water contents measured by sensors 6#–1 m, 7#–10 m, and 9#–1 m were 19%, 18%, and 20%, respectively. These values indicate that the ore-bearing soils surrounding the sensors had not yet been penetrated by the leaching solution. The volumetric water contents measured by sensors 6#–1 m and 9#–1 m remained close to their initial levels throughout the leaching period, with no pronounced fluctuations. This observation indicates that the leaching solution did not effectively reach these locations, revealing a distinct shallow blind-zone phenomenon. This formation of these shallow blind zones is mainly attributed to three factors: the absence of injection wells within the safety pillar, the relatively large depth of nearby injection wells, and the dominance of vertical seepage over lateral spreading. Near either sensor location, no injection wells were installed at the surface. In contrast, adjacent areas were equipped with injection wells at depths of 4–5 m. Combined with the high permeability of the ore–soil, these conditions caused the injected solution to migrate rapidly downward from the base of the wells. Because lateral spreading was comparatively slow, leaching solution from adjacent zones could not sufficiently infiltrate the ore–soil within the safety pillar. Furthermore, the difficulty in maintaining a stable solution level within the injection wells further limited insufficient infiltration in the shallow zones.

The moisture response recorded by sensor 7#–10 m was the most pronounced among the three monitoring points. Following the initial injection on August 7, the volumetric water content at this location rapidly increased to 64%. As the injection intensity was decreased, the volumetric water content gradually declined and returned to its initial level after injection suspended. This response indicates a high sensitivity of the sensor to changes in injection intensity. When injection resumed on September 3, the moisture content rapidly increased to 100%, indicating that the ore–soil at this depth had reached saturation. Subsequent monitoring showed that the water content decreased when the injection intensity was reduced on September 11 and during injection interruptions on September 15 and September 19. In contrast, it increased rapidly once injection resumed. This consistent temporal correspondence between moisture variations and injection operations demonstrates that the sensor effectively captured real-time changes in the seepage field.

The moisture-sensor data clearly reflect the infiltration conditions of the ore–soil at different depths. They also accurately capture variations in injection intensity and injection duration. Under the current configuration, injection wells are relatively deep, and the ore–soil exhibits high permeability. As a result, the leaching solution does not adequately reach the shallow zones at the slope toe. This results indicates that the existing injection strategy leads to the formation of pronounced shallow leaching blind zones. It therefore highlights the urgent need to optimize injection parameters.

4.2 Evolution of leaching blind zones along the survey line profile

Figure 4 presents the inverted apparent resistivity profiles obtained during the leaching process of the experimental block. These inversion results enable clear tracking of leaching-solution migration pathways, the formation of leaching blind zones, and the evolution of low-permeability zones at the base of the ore body. As shown in Figure 4, the leaching blind zones exhibit a distinct stage-wise evolution pattern. Their dynamic evolution is jointly controlled by injection parameters, slope geometry, and geological structure.

Precipitation in the mining area was relatively concentrated in August 2024 (as reported in the 2024 Jiangxi climate bulletin). Therefore, the soil in the mining block exhibited an overall high moisture content. The inversion results (Figure 4a) indicate that the upper part of the ore-bearing soil is characterized by a green zone with an absolute resistivity of approximately 1200 Ω m. Before leaching, the three moisture sensors were located within the green zone, where the natural moisture content ranged from 18% to 20%, corresponding to a resistivity of approximately 1200 Ω m. The leaching blind zone refers to areas that have not been affected by leaching. Therefore, a moisture content of 18%–20% and a resistivity of 1200 Ω m can be used as thresholds value for identifying such zones. The injection of large volumes of leaching solution or rainfall is expected to increase the volumetric water content of the ore body and, consequently, decrease its resistivity. Accordingly, areas with a volumetric water content ≤20% and a resistivity ≥1200 Ω m can be inferred to represent unleached portions of the ore body, i.e., leaching blind zones. Several “light-green” high-resistivity zones are observed near the ridge and along parts of the right steep slope, with a maximum absolute resistivity of approximately 2800 Ω m. Scattered “yellow” high-resistivity anomalies occur within the light-green region on the right slope, with a maximum absolute resistivity of approximately 4500 Ω m. A “light-blue” low-resistivity zone occurs at the foot of the right slope, with an absolute resistivity of about 180 Ω m. This low-resistivity zone is likely related to abundant groundwater runoff and the preferential migration of tailings water from the adjacent depleted mining block. In addition, three “cyan” low-resistivity zones were identified in the lower part of the ore body, with a maximum absolute resistivity of approximately 340 Ω m. These features are preliminarily interpreted as being associated with groundwater seepage. The three cyan zones indicate relatively fragmented geometries. The groundwater flow through the fractured zones likely reduces the apparent resistivity compared with the surrounding materials. A green high-resistivity body occurs between the cyan zones at a horizontal distance of approximately 60 m (Figure 4a). This green body is inferred to represent a more intact rock mass with a lower degree of weathering based on the stratigraphic characteristics of the rare-earth deposit and regional geological information. Owing to its relatively low permeability and low moisture content, its resistivity is higher than that of the surrounding fractured zones. A steep cut face was formed at the foot of the right slope during the construction of the collection ditch. The excavation height was approximately 3 m, and the slope angle was close to 65°. Prior to leaching, this cut face corresponds to a “red” high-resistivity zone with a maximum absolute resistivity of about 12000 Ω m (field excavation characteristics shown in Figure 7). Overall, the inversion results indicate that the main body of the ore exhibited moderate to relatively low apparent resistivity. These areas appear as green zones in the profiles and reflect the background electrical characteristics of the weathered crust under natural conditions.

During the early stage of leaching, the number of injection-valve rows opened along the ridge increased from three to five. Under the combined effects of gravitational potential and seepage pressure, the leaching solution migrated preferentially downward along the ridge axis. Because capillary forces were relatively weak, lateral spreading of the solution was limited. As a result, irregular low-resistivity anomalies developed in the shallow zone. These anomalies correspond to localized infiltration areas formed during the initial stage. At this stage, infiltration was mainly confined to the middle and lower parts of the ore body. The upper weathered crust remained insufficiently wetted, as shown in Figure 4b. During the injection suspension in August, the inverted absolute resistivity distribution of the ore body showed only minor changes. This pattern is consistent with the monitoring results obtained on September 3 before injection resumed (Figure 4c).

In September 2024, precipitation in the mining area decreased, the groundwater level declined, and the duration of sunshine increased as the number of rainy days decreased. In addition, the volume of water released from the adjacent depleted mining block decreased. Consequently, the moisture content near the near-surface soil began to decline, which increased the resistivity. The inversion results show an overall increase in resistivity across the mining area from August to early September, except in the leaching-affected zones.

On 11 September 2024, the resistivity distribution in the inversion results exhibited a marked change compared with that in the previous stage. Specifically: (1) The resistivity values of the high-resistivity zones located in the middle portion of the right slope and near the slope toe increased substantially. Their maximum absolute resistivities were approximately 15,000 Ω m and 4,000 Ω m, respectively. This increase is mainly attributed to reduced rainfall and enhanced soil moisture evaporation. (2) Two previously separate circular high-resistivity anomalies connected into a single feature on the left side near the slope foot, with an absolute resistivity of approximately 1,000 Ω m. These two anomalies corresponded to two alley entrances. The alley entrances collapsed and the cross-sectional area increased after earlier rainfall, resulting in a changed resistivity pattern in this region. (3) A “yellow” high-resistivity zone appeared in the lower part of the left slope, with a maximum absolute resistivity of approximately 3,500 Ω m (4) A “green” zone was observed in the deeper subsurface, with an absolute apparent resistivity of approximately 1,200 Ω m. The increase in deep resistivity primarily reflects the groundwater level decline, which agrees with the pre-leaching geophysical monitoring results. In the inversion profile acquired on August 6, the resistivity in the same deep region was also higher than that in the surrounding area. Based on the geological structural characteristics of ion-adsorption rare earth ores reported by Li, (2014), this region is preliminarily interpreted as a rock mass with incomplete weathering, showing a lower degree of weathering than the surrounding rocks, As injection valves on both sides of the ridge were opened sequentially (Figures 4c,d), the leaching solution began to spread laterally under the influence of capillary forces. More injection boreholes were opened on the left side of the ridge, so a larger volume of leaching solution was injected into the left portion of the ore body. Consistent with this injection pattern, the inversion results also show that the infiltrated area on the left side was greater than that on the right. As a result, the deeper parts of the ore body became effectively wetted, whereas the shallow zones remained unsaturated. During this stage, shallow ore zones near the ridge were not infiltrated, leading to the early formation of the first leaching blind zone.

During the middle stage of injection (Figures 4e–g), all injection valves on the left slope were fully opened, whereas those on the steep slope were completely closed. This configuration caused the seepage field to shift toward the left. The leaching solution continued to migrate downward and laterally toward the left side. In contrast, seepage on the right side largely stagnated, leading to the formation of the second leaching blind zone. Because lateral spreading of the leaching solution remained limited, the first leaching blind zone was not effectively infiltrated during this stage. Meanwhile, the toe of the left slope along the survey profile functioned as a safety pillar where no injection wells were installed. The absence of injection wells, combined with limited lateral spreading, a large portion of the ore body in this area remained un-infiltrated for an extended period. This process ultimately led to the formation of the third leaching blind zone.

The inverted apparent resistivity distribution remained largely stable after injection began in October 2024. Rainfall decreased again, and the groundwater level continued to decline. The large volume of leaching agent injected into the mountain body, together with the headwater supply, continued to infiltrate toward the lower parts and spread laterally along the sides. Because the mine soil has relatively high permeability, the leaching solution near blind zones one, two, and three showed predominantly vertical infiltration with limited lateral diffusion during the inversion results on November 2. The spatial characteristics of these three blind zones remained essentially unchanged. Meanwhile, the liquid within the ore body continued to migrate toward the groundwater system and formed stable seepage pathways. In contrast, the deep zone beneath the left slope, which has poor permeability, consistently appeared as a high-resistivity area. This pattern is likely related to the decline in groundwater levels, the persistence of fixed seepage pathways, and pore clogging caused by particle migration. The results from three blind regions indicate that the leaching agent primarily migrates rapidly along vertical preferential flow paths. In contrast, the near-surface and edge regions form “blind zones” due to limited lateral diffusion of the leaching agent (Figures 4H–K). A red low-resistivity zone is observed at the bottom of the ore. Based on the geological structure and mineralization characteristics of the deposit (Li, 2014), this zone is interpreted to correspond to the semi-weathered layer or underlying bedrock and is therefore beyond the scope of the present study. In ion-adsorption rare earth ores, geological exploration is still commonly conducted using manual Luoyang spade drilling. The maximum exploration depth generally extends only to the bottom of the completely weathered layer because further drilling into the fractured semi-weathered layer is difficult. Consequently, geological data from the deeper portions of ion-adsorption rare earth ores remain limited, making it difficult to independently verify the nature of the low-resistivity zone at the ore bottom of the ore body.

Dynamic evolution of leaching blind zones and wetting bodies. The spatiotemporal evolution of leaching blind zones reflects the combined effects of injection layout, slope geometry, and ore–soil permeability. Under the current injection strategy, infiltration into shallow and peripheral zones is clearly insufficient. This finding underscores the need to optimize injection well placement and regulate injection timing to reduce the extent of leaching blind zones.

5 Discussion

5.1 Dynamic evolution of leaching blind zones and wetting bodies

This study indicates that real-time high-density resistivity monitoring, validated by in-situ moisture sensors, provides an effective method for accurately identifying and dynamically tracking leaching blind zones and infiltrated regions during in-situ leaching of ion-adsorption rare-earth deposits. The integrated approach provides technical support for addressing nonuniform leaching behavior. As shown in Figure 4, the inverted resistivity images clearly reveal the spatiotemporal evolution of the apparent resistivity field within the ore body. These results effectively visualize subsurface seepage processes that are otherwise difficult to observe directly. Throughout the injection period, infiltrated areas gradually transitioned from high-resistivity green zones to continuous low-resistivity blue zones. In contrast, leaching blind zones consistently exhibited as high-resistivity red zones. These resistivity patterns closely correspond to the moisture responses recorded by the co-located sensors (6#–1 m and 9#–1 m). The integration of electrical resistivity and moisture data provides a combined electrical–hydrological criterion that substantially improves the reliability and accuracy of blind-zone identification.

Based on the monitoring data, the coupled evolution of infiltrated zones and leaching blind zones is systematically characterized. The dynamic process can be divided into three representative stages. (1) Vertically dominated stage. Driven by both gravity and capillary forces, the leaching solution spreads outward, with gravity playing the dominant role. The solution migrates rapidly downward along the ridge axis and forms isolated, ellipsoidal infiltration bodies. During this stage, blind zones are mainly distributed around the margins of the infiltrated bodies and within shallow, undisturbed areas. (2) Lateral spreading stage. As the injection area expands, the influence of capillary forces become dominant and the infiltration bodies extend laterally. However, ore-body heterogeneity and topographic conditions induce pronounced seepage anisotropy. This leads to strong contrasts in infiltration between gentle and steep slopes. As a result, spatially persistent shallow blind zones develop near the ridge crest, along with blind zones associated with safety pillars. (3) Deep connectivity with shallow stagnation stage. Although deep ore zones become fully connected through bypass flow and form continuous low-resistivity pathways, shallow zones and safety-pillar regions lack sufficient driving forces for infiltration. Consequently, the upper portions of the blind zones remain in a high-resistivity state and exhibit pronounced seepage stagnation. Owing to its high resolution, non-invasive nature, and time-lapse inversion capability, high-density resistivity surveying directly captures these dynamic evolutionary characteristics. The method provides direct observational evidence for understanding the complex, multi-physics processes involved in in-situ leaching. Its results can be directly applied to real-time optimization and adjustment of field operating parameters.

The findings of this study can provide guidance for the exploitation of ion-adsorption rare earth deposits not only in China but also in other countries where similar deposits occur, such as Malaysia and the Philippines (Cris et al., 2022; Sobri et al., 2023). Electrical geophysical methods identify subsurface targets primarily by detecting anomalies in electrical resistivity. Based on the same principle, these methods can also be applied to the investigation of underground infrastructure, subsurface cavities, archaeological structures, fault zones, and hydrogeological features (Jullian and Melchizedek, 2022).

5.2 Mechanisms controlling the formation of leaching blind zones

During in-situ leaching of ion-adsorption rare-earth deposits, ore-body heterogeneity, complex topography, and suboptimal injection-parameter design often lead to the formation of leaching blind zones. In these zones, the leaching solution cannot effectively infiltrate the ore body. As a result, residual resources remain, reduced recovery efficiency, and prolonged leaching cycles. By integrating time-lapse high-density resistivity imaging with in-situ moisture sensor data, this study infers three representative leaching blind zones. These zones are located at the ridge crest of the ore-body profile, the middle portion of the steep right slope, and the toe of the gentle left slope (Figures 4, 8). The spatial delineation of blind zones is based not only on electrical anomalies but also on validation using dynamic moisture-monitoring data. This dual-validation approach substantially improves the accuracy and reliability of leaching blind-zone identification.

The inverted high-density resistivity results indicate that the three persistent leaching blind zones exhibit pronounced high-resistivity anomalies. Their apparent resistivity values are markedly higher than those of the surrounding infiltrated areas. During leaching, the injection of a low-resistivity solution into the highly porous weathered crust significantly reduces the overall apparent resistivity of the ore body (Si et al., 2026). Therefore, the persistence of high resistivity indicates that the blind zones are not effectively wetted. Water saturation in these areas remains largely unchanged, and ion-exchange reactions are essentially inactive. This interpretation is supported by moisture-sensor measurements. Sensors 6#–1 m and 9#–1 m, located within high-resistivity regions, recorded volumetric water contents that remained close to their initial values (approximately 20% RH) throughout the monitoring period. No response to injection activities was observed at these locations. Together, the resistivity and moisture data establish a mutually validating electrical–moisture criterion for reliable blind-zone identification. By integrating injection parameters with geological conditions, the formation mechanisms of leaching blind zones can be systematically classified according to their spatial locations, as discussed below.

• Shallow ridge-top blind zone: This type of blind zone is primarily distributed within the shallow portion near the crest of the ore body. During the initial stage of solution injection, the leaching solution infiltrates the ore under the combined effects of substrate suction and gravity, with flow resulting in both horizontal and vertical directions (Wang, 2023). As leaching progresses, the moisture content of the ore body increases, and the effects of gravitational and pressure potentials on vertical flow become progressively more pronounced (Wang et al., 2023b; Guo et al., 2018). When their combined influence approaches that of the matric potential driving horizontal flow, gravity gradually becomes the dominant driving force controlling solution migration.

In addition, the injection holes in the experimental ore block were relatively deep (4–5 m). Field observations showed that the leaching-solution level in the injection holes did not rise to the hole openings, causing the solution to preferentially infiltrate from the bottom of the well and develop preferential flow pathways. As a result, the upper portion of the ore body remained insufficiently wetted (Lu et al., 1997). Furthermore, the injection was performed sequentially from the upper to the lower part of the slope, beginning at the ridge and subsequently extending to adjacent injection holes. Because the capillary rise height of the ore material was limited, the ore between adjacent injection holes could not be sufficiently wetted (Wu et al., 2021), ultimately resulting in the formation of leaching blind zones, as illustrated in Figure 9.

Wang (2023a) used COMSOL to simulate the leaching process during the saturation stage in an experimental stope. The results showed that the leaching-solution flow velocity near the ridge was lower than that near the toe of the slope, resulting in incomplete leaching of the ore body in the upper slope. In the present study, a shallow blind zone was observed on the left slope near the ridge. Although leaching solution continued to be injected through the slope injection holes after the flow field within the ore body had reached a relatively stable state, the blind zone persisted. This observation suggests that the formation of Blind Zone one may be related to the relatively low flow velocity of the leaching solution in this region. Zhang (2023) employed HYDRUS-2D to investigate the effects of injection-hole spacing and diameter on the predicted area of leaching blind zones. The results indicated that the blind-zone area initially decreased and then increased with increasing hole spacing. At a given hole spacing, a larger hole diameter resulted in a smaller blind-zone area over the same leaching period. These findings indicate that the formation of leaching blind zones are strongly influenced by injection parameters. In the present study, the spacing and diameter of the injection holes were determined empirically by mine workers and were not validated through experimental testing or numerical simulation. A schematic illustration of the leaching blind zones is presented in Figure 9.

• Safety-pillar blind zone: This blind zone develops within the safety-pillar area and is characterized by the absence of surface injection wells. Its formation is controlled by both injection-system design and local geological conditions. First, the lack of injection wells results in insufficient direct hydraulic driving force for infiltration. Second, leaching solution from injection zones migrates rapidly downward under gravitational control, while lateral spreading remains limited (Wang et al., 2018; Guo et al., 2018). Consequently, the base of this blind zone becomes hydraulically connected to the slope toe. This connection forms a potential slip surface and increases the risk of slope instability. Although the deeper ore body becomes infiltrated, the overlying zone remains poorly wetted due to the absence of direct seepage and slow lateral diffusion. As shown in Figure 10, the calculated areas of Blind Zones one, two, and three are 1,193, 196.8, and 2,017 m², respectively, whereas the cross-sectional area of the ore body is 7,958.6 m². Blind Zones one and three collectively account for approximately 17.5% of the total cross-sectional area of the ore body. This substantial proportion highlights the need to optimize injection parameters.

The formation of leaching blind zones fundamentally arises from an imbalance in local seepage driving forces. In these zones, the gravitational potential of the leaching solution dominates over capillary forces within the ore–soil. In heterogeneous ore–soil systems, inappropriate injection depth, insufficient injection-well coverage, and suboptimal injection timing cause certain areas to become zones of seepage stagnation within the flow field. This study integrates multiple data sources to clarify the hypothesized mechanisms underlying blind-zone formation. It also provides a theoretical basis for optimizing injection strategies and achieving targeted leaching.

5.3 Limitations

During in situ leaching of ion-type rare-earth deposits, temporal variations in resistivity are much more complex than soil resistivity fluctuations caused solely by precipitation. In situ leaching is a long-term process in which resistivity variations are influenced not only by external factors, such as rainfall, groundwater depth, and injection interference from adjacent mining blocks, but also by leaching-related processes. These include the ionic composition of the leaching agent, ion-exchange reactions within the ore–soil, and injection conditions, such as intensity, sequence, and extent. Therefore, future studies should consider these multiple influencing factors and provide a comprehensive, systematic analysis of the formation and evolution mechanisms of leaching blind zones. In this study, a considerable amount of liquid remained in the ore body after the completion of the geophysical monitoring, and the soil moisture content exceeded its liquid limit. Under these conditions, the Luo-yang spade could not retrieve intact, slurry-like soil samples from the boreholes for verification. In addition, borehole verification could not be conducted due to budget constraints and the inability of mechanical drilling rigs to access the vegetated test mining block.

In the present study, moisture sensors 9#-1 m and 6#-1 m were installed at relatively shallow depths. Thus, the measured moisture content values represent conditions at specific points within the ore-body profile rather than the spatial characteristics of the entire profile. In contrast, the high-density electrical resistivity tomography inversion image provides the resistivity distribution across the entire survey profile. Consequently, the point-scale moisture-sensor measurements may not fully represent the resistivity characteristics of the surrounding ore body. Therefore, the boundaries of leaching blind zones identified in regions containing pre-installed moisture sensors should be regarded as inferred rather than directly measured boundaries. The geological structure of ion-adsorption-type rare earth deposits generally consists, from top to bottom, of a topsoil layer, ore layer, semi-weathered layer, and bedrock (Li, 2014). In the electrical resistivity inversion profile obtained in this study, a relatively extensive red high-resistivity zone occurs at the bottom of the profile. Inferring this zone solely from the moisture-sensor measurements could result in substantial misinterpretation. Based on the typical geological structure of ion-adsorption rare earth ores, the deeper portions of the profile are likely to comprise semi-weathered or fresh rock, which generally exhibits high resistivity. Therefore, the red high-resistivity zone at the bottom of the profile is interpreted as fresh rock or rock inclusions rather than a leaching blind zone.

The occurrence and spatial extent of leaching blind zones are inherently stochastic, making it difficult to predict their locations and boundaries in advance. Consequently, moisture sensors cannot be reliably pre-installed along the boundaries or within the central areas of potential leaching blind zones. During in situ leaching, the injection of large volumes of leaching solution is a major factor affecting the resistivity of the ore body and generally results in a substantial decrease in resistivity. Natural rainfall can similarly reduce the resistivity of near-surface soil. Both processes contribute to a reduction in ore-body resistivity, while other factors have little influence on the resistivity of ore. Therefore, high-resistivity zones identified within the ore body are inferred to represent regions that have not yet been leached.

Accordingly, future work should increase the density of sensors preinstalled at different locations and depths along the survey lines to improve the validation and reliability of the geophysical monitoring results.

6 Conclusions and recommendations

This study integrates time-lapse high-density resistivity imaging with in-situ moisture sensor monitoring. The combined approach systematically reveals the spatiotemporal evolution and formation mechanisms of leaching blind zones during in-situ leaching of ion-adsorption rare-earth deposits. Based on these findings, corresponding optimization strategies are proposed. The main conclusions are summarized as follows.

• High-density resistivity surveying effectively reveals changes in the geometry of infiltrated zones during leaching. As injection progresses, the morphology of these zones evolves through distinct stages. This evolution reflects the combined influence of injection sequence, injection intensity, slope geometry, geological structure, and engineering layout on the seepage field.
• Two representative types of leaching blind zones were inferred by integrating the electrical geophysical inversion results with the moisture-sensor monitoring data: a shallow blind zone near the slope crest and a blind zone associated with the protective pillar near the slope toe. Blind Zone one represents an inferred region, whereas Blind Zones two and three were inferred from the moisture-sensor monitoring results. The shallow blind zone near the slope crest is mainly attributed to the dominance of gravitational potential during leaching-fluid infiltration. The matric potential is insufficient to drive the leaching agent into shallow mineral-bearing soils under these conditions. The primary factor for the blind zone associated with the protective barrier pillar is the absence of designed injection boreholes at the ground surface, which reduces the availability of direct seepage-driven forcing. Moreover, the leaching agent’s gravitational potential in the mineral soil exceeds the matric potential, leading to a vertical seepage velocity greater than the lateral diffusion rate. This promotes the bypass flow of the leaching agent around the blind zone near the protective barrier pillar base. The electrical signatures of the blind zones are highly consistent with the moisture-sensor observations, indicating that these regions are characterized by high electrical resistivity. In future electrical resistivity monitoring studies, additional moisture sensors should be deployed at multiple locations along the survey line to provide more comprehensive validation of the electrical resistivity inversion results.
• The formation of leaching blind zones is fundamentally attributed to a local imbalance between seepage driving forces and ore–soil permeability. For shallow blind zones, several measures are recommended to improve infiltration efficiency. These include increasing the number of shallow injection wells, applying controlled surface spraying, and backfilling injection boreholes with soil. For safety-pillar blind zones, an “upper injection–lower drainage” strategy is proposed, provided that slope stability and structural safety are ensured. This strategy involves moderately increasing injection wells to achieve targeted infiltration and drilling drainage holes at the slope toe. These measures can promote active leachate discharge, reduce pore-water pressure, and mitigate the risk of slope instability.

The findings of this study provide a scientific basis for the visual monitoring and mechanistic understanding of leaching blind zones in ion-adsorption rare-earth mining. The proposed strategies for recovering residual resources from these blind-zone provide practical engineering guidance for improving resource recovery efficiency and promoting environmentally responsible, efficient, and safe mining operations.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Sihong Wu, Massachusetts Institute of Technology, United States

Reviewed by: Weichen Zhan, The University of Texas at Austin, United States

Jing Xie, Central South University, China