Jikui Hua,
Yufei Dengb,
Lingfen Wangb,
Huang Wanga,
Liangliang Zhaoc,
Liling Luc,
Zhongjian Liuc,
Qiang Wang*a and
Minshan Shi*a
aKey Laboratory of Coal Cleaning Conversion and Chemical Engineering Process, Xinjiang Uyghur Autonomous Region, College of Chemical Engineering, Xinjiang University, Urumqi, Xinjiang 830017, China. E-mail: wangqiang@xju.edu.cn; 458804348@qq.com
bGeology Institute of China Chemical Geology and Mine Bureau, Beijing, 100101, China
cSDIC Xinjiang Luobupo Potash CO., Ltd, Hami, 839000, China
First published on 26th August 2025
To address the technical challenges of monitoring brine seepage in salt lakes, this study pioneers the application of 10B isotope tracer to seepage detection, establishing a high-precision monitoring system that provides scientific foundations for precise seepage channel identification and flow field characterization. Systematic laboratory experiments validate the exceptional performance of the 10B tracer, including ultra-trace detection sensitivity of 10−9, a stable recovery rate of 92.8–106.5%, adsorption loss below 6.31%, and light transmittance and acid-base resistance retention rates exceeding 95%, confirming its reliability and applicability in complex high-salinity brine systems. Compared to the fluorescein control test, the 10B tracer demonstrated greater advantages in adsorption rate and buffered pH resistance. Mobility experiments further reveal inert-like migration characteristics of the 10B tracer across media, with a breakthrough curve highly similar that of chloride ions. Its rapid transport capability and low retention loss below 20% enable real-time tracking of seepage pathways, offering critical technical support for dynamic flow field analysis. Field experiments reveals that the main seepage path in the salt field leakage system is 2 → 1 → 6, dominated by a fracture system with rapid flow and high connectivity. The secondary seepage path is 2 → 1 → 4 → 6, formed by the interaction of fractures and micro-pores, characterized by fluctuating concentrations and low transmission efficiency. The rapid response and stable flow of the main path highlight the hydraulic dominance of the fracture system, while the oscillating behavior of the secondary path reflects geological heterogeneity and uneven pore distribution. For well 3, Peclet number calculation yielded Pe >1, with absent 10B tracer inflow primarily caused by an auxiliary plugging mechanism. Delayed responses and weak signals in edge wells including well 5 and well 7 further confirm permeability barriers. These findings provide critical guidance for precise seepage control engineering.
The identification of salt pan leakage poses a common technical challenge in current brine resource development, primarily manifested in three aspects. First, existing leakage monitoring techniques4,5 such as groundwater monitoring, electrical resistivity tomography, and elastic wave detection generally suffer from insufficient detection accuracy and high operational complexity. Second, the heterogeneous permeability of natural clay impermeable layers in salt pan areas, particularly the lack of engineered anti-seepage measures6 in primary evaporation units such as sodium salt ponds, exacerbates brine leakage risks. Third, dynamic hydrogeological alterations7 induced by large-scale brine extraction operations lead to spatiotemporal variability in parameters such as water abundance and permeability within mining zones, significantly complicating the precise identification of leakage pathways.
Isotope tracing technology, renowned for its strong capability in tracking material migration and high sensitivity, has been widely applied in seepage detection for hydraulic engineering projects. International studies demonstrate that Acworth et al.8 successfully measured groundwater flow rates in coastal aquifers of New South Wales, Australia, using 82Br tracer; the Grabowski team9 employed 234U/238U ratios to delineate water cycling processes in central Poland; Birkle et al.10 investigated deep groundwater movement directions in Mexico's Samaria-Sitio Grande oilfield via 18O/2H analysis; and Johnson et al.11 utilized 87Sr/86Sr ratios to study groundwater runoff in the Snake River Plain aquifer system. Zhu et al.12 tracked the magmatic process with Zr isotopes, demonstrating the primary controlling role of fluids in the fractionation of Zr isotopes during the crystal-magmatic-fluid differentiation process. Gong et al.13 used mercury stable isotopes to track volcanic activities in geological records. Tian et al.14 used potassium isotopes to trace the genesis of island arc rocks. However, a specific tracer system tailored for high-salinity brine leakage remains absent. Following the SY/T 5925-2012 standard “Guidelines for Chemical Tracer Selection in Oilfield Water Injection”,15 systematically screened 40 common elements and their isotopes, comprehensively evaluating core parameters including radioactivity, isotopic abundance, toxicity levels, and aqueous solubility characteristics, followed by a phased multicriteria filtration process.16–18 Firstly, elements with high radiation risks or acute toxicity were excluded based on radioactivity thresholds and biological toxicity limits. Secondly, targets exhibiting stable dissolution and compatibility with analytical techniques, such as spectroscopy and mass spectrometry were selected using aqueous solubility data. Subsequently, candidates with significant environmental background interference or exceeding instrumental detection limits were eliminated by comparing ambient background concentrations with detection thresholds. Finally, economic viability was optimized by assessing industrial production scale, raw material costs, and supply chain stability. Through this progressive screening protocol, the 10B isotopic tracer emerged as the optimal candidate due to its negligible natural radioactivity, controllable biotoxicity, moderate aqueous solubility, low environmental background interference, and well-established isotope separation technology coupled with a robust commercial supply chain. Furthermore, the 10B isotopic tracer demonstrates distinct advantages over nanoparticle and DNA tracers in both sensitivity and cost-effectiveness. Specifically, its primary benefit lies in ultrahigh detection sensitivity: neutron activation analysis enables precise quantification of trace 10B at parts-per-billion levels, significantly surpassing the optical detection limits of nanoparticles and the amplification dependency of DNA techniques. Economically, despite natural scarcity, mature enrichment processes ensure 10B is more cost-efficient at scale than complex synthetic nanoparticles or high-purity DNA tags. Critically, 10B exhibits chemical inertness, resisting thermal, pH, and biodegradation effects, whereas nanoparticles are susceptible to aggregation and DNA tracers are prone to enzymatic degradation. For long-term monitoring applications, the inherent physicochemical stability of 10B ensures data integrity and eliminates risks associated with nanoparticle ecotoxicity or DNA mutation. Therefore, 10B represents a technically robust and economically sustainable solution for demanding applications such as deep-earth exploration and hydrology. Through integrated laboratory and field experiments, we evaluate its leakage identification efficacy and establish a tracer methodology specifically designed for salt lake brine seepage detection, providing a reliable tool for investigating spatial heterogeneity in subsurface flow fields.
Following the preparation of 10B tracer solutions with concentration gradients ranging from 0 to 20 μg L−1 in 5 μg L−1 increments, these solutions were mixed with brine at a 1:
1 volume ratio. After 24 hours of dark incubation at 25 °C, the transmittance within the visible light spectrum was measured using an ultraviolet-visible spectrophotometer. Systems with adjusted pH levels were analyzed to determine the variation rate of the 10B concentration and the 10B/11B isotopic ratio, thereby assessing the photochemical stability of the tracer. Further validation involved injecting 10B isotopic tracer, at five concentration gradients ranging from 0 to 20 μg L−1, into both natural salt lake brine and artificially prepared Ca2+ and Mg2+ brine with comparable concentrations. Recovery rates were calculated based on the linear relationship between the inductively coupled plasma mass spectrometry response values and the calibration curve, which has confirmed the accuracy of the tracer quantification.
Finally, for fine salt particles smaller than 1 mm, coarse salt particles of 1–3 mm, and powdery samples, add a tracer solution with a concentration gradient of 0–20 μg L−1 at a mass-to-volume ratio of 1:
3. The mixture was oscillated at a constant temperature of 25 °C in a centrifuge at 300 rpm for 48 hours, followed by centrifugation at 4000 rpm for 15 minutes to separate the liquid phase. After filtration through a membrane filter, the residual 10B concentration was measured to calculate the adsorption rate of the tracer by the medium.
Fluorescein tracer was added as a parallel control in this study and detected using a fluorescence spectrometer. Changes in fluorescence intensity were monitored to obtain performance metrics for fluorescein, including linear range, pH resistance in buffer, transmittance, static retention rate, recovery rate, and adsorption rate. These metrics were then comparatively analyzed against those of the 10B tracer using radar plots.
Name | Tube weight (g) | Weight after sand filling (g) | Weight after brine injection (g) | Pore volume (mL) |
---|---|---|---|---|
Fine salt | 2214.2 | 2354.2 | 2403.0 | 48.8 |
Kosher salt | 2214.3 | 2368.7 | 2411.8 | 43.1 |
Silt | 2214.5 | 2430.9 | 2458.1 | 37.2 |
![]() | (1) |
Here, Rtruesmp denotes the corrected isotope ratio of the sample, Rmeassmp refers to the directly measured isotope ratio of the sample, Rcertstd represents the certified theoretical isotope ratio of the standard material, and Rmeasstd indicates the measured isotope ratio of the adjacent standard material.
In addition, this study calculated the Peclet number, as shown in formula (2), which is a dimensionless number characterizing the relative importance of convective and diffusive transport in a fluid.
![]() | (2) |
Here, L denotes the feature length, m; v denotes the Darcy velocity, m s−1; τ denotes the degree of curvature, dimensionless; ϕ denotes the porosity, dimensionless; D0 denotes the diffusion coefficient of the free liquid phase, m2 s−1.
To better understand geological parameters, we introduced the Tess formula, as shown in formula (3).
![]() | (3) |
After being mixed with salt lake brine for 24 hours, the 10B tracer demonstrated excellent optical stability, with a transmittance maintained at 98.2% ± 0.3%. The isotopic ratio deviation remained below 2%, indicating no turbidity or suspended particles in the solution, which reflects good stability. Additionally, under dynamic pH conditions ranging from 4 to 12, the retention rate of the 10B tracer remained stable at over 96%, highlighting its excellent acid and alkali resistance, as shown in Fig. 4.
As shown in Fig. 5, when the concentration of the 10B tracer was gradually increased from 5 μg L−1 to 20 μg L−1, the changes in the 10B tracer concentration and isotopic ratio remained below 0.38%, demonstrating its excellent stability. As the concentration increased, the isotopic ratio deviation rose from 0.02% to 0.31%, indicating that the tracer system at lower concentrations exhibits superior phase stability.24–26 Under low-concentration conditions, boric acid molecules are uniformly dispersed in the brine system in the form of monomeric B(OH)3/B(OH)4−, with weak intermolecular interactions, bringing the system closer to an ideal solution state and enhancing phase stability. When the concentration is relatively high, boric acid molecules self-assemble into polymeric associates of the [Bn(OH)3n]n− type through hydrogen bonding between hydroxyl groups. The accumulation of these polymers may lead to localized supersaturation in the solution, potentially causing spontaneous crystallization or precipitation. Therefore, the experimental results provide a basis for optimizing the tracer dosage strategy.
As shown in Fig. 6(a), the 10B tracer exhibited high recovery rates ranging from 92.8% to 106.5% across the tested concentration range of 0 to 20 μg L−1. As a key indicator for evaluating the accuracy of a method in chemical analysis, the recovery rate directly reflects the degree of closeness between the measured value and the true value. The recovery rate obtained in this study complies with the permissible range of 90–110% set by the US Geological Survey for trace element analysis standards,27 systematically verifying the reliability of the experimental system. This result confirms the accuracy of ICP-MS technology in quantitative analysis from a methodological perspective and indicates that the established method has ideal detection efficiency.
As shown in Fig. 6(b), the adsorption rate of the 10B isotope tracer increases progressively with higher concentration. Furthermore, at a concentration of 20 μg L−1, the inhibition loss rates are remarkably low, measuring 3.88% ± 0.18% in fine salt, 5.74% ± 0.31% in coarse salt, and 6.31% ± 0.42% in sludge media. These exceptionally low values demonstrate its exceptional anti-adsorption properties. The study further revealed that this anti-adsorption behaviour arises from synergistic multiscale physicochemical mechanisms. In fine salt media, the surface negative charges of particles and the electrostatic repulsion between boric acid anions [B(OH)4−] significantly reduce the probability of physical adsorption.28 While kosher salt particles possess porous structures capable of temporarily trapping certain substances, their lack of effective binding sites hinders the formation of stable chemical bonds.29 Although clay minerals in silt exhibit strong adsorption capacities, the lighter 10B isotope tends to detach more readily from mineral surfaces due to its structural characteristics.30 Consequently, 10B in salt lakes demonstrates enhanced permeability through media and maintains superior migration capacity.
![]() | ||
Fig. 7 Breakthrough curves of 10B tracer and Cl− in porous media: (a) fine salt; (b) kosher salt; (c) silt; (d) effluent concentration versus injected pore volume across packing materials. |
This study further reveals through flow experiment data that the loss rates of the 10B tracer in fine salt, coarse salt, and mud are 11.7% ± 1.2%, 13.4% ± 1.5%, and 16.8% ± 2.1%, respectively, as shown in Fig. 7(d). The gradual increase in the 10B tracer's loss rate across different media is fundamentally attributed to differences in specific surface area.34,35 The fine salt medium, composed of millimeter-sized crystal particles, has a smooth surface and homogeneous pores, resulting in the smallest specific surface area per unit mass. The coarse salt medium, with micrometer-scale roughness and secondary fractures on its particle surfaces, has a specific surface area 30–50% larger than that of fine salt, providing more interfaces for physical adsorption. The mud medium, due to its honeycomb-like pore network of clay minerals, has a specific surface area several times larger than that of coarse salt. This increase in specific surface area enhances the collision frequency and contact time between solute molecules and the medium surface, resulting in a stepwise increase in the loss rate from fine salt to mud. These results provide critical design parameters for practical applications. Under typical brine seepage conditions in salt pans, the tracer dosage needs to be increased by 20% to compensate for adsorption losses by the medium.
As shown in Fig. 8, both the 10B tracer and fluorescein exhibit good performance. However, the 10B tracer outperforms fluorescein, especially in adsorption rate and buffer pH resistance, demonstrating more advantages in these metrics.
![]() | ||
Fig. 9 Temporal variation of 10B/11B ratio in monitoring wells: (a) well 1; (b) well 4; (c) well 6; (d) well 7; (e) well 3; (f) well 5. |
The data analysis from Fig. 9(d)–(f) reveals that the 10B/11B ratio in well 5 experienced a brief peak of 0.215 before stabilizing at 0.208. This delayed response and the absence of a sustained upward trend suggest that well 5 is located at the edge of the seepage system. In contrast, well 7 exhibited a slight peak of 0.220 on the sixth day and eventually stabilized at 0.218. This limited fluctuation indicates the presence of a low-permeability geological barrier in its seepage pathway, restricting fluid flow.39,40 Notably, the 10B/11B ratio in well 6 showed a continuous upward trend without stabilizing, suggesting significant seepage retention in the area. This irregular variation implies that fluid flow may be slowed or halted in certain geological structures, leading to long-term changes in the isotopic ratio. It can therefore be inferred that well 6 is likely a leaking well, and appropriate anti-seepage measures should be implemented to prevent further leakage.
Based on the data analysis and the illustration shown in Fig. 10, the main seepage path of the salt field leakage system is characterized by fluid primarily flowing rapidly along the pathway from 2 → 1 → 6. This phenomenon indicates that the fracture system plays a dominant role in fluid migration. The isotopic ratio at well 1 rose rapidly the day after injection and remained at a high level, suggesting the existence of a dominant fracture channel. The formation of this channel is attributed to the combined effects of salt layer dissolution and geological structure, exhibiting characteristics of low resistance and high velocity in seepage.41,42 The stable isotope ratio at well 6 continued to rise steadily. Substituting the parameters s = 1.2 m, Q = 2.832 m3 d−1, t = 1.04 × 10−3 d, and r = 8.94 m into formula (3) yielded a hydraulic conductivity k of 0.0978 m d−1. This suggests the potential presence of a weakly enclosed area with medium to high permeability downstream. The rapid response and stable flow characteristics of the main seepage path further confirm the high connectivity of the fracture system. In contrast, at well no. 3, a monitoring well located 4 meters from the injection point, the signal strength was barely detectable. Therefore, this paper calculated the Peclet number for this well. Based on literature and measured data, the relevant parameters are as follows: diffusion coefficient D0 for 10B molecules at 0.45 nm = 5.45 × 10−10 m2 s−1, Darcy flow velocity v = 4 × 10−9 m s−1, tortuosity τ = 2.1, and porosity ϕ = 0.2. The Peclet number was calculated using formula (2), yielding Pe >1. These results confirm that solute transport is dominated by advection, overriding the effects of molecular diffusion. Therefore, this paper suggests that the absence of tracer 10B inflow in well no. 3 is primarily attributed to the action of the assisted plugging mechanism. This finding, verified by inversion analysis, further confirms the significant hydraulic advantage of the main seepage channel at well 1. The formation of the auxiliary path 2 → 1 → 4 → 6 stems from the synergistic effect of fractures and tiny pores. The fluid concentration at well 4 first fluctuates and then stabilizes, indicating that while the main fractures rapidly transport fluids, the surrounding tiny fractures and low-permeability pores temporarily “store and release” fluids, causing concentration fluctuations. This path is essentially a branch of the main fracture network extending from well 1, where fluid transport is accomplished through the cooperation of tiny fractures and pores, yet its transport efficiency is lower than that of the main channel. The differences between the main and auxiliary paths are mainly influenced by three factors: the orientation of fracture distribution, the uneven distribution of pores, and the dynamic changes of blockages.43,44 These findings provide vital references for precisely locating seepage plugging projects. Simple anti-seepage measures were implemented at well 6 in this study. Measurement and calculation results revealed a decrease in the leakage coefficient from 0.0684 m d−1 to 0.0447 m d−1, confirming the presence of leakage at this well. To achieve further leakage reduction, more efficient anti-seepage equipment would be required.
Field experiments successfully uncovered the spatial heterogeneity and dynamic evolution of the salt pan seepage system. The main seepage pathway 2 → 1 → 6, dominated by a highly connected fracture network, exhibited rapid breakthrough, stable flow velocity, and sustained concentration response, highlighting the hydraulic control advantages of the fracture system. The secondary pathway 2 → 1 → 4 → 6, regulated by the coupling of fractures and micropores, displayed oscillatory concentration patterns, low transmission efficiency, and periodic retention, reflecting the critical influence of geological heterogeneity on seepage pathway differentiation. For well 3, Peclet number calculation yielded Pe >1, with absent 10B tracer inflow primarily caused by an auxiliary plugging mechanism. Delayed responses and weak signals in peripheral monitoring wells such as well 5 and well 7 further confirmed the presence of permeability barriers, likely formed by dense clay layers or localized salt cementation, providing direct evidence for targeted anti-seepage engineering. Following the implementation of simple anti-seepage measures at well 6, the leakage coefficient decreased, confirming the presence of leakage at this well. More effective anti-seepage equipment is required to further reduce seepage. This technology overcomes the limitations of traditional methods in detection accuracy, real-time capability, and quantitative analysis. By integrating high-resolution tracer data with dynamic flow field, it achieves a technological leap from empirical estimation to precise diagnosis, offering an innovative tool for efficient salt lake resource exploitation, leakage risk mitigation, and ecological anti-seepage design.
Future research could focus on three directions: first, integrating artificial intelligence algorithms and IoT technologies to develop a real-time seepage monitoring and intelligent early-warning platform for dynamic leakage event detection and rapid response. Second, conducting multi-field coupled simulations to reveal the cumulative impacts of long-term mining activities on salt layer permeability. Third, expanding the application of the 10B tracer technology to other high-salinity environments to validate its universality and engineering potential. Through technological iteration and interdisciplinary collaboration, this method is poised to become a core technology for salt lake ecosystem conservation and sustainable resource utilization, supporting the efficient development of strategic salt lake resources and ecological security under the dual-carbon goals.
This journal is © The Royal Society of Chemistry 2025 |