1 Introduction
In water-rich sand strata, underground structures are highly susceptible to leakage risks [
1,
2]. In the event of water–soil gushing disasters, failure to implement immediate countermeasures may trigger severe soil erosion, surface subsidence, and damage to underground structures [
3–
7]. Currently, the primary disaster mitigation strategies for water–soil gushing disasters in underground structures include grouting sealing [
8], dewatering [
9], and structural protection [
10]. Among emergency response engineering projects, grouting sealing is the most adopted treatment method [
11]. However, without measures to reduce the water head and flow rate for the purpose of curbing soil erosion, the time required for grouting sealing to become effective can be significantly prolonged, and the outcome might not be entirely satisfactory. At the same time, it becomes extremely challenging to create favorable working conditions for structural protection measures [
12]. As a result, in numerous emergency response engineering projects, dewatering wells were utilized to lower the water head and flow rate [
13]. Nevertheless, as shown in Fig. 1(a), the traditional dewatering method suffered from low efficiency due to the continuous replenishment of water from the distant water level. Moreover, it was difficult to maintain a low water level at the leakage point for an extended period [
14]. Additionally, excessive dewatering could cause significant disturbance to the surrounding ground [
15,
16]. Therefore, in this paper, we applied the method proposed by Zhou et al. [
17] of forming a water-blocking barrier through grouting to prevent the replenishment of the water level, as depicted in Fig. 1(b). A slurry with rapid seepage reduction properties was injected around the leakage point to form a uniform and stable water-blocking barrier. Subsequently, dewatering was carried out within the water-blocking barrier. This not only significantly improved the dewatering efficiency but also better maintained a low water level near the leakage point, providing valuable time for emergency measures such as grouting sealing and structural protection.
To rapidly form a water-blocking barrier, it is essential to develop a grouting material with high injectability, high seepage reduction capacity, and high seepage resistance. Traditional grouting materials, such as cement slurry [
18,
19,
20] and resin slurry [
21], cannot diffuse rapidly and uniformly within the soil layers. Due to their high viscosity, they often employ the split grouting method, which is unable to form a uniform and stable water-blocking barrier. Inspired by podzolization, it has been found that the accumulation of the Al-OM flocs in soil pores could significantly reduce the hydraulic conductivity of the sand, and in some cases, could even achieve an almost impermeable state [
22,
23]. Therefore, this paper presented a method of using the aluminum-organic matter flocs (Al-OM flocs) as a grouting material to construct a grouting barrier. The research results indicated that the Al-OM flocs possessed unique characteristics: under high flow rate conditions, the Al-OM flocs broke due to intense agitation; once the agitation stopped, these flocs can regrow spontaneously within a few minutes [
24]. By taking advantage of this property, the Al-OM flocs could be easily injected into the ground, and after the injection was completed, the regrowth process of the Al-OM flocs effectively blocked the pore channels in the soil, significantly reducing the hydraulic conductivity of the sand to form a water-blocking barrier. Zhou et al. [
17,
25] summarized the principle of seepage reduction, applied this material to dam seepage prevention and verified the feasibility and effectiveness of the Al-OM flocs in seepage prevention applications.
This paper conducted 1D column experiments to verify whether the Al-OM flocs could be easily injected into the soil column, and whether, after the grouting was completed, the Al-OM flocs could effectively block the pore channels to exhibit high effectiveness in reducing sand permeability, and sufficient resistance against seepage flow. In addition, disaster mitigation tests were carried out to simulate the leakage disasters of underground structures in water-rich sand layers. By using the grouting device and dewatering device, this study explored how the grouting barrier enhanced the dewatering efficiency, inhibited seepage erosion of the soil, and maintained the water level at the leakage point. The research results may provide important references for engineering emergency rescue.
2 1D column experiments
To ensure that Al-OM flocs can be used as a grouting material in the practical engineering, 1D column experiments were carried out to test its performance in reducing the soil permeability.
2.1 Material
The Al-OM flocs used in the experiments was prepared by mixing aluminum chloride (AlCl
3·6H
2O) with humic acid potassium (C
9H
8K
2O
4) in a continuously stirred reactor. The dosage of AlCl
3·6H
2O and humic acid potassium in 1L water were 0.496 g and 0.472 g, respectively. The humic acid potassium (HUMIN-P775) used in the experiment was produced by Biotech, a German biotechnology company. The humic acid potassium (HUMIN-P775) was 100% water-soluble. Its main elements included nitrogen (1.025%), carbon (42.175%), hydrogen (3.475%) and sulfur (0.854%). Eder et al. [
26] found that the Al/C ratio of 0.06 is sufficient, and a relatively low aluminum dosage can minimize the impact on the ecological environment. However, our experiments showed that when the Al/C ratio was 0.125, although the aluminum dosage increased significantly, the seepage reduction effect was slightly enhanced. In emergency rescue projects, effectiveness and time are more critical; therefore, the Al/C ratio adopted in this paper was 0.125.
The sand used in the experiments was a river sand. Its main component was quartz. The specific gravity of the sand particles, Gs was 2.65. The average diameter, D50 was 0.22 mm. The coefficient of uniformity, Cu was 1.6. The gradation curve of the sand was shown in Fig. 2.
2.2 Experiment apparatus and procedures
The experiment apparatus and procedures are shown in Fig. 3. Basically, the experiment can be divided into two stages.
Stage I was a grouting test. First, the sand sample was prepared by underwater sedimentation method. As shown in Fig. 3(a), the permeameter, accommodating the sample, was made of a PVC tube, which had an inner diameter measuring 5 cm, an outer diameter of 7 cm, and a length spanning 35 cm. As shown in Fig. 4(a), the front and rear plates of the permeameter were both equipped with a filter sieve (pore diameter is 0.1 mm) to prevent the sand particles from escaping through the inlet or outlet. During the sample preparation, the permeameter was placed in a water container, as shown in Fig. 4(b). The front plate was opened, and the sand was poured into the permeameter. The permeameter was taken out of the water once it was replete with sand. The mass of the sand was recorded to calculate the porosity of the sand. After that, the permeameter was positioned on the table and linked to both the slurry container and the collection container. Before grouting, 5PV (pore volume) of pure water was injected into the permeameter to fully remove air from the sample. Then, Al-OM flocs were slowly injected into the saturated sand at 10 kPa grouting pressure to investigate their diffusion in the porous medium. After the black area basically stopped spreading, we increased the grouting pressure by 10 kPa. Once the black area expanded to the top of the sample, it was determined that the entire sample had been filled with Al-OM flocs. At this point, both grouting pressure and injection volume were recorded.
Stage II was an infiltration test. The test used a standard volume controller (SVC), which could elaborately control the injection flux, as shown in Fig. 3(b). A saturated sand sample was first tested before Stage I. Pure water was injected into the sand column at a constant flux. By continuously monitoring the grouting pressure in real time, the hydraulic conductivity of the saturated sand sample could be obtained. Following the same procedure, an assessment was then made on the sample containing Al-OM flocs that had completed Stage I to ascertain its hydraulic conductivity. The flux of the SVC was continuously increased, and when the grouted sand reached the critical hydraulic gradient, the Al-OM flocs would be flushed out, causing significant changes in the pH value and EC value in the collection container. Therefore, the hydraulic gradient at the time when the pH value and EC value changed dramatically was taken as the critical hydraulic gradient of the sample.
2.3 Experiment results
The permeameter was filled with a total of 1.273 kg of sand, and the pore ratio of the saturated sand was determined to be 0.43. In stage I, a total of 1.08 L Al-OM suspension, equivalent to 5.2 PV, was injected into the sample. As shown in Fig. 5(a), the penetration length exhibited a continuous increase with the increment of grouting pressure. When the grouting pressure reached 40 kPa, the Al-OM flocs completely permeated the sand sample in the permeameter, thereby demonstrating good injectivity.
In stage II, the flux-pressure curves of the saturated sand and the sand with Al-OM flocs were shown in Fig. 5(b), and the obtained hydraulic conductivity of the sample can be seen in Fig. 5(c). The hydraulic conductivity of the saturated sand was 9.79 × 10−4 m/s, and the hydraulic conductivity of the sand with Al-OM flocs is 3.63 × 10−5 m/s. After grouting, the hydraulic conductivity decreases by 27 times, which demonstrates that the Al-OM flocs can significantly reduce the permeability of the soil.
As shown in Fig. 5(b), when the flux was increased to the maximum value (1.5 × 10−6 m3/s) of the SVC, there was no Al-OM flocs discharged. At this time, the injection pressure was 73.6 kPa and the corresponding hydraulic gradient was 21.03. It was deemed that Al-OM flocs had sufficient resistance against seepage flow. In addition, cost analysis indicates that for industrial-scale production of slurry at this concentration, the total raw material cost is 2.3 yuan/m3. In comparison, cement slurry typically ranges from 200 to 300 yuan/m3, meaning the use of Al-OM slurry can reduce costs by approximately 100 times.
To sum up, the Al-OM flocs exhibited favorable diffusion performance under low injection pressure, high effectiveness in reducing sand permeability, and sufficient resistance against seepage flow. Consequently, it can be regarded as a feasible grouting material to mitigate water–soil gushing disasters in underground structures.
3 Disaster mitigation tests
3.1 Apparatus
Laboratory tests were conducted to test the performance of Al-OM Flocs grouting in mitigating water–soil gushing in underground structures. These physical tests, rather than scaled model tests, are designed to mirror typical engineering conditions, with a focus on verifying the underlying mechanisms. The experiment apparatus can be seen in Fig. 6(a). It comprised a model box, a water pump, a grouting tank, a series of grouting and dewatering pipes, 2 electronic scales, 2 collecting tanks, and an industrial camera. The water pump was utilized for the purpose of extracting water from the sample, namely, dewatering. The grouting tank was lifted to a certain height (1.5 m in this study) to facilitate the injection of Al-OM flocs into the sample by means of gravity. The collecting tanks were equipped with overflow pipes, and placed on the electronic scales.
As shown in Fig. 6(b), the model box was divided into 2 small sand tanks (each sand tank was 500 mm in length, 300 mm in width and 470 mm in height) by a steel plate measuring 1 cm in width. The two sand tanks were equipped with a shared outlet below the steel plate, allowing for simultaneous triggering of water–soil gushing once the outlet was opened (the opening measured 20 mm in length and 2 mm in width for each sand tank). This facilitated the execution of comparative experiments. The sand tanks were surrounded by a reservoir, which was equipped with an overflow pipe to maintain a constant water level. The notable aspect is that the rear plates of both sand tanks are impermeable, thereby only lateral supply of water was enabled. A series of piezometer tubes were installed on the base of the model box. They were used to monitor the water head along the seepage path.
To simulate the functions of dewatering wells in lowering water levels and grouting wells in grouting in real engineering, grouting and dewatering pipes were designed as shown in Fig. 6(a). They were metal pipes (350 mm in length, 10 mm in outer diameter and 9 mm in inner diameter) wrapped with a filter sieve (0.1 mm in pore diameter). Four holes with a diameter of 2 mm were drilled on the side wall of the metal pipe every 50 mm. Before sand filling, the metal pipes were fixed in their corresponding positions in advance as shown in Fig. 6(c). The spacing between two adjacent grouting pipes was set to a small enough distance (45 mm in the experiment) to allow for the formation of a complete barrier by the injected Al-OM flocs, effectively impeding seepage flow. After sample filling was completed, rubber pipes (8 mm in outer diameter and 6 mm in inner diameter) were inserted into the metal pipes, and sealing gaskets were installed on the top of the metal pipes to ensure a seal between the rubber pipes and the metal pipes. The rubber pipe in the dewatering pipe was fixed near the bottom all the time, while the rubber pipe in the grouting pipe was slowly pulled upward from the bottom to near the top of the metal pipe during the grouting process.
3.2 Programme
The test program was shown in Table 1. Basically, 4 groups of tests were conducted to evaluate the capability of Al-OM flocs in mitigating water–soil gushing disasters in underground structures. Test 1 was a benchmark test. The conditions in sand tank 1# and tank 2# were identical (only leakage was considered). In this way, the reliability of the experimental apparatus can be checked.
Test 2–4 were 3 groups of comparative tests. In Test 2, grouting was carried out first and then dewatering was performed in sand tank 1#, while only dewatering was carried out in sand tank 2#. The objective of test 2 was to address the impact of grouting on the efficiency of dewatering. In Test 3, grouting was first carried out in sand tank 1# and then leakage was triggered, while in sand tank 2#, leakage was directly triggered. The purpose of Test 2 was to study the inhibitory effect of grouting on soil erosion. In test 4, grouting was first carried out in sand tank 1#. Subsequently, both leakage and dewatering were triggered simultaneously. In contrast, grouting was absent in sand tank 2#, while leakage and dewatering were triggered simultaneously as well. Test 4 was conducted to investigate the combined effect of dewatering and grouting.
3.3 Test results
3.3.1 Reliability of the experimental apparatus
A benchmark test (Test 1) was first conducted using the model experiment apparatus to test the reliability of the experimental apparatus. To ensure that the soil erosion area was more regular, thereby better distinguishing whether the seepage conditions on both sides were consistent, the water level was kept constantly higher than the soil surface in the benchmark test. In the test, the sample was 250 mm high, and was prepared by water pluviation method. The water level within the reservoir was maintained at a constant height of 320 mm. Since the two sand tanks shared a single outlet at the bottom, upon opening the outlet, leakage was simultaneously initiated in both sand tanks.
Figure 7 depicted the erosion pattern of the soil 350 s after leakage was triggered. Basically, the erosion morphology was approximately mirror-symmetric in the two sand tanks. This indicated that the experimental conditions and setups in the two sand tanks were highly consistent, thereby demonstrating that the experimental apparatus is applicable for performing comparative experiments.
3.3.2 Impact of Al-OM grouting on dewatering efficiency
Test 2 was conducted to investigate the impact of Al-OM grouting on the dewatering efficiency. In this test, the water level in the reservoir was maintained constant at 320 mm, which was lower than the height of the sample (350 mm). This was done to prevent the water from overflowing across the top of the grouting barrier. In sand tank 1#, Al-OM flocs are injected into the soil at a constant grouting pressure of 17.5 kPa. After 180 s of grouting, an 80-mm-wide grouting barrier was formed. During the grouting process, the grout was easily injected, which might be because excessive grouting pressures caused the soil to be subjected to some kind of liquefaction. In real world engineering, it may be necessary to reduce the grouting pressure to ensure that soil liquefaction does not occur. Subsequently, dewatering was carried out simultaneously in tank 1# and tank 2# using a constant negative pressure of −15 kPa.
Figure 8 depicts the temporal variations of the water level in the two sand tanks. In sand tank 1#, the water level stabilized 143 s after the initiation of dewatering. The maximum drop in water level was observed near the dewatering pipes, approximately 199 mm, corresponding to a decline rate of 1.39 mm/s. Conversely, in sand tank 2#, the water level attained stability after 201 s. The maximum water level reduction, which was only 80 mm, also occurred near the dewatering pipes, with a decline rate of 0.4 mm/s. These test results suggest that the grouting barrier formed by Al-OM flocs enhances the drawdown by a factor of 2.5 and improves the dewatering efficiency by 3.5 times.
3.3.3 Inhibition of seepage erosion by Al-OM grouting
The preparation procedures of Test 3 were the same as those of Test 2. Initially, the Al-OM flocs were injected into the soil within Sand Tank 1#. Once the grouting barrier was formed, the grouting operation was stopped, and the leakage outlets were opened.
For the sample in sand tank 1#, the water level adjacent to the separator plate continuously decreased, and the rate of decrease was notably faster than that in sand tank 2#. Within 20 s, the water level in sand tank 1# dropped to 146 mm, and subsequently continued to decline gradually, reaching the lowest point of 21 mm at 420th second. Meanwhile, soil erosion almost ceased, and the water level remained essentially stable.
In contrast, the decrease in the water level adjacent to the separator plate in sand tank 2# was much slower. Within 200 s, the water level merely dropped to 163 mm. Moreover, extensive soil erosion occurred in sand tank 2#, causing the water level to fluctuate and rise. It rebounded to 248 mm at 420th second, and the soil erosion rate also accelerated continuously. The water level profile of Test 3 is illustrated in Fig. 9(a).
At the 420th second, the soil erosion condition is shown in Fig. 9(b). In sand tank 1# with the presence of a grouting barrier, the hydraulic head near the leakage outlet was nearly zero. This indicated that the grouting barrier increased flow resistance and reduced the seepage velocity near the leakage point, thereby limiting soil erosion. In contrast, sand tank 2# without grouting maintained a high-water level near the leakage outlet, and the water-sand mixture continued to be discharged. Without the resistance of the grouting barrier, the seepage force remained strong, leading to continuous and unobstructed soil erosion. The gradual rise of the water level in sand tank 1# was mainly due to the initially high-water level. The drop of the water level at the leakage outlet was entirely accompanied by water and sand loss, which caused the soil at the top of the soil mass (especially 150 mm away from the front glass surface) to be lost slowly, and the integrity of the grouting barrier was affected. In practical engineering, if a complete and uniform grouting barrier is installed at a relatively far distance, the erosion process will eventually be terminated over time.
3.3.4 Synergistic inhibition of seepage erosion by dewatering and Al-OM grouting
In Test 4, both leakage and dewatering were concurrently implemented. In sand tank 1#, the water level adjacent to the separator plate reached its lowest point after 98 s. The drawdown height of the water level measured 312 mm, corresponding to a drawdown rate of 3.18 mm/s. In contrast, for sand tank 2#, the water level near the separator plate dropped to its lowest point after 488 s, with a drawdown height of 205 mm and a considerably slower drawdown rate of 0.42 mm/s. The water level variations in Test 4 are illustrated in Fig. 10(a).
Upon opening the leakage outlets, a marked disparity in water levels was observed, with the water level in sand tank 1# being substantially lower. This differential in water levels led to a significant divergence in soil erosion rates on either side. As erosion progressed, the discrepancy in soil erosion areas between the two sides enlarged further. The erosion morphology of soil at 488th second after opening the leakage outlets is depicted in Fig. 10(b). Due to the synergistic inhibition by dewatering and Al-OM grouting, the erosion area was merely 24200 mm2. Conversely, the erosion area expanded to 45300 mm2 if only dewatering was adopted.
By comparing Fig. 9, it is found that in Fig. 10, the water level near the leakage outlet of Sand Tank 1# remains at a very low level, and soil leakage basically stops completely. This is because the initial dewatering measures quickly reduce the water level near the leakage outlet to 0, which greatly slows down the initial rate of soil erosion, thereby ensuring the integrity of the grouting barrier. And the complete grouting barrier will eventually terminate the erosion process thoroughly.
Upon opening the leakage outlets for a duration of 488 s, dewatering was terminated. In sand tank 2#, it was observed that the water level in the vicinity of the separator plate exhibited a rapid ascent, concurrent with a notably high soil erosion rate. A mutual reinforcing relationship between soil erosion and water level restoration became evident. Specifically, when dewatering ceased for 40 s, the water level near the separator plate had already rebounded to 231 mm. By the 59th second, the soil in the control group had largely disintegrated, and the water level close to the separator plate swiftly recovered to 292 mm. At the 69th second, a complete collapse of the soil in sand tank 2# occurred, and the water level near the separator plate had ascended to 317 mm.
In contrast, in sand tank 1#, the rate of soil erosion remained remarkably low. Over the course of 69 s, the water level near the separator plate only registered an increment of 26 mm. This disparity unequivocally demonstrates the pronounced efficacy of the Al-OM grouting barrier in retarding water level recovery. The water level variations in Test 4 after the termination of dewatering are illustrated in Fig. 11.
4 Conclusions
This study convincingly demonstrates that Al-OM flocs can be effectively employed as a grouting material for emergency rescue in the event of underground structural leakages. Moreover, the study proposes to apply the Al-OM flocs to mitigate water–soil gushing disasters in underground structures. The specific conclusions are as follows.
1) The Al-OM flocs feature outstanding injectability and anti-seepage ability, quickly forming a barrier that effectively cuts down the hydraulic conductivity of the sand, meeting the emergency rescue grouting material requirements for leaking underground structures.
2) Upon the formation of the Al-OM grouting barrier, there is a substantial enhancement in the dewatering efficiency. Additionally, the grouting barrier demonstrates remarkable effectiveness in strongly inhibiting the seepage erosion of the soil, effectively safeguarding the soil structure from being eroded by seepage forces.
3) A distinct advantage of Al-OM grouting is its effectiveness in retarding water level recuperation and maintaining low water level for a longer duration. This offers invaluable time for the swift and efficient implementation of emergency rescue measures, significantly enhancing the success rate in addressing underground structural leakages and damages.
PDF
(4896KB)
