Evaluation of Lime-Treated Lateritic Soil for Reservoir Shoreline Stabilization

: Sedimentation is one of the major problems addressed by reservoir management, and requires extensive e ﬀ ort to control it. This paper aims to evaluate the e ﬃ ciency of the soil–lime stabilization technique for reservoir shores. The treatment consisted of spraying hydrated lime in slurry form over the surface of a lateritic clay sample with 1, 2, and 4% lime solution and curing times of 1, 7, 28, and 56 days with air-drying and moist-room storage. In addition, a single test with less than 1% lime solution by weight percentage was carried out. The post-cured specimens were mapped with SEM and X-ray analyses. A wave ﬂume test was performed in samples subjected to diverse conditions of lime content, type, and curing time. The results showed that the present technique produces a Ca-rich crust by carbonation rather than stabilizing it and that the lime content and type of curing generate improvements in soil loss reduction, but the curing time does not. The technique gave relative protection against water level variation and wave impacts, but it is necessary to consider a frequent application of lime on the lateritic soil. soil loss of samples with the same preparations and curing times under air-drying. The in the water more significant shrinkage of the soil after days of curing, and the protection of the soil by lime treatment and allowed easy entry of water into a fragile structure. Besides, at 28 days, 4% lime treatment showed a lower soil loss of 16%, despite the results for 1 and 2% lime solutions. This can be explained by the formation of more calcium crust by carbonation and lower soil shrinkage, compared to other samples at 28 days of curing time. but the curing time did not have a significant impact because the water content variation was more important in the loss mass. In this way, in the field, soil moisture could influence the efficiency of this technique because when subjecting the soil samples to air-drying for a long period, the potential weakening of the soil structure was observed, which allowed mass loss. In periods of drought, the protective layer of lime can contract, leading to cracks that allow water infiltration and surface runoff at shore banks, increasing the soil erodibility. The advantages of hydrated lime slurry as limewash for shoreline soil stabilization are as follows:


Introduction
Shoreline erosion around reservoirs occurs due to an imbalance of the soil-atmosphere dynamics and soil-water conditions triggered by wind action [1], gravity [2], rainfall [3], water level fluctuations [4], wave action [5][6][7], and human activities [8]. The reservoir operation also plays a vital role in the intensity of the wave abrasion process that occurs on the bank slope [1]. The sediment yield by erosion through all the natural processes and human activities increases the sedimentation rate and, consequently, also becomes the main issue for reservoir storage capacity, power generation, irrigation, and the environment [5,[9][10][11]. Because of the lack of measures to control sediment inflow and deposition, almost 25% of the worldwide reservoir storage capacity could be lost in the next 25 to 50 years [12], and a loss of US $13 trillion to US $19 trillion in each year is estimated [13].
Soil stabilization is one of the various necessary actions taken by reservoir managers to mitigate shoreline erosion, and some methods are growing in popularity: biotechnical techniques [14], geosynthetics [15,16], and chemical soil stabilization [17,18].
Chemical soil stabilization contributes to greater internal strength and durability [17] for geotechnical purposes. Its success in stabilizing compacted soils in highways, airfields, building foundations [19], and earthen dams is undeniable [20][21][22]. However, there are fewer references on the use of chemical soil treatments for slope stability [23,24], and studies on reservoir shoreline stabilization are lacking [18]. Lime (i.e., quicklime and hydrated lime) is one of the oldest [19,25] and lowest-cost [26] chemical stabilizers used nowadays for soil stabilization [27].
The use of quicklime in the stabilization of lateritic soil was revealed to lead to better performance in terms of increasing the bearing strength and workability (e.g., plasticity and compressibility) compared to industrialized hydrated lime when both are used in the powdered form. However, hydrated lime is more applicable for soils with higher clay content [21], which is a common characteristic of tropical soils.
A low-cost hydrated lime slurry method was proposed by Nascimento et al. [18] in lateritic clay stabilization for resisting water erosion, such as sheet erosion and water level fluctuations. These works consisted of the evaluation in a laboratory of the sheet erosion with a modified Inderbitzen apparatus and the water level fluctuation by partial and total immersion of soil samples in a receptacle with clean water (i.e., degradation test). Wind-generated waves also play an important role in the shoreline erodibility through their impact actions. Therefore, it is necessary to evaluate this erosion process, which can be done by a wave flume test. This experiment allows the determination of soil mass loss according to the wave frequency and shore slope [16,[28][29][30].
The main objective of the current work is to evaluate lime-treated lateritic clay under wave flume tests to understand the behavior against wave impacts and water level fluctuations, with the aim of mitigating the effects of reservoir shoreline erosion by a non-traditional technique.
In addition, the study also aims to address the effect of varying the lime content under different curing times, the influence of curing storage (air-drying and moist room), and the influence of pH changes and to identify the depth of soil-lime microstructural and chemical reactions. The present research consists of spreading hydrated lime slurry over the surface of soil samples for shoreline stabilization, which is different from many studies that test specimens of soil-lime mixtures.

Description of the Study Area
Soil samples were collected on the shore of the Itumbiara dam reservoir, in a place with high potential for wave erosion [31], shown in Figure 1, and the analyses were carried out in the laboratories of the Federal University of Goias and Furnas. This dam is located in the Midwest of Brazil in South America and is the 14th largest hydropower plant reservoir in the world by flood area and the 12th largest Brazilian hydropower plant in installed capacity (i.e., 2082 MW) [32].
The fetch model tool based on geographic information system (GIS) software was applied to define sites with higher potential for wave-and wind-generated erosion, and a location near the dam with road access was chosen for the current study [31]. Other researchers also reported a strong influence of rainfall [3,33,34] on the watershed of the Itumbiara dam reservoir.
The average ratio of rainfall to runoff calculated by Luiz et al. [3] was 7708.00 MJ mm/ha·h during the rainy season (November to March). The parent rock consists of metamorphic rocks (schist-like), which produce residual erodible soils with a higher rate of soil degradation, which is increased by anthropic actions (i.e., agriculture and irregular settlements), according to Jesus et al. [33].
The lateritic soil (i.e., oxisols or ferrasols) is the main soil type found over the study area and results from intense weathering on a metamorphic parental rock and lies above a saprolitic soil (i.e., inceptisols). The morphology of shores around this reservoir is flat to gently lakeward-sloping (5-10 • ) platforms that create a favorable environment for sheet and rill erosion [31].
Pedological characterization and soil classification of the lateritic soil profile showed high amounts of organic matter (roots) on top of the pedon in the O horizon (3 cm thickness) and a transitional A horizon until 20 cm depth. In the deeper layers (20 to 120 cm), a homogeneous B horizon was found with high contents of Fe and Al. The colors of the A and B horizons were in slightly similar ranges Water 2020, 12, 3141 3 of 15 of hues from dark to medium red-orange (Soil Color 2.5 Yellow-Red 2/6-6/10). These horizons had a fine sandy loam texture with a few tubular pores and were slightly sticky; the drainage condition was well-drained. Pedological characterization and soil classification of the lateritic soil profile showed high amounts of organic matter (roots) on top of the pedon in the O horizon (3 cm thickness) and a transitional A horizon until 20 cm depth. In the deeper layers (20 to 120 cm), a homogeneous B horizon was found with high contents of Fe and Al. The colors of the A and B horizons were in slightly similar ranges of hues from dark to medium red-orange (Soil Color 2.5 Yellow-Red 2/6-6/10). These horizons had a fine sandy loam texture with a few tubular pores and were slightly sticky; the drainage condition was well-drained. Table 1 summarizes the soil physical and chemical properties, and Table 2 shows the general chemical composition of this lime and the selected lateritic soil. Kaolinite and quartz were the main minerals observed in the soil specimen; muscovite and iron oxides were also present. The lime samples presented calcite, portlandite, brucite, and periclase. The methylene blue method [35] and tablet method [36] were used to characterize the properties of tropical laterite soil.   Table 1 summarizes the soil physical and chemical properties, and Table 2 shows the general chemical composition of this lime and the selected lateritic soil. Kaolinite and quartz were the main minerals observed in the soil specimen; muscovite and iron oxides were also present. The lime samples presented calcite, portlandite, brucite, and periclase. The methylene blue method [35] and tablet method [36] were used to characterize the properties of tropical laterite soil.

Sample Preparation for Soil-Lime Stabilization
The tests were carried out on the untreated (natural soil) and lateritic soil samples treated with 1, 2, and 4% lime by weight percentage (i.e., water-lime solution). The lime content was determined by the sample volume and not by the total dry mass of the soil. The industrialized hydrated lime powder was added to 950 mL of distilled water in a beaker following the respective lime content and was then constantly agitated with a glass rod to achieve the slurry form (i.e., milk of lime), which was carefully spread over the soil. In addition, another specimen was treated with a suspended liquid mixture of water and 1% lime content that had not been agitated. This procedure permitted the precipitation of lime and used the minimal suspended lime fraction dissolved in the water to test the reaction of the lowest stabilizer content.
The curing times employed in the present study were 1, 7, 28, and 56 days using air-drying storage (in an air-conditioned laboratory room). For the 56-day curing time, test storage in a moist room, with relative humidity (RH) higher than 90% and temperature of 24 ± 3 • C, was included to produce uniform wetting of the sample and to compare the curing conditions. The lime contents and curing times used in the present work were based on the same methods adopted by Nascimento et al. [18] except for the curing storage and the solution with 1% lime content.

Soil-Lime Stabilization Mapping Analysis
The microstructure of untreated soil was investigated by the scanning electron microscopy (SEM) technique. To realize the elemental (i.e., calcium) mapping analysis of the treatment range, an energy-dispersive X-ray spectrometer (EDS) fitted on the SEM was used to scan the occurrence of calcium on the surface of treated soil before the SEM analysis. Furthermore, the powder X-ray diffraction (XRD) technique was used over the areas defined by the EDS mapping analysis to identify new crystalline compounds due to long-term processes. Furthermore, the results of the X-ray fluorescence spectroscopy (XRF) analysis of the lateritic soil and lime supported the interpretation of the XRD data. In addition, the pH of lateritic soil for all sets of lime contents and time periods was examined by a pH meter to verify the alkalinity condition of the soil to establish the pozzolanic reactions.

Erodibility Test
The erodibility test was performed in a wave flume, constructed by Menezes [29], which simulates the erosive effect of wave impacts on reservoir banks with a certain slope inclination. The tests were carried out according to the weighing method proposed by Schliewe [30], with a slope of 45 • , a frequency of 0.5 Hz, and a 6 h duration. Those parameters were defined by analysis of the results from which the medium erodibility values were selected [30]. The quantification of soil mass loss was carried out by three different techniques. The first method of determining the total mass of eroded sediments was performed by finding the difference between the initial and final masses of the soil, obtained by Equation (1): where m t is the total mass of sediments in grams, m i is the initial soil mass in grams, w i is the initial moisture content, m f is the final soil mass in grams, and w f is the final moisture content. The second method began with the collection of sediments using an industrial vacuum cleaner. This collected mud-like sediment was poured onto 2.00-mm and 75-µm sieves and retained in a bucket. A small part of the suspended sediments (particle size smaller than 2.00 mm) was collected in the bucket (150 mL) and dried in beakers on a hot plate, and soil particles that were retained by the sieves (particle size larger than 2.00 mm) were weighed and dried in an oven at a temperature of 105 to 110 • C for 24 h. The particle size distribution of all the collected sediments was found by a laser diffraction Water 2020, 12, 3141 6 of 15 particle analyzer (LDPA) device. The total mass of sediments was determined by the sum of the sediments collected in the channel, expressed by Equation (2): where m t is the total mass of sediments in grams, m(>2 mm) is the dry mass of sediment retained in the 2-mm sieve, m(>75 µm) is the dry mass of sediment retained in the 75-µm sieve, and m(<75 µm) is the calculated mass of sediments smaller than 75 µm, all measured in grams. The third method, using the 3D laser scanner, allowed the determination of the total mass of sediments based on the cubage of the volumes between scans before and after the test multiplied by the specific gravity of the soil, expressed by Equation (3): where m t is the total mass of sediments in grams, z is the z-axis of the Cartesian coordinate system, z 1a is the minimum z-value, z 2 is the maximum z-value, V i is the total volume given by the initial surface in cubic centimeters, V f is the total volume given by the final surface in cubic centimeters, and ρ d is the dry density of the soil sample in grams per cubic centimeter.
Laser scanners can record small-scale soil losses with high precision [37]. This work was performed using a mesh resolution of 0.5 mm to obtain a minimum data value that gives an accurate image of the soil sample and for a better computational processing time. Figure 2 shows the results of 3D scanning techniques of soil treated with lime in the sample holder before and after the wave flume test. Figure 2b presents a large erosion (soil loss of 78.24%) of the soil generated by the loss of calcium carbonate crust during the erodibility experiment.
Water 2020, 12, x FOR PEER REVIEW 6 of 16 where mt is the total mass of sediments in grams, m(>2 mm) is the dry mass of sediment retained in the 2-mm sieve, m(>75 μm) is the dry mass of sediment retained in the 75-μm sieve, and m(<75 μm) is the calculated mass of sediments smaller than 75 μm, all measured in grams. The third method, using the 3D laser scanner, allowed the determination of the total mass of sediments based on the cubage of the volumes between scans before and after the test multiplied by the specific gravity of the soil, expressed by Equation (3): where mt is the total mass of sediments in grams, z is the z-axis of the Cartesian coordinate system, z1a is the minimum z-value, z2 is the maximum z-value, Vi is the total volume given by the initial surface in cubic centimeters, Vf is the total volume given by the final surface in cubic centimeters, and ρd is the dry density of the soil sample in grams per cubic centimeter. Laser scanners can record small-scale soil losses with high precision [37]. This work was performed using a mesh resolution of 0.5 mm to obtain a minimum data value that gives an accurate image of the soil sample and for a better computational processing time. Figure 2 shows the results of 3D scanning techniques of soil treated with lime in the sample holder before and after the wave flume test. Figure 2b presents a large erosion (soil loss of 78.24%) of the soil generated by the loss of calcium carbonate crust during the erodibility experiment.  Figure 3 shows the photomicrographs of this sample at three different magnifications. The microfabric appearance of lateritic soil was a granular particle matrix, in which the presence of micropores and macropores could be observed in an agglomeration fabric of euhedral quartz grains ( Figure 3a) associated with clay minerals. This soil characteristic showed an advanced stage of weathering since the greater the degree of weathering, the greater the aggregations and the fewer the dispersed clay particles [36,38]. Figure 3b shows a larger macropore between the aggregations. The fine particles of layered kaolinite and oxyhydroxides were assembled to quartz grains in an aggregation of soil particles, as presented in Figure 3c. Both parallel and perpendicular observations of the surface sample showed non-typical clay crystal morphologies.  Figure 3 shows the photomicrographs of this sample at three different magnifications. The microfabric appearance of lateritic soil was a granular particle matrix, in which the presence of micropores and macropores could be observed in an agglomeration fabric of euhedral quartz grains ( Figure 3a) associated with clay minerals. This soil characteristic showed an advanced stage of weathering since the greater the degree of weathering, the greater the aggregations and the fewer the dispersed clay particles [36,38]. Figure 3b shows a larger macropore between the aggregations. The fine particles of layered kaolinite and oxyhydroxides were assembled to quartz grains in an aggregation of soil particles, as presented in Figure 3c. Both parallel and perpendicular observations of the surface sample showed non-typical clay crystal morphologies. A comparison of EDS images of samples treated with 1, 2, and 4% lime at 7 and 56 days of curing in the first 1.5 mm depth obtained by SEM inspection is shown in Figure 4. It was observed that the higher the lime content added to the soil, the greater the thickness of the protective layer (calcium carbonate crust formed by the carbonation of lime by mineral precipitation over the soil surface during application of a lime solution). The treatment range did not exceed 0.5 mm for lime content up to 2%, and for the 4% content, it could settle for approximately 1 mm, suggesting that the effect of stabilization with lime occurs only in the superficial part of the soil by quick carbonation of lime in a Ca-rich crust. The depth of treatment range observed in the 12 samples tested did not exceed 1.5 mm.

Mineralogical and Microstructural Soil-Lime Stabilization Mapping
SEM images of typical calcium carbonate morphologies that occurred at various curing times are shown in Figure 5. The magnified SEM image in Figure 5a shows the calcite mineral in a rhombohedral crystal form (1.25 μm) present in 1% lime at 7 days of curing, which indicated the occurrence of the carbonation process of the lime, with precipitation of calcium carbonate, even from low lime contents. Figure 5b shows several flower-shaped crystals as hexagonal platelets of calcium carbonate polymorph, identified in soil treated with 2% lime at 7 days of curing. SEM micrographs in Figure 5c revealed a disturbing irregular aggregated matrix near the top of the soil (1 mm deep) treated with 4% lime at 56 days.
This calcium carbonate crust produced by carbonation of lime fills and coats large and small voids, causing the formation of whitish cementation compounds binding soil particles. Furthermore, this indicates a possible increase in strength with a reduction in the coefficient of permeability also observed in other research [18,39].
The micromorphological features of pozzolanic reactions generated CSH, CAH, and CASH (C = Ca, S = SiO2, A = Al2O3, and H = H2O), which are already known for their cementitious improvement. These new minerals were not seen in any of the samples analyzed by SEM and XRD. This behavior is because carbonation generates insoluble calcium carbonate (i.e., calcite) from the reaction of hydrated lime [Ca(OH)2] with carbon dioxide (CO2) from the atmosphere. When the lime in slurry form is spread over the soil during the treatment, carbonation occurs almost immediately. However, the free calcium remaining after the carbonation may not available to produce this pozzolanic reaction. A comparison of EDS images of samples treated with 1, 2, and 4% lime at 7 and 56 days of curing in the first 1.5 mm depth obtained by SEM inspection is shown in Figure 4. It was observed that the higher the lime content added to the soil, the greater the thickness of the protective layer (calcium carbonate crust formed by the carbonation of lime by mineral precipitation over the soil surface during application of a lime solution). The treatment range did not exceed 0.5 mm for lime content up to 2%, and for the 4% content, it could settle for approximately 1 mm, suggesting that the effect of stabilization with lime occurs only in the superficial part of the soil by quick carbonation of lime in a Ca-rich crust. The depth of treatment range observed in the 12 samples tested did not exceed 1.5 mm. Water 2020, 12, x FOR PEER REVIEW 8 of 16  SEM images of typical calcium carbonate morphologies that occurred at various curing times are shown in Figure 5. The magnified SEM image in Figure 5a shows the calcite mineral in a rhombohedral crystal form (1.25 µm) present in 1% lime at 7 days of curing, which indicated the occurrence of the carbonation process of the lime, with precipitation of calcium carbonate, even from low lime contents.   According to Bell [19], the maximum modification of soil properties proceeds at an optimum lime content of 1-3% by weight of soil, which is far above the quantities of lime applied by weight percentage in this work. In addition, the advanced degree of weathering [40] may affect the amount of calcium available due to the high contents of iron and aluminum oxide-hydroxide [41].

Erodibility Test Results
The scanning method takes into account only the volumetric variation obtained between the 3D meshes if there are losses below the carbonate crust that cannot be registered with the handheld scanner. The values will be smaller than the real one for negative slopes, for example. The method of weighing can be influenced by the variation of the specific weight of the soil, microstructures, and humidity. The sieving method is the closest to weighing but is strongly influenced by unwanted This calcium carbonate crust produced by carbonation of lime fills and coats large and small voids, causing the formation of whitish cementation compounds binding soil particles. Furthermore, this indicates a possible increase in strength with a reduction in the coefficient of permeability also observed in other research [18,39].
The micromorphological features of pozzolanic reactions generated CSH, CAH, and CASH (C = Ca, S = SiO 2 , A = Al 2 O 3 , and H = H 2 O), which are already known for their cementitious improvement. These new minerals were not seen in any of the samples analyzed by SEM and XRD. This behavior is because carbonation generates insoluble calcium carbonate (i.e., calcite) from the reaction of hydrated lime [Ca(OH) 2 ] with carbon dioxide (CO 2 ) from the atmosphere. When the lime in slurry form is spread over the soil during the treatment, carbonation occurs almost immediately. However, the free calcium remaining after the carbonation may not available to produce this pozzolanic reaction.
According to Bell [19], the maximum modification of soil properties proceeds at an optimum lime content of 1-3% by weight of soil, which is far above the quantities of lime applied by weight percentage in this work. In addition, the advanced degree of weathering [40] may affect the amount of calcium available due to the high contents of iron and aluminum oxide-hydroxide [41].

Erodibility Test Results
The scanning method takes into account only the volumetric variation obtained between the 3D meshes if there are losses below the carbonate crust that cannot be registered with the handheld scanner. The values will be smaller than the real one for negative slopes, for example. The method of weighing can be influenced by the variation of the specific weight of the soil, microstructures, and humidity. The sieving method is the closest to weighing but is strongly influenced by unwanted sediments from previous tests. Figure 6 shows the relation between these methods (Figure 6), and a value of R 2 (coefficient of determination) of about 95% was achieved, showing a good correlation between them. The weighing and sieving methods presented the best approaches.
When the soil loss indicated low erosion, with values below 5 kg/m 2 , the scanning technique tended to approximate what occurred in the visual aspect of the sample. Therefore, the scanning and weighing values were closer when the loss was above 70 kg/m 2 (high erosion). Medium erosion, with results between 5 and 70 kg/m 2 , showed a great variance within each methodology. This occurs because the dry density of the soil considered in Equation (3) was obtained from a small fraction of the undisturbed soil block, and the whole sample was represented by microtubules and macropores, which could generate small density variations [30].
Water 2020, 12, x FOR PEER REVIEW 9 of 16 sediments from previous tests. Figure 6 shows the relation between these methods (Figure 6), and a value of R² (coefficient of determination) of about 95% was achieved, showing a good correlation between them. The weighing and sieving methods presented the best approaches. When the soil loss indicated low erosion, with values below 5 kg/m², the scanning technique tended to approximate what occurred in the visual aspect of the sample. Therefore, the scanning and weighing values were closer when the loss was above 70 kg/m² (high erosion). Medium erosion, with results between 5 and 70 kg/m², showed a great variance within each methodology. This occurs because the dry density of the soil considered in Equation (3) was obtained from a small fraction of the undisturbed soil block, and the whole sample was represented by microtubules and macropores, which could generate small density variations [30].
A one-way ANOVA model (α = 0.05) was used to determine the interaction between the results of the three methods of evaluation of the soil loss, and there was no significant difference in the amount of soil erosion according to the value of p = 0.47 (p > 0.05). However, the values had a high dispersion between methods according to the coefficient of variation among groups of 119%, due to the different characteristics of the methods and the use of the specific gravity of the soil, which has great dispersion in the case of natural soil. It is assumed that the values are, in general, within the ranges verified by Schliewe [30].
Based on the results obtained from the three methods of soil loss calculation along with the simplicity of the test, the weighing method was selected for the subsequent analyses of the influence of curing conditions, curing time, and lime content. However, the use of a 3D scanner creates a digital morphology of soil erosion, which is a technological advantage in comparison with other methods because it expands the possibilities of the analyses, such as the evaluation of the dynamics of the erosion process through microtopographic changes.

Effects of Curing Conditions
The results of the comparison of samples cured for 56 days with lime contents of 1, 2, and 4% by weight percentage are shown in Figure 5. Air-drying storage led to a reduction in water content because the treated sample surface was exposed to air (red line in Figure 7), and there was an increase in water content due to hydration of the soil in the moist room (blue line in Figure 7a). It was observed that the curing condition of air-drying storage showed a large soil mass loss (over 40%) during the erodibility test compared to curing in a moist room (below 10%) due to the variation of the water content of the samples, which resulted in suction variation and shrinkage (Figure 7b). According to Almeida et al. [42], the erodibility rates rose due to the increase in soil suction (reduction in water content), and with the shrinkage, the soil structure becomes more susceptible to the loss of mass by the action of the wave. A one-way ANOVA model (α = 0.05) was used to determine the interaction between the results of the three methods of evaluation of the soil loss, and there was no significant difference in the amount of soil erosion according to the value of p = 0.47 (p > 0.05). However, the values had a high dispersion between methods according to the coefficient of variation among groups of 119%, due to the different characteristics of the methods and the use of the specific gravity of the soil, which has great dispersion in the case of natural soil. It is assumed that the values are, in general, within the ranges verified by Schliewe [30].
Based on the results obtained from the three methods of soil loss calculation along with the simplicity of the test, the weighing method was selected for the subsequent analyses of the influence of curing conditions, curing time, and lime content. However, the use of a 3D scanner creates a digital morphology of soil erosion, which is a technological advantage in comparison with other methods because it expands the possibilities of the analyses, such as the evaluation of the dynamics of the erosion process through microtopographic changes.

Effects of Curing Conditions
The results of the comparison of samples cured for 56 days with lime contents of 1, 2, and 4% by weight percentage are shown in Figure 5. Air-drying storage led to a reduction in water content because the treated sample surface was exposed to air (red line in Figure 7), and there was an increase in water content due to hydration of the soil in the moist room (blue line in Figure 7a). It was observed that the curing condition of air-drying storage showed a large soil mass loss (over 40%) during the erodibility test compared to curing in a moist room (below 10%) due to the variation of the water content of the samples, which resulted in suction variation and shrinkage (Figure 7b). According to Almeida et al. [42], the erodibility rates rose due to the increase in soil suction (reduction in water content), and with the shrinkage, the soil structure becomes more susceptible to the loss of mass by the action of the wave.
The curing conditions of the soil (i.e., temperature and moisture) play an important role in the increase in soil strength [18,43]. The results of this work showed similarity to the effect of curing in concrete, where moist curing promoted a strength gain in concrete compared to air-drying [44], and a reduction in the moist-curing period resulted in lower strengths [45].
The behavior of the soil in the field should be in an intermediate condition between wetting and drying cycles because the temperature and moisture below ground on the shores of the reservoir are greatly influenced by the water table. When keeping the soil in a wet condition, an increase in the lime content can reduce the erosiveness of wave action. Furthermore, the dry soils are more vulnerable to volumetric changes and enable a breach in this protective lime layer. Once the crust is cracked, soil loss is facilitated by the flow of water and rapid saturation in porous media with low water content, allowing sediment detachment. The curing conditions of the soil (i.e., temperature and moisture) play an important role in the increase in soil strength [18,43]. The results of this work showed similarity to the effect of curing in concrete, where moist curing promoted a strength gain in concrete compared to air-drying [44], and a reduction in the moist-curing period resulted in lower strengths [45].
The behavior of the soil in the field should be in an intermediate condition between wetting and drying cycles because the temperature and moisture below ground on the shores of the reservoir are greatly influenced by the water table. When keeping the soil in a wet condition, an increase in the lime content can reduce the erosiveness of wave action. Furthermore, the dry soils are more vulnerable to volumetric changes and enable a breach in this protective lime layer. Once the crust is cracked, soil loss is facilitated by the flow of water and rapid saturation in porous media with low water content, allowing sediment detachment.

Effects of Curing Time
The effect of curing time on soil-lime stabilization showed an increase in mechanical properties (mostly compressive strength) of the soil using dry soil-lime mixtures, higher quantities of lime, and dry curing, due to the formation of new cementitious compounds [19,27,39,46]. Unlike other works on soil-lime treatment, this study verified a partial mechanical improvement in erosion resistance on the soil surface by the influence of the lime carbonation.
The fact that the air-drying and the curing time influenced the results of the wave tests can be verified in Figure 8. The higher soil losses were due to the loss of water content stored at an extended period under air-drying conditions. However, Figure 9 reveals the opposite, a decrease in the mass loss for samples stored in a moist room at a curing time of 56 days. Nascimento et al. [18] verified a slight decrease in soil loss of samples with the same preparations and curing times under air-drying.
The reduction in the water content caused a more significant volume shrinkage of the soil after 7 days of curing, and this completely ignored the protection of the soil by lime treatment and allowed easy entry of water into a fragile structure. Besides, at 28 days, 4% lime treatment showed a lower soil loss of 16%, despite the results for 1 and 2% lime solutions. This can be explained by the formation of more calcium crust by carbonation and lower soil shrinkage, compared to other samples at 28 days of curing time.

Effects of Curing Time
The effect of curing time on soil-lime stabilization showed an increase in mechanical properties (mostly compressive strength) of the soil using dry soil-lime mixtures, higher quantities of lime, and dry curing, due to the formation of new cementitious compounds [19,27,39,46]. Unlike other works on soil-lime treatment, this study verified a partial mechanical improvement in erosion resistance on the soil surface by the influence of the lime carbonation.
The fact that the air-drying and the curing time influenced the results of the wave tests can be verified in Figure 8. The higher soil losses were due to the loss of water content stored at an extended period under air-drying conditions.   However, Figure 9 reveals the opposite, a decrease in the mass loss for samples stored in a moist room at a curing time of 56 days. Nascimento et al. [18] verified a slight decrease in soil loss of samples with the same preparations and curing times under air-drying.

Effects of Lime Content
The lime content wielded a great influence on the reduction of soil mass loss, as observed in Figure 9. All values for 4% lime samples, except for the 56-day air-dried sample, showed soil loss below the untreated soil reference line (dashed black line). Nascimento et al. [18] obtained significant reductions, with only 1% lime by weight percentage when simulating runoff erosion. A reduction of about 72% was obtained concerning the natural sample and the saturated condition.
As the lime content increased, the crust of calcium carbonate generated by carbonation rose as well. The major reduction of soil loss verified in Figure 7 was due to the thickness of slaked lime spread over the soil, as shown in the SEM micrographs in Figure 4. This crust gave protection to the soil surface, allowing resistance to the impact of waves and variation of the water level during the period of the test.

Influence of Soil-Lime Treatment on Water pH
According to policy and regulations under Brazilian law [47], the National Council of Environment (CONAMA, in Portuguese), which provides the groundwater classification and environmental guidelines, it is established that fresh water for human consumption from groundwater, lakes, and rivers must have a pH value between 6.0 and 9.0. In addition, the Brazilian Ministry of Health recommends monitoring by sanitation companies and controls the pH of fresh water for human consumption in the range of 6.0 to 9.5 [48]. The reduction in the water content caused a more significant volume shrinkage of the soil after 7 days of curing, and this completely ignored the protection of the soil by lime treatment and allowed easy entry of water into a fragile structure. Besides, at 28 days, 4% lime treatment showed a lower soil loss of 16%, despite the results for 1 and 2% lime solutions. This can be explained by the formation of more calcium crust by carbonation and lower soil shrinkage, compared to other samples at 28 days of curing time.

Effects of Lime Content
The lime content wielded a great influence on the reduction of soil mass loss, as observed in Figure 9. All values for 4% lime samples, except for the 56-day air-dried sample, showed soil loss below the untreated soil reference line (dashed black line). Nascimento et al. [18] obtained significant reductions, with only 1% lime by weight percentage when simulating runoff erosion. A reduction of about 72% was obtained concerning the natural sample and the saturated condition.
As the lime content increased, the crust of calcium carbonate generated by carbonation rose as well. The major reduction of soil loss verified in Figure 7 was due to the thickness of slaked lime spread over the soil, as shown in the SEM micrographs in Figure 4. This crust gave protection to the soil surface, allowing resistance to the impact of waves and variation of the water level during the period of the test.

Influence of Soil-Lime Treatment on Water pH
According to policy and regulations under Brazilian law [47], the National Council of Environment (CONAMA, in Portuguese), which provides the groundwater classification and environmental guidelines, it is established that fresh water for human consumption from groundwater, lakes, and rivers must have a pH value between 6.0 and 9.0. In addition, the Brazilian Ministry of Health recommends monitoring by sanitation companies and controls the pH of fresh water for human consumption in the range of 6.0 to 9.5 [48]. Figure 10 presents the results obtained by the direct measurement of water before and after the test in three sectors. The pH of the water supply system was 7.87. The mean values obtained before and after the test were below the natural pH value of the water, but, in general, the pH values werre higher after the test. In addition, the values of water pH nearest to the treated soil (<2.00 m) were higher than the values collected far away from the sample in the wave flume (>4.00 m). Figure 10 presents the results obtained by the direct measurement of water before and after the test in three sectors. The pH of the water supply system was 7.87. The mean values obtained before and after the test were below the natural pH value of the water, but, in general, the pH values werre higher after the test. In addition, the values of water pH nearest to the treated soil (<2.00 m) were higher than the values collected far away from the sample in the wave flume (>4.00 m). Another test was conducted to verify the influence of lime on increasing the pH of the water, and it was found that upon the addition of 0.003 g (1% lime) of hydrated lime in 100 mL of water, the pH increased by 3%. With 0.013 g (4% lime) of hydrated lime in the same amount of water, the pH rose by 12%.
This work was concerned that the soil-lime technique could bring more damage to the reservoir shoreline through environmental issues. However, it seems that there will be no problems with dust since the lime is applied in liquid form directly on the ground (without vegetation). The white visual aspect can be circumvented by the use of brown color limewash pigments. The soil becomes less acidic, and the water pH rises within the required values by the water quality control regulations.

Conclusions
A thin layer of calcium carbonate (<2 mm) was the major result of spreading lime in slurry form over the soil with the aim of soil stabilization without the formation of cementitious compounds between the edges of the clay particles. Thus, the stabilization of the soil was restricted only by the efficiency and durability of this white crust of calcium carbonate. The use of lime of soil-lime stabilization at the shoreline will not affect the chemistry of fresh water for human consumption and will probably have little influence on the reservoir's biological life, considering the values stipulated by Brazilian policy.
The lime content was important for the thickening of the calcium carbonate crust together with wet curing conditions, which contributed to the improvement in the resistance of the soil against the erosivity of the waves, but the curing time did not have a significant impact because the water content variation was more important in the loss mass. In this way, in the field, soil moisture could influence the efficiency of this technique because when subjecting the soil samples to air-drying for a long period, the potential weakening of the soil structure was observed, which allowed mass loss. In periods of drought, the protective layer of lime can contract, leading to cracks that allow water infiltration and surface runoff at shore banks, increasing the soil erodibility.
The advantages of hydrated lime slurry as limewash for shoreline soil stabilization are as follows: (a) The formation of a protective crust occurs, reducing the soil mass loss due to the water Another test was conducted to verify the influence of lime on increasing the pH of the water, and it was found that upon the addition of 0.003 g (1% lime) of hydrated lime in 100 mL of water, the pH increased by 3%. With 0.013 g (4% lime) of hydrated lime in the same amount of water, the pH rose by 12%.
This work was concerned that the soil-lime technique could bring more damage to the reservoir shoreline through environmental issues. However, it seems that there will be no problems with dust since the lime is applied in liquid form directly on the ground (without vegetation). The white visual aspect can be circumvented by the use of brown color limewash pigments. The soil becomes less acidic, and the water pH rises within the required values by the water quality control regulations.

Conclusions
A thin layer of calcium carbonate (<2 mm) was the major result of spreading lime in slurry form over the soil with the aim of soil stabilization without the formation of cementitious compounds between the edges of the clay particles. Thus, the stabilization of the soil was restricted only by the efficiency and durability of this white crust of calcium carbonate. The use of lime of soil-lime stabilization at the shoreline will not affect the chemistry of fresh water for human consumption and will probably have little influence on the reservoir's biological life, considering the values stipulated by Brazilian policy.
The lime content was important for the thickening of the calcium carbonate crust together with wet curing conditions, which contributed to the improvement in the resistance of the soil against the erosivity of the waves, but the curing time did not have a significant impact because the water content variation was more important in the loss mass. In this way, in the field, soil moisture could influence the efficiency of this technique because when subjecting the soil samples to air-drying for a long period, the potential weakening of the soil structure was observed, which allowed mass loss. In periods of drought, the protective layer of lime can contract, leading to cracks that allow water infiltration and surface runoff at shore banks, increasing the soil erodibility.
The advantages of hydrated lime slurry as limewash for shoreline soil stabilization are as follows: (a) The formation of a protective crust occurs, reducing the soil mass loss due to the water erosion process. (b) The low lime content already allows the formation of a protective layer. This technique may be feasible considering some factors: (i) Soil protection should be preserved by maintaining the carbonate layer and controlling eventual cracks. It is necessary to consider the frequent application of lime, according to Santos [24]. (ii) The soil moisture condition should be maintained. (iii) The access of people, vehicles, and large animals should be restricted to avoid damaging the protective crust.
In future works, the soil suction process should be evaluated, and experimental field programs must be realized. It is also necessary to find the optimum lime content (OLC) for field application because lime slurry causes clogging of the pressure pump, making it difficult to pump and spray the lime (about 32% content) with an immobile slurry system. The use of lime allowed the reduction of soil loss under the effect of waves under specific conditions, and this result supports other studies of soil erosion control [24].
Although the experimental procedures presented here were applied to a lareritic clay from the Midwest Brazilian Region, the proposed technique can be applied to other types of clay soils. For this, studies are needed to define the solution contents and optimal preparation conditions. Laterite soil is widespread in wet equatorial parts of South America, Africa, Asia, and the Pacific.