1 Introduction

Soil excavation is a frequently employed practice for construction projects, which may involve the office and parking structures, storage spaces, mechanical and electrical compartments, and residential dwellings. Soil excavation activities often involve the utilization of a stable slope as a precautionary measure to reduce the likelihood of soil collapse in the vicinity of the excavation site. The process of soil excavation within the boundaries of a construction site is frequently limited by the property lines and the surrounding conditions, particularly in urban area. The utilization of a temporary earth-retaining structure as a proactive measure to mitigate soil collapse is a widely employed practice. The design of excavation operations is influenced by various factors, such as the depth of excavation, geotechnical characteristics, groundwater levels, vertical and lateral loads exerted by adjacent structures, and the presence of construction equipment and material storage. Different varieties of retaining walls can be utilized depending on the particular construction procedures and methodologies employed.

In recent years, there has been a noticeable rise in the adoption of recycled water tanks in the Bangkok region. The increase in numbers that has been observed can be attributed to the rapid process of urbanization in a limited area, leading to a greater frequency of excavation activities characterized by shallow depths (ranging from 3 to 5 m) and narrow dimensions. The utilization of a temporary retaining wall structure plays a crucial role in underground excavation projects, primarily owing to the distinctive characteristics of soft Bangkok clay, particularly its low shear strength and stiffness. The use of steel sheet pile walls is a common method for constructing temporary retaining walls in various excavation projects in the city of Bangkok. However, the constrained mobility of steel sheet piles and large machinery frequently presents a substantial challenge in construction areas characterized by limited spatial capacity. In addition, the installation of sheet piles has the potential to induce ground vibration and soil displacement, which can have adverse impacts on adjacent structures. The application of the deep cement-soil mixing (DCM) technique in the construction of soil–cement column (SC) walls offers a feasible alternative for the development of temporary retaining walls. The SC walls use local soils as the primary material, mixed with a certain percentage of cement, which can reduce the disruption to the local ecosystem. In addition, the installation of the SC walls often requires less energy. In terms of environmental impact, it contributes to the reduction of emissions associated with material production, transportation, and construction phases.

The Deep Cement Mixing (DCM) method is a well-established and effective technique utilized to improve soil characteristics by introducing cementitious materials, such as lime, cement slurry, and cement mortar, through the injection process [1,2,3,4]. The DCM technique has been successfully employed in various projects such as buildings, roads, railways, embankments, highways, and airports [3,4,5,6,7]. The successful implementation of this method in Thailand's Bangna-Bangpakong highway is documented in reference [8]. The aforementioned technique was employed in order to enhance the load-bearing capacity and address settlement concerns. Jamsawang et al. (2015) [9] utilized two kinds of piles in the construction of Rama Hospital, located in Bangplee, Samutprakarn Province, in close proximity to the urban region of Bangkok. Type 1 was characterized by SC, featuring a diameter measuring 0.5 m and a length spanning 10.0 m. On the other hand, Type 2 was (SSC), which involved the inclusion of eucalyptus wood with dimensions of 0.15 m in diameter and 6.0 m in length in the central area of the SC. It was observed that SSC demonstrated a recorded ultimate load capacity of 250 kN, indicating a notable enhancement of approximately 30% in comparison to the load capacity exhibited by SC. A comparative analysis was undertaken to assess the efficacy of SC and SSC by means of full-scale pile load tests carried out in the context of soft Bangkok clay [10]. The diameter of the SC was 0.60 m, while its length measured 7.0 m. The SSC system consists of SC that incorporates a precast reinforced concrete cored pile with a diameter of 0.22 m. The research conducted revealed that the SSC demonstrated a vertical load capacity that was approximately 2.2 times higher than that of the SC. Additionally, the lateral load capacity of the SSC was determined to be approximately 15 times greater than that of the SC.

Researchers have investigated the behavior of a test embankment that was constructed using SC in the Bangkok clay [11,12,13]. Vootipruex et al. (2011a) undertook a thorough examination comprising a full-scale embankment load test and finite element simulation. The objective was to compare the performance of SSC with precast reinforced concrete core piles with that of SC [14]. The findings of the study indicate that the vertical and laterial load capcities of SC was approximately 2.2 and 15 times lower than those of SSC.

The application of SC was also documented in the context of excavation methodologies for erecting the foundation and basement of high-rise buildings [15]. The research conducted by Jamsawang et al. (2017) [16] investigated the efficacy of a SC wall in the context of an unbalanced deep excavation in soft Bangkok clay, employing a top-down support system. According to the findings presented in the report, the utilization of the SC wall exhibited a notable reduction in the impact of uneven pressure. The observed ground displacement during the excavation process remained within an acceptable range, thereby avoiding any structural damage to the unrestrained pile. Tanseng (2012) [17] reported that the utilization of a wall-strut system incorporated with SC walls held promise in alleviating ground displacement during tunnel excavation in the soft clay deposits. The lateral displacement observed in this study was comparatively smaller when compared to the lateral displacement exhibited by the conventional sheet pile and diaphragm wall.

Extensive research has been conducted to primarily examine the stability characteristics of SC and SSC walls installed in Bangkok clay via finite element modelling (FEM) techniques and executed full-scale tests. Furthermore, a considerable proportion of the academic literature pertaining to composite soil–cement retaining walls in different countries has predominantly focused on examining the mechanical characteristics, such as soil movement, the lateral displacement of the retaining wall, and the safety factors of retaing walls [18,19,20,21,22,23,24]. To the best of the authors' knowledge, no previous research has investigated the relationship between execution time, cost, and the stability of SC and SSC. Understanding the relationship between these factors can help optimize the deep mixing method for SCC and SSCC walls, leading to improved efficiency, reduced construction costs, and better overall performance. In addition, it can contribute to better project planning and decision-making in geotechnical engineering projects, which offers environmental, social, and economic benefits that contribute to the overall sustainability of construction projects.

The current investigation employed the finite element (FE) method to perform a stability analysis, specifically comparing SC and SSC walls with the conventional sheet pile wall system. The initial calibration of the finite element (FE) model entailed a comparative analysis between the simulated results of SC and SSC walls that were constructed at a site in Bangkok, and the corresponding measurements taken in the field. The calibration process refined the FE model and improved its accuracy and reliability for research purposes. This ensures that the FE model is robust and reliable for predicting the behavior of SCC and SSCC walls. After calibrating the finite element (FE) model, a subsequent parametric analysis was conducted to examine the impact of different factors on wall displacement, stability, and the associated costs and construction duration. The parameters investigated for their impact were the structural attributes of the SC and SCC walls, specifically their patterns and lengths. The results of this research will contribute to the improvement of the efficacy in the design and selection process of SC and SSC walls, along with sheet pile walls, for their application as temporary support systems in shallow and narrow excavation aplications within the context of soft Bangkok clay. This will guarantee that these systems maintain an adequate safety factor while also being cost-effective and time-efficient.

2 Field case analysis

2.1 Comprehensive description of the project

The construction site was situated within the Bangkok-Noi District of Bangkok, Thailand. The geometric depiction of the location can be observed in Fig. 1a. The implementation of three substantial recycled water tanks required an excavation that extended to a maximum depth of 4.5 m. The excavation being discussed was classified as a shallow and narrow excavation, typically seen in recycled water tank initiatives. The application of a sheet pile wall with bracing, commonly utilized in Bangkok area, was determined to be inappropriate for the particular site under consideration. The primary reason for this was the hindrance caused by preexisting structures situated at the entrance of the installation device (Fig. 1a). Two distinct versions of soil cement column walls were devised as temporary structures for implementation in this specific project, taking into account the constraints imposed by the property boundary of the construction site. The aforementioned variants are commonly known as the soil–cement column (SC) Wall and the stiffened soil cement column (SSC) Wall. The SC-IIIRow Wall, comprising of the SC with a diameter measuring 0.6 m and a length of 12.0 m, was arranged into three rows. The process of installing the SSC-IRow Wall entailed the insertion of a steel pipe measuring 0.2 m in diameter and 9.0 m in length into the SCC. This practice was implemented in regions where the construction of a thick SC wall was deemed impermissible. Figures 1b and c illustrate the configurations and divisions of SC-IIIRow Wall and SSC-IRow Wall, respectively. This study sought to examine the enhancement of wall design through stability and deformation analyses. The optimization process involved refining the layouts, patterns, and embedded length of SC and SSC in order to maximize the performance of the retaining wall. The goal is to achieve the desired structural performance while minimizing material usage, construction time, and costs.

Fig. 1
figure 1

a Site geometry, and types of retaining wall that construction: (b) SC-3Row Wall and (c) SSC-1Row Wall

Full size image

Figure 2 depicts the sequential arrangement of construction activities executed at the project site. The combined duration for the construction of the retaining wall and the installation of recycled tanks amounted to approximately 56 days. The construction process initiated with the implementation of bored piles, a procedure that spanned approximately seven days. The process of mobilizing equipment and setting up the mixing plant was successfully accomplished within a span of five days. The initial installation of the SC-IIIRow Wall was conducted over a duration of 8.5 days, followed by the subsequent installation of the SSC-IRow Wall, which was completed within a period of 2.5 days. Following that, inclinometers were employed to quantify the horizontal displacements that occurred behind the SC-IIIRow Wall and SSC-IRow Wall throughout the excavation process and subsequent installation of recycled water tanks (refer to Fig. 1a). After the inclinometers were installed for a period of 10 days, the excavation procedure was commenced.

Fig. 2
figure 2

Construction sequence

Full size image

The soil excavation process was carried out in two separate phases. The first phase of the project entailed the excavation of the soil to a depth of 2.0 m, which was carried out over a period of two days. Following this, the subsequent phase involved conducting further excavation in order to achieve a final depth of 4.5 m. This process necessitated an additional three days to be fully accomplished. The lateral displacements were initially measured at two distinct time intervals: 28 days following the implementation of the SC-IIIRow Wall and 20 days following the implementation of the SSC-IRow Wall. The duration of the tasks involving the cutting of the pile head, the construction of the lean concrete, and the building of the basement amounted to approximately five days. The process of installing columns to provide structural support for the recycled water tanks required a duration of approximately four days. This was followed by an additional four days for the subsequent installation of three recycled water tanks. The assessment of the second horizontal displacement was performed subsequent to the examination of leakage in the tanks containing recycled water. The evaluation was conducted at two specific time points: 41 days after the installation of the SC-IIIRow Wall and 33 days after the installation of the SSC-IRow Wall. The utilization of the filling sand followed in order to backfill the specified area, thereby facilitating the construction of the concrete pavement atop the water tanks. The aforementioned procedure was carried out within a time frame of three consecutive days.

2.2 Soil and SC properties

The soil profile observed at the construction site displayed a variety of characteristics, including the occurrence of both soft and stiff clay. The study identified a layer of soft clay that extended over a distance of approximately 10 m. The observed layer demonstrated a notably diminished undrained shear strength (Su) of approximately 20 kPa. Furthermore, it was observed that this stratum of malleable clay exhibited a significant moisture content, as illustrated in Fig. 3. The soil profile observed in the central region of Bangkok exhibits characteristics that align with the findings reported by Horpibulsuk et al. (2007) [25]. The soft clay demonstrated an inherent moisture content of approximately 80%. The liquid limit exhibited a range of 60% to 80%, whereas its plastic limit ranged from 20 to 30%. In addition, the coefficient of compression was determined to be 0.35, while the friction angle was measured to be 23 degrees.

Fig. 3
figure 3

The soil profile at the construction site, Bangkok-Noi District, Thailand

Full size image

After a construction period of 28 days, core samples were obtained from the central region of the SC wall at various depths utilizing a coring apparatus. The specimens were subsequently prepared for unconfined compressive strength (qu) testing. The qu samples were modified in order to attain a diameter-to-length ratio of 1:2, wherein the diameter measured 50 mm and the length measured 100 m. Figure 4 indicates the undrained shear strength, \({S}_{{u}_{SCC}}\left({S}_{{u}_{SCC}}=1/{2q}_{u}\right)\) values at various depths whereby the average \({S}_{{u}_{SCC}}\) value was about 535 kPa. The correlations between \({S}_{{u}_{SCC}}\) and the secant modulus of SCC assist in the estimation of the stress–strain properties used in the finite element simulation of excavation and retaining wall performance.

Fig. 4
figure 4

A relationship between depth and undrained shear strength of SC

Full size image

2.3 Calibration of the finite element model using field observations

Calibrating a finite element (FE) model of the deep mixing method for SCC and SSCC Walls with field measurements involves comparing the results of the FE analysis to actual measurements from field monitoring data. This process helps refine the FE model and improve its accuracy and reliability. This study applied three steps to calibrate the FE model with field measurements. First, field tests and monitoring data during and after the construction of SCC and SSCC walls were obtained. The data included in-situ soil properties, SCC and SSCC properties, and wall deformations. Next, a FE model was created to represent the actual construction site, which included defining the geometry boundary conditions, selecting suitable constitutive models for the soil and SCC/SSCC materials, and applying loads incorporating the stages of construction. Then, the FE analysis was performed to obtain predictions of the SCC and SSCC walls’ deformations. The accuracy of the finite element (FE) model was evaluated by comparing the predictions with the measured field data. If the FE model predictions deviated significantly from the field measurements, the model’s parameters or constitutive models were adjusted to improve the agreement between the two. This involved updating material properties, refining the mesh, or changing the boundary conditions. The FE calculations were repeated until the simulated results were in very good agreement with the field data. Once the best suitable FE model was obtained, this calibrated FE model was used to analyze the performance of SCC and SSCC in various scenarios (pattern and length of SCC and SSCC walls) and optimize the deep mixing methods, such as time and cost of construction, and stability aspect.

The utilization of a finite element (FE) model, which has been calibrated through the analysis of field monitoring data, can serve as a robust and effective method for forecasting the performance of retaining wall structures and systems. However, the FE model in this research involved simplifications and assumptions to make the problem more practical, such as linear elasticity, simplified constitutive models, or homogeneous soil layers and SCC and SSCC materials. These simplifications can introduce limitations in the model’s ability to accurately represent the true behavior of the system.

The research employed the Finite Element (FE) technique and the PLAXIS 2D software to examine and assess the effectiveness of the SC Wall and the SSC Wall that were constructed. The simulation of the soil, SC Wall, and SSC Wall was conducted using 15-node wedge elements. Voottipruex et al. (2011a; 2011b) [14, 26] investigated the behavior of deep cement mixing (DCM) and stiffened deep cement mixing (SDCM) piles on soft Bangkok clay by full-scale embankment loading and finite element analysis using PLAXIS program. In the context of finite element analysis, the Mohr–Coulomb failure criterion was employed to represent the mechanical behavior of DCM and SDCM piles. This criterion was assumed to be applicable to linearly elastic to perfectly plastic materials. Conversely, the soft soil model was utilized to characterize the behavior of soft clay in the finite element analysis. Results indicated that the field measurement and simulated measurement of both compression settlement and lateral movements were in good agreement. Similar case studies that used the Mohr–Coulomb model for DCM and SDCM walls in soft clay also confirmed the good agreement results between the measured lateral movement during and after construction and simulated lateral movement [9, 16, 27,28,29,30,31,32]. In this study, the Mohr–Coulomb model was utilized to examine the performance of the SC Wall and the SSC Wall. The Mohr–Coulomb model requires the consideration of five key soil parameters: cohesion (c), friction angles (ϕ), dilatancy angle (ψ), elastic modulus (E), and Poisson's ratio (v). These parameters are determined through conventional laboratory tests.

Prior research has employed the Mohr–Coulomb model to simulate soft clay owing to its inherent simplicity [28,29,30,31,32]. The Soft Soil model was utilized in this research to replicate the behavior of soft clay. The Soft Soil model is appropriate for materials that demonstrate a distinct level of compressibility, such as clays that are frequently consolidated [33]. For undrained behavior, the pore pressure increased during the short loading of a soft soil material. Consequently, the effective stress undergoes a reduction rather than an augmentation, thereby influencing the attainable maximum shear resistance. Precise assessment of the rise in pore pressure, decline in mean effective stress, and consequent diminution in undrained shear strength are pivotal considerations in the analysis of stability within the framework of this study.

The Soft Soil constitutive model was effectively employed in the computational examination of the mechanical characteristics of soft Bangkok clay, as demonstrated in references [34, 35]. The Soft Soil model necessitates the inclusion of three essential soil parameters as inputs: the modified compression index (λ*), the modified swelling index (κ*), and the unloading or reloading Poisson’s ratio. Furthermore, it should be noted that the Soft Soil model demonstrates a failure mechanism that conforms to the Mohr–Coulomb criterion. The determination of this criterion is influenced by the effective strength parameters, specifically the effective cohesion (c’) and effective friction angle (ϕ’). The study reported a range of friction coefficients, specifically between 0.03 and 0.05, at the interface between the soil cement column and the soft Bangkok clay [36].

Table 1 presents the model parameters pertaining to the Shear Strength of Cohesive (SCC) and Soft Clay materials. Table 2 provides a comprehensive summary of the diverse parameters linked to steel pipe. Figure 5 depicts the finite element mesh that was utilized in the simulation of the SCC Wall and SSCC Wall. The construction phases of the SC Wall were implemented in the subsequent manner: The process can be classified into five discrete stages: (1) the initiation of preliminary stresses, (2) the implementation of SC at a depth of 12 m, (3) the excavation of a depth measuring 2.0 m, (4) the excavation of a depth measuring 4.5 m, and (5) the calculation of the factor of safety (FS). The construction process of the SSC Wall encompassed multiple stages. Firstly, the generation of initial stresses occurred. Additionally, SCs were installed at a depth of 12 m. In addition, steel pipes were positioned within the core of the SCs. In addition, a dig with a depth of 2.0 m was carried out as the fourth stage, which was subsequently followed by a dig with a depth of 4.5 m in the fifth stage. The determination of the factor of safety (FS) was ultimately carried out during the sixth stage.

Table 1 Soil parameters for finite element analysis
Full size table
Table 2 Parameter of steel pipe and sheet pile (Plate) for finite element analysis
Full size table
Fig. 5
figure 5

FE mesh used for the back-analysis soil stiffness: a) SC-IIIRow Wall and b) SSC-IRow Wall

Full size image

The factor of safety (FS) was determined using the shear strength reduction (phi-c reduction) method implemented in the PLAXIS 2D software. The experimental procedure entailed a gradual reduction of the soil’s strength parameters, namely tanϕ and c, until the point of structural failure was observed. The calculation of the factor of safety (FS) can be obtained from the following equations:

$$\mathrm{FS}=\mathrm{value}\;\mathrm{of}\sum Msf\;\mathrm{at}\;\mathrm{failure}=\frac{\mathrm{available}\;\mathrm{strenght}}{\mathrm{strenght}\;\mathrm{at}\;\mathrm{failure}}$$
(1)
$$\sum\;Msf=\frac{{\tan\phi}_{input}}{{\tan\phi}_{reduced}}=\frac{c_{input}}{c_{reduced}}$$
(2)

where \(\mathrm{tan}{\varnothing }_{input}\) and \({c}_{input}\) are the strength parameters of the material specified as inputs during the modeling process and \(\mathrm{tan}{\varnothing }_{reduced}\) and \({c}_{reduced}\) are the diminished values acquired from the finite element program. \(\sum\,Msf\) was first set as 1.0 at the start of the calculation and the incremental \(Msf\) was used to specify the reduction of the strength from the first calculation step until its unreduced values (strength at failure).

3 Parametric study

3.1 Influence of stiffness on lateral movement

After the completion of the validation process for the finite element modeling, it was subsequently utilized to carry out a parametric study. The aim of this research was to investigate the impact of various influencing factors on the cost, duration, and structural integrity of different types of retaining walls, namely SC, SSC, and sheet pile with bracing. The implementation of bracing is necessary to mitigate significant lateral displacement due to the low stiffness of sheet piles. The act of inserting a steel pile into the wall of the SC has the potential to result in a reduction in the number of rows of SC. The incorporation of bracing further improves the structural integrity of the FS. The current study examined the SC-IIRow Wall, SSC-IRow Wall with and without a bracing system, and a sheet pile wall (Fig. 6). In the context of the SSC Wall with a bracing system, the steel piles were strategically placed along the perimeter of the SC structure. Each pile was installed at alternating C locations. The steel piles were subsequently connected to the wale and strut system via the process of welding. The tables provided in this study include the soil and pile structure parameters that were employed for finite element (FE) analyses. The tables, specifically Tables 1, 2, and 3, present a comprehensive compilation of the specific parameters utilized in the finite element analyses of the soil and pile structures. Figure 7 illustrates the finite element (FE) model and FE mesh of three distinct wall configurations, specifically the SC-IIRow Wall, SSC-IRow Wall with a bracing system, and sheet pile wall. The research study additionally incorporated the utilization of the phi-c reduction technique in finite element analyses to investigate the stability and lateral displacement properties of various categories of SC and SSC walls. Subsequently, a comparison was made between the aforementioned findings and those derived from the sheet pile wall system. To facilitate the execution of this parametric investigation, a depth of excavation measuring 4.5 m was selected. This specific dimension is commonly utilized in projects that incorporate repurposed water tanks. The soft clay thickness was consistently recorded as 10 m, which is indicative of the central region of Bangkok [25].

Fig. 6
figure 6

Layout and section of retaining wall: a) SC-IIRow, b) SSC-IRow with bracing, and c) sheet pile

Full size image
Table 3 Parameter of strut (Anchor) for finite element analysis
Full size table
Fig. 7
figure 7

FE mesh: a) SC-IIRow wall, b) SSC-IRow Wall with bracing, and c) Sheet pile wall

Full size image

4 Finite element model and its validation

The primary and foremost consideration in the design of retaining walls is the assurance of stability, as an excavation that lacks stability possesses the capacity to lead to catastrophic failure. The evaluation of the stability of SC and SSC Walls can be carried out by employing the Factor of Safety (FS) approach. The stability of a retaining structure can be influenced by various factors, which may differ based on international or local standards. The current investigation utilized a factor of safety (FS) of 1.30, adhering to the minimum requirement established by the Department of Public Works and Town & Country Planning in Thailand. This requirement serves as a reference point for evaluating the stability of temporary structures.

The lateral movements depicted in Fig. 8 are being juxtaposed with the field measurements obtained at two distinct time intervals. The initial measurement was performed 28 days following the installation of the SC-IIIRow Wall and SSC-IRow Wall, whereas the subsequent measurement was taken 20 days post-installation. Furthermore, two additional measurements were conducted subsequent to the implementation of recycled water tanks. The initial measurement was carried out 41 days following the installation of the SC-IIIRow Wall, while the subsequent measurement took place 33 days after the installation of the SSC-IRow Wall. During both simulation periods, the maximum lateral displacement (δmax) was observed at a depth of 4.5 m. The horizontal displacements of the soil observed during the excavation and installation of recycled water tanks showed a strong agreement with the results obtained from finite element (FE) analysis for both the SC-IIRow Wall and SSC-IRow Wall configurations. Hence, the selected soil model exhibits the capacity to be employed for the purpose of conducting parametric investigations. The results obtained from the fixed effects estimation demonstrated that the factor of safety (FSs) for the SC-IIRow Wall and SSC-IRow Wall were determined to be 2.05 and 1.45, respectively. The SC-IIIRow Wall exhibited decreased lateral displacement in comparison to the SCC-IRow Wall due to an increase in factor of safety (FS). The SC-IIIRow Wall exhibited a maximum lateral displacement of approximately 24 mm after a duration of 41 days, while the SSC-IRow Wall demonstrated a maximum lateral displacement of approximately 34 mm within a period of 33 days.

Fig. 8
figure 8

Soil movement a) SC-IIIRow Wall and b) SSC-1Row Wall

Full size image

4.1 Results of parametric study

Based on the findings of Jamsawang et al. (2015) [9], the research established a clear correlation between the maximum lateral displacement, denoted as δmax, of a SC wall and the factor of safety (FS). The relationship between the maximum displacement (δmax) and the factor of safety (FS) across different classifications of SC and SSC walls is illustrated in Fig. 9. The observed discrepancy in the relationship was clearly discernible and contingent upon the arrangement of the retaining structures. The SC-IIRow Wall exhibited a greater maximum displacement (δmax) than the SC-IIIRow Wall when subjected to identical FS. This empirical evidence indicates that augmenting the rigidity of the retaining wall leads to a reduction in lateral displacement while maintaining the same level of safety. The findings of this study align with previous research that has investigated the improvement of ground conditions in the soft soil layer [37,38,39]. The SSC-IRow Wall exhibited the smallest maximum deflection (δmax) while maintaining the same factor of safety. This implies that the reduction in the δmax of the retaining wall can be achieved by either augmenting the quantity of SC rows or by enhancing the rigidity of the SC by incorporating rigid piles. Furthermore, the integration of bracing into the SSC-IRow Wall presents the possibility of significantly reducing the δmax without compromising the factor of safety.

Fig. 9
figure 9

Correlation between lateral maximum lateral movement and factor of safety of various walls

Full size image

It is important to acknowledge that enhancing the factor of safety of the various wall types examined can also be achieved through an increase in the pile length (L). Figure 10 illustrates the relationship between the maximum displacement (δmax) and the factor of safety (FS) for different wall types, each characterized by varying pile lengths ranging from 5 to 13 m (L = 5 to 13 m). It was observed that the value of FS exhibited an increase, while the value of δmax demonstrated a decrease as the parameter L was increased. This trend persisted until the critical value of L was reached, at which point the maximum FS and minimum δmax were attained. The stability of the SC-IIIRow Wall was observed to be relatively consistent when the length (L) of the wall exceeded 12 m, as indicated by the factor of safety (FS) and maximum displacement (δmax). Nevertheless, when the L values were less than 12 m (particularly, at the critical L = 12 m), the factor of safety gradually decreased while the maximum displacement, δmax, increased. The critical L value of the SC-IIRow Wall was found to be less than 10 m, particularly with respect to the stiff clay layer. Although the factor of safety (FS) of the SC-IIRow Wall was determined to be lower in comparison to that of the SC-IIIRow Wall, it is important to highlight that the FS of the SC-IIRow Wall at a length of 7 m surpassed the minimum threshold of FS > 1.3. This implies that the temporary support of a narrow excavation in soft Bangkok clay can be achieved by utilizing a temporary retaining wall composed of two rows of SC with a length (L) of 7 m, as assessed in terms of the factor of safety (FS). However, it was observed that the maximum displacement, denoted as δmax, of the aforementioned structure exhibited a slightly higher value when compared to that of the SC-IIIRow Wall.

Fig. 10
figure 10

The relationship between factor of safety and lateral movement varied with lengths of piles

Full size image

In a similar manner, the factor of safety (FS) and maximum deflection (δmax) of both SSC Walls, with and without bracing, demonstrated an upward and downward trend, respectively, with the increase in length (L). The incorporation of a bracing system led to a significant enhancement in the factor of safety (FS) and a reduction in the maximum displacement (δmax) of the SSC Wall, as depicted in Fig. 10, in comparison to the SSC Wall lacking any form of bracing. The introduction of the bracing system resulted in a 50% reduction in the number of steel piles required for the SSC Wall, as compared to the SSC Wall without bracing. In the case of SSC-IRow, the factor of safety (FS) for the single-row wall without bracing exceeded 1.3 when the length (L) exceeded 7 m, which is consistent with the observed behavior of the SC-IIRow Wall. Nevertheless, when a bracing system was implemented, a mere length of 5 m proved to be adequate in attaining a factor of safety (FS) exceeding 1.3 for the SSC-IRow Wall.

Figure 11 illustrates the correlation between the factor of safety (FS) and the construction cost in relation to different lengths (L) of various retaining wall configurations. The determination of construction expenditures was obtained through the evaluation of material and labor expenses in Thailand in the year 2020. The recorded cost of implementing SC was 10.95 USD per meter, with the material cost representing 5.50 USD per meter and the labor cost representing 5.45 USD per meter. On the contrary, the cost of executing steel pipe was recorded at 34 USD/m, with the material cost and labor cost both totaling 17 USD/m. It is imperative to acknowledge that the exchange rate utilized in this analysis is 1 United States Dollar (USD) equivalent to 32 Thai Baht (THB). The projected expense of the construction was determined by evaluating the cost per meter of the retaining wall. In the case of the SC-IIRow Wall, there were four SC units present within each one-meter length. Conversely, the SC-IIIRow Wall exhibited six SC units within the same one-meter length. In the specific context of Thailand, the expenditure associated with the implementation of sheet pile installations is estimated to be around 14 USD per meter. This figure represents the combined costs of labor and rental expenses over a period of one month. Hence, the implementation of a sheet pile wall system may not be economically viable for a project of prolonged duration, primarily due to the increased costs associated with renting.

Fig. 11
figure 11

The relationship between factor of safety and cost of construction varied with lengths of piles

Full size image

The expenses associated with the construction of SC walls demonstrated a progressive rise in correlation with the length (L) of the walls, as illustrated in Fig. 11. The observed phenomenon can be ascribed to the concurrent rise in the demand for construction materials and the duration of the construction process. After evaluating different alternatives, it was noted that the SSC-IRow Wall without bracing exhibited the highest cost, while the SC-IIRow Wall was identified as the most cost-effective option. Subsequently, the SSC-IRow Wall with bracing and the SC-IIIRow Wall were assessed in terms of their respective costs, revealing that the former incurred higher expenses compared to the latter. Upon conducting a comparison between the SC-IIRow Wall and the conventional sheet pile, it was observed that the construction cost of the SC-IIRow Wall was found to be lower for values of L that were less than 10 m. Furthermore, the incorporation of a bracing system in the construction of the SSC Wall has the potential to result in a decrease in associated construction costs, while also enhancing stability. The findings of the study indicate that the construction cost of the SC-IRow Wall with bracing was lower in comparison to the SC-IRow Wall without bracing for all values of L. Additionally, the FS demonstrated a significantly higher value. When comparing the SC-IIIRow Wall to the SSC-IRow Wall with bracing, it was observed that the latter demonstrated a greater factor of safety (FS) and a reduced construction cost. Based on the FS > 1.3 criterion, it was observed that the construction costs of the SC-IIRow Wall, with a length (L) ranging from 7 to 10 m, the SC-IIIRow Wall with a L of 7 m, and the SSC-IRow Wall with bracing and a L of 7 to 8 m, were comparable to those of the conventional sheet pile. The study determined that the factor of safety (FS) for the SSC-IRow Wall with bracing, which had a variable length (L) ranging from 7 to 8 m, fell within the range of 1.95 to 2.15. The factor of safety (FS) for the conventional sheet pile, which had a length of 12 m, was found to be 2.60 in comparison.

Figure 12 depicts the relationship between the duration of construction and the corresponding expenses for all walls under examination, which vary in length and are denoted as L. The temporal extent of the construction project was quantified in terms of the number of hours necessary to finish a one-meter segment of the retaining wall. Based on the data depicted in Fig. 12, it is evident that both the quantity of SC rows and the extent of the SC Wall exert a noteworthy influence on the duration of construction. In the context of comparing walls with equal lengths, it was observed that the construction duration of the SC-IIIRow Wall was significantly greater in comparison to that of the SC-IIRow Wall. The SSC-IRow Wall, which was built without any form of bracing, demonstrated the most efficient construction time, despite incurring the highest construction expenses. The SC-IIRow Wall demonstrated cost-effectiveness and accelerated construction, similar to the SC-IRow Wall, without requiring supplementary bracing. Upon comparing the SC-IIRow Wall with the SC-IIIRow Wall, it was noted that the correlation between construction time and L exhibited a greater inclination for the SC-IIIRow Wall. This phenomenon can be ascribed to the influence exerted by the quantity of SC rows on the duration of the construction process. The study revealed that the construction time for SSC-IRow with bracing was comparatively longer than that of SSC-IRow without bracing across a range of L values from 5 to 13 m. However, it was noted that the rate of increase in construction time with respect to L was less pronounced in the former case. Hence, the discrepancy in the construction timeframe was mitigated to a lesser extent with the elongation of the length denoted as L. The duration of construction for the SSC-IRow with bracing was observed to be the most extensive for L < 7 m, with the SC-IIIRow ranking second in terms of construction time. However, it was observed that when the length (L) exceeded 7 m, the SC-IIIRow construction method demonstrated the longest duration for completion, with the SSC-IRow method with bracing following suit.

Fig. 12
figure 12

The relationship between time and cost of construction varied with lengths of piles

Full size image

The study centered on the comparative analysis of the construction duration, expenses, and factor of safety (FS) associated with various types of SC Walls, as illustrated in Figs. 11 and 12. The study determined that the factor of safety (FS) for the SC-IIIRow Wall and SSC-IRow with bracing was observed to be higher than 1.3 when the length of the wall (L) exceeded 5 m. In a similar vein, it was observed that the SC-IIRow Wall and SSC-IRow, when not equipped with bracing, demonstrated a factor of safety that surpassed 1.3 for L values that exceeded 7 m. In the comparison between SC-IIRow and SC-IIIRow, it was noted that SC-IIRow exhibited lower construction costs and time requirements when the length (L) exceeded 7 m, provided that the factor of safety (FS) remained above 1.3. A comparative analysis was conducted to examine the construction time of SSC-IRow with and without bracing, specifically focusing on cases where the factor of safety (FS) exceeded 1.3. The study revealed that the construction time of SSC-IRow with bracing was found to be longer than that of SSC-IRow without bracing when the length (L) exceeded 7 m. Additionally, it was observed that the relationship between construction time and L exhibited a more pronounced decrease for SSC-IRow with bracing in instances where L was exceptionally long. In contrast, the expense associated with the construction of the SSC-IRow Wall without bracing was significantly higher compared to its construction with bracing at all measured distances.

Similar values of the factor of safety (FS) were observed for both the SC-IIRow Wall and the SSC-IRow Wall without bracing when the length (L) exceeded 7 m. Nevertheless, the duration of construction and the rate at which the construction time increased in relation to L were greater for the SC-IIRow Wall. The construction expenses for the SSC-IRow Wall, when not utilizing bracing, demonstrated a significant rise in comparison to the SC-IIRow Wall. This increase can be primarily attributed to the elevated costs of materials involved in the construction process. The study found that the factor of safety (FS) for SC-IIRow remained consistent for lengths exceeding 10 m. In contrast, the FS for the SSC-IRow Wall without bracing displayed a linear growth. When considering the FS > 1.3 criterion, it is evident that SC-IIRow demonstrated a relative superiority compared to SSC-IRow Wall without bracing for critical L values less than 10 m, particularly with regards to the cost of construction. Nevertheless, it is important to acknowledge that SC-IIRow also exhibited a drawback in relation to the duration required for construction. The stability of these two types of walls can be improved by increasing their stiffness through the incorporation of either a bracing system or a higher quantity of SC rows.

During the comparative analysis of the SC-IIIRow Wall and the SSC-IRow Wall with bracing, it was noted that the construction cost of the SSC-IRow Wall with bracing was marginally lower than that of the SC-IIIRow Wall. However, the factor of safety (FS) of the SSC-IRow Wall with bracing demonstrated a considerably higher magnitude compared to that of the SC-IIIRow Wall for all values of L. The suggestion is that the SSC-IRow Wall with bracing presents certain benefits in comparison to the SC-IIIRow, specifically in terms of stability and construction duration. In this specific application, it is advisable to utilize a length of L = 5 m for both variants of the retaining wall.

In conducting a comparative analysis of the SC-IIIRow Wall, the SSC-IRow Wall with bracing spanning a length of 5 m, and the SC-IIRow Wall spanning a length of 7 m (all designed with a factor of safety exceeding 1.3), it was noted that the SC-IIRow Wall exhibited the lowest construction cost, amounting to 306.6 USD. Subsequently, it can be observed that the SSC-IRow Wall with bracing incurred a construction expenditure amounting to 361.4 USD, whereas the SC-IIIRow Wall exhibited the highest cost of 394.2 USD. The SC-IIRow exhibited the most efficient construction process, with a construction rate of 1.03 h per meter. The subsequent model, known as the SC-IIIRow, exhibited a construction time of 1.54 h per meter. Additionally, the SSC-IRow, which incorporated bracing, demonstrated a construction time of 1.56 h per meter. To provide further elucidation, it can be asserted that the SC-IIRow, positioned at a distance of 7 m, demonstrated enhanced efficacy in terms of both time and cost, based on the FS > 1.3 criterion, in comparison to alternative variations of SC and SSC walls.

This study investigates the expenses and time required for construction projects during the construction phase by analyzing various case studies. However, it is crucial to undertake further research in the future in order to comprehensively examine the long-term maintenance, life-cycle assessments, and uncertainties linked to the construction process.

5 Suggested retaining wall system

The utilization of a standardized 12-m long steel sheet pile accompanied by a bracing system is a widely employed technique within the construction industry. This methodology functions as a transitory retaining system utilized in excavation procedures. During the investigation of the factor of safety (FS) for various walls analyzed at a length of 12 m, it was noted that the sheet pile wall with bracing demonstrated a greater FS in comparison to the SC-IIIRow Wall, SC-IIRow Wall, and SSC-IRow Wall without bracing. The factor of safety (FS) for the sheet pile wall with bracing was found to be lower compared to the SSC-IRow Wall with bracing, as illustrated in Fig. 11. After evaluating multiple alternatives, it was observed that the sheet pile wall, measuring 12 m in length, demonstrated the most cost-effective construction expense. Following that, the SC-IIRow Wall, SSC-IRow Wall with bracing, SC-IIIRow Wall, and SSC-IRow Wall without bracing, in that order, exhibited similar costs. The duration of construction for the sheet pile wall was similar to that of the SC-IIRow Wall, but it was longer than the construction time for the SSC-IRow Wall without bracing, and shorter than the SSC-IRow Wall with bracing and SC-IIIRow Wall. The cost advantage of sheet pile in terms of construction expenses becomes apparent when considering a standard length of L = 12 m. Furthermore, it exhibits similar advantages in relation to the factor of safety (FS) and duration of construction in comparison to alternative forms of walls.

The limitation imposed by the confined construction area in Bangkok clay pertains to the utilization of a sheet pile wall system that adheres to a standard length of 12 m and requires appropriate bracing. The wall with bracing in the SSC is considered unsuitable for the current condition. The SC Wall is presented as a feasible alternative for shallow and narrow excavations in the soft clay of Bangkok. The system is applicable in both unconfined and confined spaces, providing versatility in terms of stability, lateral displacement, construction expenses, and duration. Furthermore, the SC Wall has the capability to be customized to accommodate various desired pile lengths. The length and number of SC rows were observed to have a substantial impact on the stability, construction cost, and construction time of SC walls.

The research findings indicate that the SC-IIRow Wall, which had a length of 7 m (equivalent to approximately 1.5 times the depth of excavation), demonstrated superior cost and time efficiency compared to the SSC Wall and sheet pile. The conclusion was derived from the application of the criterion of FS > 1.3, as depicted in Figs. 11 and 12. Although SC-IIIRow, with a height of 5 m, demonstrated a factor of safety exceeding 1.3, making it a feasible option for a temporary retaining wall, it is important to acknowledge that its construction costs and duration were slightly higher in comparison to SC-IIRow, which had a height of 7 m. Nevertheless, the augmentation of SC rows has the potential to enhance the structural stiffness of the wall, resulting in a reduction in the maximum displacement (δmax) as illustrated in Fig. 10. According to the criterion of FS > 1.3, it is recommended to employ the SC-IIRow Wall with a length of 7 m for the recycled water tank project located in the soft clay. The soft clay layer is estimated to have a thickness of approximately 10 m, and it is anticipated that the construction process will be completed in less than one month.

For the constrained configuration of retaining wall (the thick SC Wall is not allowed), the SSC-IRow Wall without bracing at L = 7 m is the best choice in terms of both cost and time for a confined construction site that the immobility of the bracing system is not possible. For the unconfined construction site, the SSC-IRow Wall with bracing at L = 5 m was more economical than the SSC-IRow Wall without bracing at 7 m but took almost twice longer construction time. However, the bracing system might obstruct the construction of recycled water tanks.

When soil excavation is constructed in a soft ground stratum, suitable permanent and/or temporary retaining walls are required to improve stability and prevent soil collapse due to large settlements of the soft soil. Numerous techniques for ground improvement have been developed with the aim of augmenting the stability of soft soil in the context of deep excavation. Among these methods, the deep cement mixing (DCM) technique utilizing soil cement columns has gained widespread practical application on a global scale. For shallow to medium deep excavation, a stiffened soil cement column (SSC) pile can be used to enhance the flexural stiffness against the large lateral deformation instead of the increase of the row number of SCC piles. In this study, the advantages and disadvantageous of selecting different types of temporary retaining walls including SC, SSC, and sheet pile were investigated and reported in terms of quality control (wall stability and lateral movement) and construction project management control (construction cost and time), which are useful for geotechnical engineers who work in similar site condition.

The procedures for the selection and design of the SC Wall and SSC-1Row Wall for recycled water tank in soft Bangkok clay are suggested as follows:

  1. (1)

    Undertake an on-site soil investigation and conduct geotechnical laboratory testing in order to obtain the soil profile and relevant parameters.

  2. (2)

    Select the retaining wall system based on construction constraints:

    • (2.1) For flexible configuration of excavation, the SC Wall is recommended due to its cost- and time-effectiveness for both confined and unconfined sites.

    • (2.2) For the limited thickness of the retaining structure:

      1. (2.2.1)

        With the immobility of bracing, the SSC-IRow Wall without bracing is recommended.

      2. (2.2.2)

        With the mobility of bracing, both SSC-IRow Wall with bracing and without bracing can be adapted. The SSC-IRow Wall with bracing is more cost advantageous but less time disadvantageous.

  3. (3)

    From the target excavation depth (dependent on the dimension of the recycled water tank) and soil profile, calculate the dimension of either SC or SSC according to step (2) to meet the minimum FS and lateral movement.

  4. (4)

    Calculate the construction cost and time of SC Wall/SSC-1Row Wall at various dimensions.

  5. (5)

    Select the optimized dimension of SC Wall/SSC-1Row Wall by a trade-off between time and cost using similar relationships presented in Figs. 10, 11, and 12.

Overall, the adaptability of SC and SSC walls for use in diverse ground conditions allows for optimized resource usage and minimal environmental impact. For a particular soft soil condition such as the Bangkok area, the optimization method for selecting the SC and SSC wall can contribute to resource-efficient management as they require less energy-intensive materials and installation processes.

In addition to the suggested retaining wall system, to make the construction of SC and SSC more sustainable in terms of reducing environmental impacts, conserving resources, and promoting long-term performance, the other recommendations can be considered for future studies as follows:

  • It is recommended to investigate the application of alternative cementitious binders that exhibit diminished environmental impacts, such as fly ash, ground granulated blast furnace slag, or rice husk ash. These materials have the potential to serve as partial substitutes for cement, thereby contributing to the mitigation of environmental impacts associated with carbon emissions and resource consumption typically associated with cement production.

  • Utilize energy-efficient construction equipment and machinery to reduce fuel consumption and greenhouse gas emissions during the SC and SSC construction process.

  • Conducting a life-cycle assessment (LCA) is essential for assessing the environmental consequences associated with various methods and materials employed in the construction of SC and SSC.

6 Conclusions

The present research study aimed to conduct a comparative analysis between soil–cement column (SC) and stiffened soil–cement column (SSC) walls, in relation to the conventional sheet pile wall system. The evaluation centered on three primary dimensions: execution time, cost, and stability, encompassing factors such as the factor of safety and lateral movement. The stability analysis was conducted by employing finite element (FE) simulation through the utilization of PLAXIS 2D software.

A case study was undertaken to investigate the implementation of a narrow and deep excavation within the context of soft clay soil conditions in Bangkok. The study focused on the utilization of SC and SSC walls as temporary retaining structures. The finite element (FE) model was calibrated using field measurements. The calibrated finite element model was subsequently employed to analyze the stability of the investigated SC and SSC walls at various lengths, and compared to that of a conventional sheet pile wall with a length of 12 m. An examination and evaluation of the construction expenses and duration for each type of retaining wall system were also undertaken. The findings were evaluated in order to assess the relative strengths and weaknesses of the SC, SSC, and sheet pile wall in terms of stability, construction cost, and time across different site conditions.

The project derives advantages from the uncomplicated installation procedure and cost efficiency of sheet pile walls in an unconstrained construction site, especially when compared to alternative forms of SC and SSC walls that employ piles of standard length (L = 12 m). However, the extended construction project is encumbered by the unfavorable rental expenditures of sheet piles throughout the construction period and the subsequent expenses linked to their removal upon the completion of the project. In addition, the implementation of a bracing system is crucial in reducing the risk of failure caused by substantial lateral displacement due to the low stiffness of the sheet pile. Implementing this precautionary measure is of utmost importance in order to mitigate any potential negative impacts on nearby infrastructure. The calculation of the optimal length of sheet piles and the evaluation of the need for a bracing system are critical considerations when constructing a sheet pile wall within a constrained construction site.

The achievement of stability requirements, such as factor of safety and lateral movement, for SC and SSC Walls can be attained through the simultaneous minimization of construction cost and time. The utilization of these walls presents several benefits in comparison to traditional sheet pile systems for excavations that are narrow and shallow to medium-deep, such as the project involving the construction of a recycled water tank. From an environmental perspective, it can reduce the emissions associated with less material production, equipment mobility and transportation, and construction phases. The results of this study suggest that the SC-IIRow Wall, measuring 7 m in length, exhibited enhanced effectiveness in terms of both time and cost in comparison to the SSC Wall with and without bracing, as well as the sheet pile wall system. The phenomenon described was noted in both restricted and unrestricted construction sites, with particular emphasis on its correlation with the FS > 1.3 criterion.

When the thick SCC Wall is not allowed, the application of an SSCC Wall is the best alternative. The SSCC-1Row Wall without bracing is recommended for the confined construction site. However, for the unconfined construction site, where the bracing system can be mobilized, both SSCC-1Row Wall without bracing and without bracing can be adapted. The SSCC-1Row Wall with bracing is more cost advantageous but less time disadvantageous. The stepwise procedure for selecting and designing the SCC and SSCC Wall systems was suggested based on the critical analysis of the study results.

The results of this research make a valuable contribution to the advancement of the selection and design procedures for a temporary retaining structure that is appropriate for shallow to medium deep excavation in the soft Bangkok clay. The knowledge gained from this research can be efficiently applied in similar excavation endeavors carried out in other soft soils. Utilizing FE simulations provides an opportunity to employ smart algorithms to predict wall behavior under various conditions. The results obtained from this research can be used for machine learning or artificial intelligence (AI) tools to further enhance the soil models and to accurately predict the behavior of the retaining walls. By integrating modern sensors and FE simulations, there is an opportunity to develop a feedback loop. Real-world data can refine FE models, while FE models can guide the placement and types of sensors used. Future research should investigate the integration of AI tools with real-time data from embedded sensors in retaining walls to continuously refine FE models and predictive algorithms. This would lead to an iterative improvement cycle, combining theatrical models with real-world data for better outcomes.