Impact of ground conditions and excavation depth on selection of excavation support system

Abstract Inappropriate selection of an excavation support system (ESS) may lead to delays, cost overruns and quality and safety problems. Soil conditions, excavation depth and groundwater level are important factors that significantly affect the selection of ESS. However, a guideline that determines the effects of these three factors and accounts for their variation in the selection of ESS is not available. To evaluate such effects, a questionnaire was sent to excavation support experts worldwide. The results of the conducted survey are presented and discussed in the current paper to determine the favourable soil conditions and maximum excavation depths for ESS in dry and wet soils. The focus is on the soldier pile, secant pile, diaphragm and jet grout walls because of their common use worldwide. Clear guidelines that can help in the selection of an appropriate ESS for given site conditions could be established. The obtained findings are compared with the conditions and adopted ESS of reported case histories.


Introduction
Urban construction projects include deep vertical excavation that usually needs to be supported.Many different systems are commonly adopted in the current practice to support the sides of the excavation.Examples of such excavation support systems (ESSs) include sheet pile walls, soldier pile walls, bored pile walls, micropile walls, diaphragm walls, jet grout walls, soil mixing walls and caissons.Each system has its favourable conditions, advantages and limitations.Improper selection of ESS may lead to serious schedule delays (Moungnos and Charoenngam 2003), cost overruns (Shin et al. 2009;Choi and Lee 2010;Mary and Sankar 2016), major quality problems (Shin et al. 2009) and safety problems (Choi and Lee 2010).
The selection of appropriate ESS for a given project and site conditions is essentially affected by many factors such as ground conditions, groundwater level, excavation depth, cost, construction duration, constructability, construction-induced noise and vibration, distance and condition of nearby structures and client requirements (e.g.Mojahed and French 2008;Pan 2009;Ali 2011;Elnabolsy 2015;Hassan and El-Kelesh 2021).The large number of involved factors and the necessity to assess the relative importance of each factor has made the selection process complicated.In the current practice, an ESS is commonly decided for a given job based on the experience of practitioners.This has proved ineffective (Ali 2011;Lee 2012;Dong 2014;Hassan and El-Kelesh 2021).Therefore, several models were developed to help decision-makers in this regard (e.g.Pan 2009;Shin et al. 2009;Choi and Lee 2010;Lee 2012;Mary and Sankar 2016;Mahdi et al. 2021).However, these models do not agree on the significantly controlling factors (e.g.Yang 2004;Pan 2009;Choi and Lee 2010;Mary and Sankar 2016).Nonetheless, these models as well as other studies (e.g.Geotechnical Engineering Office -Hong Kong 1993; Ratay 1996;Mojahed and French 2008;Elnabolsy 2015) consider soil conditions, groundwater level and excavation depth among the most important factors that significantly control the selection process.
It should be mentioned that a variation in any of these three factors affects other important factors.For instance, a variation in the soil condition or excavation depth results in significant changes in the design and thus the construction duration and cost of a given ESS.This variation may also affect other factors such as the displacement of surrounding soils and the condition of nearby structures.It is also worth mentioning that a change in any of the three factors may limit or discard the use of an ESS.For example, the use of soldier pile walls has certain depth limitations (e.g.Union Pacific Railroad 2001; Choi and Lee 2010;Ali 2011;Mary and Sankar 2016).The presence of groundwater is not preferred for the installation of soldier pile walls because of their high permeability (e.g.Union Pacific Railroad 2001; Mojahed and French 2008;Ali 2011;Mary and Sankar 2016).On the other hand, the effectiveness of adopting a given ESS is sensitive to variations of the three factors.For example, dense sand is more favourable for the installation of the bored piles of secant pile walls than loose sand (e.g.Union Pacific Railroad 2001; Mary and Sankar 2016).The installation of bored piles in dry soils is easier and more economical than that in wet soils (e.g.Ergun 2008;Clayton et al. 2014).The overlapping of secant piles can be well controlled to certain depths, while it may be questionable for deeper installations (e.g.Macnab 2002;Ali 2011;Clayton et al. 2014).Therefore, effective selection of ESS requires a better understanding of the effects of variations of the above three factors on the selection process and the favourable soil and groundwater conditions and excavation depths for each ESS.Unfortunately, little research effort has been made in this regard.This paper presents the results of a part of ongoing research work that aims at developing a tool for the selection of appropriate ESS for a given project and site conditions.The objective of the paper is to evaluate the favourable conditions and limits of soil, groundwater and excavation depth for the selection of ESS.The focus is on the soldier pile, secant pile, diaphragm and jet grout walls because of their common use worldwide.To approach this objective, a questionnaire was sent to excavation support experts worldwide.The paper starts with a brief description of the four ESSs.This is followed by a review of the relevant literature.Then, the adopted methodology is summarized and the obtained responses from experts are presented and discussed.The obtained findings are compared with actual conditions and the adopted ESSs of reported case histories.Finally, implications of the findings for actual construction are presented.

Excavation support systems
Sloping back the sides of an excavation to an appropriate angle is the cheapest, easiest and most common way to support an excavation section.If a site is located in an urban area, where no sufficient free space to slope back the soil is available, an ESS is usually required.ESS should provide for safe sides of deep excavation (e.g.Macnab 2002;Ali 2011;Reis and Lucko 2016) and ensure that excavation-induced displacements of soil do not significantly affect the neighbouring structures or utilities (e.g.Ali 2011;Zhu and Ibrahim 2017).Excavation sides can be supported using different techniques such as sheet pile walls (e.g.Yang 2004;Jardaneh 2006;Ali 2011 ), soldier pile walls (e.g.Macnab 2002;Pan 2009;Ali 2011), secant pile walls, contiguous pile walls (e.g.Ali 2011;Mary and Sankar 2016), diaphragm walls (e.g.Yang et al. 2003;Ali 2011;Clayton et al. 2014), jet grout walls (e.g.Ho and Tan 2012;Clayton et al. 2014), soil mixing walls (e.g.Jian et al. 2009;Denies and Huybrechts 2012;Chen et al. 2013), micropile walls (e.g.Macnab 2002;Chaudhari et al. 2015), caissons (e.g.Ergun 2008;Elnabolsy 2015) and freezing walls (e.g.Braun et al. 1979;Jessberger 1980).Soldier pile, secant pile, diaphragm and jet grout walls are commonly used worldwide.Therefore, their construction procedure, advantages and limitations are briefly described in the remaining part of this section.

Soldier pile walls
Soldier piles are normally steel vertical elements that are used to support the excavation sides.Once the piles are installed the spaces between piles (1.80-3.00m) are filled with timber lagging.The installation procedure can be summarized as follows: strike soldier piles into the soil, place laggings as the excavation proceeds, backfill the voids between soldier piles and laggings and install horizontal struts at proper places as the excavation proceeds.According to their construction method, soldier piles can be divided into four main types: driven, drilled and concreted, churn drilled and wet set (e.g.Macnab 2002;MDT Geotechnical Manual 2008;Clayton et al. 2014).
The main advantage of soldier pile walls is that the main support structure is installed before starting excavation.Applicability to most of the soil types, low initial cost, easy and fast handling and installation, minimal excavation, backfill quantities and convenience in making adjustments in the field are other advantages (e.g.Macnab 2002;MDT Geotechnical Manual 2008;Clayton et al. 2014).On the other hand, the use of soldier pile walls is primarily limited to temporary construction.They cannot be used in case of high water

Secant pile walls
With relatively low noise and vibration, bored piles can be installed in different forms in almost all types of soil.Bored piles can be included in the structural design of building foundations to support high lateral and vertical pressures (e.g.Macnab 2002;Ergun 2008;Clayton et al. 2014).Clayton et al. (2014) classified bored piles according to the spaces between them into three categories: intermittent, contiguous and secant piles.The installation procedure of secant pile walls consists of the following in order: construction of guide wall and installation of casings, augering of primary boreholes followed by concreting them, augering of the secondary boreholes and installation of their steel cages, concreting of the secondary borehole and repetition of the process till the completion of the target secant pile wall.
Secant pile walls can be constructed before the excavation works and below the groundwater (e.g.Union Pacific Railroad 2001; Ali 2011; Elnabolsy 2015).They have adjustable wall stiffness and can be installed in difficult ground (e.g.MDT Geotechnical Manual 2008;Ali 2011;Quintanar 2014).Inducing low vibration and noise are other advantages.On the other hand, secant pile walls need special requirements for the equipment and construction team.They are more expensive than other systems.In case of deep excavation sections, it may be difficult to maintain the verticality tolerance (e.g.Ali 2011;Elnabolsy 2015).Attaining complete water tightness at the joints needs careful control of the installation works (e.g.Ali 2011;Quintanar 2014).

Diaphragm walls
Diaphragm walls are cast-in-place concrete sections.With the support of nailing techniques, bracing, or struts these walls could be installed at depth and resist high lateral pressures.Recent developments have enabled the use of prefabricated sections for diaphragm walls (e.g.Ergun 2008;Clayton et al. 2014).This has resulted in improved quality and the use of thinner and lighter panels (Ergun 2008).The installation of diaphragm walls starts by forming guide walls.Then, primary panels followed by secondary ones are installed.Bentonite suspension is used to support the sides of the excavated panels.For each panel, reinforcement and stop end tubes are lowered into the excavated panels.Then, concrete is tremied.
Diaphragm walls can be formed in all types of soil (e.g.Bobet 2002;MDT Geotechnical Manual 2008;Ali 2011).They provide watertight walls and structural stiffness and economic solutions in cases where temporary and permanent supports are required.Diaphragm walls facilitate excavation below groundwater without the need for dewatering, can be easily adapted to both anchors and internal structural bracing systems, and can be used with difficult ground conditions and obstructions (e.g.Ali 2011;Quintanar 2014).On the other side, they require the use of special heavy construction equipment and skilled personnel and cause a large volume of generated spoils (e.g.MDT Geotechnical Manual 2008;Ali 2011;Elnabolsy 2015).

Jet grout walls
Jet grouting involves the erosion of soil by means of a highenergy jet of fluid.The eroding fluid can be grout, water or grout shrouded with compressed air.The eroded soil is mixed and partially replaced with grout to form a solidified element known as soilcrete.The procedure consists of drilling a borehole to the design depth, introducing the jet grouting string to the bottom of the borehole, and then eroding the soil.During erosion, the string is rotated and withdrawn at predetermined speeds to form a cylindrical element.The elements are formed at predetermined spacing and overlapping ratio to produce soilcrete walls (e.g.Burke 2004;Clayton et al. 2014).
Jet grout walls can be formed before excavation and below groundwater and in almost all types of soils (e.g.Union Pacific Railroad 2001; Burke 2004;Njock et al. 2018).They can also be formed in confined spaces with major environmental control, minimal disturbance to adjacent structures and low noise and vibration (e.g.Union Pacific Railroad 2001; Burke 2004).On the other hand, they cannot be formed in soils that contain boulders; the existence of soil obstructions can block the lateral spread of the grout jet (e.g.Burke 2004;Njock et al. 2018).They generate large volumes of spoil (e.g.Burke 2004;Njock et al. 2018).

Significance of ground conditions and excavation depth
The selection of appropriate ESS is the key to optimizing the design and execution of excavation for given site conditions.Therefore, several studies were conducted to determine the factors that control the selection and assess their relative importance.An extensive review of the literature indicates that the main factors that affect the selection of ESS include the following: Soil condition, groundwater level and excavation depth (e.g.Ratay 1996;Yang 2004;Mojahed and French 2008;Pan 2009;Elnabolsy 2015;Hassan and El-Kelesh 2021).Condition of neighbouring buildings and their distance to the target excavation and displacement in the soil around the excavation (e.g.Moungnos and Charoenngam 2003;Mojahed and French 2008;Pan 2009;Ali 2011).Depth interval between the new excavation and foundation level of adjacent structures (e.g.Hutchinson et al. 1987;Hassan and El-Kelesh 2021).Construction cost and duration (e.g.; Ratay 1996;Smith et al. 1998;Yang 2004;Ali 2011;Lee 2012;Elnabolsy 2015 ).The layout of the construction site and available working space (Smith et al. 1998;Lee 2012;Mary and Sankar 2016;Mahdi et al. 2021).Constructability in the form of the availability of execution equipment, engineering expertise, trained labour and materials (e.g.Smith et al. 1998;Pan 2009;Lee 2012;Elnabolsy 2015).Noise and vibration during the execution process (e.g.Mojahed and French 2008;Elnabolsy 2015).Aesthetics and final shape of ESS (e.g.Smith et al. 1998;Hassan and El-Kelesh 2021).Client and municipality requirements and constraints (e.g.Ling et al. 2003;Elnabolsy 2015;Hassan and El-Kelesh 2021).
The reported studies on the influencing factors include questionnaire surveys (e.g.Chini and Genauer 1997;Pan 2009;Ali 2011; Hassan and El-Kelesh 2021), quantitative analyses (e.g.Geotechnical Engineering Office -Hong Kong 1993; Ratay 1996; Lee 2012; Elnabolsy 2015) and development of selection models (e.g.Smith et al. 1998;Yang et al. 2003;Shin et al. 2009;Mary and Sankar 2016).These studies do not agree on the relative importance of the considered factors.However, almost all of them considered the soil condition, groundwater level and excavation depth among the top of the most critical factors that affect the selection of ESS.As mentioned above, a variation in any of the three factors affects other important factors and may limit or discard the use of certain ESS.It may also affect the effectiveness of the adopted ESS.A guideline that determines the effect of variation of any of the three factors on the selection process or their favourable conditions for a given ESS is not available.Ali (2011) reported that soldier pile walls are preferred in dry sand soil for excavation depths up to 6.00 m.For the same soil and groundwater conditions, Mary and Sankar (2016) and Quintanar (2014) also limited their economical depth to 5.00 m.Bobet (2002) reported that the maximum depth is 5.00 m, while it can be increased to 20.00 m if anchors are used.However, Choi and Lee (2010) documented case histories in which soldier pile walls were used to depths of approximately 24.00 m.It is reported that soldier pile walls are not effective in water tightness (Union Pacific Railroad 2001;MDT Geotechnical Manual 2008).
It has been reported that secant pile walls can be used in different soil types and fit in diverse soil conditions as well as under the groundwater level (e.g.MDT Geotechnical Manual 2008;Clayton et al. 2014;Elnabolsy 2015).Mary and Sankar (2016) recommend the use of secant pile walls in cohesive soils.Bobet (2002) and Ali (2011) reported that the maximum achievable depth of secant pile walls is about 10.00 m without anchors and 24.00 m with anchors.Quintanar (2014) reported a depth of 20.00 m.Union Pacific Railroad (2001) generalized secant pile walls as good in groundwater control.However, Elnabolsy (2015) and MDT Geotechnical Manual (2008) limited their depth for effective water tightness to 25.00 m.MDT Geotechnical Manual (2008) recommended the use of diaphragm walls in all soil types at depths of 6.00-24.00m.However, Lee (2012) recommended soft clay, stiff clay and granular soils and excluded rock.Elnabolsy (2015) reported that diaphragm walls face challenges in very hard clay and very dense sands.Bobet (2002) mentioned that the depth should not exceed 25.00 m for diaphragm walls.Choi and Lee (2010) indicated that the maximum cost-effective depth is 40.00 m. Lee (2012) set the maximum depth to 50.00 m.Quintanar (2014) suggested a depth of 20.00-80.00m.Elnabolsy (2015) reported that the maximum achievable depth is 60.00 m.
Jet grout walls can be formed in almost all types of soil (Burke 2004;Wang et al. 2013;Njock et al. 2018).They are highly effective in groundwater control (e.g.Union Pacific Railroad 2001; Burke 2004;Njock et al. 2018).A major advantage of jet grout walls is that they can be formed in close vicinity of existing structures and around existing buried structures (e.g.Union Pacific Railroad 2001 ;Burke 2004).This has qualified the method for special support applications such as underpinning walls and cut-off walls (e.g.Burke 2004;Njock et al. 2018).However, little effort has been made to compare the applicability, effectiveness and favourable soil conditions and depth of jet grout walls with those of the other support systems.
Several efforts have been made to develop selection models.Yang et al. (2003) developed a selection model that takes into account the three factors.However, the model does not include the soldier pile, secant pile and jet grout walls in the selection.The selection models developed by Shin et al. (2009) and Choi and Lee (2010) do not include the secant pile and jet grout walls.On the basis of available databases, other selection models were developed (e.g.Yau and Yang 1998;Yang 2004;Shin et al. 2009).However, these models do not consider important factors such as the construction, site and soil conditions in the selection.This has resulted in questionable results (e.g.Hutchinson et al. 1987;Geotechnical Engineering Office -Hong Kong 1993;Chini and Genauer 1997).

Study approach and methodology
The decision of appropriate ESS for given site and project conditions is not a straightforward process.The difficulty stems from the many different factors that are involved in the process.A decision guideline that considers the significant influencing factors and their relative importance is not available.The reported factors and data cannot be used to rigorously analyze or evaluate the relative importance of the factors owing to the following: (1) the factors involved in the selection process are many and thus render an analysis complicated, (2) the relative importance of influencing factors that is reported in available studies is not in agreement (e.g.Yang 2004;Shin et al. 2009;Choi and Lee 2010;Mary and Sankar 2016) and (3) details that are necessary for analysis are not sufficiently reported.Because of such limitations, the approach of the current investigation was to seek the opinions of ESS experts through exploratory sequential mixed methods of semi-structured interviews and questionnaire surveys.
In a recent study conducted by the authors (Hassan and El-Kelesh 2021), the literature was critically reviewed and all the factors that may affect the selection of ESS were collected.These factors were initially evaluated and filtered through semi-structured interviews with ESS experts representing different stakeholders (contracting companies, engineers and owners).The modified list of factors was then sent through a questionnaire to ESS experts.
The exploratory sequential mixed methods are useful in descriptive and exploratory studies (e.g.Onwuegbuzie et al. 2010;Fetters et al. 2013;Saunders et al. 2019).In addition, they can be used to investigate and explain relationships between different variables as long as the questions have the same interpretation to the respondent persons (e.g.Fetters et al. 2013;Saris and Gallhofer 2014;Robson and McCartan 2016).Based on the research objectives (Lines 69-78), an exploratory study is introduced to attain these objectives.To evaluate the usability of each ESS in different soil types, experts' opinions were collected and analyzed to set up a ranking of these ESSs in each soil type.Further, experts' experience is crucial to estimate the maximum achievable excavation depth of each system and to assess the effects of groundwater on it.The questionnaire targeted worldwide experts through personal contacts and LinkedIn media.Some of the questions were answered on a five-point Likert scale (1-5), where 1 represents 'not important' and 5 represents 'most important'.The other questions were answered in a short text form.The questionnaire was made up of three sections as shown in Online Appendix.The first section included questions on the particulars of the respondents: contact information, type of business and experience.In the second section, the respondents were asked to assess the relative importance of the factors.This has resulted in 18 factors as shown in Table 1.Details and discussions of these results are reported elsewhere (Hassan and El-Kelesh 2021).The relative weights of the factors are also shown in the table.In agreement with past studies, the results in Table 1 indicate that the soil condition, groundwater level and excavation depth represent the most significant and influencing factors.
The experts were asked in the third section to assess the impact of each factor on the performance of the four ESSs (soldier pile, secant pile, diaphragm and jet grout walls).To approach the objective of evaluating the effect of variation of the above three factors on the selection process and the favourable soil and groundwater conditions and excavation depth for the four support systems, the third section included detailed questions on the soil condition, groundwater and excavation depth.According to the classification of Moormann (2004), four soil types and conditions were considered in the questionnaire as follows (see Online Appendix The questionnaire reached 200 experts.Sixty-one responses were received, representing a response rate of 31%.According to Cochran's formula (Israel 1992;Hashim 2010), the researchers are 90% certain that the obtained responses represent the population opinion with a precision level of 10%.Cochran's formula is given by where n 0 ¼ sample size, z ¼ the abscissa of the normal curve that cuts off an area a at the tails (z ¼ 1.645 at a confidence level of 90%), e ¼ desired precision level, p ¼ 0.5 (estimated proportion of an attribute that is present in the population) and q ¼ 1Àp.The final sample size (n) is given by where n ¼ final sample size, n 0 ¼ sample size according to Cochran's formula and N ¼ population size.It is worth noting that 51 responses correspond to a confidence level of 95% and a precision level of 10%.A confidence level of 95% and a precision level of 10% would require additional 4 responses.The respondents consisted of 38 consultants (62%), 20 contractors (33%) and 3 owners (5%).They belong to both the public and private sectors with percentages of 87% and 13%, respectively.The experience of the respondents ranged from 1 to more than 16 years.It was 16 years and more for 24 respondents, ranged from 11 to 15 years for 12 respondents, ranged from 6 to 10 years for 19 respondents, and ranged from 1 to 5 years for 6 respondents.Cronbach's alpha coefficient (a) measures the reliability or internal consistency of statistical findings (Cronbach 1951).The coefficient has a value that ranges from 0 to 1; the higher the value a, the greater the internal consistency (Field 2005).It was calculated for this study using Two Factors ANOVA and found to be 0.71038 as shown in Table 2.According to Field (2005) and Sharma (2016), as shown in Table 3, a value of a ¼ 0.7 indicates acceptable internal consistency.The calculated a demonstrates the acceptable internal consistency of the obtained responses.The experts were also asked to indicate their familiarity and experience with each ESS on a fivepoint scale, where 1 means almost no experience and 5 means experience in many projects of ESS.The answers shown in Figure 1 indicate that the experts are familiar and have good experience with the four ESSs.Most of them have already executed many projects of ESS.

Soil condition
The experts were asked to assess the applicability of the four ESSs to the above-mentioned four soils (soft clay, stiff clay, granular soil and layered soil).The responses are shown in Figure 2 and range from inapplicable to highly applicable.The results in Figure 2 are replotted to show a relative assessment of the applicability for the soldier pile, secant pile, diaphragm and jet grout walls in Figures 3-6, respectively.The average value for each soil which is the summation of the experts' responses (scale 1-5) divided by 61 is shown in Figure 7 for the four ESSs.
The results in Figures 2-7 reveal useful insights on the overall favourable conditions of soil and thus the applicability of the four systems.It is seen for the soldier pile walls that stiff clay is the most favourable soil followed by granular soil and then layered soil.Soft clay is the worst for soldier piles.This is in agreement with the observations of Macnab (2002) and Ali (2011) who reported that soldier pile walls are preferred in stiff clay and granular soil.Generally speaking, driving soldier piles in stiff clay and granular soil is better controlled than in layered soil and soft clay.Driving in stiff clay is easier and more productive than in granular soil (e.g.Bobet 2002).Installation of the lagging of soldier pile walls is better controlled in stiff clay than in granular soil, especially in case of installation under the groundwater level (e.g.Bobet 2002).
For the secant pile walls, the figures show that granular soil is the most favourable soil followed by layered soil and stiff clay.Soft clay is the worst soil condition.In accordance with these results, Clayton et al. (2014) indicated that secant pile walls can be effectively installed in most different soils and that granular and layered soils are favourable.It is worth mentioning that boring for secant piles is essentially more productive in granular and layered soils than in clays (e.g.Jieh-Haur Chen et al. 2008;Sunjaya and Susilo 2020).Moreover, boring in stiff clay is better Table 3. Value of Cronbach's alpha and corresponding internal consistency of data (Field 2005;Sharma 2016).
As for the diaphragm walls, it is seen that the favourable soils are layered soil, granular soil, stiff clay and soft clay in order.Clayton et al. (2014) also reported that diaphragm walls can be an efficient solution in almost all types of soils and that granular and layered soils are favourable for their installation.Unlike the results of the other ESSs, those of the diaphragm walls in Figures 2, 5 and 7 show small differences for the four soil conditions.This may be explained by the installation procedure of diaphragm walls in which heavy equipment with hydraulically operated cutting discs is used in the excavation for the wall panels.The working mechanism and the cutting power of the machine may decrease the sensitivity to variation of soil conditions.
Figure 7 shows that granular soil is the most favourable for jet grout walls.It also shows almost identical results for the other three soil conditions.However, a careful examination of the results in Figures 2 and 6 reveals that granular soil is the most favourable soil condition followed by layered soil and then stiff clay.Soft clay is the worst among the other soils.This supports the observations of others (e.g.Burke 2004;Wang et al. 2013;Njock et al. 2018).Generally speaking, it is well known that granular soil is more erodible by jet grouting than cohesive soil (stiff clay and soft clay).This implies that jet grouting is more effective in granular soil than in cohesive soil.Jet grouting has a great advantage when used in layered soils.For a given target diameter of the soilcrete elements, the jet grouting parameters can be controlled to account for a variable condition of soil throughout the treatment depth interval.This provides for soilcrete elements of practically uniform diameter (e.g.Burke 2004;Wang et al. 2013).This most likely explains why the experts assessed layered soil as more favourable than cohesive soils for the jet grout walls.
The average values in Figure 7 provide a useful guideline on the applicability of ESSs.For soft clay, the average values are 3.89, 3.81, 3.00 and 2.94 for the secant pile, diaphragm, jet grout and soldier pile walls, respectively.This implies that secant pile and diaphragm walls are more applicable to soft clay than jet grout and soldier pile walls.The slightly higher applicability of the secant pile walls compared to that of the diaphragm walls is likely attributed to the differences in the geometry and volume of their singular elements.In other words, the secant pile walls and diaphragm walls consist of overlapped bored piles and adjacent panels, respectively.The bored piles are commonly circular in cross-section and smaller in volume, while the panels are rectangular in cross-section and larger in volume.This implies   relatively better stability of the pile borehole than that of the panel trench.This is most likely the issue that recommends secant pile walls for soft clay soils.The higher applicability of diaphragm walls than that of jet grout walls may be explained by the effectiveness of cutting and eroding the soil in the installation of diaphragm walls and jet grout walls, respectively.Compared to other soils, clay soils whether stiff or soft are less erodible by jet grouting.However, cutting soft clays is much easier in the installation of diaphragm walls.This most likely explains the significant difference between the average values for the diaphragm walls and jet grout walls.The low average value for the soldier pile walls is likely attributed to the inconvenience and difficult control that are experienced during the installation works.
For stiff clays, the average values are 4.02, 3.96, 3.56 and 3.04 for the diaphragm, secant pile, soldier pile and jet grout walls, respectively.This represents the order of applicability of the four systems in stiff clay soils.As for granular soils, the average values are 4.15, 4.07, 3.51 and 3.35 for the diaphragm, secant pile, jet grout and soldier pile walls, respectively.However, for layered soils, the average values are 4.24, 4.08, 3.26 and 3.06 for the diaphragm, secant pile, soldier pile and jet grout walls, respectively.These orders of applicability for the stiff clay, granular soil and layered soil are most likely attributed to the order of the corresponding productivity.

Depth of dry excavation
The experts were also asked to determine the maximum excavation depth that can be achieved by each ESS.The obtained values of maximum depth that largely disagree with the literature and those which are reported for special cases or conditions were excluded.An example of such excluded data is the depth of 250.00 m which was obtained for diaphragm walls; this was reported by a single respondent.The obtained responses determine maximum depths of 3.00-40.00,4.00-40.00,5.00-50.00and 1.00-40.00m with average values of 11.9, 17.88, 19.67 and 26.39 m for the soldier pile, secant pile, diaphragm and jet grout walls, respectively.
The obtained results are represented in Figure 8 for depth intervals of 10.00.Regression lines and their corresponding coefficients of determination (R 2 ) are also shown in the figure.The values of R 2 for the soldier pile, secant pile, diaphragm and jet grout walls are 0.997, 0.925, 0.980 and 0.681, respectively.These results indicate that soldier pile walls are more suitable for shallow excavations with an average value of the maximum depth of   11.90 m.This is generally in accordance with the literature (e.g.Macnab 2002;MDT Geotechnical Manual 2008, Ali 2011;Mary and Sankar 2016).For deeper excavation sections, secant pile walls or diaphragm walls are more suitable.Secant pile walls can be used for depths in the range of 4.00-40.00m with an average of 19.67 m.This exceeds the values reported by the MDT Geotechnical Manual (2008) which reports a maximum achievable depth for secant piles of 24.00 m.Diaphragm walls are suitable and highly feasible for deeper excavations.The results indicate depths in the range of 5.00-50.00m with an average value of 26.00 m.This is not in agreement with the findings of Clayton et al. (2014) who reported a maximum depth of 30.50 m for the diaphragm walls and the MDT Geotechnical Manual (2008) which reports a maximum depth of only 24.00 m.

Depth of wet excavation
The groundwater table is one of the most crucial factors that affect the applicability of ESS (e.g.Moungnos and Charoenngam 2003;Mojahed and French 2008;Elnabolsy 2015;Hassan and El-Kelesh 2021).For instance, the existence of groundwater above the excavation depth excludes the use of soldier piles because of their high permeability.On the other hand, diaphragm walls and secant pile walls are highly recommended in case of excavation below the water table because of their good water tightness.The term 'wet excavation depth' is used herein to designate the depth of excavation below the level of groundwater.
The experts were asked about the applicability and effectiveness of the four ESSs for wet excavation.The obtained responses are shown in Figures 9 and 10.It is seen that soldier pile walls are not recommended for wet excavation.However, diaphragm walls are clearly the most appropriate system.Secant pile walls and jet grout walls have intermediate applicability.The results of the question on the maximum depth that can be achieved in the wet excavation are shown in Figure 11 for depth intervals of 10.00.The obtained values that largely disagree with the literature and those which are reported for special cases or conditions were excluded.An example of the excluded data is the depth of 120.00 m which was obtained for the diaphragm walls and 45.0 m for the jet grout walls.The obtained responses determine depths of 0.00-25.00,4.00-40.00,5.00-50.00and 0.00-30.00m with average values of 8. 25, 20.11, 28.33 and 14.42 m for the soldier pile, secant pile, diaphragm and jet grout walls, respectively.Regression lines and their corresponding R 2 are also shown in Figure 11.The values of R 2 for the soldier pile, secant pile, diaphragm and jet grout walls are 1.000, 0.982, 0.674 and 1.000, respectively.Figure 12 shows a comparison between the results for dry and wet excavation.It is seen that the existence of groundwater decreases the maximum depth for soldier pile and jet grout walls.However, as intuitively expected it does not significantly affect the maximum depth for secant pile and diaphragm walls.

Comparison with case histories
In this section, the above findings are compared with conditions and the adopted ESSs for reported fourteen cases histories.The case histories were carefully selected to represent different site conditions.The cases are summarized in Table 4. Soldier pile, secant pile, diaphragm and jet grout walls were adopted in 4, 3, 5 and 2 cases.The excavation was made in different soils and the excavation depths range from 4.50 to 24.00 m.Different bracing methods were considered along with the support systems.
Soldier pile walls were used in Cases 1-4.The excavation was made in gravel, sandy silt, silty sand and stiff clay soils.These soils are classified as granular or stiff clay soils, which are the most favourable soils for soldier pile walls as concluded above.The excavation depths of the four cases range from 4.50 to 11.50 m.This is also in agreement with the above findings where the maximum depth for soldier pile walls ranges from 3.00 to 40.00 m with an average of 11.90 m.
In Cases 5-7, secant pile walls were used to support excavation in gravel and clay soils.This agrees with the above findings that the most favourable soils for secant pile walls are granular soils followed by stiff clay soils.The excavation depths in the three cases range from 7.20 to 19.30 m.This also agrees with the findings herein that the maximum depth ranges from 4.0 to 40.0 m with an average of 20.11 m for wet excavation.
Diaphragm walls were used in Cases 8-12.Excavations were made in clay and sand soils.The excavation depths range from 8.20 to 24.00 m.This is in good agreement with the above findings on the favourable soils which are stiff clay and granular soil.It also agrees with the concluded depth which   ranges from 5.00 to 50.00 m with an average of 28.33 m for wet excavation.
In Cases 13 and 14, jet grout walls were adopted to support excavations in soft clay soils.As mentioned above, jet grout walls can be installed in almost all types of soil including soft clay soils.The excavation depths of the two cases range from 6.50 to 18.00 m.This also agrees with the findings herein that the excavation depth for jet grout walls ranges from 1.00 to 40.00 with an average of 18.00 m.

Implications for actual construction
Figure 13 summarizes the main findings of the current study and provides guidelines that a decision-maker can use for the selection of appropriate ESS.The upper part summarizes the applicability of the four ESSs to different soils.For soft clays, the most favourable ESS is secant pile walls followed by diaphragm walls and jet grout walls.The least favourable ESS is the soldier pile walls.In the case of stiff clays and layered soils, the recommended order of ESS is the diaphragm, secant pile, soldier pile and finally jet grout walls.As for granular soils, the most favourable ESS is diaphragm walls followed by secant pile and jet grout walls.The least favourable ESS is soldier pile walls.
The lower part in Figure 13 shows the maximum achievable wet and dry depths for each ESS.All the four ESSs can be used to the depth of 11.90 m and after that soldier pile walls are not efficient anymore and should be excluded.After the depth of 17.88 m, jet grout walls should not be considered as a suitable ESS.Secant pile and diaphragm walls remain usable till 19.67 m and after that diaphragm walls are the most suitable ESS.
If the excavation depth is below the groundwater level, the maximum achievable depth decreases.For excavation in wet soils up to the depth of 8.24 m, the four ESSs can be used.In the range of depth 8.24-14.42m, secant pile, diaphragm and jet grout walls can be used.In the range of depth 14.42-20.11m, secant pile and diaphragm walls can be used.For excavations deeper than 20.11 m, diaphragm walls can be used.
The current study could provide useful guidelines for a better selection of ESS.The focus herein is on the effects of soil condition, excavation depth and groundwater level on the applicability and thus the selection of four ESSs.However, other issues need to be addressed in future investigations.Among these issues are the following: Understanding the effects of other important factors such as the distance to nearby structures, condition of nearby structures, depth interval between the level of new excavation and level of nearby structures, and displacement of soil around the new excavation.Understanding the effects of other types of soil that are common in the practice of excavation support such as silty sand, silty clay, sandy clay and sandy gravel.
Understanding the effects of soil condition as evaluated in terms of more comprehensive soil indices such as the N-value of the standard penetration test.The use of such an index will provide for representing much wider conditions for different soils.

Conclusions
This paper presents the results and discussion of a study conducted to investigate the favourable soil conditions and maximum excavation depths for ESSs in dry and wet soils.It focuses on soldier pile, secant pile, diaphragm and jet grout walls because of their common use worldwide.Clear guidelines that can help in the selection of an appropriate ESS for a given site condition could be established.On the basis of the presented results and discussion, the following conclusions could be drawn: For soldier pile walls, stiff clay is the most favourable soil and granular soil is the most favourable soil for secant pile walls.Soft clay is the worst soil condition.
As for the diaphragm walls, the favourable soil is layered soil and granular soil is the most favourable soil condition for jet grout walls.Secant pile and diaphragm walls are more applicable to soft clay than jet grout and soldier pile walls.The maximum achievable depths for excavation in dry soils are 3.00-40.00,4.00-40.00,5.00-50.00and 1.00-40.00m with average values of 11.9, 17.88, 19.67 and 26.39 m for the soldier pile, secant pile, diaphragm and jet grout walls, respectively.

Figure 3 .
Figure 3. Applicability of soldier pile walls to different soils.

Figure 4 .
Figure 4. Applicability of secant pile walls to different soils.

Figure 2 .
Figure 2. Applicability of four ESSs to different soils.

Figure 5 .
Figure 5. Applicability of diaphragm walls to different soils.

Figure 6 .
Figure 6.Applicability of jet grout walls to different soils.
These results align with the literature where several studies recommend diaphragm walls (MDT Geotechnical Manual 2008; Ali 2011; Lee 2012; Mary and Sankar 2016) and secant pile walls (MDT Geotechnical Manual 2008; Ali 2011; Mary and Sankar 2016) for wet excavation.Soldier pile walls are not the favourable system for wet excavation because of the voids inside the walls (MDT Geotechnical Manual 2008; Ali 2011; Mary and Sankar 2016).The average values of the responses are 2.29, 4.02, 4.58 and 3.65, and the corresponding standard deviations are 1.39, 1.07, 0.77 and 1.25 for the soldier pile, secant pile, diaphragm and jet grout walls, respectively.

Figure 8 .
Figure 8. Maximum excavation depth for four ESSs in dry excavation.

Figure 9 .
Figure 9. Applicability of four ESSs to wet excavation.

Figure 10 .
Figure 10.Applicability of four ESSs to wet excavation.

Figure 11 .
Figure 11.Maximum excavation depth for four ESSs in wet excavation.

Figure 12 .
Figure 12.Comparison between maximum excavation depths for four ESSs in dry and wet excavations.
table conditions without extensive dewatering (e.g.Union Pacific Railroad 2001; Bobet 2002; MDT Geotechnical Manual 2008).Installation of soldier piles causes significant surface settlement, vibration and noise (e.g.Union Pacific Railroad 2001; Ali 2011).The difficulty to control the basal heave of soil at the bottom of the excavation is another limitation of soldier piles (e.g.Union Pacific Railroad 2001; Ali 2011; Quintanar 2014).

Table 2 .
Calculation of Cronbach's alpha.sum of squares, df ¼ degrees of freedom, MS ¼ mean square, F ¼ F-statistic, P-value ¼ P-value for the one-way ANOVA test, F crit ¼ the F critical value (calculated based on F and df values), Cronbach's alpha ¼ 1 À (MS ERROR /MS ROWS ).

Table 4 .
Data of reported case histories.