Ground Movements in Station Excavations of Bangkok First MRT

INTERNATIONAL SOCIETY FOR SOIL MECHANICS AND GEOTECHNICAL ENGINEERING

This paper was downloaded from the Online Library of the International Society for Soil Mechanics and Geotechnical Engineering (ISSMGE). The library is available here: https://www.issmge.org/publications/online-library

This is an open-access database that archives thousands of papers published under the Auspices of the ISSMGE and maintained by the Innovation and Development Committee of ISSMGE. Geotechnical Aspects of Underground Construction in Soft Ground – Ng, Huang & Liu (eds) © 2009 Taylor & Francis Group, London, ISBN 978-0-415-48475-6

Ground movements in station excavations of Bangkok first MRT

N. Phienwej Asian Institute of Technology, Bangkok, Thailand

ABSTRACT: The characteristics of movements of diaphragm wall and ground in the excavation of 18 stations of the first Bangkok underground MRT line were evaluated. Three modes of deflected shapes of the walls were observed at different excavation depths, namely cantilever mode, braced modes with bulge in soft clay and bulge in stiff clay. The ratio of maximum lateral wall deflection developing with excavation depth and the ratio of ground surface settlement to excavation and the normalized zone of ground surface settlement varied with the mode of wall deflection. Back-calculation of undrained soil moduli for different soil layers were made from wall movement data of three selected stations using the 2-D linear elasto-plastic FEM analyses. The modulus values, which were higher than those commonly obtained from conventional triaxial tests, can be used as guideline for future works in Bangkok.

1 INTRODUCTION in soft and subsiding Bangkok ground. The contracts made its compulsory that full instrumentation program Deep excavation by means of strutted concrete be implemented during excavation for design verifi- diaphragm walls is often used for construction of cation and safety assurance. Evaluation on the actual multi-level building basements in Bangkok soft soil. performance at sites was performed to confirm the suf- It is superior to the excavation with steel sheet piles ficiency in the design of the support systems for the for control of ground movement to avoid damages to MRT station excavations. A comprehensive study was adjacent structures. Prior to the construction of the first made on the aspect of the wall and ground movements Bangkok MRT project, few studies were made on the (Hooi, 2003) and the salient points from the study are characteristics of the ground movement and its pre- reported herein. diction (e.g. Phienwej and Gan, 2003 and Teparaksa et al, 1999. etc.). However, it was the implemen- tation of the first Bangkok MRT subway line that provided systematic and comprehensive monitoring 2 PROJECT DESCRIPTION data on the excavation of station boxes that allowed in- depth evaluation on the characteristics of the wall and The ISP Blue Line is the first underground MRT line ground movements associated with deep excavation in constructed in Bangkok. It comprises 22-km-long twin Bangkok subsoil conditions using strutted diaphragm single-track tunnels, 18 stations and a depot. The hori- walls.The project involved the deepest ever excavation zontal alignment mainly follows the right of way of city made in Bangkok to date.The excavations of all station roads. Construction of the underground structures was boxes were fully instrumented. Monitoring data from implemented under two fast track design-built con- 18 station excavations were compiled, summarized and tracts, each having approximately the same amounts interpreted. of work. The South Contract involved construction of The construction of the first MRT underground a twin bored tunnels from the inter-city railway termi- project in Bangkok, the Mass RapidTransit Initial Sys- nal at Hua Lamphong eastwards for 5 km beneath the tem Project (MRT ISP Blue Line) was started in 1998. busy Rama IV road to the Queen Sirikit National Con- Prior to that there were public doubts on technical via- vention Center, then 4.5 km north beneath the narrow bility of the construction and operation safety of under- business Asoke road, and Ratchadaphisek road ending ground MRT in Bangkok soft soil. That pessimistic on the surface near the depot. Works of the North Con- outlook led to a call for an in-depth investigation on the tract continued for 4.5 km north along Ratchadaphisek application and performance of the excavation method road to Lad Phrao road then turned west to Chatuchak and support systems to be integrated in the excavation Park and finally terminated beneath the Bang Sue

181 railway station. The 18 stations of the project are as by approximately 23 m from the original hydrostatic follows: profile. The groundwater pumping has also created regional subsidence throughout Bangkok metropolis. South Contract North Contract At the locations of the project, a perched water table is typically encountered in Made Ground. Below this horizon, hydrostatic conditions are generally found to 1. Hua Lamphong Station 1. Thiam Ruam Mit Station a depth of approximately 8 m to 10 m depending on 2. Sam Yan Station 2. Pracharat Bumphen Station the location and thickness of First and Second Sand 3. Silom Station 3. Sutthisan Station layers. Typically, the upper few metres of sand layers 4. Lumphini Station 4. Ratchada Station underlying First Stiff Clay and Very Stiff Clay have 5. Bon Kai Station 5. Lad Phrao Station been dewatered. 6. Sirikit Centre Station 6. Phahonyothin Station 7. Sukhumvit Station 7. Mo Chit Station 8. Phetchaburi Station 8. Kamphaeng Phet Station 9. Rama IX Station 9. Bang Su Station 4 STATION CONSTRUCTION

The MRT stations had following features: 3 GROUND CONDITIONS – Typically,three levels of structure, with a centre platform that is fed by stairs and escalators between two Bangkok is situated on the southern part of the low lines of columns down the middle of the station. lying Chao Phraya plain, which extends north from the – Up to 230 m long and approximately 25 m wide, coast line at the Gulf of Thailand up to approximately excavated up to a depth of 25 m to 30 m below the 350 km and spans east-westward up to 150 km. The ground surface. flat topography plain is covered with a thick marine – The perimeters were of diaphragm walls, 1.0 m thick clay layer, which overlies a very thick series of allu- and 30 m to 35 m deep and solid in-situ reinforced vial deposits of alternating stiff to hard clay and dense concrete slabs, typically 0.9 m thick for the roof to very dense sand to gravel. The thick soft clay layer slabs, 0.7 m thick for the intermediate slabs and generally extends from the ground surface to a depth of 1.75 m thick for the base slabs, which were used 12 to 15 m. The soft clay which is known as “Bangkok as the excavation support system and permanent soft clay” has high water content (70–120%), high structures later on. plasticity, low strength and high compressibility. – There were three stacked stations, at Samyan, Silom The shallow subsoil of the upper 35 m zone is rela- and Lumphini Stations, due to space constraints tively uniform and generally consists of layers of soft caused by the existence of the foundation piles of the to medium clay, stiff to hard clay and sand. Below this long road flyover and a water transmission tunnel at shallow zone, alternating layers of stiff to hard clay and these busy intersections. dense sand exist to a great depth. Bedrock is found at – There was a side platform station at Bang Su Sta- depths more than 450 m. The typical subsoil profile tion, with two levels only, to accommodate the track for the first 50 m depth is listed as follows: alignment for future elevated extension of the line to the north. – Made Ground – The remaining stations were constructed as Centre – Bangkok Soft Clay Island Platform stations. – First Stiff Clay – Pracharat Bamphen and Sutthisan Stations incorpo- – Medium Dense Clayey Sand, Sandy Clay and Silty rated intersection road underpasses on the roofs of Clay the stations. – Very Stiff Sandy Clay/First Bangkok Sand – Silom, Lad Phrao and Phahonyothin Stations were – Second Hard Clay excavated underneath foundation piles of existing – Second Bangkok Sand road flyovers and thus, the station structures and foundation were designed to support the flyovers Maconochie (2001) summarized the general soil via cross-beam and underpinning bored piles. profile and properties at the Bangkok MRT ISP Blue Line. The variation in soil profile along the align- The top down construction technique was adopted ment was observed primarily in the Very Stiff Sandy for all station box excavations with diaphragm walls Clay/First Bangkok Sand layer and the soils immedi- and concrete slabs as the excavation support system, ately below it. Figure 1 show the soil profile along the which was later utilized as the permanent structures MRT alignment and at the stations. of the stations. The designs were made with an aid Deep well pumping in Bangkok and its environs of FEM analyses. The excavation depths and the toe has reduced the pore water pressures in the sand layers depth of the diaphragm walls of all station excavations

182 Figure 1. Soil profile along MRT alignment and at the stations.

are summarized in Fig. 2. It should be noted that the The data were screened to preclude movements not ratio of the depth of the wall embedment to the depth expressly related to the excavation and support instal- of excavation was significantly different between the lation, such as diaphragm wall construction, tempo- two contracts, primarily due to the difference in the rary decking works and the initial 2.3 m excavation that design criteria adopted by the two different designers. involved driving sheet piles, backfilling and extract- ing the sheet piles subsequently.A detailed study of the instrumentation data obtained was undertaken by com- 5 INSTRUMENTATION DATA paring observed ground movement among the station excavations. Factors that may result in such patterns The measurement data from inclinometers and sur- of data was examined and deduced, such as: face settlement points were compiled and interpreted to evaluate the overall performance of the station exca- – Station box configuration and dimension vations in Bangkok subsoil using diaphragm walls. – Construction sequences

183 NORTH CONTRACT Maximum Lateral Wall Movement/Excavation Depth, d /H Hmax 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0

Excavation Depth, H (m) 20

25 Cantilever mode (H=1.6m-4.0m) Braced mode-soft soil (H=6.5m-10.8m) Braced mode-stiff soil (H=12.4m-16.7m) Braced mode-stiff soil (H=20.0m-32.6m)

SOUTH CONTRACT Maximum Lateral Wall Movement/Excavation Depth, d /H Hmax Figure 2. Depth of excavation and toe depth of D-wall. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 0

Excavation Depth, H (m) 30

35 Cantilever mode (H=1.6m-4.0m) Braced mode-soft soil (H=6.5m-10.8m) Braced mode-stiff soil (H=12.4m-16.7m) Braced mode-stiff soil (H=20.0m-32.6m)

Figure 4. Normalized maximum wall deflection versus excavation depth.

stiff soil layers was more predominant. This different behavior may be attributed to the difference in soil profiles. In addition, South Contract had three deep station excavations of vertically stacked platforms. The maximum lateral wall deflection of the 18 station excavations were in the range of 10–47 mm. The Figure 3. Modes of wall deflection. monitoring data also showed that there were signif- icant variations in the shape and magnitude of the – Variation in soil profile and properties lateral wall movement within some of the station exca- – Temporary works or presence of structures adjacent vations. The variation may be attributed to a number to excavation box of factors including the variation in soil profile and properties over the plan area of the station, adjacent temporary surface works, and confinement from road 5.1 Lateral wall movement pavements and buried utility structures. The area of In general, most inclinometers all station excavations the excavation was quite large (about 25 m wide and showed that the cantilever mode was the most predom- 200 m long). For the three stations adopted for the inant of wall deflection shape at the initial excavation detailed analysis in this study, the variations in the stage, while the braced excavation mode developed in maximum lateral wall movements were as follow: 28– the subsequent stages as the excavations were deep- 38 mm for the deepest Silom Station, 15 to 40 mm ened. Figure 3 shows the three modes of deflected for Sirikit Station, and 18–27 for Thiam Ruam Mit shape of wall movement, which occurred at differ- Station. Figure 4 summarizes the ranges of the normal- ent excavation depths. The cantilever mode was most ized maximum lateral wall movement with excavation common during the first excavation stage. The braced depth (δHmax/H) versus excavation depths recorded by excavation mode with bulge in soft soil prevailed at the all inclinometers at the 18 stations. The plots show second and third excavation stages. This mode con- that in the first stage cantilever mode excavation the tinued to dominate the pattern of lateral movements value of δHmax/H was as high as 1.60. The maxi- for North Contract, but data of South Contract exhibit mum value of δHmax/H decreased with the increase that the braced mode with bulge in the underlying in the excavation depth in the braced modes of the

184 For shallow excavation under the cantilever mode the value of δvmax/H can be higher. The characteristics of ground surface settlement behind the excavation with diaphragm walls can be categorized according to the mode of wall deflection of excavation depth. The zones of ground settlement for the three modes of wall deflection in Bangkok soil are marked in figure. In addition, the three categories of ground movement in braced wall excavation for flexible walls (sheet piles or soldier piles) suggested by Peck (1969) are also shown in the plot. The level of ground settlement in excavation with diaphragm wall in Bangkok soft soil is much smaller than that predicted by Peck’s chart for flexible wall. However, the influence zones of ground settlement were wider than those suggested by Peck’s.This finding can be used as a general guideline for prediction ground surface settlement from deep excavation with diaphragm walls in Bangkok subsoil condition.

6 PREDICTION OF MOVEMENTS

For the design of the MRT station excavations of both contracts, FEM analysis were made to determine ground movement and forces on the diaphragm Figure 5. Normalized settlement versus distance from wall. walls and bracings. The analyses mainly utilized linear elasto-plastic Mohr-Coulomb soil parameters. The parameters were obtained from the soil investigation walls. The decrease in the value with depth was due to and testing program made for each station excavation. the change in soil condition at the excavation bottom Triaxial compression tests as well as pressuremeter from soft clay to stiff clays as the excavation depth tests were conducted to determine the values of soil increased. When the excavation depth reached stiffer modulus for design analysis. The instrumentation data δ soils, Hmax/H decreased and the range of the value provided valuable information to evaluate the appro- became narrow. When the excavation bottom was still priateness of the soil model and the soil parameters δ in the soft clay layer, the maximum value of Hmax/H used in the design calculation. In this study, monitor- was smaller than 0.5. The value was smaller than 0.2 ing data from three representative station excavations, when the excavations were deeper, in stiff clays. i.e. Silom, Sirikit and Thiam Ruam Mit Stations, were examined in details and suitable soil parameters were back-calculated using a continuum FEM 5.2 Ground surface settlement analysis. Computer code PLAXIS 2D was adopted in The maximum ground surface settlement was the study. Effective stress strength parameters were observed at 58 mm at Bon Kai Station and 75 mm at adopted in the undrained analysis. The drawdown phe- Pracharat Bumphen, in South and North Contracts, nomenon of the piezometric levels was considered in respectively. Figure 5 shows plot of the ratio of maxi- the simulation. mum settlement to excavation depth (δvmax /H) versus Silom Station was the deepest excavation in the distance from excavation normalized by the exca- Bangkok to date (32.6 m deep), with a vertically vation depth (D/H). For shallow excavations under stacked platforms thus it had four levels of slab below cantilever mode of wall deflection, the zone of ground the roof. The station was designed to underpin the settlement may extend up to D/H of 7 to 10. As for existing flyover roadway running over the station the excavation depths while the wall deflection was length. A dense sand layer of the first Bangkok Sand under the braced mode with bulge in soft clay layer, was encountered from depth 8.5 m above the final the δvmax/H may extend up to D/H of 7. For deeper excavation level. Hence, the excavation required dewa- excavations while the wall was deflected under the tering. The diaphragm wall was toed into the Second mode of bulge in stiff clay layer, the zone of ground Sand layer. settlement may extend up to a distance of D/H of 4 Sirikit Station had the typical configuration of the and δvmax/H value may be as high as 0.35 Under the centre island platform with three levels below the roof braced mode in soft clay, δvmax/H value may reach 0.55. slab. The first stage and final excavation depths of

185 3.65 m and 23.6 m respectively, which were similar to – The maximum lateral wall movement (δHmax)was majority of other stations. smaller than 47 mm in both contracts. The normal- Thiam Ruam Mit Station was selected because the ized wall deflection, δHmax/H in the cantilever mode soil profile was slightly different from those found in of movement was as high as 1.60, while it was the first two stations. The area had a thicker First Stiff reduced to no more than 0.60 and 0.40 in the latter Clay layer with lenses of clayey sands. In addition, the stages of excavation when the wall deflection devel- first stage excavation was very shallow with roof slab oped in the braced mode with bulge in soft clay and was only at 1.8 m depth. braced mode with bulge in stiff soil, respectively. The back-calculation using the lateral ground move- – The maximum ground surface settlement (δVmax) ment data from the excavations of the three stations was 58 mm for South Contract and 75 mm in North suggested the suitable undrained soil modulus param- Contract. The normalized maximum ground settle- eters as follows. ment with excavation depth, δvmax/H, was smaller – Soft and Medium Clay : E = 500 C kN/m2 than 0.55 and 0.35 for Modes 2 and Mode 3 wall u u deflection, respectively. The normalized distance – First Stiff Clay : E = 700 N kN/m2 u 60 from excavation of the zone of ground settlement, – Clayey Sand and Silty/ : E = 900 N kN/m2 u 60 D/H, varied from 7.0 and 4.0 for the two modes Sandy Clay = 2 of wall deflection. In the initial excavation stage – Second Hard Clay : Eu 1600 N60 kN/m of cantilever mode, the values of both normalized = 2 – Third Hard Clay : Eu 2500 N60 kN/m settlement and distance of ground movement were higher that those in the braced modes. where Cu is the corrected field vane shear strength and – Back-calculation of soil moduli of different soil N60 is the corrected SPT “N” value according to Liao and Whitman (1986). layers using monitoring data from three selected These back-calculated values of soil modulus are stations showed higher values than those com- higher than those commonly obtained from conven- monly obtained for conventional laboratory tests. = tional laboratory triaxial tests. It reflects the modulus The values are: Soft and Medium Clay: Eu 500cu, = 2 values at low strain level which would be the dominat- First Stiff Clay Eu 700 N60 kN/m , Clayey Sand = 2 ing response of soil in the excavation problem (Mair, and Silty/Sandy Clay Eu 900 N60 kN/m , Second = 2 1993). The soils would be mainly under unloading Hard Clay Eu 1600 N60 kN/m , Third Hard Clay = 2 condition of stresses. Eu 2500 N60 kN/m In similar early studies, Phienwej and Gan (2003) and Teparaksa (1999) both proposed the same modulus parameter of the soft clay as Eu = 500cu. While REFERENCES for stiff clay, the value of Eu = 1200Cu and 2000Cu were suggested, respectively. Based on the relation- Hooi, K.Y.2003. Ground MovementsAssociated with Station 2 ship of Cu = 0.6N60 kN/m typically used for Bangkok Excavations of the First Bangkok MRT Subway. Master Thesis, Asian Institute of Technology, Bangkok. subsoil, the parameters are equivalent to Eu = 720N60 and 1200N kN/m2, respectively.The back-calculated Liao, S. & Whitman, R.V.1986. Overburden Correction Fac- 60 tor for SPT in Sand. Journal of Geotechnical Engineering. values from this study were comparable to those American Society of Civil Engineers 112(3): 373–377. suggested by Phienwej and Gan (2003). Mcconochie, D. 2001. Geotechnical Completion Report MRTA Chaloem Ratchamongkhon Line. A CSC Report submitted to the MRTA. 7 CONCLUSIONS Mair, R. 1993. Developments in geotechnical engineering research: application to tunnels and deep excavations. Unwin Memorial Lecture 1992, Proceedings of Institution The following conclusions can be drawn from the study of Civil Engineers, Civil Engineering, 93, Feb: 27–41 of the wall and ground movements in the excavation of Phienwej, N. & Gan, C.H. 2003. Characteristics of the stations of the first Bangkok MRT underground. Ground Movements in Deep Excavations with Concrete Diaphragm Walls in Bangkok Soils and their Predic- – Three modes of deflected shapes of the wall move- tion. Journal of The Southeast Asian Geotechnical Society ment were observed at different ranges of exca- 34(3): 167–175. vation depth. Mode 1: Cantilever mode (H = 1.6 Teparaksa, W.,Thasnanipan, N. & Tanseng, P.1999. Analysis m–4.0 m), Mode 2: Braced mode with bulge in soft of Lateral Wall Movement for Deep Braced Excavation in clay layer (H = 6.5 m–11 m), and Mode 3: Braced Bangkok. Proc. of AIT 40th Anniversary Conference, AIT, mode with bulge in stiff soil (H = 12.4 m − 32.6 ˙m). Bangkok, Thailand.

186