Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No.

3 September 2023 ISSN 0046-5828

Performance Monitoring of Restraint Back-to-Back Mechanically Stabilized Earth

Walls for Dam Crest Rehabilitation of Mae Suai Dam, Thailand
K. Wannasiri1, S. Prempramote2, and S. Soralump3
1
PhD. Candidate, Department of civil engineering, Faculty of engineering, Kasetsart University, Bangkok, Thailand
2
Department of Civil Engineering, Faculty of Engineering, Kasetsart University, Bangkok, Thailand
3
Geotechnical Engineering Research and development center, Civil Engineering Department, Kasetsart University, Thailand
E-mail: [email protected]

ABSTRACT: The Mechanically Stabilized Earth walls (MSE walls) for dam crest rehabilitation was constructed at Mae Suai dam, Thailand.
Welding mesh gabion was used as the facing on both sides, the polymetric geogrid and rebar were used as reinforcements. Furthermore, the
MSE walls was placed on the original earth dam crest, steel sheet pile was installed at upstream and downstream side to prevent leakage and
control the settlement of the new dam crest. The instruments were installed at various test sections to careful field monitoring to obtain high-
quality data. The results obtained from 2D finite element method simulations were in good agreement with the field measurements, the lateral
deformation and settlements were very small. The axial forces in rebar reinforcement were found to be continually changing due to deformation
of foundation, external stimuli and construction factors. Likewise, the strain measured in all positions of geogrid reinforcement was very low.
Combining steel reinforcement (high stiffness) with geogrid reinforcement (low stiffness) was redundant. Most of the lateral stresses are
resisted by the former than the later. It can be concluded that in a reinforced soil wall that uses two or more types of reinforcing materials,
tensile force is developed in higher stiffness material.

KEYWORDS: Dam crest rehabilitation, Dam crest raising, MSE Walls, and Reinforced earth embankment.

1. INTRODUCTION Bergado et al. (2003) analyzed the behavior of reinforced

embankment with silty sand backfill built on soft soil. The
The reinforced earth embankment namely, Mechanically Stabilized
embankment was reinforced with galvanized and PVC coated
Earth walls (MSE walls) or Geosynthetic reinforce soil walls (GRS hexagonally shaped geogrids. Bergado and Teerawattanasuk (2008)
walls) are used extensively as earth retaining structures because of compared the effect of embankment geometry with 2D and 3D
their cost-effectiveness and ability to withstand much larger simulations and concluded that 3D analysis must be conducted for
differential settlements than conventional reinforced concrete short embankments to obtain good agreement with measured field
retaining walls (Kim et al., 2012; Watanbe et al., 2003). Various types data. Huang et al. (2009) has investigated different soil constitutive
of MSE wall facings and reinforcements are used depending on the models and their influences on the results. The paper confirmed that
specific application, soil conditions and wall. Ho and Rowe (1996) the modified Duncan–Chang model is a suitable constitutive model
found that the reinforcement stiffness, vertical spacing and length to
and that the parameters used in that model can be determined from
wall height ratio, L/H, are important parameters that influence the conventional triaxial testing. Baral et al. (2016) compared the
wall displacement response. Rowe and Ho (1998) showed that the behavior of polymeric and metallic reinforced embankments on hard
magnitude of wall lateral displacement is influenced by the soil foundation with 3D numerical simulations conducted using PLAXIS
friction angle and a reinforcement stiffness factor. The simplified 3D. The lateral displacements and settlements were very small in the
design and analysis methods of reinforced earth embankment are case of the MSEW with inextensible reinforcement. The
provided in design guidance documents such as BS8006 (BSI, 2010) corresponding lateral and vertical deformations in the RSS were
in the UK, AASHTO (AASHTO, 2012) and FHWA (Berg et al., much larger due to its extensible reinforcing materials. Furthermore,
2009) in the USA. Issues related to the design and factors affecting
many researchers have studied the behavior of Back-to-Back MSE
the performance of reinforced soil have been addressed by many walls (Benmebarek et al., 2016; Benmebarek and Djabri, 2017; El-
researches in recent times (Allen et al., 2004; Allen et al., 2003; Sherbiny et al., 2013; Lajevardi et al., 2021; Samee et al., 2021; Xu
Bathurst et al., 2008; Bathurst et al., 2009; Bathurst et al., 2006; et al., 2021; Yang et al., 2022). However, a few research has studied
Huang et al., 2010; Kongkitkul et al., 2007; Leshchinsky, 2009; the reinforcing retaining walls used for dam crest rehabilitation/
Miyata et al., 2015). Also, the behavior of reinforced earth structures raising. Hardianto Fransiscus et al. (2013) describes the design
has been comprehensively studied through field observation of full- method and construction challenges of a geosynthetic-strip-
scale physical model, laboratory model testing, and numerical reinforced mechanically stabilized earth (MSE) wall used for the
simulation. Shivashankar (1991) observed the behavior of a welded expansion of Los Vaqueros reservoir dam in Contra Costa County,
wire wall with poor quality, cohesive-friction backfills on soft CA. The site is located in a high seismic area, and with their proven
Bangkok clay. Voottipruex (2000) studied the behavior of full-scale performance under such conditions, an MSE wall with a maximum
embankment built in AIT campus which was reinforced with height of 15m was designed and installed to provide a wider dam crest
hexagonal wire mesh up to 6 m with 10° inclined of gabion facing. while being part of the embankment system to increase the dam height
Bergado et al. (2000) simulating the behavior of the full-scale test by 10.4m.
embankment were the method of applying the embankment loading A 6-m-high reinforced earth embankment for dam crest
during the construction process, the variation of soil permeability rehabilitation was constructed at Mae Suai dam, Thailand. Welding
during the consolidation process, and the selection of the appropriate mesh gabion was used as the facing on both sides, the polymetric
model and properties at the interface between the soil and geogrid and rebar were used as reinforcements, steel sheet pile was
reinforcement. Holtz and Lee (2002) made report on research installed at upstream and downstream side. This embankment was
conducted on the internal stability of reinforced soil walls. Using the fully instrumented with piezometers, settlement plates, inclinometers
results of monitoring of 6 walls with different reinforcement elements and strain gauges and subjected to careful field monitoring to obtain
and types of backfill material, they made recommendations for high-quality data. In this research, the field measurement data was
improving the modeling techniques for the level of working stress. verified with 2D FEM analysis using MIDAS GTS to determine the
27
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

performance of back-to-back MSEW for dam crest rehabilitation. dam crest and it will be able to control the normal seepage and prevent
Particular attention was given to the lateral displacements, vertical leakage along the joints.
settlements and axial forces in the reinforcements.

2. MAE SUAI DAM AND DAM REHABILITATION

Mae Suai dam is a 59 m high structure with 400 m crest length and
having a reservoir capacity of 73 million cubic meters (Figure 1). The
dam has been in operation since 2003. The RCC section is used as an
overflow spillway and was designed for 500 years return period of
flood. The RCC material is a low paste RCC covered with the
conventional concrete (CVC). Figure 2 shows the longitudinal and
transverse section of the dam. The RCC section consists of an
overflow spillway and gravity retaining wall at both sides to create (a)
flow channels and retain the earth dam at both sides.

(b)

Figure 1 Mae Suai Dam (Soralump et al., 2023)

The RCC section is surrounded by earth zone dam. Core trench of the (c)
earth zone dam was excavated to the foundation rock in the river bed Figure 2 Longitudinal and Cross Section of Mae Suai Dam: (a)
and abutments. Impervious clay consisted of low-plasticity clay (CL) Longitudinal section; (b) Cross section A-A; (c) Cross section B-
and internal filter (sand and gravel) consisted of clayey sand (SC) B (Soralump et al., 2023)
materials to reduce the water pressure and discharge the seepage
water into RCC gallery. Shell zone or random zone is made up of
semi-impervious coarse grain earth (low-plasticity clay (CL), clayey
sand (SC) and silty sand (SM)) with horizontal drain to drain out the
water during drawdown period and maintain the stability of shell
zone. The earth zone extends in both side of the abutment. The 6 m
high RCC retaining block was constructed over the earth filled
material at the downstream of dam crest to reduce the earth fill work
on downstream slope and lower the construction cost.
The transition trapezoidal RCC block (Block D) was constructed
near the joint between RCC spillway section and earth zone dam.
Furthermore, to prevent the erosion at the crest of earth dam during
the overtopping of spillway, RCC blocks A, B and C were constructed
as a water guide wall (wing wall). These blocks were placed directly Figure 3 Removing of the RCC block and dam crest (Soralump
over the earth filled material (Soralump et al., 2023). et al., 2016)
In 2004, after 1 year of operation of Mae Suai Dam, water
overflowed the spillway and leakage was observed at the downstream 3. DESCRIPTION OF THE EMBANKMENT
crest in contact area between earth fill dam and RCC spillway
structure. The water flow was clearly observed behind the RCC block A 6-m-high reinforced earth embankment for dam crest rehabilitation
where differential settlement was also clearly visible. The leakage was designed by Geotechnical Engineering Research and
was observed when reservoir reached a certain elevation near the dam development center (GERD) and constructed by Kanber Geotechnic
crest. The repair work has been done by installing the impervious (Thailand) Company limited with construction supervision by the
membrane over the surface of RCC blocks. The leakage flow was Royal Irrigation Department (Thailand) and Samart Engineering
reduced after the repair of dam but did not disappear completely Consultants Company limited acted as a project consultant. The MSE
(Soralump et al., 2016). wall was 177 m along the dam crest and 10.90 m wide at the top and
Royal irrigation department (Thailand) decided to rehabilitate was prepared by welding the mesh Gabion 1.20x1.20x1.20 m as the
Mae Suai dam and repair works was proposed to solve the leakage facing on both sides and fastened with rebar (DB16 mm) between the
problem and stability of the dam during future earthquake. It has been gabion. The steel plate of dimension 2 0 0 x2 0 0 x9 mm and nut were
proposed to remove the RCC blocks A, B, C, D and blocks on the used for the connection joints as shown in Figure 4 and Figure 5. In
crest of the earth dams (Figure 3) and replace RCC block on the addition, the polymetric geogrid reinforcement was added to the MSE
downstream crest by more flexible structure. In this case, MSE Walls wall. The spacing of the rebar reinforcement was 0.60 m. in vertical
(MSEW) will be used (Figure 4), so that there won't be any rigid and and horizontal directions. Likewise, the vertical spacing of the
brittle that crack when subjected to seismic force or differential geogrid reinforcement was 0.30 m. Furthermore, the new MSE walls
settlement. Likewise, no further significant displacement, both will be placed on the original dam crest that consist of 3 materials
vertically and horizontally, will be observed from the load of a new type: 1) Impervious Core 2) Filter and 3) Random Zone. Steel sheet

28
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

piles of 10 m and 6 m length were installed at the upstream and

downstream sides to prevent leakage and control the settlement of the
original earth dam crest.

3.1 Preliminary Design of MSE Walls/Embankment

Preliminary design of the Back-to-Back MSE wall was carried out
based on LRFD Method (AASHTO, 2012; Berg et al., 2009). The
external and internal stability has been analyzed according to the
geometry of the wall. Sand and gravel available around the
construction site is used as backfill material. The wall height assumed
for the preliminary design is 6 m and the incline facing was 18.30 Figure 5 Connection joints between the rebar and gabion
degrees from the vertical.
The internal stability, tension in the reinforcement behind the
failure surface was checked against the lateral internal earth pressures
on the assumption that each side of the wall was independent. In the
design, the designer has divided the behaviour of reinforcement load
into two parts: 1) During construction and static loads, the geogrid is
defined as a reinforcing material, failure surface is determined by
both the coherent gravity method and the coulomb method
concurrently. 2) seismic loads, rebar is defined as a reinforcing
material, failure surface is determined by the coherent gravity
method. Furthermore, finite element analysis was performed to
determine the tensile force in both reinforcing materials. The external
stability was examined using the 2D finite element analysis by
strength reduction method (SRM) as mentions in the next section

3.2 Construction Method

After the removal of original dam crest (Figure 3), steel sheet piles of
depth 10m and 6m were installed at the upstream and downstream
sides. Site clearing and levelling works were carried out for the Figure 6 Rebar reinforcements installation
marking of the position of the proposed MSE wall/embankment.
At first, drainage system was installed for the new dam crest and
Geosynthetic Clay Liner (GCL) was laid over the wall foundation to
prevent the seepage of water into the earth dam from above. The
welding mesh was placed along the dam crest on two sides and filled
the gabion with 30 cm high rock. The first layer of geogrid across the
embankment from the upstream gabion to the downstream gabion
was installed (Figure 10). The geogrid was installed and backfill layer
was compacted layer by layer until the height of wall was reached
(Figure 7). At the back of the gabion, geogrid was folded and
geotextile was installed to prevent the backfill leaked out into the Figure 7 Construction of full scale MSE wall
gabion (Figure 5). Longitudinal and transverse rebars were installed
through the gabion and attached to the gabion by steel plate and nut 4. DAM INSTRUMENT
(Figure 5 and Figure 6). While installing the gabion up to the 3rd
layer, the clay between MSE wall and the sheet pile was compacted The rehabilitation of the dam crest using the MSE wall using the
as shown in Figure 4. Finally, the upper surface with asphalt having additional instruments was carried out to investigate the behavior of
thickness of 0.10m shall be used as a traffic surface. During the the new dam crest. Figure 8 shows the locations of test section with
construction of the embankment, field density test at various selected instrumentation selected from the location with the most settlement
places were carried out using sand cone replacement method to ensure at the left and the right bank (Soralump et al., 2023) and the position
compaction was carried out to minimum of 95% standard proctor predicted that the wall would be least affected by boundary conditions
density. (Plane strain). The instrumentation consists of Inclinometer (INC),
Settlement Plate (SP), Piezometer (PI), Rebar Strain Gauges (RSG),
Geogrid Strain Gauges (GSG) and Surface Monument (SM) as shown
in Figure 9. Figure 10 shows the locations of test section with field
instrumentation at the left bank. Inclinometers were installed on the
side of the MSEW outside the steel sheet pile (at upstream and
downstream section) to determine the lateral deformation of the new
dam crest. Settlement plates were installed inside the MSEW at
impervious core and filter to determine the vertical settlement of the
wall foundation and settlement plates were installed inside the walls
facing on both sides. Vibrating Wire (VW) Piezometer were installed
at the foundation of the new dam crest between sheet pile and MSE
wall to monitor the water pressure behind the sheet pile. In the past,
VW Piezometer was installed near the contact between RCC Section
and earth dam section that has leakage (Soralump et al., 2023).
Vibrating Wire Rebar strain gauges were installed along the length of
Figure 4 Restraint Back-to-Back MSE Wall (Soralump et al., the rebar at the designed location based on the failure surface
2016) constructed using the coherent gravity method to determine the axial
29
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

forces generated in the rebar and internal stability of MSEW as shown

in Figure 11a. Ultrahigh-elongation Foil Strain gauges (Figure 11b)
were installed along the length of the geogrid reinforcement at the
designed location based on the failure surface constructed using
coherent gravity method and coulomb method to determine the tensile
strain generated in the geogrid. The Surface Monument was also
installed on the traffic surface above the new dam crest and on the
bridge over the RCC section to monitor the settlement along the dam
crest. V-Notch Weir was installed at the downstream side to
monitored the drainage of the MSEW and leakage. Figure 11 Rebar and Geogrid strain gauges installation (a)
rebar strain gauges (b) geogrid strain gauges

5. 2D FINITE ELEMENT MODEL OF MSE WALLS FOR

DAM CREST REHABILITATION
The finite element analysis was performed using a geotechnical
analysis software MIDAS GTS. In this research, 2D finite element
model of original earth dam and MSE Walls for dam crest
rehabilitation of Mae Suai Dam was created according to the dam
geometry. Stress-Seepage-Slope coupled analysis was used in the
analysis as there was sequential seepage-stress analysis and slope
stability analysis during the construction process (Midasgts, 2018).
Figure 8 Location of test section with instrumentation (left and Likewise, the slope stability analysis by strength reduction method
right bank) (SRM) was used to determine the safety factors of MSEW for dam
crest rehabilitation. Figure 12 shows the solution algorithm for 2D
Finite element analysis of MSEW for dam crest rehabilitation. At
first, the stress and deformation of the original earth dam was
analyzed using Stress-Seepage coupled analysis (Figure 13 and
Figure 14) followed by transient seepage analysis. The function of
water level change was determined from the water retention record in
first 3 years. The water level on upstream side was varied from the
Minimum Water Level (MFLs) to the Normal High Water Level
(NHWL) (Figure 2). During the first filling of a reservoir, the water
level was increased from MFLs to NHWL within 1 year and the water
level fluctuate during the period of storage in other years of operation.
Likewise, total of 14 construction stage (1 year/stage) with time step
of one month was used in analysis (Soralump et al., 2023). In the next
construction stage, the original dam crest elements in the parts has
Figure 9 Schematic diagram of the test section with been removed (Figure 3) and clear displacement was set to zero
instrumentation (Left Bank) preserving the stress history and deformation shape and it was
replaced with new dam crest rehabilitated elements (MSE walls, new
embankment and sheet piles) as shows in Figure 15 by setting the
water level to MFLs. Finally, the safety factor of new dam crest was
determined by Strength Reduction Method. Mohr-Coulomb model
was used for the foundation, filters and RCC materials while
Modified Cam-clay model was used for random and core of the dam.
Furthermore, linear elastic model was used for reinforcement and
steel sheet pile.

5.1 Geometry Mode

2D finite element modelling was performed under plane-strain
conditions. The geometry and height of the original earth dam was
selected from the cross-section of the original earth dam at the test
section (Figure 8) which was consistent with the FEA results by
Soralump et al. (2023) as shown in Figure 13 and Figure 14. To avoid
the boundary effects caused by the constraints of the numerical
Figure 10 Geogrid installation and instrumentation of test model, the foundation of the upstream and downstream embankment
section (left Bank) has been extended 120 m in both directions, twice the maximum
height of the dam (Gikas and Sakellariou, 2008; Soralump et al.,
2023). Therefore, the deformation of the dam body had very little
impact due to the constraints of the model. The plain-strain triangle
and quadrilateral elements type were used in RCC and soil elements,
the beam and embedded beam elements type were used in
reinforcement (rebar & geogrid) and steel sheet piles. Element sizes
varies from 0.10m to 20m, the smaller element size is located at the
original dam crest and MSE Walls and increases when its lowered.
The contact surface between steel plate and gabion are defined as
plate bearing joints because each part can slide and separate.

30
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

Therefore, to transfer the compression, rigid link type interface

elements have been used. The rigid link was a high stiffness element
that is only able to transfer compressive forces in a horizontal
direction and the movement of side surface was allowed in vertical
direction Figure 15 shows the 2D Finite Element Model of MSEW
for dam crest rehabilitation.

Figure 15 2D Finite Element Model of MSE Wall for dam crest

rehabilitation

5.2 Geotechnical Parameters of the Materials

The foundation of MSEW for dam crest rehabilitation was the
original earth dam that that has been in used for a while. Therefore,
the foundation of the new dam crest consists of the components of the
original earth dam, namely Impervious Core, Random Material,
Filter, Dam Foundation and RCC. Some rudimentary information
concerning the material parameters of foundation of MSE Walls used
in the analyses were taken from Final design report of Mae Suai Dam
Figure 12 Solution algorithm for 2D Finite element analysis (RID, 1998). The parameters were determined from the results of the
geotechnical tests done on the samples from the borrow areas which
has been identified during the geological survey of the site (Soralump
et al., 2023). The parameters of the original dam crest required for
Midas GTS analyses are tabulated in Table 1. The backfill materials
used in this embankment consisted of sand and gravel available near
construction site and was classified as well graded gravel (GW). The
welding mesh gabion wall-facing system was made from hot-dipped
galvanized rectangular wire mesh (RB 6 mm.), uniform square mesh
8x8 cm. with tensile strength (fy) of 2,830 kg/cm2 (Figure 5), baskets
size 1.20x1.20x1.20 m and filled with river rock of size 50 – 300 mm.
The properties of the backfill material required for Midas GTS
analyses are tabulated in Table 2. Two types of reinforcement, namely
Geogrid and Rebar reinforcement were used in the reinforced
embankment. The geogrid reinforcement was a high-strength steel
wire composite with polyethylene with the aperture dimension of 5 x
Figure 13 2D Finite Element Model of original earth dam, 57.50
5 cm (Figure 16). The material property was obtained from the wide-
m. high
width strip tensile strength test for geogrid along the machine and
along the cross-machine directions. The tensile strength along the
machine and along the cross-machine directions were 60.41 KN/m
and 56.56 KN/m respectively. The Elongation at break were 3.49%
and 4.44% respectively. The rebar reinforcement was a deformed
steel bar (DB16mm.) with tensile strength (fy) of 4,440 kg/cm2. The
properties of the reinforcement are tabulated in Table 3. The rebar and
geogrid reinforcement were considered to be a linear elastic material.
Steel sheet pile was used in that dam crest rehabilitation (Figure 17).
It was made from the hot-rolled steel and hot-dipped galvanized and
was installed using vibration machine at the upstream and
downstream side along the dam crest. The properties of the steel sheet
pile are tabulated in Table 3. The steel sheet pile was considered to
Figure 14 2D Finite Element Model of original dam crest be a linear elastic material. The interface coefficients were used in the
interaction between the backfill soil and the reinforcing materials,
namely rebar, geogrid and steel sheet pile. The model parameters at
the soil–structure interface can be generated from the soil using the
strength reduction factors (Rinter), defined as the ratio of the shear
strength of the soil–structure interface to the corresponding shear
strength of the soil. In this research, the strength reduction factors
(Rinter)suggested by Brinkgreeve and Shen (2011) were used as
shown in Table 4.

31
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

Table 1 Material properties of original dam used in FEM analyses

Name Clay Core Random Zone Filter Foundation RCC

Model Type Modified Cam Clay Modified Cam Clay Mohr Coulomb Mohr Coulomb Mohr Coulomb
2
Elastic Modulus, E (ton/m ) 3,000.00 3,000.00 5,000.00 300,000.00 1,000,000.00
Poisson’s Ratio (n) 0.30 0.30 0.30 0.20 0.20
Unit Weight, g (ton/m3) 2.02 2.02 1.85 1.94 2.40
Kx (m/sec) 2.0x10-7 2.0x10-6 1.0x10-4 3.0x10-7 1.0x10-7
Ky (m/sec) 5.0x10-8 1.0x10-6 1.0x10-4 3.0x10-7 1.0x10-7
Cohesion (ton/m2) - - - 50.00 30.00
Frictional Angle, f (Deg.) - - 35 40 40
Over Consolidation Ratio, OCR 1.000 1 - - -
Slope of NCL, λ 0.335 0.761 - - -
Slope of URL, κ 0.053 0.096 - - -
Slope of critical state line, M 0.856 1.02 - - -
Table note: NCL = normal consolidation line, URL = Unload-reloading line (Overconsolidation line)

Table 2 Properties of backfill material used in FEM analyses

Name Backfill Rock Fill

Model Type Mohr Coulomb Mohr Coulomb
2
Elastic Modulus, E (ton/m ) 5,000 5,000
Poisson’s Ratio (n) 0.30 0.30
Unit Weight, g (ton/m3) 2.07 1.95
Kx (m/sec) 1.0x10-4 1.0x10-4
Ky (m/sec) 1.0x10-4 1.0x10-4
2
Cohesion (ton/m ) - -
Frictional Angle, f (Deg.) 41 41

Table 3 Properties of reinforcing materials used in FEM analyses

Name Rebar Geogrid Sheet Pile Welding Mesh Steel Plate

Model Type Linear Elastic Linear Elastic Linear Elastic Linear Elastic Linear Elastic
7
Elastic Modulus, E (ton/m ) 2
2.04x10 7
28,550 2.04x10 7 2.04x10 2.04x107
Poisson’s Ratio (n) 0.30 0.45 0.30 0.30 0.30
Unit Weight, g (ton/m )3 7.85 0.93 7.85 7.85 7.85
Thickness (mm) - 1 126.33 - 9
Diameter (mm) 16 - - 6 -

32
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

supported the weight of the old dam crest which was heavier than the
new dam crest. The analytical results are consistent with the measured
results from the settlement plate as shows in Figure 21. The settlement
occurred only at beginning of the wall construction and at the end
when the wall height was 0.90 m. for SP1& SP3 and 1.50 m. for SP4.
Likewise, no settlement was detected in SP2. The maximum
settlement measured at position SP4 was 0.05 m. It can be seen that
the settlement that occurred was an immediate settlement caused by
construction conditions (insufficient preparation of the foundation
surface), especially in the area where the measuring equipment was
installed.

Figure 16 The geogrid reinforcement

Figure 18 Vertical displacement of original earth dam

Figure 19 Vertical displacement of earth dam rehabilitated with

MSE walls
Figure 17 The hot-dipped galvanized steel sheet pile

Table 4 Strength Reduction Factors (Rinter)

Interaction Rinter
Backfill / Rebar 0.60
Clay / Steel sheet pile 0.50
Backfill / Geogrid 1.00

6. RESULTS AND DISCUSSIONS

Figure 20 Vertical displacement of earth dam rehabilitated with
6.1 Vertical Displacement MSE walls
Settlement plates were installed in the embankment (Figure 8, Figure
9 and Figure 10) to measure the vertical settlements and the results
were verified with 2D FEA. Figure 18 shows the vertical
displacement contour of original earth dam obtained from Stress-
Seepage coupled analysis at the end of time step (2016) with
maximum vertical displacement of 0.70 m. The results from the
analysis are consistent with the geodetic monitoring data and FEA
results by (Soralump et al., 2023). Figure 19 and Figure 20 shows the
vertical displacement contour of dam crest rehabilitation with MSE
Walls and steel sheet pile after deactivating the original dam crest
elements and replacing with the new dam crest elements. In other
words, it was the settlement caused by the new dam crest. The
maximum vertical settlement was found to be 0.014 m at the top
surface of the backfill on the upstream side. Settlement of the
foundation of MSEW according to the location of the settlement plate
installed was 8 mm and 4 mm at the upstream and downstream sides,
respectively. The settlement pattern slightly tilted to the upstream Figure 21 Vertical displacement of foundation of new dam crest
side. However, settlement of the foundation obtained from the FEA (Settlement Plate)
was small compared to the height of the dam (59 m.) or even the
height of the MSEW (6m.) because the foundation of the MSEW

33
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

6.2 Lateral Deformation 6.3 Force and Strains in Reinforcement

The lateral deformation of the new dam crest with MSEW obtained The axial force in the reinforcement was measured using vibrating
from field measurements using inclinometers according to the height wire rebar strain meter bar. The comparison between FEA and axial
of the wall construction has been compared with the FEA results at force obtained from field measurement distribution along the rebar
the end of the construction. Figure 22a shows the net lateral length and are in good agreement (Figure 23). At the reinforcement
displacement obtained from upstream inclinometer (INC3) compared in layer 1 & 3 at downstream sides, the tensile force at the location
with the FEA results at the end of the construction. The lateral behind the gabion was low and had increased as it moved further
displacement obtained from field measurements and FE were away. The maximum measured tensile force was 1,086 kg at the
consistent; it has very lowest deformation and was bent to location behind the gabion facing 2.60 m. The tensile development
downstream side. The maximum measured lateral displacement was during the construction is as shown in Figure 24a. It shows that the
3 mm at depth of 7 m depth and the maximum lateral displacement tensile force increases with increasing wall height. Meanwhile, the
obtained from FEA was 2.7 mm at 4 m depth. Likewise, no lateral tensile force measured on the upstream side was very low (layer 1, 3
displacement occurred from 9.00 m depth. Figure 22b shows the net and 5) and the axial force in rebar reinforcement of layer 1 was
lateral displacement obtained from upstream inclinometer (INC4) compressive. The compaction of the clay between MSEW facing and
compared with the FEA results at the end of the construction. The steel sheet pile on the upstream side results in lateral earth pressure.
results from the field measurements showed a slightly bent to The tensile force developed during the construction is shown in
downstream side while the result from FEA was slightly bent to Figure 24b. The clay was not compacted at the beginning of the
upstream side. The maximum measured lateral displacement was 1.23 construction; therefore, the tensile force increases with the wall height
mm. and the result from FEA was 5 mm. Since various construction and is consistent at the downstream side. The tensile force decreases
activities (upstream soil compaction, temporary road construction and and is converted to the compressive force as the height of wall
its use) were being performed at the location where inclinometer was increases.
installed, the differences occur due to the construction phase However, the tensile stress in rebar from both field measurements
designation in the FEM. However, the measured and modeled lateral and FEA was low as compared to the yield strength of rebar. The
displacement was very low. tensile strain in the geogrid reinforcements was measured using
ultrahigh-elongation foil strain gauges. The comparison between FEA
and field measurements tensile strain distribution along the geogrid
length are in good agreement and is shown in Figure 25. The strain
measured in all the positions were very low. The maximum field
measured strain was 180 micro strain or 0.018% and the maximum
strain obtained from FEA was 350 micro strain or 0.035%, which was
very low as compared to the elongation at break of geogrid obtained
from laboratory tests, which was 3.49%. It shows that the geogrid
does not elongate according to the behavior of reinforced soil wall.

(a) (b)
Figure 22 Inclinometer reading and FEA results of test section
at the left bank (a) INC-3, Upstream (b) INC-4, Downstream

Figure 23 Axial force in rebar reinforcements

34
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

rehabilitation was 3.10. Figure 27 shows the operation of Mae Suai

Dam in year 2021 after the completion of rehabilitation process.

(a)

(a)

(b)
Figure 26 Total displacement of earth dam from the slope
stability analysis (SRM)

(b)
Figure 24 Axial force in rebar reinforcements with construction
stage (Time, Wall height (H)) (a) downstream side, D1 (b)
upstream side, U1

Figure 27 Mae Suai Dam in operation again since 2021.

7. CONCLUSIONS
A 6-m-high reinforced earth embankment for dam crest rehabilitation
was designed and constructed at Mae Suai dam, Chiangrai province,
Thailand. Welding mesh gabion was used as the facing on both sides;
the polymetric geogrid and rebar were used as reinforcements.
Furthermore, the MSE wall was placed on the original earth dam
crest, steel sheet pile was installed at upstream and downstream side
to prevent leakage and control the settlement of the original earth dam
crest. The instruments were installed at various test sections. The field
measurement data was verified with 2D FEM using MIDAS GTS for
Stress-seepage coupled analyses. Furthermore, slope stability
analysis was performed using strength reduction method (SRM) to
determine the F.S. of the dam crest rehabilitation. From the analysis
results, the factor of safety of the new dam crest was 3.10. The
Figure 25 Tensile Strain in Geogrid behavior of restraint back-to-back MSEW for dam crest rehabilitation
is as shown below:
6.4 Stability of MSE Wall for Dam Crest Rehabilitation - Vertical displacement
Stress-seepage coupled analysis in FEM has revealed that the The settlement of the foundation obtained from FEA was very low as
behavior of MSEW and new dam crest during construction and compared with the height of the dam or MSEW because the
operation are consistent with the field measurement data. In this foundation of the MSEW was an earth dam that used to support the
research, slope stability analysis was performed using strength weight of the old dam crest, which was heavier than the new dam
reduction method (SRM) which was the continuation of the Stress- crest. Likewise, the field measurement data also shows the settlement
seepage coupled analysis. Figure 26 shows the total displacement of behavior that was an immediate settlement caused by construction
earth dam from the slope stability analysis (SRM). The failure plane conditions.
occurs on downstream side along the filter passing through the bottom
of the sheet pile, the factors of safety (F.S.) of MSEW for dam crest
35
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

- Lateral deformation Benmebarek, S. and M. Djabri. (2017). “Fem to Investigate the

The lateral displacement obtained from field measurements and FE Effect of Overlapping-Reinforcement on the Performance of Back-
were consistent i.e., it has very low deformation. It shows that the to-Back Embankment Bridge Approaches under Self-Weight.”
good performance of sheet pile was able to control lateral deformation Transporta- tion Geotechnics, 11: 17-26.
which had occurred due to the new dam crest. Berg, R.R., B.R. Christopher and N.C. Samtani. (2009). “Design
- Force and strains in reinforcement and Construction of Mechanically Stabilized Earth Walls and
The measured and FEA in the rebar and geogrid reinforcement were Reinforced Soil Slopes.” Federal Highway Administration,
in good agreement. Tensile force in rebar on downstream side was Washington, D.C., USA.
higher than upstream side and the maximum tensile force was located Bergado, D.T. and C. Teerawattanasuk. (2008). “2d and 3d
at the bottom layer. The tensile force decreased with increase in Numerical Simulations of Reinforced Embankments on Soft
height. Meanwhile, the tensile force measured on the upstream side Ground.” Geotextiles and Geomembranes, 26: 39-55.
was very low; the axial force in rebar reinforcement of the bottom Bergado, D.T., C. Teerawattanasuk, S. Youwai and P. Voottipruex.
layer was a compressive caused by the lateral earth pressure of (2000). “Finite Element Modeling of Hexagonal Wire Reinforced
compacted clay on the upstream side. These measured forces were Embankment on Soft Clay.” Canadian Geotechnical Journal, 37:
found to be continually changing due to deformation of foundation, 1209-1226.
external stimuli and construction factors. Likewise, the strain Bergado, D.T., S. Youwai, C. Teerawattanasuk and P.
measured in all positions of geogrid reinforcement was very low Visudmedanukul. (2003). “The Interaction Mechanism and
which questions the performance of geogrid. Combining steel Behavior of Hexagonal Wire Mesh Reinforced Embankment with
reinforcement (high stiffness) with geogrid reinforcement (low Silty Sand Backfill on Soft Clay.” Computers and Geotechnics,
stiffness) was redundant. Most of the lateral stresses are resisted by 30: 517-534.
the former than the later as obtained in the results. It can be concluded Brinkgreeve, R.B.J. and R.F. Shen. (2011). “Structural Elements &
that in a reinforced soil wall that uses two or more types of reinforcing Modelling Excavations in Plaxis.” Power Point Presentation File,
materials, tensile force is developed in higher stiffness material. Delf, the Netherlands.
However, the tensile stress in rebar from both field measurements and BSI. (2010). “Bs 8006-1 : Code of Practice for Strengthened/
FEA was low compared to the yield strength of rebar. It can therefore Reinforced Soils and Other Fills.”
be concluded that the dam crest rehabilitation was a highly successful. El-Sherbiny, R., E. Ibrahim and A. Salem. (2013). “Stability of
Back-to-Back Mechanically Stabilized Earth Walls.” In Geo-
8. ACKNOWLEDGMENTS Congress 2013, 555-565.
Gikas, V. and M. Sakellariou. (2008). “Settlement Analysis of the
The authors would like to express their gratitude to the Geotechnical Mornos Earth Dam (Greece): Evidence from Numerical Modeling
Engineering Research and Development Center, Faculty of
and Geodetic Monitoring.” Engineering Structures, 30: 3074-
Engineering, Kasetsart University, Royal Irrigation Department, 3081.
Samart Engineering Consultants Company Limited and Kenber Hardianto Fransiscus, S., R. Lozano, E. Sankey John and K. Hughes
Geotechnic (Thailand) Company Limited for providing technical David. (2013). “Geosynthetic-Strip Reinforced MSE Wall for Dam
support throughout this research. Expansion.” Geo-Congress 2013, 543-554.
Ho, S.K. and R.K. Rowe. (1996). “Effect of Wall Geometry on the
9. REFERENCES Behaviour of Reinforced Soil Walls.” Geotextiles and
AASHTO. (2012). “Lrfd Bridge Design Specifications.” 6th Geomembranes, 14: 521-541.
edition. Washington, DC. American Association of State Highway Holtz, R.D. and W.F. Lee. (2002). “Internal Stability Analyses of
and Transportation Officials. Geosynthetic Reinforced Retaining Walls.”
Allen, T.M., R.J. Bathurst, R.D. Holtz, W.F. Lee and D. Walters. Huang, B., J. Bathurst Richard and K. Hatami. (2009). “Numerical
(2004). “New Method for Prediction of Loads in Steel Reinforced Study of Reinforced Soil Segmental Walls using Three Different
Soil Walls.” Journal of Geotechnical and Geoenvironmental Constitutive Soil Models.” Journal of Geotechnical and
Engineering. 130: 1109-1120. Geoenvironmental Engineering, 135: 1486-1498.
Allen, T.M., R.J. Bathurst, R.D. Holtz, D. Walters and W.F. Lee. Huang, B., R.J. Bathurst, K. Hatami and T.M. Allen. (2010).
(2003). “A New Working Stress Method for Prediction of “Influence of Toe Restraint on Reinforced Soil Segmental Walls.”
Reinforcement Loads in Geosynthetic Walls.” Canadian Canadian Geotechnical Journal, 47: 885-904.
Geotechnical Journal. 40: 976-994. Kim, D., S.K. Bhowmik and J.L. Willmer. (2012). “A Case History
Baral, P., D.T. Bergado and S. Duangkhae. (2016). “The Use of of MSE Wall Failure: Finite Element Modeling and Evaluation. ”
Polymeric and Metallic Geogrid on a Full-Scale MSE GeoFlorida 2010, 2232-2242.
Wall/Embankment on Hard Foundation: A Comparison of Field Kongkitkul, W., F. Tatsuoka and D. Hirakawa. (2007). “Effects of
Data with Simulation.” International Journal of Geo-Engineering. Reinforcement Type and Loading History on the Deformation of
7: Reinforced Sand in Plane Strain Compression.” Soils and
Bathurst, R.J., Y. Miyata, A. Nernheim and A.M. Allen. (2008). Foundations, 47: 395-414.
“Refinement of K-Stiffness Method for Geosynthetic-Reinforced Lajevardi, S.H., K. Malekmohammadi and D. Dias. (2021).
Soil Walls.” Geosynthetics International. 15: 269-295. “Numerical Study of the Behavior of Back-to-Back Mechanically
Bathurst, R.J., A. Nernheim and T.M. Allen. (2009). “Predicted Stabilized Earth Walls.” Geotechnics, 1: 18-37.
Loads in Steel Reinforced Soil Walls using the AASHTO Leshchinsky, D. (2009). “On Global Equilibrium in Design of
Simplified Method.” Journal of Geotechnical and Geoenviron- Geosynthetic Reinforced Walls.” Journal of Geotechnical and
mental Engineering. 135: 177-184. Geoenvironmental Engineering, 135: 309-315.
Bathurst, R.J., N. Vlachopoulos, D.L. Walters, P.G. Burgess and T. Midasgts. (2018). “Midas GTS Reference Guide (Part 1, Part 2, Part
M. Allen. (2006). “The Influence of Facing Stiffness on the 3 & Part 4).”
Performance of Two Geosynthetic Reinforced Soil Retaining Miyata, Y., R.J. Bathurst and H. Miyatake. (2015). “Performance of
Walls.” Canadian Geotechnical Journal. 43: 1225-1237. Three Geogrid-Reinforced Soil Walls Before and After Foundation
Benmebarek, S., S. Attallaoui and N. Benmebarek. (2016). Failure.” Geosynthetics International, 22: 311-326.
“Interaction Analysis of Back-to-Back Mechanically Stabilized RID. (1998). “Final Design Report : Mai Suai Dam Project.”
Earth Walls.” Journal of Rock Mechanics and Geotechnical Rowe, R.K. and S.K. Ho. (1998). “Horizontal Deformation in
Engineering, 8: 697-702. Reinforced Soil Walls.” Canadian Geotechnical Journal, 35: 312-
327.
36
 Geotechnical Engineering Journal of the SEAGS & AGSSEA Vol. 54 No. 3 September 2023 ISSN 0046-5828

Samee, A.A., M. Yazdandoust and A. Ghalandarzadeh. (2021).

“Performance of Back-to-Back MSE Walls Reinforced with Steel
Strips under Seismic Conditions.” Transportation Geotechnics, 30:
Shivashankar, R. (1991). “Behavior of Mechanically Stabilized Earth
(MSE) Embankment with Poor Quality Backfills on Soft Clay
Deposits, Including a Study of the Pullout Resistances.” D.Eng,
Asian Institute of Technology, Bangkok, Thailand.
Soralump, S., C. Thongthamchart, M. Jinagoolwipat and A. Boonpo.
(2016). “Rehabilitation of Leakage and Seismic Damaged Problem
of Mae Suai Earth Zone Composited RCC Dam.” In. Kuala-
Lumpur, Malaysia.
Soralump, S., K. Wannasiri and S. Prempramote. (2023). “Geometry
and Roughness Effect at the RCC Interface-Earth Dam Causing
Leakage in Mae Suai Composite Dam, Thailand.” Geotechnical
Engineering Journal of the SEAGS & AGSSEA, Vol. 54, 23-35.
13p.
Voottipruex, P. (2000). “Interaction of Hexagonal Wire Reinforce-
ment with Silty Sand Backfill Soil and Behavior of Full-Scale
Embankment Reinforced with Hexagonal Wire Mesh.” Ph.D.,
Asian Institute of Technology, Bangkok, Thailand.
Watanbe, K., Y. Munaf, J. Koseki, M. Tateyama and K. Kojima.
(2003). “Behaviors of Several Types of Model Retaining Walls
Subjected to Irregular Excitation.” Soils and Foundations, 43: 13-
27.
Xu, P., G. Yang, T. Li and K. Hatami. (2021). “Finite Element Limit
Analysis of Bearing Capacity of Footing on Back-to-Back Re-
inforced Soil Retaining Walls.” Transportation Geotechnics. 30:
Yang, G., Y. Zhao, H. Wang and Z. Wang. (2022). “Analysis of the
Working Performance of a Back-to-Back Geosynthetic-Reinforced
Soil Wall. ” Applied Sciences. 12:

37