Sandwich wall beneath Amsterdam Centraal ... - Royal Haskoning

North/South Line Amsterdam Date: 28-11-2005
Sandwich wall beneath Amsterdam Centraal Station
An innovative approach for jet grouting under difficult conditions
Directors of the Steering Group:
J. de Wit , J. Bogaards : Adviesbureau Noord/Zuidlijn Amsterdam (‘Amsterdam North/South Line
Consultants’), Royal Haskoning, Rotterdam
R.D. Essler, J. Maertens : Consultants
O. Langhorst : VOF Stationseiland Amsterdam, Holland Railconsult, Utrecht
B. Obladen, C. Bosma : Principal Contractor, CSO, Combinatie Strukton Betonbouw van Oord ACZ
Y. Sleuwaegen, H. Dekker : Jet grouting Subcontractor, Smet Keller
Outline
An excavation, within which a tunnel is sunk as a part of the Centraal Station underground station (platform
section) of the Noord/Zuidlijn (‘North/South Line’), is to be created beneath the Amsterdam Centraal
(‘Amsterdam Central’) station. The section beneath Amsterdam Centraal Station is characterised by the
application of a particular technology in the form of, inter alia, the so-called ‘sandwich wall’. This is a
composite wall consisting of two rows of Tubex piles with a body of jetgrout columns in-between. This wall acts
both as an excavation support wall and also provides vertical bearing. The installation of the wall, including both
Tubex piles and jetgrout columns, within these specific conditions (limited height, sensitive historical building),
within the design requirements set in terms of construction tolerance and water and soil retention, may be
regarded as being a pioneering achievement.
The environmental constraints are that the trains should continue to run, the inconvenience to passengers should
be kept to a minimum and the historic station should not suffer any damage.
Because of the complexity of the working conditions, the project is being overseen by a steering group of experts
in the specified construction processes. In order to manage the project risks, jet grouting trials were initially
carried out, in order to highlight and control execution problems; the practicability of the design is then tested.
Work commenced on the sandwich wall in 2003, when the wooden piles were extracted at the locations where
the sandwich wall was to be constructed; in 2004, the steel Tubex piles for the southern wall sections were
installed.
1

From May 2005, the construction of the sandwich wall got under way; in addition to the described process
supervision, an extensive measuring programme was established, so that the process could be adjusted as
necessary and the quality of the finished product determined. This applied to both individual columns and for the
end product, the sandwich wall. The measurement results were processed for each column, interpreted and the
effect on the work still to be conducted was assessed for each future column. Preventative and corrective
measures were, if necessary, taken in order to control or counteract the measured defect.
In addition to the process-oriented and qualitative measurements, various independent measurement systems at
the station building served as a further control in order to prevent damage to the building.
Introduction
The new ‘Noord/Zuidlijn’ underground link connects
Amsterdam Noord (‘Amsterdam North’) and
Amsterdam Zuid/WTC (‘Amsterdam South/WTC’)
to the city centre. The photo shows the location of the
new underground line under the IJ waterway and the
island station. The new underground station is shown
to be constructed beneath the Amsterdam Centraal
Station train station.
The station building was built around 1880 on an
island dredged beforehand in the IJ waterway. The
building is supported on some 9,000 wooden piles
and there has been approximately 18 cm subsidence
over 100 years.
It is a fundamental requirement for the construction
of the underground link that, during the building
work on the tunnel beneath the station, the trains
should continue to run and the inconvenience to
passengers should be kept to a minimum. The
renovation is being carried out over a relatively small
footprint within the largest interchange station in the
Netherlands and the project is accordingly highly
complex.
The tunnel will be constructed beneath the station building in various construction phases, as illustrated in the
following figure.
Construction phases:
1 Extract existing wooden piles and install Tubex
piles
2 Install sandwich wall
3 Carry out station building support works
4 Lower ground water level
5 Excavate in the dry
6 Install low level grouted strut
7 Install high level steel strut frame
8 Raise water level
9 Excavate in the wet
10 Float in and sink tunnel
11 Backfill with soil
In addition to the construction of the underground station, other projects are at the same time under way on the
station island, such as extending the platforms on the west side, an underpass for vehicle traffic, a new bus station
with a roof covering all of the Ruiterkade and renovation of the public area and bridges on the IJ side of the
station. It is expected that the entire reconstruction will be completed in 2010. The following figure provides an
impression.
2

Overall soil description
The soil conditions from ground level (+3m NAP [Normal Amsterdam Water Level]) consist of an approx. 8 m
thick layer of dredged sand (down to NAP -5 m) overlying the relatively weak ‘IJ clay’ containing sand bands to
approx. –15 m NAP. Below this layer down to approx. –28 m NAP to –29 m NAP, is a medium dense, locally
dense, 2 nd sand layer, consisting of slightly silty sand and clayey sand. Beneath the level from approx. –28 m
NAP to –29 m NAP, there is the ‘Eem clay’ layer down to approx. –45m NAP. At a level of –45 m NAP, is the
approx. 1 m thick layer of ‘Harting’, which is a relatively thin peat layer that may contain methane gas. From
this level down to approx. –56m NAP, there is a ‘glacial clay’ layer. The third sand layer, which has a high cone
resistance, is found from –56 m NAP; this layer has high support capacity, at least down to the investigated
depth of –70 m NAP. The highest ground water level measured is –0.25 m NAP within the dredged sands,
approx. -1.50m NAP in the second sand layer and in the third sand layer NAP –3.00 m. The geotechnical profile
is shown in the following figure.
The design of the sandwich wall
The sandwich wall that was developed for this element of the project was adapted to the specific conditions
under the Amsterdam Centraal Station. The protection of the station building, as a national monument, was, in
particular, the main priority during the design development.
The design of the sandwich wall is based on
a robust, stiff wall, which is constructed
from smaller components. As a result, the
equipment required is relatively light and
can be used within the station building
without excessive modification to the
building structure.
Geometrical aspects
The sandwich wall consists of two rows of
Tubex piles with a diameter of 457 mm, a
wall thickness of 25/16 mm and a length of
26 m to 60 m. The piles are approx. 1 m
centre to centre and the pile rows are
approx. 2.5 m centre to centre. The Tubex
piles are provided with rings (32 mm) in
order to ensure full mobilisation of shear
strength within the grout body. The piles are
installed in short sections (2 – 5 m) using a
Tubex machine with an extension (Topdrill)
specifically designed for this purpose. To
increase the speed of the work a special
screw coupling for connecting the pile
sections was developed.
The photo shows Tubex piles being installed
using the Topdrill.
3

The space between the piles is filled by jetgrout columns having a diameter from 800 mm to 1200 mm and a
length of 28.5 m. The space between the pile rows is filled by two rows of jetgrout columns with diameters from
1400 mm to 2200 mm and lengths from 26.0 to 28.5 m (the row on the outside of the excavation has a length of
26 m). The remaining wooden piles on site were extracted while backfilling the hole with sand.
Constructional aspects
The retained height of the wall is approx. 18 m, the water level difference across the wall being 3 m. The
maximum water level difference is approx. 5 m during the phase when the grouted strut is installed from a level
of approx. –4.5m NAP.
The horizontal stability of the wall is
provided by the structural interaction
between the steel Tubex piles and the
grout body, supplemented with temporary
struts at three levels. This results in an
extremely stiff wall, with predicted
deformations being of the order of 20 to
30 mm.
The sandwich wall also transfers the
vertical loads resulting from supporting
the foundations of the building at the
location of the construction pit, via a
number of Tubex piles in the Sandwich
wall that extend to the 3rd sand layer.
The selected wall system is extremely stiff and is based on the assumed interaction between the Tubex piles and
the jetgrout. However, it has the disadvantage of having a relatively brittle failure behaviour. The safety factors
used have therefore been increased considerably, with as an in-built safety factor, the fact that resistance to
failure is guaranteed by the strength of the two rows of Tubex piles (the horizontal deflection does however
increase considerably).
Additional support has been provided directly adjacent to the excavation walls in order to protect the station
building. These supports are intended to compensate the reduced capacity of the present foundations of the
station building resulting from the construction of the excavation walls.
4

Seepage and soil retention of the wall system
The design prerequisites can be summarised as the resistance to seepage, soil retention and a construction
process during which the trains continue to run and inconvenience is kept to a minimum. The realisation of these
conditions of the design resulted in the sandwich wall.
Practical experience has shown that the water-tightness of a jet grout body consisting of columns is particularly
sensitive to variations in construction tolerances. The coordination of the design preconditions and the
construction is very important in this regard. The following figure shows the elements of the sandwich wall.
Jetgrout columns
When designing the sandwich wall, the aspects mentioned above were taken into account as far as possible,
which should result in a wall construction that resists water seepage and retains ground with a limitation that the
difference in water level across the wall will be a maximum of 5 m during the phase when the grouted prop is
being constructed and approximately 3 m in the phase when the construction pit is being excavated. Four rows of
jetgrout columns will be used, with the outer row being constructed between the Tubex piles.
Risk profile
Extensive risk analyses were carried out for this particular wall during the design phase. Basically, without going
into the detail, the risks may be reduced to three main groups, which relate to the structional function, the
resistance to seepage function and the soil-tightness. The following table summarises the design risk profile of
the sandwich wall.
Risk/failure Direct consequence Cause
Constructional Large deformations of the wall
(collapsing, jetgrout wall not in order, only Tubex
piles have sufficient strength)
Seepage control Reduction in water level outside excavation
Water level inside excavation uncontrollable
Soil retention Reduction in water level outside excavation
Water level inside excavation uncontrollable
Soil movement
Strength of jet grout too low
Insufficient depth of Tubex piles
Lack of overlap within jetgrout (shadowing
effect/misdrilling)
Tubex piles with
shear rings
Insufficient overlap between Tubex piles and jet grout
owing to:
– diameters of jet grout columns too small
– excessive vertical deviation of jet grout columns
– excessive vertical deviation of Tubex piles
– obstructions in subsoil
Insufficient overlap between grout columns owing to:
– diameters of jet grout columns too small
– diameters of jet grout columns too large (misdrilling)
– shadowing effect caused by remaining wooden piles
– excessive vertical deviation of jet grout columns
– obstructions in subsoil
See causes under ‘seepage control’
A failure of the wall system may indirectly cause damage to the station building of Amsterdam Centraal Station.
It is therefore very important to limit and control the risks as far as possible. As part of a major study into soil
5

improvement technologies in preparation for the Noord/Zuidlijn, jetgrouting was examined in detail through a jet
grouting trial carried out in Amsterdam Noord. A number of columns were produced for the design of the
sandwich wall; the achievable final strengths and the minimum diameters to be produced in the Amsterdam soils
were, in particular, investigated. This trial provided sufficient confirmation for viability of the design. As a
result, the design requirements were established.
The design assumed a vertical deviation, measured over the full length of the Tubex pile, of at most 0.5 % of the
depth from ground level. For jet grouting, the vertical deviation had to be less than or equal to 200 mm over the
uppermost 20 m (1 %) and 300 mm over the total length. The variation of the actual diameter relative to the
theoretical diameter has to be less than 15 % of the diameter for diameters up to 1,000 mm, and less than 10 % of
the diameter for diameters greater than 1,500 mm (linearly interpolated percentages apply for intermediate
diameters). The characteristic (design) value of the compressive strength was determined as 1.75 N/mm² at 28
days.
From the jet grout trial to the final work
Once the work had been approved, a second jet grouting trial was carried out at the site (Voorplein). The
objective of the second jet grouting trial on the Voorplein (Station Square) was to enable the subcontractor to
adjust the process on site before starting on the sandwich wall, to confirm his expertise in the field of jet grouting
and to understand the risks more thoroughly. During the trial, measurements of the diameter, variations in
diameter using both borehole callipers and hydrophones, the verticality and strength of the jet grout columns
were carried out.
The results of the trial on site proved disappointing. The control of the diameter and its variation was found to be
insufficient and the strength only just adequate. The nature of the soil on site was found to be less predictable
than at the test location in Amsterdam Noord and this was regarded as the most important cause of the problem.
However, it was possible, on the basis of the information obtained, to make decisions for carrying out the final
work.
The jet grouting trial raised a number of points of interest in terms of technical content (design and construction)
and procedure for the final work.
The technical points were as follows:
• Separation of the ‘diameter’ and ‘strength’ construction phases. The pre-cutting phase created the full
diameter, and the post-jetting phase then provided sufficient binding agents for the strength.
• Formation of columns in sections (from 5 to 10 m in length) in order to limit the time effects, in other
words phasing of the column over the depth from ground level to –28.5 m NAP.
• Controlling the strength as a result of improved quality control of the mixes and densities on site;
• Adjusting the composition of the binder more effectively to the highly heterogeneous soil structure on site.
• Adapting strength requirements to the actual project situation, i.e. age of strength and sensitivity calculation
of missing grout volumes. The characteristic value of the compressive strength of the jetgrout at 120 days
was to be 1.5 N/mm² and the characteristic value of the splitting tensile strength of the jetgrout at 120 days
0.15 N/mm².
• Providing more latitude in the design, allowing greater deviation in diameter. The design was adapted to
allow a variation of +/- 20 %, a basic margin that all of the experts involved believed to be feasible. The
actual position of the already installed Tubex piles and remaining (parts of) wooden piles would be taken
into account in the design of the column model.
The work-oriented alterations were as follows:
• Setting up an extensive measurement programme for the jetgrouting process (process instrumentation,
diameter measurements using borehole callipers and hydrophone measurements, strength measurements);
• For the model, taking into account ‘as-built’ information (3-D measuring of Tubex piles, remaining wooden
pile (parts), successive production of jet grout columns). This required a highly focussed approach and
supervision ( observational method), in which the design and/or implementation parameters had, if
necessary, to be adjusted.
• Expansion of the organisation during the jet grouting by means of:
o establishing a steering group/implementation monitoring team
o having an inspector on site who carried out real-time checks in order to eliminate human error as far
as possible, thus doubling control of process control
o processing ‘as-built’ data on site.
6

Management of final works operations
General
When adopting the observational method, information was obtained from work completed and used to adjust
work still to be carried out. Prior to starting the jet grouting process, this meant that the basis of the jet grout
model was not only adapted to the actual position (and deviation) of the Tubex piles that had already been
installed, but also to the wooden piles or pile sections that could not be extracted. When carrying out the jet
grouting of the sandwich wall, the focus was on both the individual products (the separate columns) and the final
product (the wall system). For both processes, measurements were carried out and options for verification and
adaptation provided. Maximum possible control on site was thus obtained in each phase and in each component.
The following figure is a plan view indicating the Tubex piles in red, the jet grout perimeter columns in blue, and
the jet grout infill columns in yellow/brown.
The basic premise of the methodology applied is that if a condition of a particular component of the wall satisfies
the requirements set, the chance of failure of the wall system (end product) is minimal. In order to meet this
condition, it is important that all of the individual columns are produced with the utmost precision, whereby
every column is treated individually. A work plan is accordingly drawn up for every column. This plan may be
regarded as a script for the production and contains all of the relevant information for the implementation, such
as design data, the jet grouting parameter plan, measuring regime, risk table and implementation protocols
(method for dealing with deviations observed during implementation). The production of the column may be
based very closely on the parameters set (on the machine) or, in the event of deviations, on the established
implementation procedures that control the redesign of each column construction depending on information
available at the time..
The best picture possible must be gained of the diameter and strength of the constructed columns by means of
the recorded machine data (lift speeds, rotational speeds, pump pressures and outputs, etc.), measurements
carried out (inclination measurements, diameter measurements using borehole caliper and hydrophones, spoil
densities, etc.). The logbooks of the works also list important events and the records of additional changes
sanctioned by the inspectors. The result is processed in the ‘as-built’ layout to determine the actual quality, in
order thus to be able to assess whether adaptations are required to the design and/or implementation and, if so, to
provide proposals to do so (verification and adaptation).
The following figure provides an impression as to how the ‘as-built’ information is processed in diagrammatic
form for the columns still to be produced. The position and the dimensions of the jetgrout column are
represented for each level (NAP, -5, -10, -15, -20, -25 m).
7

Implementation
Determining the various jet grout parameters proved to be a difficult task for this project, in view of the complex
preconditions. The following factors had to be taken into account:
• The significant depth of jetting;
• The highly heterogeneous and variable nature of the soil, both in its depth and across the site; the Tubex
piles present and remaining wooden piles or pile sections further complicating factors in this regard.
• The various specified diameters, including the requirement that it should be possible to correct some of the
deviations from the verticality during drilling by adjusting the column diameter during the production
process;
• The introduction of the necessary quantity of cement to obtain the desired strength;
• The provision of limited tolerances in the diameter (< 20 %) of the column in order to prevent a shadowing
effect in the production of the neighbouring columns;
• The determination of the correct fluid density of the jetgrout slurry used during pre- and post-cutting in
order to ensure smooth discharge of the return liquids and to prevent blockages;
• The use of as little air as possible in order to prevent local variations in ground water pressure (in order to
limit the effect on the building to a minimum).
The mono-jet system was used for the smaller diameters from 800 mm to 1200 mm, and the bi-jet system for the
diameters from 1400 mm to 2200 mm. The columns, on the other hand, were produced in two operations,
namely pre- and post-cutting. In the case of pre-cutting, the entire diameter was cut using a low density grout
slurry, whereas in the case of post-cutting, the column was homogenised using a higher density grout slurry, the
required cement content of the column thus being achieved. The important thing, in this case, was to produce the
column in sections, so as little time as possible elapsed between the two phases, as the trials revealed that the end
result was influenced favourably using this method. In order to accommodate the highly variable nature of the
soil, individual layer divisions and associated parameter sets were established for each layer. At some locations,
up to seven sets of parameters had to be defined for one specific diameter. Everything was set out in advance for
the engineers in working plans and protocols were also provided for the deviations that could be expected.
Adapted parameters were provided in these protocols for these divergent situations. The agreed parameters were
tested in two further trial phases and during the works and, if necessary, adjusted in response to the numerous
measurements. For the purposes of control, diameter calculations were also carried out on the basis of the density
of the return liquid.
The following figure illustrated the specific kinetic energy of each soil layer and the division of the parameter
sets.
Energy Level
7
6
5
4
3
2
1
0
0
S308
-2
-4
-6
-8
Specific Kin_Energy per groundlayer & Parameter sets
-10
-12
Jet Parameter Sets
-14
-16
Depth in N.A.P. [-m]
-18
-20
-22
-24
-26
Energy Level S308
-28
14.00
12.00
10.00
8.00
6.00
4.00
2.00
0.00
-30
Cone Resistance [MPa]
8

Assessment of the columns
Each column was intensively assessed immediately after implementation; this was necessary in order to be able
to determine whether work could continue on the basis of the jet grout model or whether corrective measures
were necessary. If a column met the design requirements, adjustments were not necessary and work could
continue on the basis of the grout model. If the column did not meet the design requirements, corrective
measures could be necessary. Corrective measures were taken from within the design tolerances or the works
process. In the case of incidental or local deviations, the corrective measure was usually initiated from within
the design process. For a corrective measure, a new work plan was drawn up and adaptation of the column
layout, the addition of an extra column or directed injection behind the wall was considered. Adjustment could
also take place from the implementation process by adapting jet grout parameters on the basis of the information
obtained to obtain larger diameters for example.
The assessment related to the quality of the produced column. The analysis consisted of various components,
namely:
• Monitoring the work process; producing the recorded machine data and monitoring it using the previously
established work plan. The ‘as-built’ parameters are thus determined and a check is made as to whether
deviations have occurred in the implementation. An inspector, who monitors all of the settings of the jet
grout equipment during implementation on a separate display unit, is appointed in order to prevent human
error in the process.
• Analysis of the column log, drawn up by the inspector. This indicates the particular details of the
implementation process.
• Analysis of the measurement data; the density of the return slurry, which is measured during the process,
provides an indication of the amount of excavated soil, and therefore an estimation of the diameter.
• Monitoring and calibration of the applied jet grout parameters by comparing the amount of energy
introduced for each column volume relative to the actual column volume.
• Analysis of the borehole calliper measurements with which the diameter is measured.
• Analysis of the hydrophone measurements in Tubex piles and specially installed tubes. This allows a
diameter indication to be provided and/or allows it to be established whether the grout jet has reached the
Tubex piles (indirectly determining whether there is a connection between the pile and the grout body
produced).
• Monitoring the compressive strengths of the return slurry from the column (the test results become available
at a later stage and do not influence the model). Nevertheless, it is important that the results are available on
time in order to be able to carry out any adjustments to the amount of binding agents.
For the purposes of the analysis, the density of return, adapted jet grout parameters, borehole calliper
measurements and compressive strengths are summarised in graphical form and plotted with depth. This creates
an overview plotted against depth, which can be used to make a visual comparison with the strata, based on the
additional information from the soil drilling tests.
Assessment of the wall
The monitoring of the end product is oriented more to the functioning of the wall as a system. This means that
the monitoring emphasis and associated measurements are oriented more at the seepage, soil-retention and
strength of the wall. This means that these aspects were tested as soon as possible in order to make any necessary
adjustments within the working process. In this case, the information that became available during the
production of the individual columns was used. Leakage detection measurements were then carried out in order
to trace any deficiencies in the wall.
The figure on page 8 [sic] is a sectional view of a wall of the first row of completed peripheral columns; the
results of leakage detection measurements, hydrophone measurements and geometrical measurements provide an
impression of possible imperfections, which are to be filled with the rear infilling columns. The possible
imperfections in which, at the minimum and maximum diameters, there is no overlap with the Tubex piles are
indicated in red, those which at minimum diameter there is no overlap and at maximum just some overlap in
orange, and those which at minimum diameter there is no overlap and at maximum diameter there is overlap in
light brown. Texplor measurements indicate the most probable leakage zone in the dark blue zone, light blue is a
leakage risk zone, purple a low risk and yellow a very low risk. The results of the hydrophone measurements (Gtec)
are indicated in terms of a weak jet impact (-J) and no jet impact (N).
9

B e oord elin g jetgro utrand k olom m en S -ZW R K 2 50 t/m 4 80 versie 14-9-200 5
t .o . v . N A P
T u b e x 6 7
R K 2 5 0
T u b e x 7 3
R K 2 6 0 / 2 6 1
T u b e x 7 4
R K 2 7 0
T u b e x 7 5
R K 2 8 0
T u b e x 7 6
R K 2 9 0
T u b e x 6 8
0 [m ] - -J N - - - -
- -J N - - - -
-1 [m ] - -J N - - - -
- -J N - - - -
-2 [m ] - -J N - - - -
- N N - - - -
-3 [m ] - N N - - - -
- -J N - - - -
-4 [m ] -
-
-J N
N
A
-
-
-
-
B
-
C ? -
-
-
-5 [m ] - N - - - -
- -J - - - -
-6 [m ] - N - - - -
R K 3 0 0 / 3 0 1
T u b e x 7 7
- -J - - - -
-7 [m ] - - - - -
- - - - - N
-8 [m ] - N - - - -
- N - - - -
-9 [m ] - N - - -J - -
- N - - - -
-10 [m ] - N - - - -
- -J - - - -
-11 [m ] - N - - - -
-12 [m ]
-
-
A -
-
-
-
B
C ?
-
-
-
-
D
- -J - - - -
-13 [m ] - -J - - - - -J
- -J - - - -
R K 3 1 0
T u b e x 7 8
R K 3 2 0
T u b e x 7 9
-14 [m ] - -J - - - -
- -J - - - - N N
-15 [m ] - N - - - - -J
- N - - - - -J
-16 [m ] - J - - - - -J
- -J - - -J - - -J
-17 [m ] - N - - -J - - -J
- -J - - -J - - -J N N
-18 [m ] - -J - - -J - - -J N N
- -J - - -J - -
-19 [m ] - -J - - - - N
- -J -J - - - - N
-20 [m ] - -J -J - - -J - - N N
- N - - - - N N
-21 [m ] - N - - - - N N N
-22 [m ]
-
-
-J
N
A -
-
-
-
B -J -
-
-
-
D
S 2
-J -J
- N -J - - - - -J -J
-23 [m ] - N -J - - - - -J -J
- -J -J - - - - -J
-24 [m ] - -J - - -J - - -J -J
- -J - - -J - -
-25 [m ] - -J -J - - -J - - -J
- -J - - -J - - -J
-26 [m ] - -J - - -J - - -J
- -J -J - - -J - - -J
-27 [m ] - - - -J - - -J -J -J
- -J J - - -J - - N -J -J
-28 [m ] - -J - - - - N -J -J
- N - - - - N N N -J -J
-29 [m ]
LE G E ND A
G-tec -J : Zw akke a anstraling ge m e ten do or G -te c
N : G ee n aa nstra ling gem eten doo r G -tec
- : G ee n m eettesulta te n H ydrofo ne n besch ikba ar
m in /m a x dia m eter O p b a sis va n g erea lisee rd grou tp atroo n m e t m in im ale / m axim a le diam eters g ro u tkolo m m e n:
: B ij m inim ale d iam ete r ge en o verlap en bij m a xim ale diam eter w e l ove rlap m et tub e x-p aal.
: B ij m inim ale d iam ete r ge en o verlap en bij m a xm ale d iam eter net a an overla p m et tubex-p a al.
: B ij m inim ale e n m axim ale dia m e ter g ee n ove rla p m et tube x-p aal.
Tex plor : R isico-zo ne va n lekk age op bas is van E FT-m etin gen
(2 -9 -2 00 5 ) : M ees t wa a rschijnlijke zone va n lekka ge op b asis van E F T -m eting en
R K 3 3 0
Seepage and soil retention tested using leakage detection systems
In this context, leakage detection measurements using the EFT system from Texplor and pump tests were carried
out. The zone in which a possible imperfection is located could be identified reasonably accurately by means of
leakage detection measurements. The pump tests primarily provide information regarding the degree of seepage
or non-water tightness.
The EFT (electro-flux tracking) system is used for leakage detection measurements. In this system, a controlled
and defined electrical signal is introduced into the ground water on the outside of the sandwich wall and is
detected on the other side of the construction by means of a sensor. If there are gaps in the watertight
construction, the electrical energy will be concentrated through them. An increased electrical potential relative to
the watertight sections was measured on site. Two types of measurements were carried out. In the first type of
measurement, an electrical signal was introduced into each of the three determined soil layers. The sensors were
a grid at ground level on the other side of the sandwich wall. The locations at which there was an increased
chance of a potential leakage path were thus determined for each soil layer. The results of the measurements
were repoted for each soil layer by means of isolines (see figure on page 11).
The second type of measurement was a supplementary measurement to the first type. In this case, the electrical
signal is introduced at a previously defined level for each soil layer. On the other side of the sandwich wall, the
sensor is suspended at the same level. This measuring method is intended, as a supplement to the first
measurements, to introduce a refinement at the level of the locations having an increased risk of imperfections.
The above-mentioned measurements are carried out for each wall section at the following points in time:
1. Once the row of jet grout columns located between the Tubex piles (on the inside of the sandwich wall) is
complete.
2. Once the entire wall section of the sandwich wall is complete.
T u b e x 8 0
R K 3 4 0
T u b e x 8 1
R K 3 5 0 /3 5 1
T u b e x 6 9
R K 3 6 0
T u b e x 8 2
: Zo ne m et zw a kke vo rm van lekkage o p b asis va n E F T m etingen M eetm e thode B b eho e ft toe lichting
: Zo ne m et ze er zw akke vorm van le kkage o p b asis va n c ontourlijn en E F T-m etinge n
: K olo m in tw ee d elen gem aa kt
R K 3 7 0
T u b e x 8 3
R K 3 8 0 / 3 8 1
T u b e x 8 4
R K 3 9 0
T u b e x 8 5
R K 4 0 0
T u b e x 8 6
R K 4 1 0
T u b e x 8 7
R K 4 2 0
10

Up until now, the measurements specified under 1) have been carried out for the wall section on both the east
side and the west side. The first type of measurements provided an image that was reasonably compatible with
the data that was obtained from the hydrophone and geometrical measurements. The supplementary
measurement, for distinguishing the depth location of the places with an increased risk of imperfections, has in
the past had only limited additional value.
Once the entire wall has been completed, pump tests will then be carried out within the excavation. If, at that
stage, there are still substantial imperfections in the wall, it is is still possible to take back-up measures, for
example local freezing, further limiting of the water level differences along the wall, etc.
Assessment of the thickness
The strength of the wall is monitored by taking numerous samples (return spoil samples from each column, core
drillings per number of columns). The strength is determined from these samples by means of load tests. If the
set grout does not satisfy the design requirements as far as strength is concerned, back-up measures may be
taken.
Once the test excavation had been performed, it was found that the strength of the set grout fluctuated over the
depth and barely met the required strength. A binding agent test was therefore initiated in order to optimise the
grout mix. The highly diverse soil structure was found to have a marked influence on the strength development
of the binding agent. In order to be able to determine this influence, two soil drillings were carried out at the site
of the sandwich wall on the Voorplein. In the material supplier’s laboratory, the soil samples from two different
depths were selected and mixed with the binding agent in the same proportions as on site. In order to optimise
the binding agent, four different compositions were tested and mixed with soil. A neat cement mix was used as
a reference mix. Once the binding agent had been mixed with soil, 72 test samples were taken and stored in a
conditioned manner, in order to be tested after 1, 7, 28 and 90 days. On the basis of these results, it was decided
to use mix BM-2 (see figure on page 12) on site.
11

Compression Strength [MPa]
Compression Strength [MPa]
30.00
25.00
20.00
15.00
10.00
5.00
0.00
24
22
20
18
16
14
12
10
8
6
4
2
Compression Strength development Grout mixes Blanco and mixed with Ground
0 7 14 21 28 35 42 49 56 63 70 77 84 91
Curing Time [Days]
Binder-1 Blanco Binder-1 Ground Binder-2 Blanco
Binder-2 Ground Binder-3 Blanco Binder-3 Ground
Influence Ground on Compression Strength
0
0 5 10 15 20 25 30
Depth [-m NAP]
Conus Friction CompR CoreDrilling
Various core drillings were carried out in order to determine the compressive strength of the columns on site.
These clearly revealed that the soil structure has a significant influence on the strength development (see the
following figure). In view of the fact that the grout strength is an important parameter in the behaviour of the
wall as a whole, efforts were made during the works to maintain the injected quantity of binder per m3 to as
constant a value as possible
Both the compressive strength of the spoil return and the core drilling provide roughly the same picture: a
reasonable strength in the dredged sand at approx. –3 m NAP, then a reduction in the IJ clay. The clay is
followed by a sandy layer, as a result of which the strength increases again. In the sand layer between -20 and
25 m NAP, the strength increases once more. After -25 m NAP, the strength decreases again owing to the
influence of the Eem clay. The accompanying soil-drilling test, showing the cone resistance and friction ratio, is
reproduced for the sake of clarity, in order to indicate where the genuine clay layers are located.
Recent tests have revealed that the strength increases from 28 days to 120 days by approximately 50 %, by
which stage the strength development is not complete. On the basis of current knowledge, it may provisionally
be concluded that the strength requirements have been met.
12

Adapted worksite organisation through the steering committee
In order to be able adequately to carry out the jet grouting process, with the associated high requirements for
quality and necessary flexibility for interim adaptation, the organisation had to be adapted accordingly, so there
was maximum control on site and it was possible to react immediately to the results of the implementation. The
client, the contractor and subcontractor therefore formed a steering group during the jet grouting process. The
steering group operated at management level and was responsible for the quality and progress of the jet grouting
process. In the steering group, the Adviesbureau Noord/Zuidlijn (‘North/South Line Consultants’) and VOF
Stationseiland were represented on behalf of the municipality. Internationally recognised experts in the field of
jet grouting have also been called in by the client. The principal contractor, CSO, and the subcontractor, Smet
Keller, also formed part of the steering group and contributed specific knowledge.
In addition, the work on site was monitored intensively by an implementation monitoring team: a team
composed of supervisors representing the client, engineers representing the principal contractor and production
monitors representing the subcontractor, who were specifically concerned with jet grouting. This entailed,
among other things, adapting and monitoring the work plans and putting the operation protocols into place.
Monitoring
The station building was monitored continuously during the operation phase, whereby three independent systems
were used, namely:
Geotechnical ground monitoring, in which a system of measuring instruments inserted into the soil can
measure ground water pressure and soil deformations (inclinometers, extensometers, pore pressure
transducers);
Water level system for continuous measurement of the movements of the constructions inside the building;
External building monitoring, in which the façade of the station building was continuously measured by
means of ‘total stations’ positioned outside the building (component of the overall Noord/Zuidlijn
monitoring plan).
The fact that these measuring systems operated independently of one another provided an optimum risk
limitation situation, in which the systems monitor and validate one another. Furthermore, failure of one system
does not affect the continuation of the work, because the building can still be monitored.
In addition, during the jet grouting there was extra process checking for local effects resulting, for example, from
blockages (task of the inspector). Movements in the subsoil or the building are therefore quickly noted.
Conclusion
After production of approx. 65 % of the perimeter columns and approx. 20 % of the infill columns, it was found
that intensive process control was necessary and fruitful, and has therefore been continued undiminished. On-site
adjustment and adaptation to the operational methods remain daily activities, despite the learning experience
undergone. As a result of the intensive process control, this method of working is time-consuming, but has
nevertheless proven necessary in order to obtain optimum quality of the desired end product.
13