TBM Basics TUdelft

Starting a large diameter TBM from
surface
Feasibility study for the Netherlands
Master of Science Thesis
A.H. Oldenhave
April 2014
URUP method applied for large
diameter tunnels
Feasibility study for the Netherlands
Master of Science Thesis
A.H. Oldenhave
1387286
April 2014
PUBLIC REPORT
Technical University of Delft
Faculty of Civil Engineering and Geosciences
Department of Geotechnical Engineering

Starting a large diameter TBM from surface 28 april 2014
Personal information
Student:
Angela Oldenhave
Studienr.: 1387286
Stevinweg 27, Bergschenhoek
Tel.: 06-29442689
angela.oldenhave@gmail.com
Graduation committee:
TU Delft:
Prof. Ir. J.W. Bosch
Sectie Geotechniek, Faculteit Civiele Techniek en Geowetenschappen, TU Delft
j.w.bosch@tudelft.nl
Tel.: +31 (0)15 27 82844
TU Delft:
Ir. S. van der Woude
s.vanderwoude@tudelft.nl
svanderwoude@vhbinfra.nl
Tel.: +31 (0)15 27 85266
TU Delft:
Dr. Ir. W. Broere
w.broere@tudelft.nl
Tel.: +31 (0)15 27 81545
TU Delft:
Dr. Ir. C.B.M. Blom
Afdeling Bouw, Faculteit Civiele Techniek en Geowetenschappen, TU Delft
C.B.M.Blom@tudelft.nl
Tel.: +31 (0)15 27 84324
Volker InfraDesign:
Philip Labouchere
Ontwerpmanager
plabouchere@volkerinfradesign.nl
Tel.: +31 (0)348 435111
A.H. Oldenhave 2
Preface
This report represents my master thesis for the completion of the master Geotechnical Engineering
at the faculty of Civil Engineering and Geosciences, Delft University of Technology.
For my master thesis different projects were available. In retrospect, I am very glad I decided to
choose this subject. This thesis has allowed me to gain a considerable amount of knowledge; not only
in the tunnelling technique, but also in other geotechnical subjects and implementations of concrete.
There are numerous persons I would like to thank. First of all I would like to thank Sallo van der
Woude, who convinced me to engage into this subject and for his dedication and encouragement.
Further I would like to thank Kees Blom and Philip Labouchere who helped me with insightful
knowledge about topics outside my master scope. Also the suggestions, interest and input of Johan
Bosch and Wout Broere during meetings were very helpful during my thesis.
I would also like to thank the colleagues at Volker InfraDesign, who assisted and supported me by
explaining and gaining me information. Finally, special thanks are extended to my friends and family
and particularly my partner Wiebe Eikmans for their encouragement, patience, support, advice and
contribution.
April, 2014
Angela Oldenhave
A.H. Oldenhave 3
Summary
In the Netherlands, when a bored tunnel has to be constructed, a launch and a reception shaft have
to be excavated to guarantee sufficient cover. This is necessary to ensure face stability, prevent uplift
and risk of blow out. Moreover, especially in the western part of the Netherlands, the first 10 to 20
meters of soil consists of the Holocene layer consisting of soft clays and peat. By starting the
tunnelling process at sufficient depth, the tunnel will be driven through the Pleistocene and and the
soil deformations will remain small.
In 2009, a new method was introduced in Japan: the Ultra Rapid UnderPass (URUP) method. With
this method the tunnel boring machine (TBM) starts tunnelling from ground surface level.
Consequently there is no more need for a launch and reception shaft. This eliminates the need for
expensive shafts which are time-consuming to build, which makes the URUP method interesting to
apply in the Netherlands. However, applying this method will result in small covers at the beginning
and the end of the tunnel lining. Therefore complications are to be expected.
In this report, the feasibility of this method, applied for large diameter tunnels, is assessed by
investigating the main complications associated with this method (tunnelling with marginal cover
method) in the Netherlands. As a basis for the feasibility study a reference project, which is planned
for the near future, is used. The reference project “Rijnlandroute” is situated in the west of the
Netherlands (between Leiden and Voorschoten), where the first 11 meters consist of Holocene soft
soil. For this thesis only the eastern start of the tunnel is considered, that is located in a polder. The
groundwater level is located approximately one meter below surface level and the surrounding area
is largely greenfield.
Firstly, the soil deformations in the Holocene layer, when placing the TBM at surface level, are
determined. It can be concluded that without ground improvement the settlements will become too
big. It is therefore recommended to apply Mixed In Place columns, that have been applied
successfully several times in the past for similar cases. When the TBM reaches the Pleistocene layer,
ground improvement is no longer required and deformations remain small.
Due to the high groundwater table an embankment is necessary in permanent situation to prevent
uplift. Moreover, the tunnel invert has to be filled with a sand ballast as well. During construction
(ballast is not constructed yet) the water level will be drawn down temporarily with a maximum of
two meters.
Due to the absence of cover at start of the tunnelling process, the force distribution in the lining
differs from the conventional method. The axial loads are negligible, whereas bending moments still
exist resulting in opening up of the joints, which might cause leakage. Therefore it is required to
increase the capacity to withstand high bending moments when tunnelling with shallow covers. The
required embankment will increase the normal force. However, this increase is too low to counteract
the occurring bending moments in the lining. The Japanese box, which is a bolt type where the
tension forces are transferred directly to the reinforcement, will be applied (permanently). This bolt
type can withstand higher forces, though this bolt type will increase the construction costs.
It can be concluded that despite these issues, tunnelling with marginal covers is feasible in the
Netherlands. To gain more insight in the relative benefits of this method compared to the
conventional method, both are assessed, considering the costs as well as the construction time. The
A.H. Oldenhave 4
comparison shows that the total construction costs only differ 1.8 % (the costs of the conventional
method will be 1.8 % higher). However, the construction of the tunnel with the marginal cover
method will result in a shorter construction time. The construction time of the conventional method
is approximately four years whereas the construction time of the tunnel applying the tunnelling with
marginal cover method only amounts approximately three years. Therefore, it will be efficient to
apply this method in the Netherlands. However, despite the feasibility and efficiency of this method
in the Netherlands, it cannot yet be concluded if tunnelling with marginal cover will be
recommended in future. Further research is still required to conclude if the marginal cover method
will be successfully. With this thesis it can only be concluded that the tunnelling with marginal cover
method, applied for large diameter tunnels, is feasible in the Netherlands.
A.H. Oldenhave 5
Samenvatting
Wanneer in Nederland een geboorde tunnel wordt geconstrueerd, dienen een starten
ontvangstschacht gerealiseerd te worden alvorens met het boorproces begonnen kan worden. Deze
schachten zijn noodzakelijk om voldoende dekking boven de tunnel lining tijdens het boren te
kunnen waarborgen. Dit is nodig om opdrijving van de tunnel te voorkomen en voldoende
normaalkrachten in de tunnel lining te kunnen garanderen. Bovendien bestaat, vooral in het westen
van Nederland, de eerste 10 à 20 meter uit slappe Holocene grondlagen (zachte klei en veen).
Wanneer het boorproces op voldoende diepte begint, zal de tunnel grotendeels in de sterke
Pleistocene zandlaag geconstrueerd worden, wat resulteert in geringe zakkingen. Wanneer de
geboorde tunnel is voltooid, worden de toeritten naar het geboorde gedeelte aangelegd.
In 2009 is in Japan een nieuwe boormethode gepresenteerd: de Ultra Rapid UnderPass (URUP)
methode. Bij deze methode start de tunnelboormachine (TBM) op maaiveldniveau, boort en
construeert vervolgens de tunnel lining en eindigt weer op maaiveldniveau. Dit resulteert erin dat de
toeritten achteraf niet meer geconstrueerd hoeven te worden, maar al bij het boorproces zijn
inbegrepen en verkort de totale constructietijd van de tunnel. Wegens deze voordelen, wordt het
interessant om deze methode ook in Nederland toe te passen, maar deze methode resulteert in
geringe dekking aan het begin en eind van de tunnel. Daarom zullen bij het toepassen van deze
methode in Nederland complicaties verwacht worden.
In deze thesis wordt een haalbaarheidsstudie van deze methode (boren met marginale dekking)
verricht, toegepast voor grote diameter tunnels in Nederland. Voor deze studie worden de grootste
verwachte complicaties onderzocht, voor een referentieproject genaamd de “Rijnlandroute”. De
Rijnlandroute bestaat deels uit een geboorde tunnel in het westen van Nederland (tussen Leiden en
Voorschoten), waar de eerste 11 meter van de ondergrond bestaat uit slappe Holocene lagen. Voor
deze thesis wordt alleen de oostkant van de start van de tunnel (inclusief open en gesloten toerit) in
beschouwing genomen, gelegen in een polderlandschap. De grondwaterstand ligt gemiddeld één
meter onder maaiveld en er is geen bebouwing in de directe omgeving aanwezig.
Voor de haalbaarheidsstudie zijn allereerst de zettingen in de Holocene laag bepaald wanneer de

TBM op maaiveldniveau met het boorproces zal starten. Er kan worden geconcludeerd dat zonder
een grondverbetering de zakkingen in de slappe laag te groot worden en de TBM in de grond zal
zakken. Als grondverbetering worden Mixed In Place kolommen aangeraden. Deze grondverbetering
is al meerdere malen met succes in Nederland toegepast voor vergelijkbare projecten. Wanneer de
TBM het Pleistoceen kruist, is de grondverbetering niet langer meer nodig wegens de hogere
draagkracht van het Pleistoceen. De zakkingen blijven vanaf hier binnen de maximaal toelaatbare
waarde.
Om opdrijven van de tunnel lining door de hoge grondwaterstand te voorkomen, dient een dijk op
maaiveldniveau te worden aangelegd. Bovendien zal het gedeelte onder het wegdek in de tunnel
gevuld worden met een zandballast. Tijdens de constructie van de tunnel (op dit moment kan de
zandballast nog niet worden aangelegd) zal de grondwaterstand tijdelijk maximaal twee meter
verlaagd worden.
De afwezigheid van een dekking boven de tunnel lining zorgt ook voor een andere krachtverdeling in
de lining. De normaalkrachten zijn klein, waardoor de hoge buigende momenten niet kunnen worden
opgenomen. Dit resulteert in het openstaan van de voegen tussen de verschillende segmenten. Er is
A.H. Oldenhave 6
dus een noodzaak om de momentcapaciteit te vergroten bij boren met marginale dekking. De dijk op
maaiveldniveau zal de normaalkracht verhogen, maar dat zal niet genoeg zijn om de buigende
momenten, met mogelijk lekkage tot gevolg, op te kunnen nemen. Hiervoor zal een ander bouttype
toegepast worden; de Japanse box. Dit bouttype brengt de trekkrachten rechtstreeks naar de
wapening in het beton waardoor het grotere trekkrachten aankan. Een nadeel van dit type zijn de
hogere constructie kosten.
Ondanks bovengenoemde complicaties kan toch geconcludeerd worden dat het mogelijk is om boren
met marginale dekking in Nederland toe te passen. Om ook inzicht te krijgen in de efficiëntie van
deze methode, is een vergelijking gemaakt tussen de conventionele methode en de nieuwe methode
met het oog op de constructiekosten en constructietijd. Uit deze vergelijking blijkt dat de
constructiekosten van de marginale dekking methode 1.8 % minder zal bedragen. De constructietijd
van beide methoden scheelt meer. De constructietijd van de totale tunnel, wanneer de
conventionele methode wordt toegepast, bedraagt ongeveer vier jaar, terwijl de constructietijd voor
het boren met marginale dekking ongeveer drie jaar bedraagt. Dit betekent dat het efficiënt zal zijn
wanneer deze methode kan worden toegepast. Toch kan nog niet geconcludeerd worden dat boren
met marginale dekking in de toekomst in Nederland aangeraden wordt. Hiervoor is meer onderzoek
noodzakelijk alvorens deze conclusie getrokken kan worden. Met deze thesis kan alleen
geconcludeerd worden dat boren met marginale dekking voor grote diameter tunnels in Nederland
mogelijk is.
A.H. Oldenhave 7
Table of contents
Personal information ..........................................................................................................................2
Preface ...............................................................................................................................................3
Summary ............................................................................................................................................4
Samenvatting......................................................................................................................................6
1. Introduction .............................................................................................................................. 11
1.1. Motivation......................................................................................................................... 11
1.2. Approach methodology of master thesis ........................................................................... 11
2. Literature study......................................................................................................................... 13
2.1. General information .......................................................................................................... 13
3. Reference design “The Rijnlandroute” [4], [25] .............................................................................. 18
3.1. Introduction of the bored tunnel design ............................................................................ 18
3.2. Soil conditions ................................................................................................................... 19
3.3. Lining geometry of bored part ........................................................................................... 19
3.4. Total length of tunnel ........................................................................................................ 19
3.5. Integration of trajectory .................................................................................................... 20
3.6. Tolerance in alignment ...................................................................................................... 20
3.7. Limitations......................................................................................................................... 21
4. Consequences driving with low cover ........................................................................................ 22
4.1. Deformation of Holocene layers due to driving with marginal cover .................................. 22
4.2. Effects on buoyancy of bored tunnel due to driving with marginal cover ........................... 22
4.3. Effects on tunnel lining due to driving with marginal cover ................................................ 22
4.4. Effects on face stability whilst driving with marginal cover................................................. 23
4.5. Choice of TBM ................................................................................................................... 27
4.6. Effects on torque of TBM due to driving with marginal cover ............................................. 28
4.7. Effects on grouting of the tail void for marginal cover ........................................................ 28
4.8. Launching procedure for driving with marginal cover ........................................................ 29
4.9. Cross-connections ............................................................................................................. 30
4.10. Summary ....................................................................................................................... 32
5. Deformations of Holocene layer due to shallow driving ............................................................. 33
5.1. Introduction ...................................................................................................................... 33
5.2. Pressure caused by the TBM .............................................................................................. 33
5.3. Soil stresses ....................................................................................................................... 34
5.4. Moment and point of application of the TBM .................................................................... 36
A.H. Oldenhave 8
5.5. Soil investigation ............................................................................................................... 38

5.6. Soil deformation [29] ........................................................................................................... 39
6. Effects on buoyancy due to shallow driving ............................................................................... 49
6.1. Introduction ...................................................................................................................... 49
6.2. Safety factors .................................................................................................................... 50
6.3. Temporary measures to prevent uplift............................................................................... 50
6.4. Permanent measures to prevent uplift .............................................................................. 53
7. Effects on tunnel lining due to shallow driving........................................................................... 58
7.1. Introduction ...................................................................................................................... 58
7.2. Calculation models [18] ....................................................................................................... 58
7.3. Loads acting on the tunnel lining [18]................................................................................... 59
7.4. LDesign [39] ......................................................................................................................... 62
7.5. Output ............................................................................................................................... 62
8. Costs and planning .................................................................................................................... 79
8.1. Costs ................................................................................................................................. 79
8.2. Planning ............................................................................................................................ 79
8.3. Conclusion ......................................................................................................................... 80
9. Conclusion and recommendations ............................................................................................ 82
9.1. Conclusion ......................................................................................................................... 82
9.2. Recommendations............................................................................................................. 83
10. Sources ................................................................................................................................. 85
Books and reports ......................................................................................................................... 85
Readers......................................................................................................................................... 85
Papers and theses ......................................................................................................................... 85
Internet sites................................................................................................................................. 86
Regulations ................................................................................................................................... 86
Software ....................................................................................................................................... 86
Received information .................................................................................................................... 86
Appendices ....................................................................................................................................... 87
Appendix A: Methods to determine soil stresses [6] ....................................................................... 88
Appendix B: CPT results [37]............................................................................................................ 91
Appendix C: Geological results [37] ................................................................................................. 98
Appendix D: Settlements of Holocene layer when applying soil replacement ................................ 99
Appendix E: Buoyancy ................................................................................................................. 102
A.H. Oldenhave 9
Appendix F: LDesign [39] ............................................................................................................... 108

Appendix G: Calculation of reinforcement [12] .............................................................................. 111
Appendix H: Calculation of bolts.................................................................................................. 114
Appendix I: Planning [43]............................................................................................................... 117
A.H. Oldenhave 10
1. Introduction
In the last century, a variety of bored tunnels has been constructed worldwide. Due to lack of
understanding of tunnelling in soft soil, the first bored tunnel was started in 1995 in the Netherlands
(see Chapter 2). Since then more than ten bored tunnels have been constructed in the Netherlands.
For a bored tunnel a launch and reception shaft have to be excavated to create sufficient cover on
top of the bored tunnel lining. Recently a new method has been developed in Japan called the Ultra
Rapid UnderPass method (URUP method). With this method the TBM starts from ground level
instead of from a launching shaft. Consequently the launching and reception shafts are no longer
necessary potentially reducing costs and construction time. Therefore it is of interest to examine the
feasibility of the URUP method in Dutch ground conditions.
This thesis builds on an earlier study of the URUP method in the Netherlands of Beijer (2010).
However Beijer considered only a small rectangular envelope for a cyclists and pedestrians tunnel
with small lengths (which was concluded as being feasible). In this thesis, the adaptation of the URUP
method in the Netherlands will be extended to large diameter tunnels, called the “tunnelling with
marginal cover method” in this thesis.
1.1.Motivation
By applying the new method (tunnelling with marginal cover method) instead of the conventional
method, it is expected that the construction time and construction costs will decrease. Therefore it
becomes interesting to determine the technical feasibility and efficiency of this method in the
Netherlands. However due to differences between both methods, the expected complications and
possible solutions have to be determined to define the feasibility. In this thesis, the feasibility and
efficiency of tunnelling with marginal covers, applied in the Netherlands for large diameter tunnels
will be determined.
1.2.Approach methodology of master thesis

This report is master thesis for the faculty of Geotechnical Engineering. This thesis will therefore
focus on the geotechnical aspects of the marginal cover method. For completeness the structural
aspects will also be considered. This thesis is a feasibility study for the use of tunnelling with marginal
covers in the Netherlands, specifically applied for a case called Rijnlandroute, described in Chapter 3.
After the literature study (Chapter 2 and 3), the geotechnical issues for driving the TBM with shallow
or no cover in comparison with the traditional method (sufficient cover) are determined (Chapter 4);
thereafter three major issues will be explored in a broader sense (Chapter 5, 6 and 7). In Chapter 5
the deformations of the soil due to driving through the Holocene layer are determined. This
determines the required soil structure for tunnelling with marginal covers and has to be used for
further calculations. The following major issue, when tunnelling with low cover, is the buoyancy due
to high groundwater tables in the Netherlands, calculated in Chapter 6. Again the measures are
determined required to prevent uplift of the tunnel lining. The results of Chapter 5 and 6 are used as
input for the calculations for the tunnel lining, explained in Chapter 7.
After determination of the minimum requirements for the tunnelling with marginal cover method,
the costs as well as the planning are evaluated (Chapter 8). This gives insight in the efficiency of the
method in comparison with the traditional method. In order to obtain meaningful results for the
A.H. Oldenhave 11
traditional method, the information about the Rijnlandroute, applied by the Province of South
Holland, is used. Hereafter the conclusion of this thesis as well as the recommendations is given.
A.H. Oldenhave 12
2. Literature study
As start of this thesis project, the applied literature study is mentioned in this chapter, as well as the
motivation for this thesis. The literature study is divided in a study of the traditional method and a
study of the “tunnelling with marginal cover” method where the tunnelling process starts at surface
level.
2.1.General information
To determine the differences between the conventional tunnelling method and the tunnelling with
marginal cover method, both methods will be described in this paragraph. Hereafter the differences
are described.
2.1.1. Conventional method [1], [2], [9], [23]

The first bored tunnels date from the 19th century, and are mostly constructed in rock and stiff
London clay. In these ground conditions settlement was limited and knowledge of tunnelling through
hard soils was present. In the Netherlands mostly soft ground is present (soft sand, clay, peat) which
also resulted in preference of immersed tunnels (first immersed tunnel is Maastunnel, constructed in
1937 – 1942). However due to further development of tunnelling techniques (more knowledge about
tunnelling through soft soil), increasing lack of space in the Netherlands and growing resistance
against environmental pollution, the bored tunnel became more interesting in the Netherlands. This
resulted in the first bored tunnel in the Netherlands, the second Heinenoordtunnel, started in 1995
and completed in 1999, followed by ten more bored tunnels, see Table 1.
Table 1: Large diameter bored tunnels in the Netherlands
Bored tunnel Completed in Type of shield

2nd Heinenoordtunnel 1999 Slurry
Westerscheldetunnel 2003 Slurry
Groene Harttunnel (HSL-Zuid) 2007 Slurry
Botlekspoortunnel 2007 EPB
Sophiaspoortunnel 2007 Slurry
Tunnel Pannerdensch Kanaal 2007 Slurry
Hubertustunnel 2008 Slurry
Statenwegtracé (Randstadrail) 2010 Slurry
Noord/Zuidlijn 2017 (planned) Slurry
Sluiskiltunnel 2015 (planned) Slurry
Different ducts tunnels - Slurry
2.1.1.1. Technique
The tunnel boring machine (TBM, see figure below) consists of a steel cylinder (mentioned shield
hereafter) with a cutter head on front and a mixing chamber, a conveyor system, a thrust system and
trailing support mechanisms inside. The cutter head, which is the rotating cutter wheel at the front,
removes the soil on the front. The excavated soil, collected in the mixing chamber and acting as
support pressure, is then transported by a conveyor belt and a screw conveyor or pump system,
depending on the type of shield (see further). The lining consists of precast reinforced concrete
segments, which are transported through the already constructed tunnel by a trailer train. These
segments are then placed by an erector and thrust cylinders keep the segments in place.
A.H. Oldenhave 13
Furthermore, the thrust cylinders or hydraulic jacks are used to advance forward by pushing off
against the concrete segments. The segments will be connected with adjacent segments, by bolts
(permanent at start and end, temporary in between) and dowels, to complete the segmental tunnel
rings of the lining.
Figure 1: Tunnel boring machine [23]
The shield of the TBM is slightly larger (about 10 cm) than the tunnel lining and results in gaps
between the tunnel lining and the soil, when the lining leaves the cylinder. To avoid settlements of
the surface, this so called tail void is filled under pressure with grout, which normally takes place in 4
to 6 points around the tunnel lining.
2.1.1.2. Type of TBM

Two different types of TBM are often used in the soft ground conditions of the Netherlands: The
Earth Pressure Balance (EPB) shield and the Slurry shield. When a highly permeable soil (i.e. sand) has
to be excavated, the Slurry shield is recommended. This type of TBM can add a slurry under pressure
(generally bentonite) in the excavation chamber of the TBM which stabilises and supports the bore
front. The excavated soil is mixed with the slurry and transported to surface. A separation plant on
surface level separates the soil from the support fluid and the slurry can be reused again.
When cohesive soils are present, the soil might stick to the excavation wheel and will slow down the
progress when using the Slurry shield. For these soils the EPB shield is often used. This shield
contains special cutters at the front to prevent sticking of the soil to the bore front. This type of TBM
does not use slurry and therefore a separation plant is not required (reduced construction costs). The
excavated soil itself is capable to maintain the water pressure and horizontal effective ground
pressure at the front, see Figure 2. The excavated soil is mixed in the excavation chamber to become
homogeneous and easy to transport with a screw conveyor. By means of the screw conveyor the
pressure at the bore front is reduced to atmospheric pressure inside the tunnel. A disadvantage of
this shield type is the setting of the support pressure in the mixing chamber. The soil is difficult to
compress, which means that small changes of volume lead to large changes in pressure. To apply
“real-time monitoring” the small changes can be limited, according to Bakker (2008) [9]. This includes
regulating the discharge of the soil and use of additives to set the pressure.
A.H. Oldenhave 14
Figure 2: Required support pressure in EPB shield [9]
In the Netherlands both types of machines have been used (Table 1). The most important difference
between both types of TBM’s is the difference in pressure of the support fluid, required in the Slurry
shield when driving through sand layers, see next paragraph.
2.1.1.3. Minimum cover

When applying the Slurry shield, there is a probability of a blowout (escape of overpressure to the
surface), due to higher pressures of the fluid. This results in a minimum required covering. Moreover,
a cover also has to prevent buoyancy of the tunnel due to the upward force of the high groundwater
table in the Netherlands. As a guideline, it can be assumed that the minimum required cover is 1 to
1.5 times the tunnel diameter, depending on the weight of the soil on top [9]. The construction of a
bored tunnel is generally executed from shaft to shaft, starting at the required depth. The ramps are
completed after the tunnelling process when the shafts are no longer required for the construction
of the bored part.
When an EPB shield will be used for construction in theory no blowout will occur. However, when air
support pressure might be needed in case of repair works in the excavation chamber (for instance
the replacement of the cutting teeth on the excavation wheels), the minimum cover is required as
well.
2.1.1.4. Start and arrival of the tunnelling process

The TBM will be assembled in the launch shaft on a cradle. The initial thrust forces are supported of a
launching frame (connected to the concrete floor and the wall of the shaft), constructed in the
building pit [2]. Behind the sheet pile wall, at the front a low strength concrete block is installed to
prevent groundwater inflow when starting the tunnelling process. When the tail sealing (just behind
the shield) reaches the ground, the concrete block is not necessary any more. The water resistance
will be guaranteed by the first installed tunnel segments. The first rings will be installed in the
building pit, resting on the shield cradle, called the blind rings. These rings are required for the
thrusting of the TBM for the first meters. When about 200 meters is bored, the blind rings and the
launching frame can be removed [9].
To maintain water tightness at the reception shaft, the TBM will stop in the sealing block (low
strength concrete). The space between the sealing block and tail void will be grouted, that will lead
to a watertight connection. After hardening, the last rings can be constructed. A compression lid is
placed in the reception shaft or the building pit is filled with water to maintain the pressure for the
TBM [9].
A.H. Oldenhave 15
2.1.1.5. Cross connections

When two tunnels are bored next to each other, cross connections are required for the safety of the
users. By freezing, pressurisation or soil grouting, the connections can be constructed.
2.1.2. Tunnelling with marginal cover method[14], [15], [24]

This thesis describes and investigates the feasibility of a new method (in this thesis research
appointed as “tunnelling with marginal cover method”) to construct bored tunnels in the
Netherlands. This method will start at surface level, immediately with the bored part, and therefore,
it will not require a launching and reception shaft and ramps towards the tunnel.
2.1.2.1. Origin of the method: URUP method

In the field of bored tunnels, Japan already has a lot of expertise, also for soft soil conditions.
However, in 2009 at the ITA (International Tunnelling and Underground Space Association) congress
a new method was announced: the URUP
(Ultra Rapid UnderPass) method [15]. With this
method a TBM starts at the surface level, bores
an underpass and ends back on the surface,
see Figure 3. Therefore no launching and
reception shaft are required. The machine
pushes itself off against a launching cradle, Figure 3: Principle of the URUP method [24]
installed on surface level.
The method was first developed for the purpose of constructing an underpass in a short period of
time to cross an urban area with a high traffic density. The elimination of launching and reception
shafts enables rapid construction of the underpass (installation of shafts, excavation equipment or
transportation of the soil is no longer required), as well as continuously construction of the lining of
the entire underpass during driving of the TBM. Moreover, it alleviates traffic congestion when the
underpass is constructed. In the thesis of Beijer (2010) a description of a pilot project, executed in
Tokyo, is given. The bored tunnels using the URUP method are executed with an EPB shield, due to
the shallow cover on top to delete the risk of blow out of a slurry.
Meanwhile due to further development of this technique, even circular shaped large diameter
tunnels are constructed nowadays, see Figure 4 and Figure 5. The main difference with the
traditional method is the start at surface level resulting in construction of the tunnel with shallow or
even no cover. The TBM launches itself from a launching frame (see also Paragraph 4.8), installed at
surface level as well. In Figure 4 it is shown that still a shallow building pit is necessary. This is due to
the fact that the road alignment inside the tunnel is located higher than the bottom of the tunnel.
Therefore a shallow depth is still required. However, cover on top of the TBM is missing at the start
of the tunnelling process.
A.H. Oldenhave 16
Figure 4: TBM launching at surface level [3] Figure 5: TBM arrival at surface level [3]
Also other projects and tunnel shapes are constructed with the URUP method. For more information,
a reference is made to the report of Beijer (2010).
2.1.2.2. Tunnelling with marginal cover method in the Netherlands

When the URUP method was announced in the Netherlands, the question arose about the feasibility
of this method in this country. As a first research project, Beijer investigated the adaptation of the
URUP method for a reference project; a constructed underpass to cross a railway in the western part
of the Netherlands (soft soil at surface level). In the Netherlands the groundwater tables are higher
than in surrounding countries, and are only a few decimetres below surface level. In Japan the
groundwater levels are much lower and this resulted in more expected complications when applying
the URUP method in the Netherlands. Still, Beijer could conclude that constructing underpasses
below roads and railways for cyclists and pedestrians with the URUP method is technically feasible.
To extend above mentioned feasibility project, this thesis will investigate the possibility of
constructing tunnels with large diameters in the Netherlands (indicated as “tunnelling with marginal
cover method”). Again a reference project will be used where the method will be applied, see
Chapter 3.
2.1.3. Differences between conventional method and tunnelling with marginal

cover method
As previously mentioned, the main difference between the conventional method and the “tunnelling
with marginal cover” method is the elimination of the launching and reception shafts. The TBM starts
and ends at surface level. Moreover, when applying the traditional method, ramps have to be
constructed after the tunnelling process. However, when the tunnelling with marginal cover method
will be applied, the ramps are included in the bored part. Therefore it will be expected that the
tunnelling with marginal cover method will result in a shorter construction time.
When starting at surface level, one has to deal with other soil conditions (mostly soft soil), since the
new method starts at the surface instead of the Pleistocene sand layer when applying the
conventional method in the Netherlands. Especially the soil in the western part of the Netherlands
consists of a Holocene layer on top of the Pleistocene.
A.H. Oldenhave 17
3. Reference design “The Rijnlandroute” [4], [25]

In order to determine the feasibility of the “tunnelling with marginal cover” method, a case study in
the Netherlands is chosen as reference project. Currently plans are developed to construct a new
bored tunnel in the western part of the Netherlands called the “Rijnlandroute”. This will be used as a
reference design to investigate the possibility of the tunnelling with marginal cover method in the
Netherlands.
3.1.Introduction of the bored tunnel design

The “Rijnlandroute” is a new provincial road of the province Zuid-Holland connecting the highway
A44 in direction to the coast with the A4 at the level of Leiden, see Figure 6. This new connection
increases the accessibility of adjacent municipalities as Wassenaar and Leidschendam-Voorburg, as
well as new housing in the future. Furthermore the intensity of the old main road through Leiden will
be reduced.
The new road contains a connection between two highways, the A44 (north-west of Leiden) and the
A4 (south east of Leiden). To limit nuisance due to the new connection, the part between Leiden and
Voorschoten will be constructed as a tunnel, see white arrow in the figure below. The bored tunnel
starts at the east side just before the Schiekanaal and ends on the west side behind the railway. The
ramps, given in yellow and orange in the figure, connect the tunnel with the highways and are
located in agricultural area without constructions. In this thesis only the eastern ramp of the project
will be considered.
Figure 6: Overview Rijnlandroute with optimized tunnel location (white arrow) [4]
A.H. Oldenhave 18
3.2.Soil conditions
Only limited information is available on the different soil types. The subsurface at the location
consists of a Holocene layer (clay with peat layers) from surface level until approximately NAP -11.0
m. On this level the Pleistocene layer starts. Between NAP -2.5 m and NAP -6.5 m sand layers can be
present as well. The surface level is located at NAP -1.0 m to NAP -1.5 m.
3.2.1. Ground water levels

The normal groundwater level at the west side is located at NAP -1.70 m whereas the groundwater
level at the east side of the tunnel amounts NAP -2.40 m. For the highest occurring groundwater
level at the east side a level of NAP -2.10 m is assumed.
3.3.Lining geometry of bored part

For the determination of the lining geometry the Dutch safety regulations VRC (Veiligheidsrichtlijnen
part C) are implemented [27]. This results in the following determined geometry for the bored tunnel
of the Rijnlandroute (see also Figure 8):
 Inner diameter bored tunnel: 9.73 m;

 Outer diameter bored tunnel: 10.53 m;
 Lining thickness: 0.40 m;
 Number of segments: 7 + 1 keystone;
 Maximum gradient 4.5%.
When applying the above mentioned geometry and the geometry given in Figure 8, the top surface
of the road will be located approximately 2.5 meters above the bottom of the tunnel lining. This area
will also be used for the required cables and ducts. The concrete quality will be assumed to be
C35/45. Both tunnels (one for each direction) will be designed as a provincial road with maximum
allowable speed of 80 km/h. For further information a reference is made to the feasibility report of
the bored part of the Rijnlandroute of Movares (2012) [4].
3.4.Total length of tunnel

The tunnel consists of the bored part, the closed part and the open part of the ramps. The lengths of
each part are given below:
 Open part western ramp: 169 m;

 Closed part western ramp: 125 m;
 Bored tunnel: 1906 m;
 Closed part eastern ramp: 93 m;
 Open part eastern ramp: 251 m.
Summed, the total length of the tunnel is 2604 m, including two service buildings, constructed after
the tunnelling process in the building pits. In total seven cross connections will be constructed
between the two tunnels. The locations of the cross connections depend on present buildings at
surface level. The maximum distance between two cross connections amounts 250 m, which is in line
with the Dutch regulations [27]. The dimensions of the cross connections can be found in the report
of Movares (2012) as well, but will not be further discussed.
A.H. Oldenhave 19
3.5.Integration of trajectory
When the tunnelling with marginal cover method will be applied for this case, sufficient area has to
be present to start the tunnelling process at surface level. In the figure below the start location of the
tunnel is given. When applying the traditional method, the bored part will start just before the
waterway. When applying the tunnelling with marginal cover method, the tunnelling process will
start 300 meters earlier.
In the figure it is clear that at the location of the

start of the tunnelling process, the sloping roads
of the highway A4 are just coming together. By
constructing both ways with adequate length in
between, the tunnelling process of the second
tunnel will not have any influence on the first
tunnel lining.
According to guidelines the first 250 meters of

the ramp (10 seconds rule, see Movares (2012)),
before entering a tunnel, has to be a straight
way without change in the number of lanes. The
final tunnel will start 33 meters after the
launching shaft. Thus the exits of the highway
will come together 217 meters before start of
the tunnel. This length is present and therefore
the tunnelling with marginal cover method can
Figure 7: Integration of trajectory
be integrated at location without changes of the
present situation.
3.6.Tolerance in alignment
When tunnels are constructed, deformations will appear in soil (subsidence) as well as in tunnel
lining. The maximum deformations depend on the tolerance in the alignment. In the figure below the
gauge of the tunnel is given; the height is 4700 mm and the width is 7850 mm. According to
Rijkswaterstaat [28] dimensions are dependent on the category of a road. The Rijnlandroute is
designed as a primary road with a design speed of 80 km/h. This leads to category IV and a lane
width of 3.10 meters. However, when the tunnel is located in agricultural area, other (wider) vehicles
can be expected as well. It will be expected that no tolerance is acceptable in horizontal direction.
The minimum required height for a tunnel amounts 4.50 meters. This consists of a maximum height
of a truck of 4.00 m, a vertical movement of 0.20 m and a safety margin of 0.30 meter. However, in
Figure 8 4.70 meters is present. This means a tolerance in vertical direction of 20 cm is allowed.
To calculate the maximum tolerance in diagonal direction, the influence of the minimum required
height and the present height are compared:
Minimum diameter is √ m.
Present diameter when applying 10.53 m is √ m
A.H. Oldenhave 20
This results in an allowable tolerance of 9.15 – 9.05 = 0.10 m = 100 mm. This is in accordance with
the report of the Rijnlandroute.
However, the total deformations are the result of settlements of the tunnel in the soil (when the
tunnel will be drilled at an angle of 4.5 % but the tunnel lining will settle and results in adjustment of
the height of the lane) and deformations of the tunnel lining. For the tolerance in deformations of
the tunnel lining typically 1 % of the radius will be assumed, amounting 50 mm. It will be stated that
the maximum settlements in the soil amounts 50 mm as well as the maximum deformations in
tunnel lining.
Figure 8: Cross section tunnel Rijnlandroute with dimensions [4]
3.7.Limitations
On the next page the scope of this thesis is drawn (see blue part). For this feasibility project only the
length of the eastern ramp (start side) of the Rijnlandroute is considered. It is assumed that when
constructing the tunnel with the tunnelling with marginal cover method on this side is feasible, it will
not give complications at the west side of the tunnel (finishing side). The bored part of the traditional
method is pretended to remain the same. Moreover, as limitation the data (dimensions, location
etc.) are assumed to remain the same as the original Rijnlandroute. Therefore no other tunnels
(shape, diameter) or locations are considered.
It has to be said that this thesis only serves as a first investigation of the adaptation of the tunnelling
with marginal cover method in the Netherlands. Further research is strongly recommended.
A.H. Oldenhave 21
Figure 9: Scope of the thesis (eastern ramp)
A.H. Oldenhave 21
4. Consequences driving with low cover

Driving a tunnel with low or even no cover is very different from the convention situation with a
minimum of 1D cover. Nine effects will be summarised in this chapter. This thesis is intended as a
feasibility report only.
For this thesis it is assumed that the first three major effects will determine the feasibility of the
method in the Netherlands. When these three major effects cannot be solved, feasibility in the
Netherlands is assumed to be unattainable. The three main consequences will be discussed
respectively in more detail in subsequent chapters.
4.1.Deformation of Holocene layers due to driving with marginal cover

In the west of the Netherlands, the TBM sits be located on top of a thick Holocene layer when
starting from ground level. Consequently large settlements are expected that could exceed the
maximum allowable construction tolerances (see Paragraph 3.6). To limit these settlements different
measures can be taken; Preloading by means of fill material accelerating consolidation before the
start of the tunnelling process, soil replacement or ground improvement. In the following chapter,
these deformations will be calculated and advice will be given with regards to the decisive measure
and its allowable settlement.
The measures result in a modification of the soil conditions (stresses, strength). The modified soil
conditions will be taken into account for further calculations. Therefore the deformations of the soil
will be determined at the start.
4.2.Effects on buoyancy of bored tunnel due to driving with marginal

cover
In the Netherlands the groundwater table is located just below surface level. When applying
underground constructions, buoyancy always plays a major role to determine the depth of the
construction. Due to these buoyancy forces normally a bored tunnel starts at sufficient depth to
counteract these upward forces (cover on top has sufficient weight). When applying a tunnel with
shallow or no cover, the downward forces of the cover are not included. It is expected that the
tunnel lining will float and will not be stable anymore.
To prevent uplift of the tunnel lining, measures can be applied as well. This can be done by applying
extra weight on top, lowering the groundwater table, applying thicker lining segments or applying
extra ballast inside the tunnel. For this thesis however it will be assumed that all assumed dimensions
for the traditional method will also be applied for the new method (tunnelling with marginal covers).
Therefore, the tunnel lining segments as well as the diameter of the tunnel remain the same.
In Chapter 6 the upward forces and downward forces will be calculated for the first part of the tunnel
where shallow or no covers are present. Safety factors will be taken into account as well. With these
calculations it can be concluded if the buoyancy forces can be counteracted to retain the feasibility of
the tunnelling with marginal cover method.
4.3.Effects on tunnel lining due to driving with marginal cover

When applying a tunnel lining at sufficient depth, the soil forces, acting on the lining, result in high
normal forces and low bending moments. When shallow or no cover is present above the lining, this
A.H. Oldenhave 22
will result in unequal force distributions (low normal forces and high bending moments). The normal
force on top is missing which might result in high deformations in the lining. Moreover, due to high
bending moments this can result in gaps between the joints (the pressure zone decreases). These
two effects have to be countered.
In Chapter 7 the effects on the tunnel lining will be calculated at different depths. Due to the number
of calculations only two critical cross sections will be used as well as the cross section where the
traditional conditions apply (+ 1D cover). The measures that are discussed in Chapters 5 and 6 will be
taken into account. The deformations for each depth as well as the amount of reinforcement and
bolts will be determined with the help of a framework analysis to calculate the normal forces, the
shear forces and the bending moments in the tunnel ring.
4.4.Effects on face stability whilst driving with marginal cover

Launching a bored tunnel partly above surface level, results in a partially supported tunnel face. In
addition, the friction between the TBM and the ground is not equally distributed over its
circumference. It requires special attention to determine the exact forces acting on the tunnel during
driving of the TBM without cover. The forces, consisting of face support pressure (paragraph 4.4.1)
and friction below the TBM (paragraph 4.4.2), are described in this paragraph and determine the
required jacking forces, together with the minimum force to guarantee water tightness by the rubber
gaskets between the rings (paragraph 4.4.3). The jacking forces, consisting of jacks behind the TBM,
continue the tunnelling process by pushing off the TBM from the tunnel lining. When sufficient
distance between the TBM and tunnel lining is achieved, the segments can be placed where after the
jacks will push itself from the new installed ring. The recommended jacking forces conclude this
paragraph.
4.4.1. Support pressure

The support pressures at the bottom of the bore front are higher than on top (even zero when the
tunnel lining does not have a cover yet). These support pressures have to be exceeded to guarantee
progress in the tunnelling process. However, when the jacking forces become too high, this will result
in pushing of the soil instead of excavation of the soil. The support pressure on front is depending on
the movement of the soil. When the soil (and the TBM) does not move, the support pressure can be
calculated using the so called neutral earth pressure coefficient. When the soil is pushed away by the
TBM, i.e. when the soil is passive and the TBM is active, the support pressure can be calculated with
the passive earth pressure coefficient. To guarantee the driving progress, the jacking pressures have
to be higher than the support pressure when the soil is moving towards the TBM (the soil is acting),
which can be calculated with the active earth pressure coefficient. In Paragraph 4.4.1.1 the minimum
and maximum support pressures are calculated. The jacking forces have to remain between these
limits.
The jacking forces can be arranged separately for each jack. This results in lower jacking forces in the
top of the ring and higher forces at the bottom (the soil pressures increase by depth). For each depth
the minimum and maximum support pressures can be calculated. As starting point the starting depth
of 2.5 meter below surface will be calculated. Furthermore the three cross sections used for
calculations in Chapter 7 will be calculated as well.
A.H. Oldenhave 23
4.4.1.1. Active and passive support pressures

To prevent failing of the bore front and ingress of ground water, the jacking pressure has to be higher
than the active support pressure. The active support pressure on top is [10]:
( ) ( )
( ) ( )
To prevent collapse of the soil above the top or deformations at surface level, the jacking pressure
has to be smaller than the passive support pressure. The passive support pressure on top is [10]:
( ) ( )
( ) ( )
Above calculated support pressure give the limits in which the jacking pressures have to remain.
When the jacking pressure becomes too high, the soil will not be excavated which might lead to
deformations. When the jacking pressure becomes lower than the active support pressure, the soil
will react towards the TBM.
When applying the four different cross sections (see Table 2 and Chapter 7; start, half below surface,
small cover and conventional start), the minimum and maximum pressures can be calculated for
each depth (in Table 18 the estimated values for each depth are given. For the starting point the
tunnel is only located in the first layer, for these values see Chapter 5). The mean, minimum and
maximum groundwater levels are all calculated as well as the minimum and maximum pressures on
top and at the bottom of the tunnel lining. The required ground improvement is taken into account.
The results are given in the table below.
Table 2: Minimum and maximum jacking pressures for different depths
Groundwater levels [m Pmin, top Pmax, top Pmin, bottom Pmax, bottom
below surface] [kN/m2] [kN/m2] [kN/m2] [kN/m2]
bottom tunnel 0.6 0 0 30.5 52.1
2.5 m below 1 0 0 28.8 54.9
surface level 1.4 0 0 27.2 57.7
bottom tunnel 0.6 0 0 57.3 228.2
2.5 m + 3.82 m 1 0 0 55.7 231.0
embankment 1.4 0 0 54.0 233.8
bottom tunnel 0.6 0 0 106.4 350.0
6 m below 1 0 0 104.7 352.8
surface level 1.4 0 0 103.1 355.6
0.47 m cover 0.6 16.7 93.9 162.1 391.3
1 16.7 93.9 160.4 394.0
1.4 16.7 93.9 158.8 396.8
10.47 m cover 0.6 126.4 234.3 257.9 563.6
1 124.8 237.1 256.2 566.4
1.4 123.1 239.9 254.6 569.2
It is clear that especially at the start the range between the minimum and maximum pressures is the
smallest. Here the jacking pressures have to be controlled very precisely in such a way that the
pressure outside the boundaries will not be reached. When the depth of the TBM increases, the
range for possible jacking pressures becomes wider due to the larger range of active and passive
A.H. Oldenhave 24
support pressure and fewer problems will be expected. To increase the range between the minimum
and maximum pressure, the required embankment to prevent uplift (see Chapter 4) is extended
towards the start location. This results in a bigger range due to higher soil pressures on the bore
front, see Table 2, and therefore an extension of the required embankment will be recommended to
prevent failure of the support pressure.
For this calculation it is assumed that the groundwater tables are not lowered. However, after
calculation it appeared that a groundwater level decrease of 2 meters is necessary. This is not taken
into account in this calculation, because this effect will only increase the range between the active
and passive support pressure. Moreover, the critical point is located at the start where the
groundwater level is not lowered yet. However, the required embankment (see Chapter 6) and
ground improvement (see Chapter 5) are included.
4.4.2. Shield friction

Another force which has to be overcome by the jacking forces is the friction between the TBM and
the soil. Therefore the following formula can be used (partly in soil) [11]:
( )
( ( ) )
 r’ is the horizontal effective soil pressure at the centre of the bore front (or at the centre of
the part below surface level);
 fw is the reduction factor due to friction between soil and steel, about tan() = tan(22) = 0.4
according to Tan et. al. (2008) [16];
 f is the stress reduction factor due to the conical shape of the TBM (0.8);
  is the internal friction angle for the soil;
 a is the adhesion between the soil and the steel frame of the TBM, which amounts 0.853 kPa
when the temperature of the soil is 5C [17];
 fbent is the reduction factor for smearing medium (assumed to be 1 for EPB);
 lTBM is the length of the TBM;
 DTBM is the diameter of the TBM;
 R is radius of the TBM;
 H is depth of the bottom of the TBM below surface level.
The horizontal effective soil pressure at the centre of the TBM depends on the depth. In the table
below the friction force for the four different depths (see Table 2) is calculated. By dividing this force
by the area of the tunnel lining, the minimum required jacking pressure to overcome the friction is
calculated, see last column in Table 3. When the TBM is not totally covered, the part above surface
level will not contribute to the friction force. This is taken into account as well (last part of the
formula).
A.H. Oldenhave 25
Table 3: Friction force for the different depths
Depth bottom TBM [m] Friction part [-] Fshield, friction [kN] Pshield, friction [kN/m2]
2.5 (no embankment) 0.324 257 20
2.5 (with embankment = 6.32) 0.564 1122 88
6 (with embankment = 9.82) 0.833 2069 163
11 (with embankment = 12.77) 1 2616 206
21 1 2995 235
4.4.3. Required force to guarantee water tightness

The jacking forces are dependent on the active and passive support pressures, calculated in
paragraph 4.4.1, and the friction between the TBM and the soil, calculated in paragraph 4.4.2.
Moreover, the jacking forces need to guarantee water tightness of the tunnel by applying enough
force to seal the rubber gaskets between the tunnel rings. According to Taillebois (1999), a gasket
compression of 30 kN/m is recommended. Multiplied with the circumference of the tunnel ring, this
results in a force of 30 x 31.8 = 955 kN. Dividing this force again by the area of the tunnel ring results
in a minimum required pressure to guarantee water tightness of the tunnel of 75 kN/m2. Only when
applying the embankment at start position, water tightness can be guaranteed.
4.4.4. Alignment control

During driving of the tunnel, small deformations will occur (due to the weight of the TBM). By
applying higher jacking forces at the bottom of the tunnel lining and smaller jacking forces in the top,
the deformations can be decreased. However, due to the ground improvement the extra jacking
forces to correct the deformations are small. It will not be expected that the extra jacking forces will
result in problems.
4.4.5. Recommended jacking forces

In the previous paragraphs the pressures that influence the required jacking pressure are calculated.
The jacking pressures may not exceed the passive support pressure to guarantee excavation of the
soil. On the contrary, the jacking pressure needs to exceed the active support pressure and the
friction between the soil and the TBM. It will be assumed that the extra jacking pressures to steer the
TBM (alignment control) and decrease the deformations of the soil below the TBM due to uneven
load distribution (first part of the TBM is heaviest) are small and are therefore not taken into account
to determine the required jacking pressure.
The decisive cross section is the starting point. To increase the range of the possible jacking pressure
(during the tunnelling process, the jacking pressure can fluctuate) and to guarantee water tightness,
the embankment has to be extended towards the starting point of the tunnelling process. In the
table below the pressures are transferred into forces (when the tunnel is located only partially in the
ground, it is easier to calculate the forces instead of pressures). The minimum jacking force to
guarantee water tightness is 955 kN. The minimum force to overcome the friction between the steel
frame and the soil and embankment is 1122 kN. When the maximum support pressure at the bottom
amounts 228 kN/m2 (see Table 2) and decreases to 0 where the TBM crosses the embankment, this
results in a force of approximately 6220 kN. It can be concluded that the minimum force to overcome
the friction and guarantee water tightness of the segments is smaller than the maximum support
A.H. Oldenhave 26
force, see table below. It is also clear that without an embankment the friction forces and jacking
forces cannot be counteracted by the maximum support pressure.
Table 4: Total forces acting in/on TBM for different depths
Depth bottom TBM [m] Fmin, jack [kN] Fshield, friction [kN] Fmin, support [kN] Fmax, support [kN]
2.5 (no embankment) 955 257 240 410
2.5 (with embankment = 6.32) 955 1122 1570 6220
6 (with embankment = 9.82) 955 2069 2990 9860
11 (with embankment = 12.77) 955 2616 4700 11330
21 955 2995 7480 16340
In the table above also the minimum support forces are given. The jacking forces have to overcome
this minimum support force to guarantee excavation of the soil plus the friction of the shield. When
the maximum jacking forces are in the order of some MN (compare with minimum jacking force of 55
kN when applying two jacks per segment), it will not be expected that the friction and the minimum
support pressure cannot be overcome by all the jacking forces.
When applying the embankment at start level, the jacking forces can provide enough force to
counteract the friction and the soil pressure on front. It will be recommended to apply a jacking
pressure equal to the mean support pressure on front Pmean, support (see formula below) + Pfriction if this
jacking pressure is higher than the minimum jacking pressure of 75 kN/m2 to guarantee water
tightness.
( ( )) ( ( ))
kN/m2
Remark
When the minimum jacking pressure of 75 kN/m2 will be applied on top of the TBM, the moment will
increase. That influences the total settlements underneath the TBM (see Paragraph 5.4.1 and
further). However, after calculation, the maximum settlements below the TBM will only increase by 1
mm. Due to the high weight of the TBM the extra jack force will not influence the moment and
settlements significantly.
4.5.Choice of TBM
Originally the bored part of the Rijnlandroute would start approximately 16 meter below surface. The
top of the tunnel is then still located in the Holocene. However, most of the tunnel lining is already
located in the Pleistocene sand, see Figure 9 (grey part). Therefore for the Rijnlandroute the slurry
shield will be used. However, when applying the tunnelling with marginal cover method, the bored
part will already start entirely in the Holocene. Moreover, to delete the risk of a blowout an EPB
shield will be used.
The application of the EPB shield in the Holocene will not result in problems. However, when the
TBM reaches the permeable Pleistocene, requirements to prevent inflow of groundwater are
necessary. For the Botlekspoortunnel an EPB shield was used (Table 1). The launching and arrival
A.H. Oldenhave 27
shaft were still located in the Holocene, but during the tunnelling process the TBM was driving full
face through medium to coarse grained Pleistocene sands. For this project foam conditioning was
successfully applied to support the Pleistocene sands with higher water pressures. This foam is
biological and therefore separation of the soil is not necessary. The foam ensures good
transportation through the excavation chamber and more important creates impermeability of the
earth paste in the excavation chamber. When the Pleistocene at the Rijnlandroute becomes too
permeable for transportation of the soil and the tunnelling process, foam will be added as well.
4.6.Effects on torque of TBM due to driving with marginal cover

When the cutting wheel is rotating, this results in a torque which can influence the stability of the
TBM. However, in Paragraph 4.4 it is concluded that an embankment at start is necessary to absorb
the minimum required jacking forces. The TBM will therefore start in more than 6 meters soil
resulting in free space of only 4.5 meters. The friction between the TBM and the soil will therefore
become high enough to counterbalance the torque. It will be assumed that torque of the TBM will
not result in any problems during the start of the tunnelling process.
4.7.Effects on grouting of the tail void for marginal cover

The TBM consists of a shield wherein the latest rings are constructed before these rings are
subjected to the soil. This steel frame has a conical shape resulting in a void between the ring and the
soil when the TBM frame moves forward. To prevent settlements due to the arising “tail voids”,
commonly this tail void is filled with grout. By exactly measuring and monitoring the amount of
injected grout and the deformations of the soil, the settlements can be limited to the minimum.
When the tunnel lining is located partly above surface level, this results in complications to inject the
grout in the tail voids. The grout will flow out on surface level. By applying fast curing mortar instead
of normal grout, the outflow will decrease. Also when more locations for injecting the mortar will be
applied (normally 4 to 6, see Chapter 2) and will be arranged separately, only the bottom will be
injected at start. During increase of depth of the tunnel more and more injections will be activated,
see Figure 10.
A.H. Oldenhave 28
Figure 10: Activated injection locations when tunnel lining is partly above surface
Further research to prevent flow out of the mortar is recommended.
4.8. Launching procedure for driving with marginal cover

For the traditional method a launching frame behind the TBM is installed which is connected to the
start shaft, see for instance Figure 11 and Figure 12. For the tunnelling with marginal cover method,
there is no need for a sheet pile wall since the launching shaft is not required. However the thrust
forces required for launching the TBM have to be transferred to the soil via the thrust frame. A thrust
frame will be constructed which will transport the forces towards the Pleistocene sand by foundation
piles. The thrust frame can be poured into a concrete floor on top of the foundation.
Figure 11: Launching of the TBM at steel frame (yellow) in Figure 12: Steel frame (yellow) connected to the sheet pile
launching shaft [40] wall on front [40]
When the TBM will launch on the steel frame, this will result in a force distribution given in Figure 14.
According to this distribution, the steel frame will have tension forces at front resulting in failure of
A.H. Oldenhave 29
the concrete floor when reinforcement will not be constructed. However, when the steel frame will
be poured into the concrete and connected with reinforcement present in the concrete floor (see
Figure 15), the structure will only fail when the yielding stress of steel is reached. For this thesis it will
be assumed that the forces acting on the steel frame are below the boundaries of failure (yielding of
steel or failure of the concrete floor by the pressure forces). However further research is
recommended. To counteract the pressure forces on the concrete floor, foundation piles will
transport these forces towards the strong Pleistocene layer, see figure below.
Figure 13: Launching procedure “tunnelling with marginal cover” method
Figure 15: Steel frame connected (welded) to

Figure 14: Force distribution in launching frame reinforcement
4.9.Cross-connections
According to the Dutch regulations the maximum distance in longitudinal length between two cross
connections in a tunnel amounts 250 meters [27]. The “first” cross connection will be constructed at
A.H. Oldenhave 30
surface level where the alignment of the road is at the same height, see Figure 16. At this location
the bottom of the tunnel is constructed at 2.5 meters below surface, i.e. the launching shaft. At this
location the tunnel lining above surface level will be removed after completion of the bore process.
Therefore, for this cross connection no real constructions are needed. Thus, at this location the two
tunnels can be constructed with enough distance to minimize any influence on the other tunnel
(unknown for the new procedure).
Figure 16: “First” cross connection at surface level
The real tunnel part starts when the tunnel lining above surface level will not be removed anymore.
According to Table 52 the tunnel lining removal ends when the bottom of the tunnel is located 4
meters below surface level; 1.5 meters deeper. This results in a start of the tunnel at 33 meters
distance from the launching shaft. Therefore, the next cross connection will be constructed 283
meters forward from launching shaft. The bottom of the tunnel is located at 15 meters below
surface; on top of the tunnel lining a cover of approximately 5 meters is present. It will be expected
that the cross connection at this location can be constructed as well as the other cross connections
designed by Movares (see Figure 17). At this location the tunnels are less far removed from each
other to minimize the construction costs of the connections.
Figure 17: Cross connection as designed by Movares [4]
For the construction of the second cross section with shallow cover, the freezing method will be
used. The shallow depth is advantageous for this method; the freezing lances can be injected
underneath the tunnel lining more easily, resulting in better freezing of the whole soil around the
tunnel. When the lining is located deeper into the soil, the freezing method becomes less reliable.
Then it is more unlikely that the whole soil underneath the lining is thoroughly frozen. This can result
in groundwater inflow during construction of the connection. For the deepest cross connections it is
therefore recommended to inject the soil with jet grouting, which is also advised by Movares for the
Rijnlandroute.
A.H. Oldenhave 31
4.10. Summary
In this chapter nine consequences are described when applying the new method (tunnelling with
marginal covers) instead of the traditional method:
1. Effects on deformations in Holocene layer;

2. Effects on buoyancy of bored tunnel;
3. Effects on tunnel lining;
4. Effects on bore front stability;
5. Effects on EPB driving through Pleistocene sand;
6. Effects on torque of TBM;
7. Effects on grouting of the tail void;
8. Effects on launching procedure;
9. Effects on cross connections.
The first three mentioned consequences are expected to determine the feasibility of this method in
the Netherlands; when these three major effects cannot be solved, the tunnelling with marginal
cover method cannot be applied in the Netherlands. For this thesis these three major consequences
will be further evaluated in Chapter 5, 6 and 7. The other six aspects are mentioned in this Chapter
and can be solved relatively easily. Therefore, it can be concluded that when the first three major
consequences can be solved, the remaining six aspects will not result in problems for the feasibility
of the tunnelling with marginal cover method in the Netherlands.
A.H. Oldenhave 32
5. Deformations of Holocene layer due to shallow driving

In this chapter the deformations of the Holocene layer will be determined when the bored process
will start at surface level. When the deformations become too big, ground improvement or soil
replacement is required. In this chapter it will be concluded whether a ground improvement or
replacement will be recommended. References are made to likewise projects.
5.1.Introduction
When starting the tunnelling process at surface level, i.e. on top of the Holocene layer, large
settlements will be expected due to the extreme weight of the TBM. As start the soil characteristics
will be determined. Unfortunately soil investigation was not available. With help of DINO loket [37],
which is a Dutch software program, the soil characteristics and its profile are estimated (see
Appendix B). This program contains all the information of soil investigations of the past. This
estimation will be used to calculate the settlements and to determine possible solutions.
5.2.Pressure caused by the TBM

To determine the deformations in the Holocene layer caused by the TBM, first the weight of the TBM
has to be determined. Therefore, the data of the Hubertustunnel is used:
Table 5: Weights different parts TBM [5]
Description Weight Contributes to (see Figure 18)

Part Length [mm]
Cutting wheel 122 ton O 1010
Cutting wheel drive 75 ton A 2560
Front shield 274 ton A 2560
Middle shield 175 ton B 4080
Rear shield 70 ton C 3595
Erector frame 23.5 ton C 3595
Total incl. pipes and installations 900 ton A+B+C 10235
Screw conveyor1 55 ton B 4080
Total 955 ton A+B+C 10235
1
Was not applied at the Hubertustunnel (slurry shield was used) so the weight of the screw conveyor at the
Botlekspoortunnel is applied.
A.H. Oldenhave 33
Figure 18: Weights and lengths TBM (EPB shield)
It is assumed that the part A+B+C causes the decisive deformations. The cutting wheel is connected
to this part via a bearing structure and therefore also contributes to the deformations. So the total
weight is 955 ton which works over a length of 10.235 meters. Furthermore, it is assumed that the
same width which contributes to the break-up calculation (see Appendix E), contributes here to the
soil deformations. So width = ( ) ( ) m.
Now all the data are known to determine the pressure caused by the heavy part of the TBM. The
pressure for this part becomes:
5.3.Soil stresses
Different methods can be used to estimate the stresses of the soil below a foundation. In Appendix A
Newmark’s chart is explained as well as the Boston Rule for uniform footing (B x L) [6]. It can be
concluded that both methods result in the same order of soil stresses. For this thesis the Boston Rule
will be applied to determine the required soil stresses.
A.H. Oldenhave 34
5.3.1. Boston Rule

Boston’s approximation assumes that stresses at certain depth are constant below the object in the
influence area, see Figure 19. This means the stress distribution is equal on the whole length of the
foundation at a certain depth.
Figure 19: Boston Rule (approximation method) [26]
When the depth increases, the area of the stress distribution also increases and therefore the value
of the additional stresses 'z will decrease. Therefore 'z is depending on the depth. With the
following formula 'z can be calculated:
, where B = 8.066 m (see previous paragraph) and L = 10.235 m.

( )( )
So when depth z increases, the additional stresses decrease. The total stresses in the new situation
become 'z = 'z;0 + 'z.
This results in Table 6. The value for the effective specific weight, ’, is estimated to be 14 kN/m3 for
the dry Holocene layer and 14-10=4 kN/m3 for the saturated Holocene layer.
Table 6: (increase of) stresses of the soil
' [kN/m3] z [m] 'z;0 [kPa] 'z [kPa] 'z [kPa]

14 0 0 115.7 115.7
14 1 14 93.8 107.8
4 2 18 77.6 95.6
4 3 22 65.2 87.2
4 4 26 55.6 81.6
4 5 30 48.0 78.0
4 6 34 41.8 75.8
4 7 38 36.8 74.8
4 8 42 32.6 74.6
4 9 46 29.1 75.1
4 10 50 26.1 76.1
A.H. Oldenhave 35
The increase of the stresses in the soil is smaller when it increases with depth. Therefore the
additional stresses (due to additional weight of the TBM) have the biggest consequences in the first
meters.
5.4.Moment and point of application of the TBM

In previous paragraph the mean stresses were calculated for the soil below the TBM. However, when
the resultant force does not apply in the centre of the TBM, this will result in a moment and the
stresses will not be equally divided under the whole length. In this paragraph first the point of
application is determined. When this point is known, the differences in stress underneath the TBM
can be calculated, see Paragraph 5.4.1.
When, per part of the TBM, the weights are known, the total force and the point of application can
be determined, see Figure 20 and Figure 21. Here, only the known forces will be used to calculate the
point of application. The real force working on the soil will be assumed to be 9,550 kN. The width of
the TBM will be disregarded.
The total moment in point A, caused by the four separate weights, is
∑ ( ) ( ) (
) tonmm
Figure 20: Schematic overview forces acting on soil
The total force FR (weight) is ton kN
To determine the exact location where this resultant force works, the moment has to be divided by
the total force: mm.
The total length of the TBM is 11,245 mm = 11.245 m. This means the geometric centre of the TBM
lies at 5.6225 m (see figure). The length between the point of application and the centre is 5.6225 -
3.8193 = 1.8032 m.
Figure 21: Resultant force causing a moment
A.H. Oldenhave 36
The TBM causes a moment of kN kNm.
Remark
The jacking forces influence the moment. However, it will be expected that these forces are in the
same order and for this reason disregarded. Nevertheless, when the first segments of the tunnel ring
have to be placed, the jacks temporary have to be released and no jacking forces can be carried out.
The ring consists of seven segments + one keystone. Due to this, it will be assumed that the
remaining jacks will carry the released jacking force. Therefore this influence can be disregarded.
5.4.1. Tilting of the TBM

The soil layer under the TBM has to prevent that this moment will result in a downwards tilt by
having enough bearing capacity. It is expected that the shield will not contribute in the soil stresses;
therefore a length of 10.235 meter is assumed. The resultant force Fr is not located in the centre of
the TBM; therefore the force will result in a counter pressure under a part of the TBM. According to
Türke (1993) [7] when the eccentricity of the resultant force e is b/6 < e < b/3, the maximum
pressure can be calculated with the following formula.
Figure 22: Stress distribution due to eccentricity
The length between FR and the centre

amounts 1.80 m, so the eccentricity is b/6
< e < b/3.
In Figure 22 the meaning of the

parameters is explained (a is the width).
The maximum pressure 0,max amounts
( )
kN/m2, see Figure 23.
Figure 23: Prevention of tilting
A.H. Oldenhave 37
5.5.Soil investigation
The deformation of the soil depends on the stiffness of the different soil layers (see Terzaghi’s
formula, next paragraph) and therefore a soil investigation is required.
To determine the structure of the soil at the desired location, six different cone penetration test
results were used from DINO loket [37], which collects all kind of information about the underground
of the Netherlands. Unfortunately no information at the exact location is present yet. So by
interpolating the information (CPT results) of the surroundings a reliable estimation of the soil can be
made.
Table 7 is used to describe the different soil layers.
Table 7: Friction ratio and cone resistance for different kind of soil types [6]
Soil type Friction ratio Cone resistance

Sand, medium – coarse 0.4%
Sand, fine – medium 0.6% 5 – 30 MPa
Sand, fine 0.8%
Sand, silty 1.1%
Sand, clayey 1.4% 5 – 10 MPa
Sandy clay or loam 1.8%
Silt 2.2%
Clay, silty 2.5%
Clay 3.3% 0.5 – 2 MPa
Clay, peaty 5.0%
Peat 8.1% 0 – 1 MPa
The results of each CPT can be found in Appendix B: CPT results. The six different locations show
diversity in the layer structure of the soil. Especially the CPT’s measured along a road have stronger
layers at the top. The two CPT’s measured in the polder (S30H00674 and S30H00677) contain
especially clay and peat with a stronger silty layer between approximately 4 and 7 meters below
surface. This is in accordance with the geological information of DINO loket, see Appendix C:
Geological results. Therefore, the two CPT’s of the polder are used to estimate the mean soil
structure.
In Figure 24 and Table 8 the estimated soil structure and its description can be found. It will be
assumed that this soil structure is plausible for the location.
A.H. Oldenhave 38
Table 8: Soil structure of location
Type of Layer Depth below Thickness Mean cone

soil surface [m] [m] resistance qc [MPa]
Peat 1 0–3 3 0.2
Clay, silty 2 3–4 1 1.0
Silt 3 4–7 3 2.0
Clay 4 7 – 11 4 0.5
Sand, silty 5 11 – 20 9 10.0
Figure 24: Estimated soil

structure
5.6.Soil deformation [29]

To determine the soil deformation different calculation models are present. For this thesis the
formula of Koppejan is used, who combined the formula of Terzaghi with the Dutch method of
Keverling Buisman:
( ) ( )
With this formula the deformations at each depth can be calculated using the stress in that layer and
the primary compression coefficient Cp and the second compression coefficient Cs.
The formula can be rewritten as:
( )
Where:
( )
The values for the compression coefficients are given in NEN 6740 (Table 1) [30] and below in Table
9. It is assumed that after one month (i.e. 30 days) the compression is equal to at least 90 % of the
total settlements of the cohesive layer. Another assumption is the infinity of the stiffness of the
Pleistocene layer. It will be assumed that the settlements will only be located in the Holocene layer.
It appears that when a layer of 10 meters sand will be added to the calculation, this layer will only
settle 2 mm. It can be concluded that assuming the Pleistocene sand layer as an infinite stiff layer is a
correct assumption. The calculations will be executed for the Serviceability Limit State (SLS). For the
Ultimate Limit State (ULS) the compression constants have to be divided by the partial factor m =
1.3.
The TBM will start at 2.5 meters depth. This means only 0.5 meters of peat remains below the TBM
and part of the TBM is present in the groundwater (the differential effective stresses ’z will
decrease, therefore total settlements will decrease). For this calculation the lowest groundwater is
taken into account (1.4 meters below surface level). This means the smallest upward pressures and
A.H. Oldenhave 39
therefore the maximum settlements. It can be concluded that even without the peat layer the
settlements remain too big, see Table 9. The tunnel cannot be constructed without any ground
improvement.
The calculations are executed for the centre of each layer. For the Boston rule the round TBM is
supposed to be a plate with dimensions 8.066 m x 10.235 m. When the bottom of the TBM is below
groundwater level, the buoyancy forces are very small due to the small volume below groundwater
level. For this calculation the height from bottom TBM to width 8.066 m is removed from the depth
below groundwater level. The figure below clarifies this phenomenon.
Figure 25: Height bottom tunnel - plate assumption
According to the figure, only when the bottom of the TBM is located more than 1.88 meters below
groundwater level, the buoyancy forces will counteract the weight of the tunnel. For the start (2.5 –
1.4 = 1.1 meters below groundwater level) no buoyancy forces are included.
For example the differential effective stress of layer 3 (see Table 9) is calculated as:
( ) ( )( ) ( )( )
kPa.
Table 9: Total settlements Holocene layer with TBM at 2.5 meter depth
Soil type Layer Thickness Mean ' 'z;0 'z 'z Cp Cs h

[m] depth [m] [kN/m3] [kPa] [kPa] [kPa] [-] [-] [m]
Peat (dry) 1 1.4 0.7 12 8.4 0.0 8.4 7.5 30 0
Peat (wet) 2 1.1 1.95 2 17.9 0.0 17.9 7.5 30 0
Peat (wet) 3 0.5 2.75 2 19.5 266.1 285.6 7.5 30 0.25
Clay, silty 4 1 3.5 7 23.5 227.8 251.3 15 160 0.18
Silt 5 3 5.5 9 40.5 158.4 198.9 45 1300 0.11
Clay 6 4 9 4 62.0 95.2 157.2 7 30 0.71
Sum h [m] 1.25

Clearly ground improvement is necessary to keep the settlements below the limits.
5.6.1. Ground improvement

Without any ground improvement and starting at 2.5 meters depth the shield will sink into the soil
for more than one meter causing problems for the construction of the tunnel lining. Possible ground
A.H. Oldenhave 40
improvements are a soil replacement, an embankment, jet grouting or mixed in place. These
phenomena will be summed below.
5.6.1.1. Embankment and soil replacement

Applying an embankment on top of the Holocene layer is an inexpensive option. Moreover when
applying an embankment of 5 meters and a soil replacement of the first three meters by sand, the
total settlements will still amount 47 cm, see Table 10. Therefore this ground improvement will not
be recommended.
Table 10: Total settlements Holocene layer with embankment (5 meter) and soil replacement

3
[m] depth [m] [kN/m ] [kPa] [kPa] [kPa] [-] [-] [m]
Sand (dry) 1 6 3 18 54.0 0.00 54.00 600 0.00
Sand (wet) 2 1.5 6.75 10 115.5 0.00 115.50 600 0.00
Sand (wet) 3 0.5 7.75 10 125.5 266.05 391.55 600 0.00
Clay, silty 4 1 8.5 7 131.5 227.75 359.25 15 160 0.08
Silt 5 3 10.5 9 148.5 158.39 306.89 45 1300 0.05
Clay 6 4 14 4 170.0 95.17 265.17 7 30 0.34
Sum h [m] 0.47
5.6.1.2. Jet grouting

Another method to improve the strength and reduce the settlements is jet grouting. With this
principle grout will be implied in the soil under high air pressures and the soil is compressed. Jet
grouted columns with high compressive strengths are the result. A disadvantage is the high
construction costs and the inadequate knowledge about this method due to few applications in the
Netherlands. Moreover, the soil becomes very strong and may lead to difficulties for the bored part.
Therefore this method will not be recommended.
5.6.1.3. Mixed-In-Place [31], [32]

A more commonly used method in the Netherlands nowadays is the mixed in place method (MIP).
With this method the soil is stabilized with binder suspension to increase the stiffness of the soil. The
surrounding soil is interrupted by an auger and the pores are filled with the suspension. This will
result in minimal settlements.
The soil can be stabilized by columns (column stabilization), see figure, or by applying MIP for the
whole ground volume (block stabilization). Both methods can be combined as well, but require
different machines (the block stabilization does not consist of columns). The columns can be applied
until a depth of 25 meters. The block stabilization (mixing a ground volume) can only be applied for 5
meters depth. The columns can be constructed using the ‘dry’ and the ‘wet’ method. With the dry
method binding suspension will be injected with air pressure; for the wet method the suspension will
be mixed with water (it becomes a slurry) before it enters the soil.
A.H. Oldenhave 41
Figure 26: Possible configurations of stabilized MIP columns
When applying stabilized columns the strength will contribute to the surrounding non-stabilized soil
when the strength of the column is not higher than circa 20 times the strength of the surrounding
soil (to prevent that the columns will operate as a pile foundation). When a load is applied on top of
the mixed soil, the columns will contribute more in the load distribution, causing an arc effect, see
Figure 27 (the arrows show the contribution of the non-stabilized soil to the MIP columns).
Figure 27: Force distribution in the soil (brown is MIP columns, blue is non-stabilized soil)
The advantages of applying MIP as ground improvement are the low construction costs (production
rate, no soil removal) and the improvement of the geotechnical properties of the Holocene layer.
Disadvantages are the limitations in depth and hardening time of the mixed soil. Moreover, it is
difficult to monitor the increased strength of the soil. This is depending on the homogeneity of the
mixing process and the properties of the soil layers. However, different reference projects show good
results.
5.6.1.3.1. Botlekspoortunnel[33]
The first reference project is the Botlekspoortunnel. This tunnel is mostly located in the Pleistocene
layer. The ends however, connecting the bored part to the shafts, are located in the Holocene. At the
east side the strength of the clay and peat layers were too small and ground improvement was
necessary to decrease the settlements to an acceptable level and create sufficient support for the
concrete tunnel lining. Initially jet grouting was scheduled. However, due to financial benefit, better
soil conditions to bore through the improved part and less removal of soil, the MIP method became
an interesting option.
A.H. Oldenhave 42
The increase of strength is depending on the soil structure. The soil

consisted of:
 Surface level at NAP +4.5 m;

 Sand layer NAP +4.5 m – NAP -0.5 m;
 Clay layer NAP -0.5 m – NAP -5.0 m;
 Sand layer NAP -5.0 m – NAP -6.0 m;
 Peat layer NAP -6.0 m – NAP -10.5 m;
 Silty sand layer NAP -10.5 m – NAP -12.0 m;
 Soft clay layer NAP -12.0 m – NAP -19.0 m;
 Pleistocene layer starts at NAP -19.0 m.
Figure 28: Soil structure Botlek
The two possible methods for the MIP (wet and dry method) were prior to the construction on site
tested in a pilot project next to the location of the tunnel. For the wet method 100 % cement was
injected in the soil. The dry method used 80 % cement and 20 % anhydrite, together blown into the
soil. After the pilot project and different tests it was concluded that the wet method would be
recommended. This method ensures more homogeneity in the improved soil, as well as in horizontal
direction as vertical direction, and the quality of the ground improvement can be checked more
easily.
The stabilized area was 11 x 30 x 125 meters and the columns were executed from NAP – 8.0 meters
to NAP – 19.0 meters (23.5 meters below surface). In total 4,500 columns are installed with a
diameter and hart to hart distance of 800 mm. For every kind of soil the amount of cement is
measured. After installation of the columns measurements were performed in order to determine
the compressive strength of the improved layers. The measurements are given in the table below.
Table 11: Results of measurements at Botlekspoortunnel [32]
Soil Amount of cement Compressive strength

Peat 340 kg/m3 440 kPa
Organic clay 340 kg/m3 520 kPa
Clay with thin sand layers 230 kg/m3 500 kPa
Peat with sand 230 kg/m3 500 kPa
After this project it was concluded that the required strengths are achieved. The homogeneity of the
soil was enough and the settlements remained below the limits. Yet, additional investigation is
recommended to determine the influence of the surrounding area and the variation in the results.
The construction was carried out in 10 months, by two machines fulltime. A third machine was not
used to decrease the amount of heave. The total settlements remained below 50 mm.
5.6.1.3.2. Randstadrail Rotterdam

For this project the ground improvement was necessary to decrease the settlements for the in future
realising railroad above the tunnel of the Randstadrail. Moreover, the soil layers next to the tunnel
A.H. Oldenhave 43
were too weak to construct a tunnel lining. The stability of the soil had to be increased by MIP
columns.
The columns had to be constructed singular, in rows with overlap, in a triangle pattern and as block
stabilization. The columns were 14 to 18 meters long, with diameter 600 and 800 mm. The block
stabilization was executed until the Pleistocene layer. For this, it was expected that the settlements
were minimal.
In the triangle pattern the hart to hart distance

amounted 1500 mm, whereas the diameter was only
600 mm, see Figure 29. This meant that soft soil was
still present. However (as previously mentioned) due to
the arc effect the soil was strong enough to limit the
settlements.
Figure 29: Triangular pattern
After installation of the columns some measurements were executed. It could be concluded that due
to the MIP method the strength of the soil was increased.
5.6.1.4. MIP applied for the Rijnlandroute [34]

Due to good results in the past, it will be recommended to improve the strength of the soil by
applying the Mixed In Place method. For the calculations the CUR report 2001-10 (Diepe
grondstabilisatie in Nederland) is used. This report especially explains the methods to calculate the
settlements of the soil as well as the columns with respect to the hart to hart distance and diameters
of the applied columns. When taking into account a maximum driving tolerance of 50 mm, the hart
to hart distance and diameter of the MIP columns can be calculated.
For the Rijnlandroute the wet method will be recommended, especially due to the more
homogeneity in vertical direction. The differences in strength of the layers become smaller and
therefore less differential settlements will be expected. The last clay layer above the Pleistocene
layer determines the greatest settlements. Therefore it will be recommended to apply the MIP
columns until the Pleistocene layer. The Pleistocene layer will be assumed as an infinitely stiff layer.
When assuming the same compressive strength of the Botlekspoortunnel (Table 11), the maximum
present pressure (281.0 kPa) remains lower than the strength of the soil. This means the soil deforms
elastically and it follows . Again in Table 12 below, the soil characteristics are estimated
with table 1 of NEN 6740. The TBM will start at 2.5 meter depth, resulting in a remaining peat layer
of 0.5 meter. To estimate the bedding stiffness Eoed of the soil, the following formula is used:
( )
( )( )
Table 12: Soil characteristics
Layer h [m]  [-] Es [kPa] Eoed [kPa]

Peat 0.5 0.15 500 528
Clay, silty 1 0.3 2000 2692
Silt 3 0.3 15000 20192
Clay 4 0.4 1000 2143
A.H. Oldenhave 44
h 8.5 m Mean Eoed; soil 8483 kPa

Mean Eoed; column 127244 kPa
As earlier mentioned the difference between the strength of the soil and the strength of the columns
may not differ more than 20 times to keep the contribution of the strength of the soil. The strength
of the soil or the stress is linear with Young’s Modulus according to Hooke’s law and therefore the
difference between Eoed; soil and Eoed; column is smaller than 20 times as well. According to CUR report
2001-10 it is assumed that column = soil. When assuming a difference of 15 times, the mean Eoed of the
columns can be assumed: kPa.
The strength increase of the soil depends on the amount of columns (hart to hart distance) and the
diameter of the columns. This is reflected in an enhancement factor a. When applying a triangular
pattern, this factor can be calculated according to:
Where:
 a = enhancement factor for triangular pattern;

 A = Area of the column; ;
 s = hart to hart distance of the columns.
In final situation the settlements of the soil is equal to the settlements of the columns (the system is
in balance), due to the higher stresses on top of the columns and smaller stresses on top of the soil.
The total settlement of the surface can be calculated with:
∑
( )
Where q is the load applied on top of the soil (281 kPa due to weight of the TBM) and w m is the
settlement of the soil as well as the columns. When now assuming a hart to hart distance and a
diameter of the columns, the settlements can be calculated, see further. These settlements have to
remain below 50 mm (see Paragraph 3.6).
According to the CUR report the settlements for the soil and columns can also be calculated
separately, depending on the load for each kind. By assuming that w m = w1 = w2, where w1 is the
settlement of the column and w2 is the settlement of the soil, the loads q1 and q2 can also be
calculated.
kPa.
A.H. Oldenhave 45
When assuming a maximum bearing capacity ult;col of 440 kPa, the maximum allowable load for one
column amounts . So q1 < q1;max.
With above mentioned conditions finally the optimal dimensions of the columns can be calculated,
see bold row in Table 13. It appears that the maximum allowable load q1;max is decisive for the
calculations. When applying 800 mm columns with a hart to hart distance of 900 mm the total
settlements only amount 26 mm (< 50 mm). However, q1;max is 284 kPa and q1 is 274 kPa. The
additional soil stress q2 is 7 kPa. In the table below other possibilities are given as well.
Table 13: Optimal MIP column pattern and other possibilities

s [m] D [m] a [-] q1;max [kPa] q1 [kPa] q2 [kPa] wm [m]
0.65 0.60 0.77 306 276 5 0.024
0.90 0.80 0.72 284 274 7 0.026
1.10 1.00 0.75 297 275 6 0.025
Figure 30: Optimal MIP column
1.00 0.90 0.73 291 274 7 0.025 pattern
5.6.2. Required area of ground improvement

To determine the minimum length in longitudinal direction where ground improvement is required,
for different steps in depth the settlements are calculated, taking into account the buoyancy. Due to
the bigger volume of the TBM below groundwater level, the upward forces become bigger and the
total settlements will decrease. However, according to the Boston rule, when applying a load at
bigger depth, the load distribution is smaller and therefore the additional soil stresses remain bigger.
In Appendix D the settlements for different depths of the TBM are calculated. When depth increases
indeed the additional stress becomes smaller, however even when applying the TBM half a meter
above the Pleistocene in the clay layer, the settlements remain too big (80 mm > 50 mm), see Table
47. Therefore it can be concluded that ground improvement (Mixed In Place) is necessary when the
bottom of the TBM is still located in the Holocene layer. To conclude whether the bearing capacity of
the Pleistocene is still high enough, an extra calculation is added, see Table 48. When the bottom of
the TBM rests at the Pleistocene the TBM will only settle 1 mm. The assumption that the Pleistocene
layer is infinitely stiff is therefore correctly assumed.
When the TBM starts at 2.5 meters depth, reaches the Pleistocene at 11 meters depth and the TBM
bores at an angle of 4.5 %, the total length the ground improvement has to be applied amounts
( )
m.
According to the Boston rule, the distribution of the stresses in the soil develops like the figure below
(2:1 in depth). At the starting point the depth of the Holocene layer is 8.5 meters. However, when
assuming a rectangular plate this plate will be lying 1.88 meters higher. Therefore the thickness of
the Holocene layer amounts 8.5 + 1.88 = 10.38 m. The total width of the influence area becomes
meter. The total area of MIP for the east side is 200 x 18.5 m2
for each starting point (two areas at the east side, see also Chapter 8). At the west side the tunnels
will end in the same building pit, 40 meters width. There the total width for applying MIP is 5.19 + 40
+ 5.19  50.5 meters.
A.H. Oldenhave 46
Figure 31: stress distribution in Holocene layer
On the next page the required ground improvement is schematized in a figure (Figure 32),
representing the total research area of this thesis, i.e. the eastern ramp of the Rijnlandroute. It can
be concluded that when applying MIP columns, the deformations remain below the maximum of 50
mm. Therefore, the jacking forces do not have to participate to control the alignment of the lining
with high forces. Still small differences will be applied due to the higher support pressures and
friction at the bottom of the lining.
A.H. Oldenhave 47
Figure 32: Length of required ground improvement for application of the “tunnelling with marginal cover” method
A.H. Oldenhave 48
6. Effects on buoyancy due to shallow driving

In previous chapter it is concluded that ground improvement is necessary to avoid too large
deformations to construct a tunnel. In this chapter the effects on the buoyancy will be calculated.
The improved soil characteristics are used for the calculations. Distinction is made between
temporary situation and permanent situation. For each depth (in steps of 1 meter) the calculations
are executed. These can be found in Appendix E.
6.1.Introduction
The main reason to start the bore process on a specific depth in a building pit is to prevent uplift
forces. When starting the bore process on surface level, the high groundwater table (0.6 m below
surface is decisive) will cause uplift forces. Therefore, when starting at surface sufficient
countermeasures have to be taken into account to prevent any upwardly directed forces at any
location of the tunnel.
The road alignment in the tunnel will be located 2.5 meters above the bottom of the tunnel lining.
Therefore the actual bore process will start at 2.5 meters depth instead of surface level (see Figure
33).
Figure 33: Starting depth tunnelling process
For the calculations the ground improvement is taken into account. In CUR 2006-2 the amount of
added cement is given for each soil type (see fifth column in Table 14). This added by the initial
weight factors leads to the following results, required for the uplift calculation.
A.H. Oldenhave 49
Table 14: Soil structure of location [30]
Type of Depth below Thickness initial concrete total ’

soil surface [m] [m] [kN/m ] [kN/m ] [kN/m ] [kN/m3]
3 3 3
Peat 0–3 3 12.0 3.4 15.4 5.4

Clay, silty 3–4 1 17.0 2.3 19.3 9.3
Silt 4–7 3 19.0 2.3 21.3 11.3
Clay 7 – 11 4 14.0 3.4 17.4 7.4
Sand, silty 11 – 20 9 20.0 - 20.0 10.0
Figure 34: Estimated soil

structure
The soil structure given in the table above is only applied for the part where the tunnel lining is still
located above the Pleistocene layer. When the top of the tunnel is located 0.47 meter below surface,
the lining will cross the Pleistocene and the ground improvement is not necessary any more (4th
column of Table 14 will then be used for calculation). This results in less weight of the untreated soil
on top of the lining and therefore higher buoyancy forces, see tables in Appendix E.
The possible countermeasures (embankment, lowering of groundwater table, ballast) are different
for the temporary and the permanent situation, see next paragraphs.
6.2.Safety factors
In the Netherlands directives for the design of constructions (in Dutch: “Richtlijnen Ontwerp
Kunstwerken) are drawn. In this report a description of guidelines can be found for each kind of
construction including safety factors. According to these directives the following safety factors have
to be used for the uplift calculations [35]:
  = 1.0 for permanent unfavourable loads;

  = 0.9 for permanent favourable loads;
  = 1.5 for variable unfavourable loads.
The safety factor of the groundwater level is 1.0. This safety factor applies for the “worst case
scenario” or the highest groundwater table for buoyancy calculations. According to the report of the
Rijnlandroute the highest occurring groundwater table lies at NAP -2.10 m. The surface level varies
between NAP -1.0 m and NAP -1.5 m. Therefore for the uplift calculation a groundwater level of 0.6
m below surface level will be assumed.
Moreover, according to the Dutch “Richtlijnen Ontwerp Kunstwerken, ROK” [35] a temporary
removal of a soil layer with a thickness of 0.5 m above the tunnel should be taken into account. This
is only taken into account for the permanent situation.
6.3.Temporary measures to prevent uplift

For the temporary solutions the following measures are taken into account:
 Embankment (max. is height in permanent situation), with dry is 16 kN/m3;

 Water level decrease (max. 2.0 m).
A.H. Oldenhave 50
Two different starting points are possible; to minimize the change in groundwater level and to limit
the height of the temporary embankment. Both scenarios are calculated, including the above
mentioned safety factors. However, in final situation an embankment is always necessary and
therefore this will be used as starting point. As limitation the height of the permanent embankment
is taken into account. When further measures are required, the water level will be decreased. This
will lead to the following calculation; see Appendix E and Table 15.
With the calculations it can be concluded that as well as an embankment as water level decrease is
necessary to avoid uplift. In the table below a summary for the temporary situation is given. In Figure
35 the summary of the minimum required measures is more clearly presented in a drawing. For the
construction of the tunnel with the tunnelling with marginal cover method, a higher embankment
will be recommended. In the figure only the minimum required heights of the embankment are
drawn which clarifies the winding line.
Table 15: Summary of the temporary situation
Zones Top of tunnel below surface [m] Temporary measure

Smallest height of embankment
Zone A 10.0 to 7.5 Water level decrease
Zone B 7.5 to -4.0 Embankment + water level decrease
Zone C -4.0 to -6.5 Water level decrease
Zone D -6.5 to -8.0 Nothing
A.H. Oldenhave 51
Figure 35: Temporary situation
A.H. Oldenhave 52
6.4.Permanent measures to prevent uplift

In the permanent situation it is assumed that a decrease of the groundwater level is not desirable
and therefore needs to be avoided. According to the cross-section of the tunnel of the Rijnlandroute,
a volume with a height of 2.5 meters is present below the alignment which has to be used for cables
and pipelines (including a drainage system for the incoming rainwater), see Figure 36. The required
area for these cables and pipelines is assumed to be 2 m 3/m. Without this area a total area of about
13 m3/m remains which can be used for ballast.
Figure 36: Cross section tunnel with present volume for ballast [4]
For the ballast three different materials are assumed:
 Sand ( = 16 kN/m3);
 Slag ( = 19 kN/m3);
 Concrete ( = 23 kN/m3).
For each kind of ballast a calculation is made to determine whether or not an embankment is
required and whether the tunnel lining above surface level can be removed (see Appendix E). For this
calculation the safety factor of 0.9 will be used for all the downward loads (tunnel lining, ballast and
weight of the ground). In Appendix E.2 the measures for each zone can be found. Due to the missing
ground improvement, a ballast is required in the whole assumed area (even when 1D cover is
present). In Table 16 a summary of the three tables is given. Again, in Figure 37, Figure 38 and Figure
39 the situation for each kind of ballast is presented in a drawing as well.
Table 16: Summary of the different zones
Zones Top of tunnel below surface [m] Permanent measure

Concrete ballast
Zone 1 10.0 to 7.5 Sand ballast
Zone 2 7.5 to 6.0 Concrete ballast
Zone 3 6.0 to -5.0 Concrete ballast + embankment (+ 225 m long)
Zone 4 -5.0 to -6.0 Concrete ballast
Zone 5 -6.0 to -8.0 Sand ballast
A.H. Oldenhave 53
Slag ballast
Zone 2 7.5 to 6.5 Slag ballast
Zone 3 6.5 to -5.0 Slag ballast + embankment (+ 250 m long)
Zone 4 -5.0 to -6.0 Slag ballast
Sand ballast
Zone 2 7.5 to -6.0 Sand ballast + embankment (+ 290 m long)
6.4.1. Proposed measure

The outcome for the different types of ballasts does not differ significantly. When applying concrete
ballast the embankment on top will be applied over 225 meters length. When applying sand ballast,
the embankment has to be 290 meters long. For this case it will be recommended to apply sand
ballast. This can be constructed more easily. Moreover, the costs will decrease despite the longer and
higher embankment (4.19 meters).
To determine if a height of an embankment of 4.19 meters is accepted, different heights of the

surface level of the Netherlands are checked [38]. It can be concluded that differences in height of up
to 5 meters is normal at the start of bored tunnels. Also next to overpasses an embankment of 5
meters is ordinary.
A.H. Oldenhave 54
Figure 37: Permanent situation with concrete ballast
A.H. Oldenhave 55
Figure 38: Permanent situation with slag ballast
A.H. Oldenhave 56
Figure 39: Permanent situation with sand ballast
A.H. Oldenhave 57
7. Effects on tunnel lining due to shallow driving

Constructing a tunnel at shallow depth can also result in problems with respect to tunnel lining
stability. This phenomenon is explained and calculated in this chapter. The forces and bending
moments in the tunnel ring are determined in three cross sections. With these forces and the
required measures it can be concluded if a tunnel construction with shallow covers is feasible.
Asymmetrical situations (for example eccentric vehicles in the lining) are not taken into account. For
this thesis only the situation during construction is taken into account.
7.1. Introduction
When applying a round shaped tunnel lining with small covers on top of the lining, the normal force
in the elements will decrease. When the normal force with respect to the bending moments in the
lining will become too small, this can result in high tension stresses. These stresses can be absorbed
by reinforcement. Still, an upper limit for the applied reinforcement in percentages of the total area
of the concrete element does exist (to prevent failure of the pressure zone of the concrete before
yielding of the reinforcement, this will result in spontaneous failure). Therefore the exact amount of
required reinforcement has to be calculated to check if this requirement is fulfilled.
Between the segments in ring direction the forces have to be translated by bolts. To determine the
required amount of bolts as well as the required reinforcement two critical cross sections will be
chosen for calculation as well as the starting depth of the traditional method for a good comparison
between the different methods. Hereafter it can be concluded if a construction with small covers is
feasible with respect to the tunnel lining.
7.2.Calculation models [18]

At first the normal forces and bending moments for each depth have to be determined. This can be
calculated by analytical and numerical methods. When applying a numerical calculation (for example
with Plaxis) the ring model (lining and soil) will be divided in finite elements (Finite Element Method,
FEM) and the soil is represented by a soil continuum, see Figure 40. The calculation, executed by
computer, gives very accurate results. This calculation method is also very complex and therefore will
be applied for difficult situations which analytical methods cannot simulate. Examples of such
situations are determining the influence of two adjacent tunnels or the influence of plasticity of
different soil layers.
Figure 40: Continuum model Figure 41: Beam spring model Figure 42: Crown reduction
A.H. Oldenhave 58
A more common method is the analytical method where the ring will be modelled as a beam spring
model (also used at framework programs, see Figure 41). The lining has certain stiffness (i.e. EI and
EA) and the springs will represent the stiffness of the soil (instead of a soil continuum). The horizontal
and vertical soil loads acting on the lining can be converted into radial and tangential loads, see
Figure 43. In theory it is expected that both loads are acting on the lining. However, in practice it
appears that when assuming both radial and tangential loads, the lining is sometimes over
dimensioned. When tangential load is omitted, calculations still satisfy according to different results
in practice. Apparently the lining does not give tangential support in some cases. According to Blom
(1995) it can be assumed that for deep locations (H > 3*D) the tangential support is present (‘bond’)
while for shallow locations of the tunnel (H < 2*D) the tangential support is omitted (‘slip’ or ‘no
bond’). When for this case only the shallow locations are within the scope, it will be assumed that
tangential support is omitted and therefore only the radial load works on the tunnel lining. H is the
length between the centre of the tunnel and the surface.
For this case it is assumed that the analytical beam spring model will give reliable estimations of the
forces and bending moments in the tunnel lining. A comparison with a numerical model will be
recommended when applying this method.
Figure 43: Transformation of the horizontal and vertical loads to radial and tangential loads [11]
Another assumption is the omitting of the soil support on top of the tunnel lining for shallow
locations, see Figure 42. At these depths the cover is minor resulting in too low resistance along the
sliding lines to counteract deformations and loads with the soil stiffness. By omitting the springs on
the top part over an angle of 90 degrees, the soil stiffness will be even estimated to be zero (no
resistance). According to Blom (1995) the soil directly above the tunnel lining does not influence the
stiffness when H < 2D (cover is 1.5D).
7.3.Loads acting on the tunnel lining [18]

To calculate the forces and bending moments acting on the tunnel lining, a framework program can
be used where the tunnel lining is simplified as the earlier mentioned beam spring model. To create
such a model, the lining is subdivided in different points and between the segments the points are
simplified as joints. The loads acting on each point are described below.
7.3.1. Soil loads

On each point horizontal and vertical soil stresses are working. Converting these stresses into radial
stresses and multiplying with the area of the lining where the stresses are working, the soil loads can
A.H. Oldenhave 59
be derived. As earlier mentioned it will be assumed that no tangential support will occur. Therefore
only the radial stresses are taken into account.
The radial soil stress can be calculated with (see also Figure 43):
( ) ( )
7.3.2. Bedding stiffness [19]

The soil has a certain stiffness to oppose forces and bending moments (resulting in deformations) of
the tunnel lining. To determine this soil stiffness, simplified as springs along the tunnel lining, the
radial and tangential bedding constant according to Duddeck are used:
[kN/m3];
[kN/m3].
Duddeck (1980) also mentions that soil support may only be assumed when the tunnel lining
deforms externally to the surrounding soil for depths H < 2D. According to Ahrens (1982) and
Erdmann (1983) the soil support on top of the tunnel can be even assumed to be zero in the worst
case, regardless of depth. As earlier mentioned, for small covers (H < 2D) the so called crown
reduction (over an angle of 90 degrees) of the soil stiffness will be assumed.
To estimate the bedding stiffness Eoed of the soil, the following formula is used:
( )
( )( )
For this case a ground improvement is applied (MIP columns). Therefore the Young’s modulus of the
soil is different for the columns and for the soil. To estimate the mean Es for the total soil (columns
included) the area of the columns and the soil is calculated for one triangle (the columns will be
applied in triangular pattern). The area of the columns becomes:
m2;
The total area of the triangle can be calculated with:
√ √ m2;
The part of the columns becomes:

Figure 44: Area column – soil in
triangular pattern
.
This means Ecolumn (which is 15 times Es) has to be multiplied with this factor and the remaining 28 %
is determined by Es:
The required parameters for the calculations of this chapter are summarized in the table below:
A.H. Oldenhave 60
Table 17: parameters of the different soil layers (according to NEN 6740)
Soil Depth [m] Es Etotal -] qc dry sat  []

[MPa] [MPa] [MPa] [kN/m3] [kN/m3]
Sand embankment Variable 25.0 25.0 0.30 10.0 18.0 20.0 26.0
Peat 0–3 0.5 5.5 0.15 0.2 15.4 15.4 15.0
Clay, silty 3–4 2.0 22.0 0.30 1.0 19.3 19.3 22.5
Silt 4–7 15.0 165.0 0.30 2.0 21.3 21.3 27.5
Clay 7 – 11 1.0 11.0 0.40 0.5 17.4 17.4 17.5
Sand, silty 11 – >20 25.0 275.0 0.30 10.0 18.0 20.0 26.0
It is clear that many parameters vary with depth (bedding stiffness Eoed, natural earth pressure
coefficient K0 and the saturated soil weight sat). Moreover, for each depth the height of the
embankment on top is different as well (see previous chapter and Appendix E.2). As a simplification
for the calculations, the parameter in the centre of the tunnel lining is assumed to be valid for all the
points of the tunnel lining. An estimation of Eoed can now be calculated for each depth as well as the
required bedding stiffness kr representing the soil support.
Table 18: Estimated parameters for the three calculated cross sections (see further)
Mean dry Mean sat Mean  Mean K0 Mean Es Mean Mean Eoed
[kN/m3] [kN/m3] [°] [-] [MPa]  [-] [MPa]
bottom tunnel 15.4 18.31 22.59 0.616 20.367 0.254 24.634
6m under
surface level
0.47 m cover 15.4 15.23 20.22 0.654 54.598 0.362 92.67
10.47 m cover 15.4 19.70 25.57 0.568 23.792 0.305 32.493
7.3.3. Longitudinal joint stiffness [20]

The longitudinal joints require more explanation. The ring is divided in different points. Seven of
those points are located at the longitudinal joints (the ring contains 7 segments + 1 keystone, but the
keystone is omitted in this calculation, because the influence is negligible). The longitudinal joints
reduce the stiffness of the ring (a homogeneous ring will act stiffer), but still contain some rotational
stiffness (with small forces the joint will not collapse). The rotational stiffness is different for the
following two situations: 1. there is no gap present yet between the two segments and 2. a gap starts
in the joint. For the first calculations it will be assumed that the joint is still closed. The constant
rotational stiffness cr amounts where lt is the contact length in the longitudinal joint, b is
the width and Ec is the Young’s modulus of concrete. Reference is made to Blom (2002) for the
derivation. The other points are not modelled as joints and therefore no additional stiffness has to be
applied.
7.3.4. Self-weight
The last load acting on the tunnel lining is the weight of the lining itself. Assuming a self-weight of 24
kN/m3 for concrete, the total weight can be calculated by multiplying this value with the volume of
the ring. In theory the self-weight does not influence the results of the forces acting in the tunnel
A.H. Oldenhave 61
lining for deep locations. However, this scope especially focuses on shallow depths, so this load shall
be taken into account as well.
7.3.5. Safety factors

For this thesis only the ULS (ultimate limit state) will be calculated. Still, it will be recommended to
calculate the serviceability limit state as well.
For the ULS the following safety factors are given by Walraven (2006) [12]:
Table 19: Safety factors
Permanent load Variable load

a Combinations Unfavourable Favourable
Class 1 1.2 0.9 1.2
Class 2 1.2 0.9 1.3
Class 3 1.2 0.9 1.5
b Only permanent loads 1.35 0.9
c Serviceability state for all classes 1.0 1.0 1.0
For the calculations only the permanent loads will be taken into account. Therefore the unfavourable
loads (tensile forces and bending moments) will be multiplied by  = 1.35 and the favourable loads
(pressure forces) by  = 0.9.
7.4.LDesign [39]
Nowadays a variety of framework programs is available where simple beam spring models can be
modelled and calculated. Unfortunately creating a ring model is more complex in these programs.
For this, Dr. Ir. C. Blom decided to make an own framework program especially for tunnel linings,
called LDesign. This program calculates with the input the different loads acting on each point and
determines the normal forces, shear forces, bending moments and deformations for each point. In
Appendix F an explanation of the working of the program is given.
7.5.Output
With the generated input in LDesign (data of the Rijnlandroute are used) the model can be calculated
and the different forces in the tunnel lining can be presented by N, M and V lines (normal forces,
bending moments and shear forces). These forces will act when the segments are installed. During
installation, transportation or storage of the segments, other forces and bending moments will also
work on the segments. However, these forces are not taken into account.
To limit the amount of output three different sections have been analysed; two critical sections
which can be compared with the traditional method starting at approximately 1D cover. Furthermore
the minimum and maximum groundwater levels are taken into account as well.
7.5.1. Tunnel lining half in surface

As a first critical section the starting point of the tunnel (complete ring in permanent situation) has
been chosen. At this section the tunnel is bored through the improved first three soil layers (MIP
columns included) and the required embankment (to prevent buoyancy), see Table 17, Figure 45 and
Figure 46. For both groundwater tables (0.6 and 1.4 m below surface level) the calculations are
made.
A.H. Oldenhave 62
Figure 45: Situation for lowest groundwater level Figure 46: Situation for highest groundwater level
After calculation it appears that the lowest groundwater table (1.4 m below surface level) is decisive
for calculations. The N, M and V-lines are given in the figures below.
Groundwater level 1.4 meter below surface
Figure 47: Normal forces [39] Figure 48: Shear forces [39] Figure 49: Bending moments [39]
It is expected that the most onerous condition is located in the top of the lining (due to the extreme
values located in the top, see moment line) or at the bottom. To eliminate doubts the forces are
compared for each point in the lining, see tables below. Safety factors are included.
The forces are depending on the length of the longitudinal joint; when the longitudinal joint is more
opened i.e. lt is small, the deformations will increase but the forces will decrease. On the other hand,
when the tunnel is acting more stiff and the longitudinal joint remains closed (l t = 170 mm), the
forces will increase.
As previously mentioned the maximum allowable deformation of the tunnel lining caused by the
internal forces is stated as 50 mm. Moreover, according to Beijer (2010), when the length of the
longitudinal joint lt is 100 mm, water tightness is guaranteed.
Underneath the forces for as well as a joint length of 170 mm as for 100 mm is given. This will only
influence the magnitude. Therefore above figures are given for only one case.
A.H. Oldenhave 63
Table 20: Forces acting in the lining when lt = 170 mm Table 21: Forces acting in the lining when lt = 100 mm
Hsoil - Hwater = 1.4 m Hsoil - Hwater = 1.4 m

H=0.735 (dead weight included), lt = 170 mm H=0.735 (dead weight included), lt = 100 mm
Node i Node j N [kN] V [kN] Mi [kNm] Mj [kNm] Node i Node j N [kN] V [kN] Mi [kNm] Mj [kNm]
1 2 -95 6 113 -103 1 2 -85 5 90 -82
2 3 -94 19 103 -74 2 3 -85 16 82 -58
3 4 -93 33 74 -23 3 4 -85 27 58 -17
4 5 -92 39 23 36 4 5 -85 31 17 31
5 6 -95 13 -36 56 5 6 -89 5 -31 38
6 7 -102 7 -56 67 6 7 -98 0 -38 37
7 8 -109 18 -67 95 7 8 -106 16 -37 62
8 9 -115 13 -95 114 8 9 -113 13 -62 81
9 10 -124 -4 -114 108 9 10 -121 0 -81 81
10 11 -138 -47 -108 36 10 11 -134 -40 -81 20
11 12 -160 -92 -36 -105 11 12 -154 -82 -20 -105
12 13 -175 11 105 -88 12 13 -166 30 105 -59
13 14 -175 22 88 -54 13 14 -164 25 59 -21
14 15 -174 10 54 -39 14 15 -163 9 21 -7
15 16 -174 -10 39 -54 15 16 -163 -9 7 -21
16 17 -175 -22 54 -88 16 17 -164 -25 21 -59
17 18 -175 -11 88 -105 17 18 -166 -30 59 -105
18 19 -160 92 105 35 18 19 -154 82 105 20
19 20 -138 47 -35 108 19 20 -134 41 -20 81
20 21 -124 4 -108 114 20 21 -121 -0 -81 81
21 22 -116 -13 -114 95 21 22 -113 -13 -81 62
22 23 -109 -18 -95 67 22 23 -106 -16 -62 37
23 24 -102 -7 -67 57 23 24 -98 0 -37 38
24 25 -95 -13 -57 36 24 25 -89 -5 -38 31
25 26 -92 -39 -36 -23 25 26 -85 -31 -31 -17
26 27 -94 -33 23 -74 26 27 -85 -27 16 -58
27 28 -95 -19 74 -103 27 28 -85 -16 58 -82
28 1 -95 -6 103 -113 28 1 -85 -5 82 -90
ymax = 13 mm ymax = 17 mm
It is clear that despite higher deformations for lt = 100 mm (which is the most onerous situation,
smallest forces), the deformations remain below the limit of 50 mm and therefore these values will
be used for calculation.
7.5.1.1. Results
In Table 21 the decisive normal forces and bending moments are given. It appears that the bottom is
decisive (M=105 kNm). These values determine the minimum required reinforcement in each ring.
A.H. Oldenhave 64
The shear force determines the amount of reinforcement in longitudinal direction. Both calculations
are given in Appendix G.
Between the joints reinforcement is absent and bolts have to prevent failure caused by the forces
acting in the tunnel lining. Again, calculations are explained in Appendix H.
Reinforcement in ring direction

The required reinforcement is depending on the local forces which are different for every point in the
tunnel lining. To prevent construction failures the decisive forces will be used to determine the
required reinforcement.
According to the calculations the decisive points are located at node 12 and 18 (sides of the bottom).
Here the normal force amounts 154 kN (pressure force) and the bending moment 105 kNm.
Reference is made to Appendix G for more explanation about the calculations. With these forces the
minimum amount of reinforcement amounts 486 mm2 for one side. However, according to the
regulations a minimum reinforcement is always required to prevent a sudden collapse of the
concrete (see Table 23 for further explanation). The minimum reinforcement is depending on the
concrete class and the height of the element and amounts 726 mm2. For prevention of construction
failures reinforcement will be applied on both sides of the element. The total reinforcement A s,total
will be As times two (1452 mm2). This is equal to two rows (see figure) of 10 x 10. The length
between two bars is 92 mm which is lower than the maximum bar distance of 120 mm (see section
7.5.5). Also the diameter remains below the maximum of 13 mm (again see section 7.5.5) and
therefore the crack width will remain below the maximum width to protect the reinforcement.
Figure 50: Cross section segment with required ring reinforcement
Reinforcement in longitudinal direction

In order to create a rigid reinforcement cage, brackets will be applied as well. The dimensions are
depending on the shear forces in the tunnel lining. Again, in Appendix G the calculation to determine
this reinforcement is given.
The maximum shear force in the tunnel lining is 82 kN. When assuming a concrete class C35/45, the
maximum shear force without reinforcement is N (see Appendix G), so
no reinforcement is required. However, in order to guarantee dimensional stability of the
reinforcement cage, brackets will be applied with a maximum distance of 300 mm. To prevent
A.H. Oldenhave 65
construction failures 12 will be applied as well. By constructing 16 bars in one segment, the
distance between each bar is 282 mm. Again on both sides the longitudinal reinforcement will be
applied to connect the longitudinal reinforcement to the ring reinforcement, see figure below.
Figure 51: Cross section segment with required longitudinal reinforcement
Bolts in the longitudinal joints

When applying regular bolts, the pre-stressing force depends on the dimensions of the bolt. The high
tension forces in these bolts can result in failure due to inadequate forces of the concrete (bolts are
pulled out of the concrete). For the different types of bolt connections and its advantages reference
is made to the report of Beijer (2010) [14]. He concluded that the Japanese box method is the most
onerous bolt connection, especially when the bending moments with respect to the normal forces
become high. The other bolt types are not further taken into account. By applying this Japanese box
method, the forces will not be transferred through the concrete, but are connected with the
reinforcement, see Figure 52, Figure 53 and Figure 54. Therefore this connection can transfer higher
forces and especially higher bending moments. The steel box is poured into the concrete of the
segment. Per joint two boxes are applied.
Figure 52: Cross section Japanese box method, dimensions in mm Figure 53: Cross section Japanese box method with
[14] reinforcement bars (connected to applied reinforcement),
dimensions in mm [14]
A.H. Oldenhave 66
Figure 54: Japanese box method applied in tunnel segments [41]
For this connection two bolts and six reinforcement bars per box will be applied, in total two boxes
per joint (see Figure 53). The bolts are constructed in the middle of the steel plate to prevent
construction failures. The most generous location for the decisive point (the top) will be at the
outside of the lining, see moment line; however construction at this location is impossible. By
applying the bolts in the centre of the lining, these bolts can counteract the present bending moment
in the joint. The rotation point is located at the centre of the pressure zone of the concrete.
However, after calculating the reinforcement and its pressure zone, it was stated that the pressure
zone of the concrete is small with respect to the total thickness of the lining. For this calculation it
will be assumed that the length between the bolts and the rotation point is ½ h - 10 mm (190 mm).
The pressure forces will be transferred through the surrounding concrete.
In Appendix H the total calculation is given. For this cross section M30 bolts of class 8.8 will satisfy,
together with a steel plate with a thickness tp of 20 mm. The six reinforcement bars behind the steel
plate are of 20 (see figure).
Bolts between the rings

When bolts will not guarantee balance of the single ring, the forces will be transferred to the
adjacent rings. However, in this case the bolts can guarantee the balance of the system. Bolts
between the rings will only be applied to prevent shearing of one ring with respect to adjacent rings
and maintain water tightness. The regular oblique bolts will be used, see Figure 55, Figure 56 and
Figure 57.
A.H. Oldenhave 67
Figure 55: Applied bolts Figure 56: Location and Figure 57: Oblique bolt [14]
in Sluiskiltunnel number of bolts in
Sluiskiltunnel
According to the TNO report of Siemes (2001) [8] the maximum friction force between two profiles
during construction is 50 kN/m. Per element this is equal to 236 kN. When the bolts can guarantee
this force, the water tightness will remain. The oblique bolts will be constructed under an angle of 30
degrees, see Figure 57; therefore the required force is √ kN. According to Figure 56
above, four bolts between the rings will be placed over one element. Therefore the total force of one
bolt has to be 66 kN. An M20 of class 6.4 will satisfy.
When also taking into account the maximum absorbing bonding stress of the concrete (prevention of
reinforcement pulled out of the concrete), the minimum length can be determined (safety factors
included, for calculation see Appendix H.2). The length of the bolts will be 450 mm.
Summary
Underneath in Table 22 the minimum reinforcement and bolts are given for one segment.
Table 22: Requirements for one segment
As ring 10 x 10 x 2
As longitudinal 16 x 12 x 2
Bolts in longitudinal joints 2 x M30, class 8.8
tp 20 mm
reinforcement 20 mm
Bolts in ring joint 4 x oblique bolt M20, class 6.4, length 450 mm
The amount of required measures (reinforcement and bolts) can be found in the table above. In this
table the minimum and maximum reinforcement is used as well. In order to prevent a sudden
collapse (concrete behaves brittle) a minimum reinforcement is always necessary. In the table below
the values for the minimum reinforcement percentage are given for each concrete quality and
height. According to this table the minimum amount of reinforcement per side is 0.22% =
mm2 = As,min (see Appendix G).
A.H. Oldenhave 68
Table 23: Minimum and maximum reinforcement [12]
min (%) max (%)

h = 200 mm 400 mm 600 mm 800 mm > 1000 mm
C20/25 0.19 0.16 0.14 0.135 0.13 1.38
C28/35 0.23 0.19 0.17 0.16 0.15 1.93
C35/45 0.26 0.22 0.20 0.18 0.17 2.48
C45/55 0.30 0.24 0.22 0.21 0.20 3.03
C53/65 0.33 0.27 0.25 0.23 0.22 3.59
A maximum reinforcement is also necessary to prevent failing of the pressure zone of the concrete
before the yield stress is reached. The maximum reinforcement is independent from height and given
in the last column of Table 23. This leads to a maximum As =
mm2 for each side of the column.
At last when deformations in the tunnel lining become too high, this can cause problems with the
alignment. According to the outcome (Table 20 and Table 21) these deformations remain below the
maximum of 50 mm.
7.5.2. Tunnel lining with smallest cover

The second critical section is the location where the tunnel lining is located under surface with the
lowest cover on top; see Figure 58 and Figure 59. The tunnel lining now rests on the Pleistocene and
therefore the ground improvement is not necessary any more. Here the undisturbed soil will be
assumed again resulting in a low mean bedding stiffness Eoed. The embankment on top works
efficient to decrease unfavourable forces in the lining.
Figure 58: Situation for highest groundwater level Figure 59: Situation for lowest groundwater level
After calculation it appears that without ground improvement the values become too high.
The mean Eoed without MIP columns amounts 8.4 MPa but with ground improvement it increases to
92.7 MPa. This results in stiffer soil springs. Moreover, the soil weights also increase due to ground
improvement. The calculation without ground improvement results in a displacement of 259 mm
which is absolutely not acceptable any more. It can be concluded that ground improvement is not
only necessary to limit the settlements but also to reduce the deformations in the tunnel lining and
A.H. Oldenhave 69
to guarantee feasibility of construction of the tunnel lining. Thus, for this cross section ground
improvement will be applied as well.
For this situation the highest groundwater table (0.6 m below surface) is decisive. Dead weight is
included and with LDesign the following data is calculated. See figures below and Table 24 and Table
25.
A.H. Oldenhave 70

1 2 -406 24 278 -241 1 2 -388 22 256 -222
2 3 -398 71 241 -132 2 3 -382 64 222 -123
3 4 -384 109 132 35 3 4 -369 99 123 28
4 5 -366 136 -35 243 4 5 -353 122 -28 215
5 6 -349 105 -243 404 5 6 -339 95 -215 360
6 7 -346 10 -404 420 6 7 -337 12 -360 379
7 8 -380 -251 -420 35 7 8 -366 -201 -379 70
8 9 -438 -220 -35 -302 8 9 -414 -173 -70 -194
9 10 -462 78 302 -182 9 10 -435 59 194 -104
10 11 -456 65 182 -82 10 11 -434 42 104 -40
11 12 -452 53 82 -1 11 12 -434 30 40 6
12 13 -449 40 1 59 12 13 -435 22 -6 39
13 14 -446 26 -59 98 13 14 -436 15 -39 61
14 15 -445 9 -98 112 14 15 -436 5 -61 69
15 16 -445 -9 -112 98 15 16 -436 -5 -69 61
16 17 -446 -26 -98 59 16 17 -436 -15 -61 39
17 18 -449 -40 -59 -1 17 18 -435 -22 -39 6
18 19 -452 -53 1 -82 18 19 -434 -30 -6 -40
19 20 -456 -65 82 -182 19 20 -434 -42 40 -104
20 21 -462 -78 182 -302 20 21 -435 -59 104 -194
21 22 -438 220 302 35 21 22 -414 173 194 70
22 23 -380 251 -35 420 22 23 -366 201 -70 379
23 24 -346 -10 -420 404 23 24 -337 -12 -379 360
24 25 -349 -105 -404 243 24 25 -339 -95 -360 215
25 26 -366 -136 -243 35 25 26 -353 -122 -215 28
26 27 -384 -109 -35 -132 26 27 -369 -99 -28 -123
27 28 -398 -71 132 -241 27 28 -382 -64 123 -222
28 1 -406 -24 241 -278 28 1 -388 -22 222 -256
According to the outcome, the deformations, when assuming a longitudinal joint opening of 100 mm,
are too high (> 50 mm). Therefore, the values when assuming an l t of 170 mm will be used for
calculation. It is clear that this cross section is decisive for determination of the feasibility to
construct a tunnel lining at shallow depth.
In Table 24 the maximum bending moment amounts 420 kNm and the associated normal force
amounts 346 kN. To determine whether this normal force reduces the bending moment, the stresses
due to these forces are calculated:
N/mm2
A.H. Oldenhave 71
kNm.
According to above mentioned calculation the bending moment will not decrease despite the
occurring normal force. Therefore the bending moment will be used for calculation only.
7.5.2.1. Results
The amount of reinforcement and bolts are determined with the measures.

The decisive point is located at the sides where the bending moment amounts 420 kNm and the
normal pressure force 346 kN. These forces result in a required reinforcement area of 2442 mm2,
which is higher than the minimum reinforcement. 13 Bars will be constructed with diameter 16
mm, on both sides (see Figure 63). The distance between two bars is 60 mm. The diameter is higher
than 13 mm, but the distance between two bars remains below the maximum bar distance of 120
mm. Therefore reinforcement will not be affected (again, see section 7.5.5).
Figure 63: Cross section segment with required ring reinforcement

The maximum shear force is 251 kN which cannot be incorporated by the concrete any more
(maximum shear strength of concrete is 217.8 kN). The remaining 33.2 kN has to be incorporated by
the reinforcement, equal to 76 mm2. However, this is smaller than the minimum reinforcement.
Therefore, again minimum reinforcement for the longitudinal direction will be applied (726 mm 2, 16
x 12), on both sides of the element, given in Figure 51.

In this cross section the maximum bending moment is 420 kNm, which has to be incorporated by the
bolts. Again the Japanese box will be applied. In this case M48 bolts of class 10.9 will satisfy, together
with a steel plate of thickness tp 40 mm. This will increase the costs significantly. Therefore further
research is recommended to find the optimal solution. The six reinforcement bars become 40.

Again, oblique bolts M20 of class 6.4 will be applied to prevent shearing of one ring with respect to
adjacent rings and guarantee water tightness (by rubber gaskets between the rings under
compression). The length amounts 450 mm to guarantee enough bonding strength.
A.H. Oldenhave 72
Summary
Comparing this cross section with previous critical cross section shows that this cross section will be
the decisive section. The forces are higher resulting in more reinforcement and bolts, given in Table
26 for one segment.
Table 26: Requirements for one segment
Bolts in longitudinal joints 2 x M48, class 10.9
tp 40 mm
reinforcement 40 mm
Bolts in ring joint 4 x oblique bolt M20, class 6.4, length 450 mm
7.5.3. Tunnel lining with 1D cover

In order to map the differences between the traditional method and the method with low covers, the
traditional start is taken as the third cross section. The section is schematised in Figure 64 and Figure
65. The lining is located in the Pleistocene silty sand layer. Therefore the mean saturated weight and
Eoed increase again. The embankment on top and the soil replacement is not necessary any more.
Figure 64: Situation for highest groundwater level Figure 65: Situation for lowest groundwater level
The highest groundwater table is decisive again. Dead weight is included and the data given in the
figures below and Table 27 and Table 28 are calculated with LDesign.
A.H. Oldenhave 73
When reinforcement is required, the top of the lining will be decisive (due to crown reduction).
However, at this depth the traditional method starts. Therefore it is expected that less or even no
reinforcement is required to guarantee stability in final situation (due to adequate normal forces to
balance the bending moments).
A.H. Oldenhave 74

1 2 -1101 14 125 -109 1 2 -1089 12 112 -98
2 3 -1099 41 109 -62 2 3 -1088 36 98 -58
3 4 -1095 67 62 15 3 4 -1085 58 58 8
4 5 -1089 91 -15 118 4 5 -1081 79 -8 97
5 6 -1090 3 -118 122 5 6 -1083 3 -97 101
6 7 -1102 -44 -122 72 6 7 -1094 -37 -101 59
7 8 -1119 -56 -72 9 7 8 -1110 -46 -59 7
8 9 -1136 -45 -9 -42 8 9 -1126 -37 -7 -35
9 10 -1151 -22 42 -67 9 10 -1140 -20 35 -58
10 11 -1161 5 67 -61 10 11 -1150 3 58 -54
11 12 -1167 25 61 -33 11 12 -1156 22 54 -29
12 13 -1168 33 33 5 12 13 -1158 30 29 4
13 14 -1168 27 -5 36 13 14 -1158 25 -4 33
14 15 -1167 11 -36 48 14 15 -1158 10 -33 44
15 16 -1167 -11 -48 36 15 16 -1158 -10 -44 33
16 17 -1168 -27 -36 5 16 17 -1158 -25 -33 4
17 18 -1168 -33 -5 -33 17 18 -1158 -30 -4 -29
18 19 -1167 -25 33 -61 18 19 -1156 -22 29 -54
19 20 -1161 -5 61 -67 19 20 -1150 -3 54 -58
20 21 -1151 22 67 -42 20 21 -1140 20 58 -35
21 22 -1136 45 42 9 21 22 -1126 37 35 7
22 23 -1119 56 -9 72 22 23 -1110 46 -7 59
23 24 -1102 44 -72 122 23 24 -1094 37 -59 101
24 25 -1090 -3 -122 118 24 25 -1083 -3 -101 97
25 26 -1089 -91 -118 15 25 26 -1081 -79 -97 8
26 27 -1095 -67 -15 -62 26 27 -1085 -58 -8 -58
27 28 -1099 -41 62 -109 27 28 -1088 -36 58 -98
28 1 -1101 -14 109 -125 28 1 -1089 -12 98 -112
7.5.3.1. Results

With a maximum bending moment of 112 kNm and a normal force of 1089 kN, no reinforcement is
required any more, which was expected due to high pressure forces. However, minimum
reinforcement will be applied to prevent sudden collapse (10 x 10), see Figure 50.

Again minimum reinforcement will satisfy due to the small shear forces, see Figure 51 (726 mm2, 16 x
12).
A.H. Oldenhave 75

For this case the normal force is not negligible any more with respect to the bending moment (M/W
+ N/A = 4.2 – 2.7, see Figure 69). Therefore only a part of the bending moment has to be
compensated by bolts.
Figure 69: Stress distribution
The remaining bending moment which has to be incorporated by the bolts is ( )

( ) kNm. For this case the length of the pressure zone of the
concrete is 70 mm, so the rotation point will be located 35 mm below the end. By applying two bolts
in the middle, the mean force of the two bolts is kN. Two M20 bolts of class 8.8 will
satisfy, however when taking into account the bonding strength, the bolts have to be of M30, length
500 mm.

Again oblique bolts M20, class 6.4, length 450 mm will be applied at four locations per segment to
maintain water tightness of the tunnel lining and keep the elements in place during relaxation of the
jacking forces.
Summary
In the table below again the minimum reinforcement as well as the dimensions of the bolts per
segment are given.
Table 29: Requirements per segment
Bolts in longitudinal joints 2 x M30, class 8.8, length 500 mm
Bolts in ring joint 4 x Oblique bolt M20, class 6.4, length 450 mm
7.5.4. Conclusion
It can be concluded that construction at sufficient depth will decrease the amount of required
measures. Regular bolts will satisfy which will decrease the costs as already expected.
On the next page the total measures, required to prevent collapse of the tunnel lining, are
schematized in one figure, see Figure 70. It has to be remarked that due to the fact that only three
cross sections are used for calculation, the length of the zones are a rough estimation. Further
research is recommended.
A.H. Oldenhave 76
Figure 70: Overview requirements for tunnel lining
A.H. Oldenhave 77
7.5.5. Cracking
As already mentioned, the design of the reinforcement is depending on different criteria. The most
decisive criterion is the cracking width.
When reinforced concrete segments are subjected to bending, the concrete will crack before the
reinforcement will adopt the forces. If these cracks will become too wide, this can influence the
water tightness and the protection zone of the reinforcement (this can result in rust of the
reinforcement). A good crack distribution (more cracks, smaller crack width per crack) will reduce
these influences. The amount of cracks is depending on the bonding strength between the steel and
concrete. To increase this bonding the reinforcement bars can be roughened. Moreover, more bars
and a higher amount of reinforcement will also increase the bonding. This taken into account by the
detailing of the reinforcement can influence the amount of cracks and therefore the water tightness
of the tunnel lining.
In the reader of Walraven (2006) tables are given for the maximum allowable bar diameter and
maximum bar distance. If these requirements are met, it can be assumed that the crack width is
permissible. The tables are depending on the environmental classes. For a tunnel lying (partly) below
groundwater level, the environmental class is XC2: exposure caused by carbonation; wet, rarely dry.
A mean steel stress of 300 N/mm2 will be a reliable value for the serviceability limit state. This gives
according to Table 30 and Table 31 a bar diameter km of 13 mm and a bar distance s of 120 mm. In
theory only one of the requirements is enough to assume that the crack width will remain below the
maximum width.
Table 30: Maximum allowable mean bar diameter km [mm]
Environmental Steel stress s [N/mm2]

Classes 100 125 150 200 250 300 350 400
XC1 50 40 33 25 20 17 14 12
XC2 – XC4 37 30 25 19 15 13 11 9
XF1 and XF3
XD1 – XD3
XS1 – XS3 25 20 17 13 10 8 7 6
XF2 and XF4
XA1 – XA3
Table 31: Maximum allowable bar distance s [mm]
Environmental Steel stress s [N/mm2]

Classes 100 125 150 200 250 300 350 400
XC1 870 670 540 370 270 200 155 120
XC2 – XC4 620 470 370 245 170 120 85 55
XF1 and XF3
XD1 – XD3
XS1 – XS3 370 270 200 120 70 40 - -
XF2 and XF4
XA1 – XA3
A.H. Oldenhave 78
8. Costs and planning

In order to get insight in the efficiency of the tunnelling with marginal cover method with respect to
the traditional method, the costs and planning are given in this chapter.
8.1.Costs
By adding the differences in costs between the traditional method and the tunnelling with marginal
cover method, insight in the efficiency of this new method is given. For a rough estimate of the
differences in total costs, the Sluiskiltunnel project is used as reference [42]. In this thesis only the
difference in percentages will be mentioned, where the costs of the traditional method will be used
as reference year.
8.1.1. Differential costs

When comparing the construction costs of both methods, the difference amounts 3.3 %, see Table
32 (“-” indicates smaller costs for marginal cover method, “+“ indicates higher costs for marginal
cover method). Due to the early start of the bored process the costs of the measures before
tunnelling will decrease with respect to the construction of the ramps. However, the high tension
forces in the tunnel lining (due to small or no cover on top) result in an increase of the costs of the
bored part. Still, when applying the tunnelling with marginal cover method, the construction costs
will decrease with approximately 3.3 %. The values do not differ much, however it is expected that
the building time will differ more significantly. In the next paragraph a planning for the building time
is given.
Table 32: Summary costs and differences between both methods
% of total Traditional Tunnelling with Differences

method marginal cover method
Total ramps 21 % 5% - 76 %
Total tunnelling 53 % 66 % + 20 %
Total additional 26 % 29 % +9%
Total 100 % 100 % -3%
Contingencies etc. 63.36%

TOTAL BUDGET 163.36 % 163.36 % - 3.3 %
Difference in % - 3.3 %
8.2.Planning
With previous paragraph it can be concluded that the total costs between both methods do not
differ much. However, the building time of a method also influences the efficiency (and the costs).
With the new method (tunnelling with marginal covers) the tunnelling process already starts at
surface level and deep ramps are not necessary any more. This decreases the building time.
For both methods a planning is made [43]. In Appendix I the total building time for each method is
given, including all required proceedings. In Table 33 both methods are summarized and in Figure 71
the critical paths are shown. The critical path implies that these proceedings determine the minimum
construction time due to requirement that the first proceeding is finished before the start of the next
A.H. Oldenhave 79
proceeding. The preparations include the procurements of the TBM. In the Appendix the planning is
more precisely explained.
Table 33: Comparison building time traditional and “tunnelling with marginal cover” method (1)
Traditional method
Start Work days Calendar days End date
Total building time 1-1-2015 1065 1491 31-1-2019
Preparations 1-1-2015 306 428 4-3-2016
Bored part 4-3-2016 576 624 18-11-2017
Finishing ramps 18-11-2017 314 439 31-1-2019
Tunnelling with marginal cover method

Start Work days Calendar days End date
Total building time 1-1-2015 809 1133 6-2-2018
Preparations 1-1-2015 215 302 29-10-2015
Bored part 29-10-2015 634 682 10-9-2017
Tunnel lining removal 16-9-2017 102 143 6-2-2018
Figure 71: Comparison building time traditional and tunnelling with marginal cover method (2)
It can be concluded that the “tunnelling with marginal cover” method will take less time to build. The
total building time when constructing the tunnel with the traditional method will amount 4 years and
1 month. When applying the tunnelling with marginal cover method, the total building time will
amount 3 years and 1 month. This represents a gain of 1 year when starting the tunnelling
proceeding at surface level. Especially the finishing of the tunnel results in the smaller construction
time of the tunnelling with marginal cover method, see figure above.
8.3. Conclusion
With above mentioned costs and planning, it can be concluded that the tunnelling with a marginal
cover method will be more efficient due to less building time. The costs however will not differ very
much. Still, when construction time will be less for the tunnelling with marginal cover method, this
A.H. Oldenhave 80
will also decrease the total construction costs. Moreover, the indirect costs are not taken into
account. These costs include the costs, depending on the building time (wage of staff, rental costs of
equipment, etc.). Probably when applying the new method (tunnelling with marginal covers), this will
result in smaller indirect costs (due to smaller time of construction). Moreover, the site is smaller and
less staff is required during the tunnelling process than during construction of the ramps. For this
thesis only the construction costs are taken into account. When the indirect costs will be calculated
as well, the tunnelling with marginal costs method will probably differ more than 3.3 % with the
traditional method.
A.H. Oldenhave 81
9. Conclusion and recommendations

In this chapter the conclusions and recommendations for the application of the “tunnelling with
marginal cover” method for large diameters in the Netherlands are given.
9.1.Conclusion
In Chapter 4 nine expected complications when applying the tunnelling with marginal cover method
in the Netherlands were mentioned:
1. Effects on deformations in Holocene layer;

2. Effects on buoyancy of bored tunnel;
3. Effects on tunnel lining;
4. Effects on bore front stability;
5. Effects on EPB driving through Pleistocene sand;
6. Effects on torque of TBM;
7. Effects on grouting of the tail void;
8. Effects on launching procedure;
9. Effects on cross connections.
The first three consequences are of major interest when the feasibility of the new method (tunnelling
with marginal covers) will be determined. Therefore these three were further explored for the scope
of this thesis (see Figure 9). It can be concluded that the remaining six effects will not result in
complications for the construction of the tunnel with the tunnelling with marginal cover method.
Starting with the soil deformations in the Holocene layer, it can be concluded that ground
improvement is necessary when starting the tunnelling process on surface level. This is due to the
fact that especially in the western part of the Netherlands the first 10 – 20 meters consists of
Holocene deposits (soft soils, clay, peat). When applying a TBM on top of the Holocene layer, this will
result in high deformations that are not allowed for the control of the starting process. For this case
especially the last layer in the Holocene was very soft. Only when the bottom of the TBM reaches the
Pleistocene layer, which is a strong sand layer, the deformations will remain below limits without
ground improvements. Due to experience in applying MIP (Mixed In Place) columns for decreasing
settlements in the soil, this ground improvement is recommended.
A second and maybe the main complication in the Netherlands due to the high groundwater tables
are the buoyancy forces acting on the tunnel lining. When a cover on the tunnel lining is not yet
present, the upward forces cannot be counteracted by the soil and tunnel lining. Therefore measures
have to be applied. A sand ballast in the bottom of the tunnel lining with a maximum of 2.5 meters
and an embankment on top is recommended. Other kinds of ballast can be applied as well. However,
due to increase of costs it will not be advised. During construction of the tunnel the groundwater
level will be lowered temporarily, because of the missing ballast inside the tunnel during construction
(ballast will be applied when TBM is at sufficiently distance).
Constructing a tunnel without cover will also result in complications of construction of the tunnel
lining. The soil forces acting on the lining are absent on top. This results in high bending moments in
the lining and a missing normal force. To prevent open joints between the segments (and infill of
groundwater) bolts of sufficient dimensions have to be applied. For the shallow locations of the
tunnel normal bolts will not satisfy. For these locations the longitudinal joints will be connected using
A.H. Oldenhave 82
the Japanese box connection. This connection ensures force transmission directly towards the
reinforcement resulting in the applicability of transmitting high bending moments. However, other
bolt types are not taken into account and may be possible as well.
It can be concluded that despite above mentioned main complications the tunnelling with marginal
cover method is feasible in the Netherlands. However, despite the omitted building pits, the
construction costs will not differ for the two methods. Still, the construction time of the tunnelling
with marginal cover method will be almost one year shorter than the traditional method. Therefore it
can be concluded that application of the tunnelling with marginal cover method in the Netherlands
will be feasible and efficient.
9.2.Recommendations
As previously mentioned, this thesis is only intended to determine the adaptation of the tunnelling
with marginal cover method for large diameters in the Netherlands by applying this method for a
specific reference project. The optimal solution has not yet been reached. The calculations are made
in the same order as Chapter 4. This means first the settlements are calculated, determining the
input (soil structure) for further calculation. Subsequently, the measures to prevent uplift of the
tunnel lining are determined, resulting in a required embankment and temporarily lowering of the
groundwater table. The soil structure and embankment are then used for the calculations of the
tunnel lining. These three subjects are the main part of this thesis. After determination of the
required measures for these three subjects other calculations were made, see Chapter 4. It appeared
that some outcomes resulted in new input for the first three calculations (expanded embankment
necessary for the jacking forces for example). However, this thesis is only a first rough research to
determine if the tunnelling with marginal cover method is plausible for a Dutch situation. To
determine the optimization when applying this method in the Netherlands, further research is strictly
recommended. However, the scope of this thesis is to determine the feasibility of the tunnelling with
marginal cover method in the Netherlands. This report cannot be used as a design guide for the
applied method.
For this thesis some limitations are considered. Primarily, when for this thesis only the Rijnlandroute
(even only the eastern side) is considered, it cannot be concluded that the new method (tunnelling
with marginal covers) is feasible for every location in the Netherlands. When this method will be
applied in the Netherlands, again a feasibility research is recommended for the specific location in
the Netherlands. Moreover, due to inadequate geotechnical data of the soil, a rough assumption of
the soil structure is made. When applying the tunnelling with marginal cover method at this location,
site investigation (CPT’s, groundwater measurements) is recommended.
In the scope of this project only three complications are explored in detail. When applying the
tunnelling with marginal cover method, further research for the other complications will be
recommended as well. Moreover, the tunnel lining analysis is only executed in 2D for a single ring. It
will be recommended to apply the calculation with continuum models and 3D models as well.
Additionally, for the tunnel lining analysis only three cross sections were considered. More
calculations for extra cross sections will be suggested. This can also result in smaller application areas
of the measures required for this complication. Moreover, other bolt types are not considered in this
thesis. It will be recommended to analyse if another bolt type instead of the Japanese box will also be
probable. This can reduce the construction costs as well.
A.H. Oldenhave 83
To conclude, if this method will be recommended in future in the Netherlands, the risks have to be
determined as well. It is concluded that the total construction time will be decreased by one year
when applying the tunnelling with marginal cover method instead of the traditional method,
however when adapting a new method this will also entail additional risks. Therefore it cannot be
concluded yet if this method will be recommended instead of the traditional method.
A.H. Oldenhave 84
10. Sources
Books and reports

[1] Huiskes, G. & van der Kroef, R. (2005). Tunnel naar de toekomst. Rijkswaterstaat
[2] Heijboer, J., van den Hoonaard, J. & van de Linde, F.W.J. (2004). The Westerschelde Tunnel.
Approaching Limits. Lisse: Swets & Zeitlinger B.V.
[3] CSR Department, Head Office (2011). CSR report. Obayashi Corporation
[4] Movares (2012). Rijnlandroute tracedeel A44-A4. Boortunnel, korte variant. Rotterdam:
Movares
[5] Herrenknecht AG (2005). Final Design Report S-309. Mixshield 10.510 mm. Schwanau:
Herrenknecht
[6] Verruijt, A. (2005). Grondmechanica. 7th edition. Delft: VSSD
[7] Türke, H. (1993). Statik im Erdbau. 2nd edition. Berlin: Ernst & Sohn
[8] Siemes, A.J.M. (2001). Westerscheldetunnel – Onderzoek naar de relatie tussen reparaties
aan de tunnelwand en de levensduur. Delft: TNO
Readers
[9] Bakker, K.J., et al., (2008). Bored and Immersed Tunnels. Delft: TU Delft
[10]Broere, W. & van der Woude, S. (2010). Trenchless Technology. Delft: TU Delft
[11]Blom, C.B.M. (2009). Concrete linings for shield driven tunnels. 2nd edition. Delft: TU Delft
[12]Walraven, J.C. (2006). College gewapend beton. Delft: TU Delft
[13]Abspoel, R., Bijlaard, F.S.K. & Samsor, R. (2008). Constructieleer 2B. Deel Staalconstructies.
Delft: TU Delft
Papers and theses

[14]Beijer, A.J. (2010). Ultra Rapid Under Pass. The adaptation of the URUP method for the
Netherlands. Delft: TU Delft
[15]Fujiki, H., Nakumara, T. & Izawa, M. (2009). URUP (Ultra Rapid Under Pass) method. The first
shield tunneling method for launching and arrival at the ground level. ITA
[16]Tan, W.H., Lee, C.Y. & Sivadass, T. (2008). Behaviour of Clay-Steel Interfaces. ICCBT 2008.
page 11 – 20
[17]Azadegan, B. & Massah, J. (2012). Effect of temperature on adhesion of clay to steel.
Cercetări Agronomice, Vol. XLV, No. 2 (150) / 2012. Page 21 – 27
[18]Blom, C.B.M. (1995). Liggerwerking van boortunnels. Krachtswerking in de tunnelmantel door
ongelijkmatige zettingen. Delft: TU Delft
[19]Lengkeek (1997). Analyse grond – tunnelinteractie. Analytische en numerieke bepaling van de
veerstijfheid van de grond voor toepassing in ring- en liggermodel. Delft: TU Delft
[20]Blom, C.B.M. (2002). Background document “Lining behaviour – analytical solutions of
coupled segmented rings in soil”. Delft: TU Delft
[21]Bouygues Construction (2009). Detailed Method Statement for Excavation of Down Track
Tunnel. Hong Kong
[22]Taillebois, A. (1999). Segmental Gaskets. Tunnels et ouvrages souterrains. Recommandations
de l’AFTES. Page 167 – 169
A.H. Oldenhave 85
Internet sites
[23]Herrenknecht AG. (2011). TBM/Tunneling Video.
http://www.youtube.com/watch?v=qx_EjMlLgqY. Consulted 15-05-2013
[24]CSR News (2013). URUP (Ultra Rapid Under Pass) Method wins the Prime Minister’s Prize in
the Fifth Monodzukuri Nippon Grand Award Enable tunnel boring machine to be launched
and arrive at surface level. http://www.obayashi.co.jp/english/news/csr/2013101000.html
Obayashi Corporation. Consulted 21-02-2014
[25]Provincie Zuid Holland (2011). Rijnlandroute. http://www.zuid-holland.nl/rijnlandroute
Consulted 09-02-2014
[26]University of waterloo. Vertical Stress in a Soil Mass.
http://www.civil.uwaterloo.ca/maknight/courses/CIVE353/Lectures/stress-load.pdf
Consulted 08-06-2013
Regulations
[27]VRC. Veiligheidsrichtlijnen deel C (2009). Rijkswaterstaat
[28]Specifieke Aspecten TunnelOntwerp (2005). Rijkswaterstaat
[29]CUR 162 (1993). Construeren met grond. Grondconstructies op en in weinig draagkrachtige
en sterk samendrukbare ondergrond. Gouda
[30]NEN 6740 (2006). Geotechniek – TGB 1990 – Basiseisen en belastingen.
[31]CUR-report 2001-10 (2001). Diepe grondstabilisatie in Nederland. Handleiding voor
toepassing, ontwerp en uitvoering.
[32]CUR 2006-2 (2006). Innovatieve aardebaan. Snel gebouwd, blijvend vlak.
[33]CUR 199 (2001). Handreiking toepassing No-Recess technieken.
[34]CUR-report 95-7 (1995). Preadvies omtrent het gebruik van kalk-cementkolommen in
Nederland.
[35]ROK 1.0 (2011). Richtlijn Ontwerp Kunstwerken. Rijkswaterstaat
[36]NEN-EN 1992-1-1+C2 (1992). Eurocode 2: Ontwerp en berekening van betonconstructies –
Deel 1-1: Algemene regels voor gebouwen
Software
[37]TNO (2014). DINO loket. Data en Informatie van de Nederlandse Ondergrond.
www.dinoloket.nl Consulted 13-08-2013
[38]Geodan (2013). Actueel Hoogtebestand Nederland. http://ahn.geodan.nl/ahn/ Consulted 06-
11-2013
[39]Blom, C.B.M. (2002). LDesign (v5). Delft: TU Delft
Received information
[40]Pictures of Botlekspoortunnel of S. van der Woude
[41]Pictures of Japanese box method of A.J. Beijer
[42]Cost estimation of project Sluiskiltunnel of J. Schenk
[43]Planning estimation of project Sluiskiltunnel of S. Wierikx
A.H. Oldenhave 86
Appendices
In this chapter the appendices can be found, which contain more detailed calculations of the
complications mentioned in the report. This chapter contains:
 Appendix A: Methods to determine soil stresses

 Appendix B: CPT results
 Appendix C: Geological results
 Appendix D: Settlements of Holocene layer when applying soil replacement
 Appendix E: Buoyancy
 Appendix F: LDesign
 Appendix G: Calculation of reinforcement
 Appendix H: Calculation of bolts
 Appendix I: Planning
A.H. Oldenhave 87
Appendix A: Methods to determine soil stresses [6]

In this Appendix two methods will be explained which can be used to determine the soil stresses; the
method of Newmark’s chart and the Boston rule.
A.1. Newmark’s chart

To determine the stresses of the soil, caused by the weight of the TBM, the equilibrium of Boussinesq
is used. With this method the stress distribution into the soil is calculated and used as input to
calculate the differential settlements for each depth.
The method of Boussinesq normally applies to a uniformly loaded circular area. Depending on the
radius of the area the increase of stresses for each depth z can be determined with
( ) where √( ) and a is the radius.
An increase in radius means an increase in stress. With this in mind Newmark developed a drawing
with different radii, each ring subdivided in 20 boxes where every box equals 0.005  p, see Figure
72–Figure 75. When the stresses on a certain depth underneath an object have to be calculated,
Newmark’s chart can be used for a first estimation.
In the figures on the next page the dimensions of the TBM (L x B = 10.235 x 8.066 m) are drawn in
Newmark’s chart to a scale where distance AB on the chart equals the desired depth. In Figure 72
and Figure 74 the stress right under the centre is determined for different depths, in Figure 73 and
Figure 75 the stresses underneath a corner at a specific depth. By multiplying the amount of marked
squares by the area of the object n (in this case the TBM) with the stress on top and 0.005 which is
the stress of each box I (Influence area), the total stress caused by the extra load can be determined.
In formula: .
Now the stresses become:
 5 m below centre (Figure 72)

130 squares  zz = 130  0.005  115.7 = 75.2 kPa
 5 m below corner (Figure 73)

45 squares  zz = 45  0.005  115.7 = 26.0 kPa
 10 m below centre (Figure 74)

60 squares  zz = 60  0.005  115.7 = 34.7 kPa
 10 m below corner (Figure 75)

30 squares  zz = 30  0.005  115.7 = 17.4 kPa
A.H. Oldenhave 88
Figure 72: Stresses at z = 5 m below centre (130 squares) Figure 73: Stresses at z = 5 m below corner (45 squares)
Figure 74: Stresses at z = 10 m below centre (60 squares) Figure 75: Stresses at z = 10 m below corner (30 squares)
A.H. Oldenhave 89
Boston Rule
In the report this method is used to determine the soil stresses. The outcome can be found in the
table below.
Table 34: (increase of) stresses of the soil
' [kN/m3] z [m] 'z;0 [kPa] 'z [kPa] 'z [kPa]

14 0 0 115.7 115.7
14 1 14 93.77577 107.7758
4 2 18 77.55656 95.55656
4 3 22 65.2176 87.2176
4 4 26 55.61073 81.61073
4 5 30 47.98376 77.98376
4 6 34 41.82698 75.82698
4 7 38 36.78495 74.78495
4 8 42 32.60363 74.60363
4 9 46 29.09749 75.09749
4 10 50 26.12848 76.12848
Comparison
When the stresses, found with Newmark’s chart and the Boston rule, are drawn and compared in a
graph (see Figure 76), the Boston rule indeed gives a good estimation of the mean stress for each
depth. Therefore, to calculate the total deformation in the Holocene layer, these stresses will be
used. Only the increase of stresses (4th column inTable 34) will cause deformations.
Soil stresses at depth z = 5 m

80
Soil stress [kPa]
60
40
20 z=5 (Newmark)
0 z=5 (Boston)
0 2 4 6 8 10 12
Length [m]
Soil stresses at depth z = 10 m

40
Soil stress [kPa]
20
z=10 (Newmark)
0 z=10 (Boston)
0 2 4 6 8 10 12
Length [m]
Figure 76: Comparison soil stresses Boston and Newmark
A.H. Oldenhave 90
Appendix B: CPT results [37]

In Figure 77 and Figure 79 the locations of the different CPT’s can be found as well as the location of
the start of the bore process. The results of these six CPT’s can be found in Appendix B.1 - B.6. The
legend used for the CPT’s is given in Figure 78.
The figures show that CPT S30H02476 and CPT S30H02477 are closest to the start of the tunnelling
process. However, these penetration tests show stronger layers in the first 10 meters than the two
CPT’s measured in the polder (S30H00677 and S30H00674). Moreover, the closest penetration tests
are measured along a road, as well as S30H00659 and S30H00668. A ground improvement can be
expected. Therefore, these results will not contribute to the expected soil layers at the start of the
tunnelling process.
Figure 77: Overview CPT locations (bird-eye)
Figure 78: Legend of the CPT’s
Figure 79: Overview CPT locations
A.H. Oldenhave 91
B.1. CPT results location 1
Figure 80: CPT location 1
Table 35: Soil structure of location 1
Type of soil Layer Depth below surface [m] Thickness [m]
Man-made 1 0 – 3.1 3.1

Clay, silty 2 3.1 – 3.7 0.6
Sand, silty 3 3.7 – 4.6 0.9
Silt 4 4.6 – 5.4 0.8
Clay, peaty 5 5.4 – 5.9 0.5
Sand, clayey 6 5.9 – 8.8 2.9
Clay 7 8.8 – 9.1 0.3
Clay, peaty 8 9.1 – 9.8 0.7
Clay, silty 9 9.8 – 12.4 2.6
Peat 10 12.4 – 13.0 0.6
Sand, fine 11 13.0 – 20.0 7.0
A.H. Oldenhave 92
Man-made 1 0 - 0.3 0.3

Sand, fine 2 0.3 - 1.5 1.2
Sandy clay or loam 3 1.5 - 2.1 0.6
Sand, fine 4 2.1 - 3.0 0.9
Sandy clay or loam 5 3.0 - 3.2 0.2
Sand, silty 6 3.2 - 4.1 0.9
Peat 7 4.1 - 7.0 2.9
Sand, clayey 8 7.0 - 10.0 3.0
Silt 9 10.0 - 13.1 3.1
Peat 10 13.1 - 13.8 0.7
Sand, silty 11 13.8 - 20.0 6.2
A.H. Oldenhave 93
Peat 1 0 – 2.6 2.6

Clay, silty 2 2.6 – 3.1 0.5
Clay 3 3.1 – 4.1 1.0
Sand, clayey 4 4.1 – 6.0 1.9
Clay, silty 5 6.0 – 6.8 0.8
Sand, silty 6 6.8 – 7.7 0.9
Peat 7 7.7 – 11.6 3.9
Sand, silty 8 11.6 – 18.8 7.2
Sand, fine 9 18.8 – 20.4 1.6
Sand, silty 10 20.4 – 29.0 8.6
Sand, fine – medium 11 29.0 – 30.0 1.0
A.H. Oldenhave 94
Clay, silty 1 0 – 1.1 1.1

Peat 2 1.1 – 2.1 1.0
Silt 3 2.1 – 3.5 1.4
Peat 4 3.5 – 4.0 0.5
Silt 5 4.0 – 5.3 1.3
Sand, silty 6 5.3 – 7.7 2.4
Clay, peaty 7 7.7 – 8.3 0.6
Clay, silty 8 8.3 – 11.3 3.0
Peat 9 11.3 – 12.2 0.9
Sand, fine 10 12.2 – 18.0 5.8
A.H. Oldenhave 95
Man-made 1 0 – 0.4 0.4

Peat 2 0.4 – 3.7 3.3
Silt 3 3.7 – 7.6 3.9
Clay, peaty 4 7.6 – 8.0 0.4
Clay, silty 5 8.0 – 11.0 3.0
Peat 6 11.0 – 11.3 0.3
Sand, silty 7 11.3 – 12.5 1.2
A.H. Oldenhave 96
Peat 1 0 – 2.8 2.8

Clay, silty 2 2.8 – 5.0 2.2
Sand, clayey 3 5.0 – 6.2 1.2
Clay 4 6.2 – 10.3 4.1
Peat 5 10.3 – 10.7 0.4
Sand, silty 6 10.7 – 12.5 1.8
A.H. Oldenhave 97
Appendix C: Geological results [37]

Some geological results were found on DINO loket as well. Here, only a distinction between clay, peat
and sand is made. The 5 figures (Figure 87) were found in the polder near the entrance of the tunnel
(see Figure 88). These 5 figures together form a reliable cross-section of the soil, see Figure 86.
Together with the previous mentioned CPT results in the polder which show equal composition of
the soil, this leads to the estimation of the underground.
Figure 86: Geological information about location
Figure 87: Geological data of DINO loket (south-east to north-west) [37]
Figure 88: Location of the 5 geological results
A.H. Oldenhave 98
Appendix D: Settlements of Holocene layer when applying soil replacement

In this Appendix the settlements of the Holocene layer are given when applying a soil replacement of
the first three meters peat and replacing this soil by sand. Moreover, due to counteract the buoyancy
forces (see Chapter 5) an embankment has to be applied as well. For each depth the height of the
embankment can be found in the next Appendix. This case is an iterative process and therefore the
height of the embankment is calculated later and applied again for the settlements.
Underneath the total settlements of the Holocene layer is given when applying the TBM at different
depths. The TBM will start at 2.5 meters below surface. When the TBM rests on the Pleistocene layer,
the settlements are almost zero. Therefore it is proven that assuming the Pleistocene layer as
infinitely stiff is a good assumption.
Table 41: Settlements when applying TBM at depth 2.5 meters. Soil replacement and embankment of 4.19 meters
Soil type Layer Thickness Mean ' 'z;0 'z 'z Cp Cs [-] h
[m] depth [m] [kN/m3] [kPa] [kPa] [kPa] [-] [m]
Sand (gone) 1 5.19 2.595 18 47 0 47 600 0
Sand (gone) 2 1.5 5.94 10 101 0 101 600 0
Sand 3 0.5 6.94 10 111 266 377 600 0
Clay, silty 4 1 7.69 7 117 228 345 15 160 0.08
Silt 5 3 9.69 9 134 158 292 45 1300 0.05
Clay 6 4 13.19 7 161 95 257 7 30 0.36
Sum h 0.49
Table 42: Settlements when applying TBM at depth 4 meters. Soil replacement and embankment of 4.19 meters
Sand (gone) 1 5.19 2.595 18 47 0 47 600 0
Sand (gone) 2 1.5 5.94 10 101 0 101 600 0
Sand 3 0.5 6.94 10 111 0 111 600 0
Clay, silty 4 1 7.69 7 117 0 117 15 160 0
Silt 5 3 9.69 9 134 201 335 45 1300 0.06
Clay 6 4 13.19 7 161 114 275 7 30 0.41
Sum h 0.47
Table 43: Settlements when applying TBM at depth 7 meters. Soil replacement and embankment of 3.44 meters
Sand (gone) 1 4.44 2.22 18 40 0 40 600 0
Sand (gone) 2 1.5 5.19 10 87 0 87 600 0
Sand 3 0.5 6.19 10 97 0 97 600 0
A.H. Oldenhave 99
Clay, silty 4 1 6.94 7 103 0 103 15 160 0

Silt 5 3 8.94 9 120 0 120 45 1300 0
Clay 6 4 12.44 7 148 163 311 7 30 0.57
Sum h 0.57
Sand (gone) 1 4.24 2.12 18 38 0 38 600 0
Sand (gone) 2 1.5 4.99 10 84 0 84 600 0
Sand 3 0.5 5.99 10 94 0 94 600 0
Clay, silty 4 1 6.74 7 100 0 100 15 160 0
Silt 5 3 8.74 9 117 0 117 45 1300 0
Clay 6 0.53 10.505 7 132 0 132 7 30 0
Clay 7 3.47 12.505 7 146 168 314 7 30 0.51
Sum h 0.51
Sand (gone) 1 3.84 1.92 18 35 0 35 600 0
Sand (gone) 2 1.5 4.59 10 77 0 77 600 0
Sand 3 0.5 5.59 10 87 0 87 600 0
Clay, silty 4 1 6.34 7 93 0 93 15 160 0
Silt 5 3 8.34 9 110 0 110 45 1300 0
Clay 6 1.53 10.605 7 128 0 128 7 30 0
Clay 7 2.47 12.605 7 142 177 319 7 30 0.38
Sum h 0.38
Sand (gone) 1 3.43 1.715 18 31 0 31 600 0
Sand (gone) 2 1.5 4.18 10 69 0 69 600 0
Sand 3 0.5 5.18 10 79 0 79 600 0
Clay, silty 4 1 5.93 7 85 0 85 15 160 0
Silt 5 3 7.93 9 102 0 102 45 1300 0
A.H. Oldenhave 100

Clay 6 2.53 10.695 7 125 0 125 7 30 0

Clay 7 1.47 12.695 7 139 187 325 7 30 0.24
Sum h 0.24
Sand (gone) 1 2.99 1.495 18 27 0 27 600 0
Sand (gone) 2 1.5 3.74 10 61 0 61 600 0
Sand 3 0.5 4.74 10 71 0 71 600 0
Clay, silty 4 1 5.49 7 77 0 77 15 160 0
Silt 5 3 7.49 9 94 0 94 45 1300 0
Clay 6 3.53 10.755 7 120 0 120 7 30 0
Clay 7 0.47 12.755 7 134 198 332 7 30 0.08
Sum h 0.08
Table 48: Settlements when TBM rests on Pleistocene layer (depth = 11 meters). No ground improvement

[m] depth [m] [kN/m3] [kPa] [kPa] [kPa] [-] [-] [m]
Peat 1 1 0.5 12 6 0 6 7.5 30 0
Peat 2 2 2 2 14 0 14 7.5 30 0
Clay, silty 3 1 3.5 7 20 0 20 15 160 0
Silt 4 3 5.5 9 37 0 37 45 1300 0
Clay 5 4 9 7 64 0 64 7 30 0
Pleistocene 6 0.47 11.235 7 80 194 273 600 0.001
Sum h 0.001
A.H. Oldenhave 101

Appendix E: Buoyancy
When calculating the amount of upward forces and the amount of downward forces acting on the
tunnel, it can be calculated if buoyancy will act (upward forces > downward forces). This is the so
called uplift calculation.
Another effect which may occur is the break-up failure, see Figure 89. This failure occurs when the
horizontal ground pressures become too high with respect to the vertical ground pressures. When
the shear force of the wedge on top of the tunnel is too small, the tunnel does not have counter
pressure on top and the tunnel lining will ovalize as presented in the figure. In this case however the
tunnel is located very shallow in the soil, resulting in small ground pressures. Moreover in most cases
a cover on top is even absent. It will be assumed that the break-up failure will not be decisive for the
shallow applied locations of the tunnel lining. Therefore only the uplift calculation is executed.
Figure 89: Schematisation of break-up of a tunnel [9]
E.1. Calculation of buoyancy [9]

In the figure below (Figure 90) the forces acting on a tunnel is schematized for traditional situation.
The weight of the tunnel G1 and the weight of the soil cover G2 have to counteract the uplift force A.
However, when the tunnel will be located without any soil cover, it is expected that the uplift fore A
becomes too high to be counteracted by the downward forces. Measures have to be taken into
account to prevent uplift for shallow depths.
Figure 90: Forces acting on tunnel under groundwater

Figure 91: Dimensions tunnel
level [9]
A.H. Oldenhave 102

In Figure 91 the situation for the Rijnlandroute is schematized where the tunnel is located just below
the highest occurring groundwater level (decisive for these calculations). In this figure also the ballast
is drawn. The maximum height amounts 2.5 meters. By applying different kind of ballasts (with
different unit weights) the remaining upward pressure is different for these situations. Therefore for
each situation and each depth separately the forces are calculated, see next paragraph.
By summing the uplift force A (negative), the weight of the tunnel lining G1, the ground force G2 and
the ballast, the remaining upwards force has to be compensated by an embankment on surface level.
Safety factors however have to be taken into account in these calculations (see Chapter 5).
Underneath the formulas are given (for this cross section as example) which are used to calculate the
total uplift force (see Table 50 where top of tunnel is 1 meter below surface). For G2 the soil next to
the top of the tunnel also has to be taken into account. By calculating the area of that part of the
tunnel (given in blue in the formula), the wet soil area can be calculated.
kN/m
( ) ( ) kN/m
kN/m
( ) ( ) ( )
( ) (( ) ) ( ) kN/m
( ( )) ( )
( ( )) ( )
kN/m
( ) ( )
Where ( ) ( ) degrees
kN/m
The height of the embankment can now be calculated:
( ) ( )
m.
E.2. Buoyancy tables

Underneath the output for the temporary and permanent situation for different ballasts are given.
For each depth in steps of one meter the downward and upward forces are calculated. The decrease
of groundwater table in permanent situation, the embankment and the ballast is taken into account.
Moreover, according to the ROK [35], a temporary removal of a soil layer with a thickness of 0.5
meter above the tunnel should be taken into account (see Chapter 6). In the tables for the
permanent situation this measure is included in the outcome. The red values show that despite this
norm an embankment is still not required.
A.H. Oldenhave 103

Table 49: Temporary situation with safety factors
Calculation for smallest height of embankment

Tempo Top of Depth tunnel Upwards Weight Weight Weight Upward - Height Temporarily
rary tunnel below pressure tunnel dry wet Downwar embankm (during
zone below groundwater [kN/m] lining ground ground d [kN/m] ent [m] construction
surface [m] level [m] [kN/m] [kN/m] [kN/m] )
Water level 0.5 m lower Zone A 10 10.53 870.9 305.5 139.0 549.7 -23.9 - Water level
Water level 1 m lower 9 10.53 870.9 305.5 202.2 509.9 -45.0 - decrease
Water level 1 m lower 8 10.53 870.9 305.5 202.2 474.2 -12.8 -
Water level 1 m lower Zone B 7 10.53 870.9 305.5 202.2 434.6 22.8 0.14 Embankmen
Water level 1 m lower 6 10.53 870.9 305.5 202.2 418.2 37.5 0.22 t + water
Water level 1.5 m lower 5.6 10.53 870.9 305.5 level
265.4 331.2 59.0 0.35
decrease
Water level 2 m lower 5 10.53 870.9 305.5 328.5 272.2 55.2 0.33
Water level 2 m lower -1 6.93 607.8 305.5 81.2 2.1 257.9 1.93
Water level 2 m lower -3.53 4.40 344.8 305.5 69.8 3.30
Water level 2 m lower Zone C -4.53 3.40 243.2 305.5 -31.7 - Water level
Water level 1 m lower -5.53 3.40 243.2 305.5 -31.7 - decrease
Water level normal again Zone D -6.53 3.40 243.2 305.5 -31.7 - Nothing
-7.53 2.40 149.4 305.5 -125.6 -
-8.03 1.90 107.0 305.5 -168.0 -
Table 50: Permanent situation with safety factors and concrete ballast
Concrete ballast
Permane Top of Depth tunnel Upward Weight Ballast Height Weight Weight Upward - Height Permanently
nt zone tunnel below s tunnel [kN/m] ballast dry wet Downward embankme (after
below groundwater pressure lining [m] ground ground with safety nt with tunnelling) with
surface [m] level [m] [kN/m] [kN]m] [kN/m] [kN/m] [kN/m] safety [m] safety
Zone 1 10 10.53 870.9 305.5 127.8 1.6 75.8 560.2 -91.6 -0.04 Sand ballast
9 10.53 870.9 305.5 156.8 2.1 75.8 531.0 -91.3 -0.04
8 10.53 870.9 305.5 196.2 2.4 75.8 495.2 -94.6 -0.06
Zone 2 7 10.53 870.9 305.5 235.3 2.1 75.8 455.6 -94.2 -0.06 Concrete ballast
6 10.53 870.9 305.5 274.3 2.3 75.8 409.8 -88.0 -0.02
Zone 3 5.6 10.53 870.9 305.5 314.6 2.5 75.8 362.8 -82.0 0.01 Concrete ballast
No removal of tunnel lining above surface
5 10.53 870.9 305.5 314.6 2.5 75.8 314.3 -38.4 0.27 + embankment
4 10.53 870.9 305.5 (length is + 225
314.6 2.5 75.8 227.5 39.7 0.74
m)
3 10.53 870.9 305.5 314.6 2.5 75.8 144.3 114.6 1.18
2 10.53 870.9 305.5 314.6 2.5 75.8 85.8 167.3 1.49
1 10.53 870.9 305.5 314.6 2.5 75.8 45.5 203.6 1.71
0.47 10.40 868.8 305.5 314.6 2.5 94.2 63.9 168.5 1.50
0 9.93 851.1 305.5 314.6 2.5 66.9 44.0 193.2 1.65
-1 8.93 787.4 305.5 314.6 2.5 33.5 18.8 182.2 1.98
-2 7.93 703.6 305.5 314.6 2.5 17.0 6.8 124.1 2.33
-3 6.93 607.8 305.5 314.6 2.5 7.1 1.6 41.9 2.69
-3.53 6.40 554.0 305.5 314.6 2.5 3.7 0.1 -7.6 2.89
-4.53 5.40 449.6 305.5 314.6 2.5 0.4 0.0 -108.8 3.26
Zone 4 -5.53 4.40 344.8 147.9 254.6 2.2 0.0 0.0 -17.5 From here Concrete ballast
Zone 5 -6.53 3.40 243.2 129.2 151.5 1.8 0.0 0.0 -9.4 no tunnel, Sand ballast
-7.53 2.40 149.4 109.5 64.5 1.0 0.0 0.0 -7.2 so no 0.5 m
required
-8.03 1.90 107.0 87.7 38.1 0.7 0.0 0.0 -6.3
A.H. Oldenhave 105

Table 51: Permanent situation with safety factors and slag ballast
Slag ballast
Perma Top of Depth tunnel Upwards Weight Ballast Height Weight Weight Upward - Height Permanently
nent tunnel below pressure tunnel [kN/m] ballast dry wet Downward embankment (after
zone below groundwater [kN/m] lining [m] ground ground with safety with safety tunnelling)
surface [m] level [m] [kN]m] [kN/m] [kN/m] [kN/m] [m] with safety
9 10.53 870.9 305.5 156.8 2.1 75.8 531.0 -91.3 -0.04
8 10.53 870.9 305.5 196.2 2.4 75.8 495.2 -94.6 -0.06
Zone 2 7 10.53 870.9 305.5 233.0 2.4 75.8 455.6 -92.1 -0.05 Slag ballast
Zone 3 6 10.53 870.9 305.5 249.1 2.5 75.8 409.8 -65.4 0.11 Slag ballast +
5.6 10.53 870.9 305.5 249.1 2.5 75.8 362.8 -23.1 0.36 embankment
(length is +
5 10.53 870.9 305.5 249.1 2.5 75.8 314.3 20.6 0.62

250 m)
4 10.53 870.9 305.5 249.1 2.5 75.8 227.5 98.7 1.09
3 10.53 870.9 305.5 249.1 2.5 75.8 144.3 173.6 1.53
2 10.53 870.9 305.5 249.1 2.5 75.8 85.8 226.2 1.84
1 10.53 870.9 305.5 249.1 2.5 75.8 45.5 262.5 2.06
0.47 10.40 868.8 305.5 249.1 2.5 94.2 63.9 227.4 1.85
0 9.93 851.1 305.5 249.1 2.5 66.9 44.0 252.2 2.00
-1 8.93 787.4 305.5 249.1 2.5 33.5 18.8 241.2 2.33
-2 7.93 703.6 305.5 249.1 2.5 17.0 6.8 183.0 2.68
-3 6.93 607.8 305.5 249.1 2.5 7.1 1.6 100.8 3.04
-3.53 6.40 554.0 305.5 249.1 2.5 3.7 0.1 51.4 3.24
-4.53 5.40 449.6 305.5 249.1 2.5 0.4 0.0 -49.9 3.61
Zone 4 -5.53 4.40 344.8 147.9 249.1 2.5 0.0 0.0 -12.5 From here no Slag ballast
Zone 5 -6.53 3.40 243.2 129.2 151.5 1.8 0.0 0.0 -9.4 tunnel, so no Sand ballast
-7.53 2.40 149.4 109.5 64.5 1.0 0.0 0.0 -7.2 0.5 m
required
-8.03 1.90 107.0 87.7 38.1 0.7 0.0 0.0 -6.3
A.H. Oldenhave 106

Table 52: Permanent situation with safety factors and sand ballast
Sand ballast
Perman Top of Depth tunnel Upward Weight Ballast Height Weight Weight Upward - Height Permanently
ent tunnel below s tunnel [kN/m ballast dry wet Downward embankment (after
zone below groundwater pressure lining ] [m] ground ground with safety with safety tunnelling)
surface [m] level [m] [kN/m] [kN]m] [kN/m] [kN/m] [kN/m] [m] with safety
9 10.53 870.9 305.5 156.8 2.1 75.8 531.0 -91.3 -0.04
8 10.53 870.9 305.5 196.2 2.4 75.8 495.2 -94.6 -0.06
Zone 2 7 10.53 870.9 305.5 209.7 2.5 75.8 455.6 -71.1 0.08 Sand ballast +
6 10.53 870.9 305.5 209.7 2.5 75.8 409.8 -29.9 0.32 embankment
5.6 10.53 870.9 305.5 209.7 2.5 75.8 362.8 12.4 0.57 (length is +
290 m)
5 10.53 870.9 305.5 209.7 2.5 75.8 314.3 56.0 0.83
4 10.53 870.9 305.5 209.7 2.5 75.8 227.5 134.1 1.30

3 10.53 870.9 305.5 209.7 2.5 75.8 144.3 209.0 1.74
2 10.53 870.9 305.5 209.7 2.5 75.8 85.8 261.7 2.05
1 10.53 870.9 305.5 209.7 2.5 75.8 45.5 298.0 2.27
0.47 10.40 868.8 305.5 209.7 2.5 94.2 63.9 262.9 2.06
0 9.93 851.1 305.5 209.7 2.5 66.9 44.0 287.6 2.21
-1 8.93 787.4 305.5 209.7 2.5 33.5 18.8 276.6 2.54
-2 7.93 703.6 305.5 209.7 2.5 17.0 6.8 218.5 2.89
-3 6.93 607.8 305.5 209.7 2.5 7.1 1.6 136.3 3.25
-3.53 6.40 554.0 305.5 209.7 2.5 3.7 0.1 86.9 3.45
-4.53 5.40 449.6 305.5 209.7 2.5 0.4 0.0 -14.4 3.82
-5.53 4.40 344.8 305.5 209.7 2.5 0.0 0.0 -118.9 4.19
Zone 3 -6.53 3.40 243.2 129.2 151.5 1.8 0.0 0.0 -9.4 No tunnel, so Sand ballast
-7.53 2.40 149.4 109.5 64.5 1.0 0.0 0.0 -7.2 no 0.5 m
-8.03 1.90 107.0 87.7 38.1 0.7 0.0 0.0 -6.3 required
A.H. Oldenhave 107

Appendix F: LDesign [39]
In this Appendix the working of the framework program LDesign is explained.
F.1. Input
When starting the program a lot of different input parameters are asked (Figure 92). Starting with
the dimensions of the lining and the amount of required points (amount of segments times amount
of segment fields) the coordinates of each point can be calculated and the ring with its points can be
drawn. In Figure 93 the black dots imply the points in the segment while the green dots indicate the
longitudinal joints. As earlier mentioned, in these longitudinal joints a contact stiffness of
is present. The contact length of the joint lt is assumed to be 170 mm which is an experience number.
Figure 93: Tunnel lining with 28 springs (purple lines) and forces
Figure 92: Input caused by the soil (red lines)
Under the input box of the lining geometry the double ring can be checked. This means that instead
of assuming a single ring analysis, the second ring and its interaction with the first ring can also be
taken into account. Again, due to previous experiences with this calculation method it can be
concluded that this interaction does not have major influences on the results. Therefore in this thesis
only the single ring analysis is performed.
As a simplification different approaches exist to determine the forces acting in the lining. One of
these approaches implies that due to the upward pressure of the groundwater (floating) the soil
stresses on top increase and the stresses at the bottom decrease. This leads in the steady state to
equal pressures on top and at the bottom. This approach therefore assumes:
.
However, in this case the tunnel lining is only partly located in the groundwater. Therefore this
approach does not apply.
Stabilizing supports as springs is nothing more than stabilizing the tunnel ring to prevent rolling of
the ring. However, when this box is not ticked, one support still remains (see blue triangle on top in
Figure 93).
As previously mentioned the key segment is of minor influence and can be omitted in the calculation.
After giving the mean Eoed at the centre, the soil characteristics can be filled in. The heights of the
different parameters are with respect to the centre of the lining.
There are different methods to calculate the required data. Duddeck is a commonly used method but
only feasible for deep positions. Therefore only increasing gravity load will be used. For the shallow
positions (H < 2D) the crown reduction will be used as well.
Given in the previous chapter the dead weight lining is of minor influence for deep positions.
However for this case the influences cannot be disregarded any more.
The last part is outside this scope. Furthermore, the values are all given in N and mm.
Eventually LDesign uses the input to generate a table which is used for the final calculations.
Underneath the table is given. Coloured rows are explained below:
1

49 28 21
Yellow (1):
0 5065 1 0 0 0 0 0 0 -9995.18 0 1 1 7.9E+10
1127.06853 4938.00988 0 0 0 0 0 0 -2779.83 -12179.2 0 1 1 1
2197.62114 4563.40732 0 0 0 0 0 0 -8305.64 -17246.8 0 1 1 1
3157.97585 3959.97646 0 0 0 0 0 0 -17840 -22370.6 0 1 1 1
3959.97646 3157.97585 0 0 0 0 0 0 -32282.5 -25744.4 0 1 1 7.9E+10
4563.40732 2197.62114 0 0 0 0 0 0 -50062.2 -24108.7 0 1 1 1
The three numbers indicate the amount of points (one 4938.00988

5065
4938.00988
4563.40732
1127.06853
-1.94E-12
-1127.06853
-2197.62114
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
-69836.7
-90403.6
-110058
-125560
-15939.8
3.46E-11
25119.95
60466.62
0
0
0
0
1
1
1
1
1
1
1
1
1
1
7.9E+10
1
3959.97646 -3157.97585 0 0 0 0 0 0 -131873 105165.1 0 1 1 1
spring is connected between two points), the amount of

3157.97585 -3959.97646 0 0 0 0 0 0 -123475 154832.8 0 1 1 1
2197.62114 -4563.40732 0 0 0 0 0 0 -96972.3 201365 0 1 1 7.9E+10
1127.06853 -4938.00988 0 0 0 0 0 0 -53604.5 234856.7 0 1 1 1
-3.88E-12 -5065 1 0 0 0 0 0 1.89E-10 247086.5 0 1 1 1
-1127.0685 -4938.00988 0 0 0 0 0 0 53604.5 234856.7 0 1 1 1
-2197.6211 -4563.40732 0 0 0 0 0 0 96972.26 201365 0 1 1 7.9E+10
tunnel lining lines and the amount of springs, respectively. -3157.9758

-3959.9765
-4563.4073
-4938.0099
-3959.97646
-3157.97585
-2197.62114
-1127.06853 2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
123475
131872.9
125560.2
110057.7
154832.8
105165.1
60466.62
25119.95 3
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
7.9E+10 4
-5065 -9.30E-13 0 0 0 0 0 0 90403.55 1.66E-11 0 1 1 1

-4938.0099 1127.06853 0 0 0 0 0 0 69836.74 -15939.8 0 1 1 1
Green (2): -4563.4073

-3959.9765
-3157.9758
-2197.6211
2197.62114
3157.97585
3959.97646
4563.40732
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
50062.24
32282.46
17839.99
8305.643
-24108.7
-25744.4
-22370.6
-17246.8
0
0
0
0
1
1
1
1
1
1
1
1
1
7.9E+10
1
1
-1127.0685 4938.00988 0 0 0 0 0 0 2779.831 -12179.2 0 1 1 1
The first column gives the x-coordinates of respectively the

5543.96704 4421.16618 1 1 0 0 0 0 0 0 0 1 1 1
6388.77024 3076.66959 1 1 0 0 0 0 0 0 0 1 1 1
6913.21383 1577.89594 1 1 0 0 0 0 0 0 0 1 1 1
7091 -2.71E-12 1 1 0 0 0 0 0 0 0 1 1 1
6913.21383 -1577.89594 1 1 0 0 0 0 0 0 0 1 1 1
6388.77024 -3076.66959 1 1 0 0 0 0 0 0 0 1 1 1
tunnel lining and the soil points; the second column gives 5543.96704
4421.16618
3076.66959
1577.89594
-4421.16618
-5543.96704
-6388.77024
-6913.21383
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
-5.43E-12 -7091 1 1 0 0 0 0 0 0 0 1 1 1
the y-coordinates of these points. (0,0) is the centre of the

-1577.8959 -6913.21383 1 1 0 0 0 0 0 0 0 1 1 1
-3076.6696 -6388.77024 1 1 0 0 0 0 0 0 0 1 1 1
-4421.1662 -5543.96704 1 1 0 0 0 0 0 0 0 1 1 1
-5543.967 -4421.16618 1 1 0 0 0 0 0 0 0 1 1 1
-6388.7702 -3076.66959 1 1 0 0 0 0 0 0 0 1 1 1
-6913.2138 -1577.89594 1 1 0 0 0 0 0 0 0 1 1 1
tunnel ring. -7091

-6913.2138
-6388.7702
-5543.967
-1.30E-12
1577.89594
3076.66959
4421.16618
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1

1.312E+10 1.7493E+14 1 2
Orange (3):
1.312E+10 1.7493E+14 2 3
1.312E+10 1.7493E+14 3 4
1.312E+10 1.7493E+14 4 5
1.312E+10 1.7493E+14 5 6
1.312E+10 1.7493E+14 6 7
1.312E+10 1.7493E+14 7 8
Here the radial forces of the soil are calculated for each 1.312E+10
1.312E+10
1.312E+10
1.312E+10
1.7493E+14
1.7493E+14
1.7493E+14
1.7493E+14
8
9
10
11
9
10
11
12
tunnel lining point.

1.312E+10
1.312E+10
1.312E+10
1.312E+10
1.312E+10
1.7493E+14
1.7493E+14
1.7493E+14
1.7493E+14
1.7493E+14
5
12
13
14
15
16
13
14
15
16
17
1.312E+10 1.7493E+14 17 18

1.312E+10 1.7493E+14 18 19
Red (4): 1.312E+10

1.312E+10
1.312E+10
1.312E+10
1.7493E+14
1.7493E+14
1.7493E+14
1.7493E+14
19
20
21
22
20
21
22
23
1.312E+10 1.7493E+14 23 24
1.312E+10 1.7493E+14 24 25
The seven values indicate the earlier mentioned constant 1.312E+10

1.312E+10
1.312E+10
1.312E+10
1.7493E+14
1.7493E+14
1.7493E+14
1.7493E+14
25
26
27
28
26
27
28
1
5 29 1 1000 9378462
rotational stiffness cr. At these locations the longitudinal

6 30 1 1000 9378462
7 31 1 1000 9378462
8 32 1 1000 9378462
6
9 33 1 1000 9378462
10 34 1 1000 9378462
11 35 1 1000 9378462
joints are positioned. 12

13
14
15
36
37
38
39
1
1
1
1
1000
1000
1000
1000
9378462
9378462
9378462
9378462

16 40 1 1000 9378462
Blue (5):
17 41 1 1000 9378462
18 42 1 1000 9378462
19 43 1 1000 9378462
20 44 1 1000 9378462
21 45 1 1000 9378462
22 46 1 1000 9378462
The first column gives the EA which is the force per unit 23
24
25
47
48
49
1
1
1
1000
1000
1000
9378462
9378462
9378462
strain; the second column is the bending stiffness EI.

 Purple (6):
The purple values give the radial soil stiffness kr.
A.H. Oldenhave 109

Officially LDesign was designed for traditional bored tunnels where the tunnel was positioned at least
1D under surface level. However, with this new method (tunnelling with marginal covers) the tunnel
lining already starts above surface level. Therefore additional adjustments in the general input were
necessary. In LDesign this is possible by manipulating the table correctly for the applied situation.
When more than the upper quarter of the tunnel lining was located above surface level, more springs
had to be removed. With this method the upper half of the springs can be removed, see Figure 94.
H=2.735 H=1.735 H=0.735
Figure 94: Removal of springs above surface [39]
However, after calculation of the buoyancy it was concluded that at shallow depths an embankment
is necessary to avoid uplift forces. For the calculations with LDesign it is therefore not necessary any
more to reduce the amount of springs in the model (the embankment covers these parts of the
tunnel lining).
Moreover, the heights of the soil (h s(soil)) and groundwater table (h water) are with respect to the
centre of the tunnel. When the centre of the tunnel is located above these heights the input has to
be manipulated by hand as well (negative values are not possible in LDesign). For H=0.735 (see
further) and a groundwater level of 1.4 m below surface, the forces of the soil acting on the different
points are calculated by hand and are replacing the forces given in the table. This will result in more
accurate results.
A.H. Oldenhave 110

Appendix G: Calculation of reinforcement [12]

G.1. Reinforcement in ring direction
In order to calculate the minimum amount of reinforcement to prevent failing of the lining, a good
understanding of the behaviour of concrete and steel is necessary. In the report one kind of load case
was visible; bending and pressure forces. Underneath calculations are found to determine the
required reinforcement.
Bending and pressure forces
When pressure forces and bending moments are present, two equations have to be valid:
In Figure 95 a cross section of a column and its forces are schematised. The equation for the sum of
the normal forces is:
(0.75 is the so called fullness factor and valid when c = cu = 3.5 ‰)
In above formulas two unknowns are present; As and x (length of pressure zone, see figure).
Therefore summation of the bending moments (at the top, right point in figure) is necessary as well:
( )
The length between Nc and top (a) is depending on the strain distribution. In most cases the
maximum strain in the concrete is reached before the steel yields. Therefore c = cu is plausible for
the calculation of the amount of reinforcement. In this case the value of a is equal to 0.39 x. Now
with two equations and two unknowns the final reinforcement area A s can be calculated. Values are
given in Paragraph 7.5. The total amount of reinforcement is .
A.H. Oldenhave 111

Figure 95: Cross section column with bending and pressure forces. On the right side the strain distribution and the forces
acting on the column are drawn. The normal force acting in the right reinforcement bars are zero (reinforcement cannot
react as pressure force).
Remark
The maximum strain of concrete amounts 3.5 ‰ and steel yields at 27.5 ‰. In most cases the
maximum strain of concrete is reached (cracks will be present) before the steel will yield. This is
desirable, i.e. the maximum strain at the tension side amounts 27.5 ‰. With the maximum pressure
strain of 3.5 ‰, the minimum length of the pressure zone can be calculated.
mm.
This implies that when the value of x < 45.2 mm, the steel will yield and the tunnel lining will fail.
However, it appears that the strain distribution is different. The strain at the pressure zone is less
than 3.5 ‰ (but bigger than c,pl = 1.75 ‰, so still cracks will occur) and the length of the pressure
zone is bigger, see Figure 96. Now the concrete will still yield before reinforcement will fail.
Despite the incorrect applied calculation

method it is proven that both methods
result in the same amount of calculation.
The length of the pressure zone will be
bigger than calculated and the amount of
reinforcement will still guarantee strength.
Figure 96: Theoretical and practical strain distribution
A.H. Oldenhave 112

G.2. Reinforcement in longitudinal direction

In order to create a rigid reinforcement cage, brackets will be applied as well. The dimensions are
depending on the shear forces in the tunnel lining.
According to Walraven (2006) the limit of concrete (i.e. maximum shear stress of concrete) to absorb
the shear forces is 0.4 fb, see Table 53.
Table 53: Limit of shear stresses in concrete
Strength classes 0.4 fb

C12/15 0.36
C20/25 0.46
C28/35 0.56
C35/45 0.66
C45/55 0.76
C53/65 0.86
To determine the maximum shear force where no reinforcement (brackets) is required, the value d =
0.4 fb has to be multiplied by the area of a segment, i.e. b x d.
When the occurring shear force is higher than the maximum shear force of the concrete, the
remaining shear force has to be taken up by the reinforcement. By dividing the remaining shear force
Vs by the yielding strength of steel (435 N/mm2), the amount of reinforcement can be calculated:
( )
A.H. Oldenhave 113

Appendix H: Calculation of bolts

H.1. Bolts in the longitudinal joints
A tunnel ring consists of several segments (in this case 7) and therefore between these segments
joints are present. These joints interfere the force distribution of a singular ring. The reinforcement is
interrupted and bolts between two segments have to transmit sufficient forces to prevent failure of
the tunnel lining.
To determine the dimension and strength of the bolts, insight of the present forces in the lining is
required. Two critical sections are chosen (see Paragraph 7.5, the lowest groundwater levels are
decisive). The forces are depending on the length of the longitudinal joint.
Limitations
As previously mentioned the maximum allowable deformation of the tunnel lining caused by the
internal forces is stated as 50 mm. Moreover, according to Beijer (2010), when the length of the
longitudinal joint lt is 100 mm, water tightness is guaranteed. Adding these two limitations to the
outcome of the two critical cross sections, the deformations remain below 50 mm when applying an
opened joint (lt = 100 mm) which is the most onerous situation (smallest forces). This leads to the
following maxima:
Table 54: Maxima critical cross sections
H = 0.735 m H = 5.735 m
lt [mm] 170 100 lt [mm] 170 100
Mmax [kNm] 114 105 Mmax [kNm] 420 379
ymax [mm] 13 17 ymax [mm] 45 54
According to the table above the cross section located just below surface level will be normative. This
is due to the fact that ground improvement is not constructed at this cross section any more.
However, in this Appendix the first cross section will be used for calculations (H=0.735).
Due to self-weight of the tunnel lining, the normal forces in the joints act like pressure forces. This is
onerous for the joints, however negligible with respect to the bending moment (for example:
M/W=6.24, N/A=0.55). Therefore for these cross sections normal bolts will not be applied.
Bolt type
For this connection the Japanese box method will be applied (see Chapter 6), which can counteract
higher bending moments. For a first calculation it will be assumed that the length between the bolts
and the rotation point is ½ h (200 mm). The pressure forces will be transferred through the
surrounding concrete.
Calculations [13]
Two different failures of the bolt connection are possible; Failure of the bolt or failure of the steel
plate (steel will punch). To determine which failure will be decisive, both mechanisms are calculated
for a M30, 8.8 bolt, screwed into a steel plate with a thickness tp of 20 mm:
Maximum tensile force in the bolts
A.H. Oldenhave 114

According to EC3 the maximum tensile force of one bolt is:
Maximum tensile force in the steel plate
According to EC3 the maximum tensile force to prevent punching of the bolt through the plate is:
The maximum tensile force in the bolts is decisive and amounts 326 kN.
When Mmax = 105 kNm and arm = 0.2 m, the bolts have to guarantee a force of kN. Divided
by two both bolts have a tensile force of 263 kN. It can be concluded that M30 bolts of class 8.8 and a
steel plate with a thickness of 20 mm will satisfy the requirements. In Figure 53 and underneath a
detailed design of the connection is given.
A more precise calculation takes into account the different lengths of the arms of the two bolts.
When applying a distance between both bolts of 70 mm (see figure), the arms will be 0.155 and
0.225 m. The rotation point is located 10 mm under the top due to the centre of the pressure zone of
the concrete. The maximum bending moment these two bolts can take into account amounts:
∑ kNm > 105 kNm.
Safety factors are included, therefore the bolts will satisfy.
Reinforcement
The force of the two bolts has to be translated by the reinforcement. In total 6 reinforcement bars
will be applied and the maximum force of both bolts amounts kN. Every bar has to
A.H. Oldenhave 115

transfer 109 kN. This is equal to mm2 reinforcement, which equals a reinforcement bar
with diameter 18. Reinforcement bars with the common applied diameter of 20 will be applied.
H.2. Maximum absorbing bonding strength

To prevent that the bolt will be pulled out of the concrete, enough bonding strength between the
bolt and the concrete has to be present. By applying ribbed bars the bonding strength will be
increased. According to Eurocode 2, paragraph 8.4.2 [36], the maximum absorbing bonding strength
can be calculated using the following formula:
Where fctd is the design value of the tensile strength of concrete according to:
MPa for C35/45.
The coefficients 1 and 2 are depending on the quality of the bonding conditions. 1 amounts 1.0 for
‘good’ conditions and 0.7 for all the other cases. When h > 250 mm, one side will always result in
‘bad’ conditions, so here 0.7 is used. For 2 a value of 1.0 may be used when the diameter is smaller
than 30 mm.
So, when assuming the first cross section where the bolt forces between the rings amount 66 kN per
bolt, the maximum bonding strength amounts:
MPa = 2310 kN/m2.
The circumference of the applied M20 bolt amounts mm = 0.063 m.
The minimum required area present in the concrete amounts m2 (the maximum force
in the bolt amounts 66 kN). This results in a minimum length of the bolt of m.
A.H. Oldenhave 116

Appendix I: Planning [43]

For a first estimation of the construction time of both methods, the planning of the Hubertustunnel
as well as the Sluiskiltunnel is used. Moreover, many amounts per day are experience numbers. This
is only a first estimate of the total construction time for both methods. When applying one of the
methods, further research will be recommended.
In this Appendix the different proceedings will be shortly explained, for both methods.
I.1. Traditional method

When applying the traditional method, the complete ramps have to be excavated before starting the
bored process. This means the preparations, consisting of placing of sheet piles, excavation,
construction of underwater concrete floor, construction of the plug and installation of the launching
frame and TBM will take more construction time than when starting at surface level. The reception
shaft (west ramp) however can be constructed during construction of the east ramp.
When the preparations of the east ramp are finished in theory the tunnelling of the tunnel can start.
However, for the cross passages it will be assumed that jet grouting is required to make construction
of the passages possible. The jet grouting will already start when preparations of the east ramp are
still in progress. This results in starting of the bored part when the preparations are finished. The jet
grouting for the last cross passages are still in operation, however due to adequate difference in
length, this will not influence the tunnelling process. The jet grouting is finished 4 months after the
start of the first bored tunnel. 6 months later the tunnelling process is finished as well. These two
proceedings will therefore not influence each other. It will be assumed that the excavation rate per
day amounts 8.5 meter, which is a reliable estimation for soft soils according to Bouygues
Construction (2009) [21]. In this rate interventions for disc cutters changes are included. Moreover,
for the North-South line in Amsterdam (corresponding soil conditions) a mean rate of 8.5 meters and
12 meters were achieved. Therefore an excavation rate of 8.5 meters per day (24 hours per day, 7
days a week) is a reliable and safe assumption.
After finishing the first tunnel, the TBM has to be disassembled and transported back to the
launching shaft. This will approximately take 8 weeks to disassembly the TBM and 8 weeks to
transport and assemble the TBM again. The second tunnel will now be drilled. When the TBM has
crossed the location of the first cross passages, these passages can already be constructed before the
tunnelling is finished.
When the tunnels are finished, the TBM will be disassembled and transported again and the
construction of the ramps can start (construction of the concrete floors, walls and decks (only closed
section)). It will be assumed that to limit the costs only one side of the ramps will be constructed at
the same time. When the east ramp is finished, the west ramp will be finished afterwards. Now only
backfill of the building pits remain. When the closed parts of the ramps are refilled with soil until
surface level the whole tunnel connection is finished.
The traditional method will take 4 years and 7 months. In Table 55 and Figure 97 the planning of this
method is given.
A.H. Oldenhave 117

Table 55: Detailed planning construction time traditional method (1)
Start date Amount Amount Work Calendar End date

per day days days
Ramps 1-1-2015 1065 1491 31-1-2019
Ramp east 1-1-2015 912 1277 1-7-2018
Insertion sheet piles 1-1-2015 15480 m2 350 44 62 4-3-2015
Excavation building pit & 4-3-2015 72000 m3 816 88 123 5-7-2015
construction struts
Soil transportation 4-3-2015 88 123 5-7-2015
Construcion underwater concrete 5-7-2015 43 60 3-9-2015
floor
Placing reinforcement 5-7-2015 8 11 16-7-2015
Pouring concrete 16-7-2015 20 28 13-8-2015
Hardening time concrete 13-8-2015 15 21 3-9-2015
Pumping of water 3-9-2015 72000 m3 7200 10 14 17-9-2015
Construction plug 17-9-2015 21 29 16-10-2015
Finishing ramp 16-10-2015 20 28 13-11-2015
Construction service building 13-11-2015 40 56 8-1-2016
Construction launching frame & 8-1-2016 40 56 4-3-2016
installing TBM
Construction concrete first floor 18-11-2017 56 78 4-2-2018
Construction concrete walls 4-2-2018 52 73 18-4-2018
Construction concrete deck 18-4-2018 40 56 13-6-2018
(closed section)
Backfill building pit 13-6-2018 8280 m3 640 13 18 1-7-2018
Ramp west 11-3-2015 1016 1422 31-1-2019
Excavation building pit & 12-5-2015 72000 m3 816 88 123 12-9-2015
construction struts
Construction underwater 12-9-2015 58 81 2-12-2015
concrete floor
Placing reinforcement 12-9-2015 11 15 27-9-2015
Pouring concrete 27-9-2015 27 38 4-11-2015
Hardening time concrete 4-11-2015 20 28 2-12-2015
Pumping of water 2-12-2015 72000 m3 7200 10 14 16-12-2015
Construction plug 16-12-2015 21 29 14-1-2016
Finishing ramp 14-1-2016 20 28 11-2-2016
Construction concrete first floor 13-6-2018 56 78 30-8-2018
Construction concrete walls 30-8-2018 52 73 11-11-2018
Construction concrete deck 11-11-2018 40 56 6-1-2019
(closed section)
Backfill building pit 6-1-2019 11250 m3 640 18 25 31-1-2019
Bored part 4-3-2016 3870 m 576 624 18-11-2017
Tunnellng 1st part 4-3-2016 1935 m 8.5 228 228 18-10-2016
A.H. Oldenhave 118

TBM return & reinstalling 18-10-2016 80 112 7-2-2017

Tunnelling 2nd part 7-2-2017 1935 m 8.5 228 228 23-9-2017
Disassembly TBM 23-9-2017 40 56 18-11-2017
Cross passages 13-11-2015 647 906 7-5-2018
Cross passage 1 13-11-2015 441 618 23-7-2017
Installation jet 13-11-2015 10 14 27-11-2015
Drilling and jet grouting 27-11-2015 15 21 18-12-2015
Construction 4-6-2017 30 42 16-7-2017
Dismantling & transportation jet 16-7-2017 4 6 22-7-2017
Cross passage 2 18-12-2015 451 631 9-9-2017
Construction 22-7-2017 30 42 2-9-2017
Cross passage 3 22-1-2016 460 644 27-10-2017
Construction 8-9-2017 30 42 20-10-2017
Cross passage 4 26-2-2016 469 657 14-12-2017
Construction 26-10-2017 30 42 7-12-2017
Cross passage 5 1-4-2016 479 670 31-1-2018
Construction 13-12-2017 30 42 24-1-2018
Cross passage 6 6-5-2016 488 683 20-3-2018
Construction 30-1-2018 30 42 13-3-2018
Cross passage 7 10-6-2016 497 696 7-5-2018
Construction 19-3-2018 30 42 30-4-2018
A.H. Oldenhave 119

Planning traditional method Rijnlandroute
Dec-16
Dec-17
Dec-18
Sep-16
Sep-17
Sep-18
Apr-15
Apr-16
Apr-17
Apr-18
Apr-19
Oct-15
Jan-15
Jan-16
Jul-15
Jul-16
Jul-17
Jul-18
Jul-19
Ramps
Ramp east
Insertion sheet piles
Excavation building pit & construction struts
Soil transportation
Construcion underwater concrete floor
Placing reinforcement
Pouring concrete
Hardening time concrete
Pumping of water
Construction plug
Finishing ramp
Construction service building
Construction launching frame & installing TBM
Construction concrete first floor
Construction concrete walls
Construction concrete deck (closed section)
Backfill building pit
Ramp west
Excavation building pit & construction struts
Soil transportation
Construction underwater concrete floor
Placing reinforcement
Pouring concrete
Hardening time concrete
Pumping of water
Construction plug
Finishing ramp
Construction concrete first floor
Construction concrete walls
Construction concrete deck (closed section)
Backfill building pit
Bored part
Tunnel boring 1st part
TBM return & reinstalling
Tunnel boring 2nd part
Disassembly TBM
A.H. Oldenhave 120

Planning traditional method Rijnlandroute
Dec-16
Dec-17
Dec-18
Sep-16
Sep-17
Sep-18
Apr-15
Apr-16
Apr-17
Apr-18
Apr-19
Oct-15
Jan-15
Jan-16
Jul-15
Jul-16
Jul-17
Jul-18
Cross passages
Cross passage 1
Installation jet
Drilling and jet grouting
Construction
Dismantling & transportation jet
Cross passage 2
Installation jet
Construction
Cross passage 3
Installation jet
Construction
Cross passage 4
Installation jet
Construction
Cross passage 5
Installation jet
Construction
Cross passage 6
Installation jet
Construction
Cross passage 7
Installation jet
Construction
Figure 97: Detailed planning construction time traditional method (2)
A.H. Oldenhave 121

I.2. Tunnelling with marginal cover method

When applying the tunnelling with marginal cover method, this results in less preparations before
starting the bored process; however the bored part will take more construction time.
To limit the settlements and to make construction at shallow depths possible, it was concluded that
ground improvement is necessary (Chapter 5), as well as an embankment at surface level. These
proceedings require most of the preparation time. Afterwards small sheet piles will be installed
(depth building pit is 2.5 meters, so the sheet piles will be 3 x 2.5 = 7.5 m) and the soil within will be
excavated. The groundwater has to be lowered temporarily as well.
For the construction time of the MIP columns, CUR rapport 95-7 is used in which the amount of
construction of MIP columns is estimated as 25 mm per rotation and 150 rotations per minute.
Assuming a work day of 8 hours, it can be calculated how much MIP columns can be constructed per
day. As mentioned in Paragraph 7.3.2 the amount of MIP columns per m2 is 71.6 %. Assuming a
ground improvement area of 200 x 18.5 m2 x 2 at the eastern side (two different building pits) and an
area of 200 x 50.5 m2 at the west side, this leads to a total ground improved area of 17500 m 2. When
71.6 % of that area will be applied with MIP columns of more than 10 meters, this results in a total of
130060 MIP columns that have to be constructed. Using above mentioned values the total
construction time for the MIP columns can be calculated:
days.
After these proceedings and installing the TBM (which is ordered and delivered on site in
approximately 10 months) and launching frame, the bored process can start. For the traditional
method a mean excavation rate of 8.5 meters per day was assumed. However, when the bored part
becomes longer, the mean rate will also increase (assumption). Therefore for this method a mean
rate of 10 meters per day is assumed.
For the traditional method 7 cross passages are necessary. For this method however (due to longer
tunnel length) 9 cross passages have to be constructed. This results in a higher construction time for
the cross passages. Only when the last cross passage is finished, the back fill of the west “ramp” can
start. Hereafter the tunnel lining above surface partly will be removed (see Appendix E).
The total construction time of the Rijnlandroute using the tunnelling with marginal cover method,
amounts almost 4 years. In Table 56 and Figure 98 the planning for this method is given, including all
proceedings.
Remark
This is only a first general assumption of the total construction time for the tunnelling with marginal
cover method. It is for example assumed that at the location of all cross passages jet grouting is
required. However, when the first and last cross passage is still located in the MIP columns jet
grouting may not be necessary any more. Moreover, freezing for shallow depths can also be
interesting. The area for the MIP columns can also differ. For a more precise assumption of the total
construction time of this method it will therefore be recommended to expand the amount of
calculations at certain depths. This will influence the amount of precautions and therefore the
planning (as well as the costs).
A.H. Oldenhave 122

A.H. Oldenhave 123

Table 56: Detailed planning construction time tunnelling with marginal cover method (1)
Start date Amount Amount Work Calendar End date

per day days days
Ramps 1-1-2015 809 1133 6-2-2018
Construction MIP columns 1-1-2015 130060 m column 1800 72 101 12-4-2015
Placing embankment 20-2-2015 60755 m3 800 65 91 22-5-2015
Excavation building pit 1-6-2015 6300 m3 816 8 10 11-6-2015
Drainage of groundwater 11-6-2015 60000 m3 2400 25 35 16-7-2015
Finishing ramp 16-7-2015 20 28 13-8-2015
Construction service building 16-7-2015 40 56 10-9-2015
Construction launching frame & 10-9-2015 40 56 5-11-2015
installing TBM
Backfill building pits east 10-9-2017 2244 m3 640 4 6 16-9-2017
Backfill building pit west 17-1-2018 2094 m3 640 4 6 23-1-2018
Bored part 29-10-2015 5146 m 634 682 10-9-2017
Tunnelling 1st part 29-10-2015 2573 m 10 257 257 12-7-2016
TBM return + reinstalling 12-7-2016 80 112 1-11-2016
Tunnelling 2nd part 1-11-2016 2573 m 10 257 257 16-7-2017
Disassembly TBM 16-7-2017 40 56 10-9-2017
Cross passages 16-7-2015 654 916 17-1-2018
Cross passage 1 16-7-2015 380 532 29-12-2016
Construction 11-11-2016 30 42 23-12-2016
Cross passage 2 20-8-2015 389 545 15-2-2017
Construction 29-12-2016 30 42 9-2-2017
Cross passage 3 24-9-2015 399 558 4-4-2017
Construction 15-2-2017 30 42 29-3-2017
Cross passage 4 29-10-2015 408 571 22-5-2017
Construction 4-4-2017 30 42 16-5-2017
Cross passage 5 3-12-2015 417 584 9-7-2017
A.H. Oldenhave 124


Construction 22-5-2017 30 42 3-7-2017
Cross passage 6 7-1-2016 426 597 26-8-2017
Construction 9-7-2017 30 42 20-8-2017
Cross passage 7 11-2-2016 436 610 13-10-2017
Construction 26-8-2017 30 42 7-10-2017
Cross passage 8 17-3-2016 445 623 30-11-2017
Construction 13-10-2017 30 42 24-11-2017
Cross passage 9 21-4-2016 454 636 17-1-2018
Construction 30-11-2017 30 42 11-1-2018
Tunnel lining removal 16-9-2017 111 m 102 143 6-2-2018
Tunnel lining removal east 16-9-2017 10 14 30-9-2017
Tunnel lining removal west 23-1-2018 10 14 6-2-2018
A.H. Oldenhave 125

Planning tunnelling with marginal cover

method Rijnlandroute
Dec-16
Dec-17
Sep-16
Sep-17
Apr-15
Apr-16
Apr-17
Apr-18
Oct-15
Jan-15
Jan-16
Jul-15
Jul-16
Jul-17
Jul-18
Ramps
Construction MIP columns
Placing embankment
Excavation building pit
Soil transportation
Drainage of groundwater
Finishing ramp
Construction service building
Construction launching frame & installing TBM
Backfill building pits east
Backfill building pit west
Bored part
Tunnel boring 1st part
TBM return + reinstalling
Tunnel boring 2nd part
Disassembly TBM
Cross passages
Cross passage 1
Installation jet
Construction
Cross passage 2
Installation jet
Construction
Cross passage 3
Installation jet
Construction
Cross passage 4
Installation jet
Construction
Cross passage 5
Installation jet
Construction
Cross passage 6
Installation jet
Construction
A.H. Oldenhave 126

Planning tunnelling with marginal cover

method Rijnlandroute
Dec-16
Dec-17
Sep-16
Sep-17
Apr-15
Apr-17
Apr-16
Apr-18
Oct-15
Jan-15
Jan-16
Jul-15
Jul-16
Jul-17
Jul-18
Cross passage 7
Installation jet
Construction
Cross passage 8
Installation jet
Construction
Cross passage 9
Installation jet
Construction
Tunnel lining removal
Tunnel lining removal east
Tunnel lining removal west
Figure 98: Detailed planning construction time tunnelling with marginal cover method (2)
A.H. Oldenhave 127

List of figures
Figure 1: Tunnel boring machine [23] ................................................................................................ 14
Figure 2: Required support pressure in EPB shield [9] ....................................................................... 15
Figure 3: Principle of the URUP method [24] ..................................................................................... 16
Figure 4: TBM launching at surface level [3] ...................................................................................... 17
Figure 5: TBM arrival at surface level [3] ........................................................................................... 17
Figure 6: Overview Rijnlandroute with optimized tunnel location (white arrow) [4] .......................... 18
Figure 7: Integration of trajectory ..................................................................................................... 20
Figure 8: Cross section tunnel Rijnlandroute with dimensions [4] ...................................................... 21
Figure 9: Scope of the thesis (eastern ramp) ..................................................................................... 21
Figure 10: Activated injection locations when tunnel lining is partly above surface ........................... 29
Figure 11: Launching of the TBM at steel frame (yellow) in launching shaft [40] ................................ 29
Figure 12: Steel frame (yellow) connected to the sheet pile wall on front [40] .................................. 29
Figure 13: Launching procedure “tunnelling with marginal cover” method ....................................... 30
Figure 14: Force distribution in launching frame ............................................................................... 30
Figure 15: Steel frame connected (welded) to reinforcement ............................................................ 30
Figure 16: “First” cross connection at surface level ........................................................................... 31
Figure 17: Cross connection as designed by Movares [4] ................................................................... 31
Figure 18: Weights and lengths TBM (EPB shield) .............................................................................. 34
Figure 19: Boston Rule (approximation method) [26] ........................................................................ 35
Figure 20: Schematic overview forces acting on soil .......................................................................... 36
Figure 21: Resultant force causing a moment .................................................................................... 36
Figure 22: Stress distribution due to eccentricity ............................................................................... 37
Figure 23: Prevention of tilting .......................................................................................................... 37
Figure 24: Estimated soil structure .................................................................................................... 39
Figure 25: Height bottom tunnel - plate assumption ......................................................................... 40
Figure 26: Possible configurations of stabilized MIP columns ............................................................ 42
Figure 27: Force distribution in the soil (brown is MIP columns, blue is non-stabilized soil) ............... 42
Figure 28: Soil structure Botlek ......................................................................................................... 43
Figure 29: Triangular pattern ............................................................................................................ 44
Figure 30: Optimal MIP column pattern ............................................................................................ 46
Figure 31: stress distribution in Holocene layer ................................................................................. 47
Figure 32: Length of required ground improvement for application of the “tunnelling with marginal
cover” method .................................................................................................................................. 48
Figure 33: Starting depth tunnelling process ..................................................................................... 49
Figure 34: Estimated soil structure .................................................................................................... 50
Figure 35: Temporary situation ......................................................................................................... 52
Figure 36: Cross section tunnel with present volume for ballast [4] ................................................... 53
Figure 37: Permanent situation with concrete ballast........................................................................ 55
Figure 38: Permanent situation with slag ballast ............................................................................... 56
Figure 39: Permanent situation with sand ballast .............................................................................. 57
Figure 40: Continuum model ............................................................................................................. 58
Figure 41: Beam spring model ........................................................................................................... 58
Figure 42: Crown reduction ............................................................................................................... 58
Figure 43: Transformation of the horizontal and vertical loads to radial and tangential loads [11] ..... 59
A.H. Oldenhave 128

Figure 44: Area column – soil in triangular pattern ............................................................................ 60

Figure 45: Situation for lowest groundwater level ............................................................................. 63
Figure 46: Situation for highest groundwater level ............................................................................ 63
Figure 47: Normal forces [39] ............................................................................................................ 63
Figure 48: Shear forces [39] .............................................................................................................. 63
Figure 49: Bending moments [39] ..................................................................................................... 63
Figure 50: Cross section segment with required ring reinforcement .................................................. 65
Figure 51: Cross section segment with required longitudinal reinforcement ..................................... 66
Figure 52: Cross section Japanese box method, dimensions in mm [14] ............................................ 66
Figure 53: Cross section Japanese box method with reinforcement bars (connected to applied
reinforcement), dimensions in mm [14] ............................................................................................ 66
Figure 54: Japanese box method applied in tunnel segments [41] ..................................................... 67
Figure 55: Applied bolts in Sluiskiltunnel ........................................................................................... 68
Figure 56: Location and number of bolts in Sluiskiltunnel .................................................................. 68
Figure 57: Oblique bolt [14] .............................................................................................................. 68
Figure 63: Cross section segment with required ring reinforcement .................................................. 72
Figure 69: Stress distribution............................................................................................................. 76
Figure 70: Overview requirements for tunnel lining .......................................................................... 77
Figure 71: Comparison building time traditional and tunnelling with marginal cover method (2)....... 80
Figure 72: Stresses at z = 5 m below centre (130 squares) ................................................................. 89
Figure 73: Stresses at z = 5 m below corner (45 squares) ................................................................... 89
Figure 74: Stresses at z = 10 m below centre (60 squares) ................................................................. 89
Figure 75: Stresses at z = 10 m below corner (30 squares) ................................................................. 89
Figure 76: Comparison soil stresses Boston and Newmark ................................................................ 90
Figure 77: Overview CPT locations (bird-eye) .................................................................................... 91
Figure 78: Legend of the CPT’s .......................................................................................................... 91
Figure 79: Overview CPT locations .................................................................................................... 91
Figure 80: CPT location 1 ................................................................................................................... 92
Figure 86: Geological information about location .............................................................................. 98
Figure 87: Geological data of DINO loket (south-east to north-west) [37] .......................................... 98
A.H. Oldenhave 129

Figure 88: Location of the 5 geological results ................................................................................... 98

Figure 89: Schematisation of break-up of a tunnel [9] ..................................................................... 102
Figure 90: Forces acting on tunnel under groundwater level [9] ...................................................... 102
Figure 91: Dimensions tunnel .......................................................................................................... 102
Figure 92: Input .............................................................................................................................. 108
Figure 93: Tunnel lining with 28 springs (purple lines) and forces caused by the soil (red lines) ....... 108
Figure 94: Removal of springs above surface [39] ............................................................................ 110
Figure 95: Cross section column with bending and pressure forces. On the right side the strain
distribution and the forces acting on the column are drawn. The normal force acting in the right
reinforcement bars are zero (reinforcement cannot react as pressure force). ................................. 112
Figure 96: Theoretical and practical strain distribution .................................................................... 112
Figure 97: Detailed planning construction time traditional method (2) ............................................ 121
Figure 98: Detailed planning construction time tunnelling with marginal cover method (2) ............. 127
A.H. Oldenhave 130

List of tables
Table 1: Large diameter bored tunnels in the Netherlands ................................................................ 13
Table 2: Minimum and maximum jacking pressures for different depths ........................................... 24
Table 3: Friction force for the different depths .................................................................................. 26
Table 4: Total forces acting in/on TBM for different depths............................................................... 27
Table 5: Weights different parts TBM [5]........................................................................................... 33
Table 6: (increase of) stresses of the soil ........................................................................................... 35
Table 7: Friction ratio and cone resistance for different kind of soil types [6] .................................... 38
Table 8: Soil structure of location ...................................................................................................... 39
Table 9: Total settlements Holocene layer with TBM at 2.5 meter depth ........................................... 40
Table 10: Total settlements Holocene layer with embankment (5 meter) and soil replacement ........ 41
Table 11: Results of measurements at Botlekspoortunnel [32] .......................................................... 43
Table 12: Soil characteristics ............................................................................................................. 44
Table 13: Optimal MIP column pattern and other possibilities .......................................................... 46
Table 14: Soil structure of location [30] ............................................................................................. 50
Table 15: Summary of the temporary situation ................................................................................. 51
Table 16: Summary of the different zones ......................................................................................... 53
Table 17: parameters of the different soil layers (according to NEN 6740)......................................... 61
Table 18: Estimated parameters for the three calculated cross sections (see further) ....................... 61
Table 19: Safety factors ..................................................................................................................... 62
Table 20: Forces acting in the lining when lt = 170 mm ...................................................................... 64
Table 22: Requirements for one segment.......................................................................................... 68
Table 23: Minimum and maximum reinforcement [12] ..................................................................... 69
Table 26: Requirements for one segment.......................................................................................... 73
Table 29: Requirements per segment................................................................................................ 76
Table 30: Maximum allowable mean bar diameter km [mm] ........................................................... 78
Table 31: Maximum allowable bar distance s [mm] ........................................................................... 78
Table 32: Summary costs and differences between both methods (rounded to hundreds of
thousands) ........................................................................................................................................ 79
Table 33: Comparison building time traditional and “tunnelling with marginal cover” method (1)..... 80
Table 34: (increase of) stresses of the soil ......................................................................................... 90
Table 35: Soil structure of location 1 ................................................................................................. 92
Table 41: Settlements when applying TBM at depth 2.5 meters. Soil replacement and embankment of
4.19 meters ...................................................................................................................................... 99
A.H. Oldenhave 131

Table 42: Settlements when applying TBM at depth 4 meters. Soil replacement and embankment of
4.19 meters ...................................................................................................................................... 99
Table 43: Settlements when applying TBM at depth 7 meters. Soil replacement and embankment of
3.44 meters ...................................................................................................................................... 99
Table 44: Settlements when applying TBM at depth 7.53 meters. Soil replacement and embankment
of 3.24 meters ................................................................................................................................ 100
of 2.84 meters ................................................................................................................................ 100
of 2.43 meters ................................................................................................................................ 100
of 1.99 meters ................................................................................................................................ 101
Table 48: Settlements when TBM rests on Pleistocene layer (depth = 11 meters). No ground
improvement .................................................................................................................................. 101
Table 49: Temporary situation with safety factors ........................................................................... 104
Table 50: Permanent situation with safety factors and concrete ballast .......................................... 105
Table 51: Permanent situation with safety factors and slag ballast .................................................. 106
Table 52: Permanent situation with safety factors and sand ballast ................................................ 107
Table 53: Limit of shear stresses in concrete ................................................................................... 113
Table 54: Maxima critical cross sections .......................................................................................... 114
Table 55: Detailed planning construction time traditional method (1) ............................................. 118
Table 56: Detailed planning construction time tunnelling with marginal cover method (1) .............. 124
A.H. Oldenhave 132

TBM Basics TUdelft

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

TBM Basics TUdelft

Uploaded by

Copyright:

Available Formats

Starting a large diameter TBM from

Master of Science Thesis

Master of Science Thesis

Technical University of Delft

Faculty of Civil Engineering and Geosciences

Department of Geotechnical Engineering

Voor de haalbaarheidsstudie zijn allereerst de zettingen in de Holocene laag bepaald wanneer de

5.5. Soil investigation ............................................................................................................... 38

Appendix F: LDesign [39] ............................................................................................................... 108

1.2.Approach methodology of master thesis

2.1.1. Conventional method [1], [2], [9], [23]

Table 1: Large diameter bored tunnels in the Netherlands

Bored tunnel Completed in Type of shield

Figure 1: Tunnel boring machine [23]

2.1.1.2. Type of TBM

Figure 2: Required support pressure in EPB shield [9]

2.1.1.3. Minimum cover

2.1.1.4. Start and arrival of the tunnelling process

2.1.1.5. Cross connections

2.1.2. Tunnelling with marginal cover method[14], [15], [24]

2.1.2.1. Origin of the method: URUP method

2.1.2.2. Tunnelling with marginal cover method in the Netherlands

2.1.3. Differences between conventional method and tunnelling with marginal

3. Reference design “The Rijnlandroute” [4], [25]

3.1.Introduction of the bored tunnel design

3.2.1. Ground water levels

3.3.Lining geometry of bored part

 Inner diameter bored tunnel: 9.73 m;

3.4.Total length of tunnel

 Open part western ramp: 169 m;

In the figure it is clear that at the location of the

According to guidelines the first 250 meters of

Present diameter when applying 10.53 m is √ m

Figure 8: Cross section tunnel Rijnlandroute with dimensions [4]

Figure 9: Scope of the thesis (eastern ramp)

4. Consequences driving with low cover

4.1.Deformation of Holocene layers due to driving with marginal cover

4.2.Effects on buoyancy of bored tunnel due to driving with marginal

4.3.Effects on tunnel lining due to driving with marginal cover

4.4.Effects on face stability whilst driving with marginal cover

4.4.1. Support pressure

4.4.1.1. Active and passive support pressures

Table 2: Minimum and maximum jacking pressures for different depths

4.4.2. Shield friction

Table 3: Friction force for the different depths

4.4.3. Required force to guarantee water tightness

4.4.4. Alignment control

4.4.5. Recommended jacking forces

Table 4: Total forces acting in/on TBM for different depths

4.6.Effects on torque of TBM due to driving with marginal cover

4.7.Effects on grouting of the tail void for marginal cover

Further research to prevent flow out of the mortar is recommended.

4.8. Launching procedure for driving with marginal cover

Figure 13: Launching procedure “tunnelling with marginal cover” method

Figure 15: Steel frame connected (welded) to

Figure 16: “First” cross connection at surface level

Figure 17: Cross connection as designed by Movares [4]

1. Effects on deformations in Holocene layer;

5. Deformations of Holocene layer due to shallow driving

5.2.Pressure caused by the TBM

Table 5: Weights different parts TBM [5]

Description Weight Contributes to (see Figure 18)

Figure 18: Weights and lengths TBM (EPB shield)