See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/288030845

Grouting the tail void of bored tunnels: the role of hardening and
consolidation of grouts

Article · March 2006

DOI: 10.1201/NOE0415391245.ch42

CITATIONS READS

5 169

2 authors:

A.M. Talmon Adam Bezuijen

Deltares Ghent University
60 PUBLICATIONS   805 CITATIONS    261 PUBLICATIONS   1,692 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

TBM tunnelling View project

Monitoring and modelling of submerged retrogressive breach flow slides View project

All content following this page was uploaded by Adam Bezuijen on 04 July 2017.

The user has requested enhancement of the downloaded file.

Grouting the tail void of bored tunnels: the role of hardening and
consolidation of grouts

A.M. Talmon
WL/Delft Hydraulics, Delft, The Netherlands

A. Bezuijen
GeoDelft, Delft, The Netherlands

ABSTRACT: The role of the physical properties of grout and operational grouting conditions on grout pressures
over tens of metres behind a TBM is quantified. For that purpose the transition from the plastic flow of the grout
behind a TBM to fixation of the tunnel lining by hardened grout further behind the TBM has been calculated
numerically. Account is given to the rheology of the grout, hardening of the grout, consolidation of the grout,
mechanical properties of the soil and movement of the tunnel lining behind the TBM. As an example the
influence of hardening and consolidation of grout is quantified for Sophia Rail Tunnel. Recommendations are
given regarding the use of grouts.

1 INTRODUCTION 2 FIELD DATA

The grout pressure behind a TBM is critical with Monitoring of the Sophia Rail Tunnel has provided
respect to the deformation of the soil and the load- accurate data on grout pressures, Bezuijen et al. 2004.
ing on the tunnel lining. Amongst other influences, Fourteen pressures gauges were mounted in one of
the physical properties of grout are an important fac- the rings of the tunnel lining. The outer diameter of
tor. Grouting research in the Netherlands on monitored the lining is 9.45 m. The pressure gauges in the lining
tunnels, with different grouts, provided an opportu- were recorded for 15 hours.
nity to determine the role of the physical properties of Figure 1 shows the pressures on the lining, orga-
grouts on the quality of grouting. nised as a function of the distance behind the TBM.
A finite difference model for the calculation of grout
pressures directly behind a TBM was first presented 0.50
crown
0 Bb
Ka
by Talmon et al. 2001. Also a separate theory for the 315
Ba
A6 45
0.45 A4b Kb
buoyancy dominated region far behind a TBM was A5
Ca
A1
A4a pressure
given. This model and the buoyancy theory have been 0.40 270
sensor 90
A3b Cb
validated against field data from the Sophia Rail Tun- A4 injection
0.35 A2
A1a
point
Grout pressure (MPa)

A3a
nel and the Botlek Rail Tunnel, Bezuijen et al. 2004 0.30
225
A2b
A3
A1b
135

A2a180
and Pennekamp et al. 2002. invert
0.25
Theory and models for the transition between the
injection dominated region close to the TBM and 0.20

the buoyancy dominated region further away were 0.15

A2a A4a
absent. In this transitory region the consolidation and 0.10
A2b A4b
A3a Ba
hardening of grouts is an important factor. A3b Bb
0.05
This paper concerns the quantification of the role
of hardening and consolidation of grouts on the pres- 0.00
-2 0 2 4 6 8 10 12 14 16
sure distribution. Grout pressures are calculated by Displacement (m)
means of a recently developed finite difference model.
The model covers both the injection dominated region Figure 1. Results of field measurement on the grout pres-
behind the TBM and the region where buoyancy and sure at Sophia Rail Tunnel as a function of the distance behind
fixation of the tunnel lining are dominant. the TBM.

319

Copyright © 2006 Taylor & Francis Group plc, London, UK

This distance is presented in metres from the moment
the pressure gauges in the lining came into the grout.
When the TBM halted the grout pressures dropped
slowly. When drilling recommenced the pressures
restored quickly, Bezuijen et al. 2004.
The grout pumps produce variations in the grout
pressure. These are visible up to the fifth ring after the
TBM, which indicates that the grout is still workable.
From the rheology it is known that chemical hardening
of the cement commences about 5 hours after prepa-
ration of the mortar. This corresponds to a distance
of about 3 m behind the TBM. For the first three tun-
nel rings the measured grout pressures are about the
same, then there is a decline. Also the pressure differ-
ence between the crest and the bottom decreases here.
For x > 6 m the grout pressures become comparable to
the pore water pressure and grout pressures converge.
It is pointed out that the vertical pressure gradient
directly behind the TBM is only 18 kPa/m (for com-
parison: the theoretical static pressure gradient of the
grout is 20.5 kPa/m). At the end of the registration the
vertical pressure gradient is about 7 kPa/m, which is
less than the pore water gradient, but corresponds with
buoyancy of the lining.
3 MODEL
3.1 Grout
Grout pressures are determined by grout injection,
soil deformation, physical properties of the grout and
movement of the lining.
Different types of grouts are used in back-fill grout-
ing: mortars, inert grouts and two-component grouts.
Physical properties vary. Mortars consist of gravel, sil-
ica, cement and additives. After a certain induction
period (few hours) cement hardening starts (time base
hardening: hours). Inert grouts generally consist of
fine sand, silica and fly-ash. The puzzolanic harden-
ing of this type of grout proceeds at a very slow pace
(time base hardening: weeks). Two-component chemi-
cal grouts harden very quickly when mixed (time base:
minutes). In permeable soil all types of grout suffer
fluid loss when pressurized. Mortars and inert grouts
are known to consolidate moreover: a grout cake is
formed on the grout-soil interface. The grout cake is
stiffer than non-consolidated grout.
The specific density of mortars and grouts is
between 1.8 and 2.2. The specific density of two-
component grouts is generally lower because of high
fluid content.
In calculating grout pressures around the tunnel
lining account has to be given to the specific den-
sity of the grout, to rheological properties, chemical
hardening, consolidation, fluid loss, continuity and
the momentum balance of grout. At some distance
behind the TBM the deformation of surrounding soil
320
Copyright © 2006 Taylor & Francis Group plc, London, UK
and the movement of the tunnel lining influence the
grout pressures.
3.2 Continuity and momentum balance
For reasons of simplicity, only one side of the tunnel
is modelled and symmetry is assumed. The continuity
and momentum equations of the grout layer are solved
in a 2-D region extending from the rear of the TBM
to far behind, say 50 m. The model advances with the
TBM and is time-independent. It applies only to the
drilling phase, during which the TBM is advancing
and grout is being back-filled. This phase is the most
critical. The continuity equation in the moving frame
of reference is:
where h = thickness of the grout layer; q = fluid
loss; vs , vn = velocities with respect to moving co-
ordinates; s = axial co-ordinate; n = tangential co-
ordinate (n = D/2 cos θ); θ = angular co-ordinate; and
D = diameter tunnel.
The two momentum balance equations are:
where p = isotropic grout pressure; τs , τn = shear
stresses; and ρg = density grout.
3.3 Fluid loss due to consolidation
In case of permeable soil, the grout consolidates
when pressurized and fluid loss occurs. Fluid loss is
accounted for in the continuity Equation (1). Fluid loss
from the grout loss is a function of physical properties
of the grout, physical properties of surrounding soil,
grout pressures and time. In the model, it is the sum
of fluid loss during drilling and fluid loss during stand
still of the TBM. Fluid loss is mimicked by:
where pw = pore water pressure; s0 = length scale fluid
loss; and q0 = empirical constant.
The role of permeability is accounted for by an
assumed linear dependency of grout pressure. The
role of consolidation is mimicked by exponential
decay with distance. This mathematical formulation
is temporarily and is to be improved.
In the specific case of stand-still conditions, the
role of permeability and mechanical behaviour of
soil was quantified by Bezuijen & Talmon 2003b.
Consolidation and fluid loss are subject of ongoing
research.
 3
Shear stress tau [kPa]
1
0
-1
-2
-3
-0,6 -0,4 -0,2 0 0,2
Strain dx/dz [-]
Figure 2. Definition sketch stress-strain model for harden-
ing grout (Ggrout = 6000 Pa).
3.4 Tunnel lining
The shape of the tunnel lining is input to the model.
In the model, the lining is assumed circular and the
vertical position of the centerline (zl ) is a function of
distance behind the TBM:
where λ = equivalent natural length of the lining; and
ztbm = vertical displacement lining at the rear of the
TBM compared to position far behind TBM. Except
for plastic flow close to the TBM, small displacements
of the lining will readily mobilise grout stresses that
balance buoyancy.
3.5 Grout: stress-strain modelling
When grout is loaded, elastic deformations are taking
place before plastic yielding. The applied stress-strain
model is shown in Figure 2.
Because of consolidation a grout cake is formed.
For the calculation of shear stresses in the tail void
account is given to both the properties of the con-
solidated grout cake and the properties of the more
deformable non-consolidated grout. Generally the
grout-soil interface has different frictional properties
than the grout-lining interface. In the model it is
assumed that there is only friction at the grout-soil
interface and no friction at the grout-lining interface,
see Talmon et al. 2001. The distribution of shear stress
and deformation across the grout layer is indicated in
Figure 3. By means of integration of the elastic defor-
mation in both layers a relation between the shear stress
at the grout-soil interface (τgs ) and the corresponding
average elastic deformation is obtained:
Copyright © 2006 Taylor & Francis Group plc, London, UK
h
0<t<5 hour hcake
6 hour
7 hour
0,4 0,6 Figure 3. Sketch of elastic grout deformation in the tail void,
shear stress and deformation profile.
where X ≡ hcake /h; Ggrout = shear modulus noncon-
solidated grout; Gcake = shear modulus consolidated
√ = layer-averaged deformation of the
grout cake; x
grout x ≡ s2 + n2
; sn = components of
grout deformation; τgs ≡ τs2 + τn2 = shear stress at
grout-soil interface; and hcake = thickness consoli-
dated grout cake.
The maximum shear stress at the grout-soil inter-
face is reached when the shear stress at the interface
between both grout layers equals the yield stress of
non-consolidated grout:
where τy = yield stress non-consolidated grout; and
τmax = maximum shear stress at groutsoil interface.
At τmax the grout reaches its maximum elastic defor-
mation: xmax . For |τgs | < τmax the deformations are
elastic. In plastic flow the grout is sheared to maximum
deformation. The change of deformations s and n
with distance behind the TBM is calculated by:
where us (=vs -vt ), un (=vn ) = flow velocities with
respect to fixed co-ordinates; and vt = velocity TBM.
The Equations (7) were obtained from the defini-
tions: ∂s/∂t ≡ us , ∂n/∂t ≡ un , and application of
a time-distance transformation: ds = vt dt.
The shear stress components are calculated by:
The pressure distribution around the tunnel lining is
calculated by means of integration of Equation (2).
3.6 Deformation of surrounding soil
The deformation of the grout-soil interface (dr) is cal-
culated with the aid of Verruijt’s 1993 simplified linear
321
stress-strain relation for a perfectly elastic soil: Table 1. Input parameters.

Parameter Symbol Value

Velocity TBM vt [m/s] 0.0007

where Gsoil = shear modulus soil; pfar = grout pres- Skewness grout supply ustop /usbot 2
sure far behind the TBM; and r = radius grout-soil Diameter lining D [m] 9.45
Thickness tail h [m] 0.2
interface.
void at TBM
The grout pressure far behind the TBM (pfar ) is Natural length lining λ [m] 10
determined by buoyancy, and is a linear function of Displacement lining ztbm [m] −0.0003
height, Bezuijen et al. 2004. at TBM
Density water ρw [kg/m3 ] 1000
Density grout ρg [kg/m3 ] 2200
3.7 Thickness grout layer
Yield stress fresh grout τy [Pa] 1000
The variation of the thickness of the grout layer is Hardening rate grout d τy /ds [Pa/m] 500
calculated by: Shear modulus soil Gsoil [Pa] 1 10+8
Shear modulus grout Ggrout [Pa] 6000
Shear modulus cake Gcake [Pa] 30000
Grout pressure crest TBM pgrout [Pa] 250000
Pore water pressure crest pw [Pa] 140000
Fluid loss coefficient q0 [m/s] 1 10−11
Adaptation length s0 [m] 6
fluid loss
Weight lining F [Pa m] 640000
The behaviour of soil underneath the tunnel is gen- + backup-up eq.
erally stiffer than the soil above. For Gsoil a variation Porosities ne , ni ne , ni [-] 0.28, 0.33
around the circumference can be inputted to simulate
soil layering and soil surface influences.
measured vertical pressure gradient behind the TBM
3.8 Thickness grout cake is about 18 kPa/m. In the model this vertical pressure
gradient is ruled by the inputted skewness of grout sup-
The thickness of the consolidating grout cake that is ply over the rear of the TBM. Grout supply at the crest
being formed on the grout-soil interface is calculated is generally higher than at the bottom. Hence there
with the aid of filtration theory. The growth of the is a downward flow, influencing the vertical pressure
grout-cake with distance behind the TBM is calculated gradient. The true distribution of grout over the injec-
by continuity: tion pipes might differ however because local flow is
not considered in the model. In design calculations the
more refined grout flow model of Talmon et al. 2001
might be used to determine the pressure distribution
directly behind the TBM before calculations with the
where ni = initial porosity fresh grout; and ne = end present model are conducted.
porosity consolidated grout. The thickness of the grout layer is taken larger than
the geometrical thickness (which is 0.16 m at Sophia
3.9 Numerical model Rail Tunnel). The reason is that the measured overin-
jection is about 30%, while physical allowable fluid
The model is implemented in a spreadsheet. All equa- losses are maximal 5 a 10 % when grout is compacted.
tions are solved iteratively. Fluid loss and the thickness The vertical displacement of the lining with distance
of the grout layer are a function of the grout pressure. has been adjusted such that the grout approximately
Therefore the velocity field has to be recalculated up balances buoyancy forces far behind the TBM. This
to convergence of the computation. The principal out- inputted vertical displacement is less than measured.
put of the model is the pressure distribution around the In between drilling phases the tunnel lining is known
tunnel lining. to move, causing plastic deformation of the grout. This
might result in displacements that are not considered
in the model.
4 RESULTS The inputted shear modulus of grout is derived from
vane tests on the grout used at Sophia Rail Tunnel. The
4.1 Input inputted shear modulus of soil agrees with measured
The input parameters for the calculation of grout pres- deformations of the soil adjacent to the lining. These
sures at Sophia Rail Tunnel are listed in Table 1. The are 1 a 2 mm. The consolidation analysis of Bezuijen

322

Copyright © 2006 Taylor & Francis Group plc, London, UK

 500 grout pressure distribution [Pa]
Grout pressure [Pa] Thousands

crest
400 middle x=0 [m] 9,4
Thousands
500 0,8 10,2
bottom
1,6 11,0
300
300 2,4 11,8
3,1 12,6
200 100 3,9 13,4
4,7 14,1
5,5 14,9
100 6,3 15,7
0 4,5 9 13,5 18 22,5 7,1 16,5
distance behind TBM: s [m] 7,9 17,3
8,6
Figure 4. Calculated grout pressure for Sophia Rail Tunnel
as a function of the distance behind the TBM.

and Talmon 2003b pointed at the same numerical value Figure 5. Calculated grout pressure distribution around the
tunnel.
for the shear modulus of soil. To mimic the differences
in soil stiffness, the value at the crest of the tunnel is
taken 50% lower and at the bottom 50% higher than 5 20
listed in Table 1. kPa/m

Buoyancy gradient [kPa/m]

The fluid loss coefficient is adjusted to give a fluid 4
dh crest 15
loss of about 5%. At measured porosities of ni = 0.33
z [m] Thousands

and ne = 0.28 this fluid loss leads to a thickness of the 3

dh middle
grout cake of about half the thickness of the tail void. dh bottom 10
The weight of the lining (F) is used to calculate the
2
theoretical grout pressure distribution far downstream
where influences of the TBM have ceased. 5
1

0 0
0 5 10 15 20 25
s [m]
4.2 Calculation results
Figure 6. Calculated deformation of grout-soil boundary
The calculated grout pressure as a function of the dis-
and calculated buoyancy force.
tance behind the TBM is given in Figure 4. There
is a good agreement with the measured pressure
distribution, compare with Figure 1. 2,5

The theoretical grout pressure far behind the TBM, 2 crest

where buoyancy dominates, is shown at s = 19.6 m for middle
reference purpose. The calculated grout pressure dis- 1,5
q [m/s] Millions

bottom
tribution around one side of the tunnel is shown in 1
Figure 5.
The calculated deformation of the grout-soil inter- 0,5

face is shown in Figure 6. By virtue of Equation (9) the 0

calculated soil deformation is taking place in concert
with grout pressures. -0,5
0 5 10 15 20
The buoyancy force, which follows from integra- s [m]
tion of calculated grout pressures, is also given in
Figure 6. A smooth transition from a pressure gra- Figure 7. Calculated fluid loss.
dient of 16 kPa/m in the injection dominated region
directly behind theTBM to a pressure gradient of about
8 kPa/m in the buoyancy dominated region further The calculated fluid loss is shown in Figure 7. The
behind the TBM is calculated. The latter corresponds fluid loss is largest at the bottom, because there the
with the weight of the tunnel lining plus back-up difference between the grout pressure and pore water
equipment. This result agrees with measured vertical pressure is the largest. The fluid loss decays as a
gradients, Bezuijen & Talmon 2003a, 2004. function of distance and grout pressure.

323

Copyright © 2006 Taylor & Francis Group plc, London, UK

 0,15 stress and consolidation are important for fixation of
the tunnel lining.
Thickness grout cake [m]

Hardening and consolidation of grout are impor-

0,1 tant. Only the first stage of hardening is relevant to the
grouting process. It varies between grout type and can
crest
be influenced by grout recipe.
0,05 1/4 Consolidation promotes the fixation of the tunnel
middle lining in the grout, because of increased resistance of
3/4 the grout layer. For fixation of the lining, the yield
bottom stress of non-consolidated grout should not be as high
0
0 5 10 15 20 as would be in absence of a grout cake. Consolida-
sm[ ] tion can be influenced by grout recipe, specifically
where it regards permeability. Low permeability of
Figure 8. Calculated thickness grout cake. surrounding soil and bentonite from the tunnel face
will influence consolidation.
5 For impermeable soil, where consolidation will be
absent, and no cake will form, the yield stress should
crest s
4 be at least 3.5 kPa (this value is for dead weights like
Tau [Pa] Thousands

middle s
at Sophia, see calculated n-component at mid-height
3 bottom s in Figure 9).
crest n Our experience indicates that fine grained grouts
2 middle n might have higher frictional resistance with the tunnel
bottom n lining than coarse grained mortars. This issue needs
1 to be resolved. It is yet unknown whether this is truly
yield stress
the case, or that other processes are involved such as
0 stresses mobilised by movement of the tunnel lining
0 5 10 15 20 25
s [m] behind the TBM. Major differences have been found
in frictional resistance of supply lines between coarse
Figure 9. Calculated shear stress components in the grout. and fine grained grouts.

The calculated thickness of the grout cake is shown

in Figure 8. At about 7 m behind the TBM, the cake 6 CONCLUSIONS
has developed nearly completely. The grout cake at the
bottom is thicker because of larger fluid loss. It is possible to quantify grout pressures continuously
The calculated shear stress components are shown over distances of tens of metres behind a TBM.
in Figure 9. As inputted, at 2 m behind the TBM chem- The pressure distribution close to the TBM is deter-
ical hardening of the grout commences. The figure mined by the injection strategy, further away this is
shows that in the plastic region the shear stresses are determined by the dead weight and buoyancy.
predominantly directed in axial direction. The transition from grout injection at the TBM to
At about 6 m behind the TBM the grout is pushed fixation of the buoyant tunnel involves elastic-plastic
back in the direction of the TBM due to shrinking of properties and consolidation of the grout.
the soil cavity. In this elastic region the direction of The pressure loading on the tunnel lining is
shear stresses changes to vertical, which is caused by calculated.
upward displacement of the tunnel lining. Here the By employing this calculation method, recommen-
shear stresses are less than the yield stress, the grout dations can be given regarding the use of different
is providing fixation to the tunnel lining. Figure 9 types of grout as a function of soil properties (perme-
provides valuable information on yield stress of non- ability and stiffness), where and how to inject different
consolidated grout versus shear stresses that fixate the grouts, and constraints on the physical properties of the
tunnel lining. grouts (yield stress, etc.).

5 RECOMMENDATIONS ACKNOWLEDGEMENTS

The yield stress of fresh grouts is important for control The authors would like to acknowledge COB research
of pressure gradients directly behind the TBM through at Sophia Rail Tunnel and Tubecon (contractor for
injection strategy. This yield stress should be about 1 Sophia Rail Tunnel) for permission to publish field
a 2 kPa. At some distance behind the TBM the yield data.

324

Copyright © 2006 Taylor & Francis Group plc, London, UK

REFERENCES
Bezuijen A., Talmon A.M., Kaalberg F.J. and R. Plugge, 2004,
Field measurements on grout pressures during tunnelling
of the Sophia Rail Tunnel, J. of Soils and Founda-
tions of the Japanese Geotechnical Society, Vol.44, no.1,
pp 39–48.
Bezuijen A. and A.M. Talmon, 2003a, Grout pres-
sure measurements during tunnelling, Proc. Reclaiming
Underground Space (ITA Amsterdam), A.A. Balkema,
Rotterdam.
BezuijenA. andA.M.Talmon, 2003b, Grout the foundation of
a bored tunnel, Proc. Int. Conf. on Foundations “Innova-
tions, Observations, Design & Practise” (ICOF Dundee),
Thomas Telford, London.
BezuijenA. andA.M.Talmon, 2004, Grout pressures around a
tunnel lining, influence of grout consolidation and loading
325
Copyright © 2006 Taylor & Francis Group plc, London, UK
View publication stats
on lining, Proc. Underground Space for Sustainable
Urban Development (ITA Singapore), Elsevier.
Maidl B., Herrenknecht M. and L. Anheuser, 1996, Mecha-
nised Shield Tunnelling, Ernst & Sohn, Berlin.
Pennekamp J.G.S., Talmon A.M. & A. Bezuijen, 2002, Muck
control and mortar backfill grouting at Botlek Rail Tunnel:
Learning from specific tests related to tunnels, Pre-proc.
of the Workshop Measuring and Predicting the Behaviour
of Tunnels, Delft University of Technology.
Talmon A.M., Aanen L., Bezuijen A. and W.H. van der Zon,
2001, Grout pressures around a tunnel lining, proc. IS-
Kyoto 2001 conference on Modern Tunneling Science and
Technology, A.A. Balkema, Rotterdam.
Verruijt A., 1993, Soil Dynamics, b82, Delft University of
Technology.