Precision tunneling in the vicinity of

World Heritage Properties in Barcelona

S. Kurze1
Ingenieursoziett Prof. Dr.-Ing. Katzenbach GmbH, Frankfurt am Main, Germany

ABSTRACT
As a part of the new Spanish high speed railway line (AVE) connecting Madrid, Barcelona and the French border, currently a
tunnel with a length of 5.6 km is built under the city centre of Barcelona. The tunnel line passes directly next to the famous
church of Sagrada Familia and a building called Casa Mil. Both buildings belong to the World Heritage Properties of the
UNESCO. To ensure the highest possible safety for these extraordinary historical buildings special requirements in control and
construction had to be fulfilled during the tunnelling process. The tunnel is built with a tunnel-boring machine using an earth
pressure balanced shield. The church of Sagrada Familia and Casa Mil have already been passed. The maximum measured
settlements of Sagrada Familia and Casa Mil induced by the tunnel construction were in the range of the measurement
accuracy. Some of the extensive data of the geotechnical and geodetic monitoring is presented focusing on possible influencing
factors for the evolution respectively the prevention of surface settlements.
Keywords: tunnel, TBM, EPB-shield, settlements, World Heritage Properties, monitoring, observational method

INTRODUCTION

As a part of the new Spanish high-speed railway

line (AVE) connecting Madrid, Barcelona and
the French border, currently a double tracked
tunnel with a length of 5.6 km is built under the
city centre of Barcelona.
The tunnel line passes directly next to the
famous church of Sagrada Familia (Figure 1) and
a building called Casa Mil. Both were
planned by the architect Antoni Gaud and
belong to the World Heritage Properties of the
UNESCO. Special requirements in control and
surveillance had to be fulfilled during the
tunnelling process to ensure the entire safety of
these two outstanding buildings.

Figure 1. Nativity faade of Sagrada Familia

Ingenieursoziett Prof. Dr.-Ing. Katzenbach GmbH, Frankfurt am Main, Germany. kurze@katzenbach-ingenieure.de

To ensure the entire safety of the church

during the tunnelling process, even in the case of
incidents, a bored pile wall was installed in the
ground between the tunnel and the Glory faade
of Sagrada Familia. This protection wall consists
of approx. 40 m long piles with a diameter of
1.5 m (Figure 2). A similar protection wall was
built in vicinity of Casa Mil.
2.3

Figure 2. Cross section Sagrada Familia and AVE-tunnel

2
2.1

PROJECT OVERVIEW
AVE-tunnel

The AVE-tunnel connects the existing station

Sants with the future station Sagrera. The tunnel
passes under the high-density urban area of
Barcelona. The tunnel has an outer diameter of
11.55 m. The bottom of the tunnel is located in a
depth of at most 40 m under the ground surface.
The average groundwater table is approx. 19 m
above the bottom of the tunnel (Figure 3).
2.2

Soil and groundwater conditions

The soil layers passed by the tunnel-boring

machine are mainly tertiary layers (Figure 3). In
the first kilometre of the tunnel the TBM passes
tertiary clay. Then, a section of tertiary silty
sands follows. In this section, the TBM passes
the world heritage properties Sagrada Familia
and Casa Mil. The soil layering in the vicinity
of Sagrada Familia is given as follows: Below
the artificial fillings with a thickness of up to 2 m
quaternary sandy silts in alternating layering with
silty sands are found with a total thickness of 410 m. Tertiary sand was observed down to the
explored depth of 60 m. Tertiary clay layers of
various thickness of 0.4-2.0 m are intercepted in
the tertiary sands.

Sagrada Familia

The AVE tunnel lies in a horizontal distance of

only 4 m parallel to the Glory faade of Sagrada
Familia. Next to Sagrada Familia, the bottom of
the tunnel is at a depth of app. 37 m (Figure 2).
The construction of Sagrada Familia began in
the year 1882 and is still not finished. The end of
the construction work is currently planned for
2026. The church of Sagrada Familia has a 50 m
high main nave with a length of 90 m and is
planned to have 18 steeples, from which the
highest is planned to have a height of 170 m
(Figure 2). Until today, the main nave, the
Nativity and the Passion faade with altogether 8
steeples are finished (Figure 1). The construction
works on the 6 central steeples and on the socalled Glory faade have started.
The church of Sagrada Familia has a pile
foundation. The piles under the main nave are
estimated to have a depth of approx. 20 m.

Figure 3. Geotechnical longitudinal section

Due to these dense clayey interceptions

various aquifers underlie the city of Barcelona.
The highest aquifer is a free aquifer, the lower
ones partly have confined groundwater
conditions. In the vicinity of Sagrada Familia,
the free groundwater level is situated about
16.5 m below the surface.

2.4

Tunneling method

The tunnel is built with a tunnel boring

machine (TBM), using an earth pressure
balanced shield (EPB) (Figure 4).
The TBM is working and monitored
continuously 24 h a day. With the chosen EPB
shield, the soil is conditioned with water and
foam injections at the cutter head. The
homogenized, excavated earth slurry is used as
support medium [4], [5]. The gap between the
TBM shield and the excavated soil is injected
with bentonite [8]. The gap behind the tail of the
shield is permanently grouted with mortar to
provide a compensation grouting procedure [6].

Figure 4. Face of the TBM

DISPLACEMENTS RELATED TO EPB

TUNNELING

In general, earth pressure balanced shield

machines are chosen particularly in an urban
environment to reduce subsidence risk in
comparison to other tunneling methods [10].
Settlements are evoked by changes in the
stress conditions or changes in pore water
pressure [8]. With an active support pressure of
the face, of the gap between shield and
surrounding soil and of the gap behind the tail of
the shield, these changes can be reduced to a
minimum [8]. Nevertheless, settlements or
ground subsidence occur in every tunnel
construction process.
To characterize the settlement trough
evolution in width and depth over a tunnel
section, the volume loss factor Vl can be used. Vl
describes the volume of the settlement trough

related to the theoretical tunnel volume [2], [6]

(Figure 5).
As can be seen in figure 5, Vl is an
instantaneous value, changing with the position
of the TBM and the analyzed tunnel section. The
final Vl usually ranges from 1 to 2% for tunnels
excavated with the conventional method. In the
case the tunnel is constructed using an earth
pressure balanced shield lower values can be
observed, sometimes below 0.5% [6].

Figure 5. Settlement trough and volume loss factor Vl [2]

The factors influencing the shape, the depth

and the length of the settlement trough related to
EPB tunneling are numerous. Basically, they can
be divided into geotechnical, geometrical and
operational parameters of the TBM [1].
3.1

Geotechnical parameters

With the geotechnical parameters, i.e. the soil

characteristics as for example rigidity, friction
angle, cohesion, deformability, permeability and
abrasiveness the boundary conditions for the
tunneling process are given. Based on a good soil
investigation, the choice of the tunneling method
and the specification of operational parameters
can be done efficiently. Good knowledge of
ground parameters and groundwater conditions
enables realistic calculations and then the
possibility to define requirements and adequate
thresholds for the operational parameters of the
TBM.

3.2

Geometrical parameters

The geometrical tunnel parameters mainly are

the depth of the tunnel, the diameter of the tunnel
and the lining geometry, meaning the thickness
and shape of the lining and the width of the gaps.
Besides the geometry of the tunnel, the distance
and geometry of adjacent buildings and
structures have a significant influence on the
magnitude of settlements [9]. This might be for
example pile foundations, another tunnel or
like in Barcelona a protection wall influencing
the settlement behavior. Also the geometry of the
TBM itself influences the development of
settlements; especially the conical shape of the
shield has to be mentioned in this context [6].
Depth
40 m [m]

0.5 cm
Final surface displacements
above tunnel axis
depth of tunnel below
groundwater

0.4 cm

0.3 cm

0.2 cm

0.1 cm
Displacements
[cm] 0.0 cm

-0.1 cm

TBM in Tertiary clay

d
n
a
s
o
t
y
a
lc
m
o
fr
e
g
n
a
h
C

30 m

20 m

10 m

TBM in Tertiary silty sand

0m

PK [km]
-10 m

-0.2 cm
-20 m
-0.3 cm
-30 m

-0.4 cm

-40 m

-0.5 cm

Figure 6. Surface settlements and geotechnical parameters

3.3

Operational parameters of the TBM

There are numerous operational parameters of

tunnel boring machines with earth pressure
balanced shields, all influencing the reaction of
the soil around the TBM. The following 10 TBM
parameters were identified to have the most
influence on the magnitude of surface
settlements [1]:
- Face pressure
- Torque on the cutting wheel
- Total thrust force
- Power excavating 1 m
- Back filling pressure
- Grouted volume of mortar
- Rate of advancement
- Time for boring and installing 1 ring
- Change in vertical angle of the TBM
- Change in horizontal angle of the TBM

With a numerical study Vanoudenheusden [11]

identified that essentially the rate of
advancement, the torque on the cutting wheel,
the face pressure and the change in vertical angle
of the TBM could be correlated to surface
settlements.
4

MONITORING OF THE AVE TUNNEL IN

BARCELONA

To ensure a safe construction of the AVE tunnel

and to give the maximum possible security for
the sensitive buildings in vicinity the
construction works had to be executed under
special control and supervision requirements
following the observational method according to
Eurocode EC 7 [7]. The monitoring in Barcelona
was realized with a dense grid of geodetic and
geotechnical measurement devices in the
surrounding of the tunnel on the one side and
with a permanent monitoring of the most
important operational parameters of the TBM on
the other side.
To analyze if a clear correlation between the
aforementioned influencing factors and the
magnitude of surface settlements can be derived,
some of the observed data from Barcelona is
presented in the following diagrams.
4.1

Influence of geotechnical parameters

In figure 6 the final surface settlements above the

tunnel axis after the TBM passage are shown for
the part that has already been constructed. The
groundwater level and the zones of significantly
different soil conditions, i.e. tertiary clay in the
first part and then a change to tertiary silty sands,
are marked in figure 6 as well.
The measured surface settlements do not
exceed 0.5 cm over the whole tunnel length. The
volume loss factor Vl is in the range of only
0.1%. The biggest settlements occurred at the
start of the TBM between PK 5+800 and PK
4+700 in the tertiary clay. It is possible to
explain the decrease of the settlements over the
tunnel length by the adaptation of the gained
experience of the soil and tunneling conditions
during the first part of the tunneling process

concerning the definition of adequate thresholds

and limits for the TBM operation.
Small heaving of less than 0.1 cm occurred
between PK 4+200 and 4+300, directly after the
change from tertiary clay to tertiary silty sands.
An influence of the groundwater height over
the bottom of the tunnel on the displacements
cannot be remarked. In all reflections about the
magnitude of the displacements the measurement
accuracy for the surface leveling of approx.
0.1 cm has to be taken into consideration.
4.2

Influence of geometrical parameters

neighbored sections occurred. Although the

maximum values are only about 0.3 cm an
influence of the complex procedure of the TBM
for entering and leaving the shafts including
measures for groundwater drawdown inside
can be remarked.
4.3

Influence of operational parameters

13/ 01/ 2011

40

Construction
time
[weeks]

TBM advance

26/ 08/ 2010

20

08/ 04/ 2010

70

In figure 7, the change of the depth of the tunnel

is shown in comparison to the measured surface
settlements above the tunnel axis.
Because the distance of the adjacent buildings
is almost the same over the whole tunnel length,
just the positions of Sagrada Familia and Casa
Mil are marked.

60
50

Advance
velocity
[mm/ min]

Advance velocity

40
30
20
10
0
20.0
15.0

Torque
on cutting
wheel
[MNm]

Torque on
cuttingwheel

10.0
5.0
0.0

Depth[m]
50 m

0.5 cm
Final displacements
depth of the bottom of the tunnel

0.4 cm

0.3 cm

0.2 cm

0.1 cm
Displacements
[cm] 0.0 cm

tf
a
h
s
t
n
a
x
in
r
T

ft
a
h
s
a
lli
d
a
P

lia
i
m
a
F
a
d
a
rg
a
S

Face pressure

40 m

30 m

fta
h
s
c
u
r
B

3.0
Face
pressure
at top
[bar] 2.0

il
M
a
s
a
C

20 m

10 m

0m

PK [km]

1.0

0.0
0.1 cm

0.0 cm

PK [km]

-0.1 cm
-0.2 cm

-0.1 cm

-10 m

-0.2 cm

-20 m

-0.3 cm

-0.3 cm

-30 m

-0.4 cm

-40 m

-0.5 cm

-50 m

Figure 7. Surface settlements and geometrical parameters

Other significant buildings are the vertical

maintenance shafts for planned maintenance
stops of the TBM that will be emergency shafts
in later operation of the tunnel. There is no
noticeable effect of the depth of the tunnel below
the surface, but it is remarkable that the
settlements in vicinity of Sagrada Familia and
Casa Mil are quite small in comparison to the
other sections. This effect may be caused by an
effect of the bored pile wall or an especially
careful operation of the TBM.
In vicinity of the maintenance shafts at the
crossing of Padilla Street (PK 3+930) and Bruc
Street (PK 2+500) bigger settlements than in the

-0.4 cm

Displacements
[cm]

Final
displacements

-0.5 cm

Figure 8. Surface settlements and operational parameters

Figure 8 shows some of the operational

parameters of the TBM that are expected to be
the most important ones, i.e. the performance
over construction time, the average advance
velocity, the torque on the cutting wheel and the
face pressure at the top of the working chamber,
in comparison to the surface settlements above
the tunnel axis.
A correlation between the comparatively low
face pressure and the comparatively high
settlements (max. 0.4 cm) in the starting section
of the tunnel can be remarked. The face pressure
was smaller according to the lower weight of the
vertical cover above the tunnel crown and the
soil parameters of the tertiary clay in the first

section of the tunnel. The other operational

parameters do not significant influence.
The three larger steps in the graph at the top
of figure 8 (construction time vs. tunnel length)
show the three planned maintenance stops in the
shafts. The smaller steps in the section beginning
at PK 4+200 show additional maintenance stops
of the TBM to change the cutting tools caused by
the higher abrasion in the sandy soil. These
works were made under hyperbaric conditions.
5

CONCLUSIONS

The presented analysis of the data from the EPB

drive in Barcelona show that the observed, very
small surface settlements, that do not exceed
0.5 cm, cannot clearly be correlated with some
special parameters. Among the geometrical
parameters the position of the maintenance shafts
seams to have an influence. Among the chosen
operational parameters the face pressure shows
the greatest influence on the displacements. The
observed data shows that the surface settlements
can be reduced to a minimum with a careful and
highly supervised TBM performance.
Not listed in the main literature sources ([1],
[6], [11]), but nevertheless expected to be very
important concerning the subsidence risk are the
hyperbaric interventions. The conditions during
these hyperbaric interventions are delicate
depending on the effectivity of the bentonite
injections to hold the hyperbaric pressure.
Possible failure can cause sinkholes or blowouts
inducing great displacements and therefore
should be closer looked at in further research.
ACKNOWLEDGEMENT
The author is grateful to supervisor Rolf
Katzenbach and to the German Geotechnical
Society (DGGT) for the opportunity to present
this topic in an international conference.

REFERENCES
[1]

R. Boubou, F. Emeriault, R. Kastner, Correlation

between TBM parameters and ground surface
settlements, neural network method, ISSMGE TC 28,
Budapest, Hungary, (2008), presentation slides.
[2] A. Burghignoli, F. Di Paola, M. Jamiolkowski, G.
Simonacci, New Rome metro line C: approach for
safeguarding ancient monuments. International
Conference GEOMOS, Geotechnical Challenges in
Megacities, Moscow, Russia (2010), 55-78.
[3] DAUB
German
Tunnelling
Committee,
Recommendations for selecting Tunnel Boring
Machines (German only), Taschenbuch fr den
Tunnelbau, (2011), VGE Verlag, Essen, 231-299.
[4] A. Gbl, The interaction of ground, TBM and segment
lining with closed shield machines, Geomechanics and
Tunneling 3 (2010), 491-500.
[5] M. Herrenknecht, G. Wehrmeyer, K. Bppler, W.
Burger, Zuknftige Mglichkeiten und Grenzen der
maschinellen Tunnel-Vortriebstechnik. 50 Jahre
STUVA:
Vergangenheit
trifft
Zukunft,
Studiengesellschaft fr unterirdische Verkehrsanlagen
e.V., Kln (2010),304-321.
[6] R. Kastner, F. Emeriault, D. Dias, Assessment and
control of ground movements related to tunnelling,
International Conference GEOMOS, Geotechnical
Challenges in Megacities, Moscow, Russia (2010), 92119.
[7] R. Katzenbach, S. Strber, Schwierige Tunnelvortriebe
im Locker- und Festgestein Anforderungen an
Erkundung, Planung und Ausfhrung, Geotechnik 4
(2003), 224-229.
[8] B. Maidl., M. Herrenknecht, U. Maidl,, G. Wehrmeyer,
Maschineller Tunnelbau im Schildvortrieb, Ernst &
Sohn Verlag, Berlin, 2011.
[9] R.J. Mair, TC28 tunneling: reflections on advances
over 10 years, Geotechnical Aspects of Underground
Construction in Soft Ground, Proceedings of the 5th
International Symposium TC28. Amsterdam (2005), 310.
[10] T.M. Saczynski, M. Pearce, A. Elioff, Monitoring Earth
Pressure Balance Tunnels in Los Angeles, FMGM:
Seventh
International
Symposium
on
Field
Measurements in Geomechanics, Boston MA, (2007),
ASCE.
[11] E. Vanoudheusden, G. Petit, J. Robert, F. Emeriault, R.
Kastner, J.-Y. Lamballerie, B. Reynaud, Analysis of
movements induced by tunnelling with an earthpressure balance machine and correlation with
excavating parameters, Geotechnical Aspects of
Underground Construction in soft ground, (2006),
Bakker et al (eds), London, 81-86.