1

Tunnel monitoring system A contribution for the preparation of guidelines

G. Lunardi
1
, M. Gatti
1

1
Rocksoil SpA, Milan, Italy

1. INTRODUCTION
In conventional tunnelling, geotechnical monitoring is of fundamental importance as an instrument
for verifying the appropriateness of the operations specified in the design and for calibrating the
intensity and sequence of those operations during construction. It is also important for recording
tunnel behaviour when it is in service, in order to check the condition of the tunnel over time,
especially in relation to the rheological behaviour of the rock mass and possible changes in the
hydrogeological conditions (fault zones, walled section, inflow, etc.).
The approach based on the analysis of controlled deformation in rocks and soils, ADECO-RS [1, 2],
considers two different stages, that of the design and that of the construction of a tunnel. In the first,
the natural pre-existing equilibriums (survey phase) are taken as the basis for predicting the behaviour
of the ground during excavation in terms of the deformation response (diagnosis phase), in order to
decide the type of action to exert confinement or preconfinement to maintain deformation within
acceptable limits for tunnel safety (therapy phase). During the second stage, the deformation
phenomena are measured and interpreted (monitoring phase), firstly to verify the accuracy of the
predictions made during the diagnosis and therapy phases and then to fine tune the design by
adjusting the intensity, location and timing of stabilisation operations and the balance of these
between the face and the perimeter of the excavation and to calibrate the speed and rate of tunnel
advance (operational phase). This is all performed with a view to industrialising the tunnel construction
process and to compliance with schedules and budgets (see Fig.1 flow chart).


Figure 1 ADECO-RS Approach

This paper clarifies the role of monitoring in the design and construction process and illustrates the
criteria followed in the choice of instrumentation and in interpreting the data acquired, with particular
reference to the analysis of the deformation behaviour of the face-core using extrusion-meters.
Some case studies taken from the construction of Apennine road and rail tunnels are examined.

2
2. DEFORMATION RESPONSE AND DESIGN APPROACH
The systematic observation of the stress-strain behaviour of a wide range of tunnels during
construction has clearly shown that the deformation response is complex and involves not just the
tunnel cavity, but also the volume of ground that lies ahead of the face, virtually cylindrical in shape,
with dimensions quite similar to that of the diameter of the tunnel to be excavated. This region, called
the advance core, is affected by a primary component of the deformation response: extrusion,
which manifests on the surface of the face along the longitudinal axis of the tunnel (bellying or rotation
of the face), and preconvergence of the cavity, i.e. the convergence of the theoretical profile of the
tunnel ahead of the face (Fig. 2). These primary components depend on the relationship between the
strength and deformation properties of the advance core and its original stress state [3, 4]. Research
has found that:
there is a close connection between extrusion of the advance core at the face and
preconvergence and convergence of the cavity. Also, deformation in the cavity normally follows
and is strictly dependent on deformation in the advance core, which is the true cause of the whole
deformation process in all its components;
it is possible to control advance core deformation (extrusion, preconvergence), and as a
consequence also to control cavity deformation (convergence), by acting on the rigidity of the core
with measures to protect and reinforce it (Fig. 3);
the rigidity of the core plays a determining role in the stability of a tunnel both in the short and long
term.

The advance core can be seen as a stabilisation instrument and its analysis - in terms of action to be
taken and the related prediction of deformation is very important at the design stage, just as the
monitoring and control of the deformation response of the core-face, and not just of the cavity, is very
important at the construction stage.



Figure 2 Deformation response Figure 3 Stabilisation action

The instruments used ahead of the face are those of preconfinement. They act to facilitate the
formation of an artificial arch effect in advance (a protective function) or to improve the natural strength
and deformation properties of the advance core (a reinforcement function). Typical examples of
preconfinement are fibre glass structural elements inserted around the cavity and in the core, jet-
grouting and full face mechanical pre-cutting. Confinement, on the other hand, acts inside the tunnel,
back from the face, to prevent deformation phenomena from developing after the passage of the face
and consists of a preliminary lining consisting of a shotcrete shell, steel ribs, radial bolts and the tunnel
invert. The composition of these instruments, based on the category of the core-face (A stable, B
stable in the short term, C unstable, [1]), determines the longitudinal and cross section types,
which guarantee the feasibility of the excavation and the long and short term stability of the tunnel.
Owing to the variability of geological and geotechnical conditions, construction section type
variabilities must be designed in order to unequivocally define the type, intensity and rates of the
operations and the range of geological-geomechanical and stress-strain (extrusion and convergence)
conditions within which a given section type can be employed. The monitoring system employs a set
3
of specific guidelines to govern the construction process and to calibrate operations so that they are
appropriate for the local geomechanical conditions.A correlation between the deformation response
and the operations used to control it, with the relative variabilities, is reported in Table 1 below.

Table 1 Deformation response Vs Interventions
Deformation response Interventions Variabilities
Face and contour Preconfinement
(FRP/J G structural elements )
Intensity (N/m
2
), Length (m),
Overlapping (m)
Drainages Number, Length (m)
Estrusion
Preconvergence
(Convergence)
Excavation step and shape Lenghth (m)
Cavity Confinement (shotcrete prelining,
steel ribs, bolts )
Thickness (m), Step (m),
Length (m)
Convergence
Invert and crown casting Face distance (m)

An example is given in Fig. 4 of the section type C2, employed for the excavation of the base tunnel of
the A1 Milan-Naples motorway, between Sasso Marconi and Barberino di Mugello, in rock masses
lying in the Scaly Clays formations, (160 sq. m. approx. of full-face excavation) [5]. Given the variability
of the geomechanical conditions, with GSI values in the 30-40 range, and the related predicted
deformation response, with extrusion values in the 40-80 mm range and diametrical convergence in
the 100-140 mm range, the section type variabilities reported in Table 2 were considered and
governed by means of an appropriate monitoring system.





Variabilities
Section
Types
Intervention
Minimum Nominal Maximum
Steel rib step 1.2 m 1.0 m 0.8 m
N VTR face 50 70 90
VTR face
overl.
10.0 m 12.0 m 14.0 m
Excavation 14.0 m 12.0 m 10.0 m
Invert-face () <2.0 <1.5 <0.5
C2
Crown-face <3.0 <5.0 <7.0
Figure 4 Base Tunnel section type C2 Table 2 C2 section types variabilities

3. MONITORING SYSTEM
Monitoring systems are designed to systematically acquire information on the geological-
geomechanical conditions of the face and the deformation response of the tunnel during excavation
and when it is in service. The design must provide detailed specifications of monitoring stations: (i) the
most appropriate instruments to detect the magnitude of the measurements to be monitored; (ii) how
frequently they are installed as a function of local conditions; (iii) instructions on how to interpret
readings so that they can be compared with design predictions.

3.1 Monitoring stations
The basic types of monitoring stations listed below are those generally used during tunnel
construction. Monitoring stations must be installed very close to the face and the zero reading taken
immediately if they are to provide useful and accurate information.

3.1.1 Geological-geomechanical survey of the face
These surveys give a graphical and numerical description of the geological, geostructural and
geomechanical face conditions during excavation. Two different types of surveys are employed:
4
detailed surveys: detailed descriptions of the structural, lithological and stratigraphic rock
characteristics, with information on the degree and type of compactness, cementation,
granulometry, alteration state, discontinuity characteristics (type, position, geometry, filling, J RC,
J CS), and remarks on sources of water and on possible gravitational detachments during
excavation. Geomechanical classifications are provided: RMR and GSI. Specimens may be taken
for laboratory tests (classification, compression, triaxial, shear joint, extrusion tests, etc.) and in
situ tests may be performed (metric pressure-dilatation, etc.);
snap-shot surveys; these are employed to acquire qualitatively structural, lithological and
stratigraphic rock characteristics only.

3.1.2 Estrusion measurement station
These measurements are used to provide a survey of the longitudinal displacements of measurement
bases set inside an extensimetric rod, located along the tunnel alignment, ahead of the face, inside a
sub-horizontal bore hole. The instruments constitute an incremental extensimeter, consisting of PVC
pipes, fitted with annular anchors at intervals of 1 metre along them and set into the bore hole with
slightly expansive grouting injections. The bore holes are usually greater than 30 m in length (2-3
times the tunnel diameter, with about 10 m of overlap). Once the grout injection has set, a zero
reading must be taken before excavation starts.

3.1.3 Convergence measurement station
These measurements provide a plot of the 3D displacements for 5-7 points set on the preliminary
lining (sometimes on the final lining), in the centre of the crown, in the spring line and the centre plane,
fitted with optical targets and read using a precision topographic instrument. If differential deformation
behaviour is observed between the shotcrete and the steel ribs, targets are also installed on the steel
ribs using special supports. Cavity convergence is considered as being the average of the three
diameter measurements. Multi-point extensimeters may also be used to measure convergence inside
the ground outside the profile of the excavation.

3.1.4 Principle measurement station
These stations are designed to give a complete map of the stress-strain state of the ground and of the
lining and to collect data on hydrological conditions (water pressure and inflow) and the extension of
the plasticised zone (plastic radius). In urban areas, or under shallow overburdens, surface settlement
and ground movements are also monitored. Stations normally consist of: extensimeters, placed in the
crown and at the centre line, to monitor ground deformation around the cavity; piezometers to monitor
water pressure; load cells to monitor bolt and steel rib forces; convergence stations both for the
preliminary lining and the final lining; strain gauges, in the preliminary lining and in the final lining, to
monitor stress in the shotcrete and concrete.

3.2 Monitoring frequencies
The siting of monitoring stations along the tunnel alignment (quantities) must be carefully studied to
achieve the best possible compromise between the need for adequate knowledge and the need to
keep costs low and not to interfere with excavation work and tunnel use when in service. The
predicted behaviour of the core-face is taken into account when deciding the frequency of the intervals
at which stations are installed, as shown in Table 3. In stable core-face conditions (category A), in
which the ground is stressed in the elastic range and the deformation response is immediate and very
small, convergence measurement stations alone are sufficient, with few principle measurement
stations. In stable core-face in the short term conditions, in which the ground is stressed in the
elasto-plastic range and deformation phenomena is mostly deferred and not negligible in magnitude,
extrusion, convergence and principle monitoring stations must be installed very frequently (extrusion
and convergence stations are recommended when tunnel advance is halted for more than seven
days, owing to the possible relaxation of the ground that may occur). Finally, in unstable core-face
conditions, where deformation of the ground develops into the failure range if not properly controlled
by preconfinement action, convergence, extrusion and principle monitoring stations must be installed
extremely frequently. In situations of this type, it is very common to pass from conditions of stability to
conditions of instability very suddenly, as a result of a very small increase in deformation, so the
monitoring data are of fundamental importance.

5
Table 3 Monitoring stations frequencies and readings
A (stable core-face) B (stable core-face s.t.) C (unstable core-face)
Monitoring stations
Rate Reading Rate Reading Rate Reading
Analytical geo survey 1/200m - 1/150m - 1/100m -
Snap-shot survey 1/20-24m - 1/10-12m - 1/10-12m -
Estrusion - 1/24-48m 1/20-24m
Convergence 1/50-60 m 1/24-36m 1/10-12m
Principle
measurement
1/400-500m
1/day up to
10 m face
distance,
1/wk up to
stabilisation
1/200-300m
1/day up to
24 m face
distance,
1/wk up to
final lining
casting (*)
1/100-200m
1/day up to
24 m face
distance,
3/wk up to
final lining
casting (*)
(*) estrusion: 1 reading before escvation, 1/day during excavation, 1 reading before and after ground intervention

Monitoring must be performed very carefully in certain contexts, such as in sections of tunnel affected
by large tectonic dislocation, or under rock burst conditions, or when a strong hydraulic gradient might
be encountered, or for tunnels under the water table, where the flow rate of drainage water and the
water pressure on the final lining must be measured. In shallow tunnels, especially in urban areas,
stations must be installed to measure deformation from the surface that develops at tunnel depth and
to measure settlement at ground level so that effects on surface constructions, such as roads,
buildings and utilities can be kept under control. In these cases the surface constructions themselves
must be monitored using appropriate instruments (surface clinometers, differential settlement gauges,
topographic levels, etc.). The information given in Table 3 gives a general idea of the frequency of the
readings that should be taken by the instrumentation installed depending of course on the predicted
behaviour of the face-core. The first data read will in any case enable the frequency of readings to be
adjusted during construction.
Automatic devices are available for in-service monitoring of tunnels with remote measurement of most
of the parameters needed to monitor the health of an underground construction (e.g. stresses on
linings), so that the problem of maintenance and repair can be approached using scientific methods.

3.3 Interpretation of data
It is fundamental for the measurements to be interpreted properly if monitoring is to be useful in the
fine tuning of the construction process. Interpretation is often not easy for a whole series of reasons,
but mainly because of the reliability of the instruments (device delicacy, wrong positioning or use,
malfunctioning), a broad data scatter and external influences (running water, variations in temperature
and humidity). It is preferable to use less sophisticated equipment, more appropriate to an
underground environment and to install more than the minimum number considered necessary. It is
important to consider all aspects that might influence data acquisition.
Core-face extrusion is normally plotted graphically to give the magnitude of the deformation as a
function of time or face advance; the abscissa of the extrusion curve gives the magnitude of the
extrusion and the angle of the tangent to the curve gives the velocity, while the curve itself describes
its acceleration. The incremental curve, which is for the entire length of the drill hole (i.e. the
extensimeter rod), and the curve for the individual sections into which the extensimeter is divided are
plotted (Figure 5). Curves of the latter type are particularly useful for comparing events and trends
that manifest in different bands of ground within the core-face, and more specifically for estimating the
radius of influence of the face, intended as the distance from the face at which changes in stress
occur caused by tunnel advance. These data are very useful for calculating the length and overlap of
the preconfinement operations. The magnitude of the incremental curve is used to verify the core-face
conditions (category A, B or C) and to calibrate the intensity of the preconfinement operations, while
the acceleration is an indicator of the onset of collapse: it therefore acts as an alarm bell to quickly
place stabilisation structures or to remove personnel and equipment.
Convergence curves can be plotted as a function of time or of distance from the face and magnitude,
velocity and acceleration can be derived from those curves, as in the case of extrusion. If
measurements are taken inside the rock-mass (multi-point extensimeters), radial deformation as a
function of the distance from the excavation profile can be plotted and - if the extensimeter is long
enough - the plastic radius of the tunnel can be derived from them. A comparison of the magnitude of
actual convergence with the design prediction for it can be used to verify pre-confinement and
confinement operations and to calibrate them during excavation, especially the stiffness of the
6
preliminary lining (shotcrete thickness, steel rib intervals, radial bolt intervals and length). Its very
important to monitor the deformation gradient, because the deformation response must present a
certain degree of deceleration: acceleration is normally at the maximum during and immediately after
the passage of the face and during the excavation of the tunnel invert. Convergence must stabilise
rapidly when this is cast. The deformation gradient is used to decide the distance from the face at
which the final lining is cast (invert and crown). If substantial settlement is recorded the use of a steel
rib invert strut is recommended.


Figure 5 Extrusion measurement

Data from strain gauge readings, the creep and the shrinkage components of strain in concrete and
shotcrete, must be considered, to avoid overestimating the stress in the structure [6].

4. DATA MANAGEMENT
Once acquired, data must be filtered and organised by a monitoring manager and made available in
real time to the principle personnel concerned (design engineer, construction manager, construction
supervisor, client) with the construction of a data bank that is constantly updated with the use of GIS
or similar IT systems. The processed data output will consist of type sheets containing all the
information needed for correct interpretation of the measurements (name, position and type of
instrument, date of the reading, zero reading, system and unit of measurement employed, graphs for
viewing data trends, etc.). It is important to be able to correlate data readings with operations in the
tunnel. A daily log of work performed must therefore also be compiled indicating the position of the
face, the preliminary lining, the casting of final linings (tunnel invert and crown) and the tunnel section
type employed. It is recommended that the monitoring manager prepares a periodic report
summarising the activities performed, with indications of any differences from design predictions.
The procedure for managing the construction process in the light of data readings is as follows
(account is taken of the variabilities for the section type provided for in the design and adjusted by
the guidelines). Excavation must commence using the main section type specified for the section of
tunnel in question and with the operations specified by the design, i.e. employing the average values
for the range specified in the guidelines. At the same time monitoring as specified in the monitoring
programme commences and measurements are acquired on the geological-geomechanical context
and the deformation response. Once the geomechanical and deformation conditions are known, they
can be compared with design predictions and the following decisions can be made:

a) confirmation of the tunnel section type specified according to the intensity of the intervention and
sequence of operational stages given in the design (average values);
b) confirmation of the tunnel section type and variation of the intensity of the intervention or the
sequence of the operational stages in order to adapt them to fit the geomechanical conditions
encountered during construction more precisely (section type variabilities);
7
c) selection of a different tunnel section type from those specified in the design, if the geological and
geomechanical conditions encountered diverge from those forecast for that section of tunnel (use
of section types along the tunnel alignment);
d) report the need to design a new tunnel section type with different technical characteristics in cases
where completely unforeseen lithological and geomechanical conditions are encountered or
unexpected deformation behaviour is encountered.

The decisions made in the light of the data acquired and of a comparison with design predictions are
taken by the design engineer, the construction manager or the construction supervisor, depending on
the contents of the contract with which the client awarded the construction work. With the contracts
generally employed in Italy, decisions of the type described in points a) and b) are taken directly by the
construction manager and checked by the construction supervisor. Tunnel excavation is only
occasionally followed continuously by the design engineers with the appointment of a resident
geomechanical engineer, in order to systematically assist with decision-making for tunnel advance
procedures. This occurs under geologically difficult conditions or for major projects. The intervention of
the design engineers is, however, required for the conditions defined in points c) and d), since they
are unexpected conditions not contemplated by design predictions. Finally any increases in
construction times and costs will require the involvement of the client.

5. CASE STUDIES
Some case studies taken from the construction of Apennine road and rail tunnels are examined and
discussed to show how an examination of monitoring data and graph summaries to interpret them can
guide methods of tunnel advance. Figure 6 relates to the section of excavation for the Base road
tunnel shown in Fig. 4 and gives the changes in geomechanical conditions (RMR and GSI) and
deformation behaviour (in terms of convergence and settlement) for a section of tunnel close to the
passage through the Scaly Clays formation and the Arenarie del Cervarola formation. The
geomechanical characteristics of the latter are much better than those for the Scaly Clays (a sharp
increase in readings for GSI from 30-35 to 55-65 can be observed, with a minimum value of
approximately 20 at the point of contact between the two formations).
BASE TUNNEL- FLORENCE SIDE
GSI-RMR vs DEFORMATIONS pk 8+875 - 8+764
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
8750 8760 8770 8780 8790 8800 8810 8820 8830 8840 8850 8860 8870 8880
pk (m)
GSI RMR
sc MPa conv. med.
settlement sx vtr
SCALY CLAY
pk 8+875 - pk 8+808
2 IPN 220 +VTR
8+802 - pk 8+764
2 IPN 220 p.1.2 - 1.5 m,
frompk
8+764
Rock Bolts
CERVAROLA SANDSTONE

Figure 6 Base Tunnel GSI-Deformation response Vs Pre-confinement actions

The application of the C2 tunnel section type was adjusted by varying the intensity of the ground
reinforcement at the face in order to maintain diameter convergence below 10 cm. This was
8
performed by increasing the number of fibre glass structural elements inserted in the face from 60 to
80-90 whenever the deformation response consisted of convergence values in the 15-20 cm range.
Once excavation reached the sandstones, tunnel advance proceeded with cavity confinement action
only, initially with 2IPN220 steel ribs at 1.20 m. intervals and then, when GSI values of greater than 60
were reached, with radial rock bolts [5]. Figure 7 shows changes in the deformation response in
terms of extrusion and convergence as a function of the overburden (i.e. the geostatic stress
conditions) for the Pianoro tunnel on the high speed Bologna-Florence railway line [7] in the
Complesso Caotico formation. Values for cumulative extrusion varying between 15 cm and 5 cm
can be observed for overburdens varying between 100-120 m. and 30 m. In this case too the
number of fibre glass structural elements in the face was varied between 40 and 100-120, in
order to maintain the deformation response within design limits (extrusion/convergence <10 cm).


Figure7 Pianoro Tunnel Deformation response Vs Overburden

5. REFERENCES

[1] P. Lunardi, Design & Costructing tunnels ADECO-RS approach, Tunnels and Tunnelling
International, special supplement, 2000
[2] P. Lunardi, Design and Costruction of Tunnels Analysis of Controlled Deformation in Rock
and Soils (ADECO-RS), Springer, Berlin, 2008
[3] M. Panet, A. Guenot, Analysis of convergence behind the face of the tunnel, Tunnelling 82,
The Institute of Mining and Metallurgy, London, 1982, pp 197-204
[4] P. Lunardi, The influence of the rigidity of the advance core on the safety of tunnel excavation,
Gallerie e grandi opere sotterranee, n. 52, 1997
[5] P. Lunardi, G. Cassani, M. Gatti, Design aspects of the new Apennines crossing on the A1
Milan-Naples motorway: the Base Tunnel, Proceeding of the AFTES Congrs International
Monaco, Le souterrain: espace davenir, 2008
[6] F. Mola, M. Gatti, Instantaneous and long term analyses of cracked r.c. sections subjected to
bending and axial load in the service stage, Studies and Researches, 16, Bergamo, 1995
[7] P. Lunardi, G. Cassani, R. Bindi, From the ADECO-RS approach to the tunnelling
industrialisation, International Symposium on Tunnelling for urban development, Pattaya City,
Thailand, 14th-15th December 2007