2nd International Conference on Tunnel Boring Machines in Difficult Grounds (TBM DiGs Istanbul)

Istanbul, 16–18 November 2016

DETAILED TBM BORING CYCLE ESTIMATION USING ROCK MASS RATING SYSTEM
Nicola Della Valle1, Stefano Fuoco2, Gabriele Brino3
1
Tunnelconsult Engineering SL, Sant Cugat del Vallés (Barcelona), Spain. Email: [email protected]
2
Galleria di Base del Brennero, Brenner Basistunnel BBT SE, Bolzano, Italy, formerly Tunnelconsult
Engineering SL, Sant Cugat del Vallés (Barcelona), Spain. Email: [email protected]
3
Tunnelconsult Engineering SL, Sant Cugat del Vallés (Barcelona), Spain. Email:
[email protected]

ABSTRACT: RMR is one of the most popular rock mass classification systems for rock mechanics
application, because of its simple calculation method, and its widespread diffusion in mining and civil
engineering practices.
Since 1950s, various prediction models have been proposed to estimate Tunnel Boring Machine (TBM)
performance, in particular penetration rate (PR) and utilization factor (UF), as a function of geology. PR and
UF are useful to define machine production, but are not sufficient to describe properly the whole boring
cycle for open, single shield and double shield TBMs in different geological conditions.
This study proposes an innovative approach: the complete machine boring cycle has been analysed for
different typologies of TBMs, and expressed as a function of RMR class. The database includes experimental
values collected during last 10 years in various tunnel projects all over the world, excavated with different
typologies of machines and geological conditions.
The proposed methodology is adopted in production forecast in case of Brenner Base Tunnel project,
construction lot “Mules 2-3”, as part of the analysis performed to select the best TBM, whose main challenge
was taking into account predicted stoppages (ground treatment, investigation, excavation of cross-passages)
in boring cycle to confirm expected TBM production.

KEYWORDS: TBM; Prediction; RMR; Cycle; Performance; Utilization.

1. INTRODUCTION

Full-face mechanized excavation with Tunnel Boring Machines (TBMs) is nowadays one of the most
used tunnelling methods all over the world, thanks to 60 years of quick evolution in various aspect of this
technology, from materials to excavation technics, e.g., the first TBM machine, developed by the American
Charles Wilson and patented in 1856, the adoption of first disc cutters in Oahe Dam project, in South
Dakota, on a Robbins main beam TBM in 1952 (Roby et al., 2008), the invention of Double Shield TBM by
Robbins in 1972, Orichella project, Italy (Guglielmetti, 2007).
One of the main aspects a tunnel designer has to estimate when it decides to adopt this excavation
method is the possible advance rate (AR) of the machine. Estimation of TBM performance involves the
study and the understanding not only of how a TBM can penetrates the rock and which parameters are
involved in excavation process from the scale of the intact rock till the behaviour of the whole rock mass, but
also how the adoption of different typologies of machine can influence the advancement.
The importance of a reliable estimation of AR as a function of geology and TBM parameters grows with
years, with advancement that can vary from over 100 m/day to less than 50 m/month in critical geological
conditions (Barla & Pelizza, 2000), and, consequently, a wide impact on the whole technical and economic
sustainability of an infrastructure project.
 The goal of this study is to have a proper estimation of the AR as a function not only of geological and
machine parameters, but also of TBM type, starting from a better understanding of how the whole industrial
process that allows TBM tunnelling influences the performance.
The rate of penetration, hereafter called ROP, is the net advancement per time unit of the machine, i.e.
length of tunnel excavated over the net excavation time:

The utilization factor, hereafter called UF is a measure of the effective time of excavation of the TBM,
and it is the ratio between net excavation time and total time:

The advance rate (AR) is the ratio between length of the tunnel excavated and total time needed to
advance:
∙ ∙
The estimation of cutters penetration by means of the rate of penetration must be integrated by the study
of utilization factor, analysing losses of time due to adverse geological conditions and aspects related to the
entire excavation process, not only regripping, segment installation, maintenance, but also breakdowns of the
mucking system, stops for ground treatment and monitoring.

2. THEORETICAL BACKGROUND

2.1 Phenomenology of rock fragmentation process

In hard rock TBM tunnelling, rock fragmentation is guaranteed by the cooperation of different
excavation tools (drag picks, disc cutters, ground tools) that attack the surface of the rock, mounted on a
cutter head and pressed against the rock by the machine itself (Hood & Roxborough, 1992).
Two different processes can cause rock breakage, single pass and multiple cutting process. Single pass
cutting process consists in the detachment of rock chips, due to the action of two disc cutters and is the
process that more frequently can occur. The effect of cutters on the rock has been studied through full-size
tests, as the linear cutting test equipment (Rostami and Ozdemir, 1993). When a single disc penetrates rock
surface, it creates a pulverised area of rock below the cutter, the so-called crushed zone, and radial tension
cracks propagation starts from the point (Figure 1).
Another possibility for rock fragmentation, especially in case of rock with very high strength is the
multiple cutting process. The chip is generated by multiple passages of different discs in the same position,
due to the cutterhead design. This solution is used when the load applied to the cutter and/or the time of
application of the force is not sufficient for the detachment of a complete chip (Rostami and Ozdemir, 1993;
Bruland, 1998).

Figure 1: Chip formation and detachment under the effect of a cutter disc (Bruland, 1998).
2.2 Analysis of main performance prediction models in literature
From 1950s different models have been developed to estimate TBM advancement. Models present in
literature can be divided in two groups:
Analytical models: they evolve from theoretical assumptions, based on study of how the cutter penetrates
the rock and force balances;
Empirical models: they are based on the back-analysis of data collected during excavation, with
statistical data treatment of excavation and rock parameters.
Analytical models usually combine intact rock characteristics (mainly uniaxial compressive and tensile
strength) with information about the cutter (e.g. cutter diameter, spacing, tip shape and thrust per cutter), but
they do not take in deep account rock mass characteristics that recent studies (Bruland 1998; Ramezanzadeh,
2006; Gong & Zhao, 2007) proved to be very important to understand TBM performance, as joint frequency
and joint orientation. The most frequently used analytical models had been modified recently to overcome
this limitation and they can be considered semi-empirical. The most used ones are the Colorado School of
Mines (CSM) model, initially developed by Rostami and Ozdemir (1993), starting from the results of linear
cutting tests and mathematical assumption on rock fragmentation process, furtherly improved by
Ramezanzadeh (2006) and Frenzel (2010), and Gehring model (Gehring, 1995), that includes also factors
derived from empirical correlations.
Empirical models are based on rock mass characteristics, and many studies propose correlations with
main rock mass classifications, widespread used in rock mechanics. Data collection programs could involve
both rock face mapping and rock coring program on-field, with significant variations among models
analysed, while TBM performance and machine data datasets in hard rock tunnelling are almost
standardised.
One of the most important empirical models for performance, utilization factor and cutter wear
prediction is Norwegian University of Sciences and Technology (NTNU) model. Last version of NTNU
model is the work proposed by Bruland (1998) in its doctoral thesis.
In literature, there are several models based on rock and machine characteristics. Non-linear multiple
regression analysis is often exploited by the great majority of these studies, while some authors tried to apply
other strategies, as the neuro-fuzzy model, based on data clustering and a back-propagation algorithm: main
examples of this approach have been proposed by Alber (1996), Alvarez Grima et al. (2000), Yagiz (2008),
Gong & Zhao (2009), Hassanpour et al. (2011), and Farrokh et al. (2012).
Another important group of models are based on rock mass classification systems, in particular Rock
Mass Rating (RMR) by Bieniawski (1989), Rock Mass Quality Index (Q and QTBM) by Barton (2000) and
Geological Strength Index (GSI) by Hoek (1995, 2007). Different authors tried to find correlations between
performance and rock mass classification, primarily because they are accepted and known worldwide. Main
examples are the one proposed by Cassinelli et al. (1982), Innaurato et al. (1990), Barton’s QTBM (2000),
Ribacchi & Lembo Fazio (2005), Bieniawski et al.’s RME system (2007a, 2007b, 2008), Hamidi et al.
(2010) and three equations proposed by Hassanpour et al. (2011).
Just Bieniawski et al.’s RME system (2007a, 2007b, 2008) takes into account advance rate variability as
a function of TBM typology.

3. TBM ADVANCE RATE PREDICTION

3.1 Description of developed database

Data for TBM penetration analysis has been obtained from statistical treatment of rough data coming
from several tunnel project in which TunnelConsult (TC) has worked as site supervisor during tunnel
excavation, excavated with open, single shield and double shield TBMs in different intact rocks and rock
mass conditions all over the world.
TC database is composed of 14 projects in Europe, Asia and America, for a total of 143.98 km excavated
with Open TBM, Single Shield (SS) TBM, Double Shield (DS) TBM and Earth Pressure Balance (EPB, in
rock considered as a SS TBM) from 2005 and 2015 in different geological conditions, with a range of
excavation diameter between 3.8 and 15.08 m.

Table 1 summarizes the characteristics of the database.

 Table 1: TunnelConsult database, 14 projects, total length 143.92 km.

Φexc L
Project Description Purpose Country TBM
(m) (km)
Madrid - Barcelona - French border high speed line. High speed France-
DS 9,96 26,70
Perthus tunnel construction site supervision. railway Spain
Emergency
Emergency exit Cadí tunnel construction site supervision. Spain Open 4,22 3,75
exit

Arroyo Maldonado flood relief tunnels, expert consultancy Hydraulic Argentina EPB 7,8 14,40

Fontsanta - Trinitat water supply tunnel Hydraulic Spain DS 6 12,00

Site supervision and follow up of Line 9 of Barcelona
Metro. Stretches III between Zona Universitaria and Metro Spain SS 12,06 9,20
Sagrera Meridiana.
Montcada Tunnel site supervision of Barcelona - French High speed
Spain EPB 11,66 2,90
Border high speed tunnel railway
Site supervision and follow up of Catalan Railway
Metro Spain EPB 6,9 18,60
extension in Terrassa.
Final and detailed design of Tunnel #1 (Aked) of the High High speed
Israel EPB 10,3 10,50
Speed Line between Tel Aviv and Jerusalem railway
Final and detailed design of headrace tunnel of Kargi
Hydroelectric Turkey DS 9,88 11,70
Hydroelectrical Project.
Technical assistance on construction works for the UAA-4
Contract (corridor 2) of Chennai Metro Project, between Metro India EPB 6,68 10,17
Egmore and Nehru Park Stations
Costa
HEP Torito site supervision Hydroelectric EPB 7,98 3,56
Rica
Roadway &
Site supervision of Caltanissetta highway tunnel Italy EPB 15,08 11,64
Highway
Site supervision of headrace tunnel of Pando
Hydroelectric Panama EPB 3,8 5,10
Hydroelectric Project
Site assistances to access tunnels to Savona harbour. Roadway &
Italy SS 13,7 3,71
Baschi, San Paolo and Cappuccini galleries. Highway

The geological parameter considered the most appropriate and reliable to describe the quality of the rock
mass, and the best one that allows estimating not only the performance, but the whole boring cycle as a
function of machine and geology, is RMR system (Bieniawski, 1989).
In fact, the use of this rock mass classification is quite common in infrastructural and mining field and it
is worldwide accepted as a parameter that measures rock mass quality. For each RMR class and each type of
machine, an average value of ROP and UF has been calculated.
In order to improve the reliability of the model, the curve obtained with TC database has been used
together with outcomes from other models and data available in literature estimating ROP and UF as a
function of RMR.
3.2 Average ROP prediction
In case same thrust and torque will be applied at the cutterhead, ROP can be considered a parameter that
is independent from the typology of the machine considered, because it derives from the interaction between
tools and rock mass.
Models considered for ROP estimation are hereby listed:
- TC model;
- Cassinelli et al. (1982) model;
- Sapigni et al. (2002) model;
- Grandori et al. (2011) model.

Table 2 and Figure 2 show the results obtained with the database for TC model in terms of ROP as a
function of RMR.

TunnelConsult model
2,5

1,5
ROP [m/h]

0,5

0
0 10 20 30 40 50 60 70 80 90 100
RMR

Figure 2: TunnelConsult model, ROP as a function of RMR classes.

Table 2: TunnelConsult model, ROP as a function of RMR classes.

Class V Class IV Class III Class II Class I

RMR 0-20 20-40 40-60 60-80 80-100
ROP [m/h] 1,5 2,2 2 1,6 1,6

Maximum ROP is obtained for RMR classes between III and IV (RMR between 30 and 50), with values
of ROP around 2 m/h; on the other hand, both in case of very good rock (classes I-II, above 70) and very bad
rock (class V, RMR lower than 20) the ROP decrease to values around 1.5 m/h.
The equation proposed by Cassinelli et al. (1982) correlates the rate of penetration with Rock Structure
Rating (RSR) system (Wickham, 1974):
0.0059 ∙ 1.59
Wickham (1974) proposed following correlation between RSR and RMR:
0.77 ∙ 12.4
The database from which the equation has been derived is composed of different hydroelectric projects
executed in Italian Alps with open TBM, total excavation length 19 km, diameter less than 4 m. Figure 3
shows the curve obtained from Cassinelli et al. model.
 Cassinelli et al. (1982) model
1,6

1,4

1,2

1,0

ROP [m/h]
0,8

0,6

0,4

0,2

0,0
0 10 20 30 40 50 60 70 80 90 100
RMR

Figure 3: Cassinelli et al. (1982) model, ROP as a function of RMR.

Sapigni et al. (2002) studied the rate of penetration as a function of the rock mass rating, starting from a
database of data collected during the excavation of three hydroelectric tunnels in Italy (Varzo, Maen and
Pieve projects) in metamorphic geology with Open and DS TBMs, range of diameter 4.05-4.2m and discs
with diameter 17”, obtaining the plot shown in Figure 4.

Figure 4: Plot by Sapigni et al. (2002), ROP as a function of RMR.

Grandori et al. (2011) analysed data from first 10 km of the Brenner Base Tunnel, excavated in granite
and granodiorite by a DS TBM with 6.3 m of diameter, obtaining the results shown in Figure 5.

Figure 5: Plot by Grandori et al. (2011), ROP as a function of RMR.

 Table 3 and Figure 6 summarize the estimation of ROP as a function of RMR, i.e. the average value
between the results of TC model and other equations from literature for each RMR class.

ROP as a function of RMR

3,5

2,5

ROP [m/h] 1,5

0,5

0
0 10 20 30 40 50 60 70 80 90 100
RMR
TunnelConsult Cassinelli et al. (1982) Sapigni et al. (2002)
Grandori et al. (2011) Average

Figure 6: Average ROP as a function of RMR.

Table 3: Average ROP as a function of RMR

ROP
Cassinelli et Sapigni et al. Grandori et al.
RMR TunnelConsult Average
al. (1982) (2002) (2011)
[-] [m/h] [m/h] [m/h] [m/h] [m/h]
10 1,5 1,5 1 0,89 1,29
20 1,85 1,4 1,4 1,12 1,54
30 2,2 1,4 1,8 1,34 1,79
40 2,1 1,3 2,1 1,76 1,97
50 2 1,3 2,3 2,18 2,12
60 1,8 1,2 2,3 2,55 2,19
70 1,6 1,2 2,2 2,93 2,23
80 1,6 1,2 1,8 2,51 1,97
90 1,6 1,1 1,4 2,08 1,72

3.3 UF base value prediction

Utilization factor is strongly dependent from TBM type and from its boring cycle: average UF value as a
function of geology is here estimated, and it has to be considered as an upper limit to be verified after TBM
boring cycle analysis.
Models considered for UF estimation are hereby listed:
- TC model;
- Dolcini et al. (1996) model;
- Sapigni et al. (2002) model;
- Innaurato et al. (1990) model.

Table 4 and Figure 7 show the outcomes of TC model for the estimation UF as a function of RMR and
TBM.
 TunnelConsult model
60%

50%

40%

UF [%]
30%

20%

10%

0%
Class I-II Class III Class IV Class V
RMR

Open TBM SS TBM DS TBM

Figure 7: TunnelConsult model, UF as a function of RMR classes and all TBM type.

Table 4: TunnelConsult model, UF as a function of RMR classes and all TBM type.

RMR Class
Machine Class I-II Class III Class IV Class V
Open TBM 48% 35% 18% 4%
Shield TBM 41% 35% 23% 15%
DS TBM 57% 53% 23% 15%

Maximum UF is obtained for high RMR classes I-II, with values of UF around 55% for DS TBM, 45%
for Open TBM and 40% for SS TBM. UF decreases with the classes, especially in case of open TBMs; DS
and SS TBMs have quite close results in bad rocks (class V, around 15%).
Dolcini et al. (1996) correlate the utilization factor with RMR in case of Open and DS TBM. The
database is composed of 29.4 km of 4 tunnels executed in Greece; Figure 8 shows the curve obtained.

Dolcini et al. (1996)

50%

45%

40%

35%

30%
UF [%]

25%

20%

15%

10%

5%

0%
Class I-II Class III Class IV Class V
RMR

Open TBM DS TBM

Figure 8: Plot proposed by Dolcini et al. (1996), UF as a function of RMR.

Sapigni et al. (2002) studied also the utilization factor as a function of the rock mass rating, just for DS
TBMs, obtaining the plot shown in Figure 9.
 Sapigni et al. (2002)
40%

35%

30%

25%

UF [%]
20%

15%

10%

5%

0%
Class I-II Class III Class IV Class V
RMR

DS TBM

Figure 9: Plot proposed by Sapigni et al. (2002), UF as a function of RMR.

The equation proposed by Innaurato et al. (1991) derives from an actualization of the database used by
Cassinelli et al. (1982):
4 ∙ 10 5.7 ∙ 10 1.4 ∙ 10 0.18
The database from which the equation has been derived is composed of different hydroelectric projects
executed in Italian Alps with open TBM, total excavation length 19 km, diameter less than 4 m. Figure 10
shows the curve obtained by Innaurato et al. (1991).

Innaurato et al (1991) model

70%

60%

50%

40%
UF [%]

30%

20%

10%

0%
Class I-II Class III Class IV Class V
RMR

Open TBM

Figure 10: Innaurato et al. (1991) model, UF as a function of RMR class.

Table 5 and Figure 11 summarize the estimation of UF base value as a function of RMR and TBM type,
i.e. the average value between the results of TC model and other equations from literature for each RMR
class and typology of machine.

UF base value
60%

50%

40%
UF [%]

30%

20%

10%

0%
Class I-II Class III Class IV Class V
RMR

Open TBM SS TBM DS TBM

Figure 11: UF base value as a function of RMR class and all TBM type.
 Table 5: UF as a function of RMR classes and all TBM type.

RMR Class
Machine Class I-II Class III Class IV Class V
Open TBM 50% 37% 21% 11%
Shield TBM 41% 35% 23% 15%
DS TBM 46% 42% 29% 15%

4. BORING CYCLE ESTIMATION

If the great impact of geological conditions and machine parameters on the advancement is a well-known
aspect, at the same time a deeper understanding of TBM cycle is necessary to understand what could be the
production of a TBM in different type of rock and rock mass, especially if theoretical upper limit of
production can be limited by some fixed losses of time already predicted at design stage, because of
vacations, interferences with existing underground structures (e.g., cross passages, underground chambers),
probing, ground conditioning.
Furthermore, it’s fundamental to correlate the behaviour of the machine and the development of the
whole industrial process involved in mechanized tunnelling to a simple and worldwide adopted parameter
that can resume all geological parameters that affects the excavation process, individuated in RMR system.
Each activity in TBM boring cycle should be estimated in order to understand what the implication of
predicted time losses is in the specific project: after general methodology for estimation of time spent in
every activity for different TBMs, Brenner Basis Tunnel tender case study will be analysed in details.
Design method is shown in Figure 12 . The innovation of proposed AR prediction methodology is the
fact it is based on a proper estimation of the whole excavation process: therefore, it can include downtime
factors connected with ground treatment, probing and interaction with other excavation works in a more
clear and systematic way, limiting average UF factor if impossible to be reached in a specific analysed
project.

Figure 12: Proposed design method for AR estimation

4.1 Analysis and estimation of average TBM cycle for UF correction
TBM cycle should be analysed decomposing each activity and sub-activity, and assigning an estimation
of the time needed to complete that action, based on daily excavation time, in general supposed equal to 24
hours a day.

Excavation time
As said in previous chapters, excavation is strongly dependent from both geological conditions and TBM
type, but it could be also limited by imposed time losses at the design stage.
UF prediction model introduced in Paragraph 0 can be considered an upper limit for UF estimation, from
which tunnel engineer can start to have a rough estimation of available excavation time, but it could be
updated at the end of the average TBM cycle estimation if other activities reduce possible utilization of the
machine.

Regripping time
Regripping is the action that allows open TBM to advance: a series of thrust jacks, called grippers,
anchors the machine to the rock, counterbalancing the thrust force generated by the machine and permitting
to push the TBM on. For DS TBMs that work in “DS mode”, i.e. in case of high rock quality, a moving part
of the shield, called telescopic shield, have the same function of bearing pads for open machines.
Both Open TBM (all RMR classes) and DS TBM (if RMR class is I, II or III, where DS will work in
“DS mode”) spend approximately 5 minutes for each boring cycle, approximately equal to the length of the
ring foreseen for shielded TBMs; the same value can be imposed for open TBM as maximum possible stroke
before regripping, around 1.5-2 m. Bruland (1998)n proposed a time per regrip around 4-5 min/stroke, while
Farrokh (2013) defined a regripping time enter 2 and 6 min/stroke for both open and DS TBMs, adding a
factor for radius of curvature R too (tcurv/Ladv=409000/R2).

Support installation time

In case of open TBM, boring cycle includes the installation of temporary supports, if needed; main
possibilities are rock bolts, steel arches and wire mesh in the area immediately behind the cutter head (4-6
m), shotcrete installation.
An example of installation time for different support is shown in Figure 11, elaborated by Farrokh et al.
(2011) analysing the data coming from four different tunnel excavated with open TBMs in Italy, total
excavation length 19.79 km (Farrokh, 2013).

Figure 13: Open TBM support installation time for different support types (Farrokh, 2011)

Figure 12 shows another interesting correlation with Austrian O-Norm F classes of excavation in rock,
resumed and compared with RMR classes in Table 6, with a database of data collected in 8 projects (34.16
km).
Table 6: Austrian Önorm F Classes and connection with RMR system (adapted from Farrokh et al., 2011 and
Palmström, 1993)
Rock RMR RMR Support Type Quantity/m
Description Support
Class class rating (Φ=5.8m) (Φ =5.8m)
F1 Stable I >80 Local Rock bolts (L=2m) 0.5
Rock bolts (L=2m) 1
Slightly
F2 II 60-80 Local Wire mesh AQ 50 1 m2
loosening
Shotcrete 5cm 0.1 m3
Rock bolts (L=2m) 1-3
F3 Ravelling III 40-60 Systematic Wire mesh AQ 50 1-1.5 m2
Shotcrete 5cm 0.1-0.5 m3
Rock bolts (L=2.5m) 3-5
Moderately Wire mesh AQ 50 5-9 m2
F4 IV 30-50 Systematic
squeezing Shotcrete 8cm 0.5-1 m3
Steel rib UNP 120 40-80 kg
Rock bolts (L=2.5m) 5-7
Plastic
Wire mesh AQ 50 9-18 m2
F5 squeezing/ IV 30-50 Systematic
Shotcrete 10cm 1-1.8 m3
swelling
Steel rib UNP 120 80-160 kg
Rock bolts (L=3m) 7-10
Highly
Wire mesh AQ 50 18-27 m2
F6 squeezing/ V 10-20 Systematic
Shotcrete 15cm 1.8-3 m3
Swelling
Steel rib UNP 120 160-300 kg
F7 Running V 10-20 Systematic Special countermeasures

Figure 14: Open TBM support installation time for different Austrian rock masses F classes (Farrokh,
2011)

However, installation time can be adapted to the needs and the recommendation of the tender, as well as
derived from experiences and literature about projects in similar rocks, e.g., based on the installation time of
rock bolts in a certain type of rock.

Segmental lining installation time

Immediately behind the cutter head, under the protection of the shield, SS and DS TBM install directly
the final lining, a ring composed of precast segments, thanks to an erector. Once arrived at the end of the
strokes, longitudinal thrust jacks are retracted in order to assembly the ring and consequently pushed against
the ring recently installed in order to contrast the thrust for a new stroke.
The installation time needed can be estimated between 30 and 45 min/ring (as suggested by Wittke et al.,
2006, and Maedl et al., 2014) in case of a SS TBM, independently from geological conditions, as well as DS
TBM in “SS mode”, RMR classes IV-V. If in classes from III to I DS TBM should install the ring without
any stop, because it advances in “open mode”, a minimum extra-time of 5 minutes per stroke has been
considered, because of unexpected criticalities during ring installation that can cause temporary stops of the
machine.

Ordinary maintenance and cutter change downtime

Ordinary maintenance can be estimated in 4 hours/day; in case of gripper and SS TBM, if the quality
of the rock is good (RMR class I-II), a lower value can be imposed, 3 hours/day. Expected time for this
activity is estimated by Bruland (1998) and Farrokh (2013) between 50 and 300 hr/km. Cutter substitution
downtime is strongly dependent from quartz content (Figure 15), and it is considered just if estimated higher
than ordinary maintenance time, because this two activities can be considered overlapped. If the quartz
content is really high, cutter consumption can be one of the activities with higher downtime ( , gneiss with
high quartzite content).

Figure 15: Cutter downtime (Farrokh, 2013)

Figure 16: Excavation in gneiss with high quartzite content (Wittke et al., 2006)

Utility extension time

 Utility extension is partially included in ordinary maintenance time because of possibility to overlap: an
indicative time loss is 5 min/stroke in case of gripper TBM and 2 min/stroke in case of DS TBM in “DS
mode”, i.e. in good rock, while for SS TBM and DS TBM in “SS mode” this activity can be considered
entirely executed during maintenance and stops for ring installation. Expected time for this activity is
estimated by Farrokh (2013) between 10 and 100 hr/km.

Mucking system breakdown

In case of muck conveyors, Dolcini et al. (1996) registered an average time loss between 1% and 2%: in
worst cases, this item can reach 350 hr/km (Farrokh, 2013).

TBM breakdown/repair
The estimation of expected time losses caused by TBM breakdown derives from the analysis of literature
data:
- Open TBM: TBM breakdown time derives from data analysis of excavation parameters collected
during Meraaker Project (Figure 17), RMR classes I-II, time losses equal to 5.4%, whereas for low
quality of the rock, i.e. RMR classes IV-V, TBM repair time is around 15% (approximate average
value from Figure 18). For class III, in absence of experimental data, a value of 7.5% has been
assumed, the weighted interpolation between previous data, closer to class I-II.

Figure 17: Boring cycle for open TBM in good rock, Meraaker Project, L=10 km (Bruland, 1998)

Figure 18: Boring cycle for open TBMs in sandstone and marl, low quality of the rock (Wittke et al.,
2006).
- SS TBM: in this case, predicted time loss derives from the excavation of two tunnels in difficult
ground conditions (Figure 19), RMR classes IV-V. TBM breakdown has been considered the
average value, i.e. 14%. In case of better quality of the rock (classes I-II and III), in absence of
experimental data, a value of 11% has been assumed, because slightly better than the one registered
for DS TBM in class I-II (McFeat-Smith & Concilia, 2000).

NFM Lot 1 NFM Lot 5
16% Excavation 17% Excavation
26% 27%
Ring assembly Ring assembly
1%
7%
Maintenance Maintenance
14%
7%
TBM breakdown TBM breakdown
(geology) (geology)
TBM breakdown TBM breakdown
(utilities) (utilities)
24% 23%
20% Other downtimes 17% Other downtimes

Figure 19: Boring cycle for SS TBMs in difficult ground conditions in France (Fontanille, 2009)

- DS TBM: DS TBM breakdown time in case of RMR classes IV-V (average value between two
values in Figure 20, downtime 15%) and classes I-II (Figure 21, downtime 9.8%) has been obtained
from excavation data published by McFeat-Smith & Concilia (2000). In absence of experimental
data, the value for RMR class III can be considered equal to the one for classes IV-V, i.e. 15%.

Poor quality rock Fault zone
11% 12%
14%
Excavation 25,1% Excavation
0,8%
Support Support
works/treatment works/treatment
Maintenance Maintenance

TBM breakdown TBM breakdown
5%

Utilities Utilities
25,8%

Other downtimes 30,8% Other downtimes

16,3% 32,1% 5,4% 21,7%

Figure 20: Boring cycle for two DS TBMs in difficult ground conditions in Hong Kong (McFeat Smith
& Concilia, 2000)
 Good quality rock
5,6%
3,2% Excavation
44%
9,8%
Support
works/treatment
Maintenance

TBM breakdown

27,3% Utilities

Other downtimes

10,1%

Figure 21: Boring cycle for a DS TBM in good rock conditions in Hong Kong (McFeat Smith &
Concilia, 2000)

Probing, surveys, improving treatment, time losses due to interference with other underground works (e.g.,
cross passages)
All the activities strongly dependent from intrinsic geological, economical and organizational conditions
of a specific infrastructure should be studied and estimated in function of the “handmade shaped” solutions
the designer can adopt, taking into account tender specifications too.
This voice of the TBM boring cycle can limit the theoretical utilization factor to a lower value, starting a
new iteration in order to define UF.

Other downtimes
This activity should complete the whole cycle and it cannot be clearly defined in a general way, but it
integrates the results of the other activities in TBM boring cycle prediction.

At the end of this iterative process, predicted UF value can be obtained , as well as updated advance rate
derived from a systematic analysis of TBM boring cycle.

4.2 Case study: TBM selection for Brenner Basis Tunnel - contract package Mules 2-3 tender
The TBM performance prediction model described was developed as the core item in TBM selection
analysis for the Brenner Basis Tunnel (BBT) Mules 2-3 lot tender.

4.2.1 Description of the project and geological context

Brenner Basis Tunnel is a 55-km-long high speed railway project that will connect Italy and Austria as
part of the Scandinavian-Mediterranean Corridor in the framework of the Trans-European Transport
Network (TEN-T): it will be the world’s longest tunnel, when completed in 2026.
Mules 2-3 package is the main lot of the project, located at the Italian section of the project; it extends
from Fortezza portal to the Italian border with Austria. It includes the completion of the exploration tunnel
and the two main line tunnels, for a total of 23 km to be excavated in conventional methods and 46 km of
mechanized excavation with TBM (Figure 22).
 Figure 22: Brenner Basis Tunnel, excavation methods

The portion of exploration tunnel to be excavated with TBM has a total length of 13.927 km in north
direction, anticipating the excavation of north twin tubes main line (12.104 km). Geology crossed is
composed of granitic gneisses, marbles, schists, phillytes, calceschists anphibolites and gneisses (Figure 23),
with majority of the rock mass classified with RMR class III (

Table 7).

Figure 23: Brenner Basis Tunnel geological profile, North portion (granitic gneiss in purple, marbles in
light blue, schists and calceschists in blue with lines, phyllites in yellow, anphibolites in green, gneisses in
brown) from tender documents (BBT, 2014)

Table 7: RMR class distribution along the alignment

Exploration Tunnel North Main Tunnel (x2) South Main Tunnel (x2)
RMR Class L [m] p [%] L [m] p [%] L [m] p [%]
Class I 400.7 2.88% 400.7 3.31% 602 17,18%
Class II 5461.9 39.22% 5069.2 41.88% 1973 56,30%
Class III 7923.3 56.89% 6497.8 53.68% 915 26,12%
Class IV 141.2 1.01% 136.3 1.13% 14 0,41%
Class V 0.0 0.00% 0.0 0.00% 0 0,00%
12104 3504
Total [m] 13927
(x2) (x2 with the same TBM)

South part of the main tunnel will be excavated in the portion of the alignment where the exploration
tunnel has been already completed, for a total length of 3.504 km per tube in the Bressanone granite (Figure
24). The great majority of the rock mass can be classified as good quality (RMR class II,
 Table 7).

Figure 24: Brenner Basis Tunnel geological profile, South portion (granitic gneiss in pink) rom tender
documents (BBT, 2014)

4.2.2 AR prediction based on detailed TBM boring cycle

TBM selection is one of the main technical challenges: if in South direction, where the exploration
tunnel already exists, an open TBM can be an option, because of the good quality of the rock mass, but also
SS and DS TBMs can be used, in North direction the choice is between single and double shield TBM,
knowing that real geological-geomechanical conditions at tunnel level are not known with acceptable
reliability until the exploration tunnel won’t be excavated.
AR prediction model here proposed has been developed at design stage in order to define the production
of different TBM typologies, taking also into account tender restrictions in terms of predicted days of
machine stops in case of:
- Vacations (25 days/yr);
- Extraordinary maintenance and predicted probing ahead of the face (48 days/yr)
- Designed rock improvement (41 days/yr for exploratory tunnel, 14 days/yr for North main tunnel, 12
days/yr for South main tunnel)
- Excavation of cross-passages (60 days/yr for exploratory tunnel excavation, 15 days/yr for North and
South main tunnels).
TBM cycle has been analysed decomposing each activity and sub-activity as described in previous
paragraphs, including time losses fixed by tender documents and adapted to BBT tunnels.
Downtimes connected to ground treatment for SS TBM has been considered higher than the one for
gripper and DS TBM, because of technical obstacles caused by the shield; total time loss for ground
treatment is always imposed as reported in tender documents by reducing this time losses for other TBMs.
The monitoring plan fixes the predicted number of probe drillings, so time loss in the execution has been
calculated with theoretical drilling advance rate equal to 15 m/h for South main tunnel (from data collected
in BBT Periadriatic Line Lot) and 22 m/h for both tunnels in North direction (from maximum value of data
collected in BBT Periadriatic Line Lot, because intact rock abrasiveness is lower than the one of the
Bressanone granite).
Figures and tables below show how TBM cycle varies as a function of RMR and TBM type (south main
tunnel in Figure 25 and

Table 8; north exploratory tunnel in Figure 26 and

Table 9; north main tunnel in Figure 27 and

Table 10): cells in yellow are values fixed by tender documents.

For exploratory tunnel, time losses deriving from rock improvement and cross-passages force the
theoretical UF to decrease from the one calculated with curves proposed in Figure 11, and a new set of
values of UF as a function of RMR class obtained.
 TBM boring cycle estimation - South main tunnel TBM boring cycle estimation - South main tunnel
Open TBM SS TBM
15 15

13 13
Class I-II Class I-II
11 11
Class III Class III
9 9

Time [h/gg]

Time [h/gg]
Class IV-V Class IV-V
7 7

5 5

3 3

1 1

-1 -1
Mucking Surveys Rock Excavation Mucking Surveys Rock Excavation
Support Maintenan Utility Breakdown Other Installazion Maintenanc Utility Breakdown Other
Excavation Regripping system and improveme of cross- Excavation system and improveme of cross-
installation ce extension TBM downtimes e dei conci e extension TBM downtimes
breakdown probing nt passages breakdown probing nt passages
Class I-II 12,0 1,2 2,3 3,0 1,2 0,2 1,3 0,5 0,8 1,1 0,5 Class I-II 9,8 5,5 3,0 0,0 0,2 2,9 0,4 0,8 1,1 0,3
Class III 8,8 1,1 2,5 4,0 1,1 0,2 1,7 0,5 0,8 1,1 2,1 Class III 8,3 6,0 4,0 0,0 0,2 2,9 0,4 0,8 1,1 0,3
Class IV-V 5,0 0,6 6,1 4,0 0,6 0,2 3,5 0,5 0,8 1,1 1,6 Class IV-V 5,4 4,0 4,0 0,0 0,2 3,6 0,4 2,4 1,1 2,9

TBM boring cycle estimation - South main tunnel

DS TBM
15

13
Class I-II
11
Class III
Time [h/gg] 9
Class IV-V
7

-1
Mucking Rock Excavation
Support Maintenanc Utility Breakdown Surveys Other
Excavation Regripping system improveme of cross-
installation e extension TBM and probing downtimes
breakdown nt passages
Class I-II 11,1 1,1 1,1 4,0 0,4 0,2 2,4 0,5 0,8 1,1 1,3
Class III 10,1 1,3 1,3 4,0 0,5 0,2 4,2 0,5 0,8 1,1 0,0
Class IV-V 6,9 0,0 5,0 4,0 0,0 0,2 3,6 0,5 0,8 1,1 1,9

Figure 25: TBM boring cycle estimation, south main tunnel

Table 8: TBM boring cycle estimation for south main tunnel

Open TBM SS TBM DS TBM
IV-
Activity Unit I-II III IV-V I-II III I-II III IV-V
V
Excavation h/gg 12,0 8,8 5,0 9,8 8,3 5,4 11,1 10,1 6,9
Regripping h/gg 1,2 1,1 0,6 - - - 1,1 1,3 0,0
Support installation h/gg 2,3 2,5 6,1 - - - - - -
Ring assembly h/gg - - - 5,5 6,0 4,0 1,1 1,3 5,0
Maintenance h/gg 3,0 4,0 4,0 3,0 4,0 4,0 4,0 4,0 4,0
Utility extension h/gg 1,2 1,1 0,6 0,0 0,0 0,0 0,4 0,5 0,0
Mucking system
h/gg 0,2 0,2 0,2 0,2 0,2 0,2 0,2 0,2 0,2
breakdown
Breakdown TBM h/gg 1,3 1,7 3,5 2,9 2,9 3,6 2,4 4,2 3,6
Surveys and probing h/gg 0,5 0,5 0,5 0,4 0,4 0,4 0,5 0,5 0,5
Rock improvement h/gg 0,8 0,8 0,8 0,8 0,8 2,4 0,8 0,8 0,8
Excavation of cross-
h/gg 1,1 1,1 1,1 1,1 1,1 1,1 1,1 1,1 1,1
passages
Other downtimes h/gg 0,5 2,1 1,6 0,3 0,3 2,9 1,3 0,0 1,9
UF % 55 41 23 64 60 39 55 53 50
19, 18,
AR m/gg 20,6 10,6 16,9 11,5 19,2 22,5 14,7
7 5
Ring/strokes per day n/gg 14 13 7 11 12 8 13 15 10
 TBM boring cycle estimation - North exploratory tunnel TBM boring cycle estimation - North exploratory tunnel
SS TBM DS TBM
15 15

13 13
Class I-II Class I-II
11 11
Class III Class III
9 9

Time [h/gg]
Time [h/gg]

Class IV-V Class IV-V

7 7

5 5

3 3

1 1

-1 -1
Mucking Rock Excavation Mucking Surveys Rock Excavation
Installazion Maintenanc Utility Breakdown Surveys Other Support Maintenan Utility Breakdow Other
Excavation system improveme of cross- Excavation Regripping system and improvem of cross-
e dei conci e extension TBM and probing downtimes installation ce extension n TBM downtimes
breakdown nt passages breakdown probing ent passages
Class I-II 6,2 3,6 3,0 0,2 0,2 2,7 0,6 2,9 4,2 0,3 Class I-II 7,4 0,7 0,7 4,0 0,3 0,2 2,7 0,7 2,9 4,2 0,3
Class III 7,0 0,9 0,9 4,0 0,3 0,2 2,7 0,7 2,9 4,2 0,3
Class III 5,1 3,8 4,0 0,3 0,2 2,7 0,6 2,9 4,2 0,3
Class IV-V 4,6 0,0 3,3 4,0 0,2 0,2 3,6 0,7 2,9 4,2 0,3
Class IV-V 4,6 3,3 4,0 0,2 0,2 3,7 0,6 2,9 4,2 0,3

Figure 26: TBM boring cycle estimation, north exploratory tunnel

Table 9: TBM boring cycle estimation for north exploratory tunnel

SS TBM DS TBM
Activity Unit I-II III IV-V I-II III IV-V
Excavation h/gg 6,2 5,1 4,6 7,4 7,0 4,6
Regripping h/gg - - - 0,7 0,9 0,0
Support installation h/gg - - - - - -
Ring assembly h/gg 3,6 3,8 3,3 0,7 0,9 3,3
Maintenance h/gg 3,0 4,0 4,0 4,0 4,0 4,0
Utility extension h/gg 0,2 0,3 0,2 0,3 0,3 0,2
Mucking system breakdown h/gg 0,2 0,2 0,2 0,2 0,2 0,2
Breakdown TBM h/gg 2,7 2,7 3,7 2,7 2,7 3,6
Surveys and probing h/gg 0,6 0,6 0,6 0,7 0,7 0,7
Rock improvement h/gg 2,9 2,9 2,9 2,9 2,9 2,9
Excavation of cross-passages h/gg 4,2 4,2 4,2 4,2 4,2 4,2
Other downtimes h/gg 0,3 0,3 0,3 0,3 0,3 0,3
UF % 26% 21% 19% 31% 29% 19%
AR m/gg 10,8 11,3 9,8 12,7 15,6 9,8
Ring/strokes per day n/gg 7 8 7 8 10 7
 TBM boring cycle estimation - North main tunnel TBM boring cycle estimation - North main tunnel
SS TBM DS TBM
15 15

13 13
Class I-II Class I-II
11 11
Class III Class III
9 9

Time [hh/gg]
Time [h/gg]

Class IV-V Class IV-V

7 7

5 5

3 3

1 1

-1 -1
Mucking Rock Excavation Mucking Surveys Rock Excavation
Installazione Utility Breakdown Surveys and Other Support Maintenanc Utility Breakdown Other
Excavation Maintenance system improvemen of cross- Excavation Regripping system and improveme of cross-
dei conci extension TBM probing downtimes installation e extension TBM downtimes
breakdown t passages breakdown probing nt passages
Class I-II 9,8 5,5 3,0 0,0 0,2 2,9 0,4 0,9 1,1 0,3 Class I-II 11,1 1,1 1,1 4,0 0,4 0,2 3,0 0,3 1,0 1,1 0,8
Class III 8,3 6,0 4,0 0,0 0,2 2,9 0,4 0,9 1,1 0,3 Class III 10,1 1,3 1,3 4,0 0,5 0,2 3,0 0,3 1,0 1,1 1,4
Class IV-V 5,4 4,0 4,0 0,0 0,2 5,6 0,4 3,0 1,1 0,4 Class IV-V 6,9 0,0 4,5 4,0 0,0 0,2 5,6 0,3 1,0 1,1 0,4

Figure 27: TBM boring cycle estimation, north main tunnel

Table 10: TBM boring cycle estimation for north main tunnel
SS TBM DS TBM
Activity Unit I-II III IV-V I-II III IV-V
Excavation h/gg 9,8 8,3 5,4 11,1 10,1 6,9
Regripping h/gg - - - 1,1 1,3 0,0
Support installation h/gg - - - - - -
Ring assembly h/gg 5,5 6,0 4,0 1,1 1,3 4,5
Maintenance h/gg 3,0 4,0 4,0 4,0 4,0 4,0
Utility extension h/gg 0,0 0,0 0,0 0,4 0,5 0,0
Mucking system breakdown h/gg 0,2 0,2 0,2 0,2 0,2 0,2
Breakdown TBM h/gg 2,9 2,9 5,6 3,0 3,0 5,6
Surveys and probing h/gg 0,4 0,4 0,4 0,3 0,3 0,3
Rock improvement h/gg 0,9 0,9 3,0 1,0 1,0 1,0
Excavation of cross-passages h/gg 1,1 1,1 1,1 1,1 1,1 1,1
Other downtimes h/gg 0,3 0,3 0,4 0,8 1,4 0,4
UF % 64% 60% 39% 55% 53% 48%
AR m/gg 16,9 18,5 11,5 19,2 22,5 14,7
Ring/strokes per day n/gg 11 12 8 13 15 9

4.2.3 Outcomes
The adoption of described methodology of AR prediction for BBT project allows considering main
organizational criticalities that can occur in full-face mechanized tunnelling in hard rock and predicting the
expected production of the machines (machine dismantling, transportation and assembling of each TBM, as
listed in

Table 11).

Table 11: TBM assembling, transportation and dismantling time estimation

TBM
Open [M] S [M] DS [M]
Dismantling 2 2,2 2,5
Transportation 1 1 1
Assembling 2 2,8 3
Total 5 6 6,5
Outcomes of the analysis are hereby resumed:

- South main tunnel: open TBM is the best choice in terms of production, with total excavation time of
27.2 months to excavate both tubes, compared with 29.1 months needed with DS TBM and 30.8
months with SS TBM (Figure 28);
;

Figure 28: Space-time diagram, south main tunnel

-  
-  
- North exploratory tunnel: DS TBM is the best choice in terms of production, with total excavation
time of 50.3 months, compared with 61.8 months needed with SS TBM (Figure 29);
-  
-  

Figure 29: Space-time diagram, north exploratory tunnel

-  
-  
- North main tunnel: DS TBM is the best choice in terms of production, with total excavation time of
32.5 months, compared with 36.8 months needed with SS TBM (Figure 30).
-  
-  
 Figure 30: Space-time diagram, north main tunnel

This analysis is just one of the aspects considered in TBM selection: the decision has been justified by
using a multicriteria analysis, together with different risks associated to each TBM typology: geological
risks, H&S in working environment, costs, consumption of material, quality of the final product, equipment.
The development of the whole analysis brought the consortium to select 4 Double Shield TBMs for BBT
Mules 2-3 construction lot.

5. CONCLUSIONS

Widespread use of full-face mechanized tunnelling with TBM as excavation method in hard rock all over
the world leads to the need of a tool that permits designer to have a proper estimation of production at design
stage as a function of geology and TBM type.
The method described is based on the definition of the net advance rate (AR) as the product of the rate of
penetration of the machine (ROP) and the utilization factor (UF).
ROP is an estimation of how the tools penetrate the rock when TBM is active, and it is function of RMR
class, but it has been considered independent from TBM type in case of same thrust and torque.
UF is an estimation of what is the real excavation time of TBM, i.e. percentage of time when the machine
is actually excavating, and it depends from geology (RMR class), from TBM type (open, SS or DS TBM),
and also from TBM boring cycle.
Indeed, the industrial process of TBM tunnelling includes other variables related with organisational
aspects, involving the investigation and the improvement of mechanical characteristics of the rock mass
(e.g., execution of probe drillings, rock improvement) and site management (interaction with the excavation
of other underground works), that should be considered in performance prediction.
Proposed method integrates data collected by TunnelConsult in 10 years of experience in tunnelling with
literature data, obtaining a more reliable and innovative tool for the prediction of advance rate, which can be
systematically adapted to characteristics and criticalities of a specific project.
Furthermore it can be considered a proper methodology to be used in a wider analysis about TBM
selection, as illustrated for BBT Mules 2-3 construction lot case study.
Future development of the model can include cutter wear prediction and correlation with other
geometrical characteristics of the machine, as the diameter, in order to take into account scale effects during
excavation.

6. REFERENCES

Alber, M. (1996). Prediction of penetration and utilization for hard rock TBMs. Proceedings ISRM
International Symposium Eurock ’96, Balkema, Turin, Italy, pp. 721-725.
 Alvarez Grima, M., Bruines, P.A., & Verhoeh, P.N.W. (2000). Modeling Tunnel Boring Machine
Performance by Neuro-Fuzzy Methods. Tunnelling and Underground Space Technology, Vol. 15, Nr. 3, pp.
259-269.

Barla, G., & Pelizza, S. (2000). TBM Tunnelling in Difficult Ground Conditions. GeoEngineering 2000,
International Conference on Geotechnical and Geological Engineering. Melbourne, Australia.

Barton, N. (2000). TBM Tunnelling in Jointed and Faulted Rock. Balkema, Brookfield, Rotterdam.

BBT (2014). Lotto Mules 2-3, General Works, Longitudinal Profile – Geomechanical Profile.

Bieniawski, Z.T. (1989). Engineering Rock Mass Classifications. Wiley, New York, U.S.A.

Bieniawski, Z.T., Celada, B., & Galera, J.M. (2007a). Predicting TBM excavability – part I. Tunnels &
Tunnelling International, September 2007, pp. 32–35.

Bieniawski, Z.T., Grandori, R. (2007b). Predicting TBM excavability – part II. Tunnels & Tunnelling
International, December 2007, pp. 15–18.

Bieniawski, Z.T., Celada, B., Galera, J.M., Tardáguila (2008). New applications of the Excavability index
for selection of TBM types and predicting their performance. Proceedings of World Tunnel Congress 2008,
pp. 1618–1630.29.

Bruland, A. (1998). Hard Rock Tunnel Boring. PhD thesis, NTNU Trondheim Norwegian University of
Sciences and Technology.

Cassinelli, F., Cina, S., Innaurato, N., Mancini, R. & Sampaolo A. (1982). Power consumption and metal
wear in tunnel-boring machines: analysis of tunnel-boring operation in hard rock. Tunnelling ’82, London,
U.K., pp. 73-81.

Dolcini, G., Fuoco, S., & Ribacchi, R. (1996). Performance of TBMs in complex rock masses. North
American Tunnelling ’96, Balkema, Rotterdam.

Farrokh, E., Rostami, J., & Laughton, C. (2011). Analysis of unit supporting time and support installation
time for Open TBMs. Rock Mechanics and Rock Engineering, Vol. 44, Issue 4, pp.431-445.

Farrokh, E., Rostami, J. & Laughton, C. (2012). Study of various models for estimation of penetration rate
of hard rock TBMs. Tunnelling and Underground Space Technology, Vol. 30, pp. 110-123.

Farrokh, E. (2013). Study of utilization factor and advance rate of hard rock TBMs. PhD dissertation,
Pennsylvania State University, U.S.A.

Fontanille, G. (2009). Tunnelling Boring Machine Design. Post graduate master course “Tunnelling and
Tunnel Boring Machines”, Politecnico di Torino, Turin, pp. 19.

Frenzel, C. (2010). Verschleisskostenprognose für Schneidrollen bei maschinellen Tunnelvortrieben in

Festgesteinen. Technical report, TU Munich. Ingenieurgeologie Hydrogeologie Geothermie. Dr. Friedrich
Pfeil (Münchner Geowissenschaftliche Abhandlungen, Reihe B).

Gehring, K. (1995). Leistungs- und Verschleissprognosen im maschinellen Tunnelbau. Felsbau, Vol. 13,
pp. 439-448.

Gong, Q.M., & Zhao, J. (2007). Influence of rock brittleness on TBM penetration rate in Singapore
Granite. Tunnelling and Underground Space Technology, Vol. 22, pp.17-24.
 Gong, Q.M. & Zhao, J. (2009). Development of a rock mass characteristics model for TBM penetration
rate prediction. International Journal of Rock Mechanics and Mining Sciences, Vol. 46, pp. 8-18.

Grandori, R., Bieniawski, Z.T., Vizzino, D., Lizzadro, L., Romualdi, P., & Busillo, A. (2011). Hard rock
extreme conditions in the first 10 km of TBM driven Brenner Exploratory Tunnel. RETC 2011 Proceedings,
Littleton, Colorado, U.S.A., pp. 667-685.

Guglielmetti, V., Grasso, P., Mahtab, A., & Xu, S. (2007). Mechanized Tunnelling in Urban Areas –
Design Methodology and Construction Control. Taylor & Francis, London, U.K.

Hamidi, J.K., Shahriar, K., Rezai, B. & Rostami, J. (2010). Performance prediction of hard rock TBM
using Rock Mass Rating (RMR) system. Tunnelling and Underground Space Technology, Vol. 25, pp. 333-
345.

Hassanpour, J. , Rostami, J. & Zhao, J (2011). A new hard rock TBM performance prediction model for
project planning. Tunnelling and Underground Space Technology, Vol. 26, pp. 595-603.

Hoek, E., Kaiser P.,K., & Bawden W.F. (1995). Support of underground excavations in hard rock.
Balkema, Rotterdam.

Hoek, E. (2007). Rock engineering, Course notes by Evert Hoek. http://www.Rocscience.com.

Hood, M.C., & Roxborough, F.F. (1992). Rock Breakage: Mechanical. In: SME Mining Engineering
Handbook, Vol.1, Chapter 9, pp.680-721.

Innaurato, N., Mancini, R., Rondena, R. & Zaninelli, A. (1991). Forecasting and effective TBM
performances in a rapid excavation of a tunnel in Italy. Proceedings of the Seventh International Congress
ISRM, Aachen, Germany, pp. 1009-1014.

Maedl, B., Thewes, M., & Maidl., U (2014). Handbook of Tunnel Engineering – Volume I: Structures and
Methods. Ernst & Sons, Berlin, Germany.

McFeat-Smith, & I.; Concilia, M. (2000). Investigation, prediction and management of TBM performance
in adverse geological conditions. Gallerie e grandi opere in sotterranee, Vol. 62, pp. 21-28.

Palmström, A. (1993). The New Austrian Tunnelling Method (NATM). In: Fjellspregningteknikk
Bergmekanikk/Geoteknikk ’93, Issue 31, pp.1-20.

Ramezanzadeh, A. (2006). Performance Analysis and Development of New Model for Performance
Prediction of Hard Rock TBMs in Rock Mass. PhD thesis, Institut national des sciences appliqués de Lyon,
France.

Ribacchi, R. & Lembo Fazio, A., 2005. Influence of Rock Mass Parameters on the Performance of a TBM
in a Gneissic Formation (Varzo Tunnel). Rock Mechanics and Rock Engineering, Vol. 38, Nr. 2, pp. 105-
127.

Roby, J., Sandell, T., Kocab, J., & Lindbergh, L. (2008). The current state of disc cutter design and
development directions. Proceeding of 2008 North American Tunneling Conference, SME C, Vol. 4, pp. 36-
45.

Rostami, J., & Ozdemir, L. (1993). A new model for performance prediction of hard rock TBM. Rapid
Excavation and Tunnelling Conference 1993, Boston, U.S.A. , pp. 793-809.

Wickham, G.E., Tiedemann, & H.R., Skinner, E.H. (1974). Ground support prediction model - RSR
concept. Proc. 2nd North American Rapid Excavation & Tunnelling Conference (RETC), San Francisco.
American Institute of Mining, Metallurgical and Petroleum Engineers (AIME), New York, . pp. 691–707.

Wittke, W., Erichsen, C., & Gattermann, J. (2006). Stability Analysis and Design for Mechanized
Tunnelling. WBI, Felsbau GmbH, Aachen, Germany.
 Yagiz, S. (2008). Utilizing rock mass properties for predicting TBM performance in hard rock condition.
Tunnelling and Underground Space Technology , Vol. 23, pp. 326-339.