A new hard rock TBM performance prediction model for project planning_图文

Tunnelling and Underground Space Technology 26 (2011) 595–603

Contents lists available at ScienceDirect

Tunnelling and Underground Space Technology
journal homepage: www.elsevier.com/locate/tust

A new hard rock TBM performance prediction model for project planning
J. Hassanpour a,?, J. Rostami b, J. Zhao c

SCE Company, Tehran, IRAN, P.O. Box 16765-3465, Iran Department of Energy and Mineral Engineering, Pennsylvania State University, University Park, PA, USA c Ecole Polytechnique Federale de Lausanne (EPFL), Lausanne, Switzerland

a r t i c l e

i n f o

a b s t r a c t
Among the models used for performance prediction of hard rock tunnel boring machines two stand out and are often used in the industry. They include the semi theoretical model by Colorado School of Mines and the empirical model by Norwegian University of Science and Technology in Trondheim (NTNU). While each have their strong points and area of applications, more accurate prediction has been sought by modifying one of the existing models or introduction of a new model. To achieve this, a database of actual machine performance from different hard rock TBM tunneling projects has been compiled and analyzed to develop a new TBM performance prediction model. To analyze the available data and offer new equations using statistical methods, relationships between different geological and TBM operational parameters were investigated. Results of analyzes show that there are strong relationships between geological parameters (like UCS, joint spacing and RQD) and TBM performance parameters specially Field Penetration Index (FPI). In this study, a boreability classi?cation system and a new empirical chart, for preliminary estimation of rock mass boreability and TBM performance is suggested. ? 2011 Elsevier Ltd. All rights reserved.

Article history: Received 12 October 2010 Received in revised form 27 March 2011 Accepted 9 April 2011 Available online 4 May 2011 Keywords: TBM performance Rock mass boreability Field Penetration Index

1. Introduction Hard rock tunnel boring has become more or less the standard method of tunneling for tunnels of various sizes with lengths over 1.5–2 km. With more ef?cient and powerful machines, TBMs have been used in various ground conditions from extremely hard and massive to broken and blocky grounds. To justify the use of TBM in any project and for planning purposes, a reasonably accurate estimation of rate of penetration (ROP), daily rate of advance (AR), and cutter cost/life estimate is necessary. Various models have been offered throughout the years for offering such estimates, which in some cases were successful with pin point accuracy, and in other instances, off by a good margin. This has been the source of interest to better understand machine-rock interaction and to develop a more accurate model for performance estimate of hard rock TBMs. Currently, two models, including Colorado School of Mines or CSM (Rostami and Ozdemir, 1993; Rostami, 1997) and Norwegian University of Science and Technology or NTNU (Blindheim, 1979; Bruland, 1998) models are the most recognized TBM performance prediction and prognosis models in use around the world. In the last couple of decades, with growing use of TBMs in the world and the necessity to accurately predict performance of machines
? Corresponding author. Mobile: +98 912 2279442; fax: +98 21 22524502.
E-mail addresses: Jafar_hassanpour@yahoo.com, Jafar.hassanpour@gmail.com (J. Hassanpour). 0886-7798/$ - see front matter ? 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.tust.2011.04.004

in different ground conditions, many researchers have worked to develop new prediction models or adjustment factors for the common existing models. Research works by Barton (1999, 2000), Yagiz (2002, 2007), Sapigni et al. (2002), Gong and Zhao (2009), Hassanpour (2009, 2010), Hassanpour et al. (2009, 2010), etc. are the most recent works on this topic. Barton (1999, 2000) reviewed a wide range of TBM tunnels to establish a database for proposing a new model based on Q rock mass classi?cation system and adding some new parameters to the existing system to be able to use it for TBM applications. This new model, namely QTBM uses many input parameters (such as RQD, joint condition, Stress condition, intact rock strength, quartz content and TBM thrust) for estimating QTBM and consequently penetration rate and advance rate of the machine. Yagiz (2002) modi?ed the CSM model adding rock mass properties as input parameters into the model. Ramezanzadeh (2005) has also followed up on this work and developed a database of TBM ?eld performance for over 60 km of tunnels. He too, offered adjustment factors for CSM models to account for joints and discontinuities. Sapigni et al. (2002) studied the empirical relation between RMR and penetration rate. Also Ribacchi and Lembo-Fazio (2005) evaluated the relationship between RMR and performance of a double shield machine in the Varzo tunnel. Yagiz (2007) performed statistical analyzes on data obtained from Queens’s tunnel, in New York and proposed an empirical model to predict TBM penetration rate. He has related four rock mass parameters (UCS, Punch test index or PTI, spacing and orientation of joints) to penetration rate of


J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603

machine. In a similar research work, Gong and Zhao (2009) by performing a nonlinear regression analysis on data obtained from two tunnels excavated in granitic rock masses in Singapore developed an empirical equation to estimate boreability of rock mass. They proposed a relationship between four rock mass parameters (UCS, brittleness, joint count number, and orientation of joints) and boreability index of the rock mass. In a more recent study by the authors (Hassanpour, 2009, 2010; Hassanpour et al., 2009, 2010), based on data obtained from main tunneling projects in Iran and investigating relationships between rock mass properties and actual machine performance, some new empirical equations have been proposed for estimation of TBM performance in the given ground conditions. The above models, of course, have their advantages and disadvantages because of their origin and background. Some of them like original CSM model don’t consider the main in?uencing parameters and some of them like NTNU model require special experiments originated from the drilling. These tests are not commonly available outside Norway. Also some of the models like QTBM are too complicated. QTBM model originates from Q system and includes too many parameters for practical application. In addition, some parameters are overlapped in this model (Gong and Zhao, 2009). On the other hand growth of TBM manufacturing technology and existence of some shortcomings in the prediction models have made it necessary to perform more research on the development of the new models. In this study, compiled ?eld data obtained from three main tunneling projects in Iran (SCE Company, 2004, 2006, 2008) as well as the Manapouri tunnel project recently completed in New Zealand (URS Company, 2003), were used to establish a new concept for rock mass boreability classi?cation and a more general model for TBM performance estimation. 2. Description of the projects used for this study For developing a more accurate TBM performance prediction model that can be applied in different geological conditions, data from different projects with different rock mass conditions were collected and compiled in a database. As mentioned above, three long water conveyance tunnels recently constructed in different
Table 1 Main characteristics of tunneling projects. No 1 2 3 4 Project Karaj water conveyance tunnel, Lot 1 (Iran) Ghomrood water conveyance tunnel, Lots 2, 3 & 4 (Iran) Zagross water conveyance tunnel, Lot 2 (Iran) Manapouri second tailrace tunnel (New Zealand) Tunnel length (km) 15.9 24.5 26 10 Available data (km) 15.9 24.5 5.3 9.7

geological zones of Iran and Manapouri second tailrace tunnel were selected for this study. The main characteristics of these TBM tunneling projects are summarized in Tables 1 and 2. As can be seen in the Table 2, these tunnels have been constructed in different rock types including sedimentary, igneous and metamorphic rocks with a wide range of rock strength. 3. TBM ?eld performance database In this study data on geological and ground conditions, TBM operational parameters and machine performance represented by rate of penetration were collected during pre-construction and construction phases. The data were arranged in a special database including 158 tunnel sections of four selected projects (Table 1) where the ground conditions and machine performance information were valid and could be veri?ed. The data sets comprised two main categories. The ?rst category included machine performance parameters like net boring time, length of mined section and also the average of machine operational parameters (thrust, RPM, power and applied torque) throughout the section. These parameters were obtained from the daily operating records and the TBM data logger. Also the most important performance parameters including average rate of penetration (ROP), penetration per revolution (P), Field Penetration Index (FPI, Tarkoy and Marconi, 1991), and speci?c energy (SE) have been calculated using formulae (1)–(4) as listed below:


Lb tb



ROP ? 1000 RPM ? 60 Fn P




SE ?

200 ? NTBM ? rmc F r ? 3 ? dTBM P


where ROP is rate of penetration (m/h), Lb is boring length (m), tb is boring time (h), P is cutter penetration in each cutterhead

TBM type and manufacturer Double shield (Herrenknecht) Double shield (Wirth) Double shield (Herrenknecht) Main beam open TBM (Robbins, Kvaerner-Markham)

Construction period 2006–2009 2003–2009 2006–present 1997–2002

TBM diameter (m) 4.65 4.525 6.73 10.05

Table 2 Geological characteristics of tunneling projects. No Project Geologic zone Formation Lithology UCS range (MPa) 30–150 25–150 Max. overburden (m) 600 700

1 2

Karaj tunnel, Lot 1 Ghomrood tunnel, Lots 2, 3 & 4 Zagross, Lot 2 Manapouri tunnel

Central Alborz Sanandaj-Sirjan metamorphic belt Zagross Simply folded zone –

Pyroclastic rocks of Karaj formation Jurassic metamorphic rocks (low to medium grade) and Cretaceous Limestone Carbonate-Argillaceous rocks of Pabdeh, Gurpi and Ilam Formations Paleozoic metamorphic and igneous rocks of the Fiordland Complex

Tuffs, Shaly and Sandy Tuffs, Agglomerate, ... Limestone, Shale and sandstone, Slate, Phyllite, Schist with quartzitic veins Limestone, Shale and Limy Shales Gneiss, calc-silicate and quartzite and the intrusive rocks (gabbro and diorite)

3 4

15–150 100–225

650 1200

J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603


Fig. 1. Distribution curve and frequency histogram of rock mass and TBM performance parameters in the database.

Table 3 Summary results of determination of regression coef?cients of different geological and geomechanical parameters with FPI. Parameter UCS (Mpa) Joint frequency Spacing (m) RQD0 (%) Regression coef?cients (R2) 0.699 0.788 0.688 Regression type Exponential Power Exponential Quadratic Exponential Power Power Exponential Exponential Relationship FPI = 6.883 e0.013UCS FPI = 63.267 Sp0.847 FPI = 3.490 e0.027RQD FPI = 0.053BRMR2 ? 4.205BRMR + 92.068 FPI = 4.619 e0.023GSI FPI = 15.309 Q0.304 FPI = 17.389 Q 0:269 C FPI = 8.317 e0.042Sigmacm 0.011RMCI FPI = 10.525 e

Rock mass classi?cation systems Basic RMR 0.531 GSI 0.455 Q 0.487 Rock mass strength parameters Qc 0.601 Sigmacm (Mpa) RMCI 0.541 0.619


J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603

Fig. 2. Correlation between different rock mass properties and FPI.

revolution (mm/rev), RPM is cutterhead revolutions (rev/min), FPI is Field Penetration Index (kN/cutter/mm/rev), Fn is cutter load or normal force (kN), SE is speci?c energy (MJ/m3), Fr is cutter rolling force (kN), dTBM is TBM diameter (m), NTBM is number of cutters on the cutterhead and rmc is the weighted average cutter distance from center of rotation (m).

The second part of database or category of information included some geological parameters such as intact rock properties (Compressive and tensile strength, quartz content, porosity), discontinuity characteristics such as spacing, NTH class or fracture class developed in NTNU (Bruland, 1998), surface condition and also results of calculation of some rock mass parameters (like RQD, RMR,

J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603


Table 4 (a) Variables and summary of the generated model for forward stepwise regression analysis; (b) signi?cance of r-value and coef?cients for generated model and (c) analysis of variance for the signi?cance of regression for generated model. (a) Model summary Model 1 (b) Coef?cients Model R2 0.788 Unstandardized coef?cients B 1 (Constant) UCS RQD Sum of squares Regression Residual Total 1.384 0.008 0.015 df 97.630 26.304 123.934 Std. error 0.093 0.001 0.002 Mean square 2 155 157 48.815 0.170 Adjusted R2 0.785 Standardized coef?cients Beta 0.494 0.452 14.914 8.658 7.925 F 287.654 0.000 0.000 0.000 Sig. 0.000 t

R 0.888

Std. error of the estimate 0.41195 Sig.

(c) Anova Model 1

Fig. 3. Comparison of actual FPI values with predicted values.

Q and GSI) and rock mass strength parameters in selected tunnel sections. Some of the most important rock mass strength parameters are calculated indices like Qc (Barton, 2000), rock mass compressive strength or UCSrm (Hoek, 2007), rock mass strength or Sigmacm (Barton, 2000) and rock mass cuttability index or RMCI (Bilgin et al., 1997) which can be calculated using Eqs. (5)–(8):

It should be noted that, since most of the selected tunnels have been excavated by shielded machines and lined with precast segmental linings, there were many limitations for mapping of the geological features in total length of the tunnels. So, just tunnel sections where reliable geological data were available were selected for the database. In these selected tunnel sections geological data were obtained directly from tunnel rock face observations and measurements or by investigating drilled boreholes logs and core boxes and related laboratory tests on core samples. Graphs presented in Fig. 1 show the histograms and distribution curves of different geological and TBM performance parameters recorded in the database. As can be seen in the graphs, geological and performance parameters have wide ranges of variations. Actually, in the developed database some tunnel sections have been excavated in weak and very poor quality rock masses (UCS < 20 MPa and RQD < 25%) and some others have been excavated in very strong and massive rock masses (UCS > 150 MPa and RQD = 100%). Also, ranges of variations of TBM performance parameters are very wide and for example FPI ranges from min. 2.75 to max. 145.6 kN/cutter/mm/ rev. These wide ranges of geological and performance parameters helped in developing a more comprehensive TBM performance prediction model which has covered different geological conditions.

4. Developing empirical equations In this study, both single and multi-variable regression analyzes were used to investigate relationship between engineering rock properties and TBM performance parameters and ?nally to develop empirical equations. As indicated in previous publications (Hassanpour, 2009, 2010; Hassanpour et al., 2009, 2010), among the selected machine parameters (including P, ROP, FPI and SE) the Field Penetration Index or FPI which is a composite parameter (Eq. (3)) shows the best correlations with geological parameters. Consequently, FPI is selected as a suitable machine performance parameter for developing empirical relationships with geological parameters in the current study as well. Results of correlations of FPI with different rock mass parameters are summarized in Table 3 and Fig. 2. As shown in Table 3, among the geological parameters, some simple parameters like Spacing, RQD and UCS show good correlations with FPI (R2 > 0.68). Also, among composite geological parameters, RMCI shows a good correlation with FPI (Table 3). This parameter (Eq. (6)) represents two of the main rock mass parameters, namely UCS and RQD and the developed equation shows a

Qc ? Q ? RMCI ?

UCS 100 UCS ?RQD?2=3 100


?6? ?7? ?8?

Sigmacm ? 5 ? c ? Q 1=3 UCSrm ? UCS ? sa

s ? Exp

100 ? GSI 9 ? 3D



1 1 ?GSI=15 ? ?e ? e?20=3 ? 2 6


In Eq. (9a), D is disturbance factor (D = 0 in TBM projects) (Hoek, 2007) and c is density of rock.


J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603

Fig. 4. Chart for estimating FPI based on rock mass properties driven from Eq. (10).

Table 5 Previous equations developed by authors for each tunnel project separately. Project Karaj tunnel Zagross tunnel Ghomrood tunnel Manapouri tunnel Main Lithology Pyroclastic rocks Carbonate-argillaceous rocks Low grade metamorphic rocks Igneous and metamorphic rocks Equation FPI = exp FPI = exp FPI = exp FPI = exp (0.005UCS + 0.002RQD + 2.129) (0.004UCS + 0.008RQD + 2.077) (0.004UCS + 0.023RQD + 1.003) (0.005UCS + 0.020RQD + 1.644) Regression coef?cients (R2) 0.523 0.645 0.874 0.488 Reference Hassanpour et al. (2010) Hassanpour et al. (2009) Hassanpour (2010) Unpublished

reasonable correlation between the performance and geological parameters. Using FPI and RMCI encompasses some of the most in?uential parameters including UCS (intact rock strength), RQD (degree of fracturing of the rock mass), average cutter thrust, which is the most important TBM operational parameter and main controlling parameter for machine performance, and penetration, which is the result of these interacting parameters. It should be noted that in calculating FPI, machine penetration or P (mm/rev) is used and that allows for scaling the results to any size machine using the cutterhead rotational speed, or RPM. Obviously, this equation can be a useful and comprehensive tool for predicting ROP for hard rock TBMs. As mentioned before, in this study, multi-variable regression analysis was also used to ?nd an empirical equation to relate FPI as a function of machine performance parameters to geological parameters. For this purpose, four rock mass properties including UCS, joint spacing, RQD, and a (angle between tunnel axis and discontinuity) were used as independent variables and the recorded FPIs were treated as dependent variable. In?uence of each variable on the FPI was evaluated using forward stepwise regression analyzes. Statistical analyzes were performed by Version 11.5 of SPSS

software (2002). After consideration of different combinations of parameters, it appears that the best results could be obtained by excluding the two parameters a and spacing. The results show good correlation between Ln(FPI) as response parameter and UCS and RQD as predictors, in a linear combination with a 95% con?dence level. As a result, a new equation was introduced as follows:

FPI ? exp?0:008UCS ? 0:015RQD ? 1:384?


As shown in Table 4, the regression coef?cient (R2) for this equation is 78.5%. It indicates that the above regression model explains 78.5% of the total variance of the 158 datasets. A simple t-test and F-test analysis of the results indicates that the correlations are real and the coef?cients are true (Table 4). To evaluate accuracy of the model, the measured and calculated values of FPI are compared in Fig. 3. As shown, most of the predicted values of FPI especially when FPI < 25 kN/cutter/mm/rev are close to actual values. To facilitate using of Eq. (10), a FPI prediction chart is also developed and presented as Fig. 4. This chart can be used for quick estimation of range of values for FPI in grounds with different rock strength and rock quality. As shown in this chart, different rock

J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603 Table 6 Summary of ground conditions for various boreability classes. Boreability class B-0 B-I B-II B-III B-IV B-V FPI range (kN/mm/ rev) >70 40–70 25–40 15–25 7–15 <7 Rock mass boreability Tough Fair-tough Good-fair Good Very good Excellent Stability condition TBM excavatability (relative dif?culty of ground for TBM use) Tough Fair Good Very good Good May be problematic Example


Completely stable Stable Minor instabilities Only local structural instabilities Some major instabilities Collapse, gripper problems, squeeze, etc.

Very strong and massive quartzitic veins , intrusive and metamorphic rocks Massive igneous and metamorphic rocks Blocky and jointed Tuffs, Tuf?tes, Limestones Alternations of Sandstones, limestones and Shales Alternations of thin bedded Shale and Sandstone layers Highly foliated and schistose metamorphic rocks (Slate, Phyllite, Graphite schist), Shale, Marlstone, thick fault zones

units in the selected tunnel projects have a speci?c range of FPI based on their strength and fracturing degree (or RQD). The proposed equation in this study (Eq. (10)) is similar to the equations of Table 5 obtained by the analysis of data for each tunnel project separately (Hassanpour, 2009, 2010; Hassanpour et al., 2009, 2010). As shown in Table 5 the coef?cient of UCS ranges from 0.004 to 0.005, while coef?cient of RQD varies from 0.002 to 0.023 and constant ranging from 1 to 2.1 but all positive. 5. Classi?cation of rock mass boreability Boreability is the term commonly used to express the ease or dif?culty of rock mass excavation by a tunnel boring machine. In tunneling projects, ground characteristics or boreability of the rock mass is an important parameter for selecting machine type and speci?cations. It is clear that proper evaluation of rock mass boreability can also play a major role in machine operation to achieve the best performance. In a given project, rock mass properties have direct in?uence on boring dif?culty of ground and FPI values. Usually stronger and less fractured rock masses are more dif?cult for cutting by disk cutters and boring by TBM and require use of higher thrusts to achieve a certain level of penetration. Therefore higher values of FPI are usually recorded in strong and massive rock masses like massive intrusive sills, dikes, and thick quartzitic veins (typically higher than 70 kN/cutter/mm/rev). On the other hand, in poor quality rock masses such as very foliated and schistose rocks (shale, slate, phyllite), there is no need to apply high thrust values for reasonable penetration and therefore FPI values are small and typically less than 10 kN/cutter/mm/rev. So, FPI can be selected as an index for categorizing rock mass boreability. In this study based on actual FPI values measured in selected tunnel sections, six rock mass boreability classes, from most dif?cult for boring or B-0 class (Tough) to easiest for boring or B-V class (Excellent) were de?ned (Table 6). Clearly, rock mass quality is different in these boreability classes. Range of UCS and RQD values in each class were evaluated by statistical analyzes on data in the available TBM ?eld performance database and the results are presented in Figs. 5 and 6. Table 6 also presents general stability conditions of different rock mass boreability classes. As shown, stability conditions in different classes vary from completely stable rock masses to problematic grounds. Tunnel wall instabilities have negative in?uences on utilization factor of the machine and operational parameters during excavation of the tunnel. So, advance rate of a TBM in rock masses with very good to excellent boreability (B-IV and B-V classes) can result in high instantaneous penetration rate but due to instability problems and some machine limitations they suffer

Fig. 5. Variation range of UCS in different boreability classes.

Fig. 6. Variation range of RQD in different boreability classes.

from lower utilization and usually result in lower daily advance rate. To better illustrate the FPI range for each boreability class and facilitate application of the model in estimation of machine performance in the future, the chart presented in Fig. 4 has been converted to a rock mass boreability prediction chart (Fig. 7). In this new chart FPI has been related to rock strength and a more visual geological characteristic of rock masses or rock mass structure, according to an idea presented by Hoek (2007) to evaluate geological strength index (GSI). Rock mass structure can be evaluated in the ?eld by an engineering geologist more easily and offer a quick sense of rock mass behavior.


J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603

Fig. 7. Rock mass boreability prediction chart with actual FPI range for different geological units in the database.

Table 7 shows examples of TBM performance estimation in rock masses with different boreability classes. In this table penetration rate and advance rate of TBM in six assumed rock masses (A–F in Fig. 7) with boreability classes from B-0 to B-V have been estimated using the proposed model. As shown, the ?rst step is to predict FPI from the chart in Fig. 4 or Fig. 7, or alternatively calculate it from Eq. (10). This offers a speci?c value of FPI for each rock mass. This is followed by assuming a practical value for average disk cutter load (Fn) and RPM (rev/min), to calculate ROP (m/h) and P (mm/rev). For example, for an assumed Fn = 200 kN/ cutter and RPM = 7 rev/min the ROP is calculated as:
Table 7 Example of machine performance estimation using developed model. Rock mass Rock mass boreability class UCS (MPa) RQD (%) A: Theoretical machine performance Theoretical advance rate FPI (kN/cutter/mm/rev) (Fig. 3) Assumed thrust (kN/cutter) Assumed RPM (rev/min) P (mm/rev) ROP (m/h) Assumed U (%) AR (m/day) B: Practical machine performance Actual advance rate FPI (kN/cutter/mm/rev) (Fig. 3) Assumed thrust (kN/cutter) Assumed RPM (rev/min) P (mm/rev) ROP (m/h) Assumed U (%) AR (m/day) A B-0 210 100 B B-I 160 100

ROP ?m=h? ?

0:06 ? 200 ? 7 FPI


Finally, by assuming an average utilization factor of 25%, advance rate or AR (m/day) of the machine in each class can be estimated. As can be seen in part A of Table 7, advance rate of the machine in different classes ranges from 5.3 m/day in class B-0 to more than 73 m/day in class B-V. Actually, there are large differences between advance rates of machine in different classes. Normally, achieving high values of advance rate in rock masses with classes B-IV and B-V is very rare and dif?cult due to ground

C B-II 100 80

D B-III 75 60

E B-IV 50 40

F B-V 30 20

96 200 7 2.08 0.88 25 5.3






3.11 1.31 7.8

6.78 2.85 17.1

11.18 4.70 28.2

18.44 7.74 46.5

29.20 12.27 73.6

96 220 7 2.29 0.96 25 5.8

64 180 7 2.80 1.18 7.1

29 180 7 6.10 2.56 15.4

18 180 7 10.06 4.23 25.4

11 80 4 7.37 1.77 5 2.1

7 50 3 7.14 1.29 1.5

J. Hassanpour et al. / Tunnelling and Underground Space Technology 26 (2011) 595–603


stability issues and due to some limitations of machine and backup system such as capacity of conveyor belt. In general, more competent and stronger rocks indicated in classes B-0 and B-I coincide with higher utilization rate due to minimal ground support requirement and related stoppages and downtime. On the opposite side of the scale, lower utilization rate is often experienced in unstable grounds typical of class B-IV and B-V which reduces the utilization rate drastically and causes higher downtime and delays. In addition, in softer grounds, operator reduces cutterhead thrust to prevent cutterhead jamming due to high torque, minimizing the chance of face collapse, and reducing the possibility of overloading of conveyor belt. This simply means that same levels of applied cutter load are not used in soft ground (as it applies to stronger rocks) and by default the use of lower applied load will ease off the extremely high rates of penetration. Therefore, in part B of Table 7, by considering more practical values for utilization factor, thrust and RPM in such grounds, more realistic values for penetration rate and advance rate of machine can be estimated. By reducing utilization rate, thrust and RPM to 5%, 80 kN/cutter and 4 rpm in class B-IV and to 5%, 50 kN/cutter and 3 rpm in class B-V, estimated penetration rate and advance rate, have reduced considerably. This is more in line with the practical application of TBMs in the ?eld. However it is concluded that highest values of advance rate can be achieved in rock masses categorized as classes B-II and B-III. In such grounds, combined conditions of rock mass boreability and stability of the surrounding rock are in the optimum condition for excavating the tunnel by TBM. 6. Conclusion Rock mass boreability depends on a number of in?uencing parameters including intact rock/rock mass properties, machine speci?cations and operational parameters. In this paper a simple model is proposed to evaluate rock mass boreability and TBM performance range. In the developed model, machine performance has been related to two main rock properties (UCS and RQD) and two operational parameters (average cutterhead thrust and RPM). These Input parameters of the model are typical parameters usually are available even in the preliminary stages of the tunnel design and planning. So, proposed model can be applied as a useful tool for quick estimation of TBM performance in projects with different geological conditions and machine diameters. This paper also introduces a new boreability classi?cation based on rock masses characteristics to allow for prediction of FPI values. Various ground conditions are categorized in six different classes from B-0 as tough boring ground to B-V as easy ground for boring. Combining stability conditions which controls machine utilization and boreability characteristics of different rock types and ground condition allows for development of a new concept for classi?cation of TBM excavatability or ‘‘Relative dif?culty of ground for TBM application’’. As mentioned, required FPI for boring the rock mass, decreases from B-0 to B-V class. It means that from B-0 to B-V class, less cutterhead thrust is required to achieve a given rate of penetration. It is also concluded that rock masses with class B-II and B-III due to optimum conditions of stability and boreability are most favorite grounds for TBM application. The Developed empirical model is based on analyzing data obtained from four tunneling projects with total length of 55 km and boring diameter of 4.5–10 m which have been excavated in different geological conditions. So, it can be considered a TBM performance prediction model which is applicable in a wide range of geological conditions including layered and jointed sedimentary rocks, blocky – jointed pyroclastic and carbonate rocks, foliated and schistose metamorphic rocks, massive igneous and metamor-

phic rocks and fractured rock masses. This model must be applied with caution in highly fractured rock masses (or fault zones) and water sensitive rocks like marlstones and mudstones. Acknowledgments Authors wish to thank SCE Company, especially our colleagues in the Tunneling Division for their help in the collection of required data. References
Barton, N., 2000. TBM Tunneling In Jointed and Faulted Rock. Balkema, Rotterdam, 173. Barton, N., 1999. TBM performance estimation in rock using QTBM. Tunnels and Tunneling International 31 (9), 30–33. Bilgin, N., Kuzu, C., Eskikaya, S., 1997. Cutting performance of rock hammers and roadheaders in Istanbul metro drivages. In: Golser, J., et al. (Eds.), Proceedings of World Tunnel Congress 97 and 23th World Assembly of ITA: Tunnels for People, Vienna, Austria, Balkema, Rotterdam, pp. 455–460. Blindheim, O.T., 1979. Boreability Predictions for Tunneling. Ph.D. Thesis, Department of Geological Engineering, The Norwegian Institute of Technology, p. 406. Bruland, A., 1998. Hard Rock Tunnel Boring. Ph.D. Thesis, vol. 1–10, Norwegian University of Science and Technology (NTNU), Trondheim, Norway. 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 Science 46 (1), 8–18. Hassanpour, J., Rostami, J., Khamehchiyan, M., Bruland, A., 2009. Development new equations for performance prediction. Geo Mechanics and Geoengineering: An International Journal 4 (4), 287–297. Hassanpour, J., 2009. Investigation of the Effect of Engineering Geological Parameters on TBM Performance and Modi?cations to Existing Prediction Models. Ph.D. Thesis, Tarbiat Modares University, Tehran, Iran. Hassanpour, J., Rostami, J., Khamehchiyan, M., Bruland, A., Tavakoli, H.R., 2010. TBM performance analysis in pyroclastic rocks, a case history of Karaj Water Conveyance Tunnel (KWCT). Journal of Rock mechanics and Rock Engineering 4, 427–445. Hassanpour, J., 2010. Analysis of actual TBM performance in Ghomrood project, Bulletin of Iranian tunneling Association, No 9 (article in Persian). Hoek, E., 2007. Rock Engineering, Course notes by Evert Hoek, p. 339, http:// www.Rocscience.com. Ramezanzadeh, A., 2005. Performance analysis and development of new models for performance prediction of hard rock TBMs in rock mass, Ph.D. Thesis, INSA, Lyon, France, p. 333. Ribacchi, R., Lembo-Fazio, A., 2005. In?uence of rock mass parameters on the performance of a TBM in a Gneissic formation (Varzo tunnel). Rock Mechanics and Rock Engineering 38 (2), 105–127. Rostami, J., Ozdemir, L., 1993. A new model for performance prediction of hard rock TBM, In: Bowerman, L.D., et al. (Eds.), Proceedings of RETC, Boston, MA, pp. 793– 809. Rostami, J., 1997. Development of a Force Estimation Model for Rock Fragmentation with Disc Cutters Through Theoretical Modeling and Physical Measurement of Crushed Zone Pressure. Ph. D. Thesis, Colorado School of Mines, Golden, Colorado, USA, P. 249. Sapigni, M., Berti, M., Behtaz, E., Busillo, A., Cardone, G., 2002. TBM performance estimation using rock mass classi?cation. International Journal of Rock Mechanics and Mining Sciences and Geomechanics Abstracts 39, 771–788. SCE Company, 2004. Geological and Engineering Geological Report for Ghomrood Water Conveyance Tunnel Project (Lots 3 & 4), Unpublished report. SCE Company, 2006. Geological and Engineering Geological Report for Karaj Water Conveyance Tunnel Project (Lot1), unpublished report. SCE Company, 2008. Geological and Engineering Geological Report for Zagross Water Conveyance Tunnel Project (Lot 2). Unpublished report. SPSS Software, 2002. Statistical Package, version 11.5, SPSS Inc., Chicago, IL, USA, <http://www.spss.com>. Tarkoy, P.J., Marconi, M., 1991. Dif?cult rock comminution and associated geological conditions. In: Tunneling 91, Sixth International Symposium, 14– 18 April 1991, Hovotel, London, England, Institute of Mining and Metallurgy, Elsevier Applied Science, London, pp. 195–207. URS Company, 2003. Manapouri Power Station Second Tailrace Tunnel Engineering Geological Construction Report, Prepared for Meridian Energy Ltd., Unpublished report. Yagiz, S., 2002. Development of Rock Fracture and Brittleness Indices to Quantify the Effects of Rock Mass Features and Toughness in the CSM Model Basic Penetration for Hard Rock Tunneling Machines. Ph.D. Thesis, Department of Mining and Earth Systems Engineering, Colorado School of Mines, Golden, Colorado, USA, p. 289. Yagiz, S., 2007. Utilizing rock mass properties for predicting TBM performance in hard rock condition. Tunnelling and Underground Space Technology 23 (3), 326–339.


Performance Prediction Models for Hard Rock Tunnel Boring Machines
rock mass properties and tbm performance prediction
Performance prediction of TBM disc cutting on granitic rock by the linear cutting test
Comparison between CSM and NTH hard rock TBM performance prediction models
The Role of Individualism and the Five-Factor Model in the Prediction of Performance in a Leaderless
Performance Prediction of Hard Rock TBM Based on Extreme Learning Machine
Performance prediction of hard rock TBM using Rock Mass Rating (RMR) system
Model for assessing building contractors' project performance
TBM performance estimation using rock mass classifications
Empirical Evaluation of Model-based Performance Prediction Methods in Software Development