Excavation of underground research laboratory ramp in granite using tunnel boring machine: Feasibility study

Hongsu M, Ju Wng,*, Ke Mn, Ling Chen, Qiuming Gong, Xinggung Zho

a CNNC Key Laboratory on Geological Disposal of High-level Radioactive Waste, Beijing Research Institute of Uranium Geology, Beijing, China

b North China University of Technology, Beijing, China

c Key Laboratory of Urban Security and Disaster Engineering of Ministry of Education, Beijing University of Technology, Beijing, China

Keywords:Underground research laboratory (URL)High-level radioactive waste(HLW)disposal Tunnel boring machine (TBM)Extremely hard rock mass Rock mass boreability Spiral layout Beishan

A B S T R A C T Underground research laboratory(URL)plays an important role in safe disposal of high-level radioactive waste(HLW).At present,the Xinchang site,located in Gansu Province of China,has been selected as the final site for China’s first URL,named Beishan URL.For this,a preliminary design of the Beishan URL has been proposed,including one spiral ramp,three shafts and two experimental levels.With advantages of fast advancing and limited disturbance to surrounding rock mass, the tunnel boring machine (TBM)method could be one of the excavation methods considered for the URL ramp.This paper introduces the feasibility study on using TBM to excavation of the Beishan URL ramp.The technical challenges for using TBM in Beishan URL are identified on the base of geological condition and specific layout of the spiral ramp. Then, the technical feasibility study on the specific issues, i.e. extremely hard rock mass, high abrasiveness, TBM operation, muck transportation, water drainage and material transportation, is investigated. This study demonstrates that TBM technology is a feasible method for the Beishan URL excavation. The results can also provide a reference for the design and construction of HLW disposal engineering in similar geological conditions. © 2020 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences.Production and hosting by Elsevier B.V.This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

As accepted globally, underground research laboratory (URL) is an integral part of research and development (R&D) program for safe disposal of high-level radioactive waste (HLW) (NEA, 2001;OECD/NEA, 2013).URL provides a scientific and technical platform for characterization of deep geological environment, implementation of full-scale multidisciplinary tests, development of methods and techniques,demonstration of disposal technologies,as well as for international cooperation, staff training,and public acceptance(SKB,1996; OECD/NEA, 2013; Wang, 2014).

There are more than 20 URLs built and in good operation around the world,e.g.Äspӧ Hard Rock Laboratory in Sweden,Grimsel Test Site in Switzerland, and Bure URL in France. A large number of experiments have been performed at the URLs and the outputs have greatly improved our understanding of both technical and practical issues on deep geological disposal of nuclear waste(Read,2004; Delay et al., 2014). In China, the deep geological disposal program for HLW has entered into a new steady development stage in the 21st century (Wang, 2008). In 2016, the Xinchang site,located in Beishan pre-selected area of Gansu Province, has been selected as the final site of China’s first URL,named Beishan URL.In May 2019,the proposal of Beishan URL project was approved by the China government, and construction of the URL will start in 2021.The preliminary design of the URL project has been proposed,with the main structure of an access ramp, three shafts and two experimental levels.At the project proposal phase,one of the important issues is to select suitable excavation methods for the URL.

Generally, there are two main excavation methods: drill-andblast method and tunnel boring machine (TBM) method. TBM has been widely used in tunnel construction over the past decades,and particularly in China, the application of TBMs for public transportation and hydropower projects has been increased significantly in recent years (e.g. Liu et al., 2016; Hong et al., 2018). Compared with drill-and-blast method, TBM has several advantages, such as higher advance rate, higher construction quality, being safer to workers and more eco-friendly to environment, and particularly the limited disturbance and damage to the surrounding rock mass,which is regarded as one extremely significant factor for the HLW disposal project.However,TBM also has disadvantages,such as low flexibility to extremely adverse geological conditions, and low flexibility to project layout (Bäckblom et al., 2004). For TBM tunneling in hard rock, several kinds of adverse geological factors influencing TBM performance, such as highly fractured/faulted zone,high in situ stress,large groundwater inflow,extremely high rock strength, and high rock abrasiveness, have been experienced in many projects(Barla and Pelizza,2000;Barton,2000;Delisio and Zhao, 2014; Gong et al., 2016a; Hong et al., 2018). Such kind of adverse ground conditions may limit TBM advance rate, and introduce high cutter consumption and damage to machine and staff.Thus,it is mandatory to conduct technical feasibility study on TBM application for Beishan URL before construction.

This paper tries to validate whether the TBM method can be used for the URL spiral ramp.In view of the site-specific geological condition with respect to the layout of spiral ramp, several technical issues,i.e.extremely hard rock mass,high abrasiveness,small curve radius, high slope inclination, difficulty of muck transportation, water drainage and material transportation for long distance, are identified as the potentially technical difficulties for using TBM in URL project.As rock mass boreability in this project is regarded as the most challenging issue,the main part of feasibility analysis in this paper will be focused on rock mass boreability. To investigate rock mass boreability for TBM excavation for the project,several approaches can be used,i.e.field study,laboratory test,numerical modeling,and prediction model.Generally,each of these approaches has its advantages as well as some limitations.

Specifically,field study can provide excellent evaluation of TBM performance-under the given geological conditions (Gong et al.,2007; Macias et al., 2017; Jing et al., 2019). However, the field study cannot be performed systematically under given TBM cutter head design and ground conditions.Moreover,field study generally means high cost for field investigations; even so, field study is barely available for most researchers-Laboratory tests with various scales have been developed in the past years (e.g. Rostami,1997;Cho et al., 2013; Yin et al., 2014; Gong et al., 2016b). Full-scale cutting test, e.g. linear cutting test, is a reliable approach to predict TBM performance under certain ground conditions (Rostami,1997; Chang et al., 2006; Balci, 2009; Ma et al., 2016; Pan et al.,2018). This approach uses full-scale rock samples and cutters, and applies a realistic range of cutting forces. Accordingly, the test results can be directly used to predict TBM performance-in field(Cho et al., 2013). However, the full-scale test is more expensive when compared to small-scale tests and numerical modeling. Another shortcoming is that it is very difficult or almost impossible to obtain the full-size samples for some projects (Cho et al., 2013).

Alternatively, many researchers numerically investigated rock fragmentations induced by TBM cutter under given ground conditions(e.g.Ma et al.,2011;Cho et al.,2013;Zhang and Wong,2013;Xiao et al., 2017). Improvements have been made to enhance the accuracy of simulation processes, e.g. crack initiation and propagation and chipping formation. However, it is still difficult to numerically reveal the real process of rock fragmentation under TBM cutters,owing to the rock heterogeneity and the complexity of interaction between machine and rock.

For the prediction models, there are generally two categories developed in the past decades: one is the theoretical approach based on the force balance, and the other is the empirical models based on the laboratory tests and field data (Rostami, 2016). An introduction to these model concepts was summarized in Rostami(2016),and a brief review and discussion on capabilities of some of the commonly used models, such as the Colorado School of Mines(CSM) model (Rostami, 1997) and the Norwegian University of Science and Technology (NTNU) model (Bruland, 1998a, b) were given in Farrokh et al. (2012). Specific rock mass boreability index(SRMBI)(Gong and Zhao,2009)is an excellent compendium of TBM performance prediction models(Rostami,2016).The SRMBI model is based on rock breakage process analysis and further developed with the database of hard rock tunnels.The SRMBI index integrates the cutting force analysis with the correlation analysis between actual penetration rate and rock mass parameters, and thus it is used for prediction of penetration rate and evaluation of rock mass boreability.However,most of the available models fail to take the in situ stress into account.In this paper,SRMBI model is improved by introducing the relationship between the thrust force and in situ stress, and then it is used to predict the TBM performance-for the URL project.

Based on the brief review of available approaches to investigate the rock mass boreability,the method of full-scale cutting test will be selected for the URL project due to the reliable result and its possibly direct application to engineering as well as availability of rock samples; moreover, the modified SRMBI model is used as a supplementary method in TBM performance prediction. Besides the rock mass boreability issue, investigations on the cutter consumption and other challenges associated with the special layout of URL project are also conducted.The results show that TBM method is feasible for the URL project. Finally, some suggestions for this project are also proposed.

2. Technical issues of TBM tunneling for Beishan URL excavation

2.1. Geological conditions of URL site

The Xinchang URL site is about 135 km away from the city of Jiayuguan in Northwest China(Fig.1).It is located in the middle of the Beishan area, the priority area for the HLW repository.Comprehensive investigations on site characterization have been conducted in Xinchang site during the past decade. Results show that the major rock types are granite and granodiorite,and there is no regional or active fault in the site area, as shown in Fig. 2. The rock mass shows very high integrity in terms of rock quality designation (RQD) data obtained from the borehole cores in the site, as demonstrated in Fig. 3, and the proportion of rock mass classified as “good” or “extremely good” is greater than 90%. The uniaxial compressive strength (UCS) of the rock mass is 110-235 MPa and the average value is 173 MPa, mostly in the range of 160-180 MPa; the tensile strength ranges in 6-15 MPa with average value of 11 MPa.Concerning the rock abrasiveness,Cerchar tests show that the Cerchar abrasivity index(CAI)value is 4.3-5.3,which is classified as very high abrasiveness. Fig. 4 shows the typical mineralogical compositions of Xinchang granite.It is mainly composed of 40%-45%plagioclase(Pl),25%-30%quartz(Qz),15%-20%microcline(Mc),and 5%-10%biotite(Bt).Field tests within the depth from ground to -600 m show that the in situ stress is low with the maximum value of 20 MPa, and the permeability of fractured rock is as low as 1 ×10-8-1 ×10-11m/s. More information related to the geological conditions of the Xinchang URL site can be found in Wang et al. (2018).

2.2. URL layout design

Fig.1. Geographical location of the Beishan area and a photo of typical topography of Xinchang URL site.

Fig. 2. Geological map of the Xinchang site (after Wang et al., 2018).

According to the URL functions and geological conditions, the preliminary design scheme of Beishan URL is proposed with the main structure of one access ramp, three shafts and two experimental levels, as shown in Fig. 5. The access ramp will be mainly used for material transport, with the tunnel diameter of 7 m,length of 7970 m,curve diameter of 400 m and the maximum inclination of 1:10.For the three shafts,one is personnel shaft with diameter of 6 m, and the other two are ventilation shafts with diameter of 3 m. Due to the limited disturbance to rock mass and preferable advance rate,the TBM method will be suggested for the ramp excavation.

2.3. Technical issues of TBM tunneling

2.3.1. Technical challenges related to geological conditions

Rockburst, jamming and large inflow are the typical unfavorable conditions for TBM tunneling. However, these risks may hardly be experienced in Beishan URL site due to the good geological conditions of low in situ stress,hard rock mass with no regional fault, high RQD and extremely low permeability. Generally, the highly integrated rock mass at the Xinchang site are favorable for underground excavation stability.However,the very high UCS makes it difficult for TBM tunneling effectively, and the high abrasiveness will lead to fast cutter abrasion and subsequently, the frequent cutter replacement will raise both in construction cost and construction time. Thus, considering the sitespecific geological conditions, the potential technical challenges for TBM are “extremely hard rock with high integrity” and “high abrasiveness”.

2.3.2. Technical challenges related to URL layout

Fig. 3. (a) RQD classification of borehole cores and (b) borehole cores sampled from borehole BS32 at Xinchang URL site (after Wang et al., 2018).

Fig. 4. (a) Photo of parts of CAI test samples and (b) micrograph of the mineral components.

Fig. 5. Preliminary design of the Beishan URL (after Wang et al., 2018).

When the TBM method is used to excavate the spiral ramp(Fig.5),some technical issues may be experienced:(i)for the curve diameter of 400 m,it may be difficult for TBM operation and muck transportation in such tight curve route; (ii) for the ramp descending at a gradient of 1:10, the feasibility for TBM operation and the efficiency of water drainage may be challengeable;and(iii)for the high inclination along the ramp with long distance of nearly 8 km,it may be difficult to transport the material effectively.Thus,the technical issues for TBM tunneling in the URL can be TBM operation with small curve radius and high slope inclination,muck transportation, long-distance water drainage and material transportation.

In total, seven technical issues have been identified for TBM tunneling related to the Beishan URL:

(1) Extremely hard rock with high integrity;

(2) High abrasiveness;

(3) Small curve radius for TBM operation;

(4) High inclination for TBM operation;

(5) Muck transportation with small curve radius;

(6) Long-distance water drainage with high inclination; and

(7) Long-distance material transportation with high inclination.

Based on analyses of experts consulting, discussion with TBM manufactures, and site investigation, it shows that the most challenging issue is “extremely hard rock with high integrity”. Thus,analysis of each of the challenges will be discussed in the following sections, with focus on tunneling in extremely hard rock mass. It should be noted that the discussion is mainly depending on the investigations performed in the project proposal phase,and further detailed design and analysis will be conducted in the following stage.

3. Feasibility analysis of TBM excavation of the URL ramp

3.1. Rock mass boreability and advance rate

For the extremely hard rock with high integrity in the URL ramp,one of the key parameters is rock mass boreability,which is highly related to the advance rate.If the rock mass boreability is poor,then the advance rate would be significantly slow,and consequently the TBM will not suitable for this project. Generally, the index “penetration rate” is an essential parameter to evaluate the rock mass boreability and to predict TBM advance rate.The term“penetration rate” is defined as the excavation distance of TBM per boring unit time.It is an indicator reflecting the interaction between TBM and rock mass, and it is also used for suitability evaluation of TBM method in a project. Penetration per revolution (PRev), or penetration (P) for short, is the excavation distance per cutterhead revolution, which is also defined as the basic penetration rate(Bruland,1998a). The value of the maximum penetration depends on ground conditions and machine parameters such as rock properties (UCS and Brazilian tensile strength (BTS)), in situ stress,cutter spacing, cutter dimension (size and tip width), thrust (represented by cutter force),and others(Bruland,1998a;Barton,2000;Gertsch et al., 2007;Yagiz, 2008; Gong and Zhao, 2009; Jing et al.,2019), as shown in Fig. 6. In this paper, research on rock mass boreability and advance rate for the URL project is conducted by cutting test,prediction model and comparison with similar project.

3.1.1. Penetration prediction by cutting test

The mechanical rock breakage experimental platform (Gong et al., 2016b) (Fig. 7) is used in this research. Two different kinds of granite samples are taken from Xinchang URL site (Fig. 8), with UCSs of 198 MPa and 172 MPa, respectively (Table 1). The dimensions of the rock samples are 780 mm × (765-780 mm) × (400-450) mm. Considering the tunnel diameter and rock strength,19-in cutters (1 in = 2.54 mm) with three kinds of cutter tip width (CTW), i.e. 8 mm,16 mm and 19 mm, are used in the cutting to study the influence of cutter shape on cutter force.Six levels of penetration,i.e.1 mm,2 mm,2.5 mm,3 mm,3.5 mm and 4 mm,are conducted for each cutter.The confining stress is 15 MPa,which approximately represents the in situ stress level at the depth of -300 m at Xinchang site.

Fig. 9 shows the relationship between the penetration and required cutter force for different kinds of rock samples and CTWs obtained by the tests. Comparing the test results of rock samples with the same CTW of 8 mm but different UCSs of 198 MPa and 172 MPa (Fig. 9), it demonstrates that for the narrower CTW, UCS has smaller influence on the cutter forces required to reach the same penetration.While for the same UCS of 172 MPa,the required cutter force to reach the same penetration with CTW of 16 mm is much higher than that with CTW of 8 mm.The results indicate that the cutter dimension has a significant influence on the TBM penetration.

Fig. 6. Factors influencing the TBM penetration.

Fig. 7. Mechanical rock breakage experimental platform.

Based on the rock fragmentation mechanism induced by TBM cutter,two aspects should be considered when analyzing whether high penetration can be reached or not during TBM tunneling:

(1) Difficulty levels for cutter crushing into rock, i.e. the cutter force required to reach low penetration. For example, as shown in Fig.9,for P=1 mm,the required cutter force with CTW of 19 mm is 68.2%higher than that with CTW of 8 mm,which means that much more resistance will be encountered for the cutter with CTW of 19 mm to crush into rock before chipping formation stage.

(2) Increasing rate of cutter force with increase of penetration that can be evaluated by the power function of fitting curve.The results illustrated in Fig. 9 show that the functions of fitting curves gradually change from power function to linear function,and the function power decreases from 3.3329 to 1 as the CTW increases from 8 mm to 19 mm. The tendency patterns obtained by cutting test show that it will be much difficult for the cutter with CTW of 19 mm to improve the penetration notably.

Based on above-mentioned two aspects, it is difficult for the cutter with CTW of 19 mm to reach high penetration. In addition,the cutter with CTW of 8 mm is not suitable for the URL project due to the extremely frequent cutter replacement in terms of the abrasive resistant property, which will lead to significant delay of tunneling advance.Therefore,the test result of the CTW of 16 mm will be used to predict the TBM advance rate at Xinchang site(see Section 3.1.5). According to the relationship between the penetration and cutter force for CTW of 16 mm with the function of y = 8 × 10-5x1.9122(R2= 0.9977) (Fig. 9), it can be calculated that the penetration can be up to 4.8 mm (the corresponding penetration rate is 4.8 mm per revolution)when using 19-in cutter with tip width of 16 mm, within the maximum force capacity of 315 kN.

Fig. 8. Photographs of (a) taking sample at URL site and (b) rock sample.

Table 1 Physico-mechanical parameters of rock samples and testing parameters of cutting test.

Fig. 9. Relationship between the penetration and cutter forces required for cutters of different shapes.

3.1.2. Penetration prediction by performance prediction models

SRMBI model(Gong and Zhao,2009)integrates the cutting force analysis with the correlation analysis between actual penetration rate and rock mass parameters, and could be improved by considering the influence of in situ stress. Besides, the NTNU model is used in penetration prediction as well for comparison purpose.

3.1.2.1. SRMBI model

SRMBI model is defined as the normalized cutter force (boreability index)at the penetration rate of 1 mm/rev.The index is not relevant to the TBM operation parameters (e.g. thrust force, revolution per minute (RPM), and torque) and avoids the effect of operational uncertainties on the rock mass boreability accordingly.It can be expressed by a power function of four parameters,i.e.rock UCS, rock brittleness index (Bi), joint spacing (Jv) and joint orientation(α):

where BI(1)is the SRMBI index; BI is the boreability index (the required cutter force) at the basic penetration rate; Bi is the brittleness index,which can be calculated as the ratio of UCS to tensile strength.

The in situ stress is not taken into account in this model because the database used for developing the model is mainly related to shallow tunnels. As the in situ stress is a significant factor that affects the TBM performance for deep tunnels(Innaurato et al.,2007;Ma et al., 2016; Pan et al., 2018), the SRMBI model should be improved by considering the effect of in situ stress. A highly correlated relationship between confining stress and cutter force for Beishan granite was obtained by linear cutting test (Ma et al.,2016), which can be used to improve the SRMBI model. As the confining stress (represented by the in situ stress) can be transferred to the parameter of “depth” based on the relationship between in situ stress and depth(Wang et al.,2018),the in situ stress can be substituted by the depth accordingly,as expressed by Eq.(3).Then, the cutter forces required for different penetrations at different depths can be calculated by Eq. (4).

where BI(1c)and BIcare respectively the SRMBI index and boreability index(the required cutter force)when considering the effect of in situ stress, and H is the tunnel depth.

In order to obtain the rock strength parameters(UCS and Bi)of Xinchang site, tests are performed on the rock samples collected from different depths of the boreholes at site, and the test results are listed in Table 2. As the rock mass of Xinchang site is of high integrity with little fractures,it is assumed that Jv= 0,and α=0°.By using the SRMBI model,rock boreability for Xinchang site can be calculated and the results are listed in Table 2. Fig. 10 shows the required cutter forces to reach different penetrations of 1 mm,2 mm, 3 mm, 4 mm and 5 mm, at different depths from ground to-560 m.The results demonstrate that the penetration of nearly 5 mm can be reached within the 19-in cutter’s capacity of 315 kN,when tunneling is within the depth of 400 m. For the depth at about-550 m level,the maximum penetration is 2 mm,due to the effect of high confining stress.The predicted result at the depth of 287 m is comparable with the results obtained from the cutting test under the stressed condition of the depth of 300 m (see Section 3.1.1).

3.1.2.2. NTNU model

NTNU model of penetration prediction is expressed as

where i0is the basic penetration rate (mm/rev), Mekvis the equivalent cutter thrust (kN/cutter), M1is the critical cutter thrust(kN/cutter) (necessary thrust to achieve 1 mm/rev), MBis the maximum gross average cutter thrust (kN/cutter), b is the penetration coefficient,kdis the correction factor of cutter diameter,kais the correction factor of cutter spacing, kekvis the equivalent fracturing factor,ks-totis the total fracturing factor,kDRIis the correction factor of drilling rate index (DRI) of the rock, and kporis the correction factor of rock porosity.

As indicated in the NTNU model, the related parameters for penetration prediction of URL project can be obtained as follows:

(1) If the 19-in cutter with the spacing (S) of 80 mm is selected for URL project,then M1=189 kN/cutter,MB=280 kN/cutter,kd= 1, and ka= 0.93.

(2) For the URL geological condition, i.e. rock with no or few fractures, ks-tot= 0.36.

(3) For granite in the URL, kDRI= 1.03, and its porosity is 0.35%.When the porosity is less than 2%, kpor= 1. Thus we have kekv= 0.37 according to Eq. (7).

(4) The parameter b is related to kekv, and here b = 2.55 for kekv= 0.37.

According to Eqs.(5)and(6),we have i0=2.3 mm/rev,which is the average penetration. If Mekvis set as the maximum cutter capacity of 315 kN, then i0= 3.7 mm/rev, which stands for the maximum penetration rate.

Compared with the cutting test result of 4.8 mm/rev in Section 3.1.1, it is found that the maximum penetration predicted by the NTNU model is relatively lower due to its underestimation, while the SRMBI result is relative closer to the cutting test result under similar condition.Generally,the result obtained by the cutting test is more reliable than the NTNU model evaluation for Beishan URL project, as the tested rock sample is taken from the site,and it reflects the realistic interaction between machine and rock in the field.

Fig.10. The required cutter force for each penetration of 1-5 mm at different depths.

It should be noted that the maximum penetration (or penetration rate) can be obtained by cutting test and theoretical calculation, but TBM does not always work with the maximum penetration during the whole boring process. Therefore, it is required to have the average penetration which stands for the real state of TBM tunneling.The average penetration can be obtained by similar project comparison as discussed in the next section.

3.1.3. Comparison of similar project with TBM straight driving

The new Ulriken tunnel in Norway is a case of great interest for the Beishan URL project, as rock mass conditions are similar at these two sites. The new Ulriken tunnel is 7.8 km long, part of the“Bergensbanen”,-the railway line-between Oslo and Bergen.It is the first time to use TBM in railway tunnel construction in Norway,which presents a milestone in Norwegian tunneling (Ramoni,2016).

The new Ulriken tunnel started to be bored in December 2015 and completed in August 2017, by a gripper TBM with a boring diameter of 9.3 m. The TBM runs through hard rock mass with a length of 6894 m (Augen gneiss, and granite gneiss). The gripper TBM is designed with a total thrust force of 27 MN and a rotation speed of 0-6.4 rpm. Sixty-two 19-in cutters with tip width of 16 mm are installed on cutterhead, and the face cutter spacing is 75 mm, as shown in Fig.11a. It is recorded that for the intact rock mass (see Fig. 11b) with high UCS of 180-200 MPa, the average penetration could be up to 2.5 mm even under the limited operation condition, and the advance rate of 80-110 m per week is experienced. The excellent driving performance of new Ulriken tunnel demonstrates the strong capability of TBM technology for the extremely hard rock,and also provides reference for XinchangURL project. By comparison, the average penetration rate is set as 2.5 mm/rev for TBM driving in straight and horizontal routes.

Table 2 Prediction of rock boreability for Xinchang site.

Fig.11. Photos of (a) TBM cutterhead for Ulriken tunnel and (b) rock mass with high integrity of new Ulriken tunnel (by Herrenknecht AG).

3.1.4. Influence of curved and inclined driving on TBM penetration

For the Beishan URL ramp, the curved and inclined length is 4.4 km. In fact, inclined driving has slight impact on penetration,while for curved driving with radius of 400 m,it has a minor impact on the penetration, as shown in the following discussion. When TBM drives in curves,the location where the maximum penetration occurs at the tunnel face is along the outer arc of tunnel alignment.Thus, for each 90°curved excavation, the impact factor kccan be obtained by comparing the lengths of outer arc and axial line:

where R is the radius of axial line, and D is the tunnel diameter.

For Beishan project, R = 400 m and D = 7 m, then we have kc= 0.875%. If the average penetration rate in straight route is 2.5 mm/rev,then the value in curved route with the radius of 400 m is 2.502 mm/rev, which indicates that the difference is so slight that the impact of curved driving on the penetration rate can be negligible.

3.1.5. TBM utilization

TBM utilization is a comprehensive factor that provides insight into the suitability of TBM operating in rock mass, construction management and adverse incidents. It involves a number of uncertainties, e.g. UCS, CAI, RQD, water condition, tunnel diameter,and contractor/crew experience (Barton, 2000; Bieniawski et al.,2008; Farrokh, 2013; Rostami, 2016). As utilization is defined as the percentage of available shift time during which excavation or rock penetration occurs(Nelson,1983),it is determined by the TBM downtime.It also shows that with the increase of rock mass quality,TBM utilization increases (McFeat-Smith,1999). In the excavation of sub-horizontal rock masses,utilization ranges between 15%and 65% (Laughton, 1998). While according to Rostami (2016), the typical utilization generally ranges from 5% (for the condition of very difficult ground and poor site management) to 55% (for the condition of moderately strong rock mass driven by gripper TBM).

For Beishan URL project, the demand for rock support will be much less than that for common projects due to the high quality of hard rock mass,thus it will notably reduce the TBM downtime and increase TBM utilization. NTNU model of machine utilization includes only small amount of rock support work,thus it seems to be applicable for Beishan URL project.

The NTNU model of machine utilization is expressed as

where u is the daily machine utilization; Tbis the boring time (h/km),and Tb=1000/In,in which Inis the net penetration rate(m/h);Ttis the time for regripping (h/km), and Tt= 1000ttak/(60ls), in which ttakis the time per regripping,and lsis the stroke length;Tcis the time for cutter change and inspection,and Tc=1000tc/(60HhIn),in which tcis the time for each cutter change,and Hhis the average cutter ring life time (h); TTBMand Tbakare the repair and service times(h/km)of the TBM and the backup,respectively;and Tais the time for miscellaneous activities,including normal rock support in good condition, waiting for transport, surveying, installation and maintenance of tracks and cables of water, ventilation and electricity, and other activities.

For Beishan URL project,if the average basic penetration rate is 2.5 mm/rev, and RPM = 6 rev/min, then In= 0.9 m/h, and Tb=1111.1 h/km;since ttak=6 min and ls=1.2 m,then Tt=83.3 h/km; as tc= 45 min and Hh= 3.14 h for 19-in cutter, we have Tc=265.4 h;as In=0.9 m/h,we have TTBM=100 h/km,Tbak=65 h/km,and Ta=150 h/km,as given by the NTNU model.Thus,the daily machine utilization u is calculated as 62.6%,which can be used for straight and horizontal driving.

For curved and inclined alignments, the u value should be reduced according to Rostami(2016).In this study,3%reduction is adopted for curved section with R = 400 m and 20% reduction for tunnel section with inclination of 10%; accordingly, the u value is modified as 39.6%.

3.1.6. Prediction of TBM advance rate for Beishan URL

Based on the above investigations, the TBM advance rate for Beishan URL project can be calculated considering the following parameters:

(1) The maximum penetration rate of 4.8 mm/rev, and the average penetration rate of 2.5 mm/rev;

(2) Average UCS of 173 MPa;

(3) 19-in cutter with tip width of 16 mm;

(4) Cutter spacing of 75-80 mm;

(5) TBM operation in 6 rev/min;

(6) Daily machine utilization of 62.6%for straight and horizontal routes,and 39.6% for curved and inclined routes;and

(7) 24 d/month on average.

The advance rate is predicted as 325 m/month in straight and horizontal routes,and 205 m/month in curved and inclined routes.The results show that,compared with the drill-and-blast method of which the advance rate is predicted as 150 m/month,TBM method could be a good choice for URL construction.

3.2. Cutterlife prediction

As cutter consumption is highly related to construction cost and time,it is necessary to conduct cutter life prediction for URL project.The NTNU method for cutter life prediction has been used extensively and is a reliable method to predict the wear of disk cutters(Oggeri and Oreste, 2012), which can be expressed as follows(Bruland,1998a, b):

where Hmand Hfare the average cutter ring lives in meter and solid cubic meter,respectively;H0is the basic average cutter ring life;for 19-in disk, H0≈-0.1425CLI2+8.305CLI+1.05 when the cutter life index(CLI)is less than 30;NTBMis the real number of disks on the TBM cutterhead;RPM is the rotation speed of cutterhead(rev/min);DTBMis the diameter of TBM cutterhead;kD(TBM),kQ,kRPMand kNare the correction factors related to the diameter of TBM cutterhead,quartz content of rock, rotation speed of TBM cutterhead, and number of cutters,respectively,which can be obtained as follows:

where qz is the quartz content of rock;N0is the typical number of disks on the TBM cutterhead,and N0=DTBM/(2S).Here N0=NTBM.

On the basis of NTNU model and considering CLI=8 for granite,DTBM=7 m,19-in cutter with cutter spacing S=80 mm,qz=27.5%,and RPM=6 rev/min,we obtain NTBM=N0=44,kD(TBM)=1.5982,kQ=1.2431,kRPM=1.1905,and kN=1.The average cutter ring life in meter (Hm) and in solid cubic meter (Hf) can be calculated for different penetrations,as listed in Table 3 and Fig.12.As shown in Table 3, excavations of 2.84 m along the tunnel and 109.29 m3of rock muck can be reached for each cutter ring.Generally,the cutter life is relatively low but still acceptable under such limited boreability condition. In fact, the situation for cutter consumption is better for the Beishan URL project, when compared with that ofYinhanjiwei Water Supply Project in China, where excavations of 1.19 m along the tunnel and 59 m3of rock muck for each 20-in cutter ring were recorded under the extremely hard and abrasive rock conditions where qz=43.6%-92.6%,and CAI=4.65-5.71(see Fig.13),for the first 2320 m driving with boring diameter of 8.02 m.

Table 3 Cutter ring life prediction for different penetrations (D = 7 m).

On the other hand, some measurements are proposed to improve the cutter life, such as optimization of cutter shape, condition of cutter material as well as optimization of cutter layout(Smading, 2017; Geng et al., 2018). In addition, it is also highly recommended to use abrasive-protection technology on the cutterhead during the TBM design.

3.3. Small curve radius for TBM operation

Generally,TBM can be in good operation if curve radius is larger than 500 m, according to the constructor experience. For some projects, specific design of TBM machine can be proposed to meet the requirement of small curve radius of 200-250 m.Table 4 shows the reference list of projects with small curve radius constructed by gripper TBMs.Several projects,such as the Tunnel Maurice-Lamaire with tunnel diameter of 6 m and curve diameter of 250 m in France,and the New York Second Avenue Subway in USA with tunnel diameter of 6.717 m and curve diameter of 185 m,have shown good performance with small curve radius by gripper TBM. Thus, it is possible for TBM regular operation with curve radius of 400 m for the Beishan URL project.

In addition, special designs of TBM for Beishan URL project are suggested to better cope with excavation in curves as follows:

3.3.1. Special design of main machine

It is suggested to shorten the length of main machine due to less demand for support system.A short stroke of 1.2 m and telescopic roof shield can be used to facilitate the excavation in curves. The main beam structure should be tailor-designed to enable the adjustment of the advance direction in a timely manner during boring process. As such, for the condition with curve radius of 400 m,and the main machine with the length of 17 m and stroke of 1.2 m,the deviation of hydraulic cylinder of gripper d is 165 mm for each stroke boring in curve.Plus the extension length of 335 mm in straight driving, the total extension length is 500 mm, which is within the cylinder stroke length of 635 mm;therefore,it is feasible for TBM operation in curve with radius of 400 m.

The value of d for each stroke boring can be obtained by process simulation with CAD software, or approximately calculated as follows:

where d denotes the deviation of center point of gripper system from the tunnel axis, R is the curve radius, Lgis the distance between the cutterhead center to gripper center,and Lsis the length of one stroke.

3.3.2. Special design of main bearing

The work status of main bearing under continuously curved condition will be informed to the manufacture,and tailored design of main bearing is required for this project.

3.3.3. Abrasive protection and wear monitoring of cutterhead

Abrasive-protection design should consider cutterhead with anti-abrasion material and abrasive-protection blocks, and the monitoring system will be utilized to monitor the wear status of cutterhead and cutter.The worn cutter should be replaced timely in order to avoid the damage to steel structure of the cutterhead.

Fig.12. Predicted relationship between average cutter ring lives Hm and Hf and penetration rate.

Fig.13. (a) Tunnel face, (b) surrounding rock mass, and (c) worn cutter.

3.4. High inclination for TBM operation

It is generally accepted that the TBM is applicable to the tunnel projects with the slope no more than 10.5% (6°), which would reduce the TBM utilization but be workable for efficient cutting,muck removal, and TBM stability. The designed inclination of Beishan URL is-10%(5.72°),thus it is feasible for the regular TBM operation. Experiences with the similar conditions have already been reported in previous TBM projects.For instance,2733 m-long inclined shaft with 9.5% downhill slope of Pulianta coal mine project in China was successfully excavated by TBM with cutterhead diameter of 7.62 m and total machine length of 165 m in 2015.For the Beishan project, tailored design such as inclined stairways and platforms is suggested to enable the safety of crew on the TBM.

3.5. Muck transportation with small radius

Generally,there are two ways to transport excavation muck,i.e.by conveyor belt and by truck. Regarding the issues on personal safety,transportation efficiency and environment protection,using conveyor belt remains the pre-selected choice for this project.However,it will be quite difficult for the common conveyor belt to transport the rock muck efficiently due to the multi-curve and limited radius of the URL ramp. Special design is strongly recommended for muck conveyor belt, and alternative design schemes are proposed such as multi-stage conveyers, curve-belt conveyor,and inclining belt conveyor. Fig. 14 shows the inclining belt conveyor which has been successfully used in some projects.Research on development of several kinds of conveyor systems with better performance is in progress.

For Beishan project, nevertheless, the requirement for the transportation capacity of belt conveyor system is not strong as the penetration rate cannot be too fast due to extremely hard rock mass, and normally the excavated muck volume is less than 200,000 kg/h. It is suggested to use the inclined conveyor system with multiple drives to keep pace with the TBM boring.Additional drives will be installed before each curve to make the belt force reduced and to ensure safe and smooth transportation of muck.

3.6. Long-distance water drainage with high inclination

The risk of large inflow should be stressed for downwardinclined tunneling. However, this kind of risk may not be.encountered in Beishan project. As introduced in Section 2.1, there is no regional fault across Xinchang site and the rock mass is featured by extremely low permeability, thus it will be unlikely to experience large inflow discharged from the rock mass during tunneling.Generally, the main demand for dewatering comes from TBM working for cutter cooling and machine cleaning. According to theoretical analysis, the maximum water consumption for TBM working is 40 m3/h. Considering the potential underground water seepage and unforeseen inflow from the rock mass, the discharge value of 100 m3/h is used for dewatering design. In addition, with respect to the large slope of 10%from ground to-560 m level and long distance of nearly 8 km, the drainage system with 3-stage high-lift boosters is recommended for the URL ramp.

3.7. Long-distance material transportation with high inclination

Generally, there are two ways for material transportation for TBM excavation method, i.e. by rail-based transportation system and by trackless vehicles.In order to ensure safety and efficiency of transportation on the large-gradient slopes of URL ramp, it is suggested to use the trackless vehicles instead of rail vehicles,as one ofthe noticeable advantages of the trackless vehicle is the excellent mobility even on steep ascents or descents (see Fig.15). The recommended type of trackless vehicles for Beishan project is rubbertyred vehicle, which has load capacities of 10-200 t in line with different needs on site.

Table 4 Reference list of projects with small curve radius constructed by TBMs (partial information provided by Herrenknecht AG).

Fig.14. Inclined belt conveyor designed for small curve radius (by Herrenknecht AG).

Fig. 15. (a) Photo of rubber-tyred vehicle and (b) its comparison of manipulation performance with train under condition of inclined slope.

4. Conclusions

Beishan URL, the first URL for geological disposal of HLW in China, will be constructed in 2021. This paper presents the preliminary study on the technical feasibility of using TBM for excavation of the URL in granite.

“Extremely hard rock with high integrity” is regarded as the most challenging issue for this project.Investigations on rock mass boreability for TBM tunneling have been conducted by cutting test,modified SRMBI prediction model,and similar project comparison.The results show that a maximum penetration rate of nearly 5 mm/rev can be reached by using the TBM with 19-in cutter. The 19-in cutter with tip width of 16 mm is suggested for this project due to its better performance in rock fragmentation and cutter abrasion resistance. Machine utilization based on NTNU model and expert experience has been estimated, and the results show the daily machine utilization of 62.6%for straight and horizontal routes,and 39.6% for curved and inclined routes. Based on the results of rock mass boreability and machine utilization, the advance rate is predicted as 325 m/month in straight and horizontal routes, and 205 m/month in curved and inclined routes.

Regarding the cutter consumption, it shows that each 19-in cutter ring could finish the excavation of 2.84 m along the tunnel and 109.29 m3of rock muck on average, according to the NTNU model. The cutter life is relatively low for Beishan URL project but still acceptable under such limited boreability condition.

Investigations on the special layout of the URL ramp have demonstrated that it is technically feasible to cope with this issue.Specifically, it is possible for regular TBM operation with curve radius of 400 m and high inclination of -10%, which is properly supported by theoretical analysis and available references. For muck transportation, special design of inclined belt conveyor system with multiple drives is required to improve the transportation efficiency. For long-distance dewatering with high inclination,multi-stage pump system can be installed for URL project. For material transportation, it is suggested to use the rubber-tyred vehicle instead of rail-based vehicle due to the large inclination of URL ramp.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

China Atomic Energy Authority is thanked for its financial support for this project.The authors would like to acknowledge China Railway Engineering Equipment Group Co., Ltd., China Railway Construction Heavy Industry Co., Ltd., Herrenknecht AG, China Railway 18th Bureau Group Co., Ltd., China Railway Tunnel Group Co., Ltd., and Liaoning Censcience Industry Co., Ltd. for their technical support on this research. The valuable comments by two reviewers are appreciated as well.