Tunnelling and Underground Space Technology 57 (2016) 173182
Contents lists available at ScienceDirect
Tunnelling and Underground Space Technology
journal homepage: www.elsevier .com/ locate/ tust
Performance prediction of hard rock Tunnel Boring Machines (TBMs)in difficult ground
http://dx.doi.org/10.1016/j.tust.2016.01.0090886-7798/ 2016 Published by Elsevier Ltd.
E-mail address: Rostami@psu.edu
Jamal RostamiDepartment of Energy and Mineral Engineering, The Pennsylvania State University, State College, PA, USA
a r t i c l e i n f o a b s t r a c t
Article history:Received 15 October 2015Accepted 11 January 2016Available online 2 March 2016
Keywords:Performance predictionRate of penetration (ROP)Advance rate (AR)Utilization (U)Tunnel Boring Machine (TBM)Rock tunnelling
Performance prediction of TBMs is an essential part of project scheduling and cost estimation. This pro-cess involves a good understanding of the complexities in the site geology, machine specification, and sitemanagement. Various approaches have been used over the years to estimate TBM performance in a givenground condition, many of them were successful and within an acceptable range, while some missing theactual machine performance by a notable margin. Experience shows that the best approach for TBM per-formance prediction is to use various models to examine the range of estimated machine penetration andadvance rates and choose a rate that best represents the working conditions that is closest to the settingof the model used for the estimation. This allows the engineers to avoid surprises and to identify theparameters that could dominate machine performance in each case. This paper reviews the existing mod-els for performance prediction of TBMs and some of the ongoing research on developing better models forimproved accuracy of performance estimate and increasing TBM utilization.
2016 Published by Elsevier Ltd.
Ever since Tunnel Boring Machines (TBM) were introduced in1950s, design engineers and contractors have been preoccupiedwith the question of accurately estimating machine performancefor a given project setting. While much strides have been madein this field for simpler cases of using TBM, recent expansions inthe use of these machines in more complex geological settingsand concepts of a Universal Machine have introduced newsources of complexities and uncertainties. This has led to the useof probabilistic approaches to performance prediction where theadvance rate (AR) of a TBM in a given geology is a range, insteadof a simple number and the completion time for a project is simi-larly a range, instead of a certain number of days/weeks, whencombining the probabilistic performance of the machine in eachsegment of a project. The sources of variation in performance ofthe machine include geological, machine operation, and site man-agement parameters, where each set of parameters follow relatedhistogram of varying input parameters. This approach has beenoriginally used in planning of the Trans-Alpine tunnels in Europein 1990s (Einstein et al., 1992; Einstein, 2001), and is graduallyadopted in other projects as part of the risk management scheme.
In general performance estimation for a TBM refers to estima-tion of certain parameters which include:
Rate of penetration (ROP) which is also referred to as penetra-tion rate (PR) and often expressed in m/h and refers to the linearfootage of excavation per unit time, when machine engages theground and is in production.
Utilization rate (U), expressed in percent (%) and representingthe ratio of boring time to the total time. Total time could referto the number of hours worked per work days, boring days, orcalendar days.
Advance rate (AR), which is the amount of daily advanceexpressed in m/day and is calculated as:
AR ROP U Ns Sh; 1
with Ns being the number of shifts per day and Sh being thenumber of hours per shift.Another parameter that is often cited as part of performance
prediction is cutter life. This parameter is typically expressed interms of average cutter life in hours, meter travelled on the face,cutters per meters of tunnel, or cutters per cubic meter ofexcavated rock. While cutter life is a cost issue and is not directlyrelated to the parameters listed above, TBM experts are expectedto offer an estimate for this item. Obviously, increased cutterconsumption will impact maintenance time and machineutilization, but this item will not be discussed in current paper.Overall, this paper is not prepared to prove any particularformula/model or be argumentative, rather as an overview of the
174 J. Rostami / Tunnelling and Underground Space Technology 57 (2016) 173182
personal experience of the author for over two decades of researchon this topic. As such, it will have limited literature survey, calcu-lations, graphs, and tables, instead, it contains highlights of thechallenges and opportunities related to TBM performance predic-tion. Naturally, there would be many that agree or disagree withthe points made in the paper and the author would welcomeand respect any and all discussions that could help clarify theissues and offer practical solutions to current and upcoming chal-lenges. Following is a brief overview of the various types of machi-nes and methodologies for performance estimation for eachcategory. This will be followed by discussion of models for estima-tion of ROP and Utilization, and ending with a brief review of chal-lenges related to performance prediction of TBM in difficultgrounds.
1.1. Machine types
Various types of TBMs are available in the industry and differentsystems are used to classify them. Some guidelines for machineselection has been offered by ITA or various tunnelling societiesaround the world (ITA-AITES, 2000; EFNARC, 2005). But as machi-nes are used in more complex geologies, sometimes there are sce-narios where various ground conditions are encountered along thetunnel alignment and as such, clear cut classifications cannot beapplied. This is the basis for development of concepts for hybridmachines, which could be between rock and soil machines, as wellas shield and gripper TBMs. One can classify machines to softground and hard rock. The former machines are always within ashield, protecting the work environment from collapses in thewalls and the face. In soft ground, cutting the face is done by dragtype tools such as scrapers and normally not an issue. Rock TBMsrepresent cases where the face comprises rock (full or partial)and requires specialty tools, commonly disc cutters, to break it intomanageable size pieces. Meanwhile, one can imagine soft groundmachines that have to excavate ground containing boulders ormixed face of soil and rock, while rock machines can run into faultzones filled with gauge and running ground. Thus, complex geolog-ical settings and mixed ground conditions require a delicate studyof machine selection and related performance issues to balance thecost and risks of operating selected machines in given ground.
That said, TBMs can be classified into two general categories ofOpen and Shielded machines. Open TBMs are recommended forrock excavation where ground is generally good and stable bothat the face and walls. These machines can tolerate some instabili-ties in the ground and short reaches of unstable ground where spe-cial provisions can be used for passing through bad ground.Shielded TBMs are classified as Single or double shield (SS or DS).DS machines have become fairly popular in many rock jobs dueto their flexibility to work in good or bad conditions but they areno match to high ground water pressures or water inflow, nor forsqueezing grounds. Single shield units are classified into open ver-sus pressurized face machines. Single shields need to push off tun-nel linings to propel forward. In dry grounds or grounds wheredrainage of groundwater can be handled within the tunnels, openface SS TBMs are used. In case of high water pressure at the face orwhere drainage of groundwater is restricted due to contractrequirements, pressurized face machines are used. These TBMshave provisions to remove the excavated muck while maintainingface pressure and preventing water from being removed from theground. Two primary methods available today for pressurizedshield tunnelling include earth pressure balance (EPB) and slurrysystems. One can find machines of each type with scrapers or fea-turing disc cutters for dealing with rock at the face. Fig. 1 is a sim-ple schematic classification of various TBMs used in the industry.
1.2. Basic philosophy of machine performance
Basic concepts for operating TBMs are different between softground pressurized face machines and those of open face or rockmachines. In soft ground tunnelling, cutting the ground is rela-tively easy and often not an issue unless rock is present at the faceas boulders or as part of the face. The main concern in this softground applications is balancing the flow of muck, in particular,assuring that the amount of soil removed from the face is the sameas the volume of soil displaced by machine advance. This is toavoid ground settlement due to over-excavation. As such, the pen-etration rate of the soft ground machines is often fixed to a con-stant value during the stroke and the flow of muck is closelymonitored to maintain face pressure by balancing the volumeremoved. This fixed rate varies from 100120 mm/min in small3 m diameter to 3040 mm in large 15 m diameter TBMs. Evenif there is full face of rock to be mined, th