Available online at www.sciencedirect.

com
Tunnelling and
Underground Space
Technology
incorporating Trenchless
Technology Research
Tunnelling and Underground Space Technology 23 (2008) 308–317
www.elsevier.com/locate/tust

Soil conditioning of sand for EPB applications: A laboratory research

Raffaele Vinai a, Claudio Oggeri a, Daniele Peila a,b,*

a
Department of Land, Environment and Geoengineering, Tunnelling and Underground Space Center, Politecnico di Torino,
Corso Duca degli Abruzzi 24, 10129 Torino, Italy
b
IGAG-CNR, Politecnico di Torino, Italy

Received 10 April 2006; received in revised form 20 April 2007; accepted 30 April 2007
Available online 18 June 2007

Abstract

EPB tunnelling requires the application of soil conditioning to increase its field of applicability particularly for cohesionless soils.
Choosing the most suitable conditioning set for the various soils requires the use of a feasible laboratory test which can permit to define
the characteristics of the conditioned soils and provide measurable data.
A series of tests has been carried out using a laboratory screw conveyor device which was designed for this purpose and which sim-
ulates the extraction of the spoil from a pressure chamber in a similar way as in EPB tunnelling.
The tested soils were medium-grain sands with varying amounts of silt and the tested conditioned mixtures were obtained with dif-
ferent water contents and amounts of foam. A simple slump test was also used to analyze the global characteristics of the conditioned
soils.
The test has shown that the proposed laboratory procedure permits a quantitative comparison to be made between different condi-
tioning amounts and agents on the basis of measurable parameters.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Tunnelling; Mechanized excavation; Soil conditioning; Earth pressure balance shields; Slump test; Laboratory test

1. Introduction Soil conditioning is obtained through the injection of

foam, polymers, water and/or fillers ahead of the tunnel
Earth Pressure Balance shield tunnelling has successfully face and into the bulk chamber and along the screw con-
been adopted for urban tunnelling in recent years in very veyor. The aim of this procedure is:
different ground conditions and at present it can be consid-
ered the most commonly used mechanized tunnelling tech- – to transform the soil into a plastic and ‘‘pulpy’’ medium,
nology, even challenging the role of the Slurry Shield both that is able to properly transmit the pressure in the exca-
as far as machine size and the geomechanical field of appli- vation chamber and along the screw conveyor (Her-
cability are concerned. This increase in the number of tun- renknecht, 1994; Anagnostou and Kovari, 1996;
nels excavated by EPB machines has been possible thanks Nomoto et al., 1999),
on one hand, to the mechanical and electronic improve- – to reduce the permeability of the ground (and conse-
ments and on the other hand to the more effective use of quently to prevent water inflows, seepage destabilizing
additives to condition the soils. forces and consolidation-related surface settlements),
– to reduce frictional forces (and consequently tool and
*
Corresponding author. Address: Department of Land, Environment mechanical wear, as well as the required mechanical tor-
and Geoengineering, Politecnico di Torino, Corso Duca degli Abruzzi 24, que of the cutting head and of the screw conveyor),
10129 Torino, Italy. – to avoid adhesion effects of sticky clayey soils,
E-mail address: daniele.peila@polito.it (D. Peila).

0886-7798/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tust.2007.04.010
 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317 309

– to maintain the above-mentioned characteristics over few tests have already been indicated (Milligan, 2000, 2001;
time in order to allow safe control of the face stability EFNARC, 2005; Vinai, 2006) for conditioned soil quality
during the whole excavation step and when the machine assessment, some of which have been derived from com-
has to be stopped due to operational or technical reasons. mon geotechnical or concrete technology measurements:
foam penetration into the ground; the mixing test; the cone
Since the use of conditioning agents is relatively recent, penetration test; the permeability test; the compressibility
very little laboratory experience on their use has been accu- test; the adhesion test; the shear test; the slump test and
mulated and the conditioning design is often based on the other tests which are less frequently used such as: the flow
workers’ experience or according to a trial-and-error pro- cone test, the segregation test, the concrete mortar flow
cedure at the work site. table. A final quantification of the conditioning treatment
The evaluation of the properties of the conditioned soil is still far from being unequivocally defined. This is mainly
should instead be assessed using laboratory tests that due to the great number of parameters and also to the great
should allow the properties of the conditioned soil to be influence of the technological features of the EPB
understood and measured thus permitting the design of machines. From this point of view, only a test that is able
the optimal conditioning to be used in a specific environ- to simulate EPB operations and which involves both the
ment, and allowing an easy comparison of the various conditioning of the soil and the interaction between the soil
additives on the market. and the could truly be significant to understand the condi-
For this reason, a laboratory device that permits the tioned soil behaviour. Although this type of test is very
extraction of the conditioned soil from a pressurized tank complex, some large scale research tests have in fact been
by a screw conveyor has been designed, constructed and carried out:
tested. This is a tool able to check the ability of the soil
to control the pressure both in the bulk chamber and along – a 1/10 scale model for EPB excavation simulation was set
the screw conveyor. up as part of the work of a National French Research
In the first part of the research a systematic series of Programme (AFTES, 2001). The model consisted of a
slump tests was carried out with different percentages of 500 mm diameter cutting head, a conical excavation
conditioning foam and water contents to verify whether chamber, a inclined screw conveyor, a horizontal screw
it is possible to find some general rules to link these two conveyor, a cylindrical shield and four thrusting cylin-
parameters to the behaviour of the mix and whether this ders. The soil that had to be excavated (an incoherent
test can be used for a preliminary assessment of the and dry fine sand) was placed in a rigid 2 · 1.3 · 1.3 m
conditioning. box. A loading device, using ten air cushions, allowed
In the second part of the research, the laboratory device an additional overburden to be simulated. The system
with the screw conveyor was used to check: was instrumented with several monitoring transducers
for the driving parameters, the soil stresses and deforma-
– The transmission of the pressure from the top plate to tion control (Branque et al., 2003). The pressure in the
the bottom of the tank. This behaviour gives an indica- bulk chamber of the EPB model and in the surrounding
tion of the ability of the material to control and apply a ground, as well as the deformation and the settlements
confining pressure inside the bulk chamber. induced on the surface, were continuously monitored.
– The conditioning influence on the torque measured on In this test it was observed that two parameters are able
the screw conveyor which is an index of the workability to govern the driving of the boring machines:
of the material and its ability to reduce the torque values  the ratio between the mass and the material extracted
on both the EBP head and the screw conveyor. from the screw conveyor and the theoretical mass;
– How the material controls the pressure inside the screw  the pressure in the ground that has to be excavated.
conveyor. This is a relevant indication of the ability of Relevant indications on the optimal functioning regime
the screw conveyor to control the water pressure that of an EPB were given in this research but very little infor-
acts as an impervious plug. mation on the conditioning sets was derived:
– A laboratory model of a screw conveyor device and a
In order to reduce the number of studied parameters, full-scale EPB machine screw conveyor was used by
the tests were carried out on a reference sand obtained Bezuijen and Schaminée (2000, 2001) to study the
from a river quarry with rounded grains and with a fairly behaviour of conditioned sand soils.They observed that
regular granulometric curve but without gravel that is a the pressure was dissipated linearly along the conveyor,
typical sand that can be found in many urban areas. that the screw torque was approximately constant and
that the pressure at the end of the conveyor depended
2. Tests on conditioned soil proposed and described in on the opening of the gate valve at the discharge point:
technical literature – A full-scale EPB screw conveyor was used by Yoshikawa
(1996) and some tests were performed using plastic soil
Although there is very little technical agreement regard- and with different screw speeds. He observed that a linear
ing optimal test methods to characterize conditioned soil, a pressure gradient was present in the screw conveyor.
310 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317

– A laboratory screw conveyor apparatus, where condi- measuring points), the plate movement, the screw torque
tioned clay soils were extracted from a tank by a sub- and the extracted material weight were recorded during
horizontal screw, was used by Mair et al. (2003) and each test.
Merritt and Mair (2006) and their results have proved The tests were carried out on two different sands that
to be relevant for this type of soil. The laboratory device have the granulometric curves reported in Fig. 2. That typ-
was made of a pressurized tank which was connected to ical siliceous homogeneous sands were obtained from a
a 1 m long and 0.1 m diameter horizontal screw con- river alluvium, mixed with different percentages of water
veyor. The screw conveyor was instrumented in four sec- and foam obtained with FER (the ratio between the
tions, each with two load cells to measure the total obtained volume of foam and the volume of the generator
normal stress and the soil-casing interface shear stress fluid: water + foaming agent) of 16 which is an average
components and a pressure transducer to measure the value in tunnelling practice.
pore water pressure in the soil. Measurements of the
screw torque were also carried out. Tests can be per- 3.1. Slump tests
formed with this device using various pressures applied
to the soil over a wide range of screw speeds, with differ- Slump cone tests performed on fresh concrete have
ent discharge outlet conditions. already been used to provide a evaluation of conditioned
soil quality by several authors (Kuribashi et al., 1993;
From the described tests it is clear that a device which Peron and Marcheselli, 1994; Maidl, 1995; Quebaud,
simulates the extraction of the soil from a pressurized tank 1996; Quebaud et al., 1998; Bordachar and Nicolas, 1998;
with a screw conveyor would be a valuable tool for the def- Jancsecz et al., 1999; Williamson et al., 1999; Herrenknecht
inition of the soil conditioning parameters in EPB tunnel- and Rehm, 2002; Peña, 2003; EFNARC, 2005; Boone
ling and it could be used as a standardized test with the et al., 2005; Langmaack, 2000; Vinai et al., 2006). For
aim to permit an easy comparison between the various con- example Quebaud et al. (1998) suggested that a slump of
ditioning products and a preliminary choice of the optimal 12 cm is necessary to provide a mixture with the optimum
amount of conditioning agents. characteristics for a plastic flow in an EPB machine, while
Williamson et al. (1999) suggested 210 mm, Jancsecz et al.
3. Conducted tests (1999) 200–250 mm and Boone et al. (2005) 80–100 mm.
This type of test has the advantage of being simple to
The research involved a test programme on conditioned perform and of giving an overall index of the behaviour
granular cohesionless sand using both a slump test for a of the conditioned material.
preliminary setting of the conditioning agents and a screw Since the optimal cone fall values reported in previous
conveyor device that consists of a 800 mm high, 600 mm researches were directly linked with the job side where
high diameter cylinder tank where the conditioned ground the tests were carried out, in the present research a system-
is pressurized and then extracted by a 1500 mm long screw atic test program was carried out. In order to carry out this
conveyor which is positioned with an angle of 30 to the
horizontal (Fig. 1) (Vinai et al., 2006). The pressures inside
the tank (top and bottom) and along the screw case (three

Fig. 1. Experimental device developed for the EPB screw conveyor

extraction simulation. Numbers refer to the installed sensors: (1)
displacement wire transducer; (2) tank top pressure cell; (3) tank bottom Fig. 2. Grain size distribution for the tested sands. The continuous line
pressure cell; (4) screw conveyor pressure cell 1; (5) screw conveyor refers to the sand used for slump tests 1–12 and all the screw conveyor
pressure cell 2; (6) screw conveyor pressure cell 3; (7) torque meter on the extraction tests and the dotted line refers to the material used for slump
screw; (8) scale. tests 13–16.
 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317 311

Fig. 3. Quality assessment of the tested material after the slump cone test. Key: (a) not suitable [too stiff and dry]; (b) suitable; (c) not suitable [too fluid
and wet].

tests, a previously dried defined soil volume was mixed with were identified: too stiff and dry behaviour due to insuffi-
water and foam. The mix was then poured into three stan- cient water or foam content; too fluid and wet behaviour
dard slump cones. The three cones were lifted at the same due to excessive water or foam content and suitable behav-
time, the cone falls were recorded and the overall quality of iour of the mix where the ground behaves plastically
the mix was analyzed. (Fig. 3).
The quality of the mix was defined as ‘‘suitable’’, ‘‘bor-
derline’’ or ‘‘not suitable’’ according to the plastic behav- 3.1.1. Performed tests and obtained results
iour of the material as well as the presence/absence of The results of the test campaign conducted on the refer-
free liquid around the cone. Three main behaviour fields ence sand, summarized in Table 1 and in Fig. 4 indicate
that:
Table 1
Slump tests and obtained results on soil n. 1 – the conditioned soil only showed plastic behaviour for
Test Water FIR Slump Comments some combinations of the investigated parameters
content (%) (%) (mm) (water content and FIR);
1 9.6 20 146.7 Not suitable – Too stiff and dry – when the slump cone fall was within the 150–200 mm
2 11.2 40 148.3 Suitable – Very little liquid loss range, none or little liquid draining was observed;
3 9.4 60 231.7 Not suitable – Too fluid and wet – for certain water contents (about 5%), the conditioned
4 22.3 20 221.7 Not suitable – Too fluid and high soil appeared too stiff and dry, even when large foam
liquid loss
injection ratios were used (tests 7 and 8);
5 18.2 20 195.0 Borderline – Very fluid and with
some liquid loss – for a certain water content (about 20%), the material
6 18.0 20 216.7 Borderline – Very fluid and with appeared too fluid and wet, even for low foam injection
some liquid loss ratios (test 4);
7 0 40 – Not suitable – Too dry – the tests that gave suitable results (tests 2, 9, 11 and 12)
8 0 80 – Not suitable – Too dry
were located in the area between two straight lines which
9 5.0 40 103.3 Suitable – A little too stiff and dry
10 14.0 20 145.0 Suitable – A little too dry and define, respectively, the upper threshold over which the
with some liquid loss material was too wet and fluid (tests 3 and 4), and a
11 14.3 40 195.0 Suitable – Slightly too wet lower threshold, under which the material was too dry
12 5 60 195.0 Suitable – No liquid loss and did not show the required plastic or pulpy features
The slump values are the average values of three tests. (tests 1 and 10).

Fig. 4. Parameter set area definition and optimum conditioning treatment choice.
312 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317

In order to verify that water and foam act independently

and on different grain size elements, finer soil was added to
the reference sand and granulometric curve 2 was obtained.
A new test programme was therefore undertaken to check
the new optimum water content, keeping the FIR constant
and equal to 40%. Table 2 summarizes the test parameters
and the obtained results . Form a comparison of the results
of the two test campaigns, it can be seen that:

– When w = 10%, both for the reference sand and for soil
n. 2, the mixes show satisfactory behaviour, as far as
plasticity and consistency are concerned.
– When w = 15%, for soil n. 2, the mix shows suitable
Fig. 5. Effect of water content and Foam Injection Ratio on slump cone behaviour while for soil n. 1 (test 11), the mix is too fluid
fall.
and wet. This result confirms that if the soil has more
fines, a higher water content is necessary to obtain suit-
It is therefore possible to define an area in which the able behaviour.
conditioning parameter sets give a suitable final mix. The – When w = 5%, for soil n. 2, the mix shows too stiff and
centroid of such an area could therefore be indicated as dry behaviour, while for the reference sand (test 9), the
an ‘‘optimum’’ conditioning parameter set which, for the mix has suitable behaviour, even thought slightly dry.
studied reference sand, corresponds to the following set This result again shows that finer particles change the
of parameters: FIR = 40% and water content = 10%. global behaviour of the conditioned mix and require a
In order to analyse the effect of the foam and water on higher water content to obtain optimal behaviour. This
the quality of the conditioning, the results obtained from last result can be confirmed from the analysis of the data
the slump tests that gave plastic and fluid behaviour of that were obtained during the excavation of the Milan
the mix were plotted: if the water content is plotted vs. Railway Junction tunnel (Peron and Marcheselli, 1994)
the cone fall, it can be seen that the obtained relationship where the EPB machine excavated a gravely soil with
is nearly linear (Fig. 5) and that increasing FIR this line an important silt content (20%). It was necessary to
moves up. Thus, a small increase in the water content leads inject a certain amount of water into the excavation
to a small increase in the slump cone fall and this effect is chamber, in the drive above the water table, in order
independent of the adopted FIR. to plasticize the soil while in the drive below the water
From this result, it can be concluded that water and table, the larger amount of water led to the necessity
foam act independently as lubricating agent and that the of injecting less foam to obtain suitable plasticity of
lubricating mechanism is different for the two fluids: water the soil.
mobilizes an ‘‘internal’’ lubrication, acting on the finer part
(clay and silt), whereas the foam gives an additional effect,
reducing the friction between the sand grains. 3.2. Screw conveyor device tests
The practical consequences of these results for EPB tun-
nelling procedures are that, in the case of a moderate-to- The tests with the screw conveyor device required a con-
high presence of fine sized materials, it could be useful to ditioned sample of about 400 kg which was obtained
inject water into the excavation chamber as was done for through the following procedure: a volume of 25 kg of soil
example, in the excavation of the tunnels for the Milan with a defined water content was placed into a mixer bowl
Railway Underground Junction (Peron and Marcheselli, and the foam was added. As soon as the mix was ready it
1994) whereas in the presence of coarse soil without fine was poured into a tank and this operation was repeated
fractions, large amounts of foam should be injected (even until the tank was full. Subsequently, the upper plate was
up to 80% or more) or a fine filler should be injected into positioned and pushed down by the piston until the desired
the ground (as in the tunnel of Lot 5 of the Turin Metro). test pressure in the tank was reached (Vinai et al., 2006).

Table 2
Slump tests and obtained results on soil n. 2
Test Water content (%) FIR (%) Slump (mm) Comments
13 10 40 120 Suitable
14 15 40 185 Suitable – Foam can be seen on the ground after the test
15 20 40 225 Not suitable – Very fluid, with some liquid loss and the foam can be seen on the ground after
the test
16 5 40 – Not suitable – Too dry
The slump values are the average values of three tests.
 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317 313

The screw conveyor was then started and the material 2. A too dry conditioning: a water content equal to 5%,
was collected at the screw discharge point where it was then FIR of 40%, FER of 16 and a pressure of 60 kPa applied
weighted. The revolving speed of the screw was kept con- to the top plate.
stant at a value of 6 rpm for the whole test programme, 3. A too wet conditioning: a water content equal to 20%,
since this value was similar to the values normally used FIR of 40%, FER of 16 and a pressure of 60 kPa applied
in real EPB tunnelling sites. to the top plate.
The test was stopped when the piston reached its maxi- 4. Saturated sand with no conditioning and a pressure of
mum stroke. Slump tests were performed on the material 90 kPa applied to the top plate.
that came out with the screw as well as on the soil in the tank
at the end of the test in order to check whether the material The most relevant measured data are summarized in
properties had changed during the extraction process. Table 3. If these results are compared, the positive effect
The following data were monitored during the test: the of the conditioning on the efficiency and the performances
pressure distribution in the tank, the required torque for of the screw conveyor device can clearly be seen and the
rotating the screw conveyor, the upper plate displacement device permits to highlight the positive effect of condition-
gradient, the discharged weight gradient and the trend of ing using measurable parameters. The behaviour of the
the pressure values along the screw conveyor. conditioned soil can be therefore assessed and the different
The shape and value of these parameters permitted the conditioning amount can be compared.
suitability of the conditioning amount to be evaluated The pressure in the tank, using the optimal conditioned
and different parameter sets to be compared. material (set n. 1), showed a regular trend and it was closed
to the theoretical value (Fig. 6) while, when the dry mate-
3.2.1. Performed tests and obtained results rial (set n. 2), the wet material (set n. 3) and the saturated
Four different test were carried out on sand n. 1 using sand (set n. 4) (Fig. 7) were used, it was more difficult to
the following test parameters: control the pressure applied by the upper plate (particu-
larly for set n. 4) and the values were irregular during the
1. Conditioning close to the optimum values as defined in test. As far as parameter set n. 2 is concerned, the pressure
the slump test programme: a water content equal to distribution matched the theoretical prediction very well
10%, FIR of 40%, FER of 16 and a pressure of but the fluctuations around the mean value were due to
90 kPa applied to the top plate. the difficulties involved in controlling the pressure since
Table 3
Summary of the results obtained from the screw conveyor test campaign
Test set Theoretical pressure (kPa) Measured pressure (kPa) Torque (N m) Cell 1 (kPa) Cell 2 (kPa) Cell 3 (kPa)
Top Bottom
1 90 80 95 6–10 20–28 12–17 4–6
2 60 65 75 19 17 12 5
3 60 65 80 3 26 13 4
4 9 150 100 30–40 4–8 4–6 2–4
Key: Theoretical pressure = pressure in the tank derived from the pressure in the cylinder; Measured pressure = pressure measured by the cells located
respectively on the top and on the bottom of the tank; Torque = torque acting on the shaft of the screw conveyor; Cell X = pressure cell on the screw case
(cell numeration goes from the closest to the furthest from the tank).

Fig. 6. Pressures measured in the tank with the conditioned set of parameters n. 1. The bottom cell is located on the bottom of the tank while the top cell is
located below the upper plate that apply the pressure to the conditioned sand.
314 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317

Fig. 7. Pressures measured in the device tank with conditioned set of parameters n. 4 (saturated sand). It is evident that the values are very irregular and
higher than the theoretical one and that parameter set n. 1 and the top cell (below the upper plate) give higher values than the theoretical ones due to the
irregular behaviour of the material during the test.

the material was too stiff. These data are relevant for a cor- an index to quantify whether a material is correctly condi-
rect choice of the conditioned material in EPB applications, tioned, with reference to the reduction in the requested tor-
as one of the required qualities of the conditioned soil is to que, both at the EPB head and on the screw conveyor.
be able to apply and to transmit the pressure inside the The pressure plate lowering during the test was regular
bulk chamber. Therefore, the closer the data are to the the- with the optimal conditioned soil, since the screw conveyor
oretical values, the better the behaviour of the conditioned extracted efficiently the soil, while it was irregular and dif-
soil is. ficult to control when using the too dry and the too wet
The screw torque trends show that, for the too dry con- soils and also with the saturated sand. When the test was
ditioned soil (set n. 2), the measured values were higher carried out with the ‘‘too wet’’ parameters (set n. 3), the soil
than the ones obtained from the optimum conditioning soil came out earlier (with reference to the moment when the
(set n. 1), but they were slightly lower (but smoother) than pressure is applied by the top plate) than in the other tests,
the ones obtained with saturated sand. The torque value since the material was pushed out rather than dragged from
recorded during the test carried out with set n. 3, after an the screw flights.
initial peak, was very low since the screw rotated inside a The pressures along the screw case were higher and
mud rather than inside a plastic material. The peak was more regular during the extraction of the conditioned
related to a flow-priming phenomenon: as soon as the materials than during the extraction of the saturated soil
material reached the discharging point, the torque value (Figs. 9 and 10) and they decreased more regularly from
showed an important decrease that was due to a flow in the tank to the discharge point. This effect was evident with
the material rather than an extraction with the screw con- the optimal conditioning, while some irregularities were
veyor (Fig. 8). This parameter could therefore be used as recorded for the too stiff and too wet conditioned soils.

Fig. 8. Measured torque values.

 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317 315

Fig. 9. Comparison of the pressure measured along the screw conveyor measured for the saturated sand (a) and for the optimal conditioning set of
parameters (b).

The measured values indicated that the pressure dissipa- The choice of the optimal conditioning amount and a
tion, for the optimal conditioning set had a mean ratio of comparison between the various types of conditioning
about 2.8 kPa/flight, which is in good agreement with the products can be made using the simple slump tests or a spe-
values usually indicated for the design of a screw conveyor. cifically designed laboratory device where the conditioned
A regular decrease in the pressure inside the screw con- soil is extracted from a pressurized tank using a screw con-
veyor is an index that it has been properly filled and that veyor that properly simulates the real condition found in
the pressure in the tank that is correctly dissipated, as an EPB machine.
desired when working with an EPB below the water table. The systematic application of the slump test to two ref-
The optimum conditioned material (set n. 1) maintained erence sands has shown that suitable behaviour (plastic and
its workability after the extraction, since the cone falls for ‘‘pulpy’’) is only found for some water content and FIR
the slump tests performed at the end were in good agree- combinations and that a slump cone fall of about 150–
ment with the data obtained from the slump test campaign 200 mm is an index of a suitable mix, as is often indicated
and the conditioned material did not lose water during the in technical literature. Furthermore, the test is simple and
extraction, as can be seen from the measurements on the quick, but it obviously only permits a qualitative descrip-
sampled material which confirmed that the extracted mate- tion of the behaviour of the conditioned material through
rial had the same water content as before the test. an observation of the shape and conditions of the slump
cone and only a preliminary definition of the conditioning
4. Conclusions amount can therefore be obtained.
On the other hand, the screw conveyor device that was
EPB tunnelling requires the application of soil condi- specifically designed and used permits a quantitative eval-
tioning to obtain a correct excavation control particularly uation of the properties of the conditioned soils and a
when working in complex choesionless soils. However, it simple and clear comparison to be made between different
can be observed that only a limited number of internation- conditioning sets. The carried out tests have shown that it
ally recognized standards and/or guidelines are available to is possible to measure the capacity of the conditioned soil
asses soil conditioning. to control and apply a pressure both inside the pressure
316 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317

Fig. 10. Comparison of the pressure measured along the screw conveyor as measured for the too dry conditioned set of parameters (a) and for the too wet
conditioning set of parameters (b).

tank and along the screw conveyor, thus permitting the io who assisted with the experimental work performed
best FIR to be chosen before starting the EPB machine during their MSc thesis.
excavation. An uniform pressure reduction can be The work reported in this paper was performed within a
observed along the screw conveyor when testing a cor- PhD research project that was supported by MAPEI SpA
rectly conditioned soil: the screw is completely filled and and by the Italian Ministry of University and Research –
it permits the control of the tank pressure. On the con- PRIN 2006.
trary if a saturated sand or a too wet mix are tested the The authors have contributed equally to the develop-
measured pressures along the screw conveyor are practi- ment of this research.
cally uniform.
Important differences are also observed by comparing
the screw conveyor torques measured using different mixes: References
the torque values measured when testing a suitable mix are
AFTES, 2001. EUPALINOS 2000. Synthèse, AFTES (ed.), Octobre 2001,
less than 25% of the one measured with a saturated sand or Paris (in French).
a too dry mix. Anagnostou, G., Kovari, K., 1996. Face stability conditions with Earth-
From these data, it appears that the use of a device that pressure-balanced Shields. Tunnelling and Underground Space Tech-
is able to simulate the extraction of the material from a nology 11 (2), 165–173.
pressurized tank at a laboratory scale is a good tool to help Bezuijen, A., Schaminée, P.E.L., 2000. Model experiments using foam
simulation the drilling with an EPB shield, GeoDelft Report No. BF
job site managers and designers to carry out round robin 51010. Delft, GeoDelft.
tests with different sets of conditioning parameters to asses Bezuijen, A., Schaminée, P.E.L., 2001. Simulation of the EPB-shield TBM
the best conditioning amount to be used on a job site. in model tests with foam as additive. Proceedings of Congress on
Modern Tunnelling Science and Technology. Balkema, Kyoto, Rot-
terdam, pp. 935–940.
Acknowledgements
Boone, S.J., Artigiani, E., Shirlaw, J.N., Ginanneschi, R., Leinala, T.,
Kochmanova, N., 2005. Use of ground conditioning agents for Earth
Special thanks are due to Prof. S. Pelizza for his sugges- Pressure Balance machine tunnelling. In: AFTES (Ed.), Proceedings of
tions and comments and to Dr. D. Gallina and Dr. L. Bor- Congrès International de Chambéry, pp. 313–319.
 R. Vinai et al. / Tunnelling and Underground Space Technology 23 (2008) 308–317 317

Bordachar, F., Nicolas, L., 1998. Fluides conditionneurs pour la pression Milligan, G., 2000. Lubrication and Soil Conditioning in Tunnelling, Pipe
de terre, Tunnels et ouvrages souterrains, 169 (Janvier/Février), Jacking and Microtunnelling State of the Art Review, Geotechnical
AFTES, pp. 21–27 (in French). Consulting Group, London.
Branque, D., Subrin, D., Boutin, C., 2003. Etude sur model réduit du Milligan, G., 2001. Soil conditioning and lubricating agents in tunnelling
creusement de tunnels par la méthode du bouclier a pression de terre. and pipe jacking, In: Proceedings of Underground Construction 2001,
Tunnels et ouvrages souterrains, 179 (Septembre/Octobre), AFTES, London, pp. 105–116.
Paris, pp. 307–312 (in French). Nomoto, T., Imaura, S., Hagiwara, T., Kusakabe, O., Fujii, N., 1999.
EFNARC, 2005. Specification and Guidelines for the use of specialist Shield tunnel construction in centrifuge. Journal of Geotechnical and
products for Mechanized Tunnelling (TBM) in Soft Ground and Hard Geoenvironmental Engineering (April), 289–300. ASCE.
Rock, EFNARC recommendations. Peña, M., 2003. Soil conditioning for sands. Tunnels and Tunnelling
Herrenknecht, M., 1994. EPB or slurry machine: the choice. Tunnels and International (July), 40–42.
Tunnelling International (June), 35–36. Peron, J.Y., Marcheselli, P., 1994. Construction of the ‘Passante Ferro-
Herrenknecht, M., Rehm, U., 2002. Newest development in mechanized viario’ link in Milan. Italy. lots 3P, 5P, and 6P: excavation by large
tunnelling, In: AFTES (Ed.), Proceedings of Journées d’études EPBS with chemical foam injection, In: IMM (Ed.), Proceedings of
internationals, Toulouse, pp. 201–206. Tunnelling ’94, London, Chapman & Hall, London, pp. 679–707.
Jancsecz, S., Krause, R., Langmaack, L., 1999. Advantages of soil Quebaud, S., 1996. Contribution à l’étude du percement de galeries par
conditioning in shield tunnelling: experiences of LRTS Izmir. In: boucliers à pression de terre: amélioration du creusement par
Alten, T. et al. (Eds.), Proceedings of International Congress on l’utilisation des produits moussants, PhD thèses, Université des
Challenges for the 21st Century. Balkema, Rotterdam, pp. 865– Sciences et Technologies de Lille, Lille (France) (in French).
875. Quebaud, S., Sibai, M., Henry, J.P., 1998. Use of chemical foam for
Kuribashi, Y., Yagi, K., Ishimoto, H., 1993. The PMF Super shield improvements in drilling by earth pressure balanced shields in granular
tunneling process – expanding applications for earth pressure soils. Tunnelling and Underground Space Technology 13 (2), 173–180.
balanced shield tunnelling. In: Burger, H. (Ed.), Proceedings of Vinai, R., 2006. A contribution to the study of soil conditioning
International Congress on Options for Tunnelling 1993. Elsevier, techniques for EPB TBM applications in cohesionless soils, PhD
pp. 411–420. thesis, Politecnico di Torino, Torino (Italy) (unpublished).
Langmaack, L., 2000. Advanced Technology of Soil Conditioning in EPB Vinai, R., Oggeri, C., Peila, D., Pelizza, S., 2006. Condizionamento con
Shield Tunnelling, MBT Publication. schiuma dei terreni per applicazioni EPB: sperimentazione mediante
Maidl, U., 1995. Erweiterung der Einsatzbereiche der Erddruckschilde un nuovo apparato di laboratorio, Gallerie e grandi opere sotterranee,
durch Bodenkonditionierung mit Schaum, PhD Dissertation, Ruhr- vol. 78(1), Patron Editore, Bologna (in Italian).
Universität Bochum, Germany (in German). Williamson, G.E., Traylor, M.T., Higuchi, M., 1999. Soil conditioning for
Mair, R., Merritt, A., Borghi, X., Yamazaki, H., Minami, T., 2003. Soil EPB shield tunneling on the South Bay Ocean Outfall. In: Proceedings
conditioning for clay soils. Tunnels and Tunnelling International of RETC 1999, pp. 897–925.
(April), 29–33. Yoshikawa, T., 1996. Soil pressure drop af the screw conveyor for shielded
Merritt, A., Mair, R.J., 2006. Mechanics of tunnelling machine screw machines. Transactions of the Japan Society for Mechanical Engi-
conveyor: model tests. Geotechnique 56 (9), 605–615. neers, Part C 62 (595), 1197–1203.