NATM in soft-ground 3/9/02 03:44 pm Page 338

NATM & SHOTCRETING FOCUS

NATM IN SOFT-GROUND:
A CONTRADICTION OF TERMS?

W
hen the New Austrian Tunnelling Method (NATM) is mentioned, it often means many different
things to even experienced tunnelling engineers. The scene has more recently been complicated
by new terms introduced to reflect certain aspects of NATM. Victor Romero, an associate with
US-based engineering practice Jacobs Associates gives his views on NATM and its application to
soft-ground tunnelling, and dispels some misconceptions about this sometimes controversial topic.
‘New Austrian Tunnelling Method’.
‘Sequential Excavation Method’. ‘Sprayed
Concrete Lining’. Not even the tunnelling
industry can give a unified name for a
tunnelling method pioneered by the
Austrians in the later half of the twentieth
century. Furthermore, we are seeing
increased use of this method in the US,
particularly in soft-ground conditions. In
order to better define the method, it is
helpful to look at its history and use in both
design and construction.

NATM in perspective
As defined by the Austrian Society of
Engineers and Architects, the NATM
“…constitutes a method where the
surrounding rock or soil formations of a
tunnel are integrated into an overall ring-
like support structure. Thus the supporting
formations will themselves be part of this
supporting structure.” In world-wide prac-
tice, however, when shotcrete is proposed
for initial ground support of an open-face
tunnel, it is often referred to as NATM. The
term NATM with reference to soft ground,
however, can be misleading.
As noted in a very thoughtful article by Emit
Brown1, NATM can refer to both a design
philosophy and a construction method. Key
features of the NATM design philosophy are:

● The strength of the ground around a

tunnel is deliberately mobilised to the
maximum extent possible.
● Mobilisation of ground strength is Generic ground reaction curve that is often the basis of NATM design (after Brown, 19811)
achieved by allowing controlled defor-
mation of the ground. ● The tunnel is sequentially excavated and were in widespread use in the US and
● Initial primary support is installed having supported, and the excavation sequences elsewhere in soft-ground applications
load-deformation characteristics appro- can be varied. before NATM was described in the litera-
priate to the ground conditions, and ● The initial ground support is provided by ture. In current practice, for soft-ground
installation is timed with respect to shotcrete in combination with fibre or tunnels which are referred to as NATM
ground deformations. welded-wire fabric reinforcement, steel tunnels, initial ground support in the form
● Instrumentation is installed to monitor arches (usually lattice girders), and of shotcrete (usually with lattice girders and
deformations in the initial support sometimes ground reinforcement (e.g., some form of ground reinforcement) is
system, as well as to form the basis of soil nails, spiling). installed as excavation proceeds, followed
varying the initial support design and the ● The permanent support is usually (but by installation of a final lining at a later date.
sequence of excavation. not always) a cast-in-place concrete lining. Simply speaking, soft ground can be
described as any type of ground requiring
Key features of NATM construction It should be noted that many of the support as soon as possible after excava-
methods are: construction methods described above tion in order to maintain stability of the
338
NATM in soft-ground 3/9/02 03:44 pm Page 339

SOFT GROUND NATM

Example of multiple, numerical model that presumes to

interacting tunnels where calculate stresses in the ground and tunnel
numerical modelling is support (i.e. lining). Traditionally, this has
appropriate for design
been done with finite element or finite
difference methods. As with any computer
modelling, the answers are only as good as
the geotechnical input parameters and the
constituent models used for analysis. The
constituent model theories should not be a
mystery, and both the Engineer and the
Client (be it an owner or a contractor)
should clearly understand the limitations of
computer modelling. Most importantly, field
observations and measurements should be

excavation. For tunnels in dense urban

areas, limiting settlement is of paramount
importance to avoid damage to overlying
structures. In order to limit settlement
and ensure worker safety, most soft
ground tunnels employ the following
measures:

● Excavation stages must be sufficiently

short, both in terms of dimensions and
duration.
● Erection of the ‘full ring’ of initial ground
support must be completed immedi-
ately after excavation.

It should be noted that the major differ-

ence between the NATM design philos-
ophy and traditional soft ground tunnel
construction measures is that deformation
of the soil is not easily ‘controlled’.
Therefore, it can be concluded that the
excavation and support planned for
sequentially excavated, shotcrete-lined
tunnels in soft ground utilises NATM
construction methods but not necessarily
NATM design methods.

NATM for soft ground?

In soft-ground tunnelling, safety dictates
that the ground support be placed immedi-
ately after excavation. As long as the
ground is properly supported, NATM
construction methods are appropriate for
soft-ground conditions. However, there are
cases where soft-ground conditions do not
favour an open face with a short length of
uncompleted lining immediately next to it,
such as in flowing ground or ground with
short stand-up time (i.e., failure to develop
a ground arch). Unless such unstable
conditions can be modified by dewatering,
spiling, grouting, or other methods of
ground improvement, then NATM may be
inappropriate. In these cases, close-face
shield tunnelling methods may be more
appropriate for safe tunnel construction.

A ‘black box’ for design?

Is a computerized ‘black box’ appropriate
for NATM design? By ‘black box’ I mean a RES No. 127
339
NATM in soft-ground 3/9/02 03:44 pm Page 340

SOFT GROUND NATM

used to confirm assumptions and calibrate

future models.
Notwithstanding the above, numerical
modelling is a useful tool for design of
sequentially excavated, shotcrete-lined
tunnels in soft ground. In particular, numer-
ical modelling is useful where interacting
tunnels, unusual geometries, or adjacent
structures are present. We prefer to use the
software FLAC or PLAXIS for numerical
modelling of tunnel linings in soft-ground.
However, if a shotcrete-lined tunnel has a
section that is nearly circular or oval with
no irregularities, and if there are no adja-
cent surface or subsurface structure inter-
acting with the tunnel, then traditional
ground-lining interaction methods can be
used for design. Indeed, the closed-form
ground-lining interaction equations provide
a prudent check on numerical modelling.
So do you need a black box? If you don’t
know what the theory behind the black box
is and cannot correlate the analysis with
tunnel behaviour in the field, then you Short headings and spiles being used to control unstable ground
should not rely on a black box for any
type of tunnel design. However, if you are models. More importantly, however, moni- 2. Define the purpose of instrumentation,
using a tested program with good toring serves to alert the designer and the and select parameters to be monitored.
constituent models and know its limita- constructor if the lining is not performing as 3. Predict magnitudes of change, and
tions, then programs such as FLAC can intended, or is in danger of collapse. In this determine threshold limits and remedial
be useful design tools. But be sure to respect instrumentation of NATM construc- actions.
verify the results in the field and with a tion is no different from other types of 4. Assign tasks and responsibilities.
traditional ground-lining interaction analysis, geotechnical construction. Therefore the 5. Select instruments and locations.
if appropriate. tried and true rules of thumb for geotech- 6. Devise methods to ensure correctness.
nical instrumentation apply to NATM, 7. Plan data collection, processing, presen-
Instrumentation and namely: tation, interpretation, and reporting.
monitoring?
As noted above, instrumentation and moni- 1. Predict mechanisms that control behav- If you get these steps in a systematic
toring play a key role in verifying design iour, and define the geotechnical ques- instrumentation and monitoring approach
assumptions and calibrating numerical tions to be answered. correct, and there is good communication
between the Engineer and the Contractor,
then you have a fighting chance of
getting good data that can be relied upon,
in order to make decisions during
construction.

Tunnel collapses?
Unfortunately there have been several
collapses or other stability failures of NATM
projects around the world including, most
recently in Turkey and the US. Perhaps the
most famous is the Heathrow Airport
collapse in October 1994, which triggered
a thorough review of the NATM by the
British Health and Safety Executive (HSE).
In a 1996 report2, the HSE examined 39
NATM failures, categorising the location (in
the tunnel) of the failure. In most cases,
the failure was a result of heading collapse.
Broadly speaking the causes of these fail-
ures were varied, from unanticipated
geologic conditions, to design errors, to
construction quality problems, to poor
management. Nevertheless NATM failures,
or for that matter any tunnel failure, have
Shotcrete used successfully for initial support in a soft-ground tunnel one thing in common: most are caused
340
NATM in soft-ground 3/9/02 03:44 pm Page 343

SOFT GROUND NATM

by human error. It’s not the fault of (NATM) Tunnels, A review of sprayed If you, the readers, have views on the
the method, but misapplication of the concrete lined tunnels with particular terminology most appropriate to ‘NATM’
method. reference to London Clay, HSE Books sequential excavation, shotcrete linings
1996. or observational methods please share
What do we call it? 3 Institution of Civil Engineers. Sprayed them with other readers in the form
On the question of terminology, it is diffi- concrete linings (NATM) for tunnels in of a Letter to the Editor of World
cult to cover all aspects of the method. The soft ground. ICE design and practice Tunnelling, or submitted articles for longer
British have proposed SCL, short for guide, Thomas Telford Publishers, arguments or examples of method
Sprayed Concrete Linings (Institution of London, UK 1996. deployment.
Civil Engineers, 19963). However, the term
‘sprayed concrete’ is not very common in Acknowledgement
the US and many other parts of the world, Most of this article was first published in
where shotcrete is the commonly used Final Liner, the Jacobs Associates news-
Victor Romero
term for pneumatically applied concrete. letter, Spring 2002. Associate – Jacobs Associates
Moreover, shotcrete is used in different
types of tunnels (rock and soft ground)
and for many different applications (as
lining, lagging, surface protection, etc.).
Since the use of shotcrete lining in soft-
ground tunnels is almost always associated
with sequential excavation, my preferred
terminology is SEM, for Sequential
Excavation Method. The use of SEM as
standard tunnel terminology also highlights
the fact that this is a construction method,
rather than a design method. I would,
however, recommend keeping NATM as
terminology for the design method used in
rock tunnels when support installation is
timed to a ground reaction curve.

References
1 Brown, E.T. Putting the NATM into
perspective. Tunnels & Tunnelling,
November 1981.
2 Health & Safety Executive. Safety of

New Austrian Tunnelling Method

Closing the ring quickly behind the bench, as

shown in this photo, is key to preventing lining
instability RES No. 132
343