• Failure of segmental linings

    Rupert Sternath of Stern Consult adds to the TunnelTalk Feedback discussion about the failure of the segmental lining behind the TBM drive through the ground freezing support for the Rastatt rail tunnel drive in Germany to enquire after the process for filling the annual gap around the segmental lining to avoid failure of the integrity of the segmental lining ring.

    David Caiden of Arup questions if failure of the lining is truly failure of the lining as designed and under extreme conditions of running or flowing ground, recalling a segmentally lined TBM tunnel collapse on the Hong Kong Metro in 1983.

    Nick Shirlaw of Golder Associates responds to the TunnelTalk report of the Rastatt TBM drive collapse and comments particularly on the failure of the segmental lining of the drive, recalling three other incidences of segmental lining failure that resulted in complete, and in one case fatal, tunnel collapse; in Hull, UK (1999); Cairo, Egypt (2009) and in Okayama, Japan (2012) – and suggests that, while segmental lining failures behind soft ground TBM drives “are very rare, the consequences are so severe that we, as an industry, need to make sure that the relevant lessons are learned and the likelihood of another incident reduced.”

    See the full Feedback contribution at the bottom of this article page and on the Feedback page and for further discussion about the failure of the Rastatt tunnel failure and its aftermath, see the Discussion Forum article.

Ground freezing TBM drive collapse in Germany 22 Aug 2017

Shani Wallis, TunnelTalk

Collapse of a TBM tunnel drive in association with ground freezing support under the main Rhine Valley freight and passenger railway line at Rastatt in Germany has halted all rail traffic between Karlsruhe and Basel, Switzerland, and demanded backfilling of the new tunnel with concrete, burying the TBM in the process.

Concrete backfills the collapsed tunnel
Concrete backfills the collapsed tunnel
Fig 1. Details of the leading east TBM drive collapse
Fig 1. Details of the leading east TBM drive collapse

The 4,270m long twin tube rail tunnel is part of a 17km long project to double the capacity on the Rhine Valley rail corridor from two to four tracks taking the underground route beneath the city of Rastatt. To limit the length of underground construction and maintain acceptable gradients for the 250km/hour rail lines, the TBM drives are aligned at depths of up to 19m maximum with the 10.97m diameter TBMs passing a minimal 4m to cross beneath the main-line surface tracks (Figs 1 and 2).

Fig 2. Details of the original design for passage of the new tunnels under the surface mainline railway

To support the shallow alignment beneath the main rail lines and through loose sand and gravel deposits, a pre-support of horizontal ground freezing to create a full round ring of frozen ground was installed for the TBMs driving through the centre of the frozen collars.

It was on Saturday 12 August that monitoring instrumentation detected subsidence under the surface rail tracks. This progressed rapidly to create a subsidence depression of 500mm beneath the tracks and buckling the rails, demanding the halt of all rail traffic on the line.

At the time, the leading TBM in the east tunnel bore had passed beneath the rail tracks by about 40m and was 3,974m into its drive and just short of breakthrough in the reception shaft which was also access for installation of the horizontal ground freezing operation.

The trailing TBM in the west tunnel is also at a standstill as a result of the event and is about 1,000m behind the leading TBM at 3,064m into its 4,250m drive.

Immediate response to the incident was to install three slick-line holes into the new tunnel to create a concrete plug about 150m behind the cutterhead and fill the new 9.5m i.d. segmentally lined tunnel from the TBM bulkhead to the plug with 10,500m3 of concrete (Fig 1).

According to a press briefing on Monday this week (21 August) by Section Manager Jürgen Kölmel for project client Deutschebahn, the segmental lining of the tunnel under the rail tracks and 40m behind the TBM did not crack or collapse but rather the seven segments in the 500mm thick x 2m long rings dislodged creating gaps and allowing infiltration of water and ground.

This indicates a failure of the freezing regime and possible over-excavation of material as the TBM passed through the zone.

Collapse sinkhole beneath the existing surface rail tracks
Collapse sinkhole beneath the existing surface rail tracks

Up to 370 trains per day are being diverted away from the closed section of rail route with buses managing the broken passenger train journeys and diversions having to be found for the 170 to 200 of long distance freight trains per day on the corridor. In addition several buildings and homes in the vicinity have been evacuated. Estimates for how long the line will remain closed range from four to six weeks.

Work on the €693 million federally-funded project to build a new double-track alignment through Rastatt started in mid-2013 and the construction of the underground sections was awarded to the Ed Züblin/Hochtief joint venture. Together with TBM tunnelling, the construction works include open cut and cut-and-cover sections, NATM cross passages and ground support methods including sheet piling, diaphragm walls, jet grouting, underwater concreting and several applications of ground freezing.

Samples of the prevailing geology
Samples of the prevailing geology

To complete the twin TBM drives, Züblin/Hochtief procured two slurry Herrenknecht Mixshields. The machines launched from the north transition portal, the first east bore TBM beginning its drive in May 2016 and the west bore machine starting four months later in September 2016. Due to the shallow alignment and first section of angled ground freezing from the surface as a canopy over each TBM tube protected against potential slurry blowouts for a distance of up to 290m as the machines passed below a cover of about 5m below the Federbach conservation area. Freezing will also be used in the waterbearing ground for open face excavation of cross passages every 500m between the two tubes.

According to technical trade press media, excavation of the section under the main railway lines by extending the operation of the TBM drives was a contractor alternative to initial proposals to use open-face NATM excavation through the ground freezing support. The alternative was said to result in construction savings and was approved after comprehensive study.

The 200m long horizontal freeze for the rail underpass was established from two 30m deep shafts installed about 100m either side of the tracks. From the shafts a circle of 42 x 100m long freeze pipe holes were drilled using horizontal directional drilling and the freeze applied to create a 2m thick collar of frozen ground around each 10.97m o.d. tunnel tube.

The exact cause of the tunnel failure is now the subject of intense and wide-ranging investigation to include freeze management, instrumentation and monitoring reviews, TBM operation, and operator logs. Reports in the local German media suspect that the integrity of the ground freezing operation was undermined by a recent period of high summer temperatures combined with periods of heavy rainfall.

Launch and working site of the twin TBM drives
Launch and working site of the twin TBM drives

As well as preparing the freezing operation for changing climate conditions, the application also involves accounting for the heat generated by the operation of the TBM machine and systems and by the slurry excavation process within the excavation chamber ahead of the bulkhead. The balance between keeping the freeze to support the ground and the heavy and regular train traffic above while preventing the TBM becoming trapped in the frozen ground or its slurry circulation system or other operating fluids freezing up in the applied freezing conditions.


Rastatt TBM drive collapse and failure of its segmental lining
Feedback from: Rupert Sternath

One can see out of the available publications that the segments of the lining have dislocated some 40m behind the TBM. This is an indication that the ring gap has not been filled properly. As this happened to an experienced contractor it may be the case that the grouting operation together with the TBM drive through a frozen soil includes some particular problems.

Mining through an ice body has the characteristic of a hard rock drive, which requires some over excavation to enable shield steering.

Most shielded TBMs use grout lines through the shield tail to fill the annual gap immediately behind the tail seal. Under hard rock conditions the mortar tends - due to the over excavation - to flow around the body of the TBM and to the front and into the working chamber and so leaving voids outside the segmental lining. These voids have to be filled by a secondary grouting operation through the segments as soon as possible from the top of one of the trailing gantries.

In case of a frozen soil outside the gap however, it may happen that the voids are being filled by groundwater, which would also freeze, and as heat is present inside the tunnel during the mining process, the ice in the gap may melt leaving the segments unsupported. In this case filling of the gap by blowing pea gravel through the segments combined with a cement grouting operation may be a better option in my view.

Anyway, the tunneling world is keen to see the outcome of the following investigations and very interested on further reports in TunnelTalk about them!

Rupert Sternarh
Stern Consult
Holzhamer Bogen 15 83624 Otterfing

Rastatt TBM drive collapse and failure of its segmental lining
Nick Shirlaw

Thank you for the write-up on a major failure, which appears so far to have had limited press coverage, despite the severe impact on train operation.

As far as I am aware, this is the fourth incidence of catastrophic segmental lining failure behind a pressurised TBM in the last eighteen years; these being:

  • Hull wastewater transfer tunnel, UK [1999]
  • Cairo, Egypt [2009]
  • Okayama, Japan [2012]
  • Rastatt, Germany [2017]

I know of two other cases of severe, local, distortion of gasketted, concrete segmental tunnel linings, in Singapore and the USA, where total failure was avoided by providing additional support in the tunnel.

Given the huge number of segmentally lined tunnels built over the last 18 years, the proportion that has failed is tiny; and in each case the failure has been local, without similar problems on the rest of the drive. However, the consequences of each of the failures have been catastrophic.

To date, the best documented of the failures is that at Hull, which was the subject of an investigation that was summarised in Grose and Benton (2005)(1). Even in this case the investigation was limited and the conclusions tentative.

The paper was the subject of a number of discussions, to which I contributed, and which elicited detailed responses that contained much additional information to that in the original paper(2). In my opinion, the conclusions were inconsistent with some of the observations made in the tunnel; I stated this in a further discussion, which was submitted, but rejected by the journal on the basis that they did not accept a second round of discussions.

The failures at Cairo and Okayama have been the subject of a number of articles in TunnelTalk, but I have not seen any definitive explanation of causation.

This limited response to these failures can be compared with that to the failure of the cut-and-cover tunnel at Nicoll Highway in 2004 in Singapore. This was the subject of a public inquiry, which published clear, extensive and detailed findings that have had a major effect on practice in Singapore.

Because the failure of segmental linings is so rare, those listed above have each occurred in different countries. As far as I am aware each has been assessed in isolation. I hope that the detailed results of the investigation into the failure at Rastatt are made public, but this will take months or years, based on previous experience. Given that there have been several failures there does appear to be a case for reviewing them together, to see if there are any common features, and lessons to be learned.

Even though these events are very rare, the consequences are so severe that we, as an industry, need to make sure that the relevant lessons are learned and the likelihood of another incident reduced.

1. Hull wastewater flow transfer tunnel: tunnel collapse and causation investigation, Grose and Benton, 158, October 2005, Issue GE4, Proceedings of the Institution of Civil Engineers, Geotechnical Engineering

2. Hull wastewater flow transfer tunnel: tunnel collapse and causation investigation, discussion report, Ground Engineering, Volume 159 Issue 2, April 2006, pp. 125-128.

Nick Shirlaw,
Golder Associates,


Add your comment

Thank you for taking the time to share your thoughts and comments. You share in the wider tunnelling community, so please keep your comments smart and civil. Don't attack other readers personally, and keep your language professional.
In case of an error submitting Feedback, copy and send the text to Feedback@TunnelTalk.com
Name :

Date :

Email :

Phone No :

   Security Image Refresh
Enter the security code :
No spaces, case-sensitive