Tunnel construction in swelling rock – Prof. Georg Anagnostou and Prof. Kalman Kovári, both from ETH Zurich

1. Introduction

In Switzerland, several railway and highway tunnels crossing geological formations with swelling potential have existed for a long time. Usually, this phenomenon can be seen in rock layers located in the middle and eastern part of the Jura. Marls in the Tertiary Molasse or Triassic sulphate-bearing rocks, called gypsum keuper or anhydrite (Gipskeuper or Anhydritgruppe), are dreaded for their swelling behaviour.

Figure 1 Estheria beds (claystone with anhydrite in gypsum keuper)

2.  Well-known tunnels in swelling rock in Switzerland

  • Railway tunnels
    • Hauenstein Base Tunnel, Tecknau – Olten, length of Gipskeuper: 210 m N, 445 m (middle); 395 m S
    • Adler Tunnel, Muttenz – Liestal; length of Gipskeuper: 45 m
    • New Bözberg Tunnel, Effingen – Schinznach-Village, length of Gipskeuper: 95 m N,75 m S
  • Highway and main road tunnels
    • Belchen Tunnel A2, Eptingen – Hägendorf, length of Gipskeuper: east tunnel 1356 m, west tunnel 1308 m
    • Bözberg Tunnel A3, Effingen – Schinznach-Village
    • Gubrist Tunnel A1, third tube, Weiningen – Affoltern, length of swelling marl (molasse): 2,000 m
    • Chienberg Tunnel, H2 Pratteln – Sissach, length of Gipskeuper: 560 m W resp. 55 m E
    • Belchen, rehabilitated tunnel A2, Eptingen – Hägendorf, length of Gipskeuper: 1296 m

Although designed for withstanding swelling pressures, many of the tunnels driven in swelling rock have experienced severe damage during their lifetime and had to undergo massive reinforcing measures.

Figure 2: Belchen Tunnel, 1968, damage caused by swelling rock: heaved lining [3]
Figure 3: Rehabilitating the Adler Tunnel (Source: Swiss Federal Railways – SBB)

3.  Development of constructive measures against swelling occurrences

In Switzerland, tunnels have been built through the Jura fold-and-thrust belt and partly the Table Jura since 1853. The Oberer Hauenstein Tunnel, the Bözberg Railway Tunnel and the Hauenstein Base Railway Tunnel cut through Triassic gypsum keuper (Mittlerer Keuper) or anhydrite formations belonging to the Mittlere Muschelkalk. Massive heaves usually occur in such rock formations when water penetrates.

Until the end of the 60s, no specific measures were taken against swelling due to a lack of understanding of swelling processes and knowledge of realistic swelling pressure values. The first countermeasure was to opt for circular tunnel profiles and massive outer and inner linings. Today, such measures are still sufficient when encountering swelling marl. However, special measures like the Modular Yielding Support are needed when tunnels are driven through sulphate-bearing rock. This solution, first applied to the Chienberg Tunnel, H2 Sissach bypass, Switzerland, was developed by Prof. Kalman Kovári from ETH Zurich. Using prefabricated yielding elements tend to limit the load on arch rings. A new equilibrium with the overburden can be maintained even if the base of the tunnel is heaved.

Figure 4          Chienberg Tunnel, Modular Yielding System

Prof. Giorgios Anagnostou, ETH Zurich, Prof. Kovári’s successor, worked on several projects as a design engineer, tunnelling expert and design reviewer. His research deals, i.e. with conventional rock pressure theories (squeezing and swelling rock). He was involved in the design and expertise of several tunnels driven through squeezing rock (Adler Tunnel, Belchen Tunnel, and Chienberg Tunnel).

Figure 5: Prof. Giorgios Anagnostou
Figure 6: Prof. Kalman Kovári

The following passages will examine two tunnel projects where the swelling rock had a massive impact on the operation of the new tunnels, which ultimately led to extremely costly rehabilitation.

4.  Adler Tunnel

The about 5.3-km long Adler Tunnel is a single-tube, twin-track railway tunnel that connects Pratteln to Liestal on the Basel-Olten line. Excavated in the Table Jura Formation between 1996 and 1999 using a TBM, this tunnel boasts a circular cross-section (excavation diameter 12.6 m) with a reinforced segmental lining, an inner lining and a seal between the segments.

Since completion, deformations (squeezing and heaving of the tunnel cross-section) occurred due to rock swelling pressures in an approximately 40-m long section of the tunnel located about 1,000 m from the west portal at Pratteln. Severely damaged by cracks, the unreinforced inner lining had to be immediately secured with steel segment arches to ensure further serviceability.

Figure 7 Securing the Adler Tunnel with steel segment arches (left), (Source: Swiss Federal Railways – SBB))

In the vicinity of the damage zone, the tunnel runs through karst structures of Bunte Mergel and gypsum keuper (anhydrite) with a 35 to 45-m deep overburden. In this section, a sinkhole appeared in front of the TBM during construction (Figure 8). It developed into a cavern and eventually stopped the advance.

Figure 8 Adler Tunnel, grout injections following cavern formation [15]

The TBM could only drive the tunnel through the geologically critical zone from the surface after a bypass tunnel had been built and grout injections made. However, due to these events and measures, water reached the anhydrite inside the tunnel floor, which led to the known swelling. In addition, the reaction modulus of the lining in the tunnel apex was weakened by the sinkhole.

In 2008, the Swiss Federal Railways (SBB) held a project competition to extensively repair the approx. 40-m long damaged tunnel section and align its service life with that of the entire structure. This competition resulted in a proposal which was implemented in 2011 and 2012 in a slightly modified form, as shown in Figure 7.

Figure 9: Adler Tunnel, the principle of tunnel base abutment (Source: Swiss Federal Railways – SBB)

The rehabilitation proposal was based on the principle of tunnel base abutment. In addition to the additional load due to the approx. 30-m high overburden, the swelling pressures occurring in the tunnel floor was to be mainly counteracted by lateral Interlocking. Therefore, it was necessary to mobilise the friction required on the laterally overlying rock. For this purpose, two layers of permanent prestressed anchors were required on both sides. A major advantage of this proposition is that the anchor drilling works take place above the swelling sensitive gypsum keuper layer.

Consequently, niches with a cross-section of 2.0 x 4.5 m were excavated at carriageway level on both sides, each for one-way railway operation, over a 40-m stretch. The niches were to be used as supports for the prestressed anchors. The invert arch was also reinforced to absorb the swelling pressures more effectively.

Figure 10: Adler Tunnel, drilling of anchors and micropiles in the newly built niches (SBB)

Since completion, track height has been checked regularly. Heaving has still been detected, albeit of a much smaller magnitude. From today’s perspective, if the measured values remain the same, it will not be necessary to check again whether countermeasures are required for the next ten years.  

4.  Chienberg Tunnel

Located around 25 km southeast of Basel, the Chienberg Tunnel is part of the main highway (H2) connecting Liestal to Sissach. The 2.3-km long tunnel bypass, which consists of a single tube tunnel with two-way traffic, reduces traffic flow in the centre of Sissach.

The 1.5 km long underground section passes through typical Jura formations. Two-thirds of the tunnel passes through rock formations exhibiting swelling potential. Half of the length of the tunnel passes through keuper containing sulphates in the form of anhydrite and gypsum, as well as weathered marl. The 400-m long west section of the tunnel lies in the Upper Keuper with a lower depth of overburden of 25 to 40 m.

After driving a small pilot tunnel, the Chienberg tunnel was built using the sprayed concrete lining technique combined with the top heading and benching method of advance. The ring closure of the inner lining was completed within 25 weeks at the latest or within a maximum distance of 450 m behind the working face of the top heading to anticipate the potential swelling behaviour in the tunnel floor.

An almost circular tunnel profile was the option retained to anticipate corresponding swelling pressures. Its inner lining had to be 0.7 m thick at the crown and 1.1 m at the invert and carried out in C55/67 concrete.

Figure 11 Chienberg Tunnel, circular tunnel profile (first design), [16]

Roughly four months after the inner lining was completed, uplift was recorded in the entire tunnel. One year later, the maximum heave in the crown reached 80 mm, while another one was noticed at ground surface level, though of a smaller magnitude. At that point, the maximum rate of heave in the tunnel crown reached 4.5 mm/month.

Figure 12 Chienberg Tunnel, tunnel profile after rehabilitation (second design), modular yielding system [16]

Based on the observations mentioned above, various alternatives were considered for rehabilitating the tunnel. Without intervention and constructive modifications, it was impossible to put the tunnel back into operation. Finally, the concept of Modular Yielding Support developed by Prof. Kalman Kovári was adopted. While heaving subsisted at the base and a new equilibrium with the overburden was maintained, the elements used definitely limited the load. These precast elements were 100 cm high, with a diameter of 90 cm. Once the yielding capacity of an element is reached, it can be replaced.

Figure 14 Chienberg Tunnel, rehabilitation works (second design), [16]

Rehabilitation was carried out over 370 m in the western part and 60 m in the eastern part of the tunnel. Extensive demolition and reconstruction work was necessary near the sidewalls and the base. Measurements to date have confirmed the success of the operation. Since the yielding elements have been installed, the heave could be brought to a standstill (Fig. 15).

Figure 15 Chienberg Tunnel: tunnel floor reinforced with sliding anchors after rehabilitation

5.  References

[1]   Huder, J. und Amberg, G. 1970. „Quellung in Mergel, Opalinuston und Anhydrit“; Schweizerische Bauzeitung, 88. Jahrgang Heft 43, 22. Oktober 1970

[2]   Schillinger, G., 1070. „Die Felsdrücke im Gipskeuper beim Bau des Belchentunnels“
Heft Strasse und Verkehr, 56. Jahrgang, Nr. 10/70

[3]   Grob, H. 1972. „Schwelldruck im Belchentunnel“, Internationales Symposium für Untertagebau, Luzern, 11. – 14.Sept. 1972, pp. 99 – 119

 [4]  Madsen, F. und Kahr, G, Zürich, 1985. „Quellende Gesteine als Ursache von Problemen im Untertagebau“, Mitteilung der Schw. Gesellschaft für Boden- und Felsmechanik, Nr. 111, Studientagung 10./11. Mai 1985. Zürich [http://geotechnikschweiz.ch.vtxhosting.ch/wp-content/uploads/2017/04/Heft115.pdf]

[5]   Kovári, K., Amstad, Ch., Anagnostou, G., Eidg. Techn. Hochschule, Zürich, 1987. „Tunnelbau in quellfähigem Gebirge“, Mitteilung der Schw. Gesellschaft für Boden- und Felsmechanik, Nr. 115, Frühjahrstagung 7. Mai 1987. Biel [http://geotechnikschweiz.ch.vtxhosting.ch/wp-content/uploads/2017/04/Heft115.pdf]

[6]   Madsen, F. Nüesch, R., Zürich, 1989. „Quellende Gesteine, Quellmechanismen und die Bestimmung massgebender Quellparameter im Labor“, Mitteilung der Schw. Gesell-schaft für Boden- und Fels-mechanik, Nr. 119, Studientagung 6./7. Mai 1989, Delémont
(in German) [http://geotechnikschweiz.ch.vtxhosting.ch/wp-content/uploads/2017/04/Heft119.pdf]

[7]   Amstad, Ch.; Kovári, K., 2001. „Untertagebau in quellfähigem Fels“, Institut für Geotechnik, Professur für Untertagebau; ETH Zürich, Auftraggeber: UVEK, Bundesamt für Strassenbau ASTRA

 [8]  Hofer, R., Chiaverio, F., Kovári, K., 2007. „Chienbergtunnel Sissach – Tunnelhebung infolge Quellen“, pp. 95 – 100, SIA Dokumentation D 0222, Band 6, Swiss Tunnel Congress 2007, Fachtagung für Untertagebau, 21. Juni 2007 in Luzern

[9]   Rota, A., 2007, Wettbewerb „Instandsetzung Adlertunnel“, Tunnel SBB Pratteln – Liestal, TEC21, Heft 41/2007

[10] Anagnostou, G., ETH Zürich, Ehrbar, H., Swiss Tunneling Society STS, pp. 346 – 349, FGU, 2013. „Chienbergtunnel, H2 Bypass Sissach“, Tunnelling Switzerland

[11] Kobler, Th., Arndt, N., Hug, R., Hertweck, M., Pagliari, G., 2017, „Third Tunnel for Gubrist – The Core Part of the Extension of the Zurich Northern Ring Road“, Band 16, pp. 16-29, FGU, Swiss Tunnel Congress 2017, Fachtagung für Untertagebau, 30. Mai 2017 in Luzern

[12] Hertweck, M., Pagliari, 2017, „Four-Track Expansion between Olten and Aarau – Eppenberg Tunnel/CH – Challenges Faced by Client and Designer“, Band 16, pp. 30 – 41, FGU, Swiss Tunnel Congress 2017, Fachtagung für Untertagebau, 30. Mai 2017 in Luzern

[13] Bühler, M., Zimmermann, A.; 2017, „Swelling Ground and Protection of the Thermal Spa – Challenges Faced in Designing the New Bözberg Twin-Track Tunnel“ Band 16, pp. 42 – 55, FGU, Swiss Tunnel Congress 2017, Fachtagung für Untertagebau, 30. Mai 2017 in Luzern

[14] Böheim, S., Chiaverio, F., Straumann, U.,“A2 Renovation Tunnel – Challenging Geology and the Resulting Project requirements, 2017, Band 16, pp. 56 – 67, FGU, Swiss Tunnel Congress 2017, Fachtagung für Untertagebau, 30. Mai 2017 in Luzern

[15] Madsen F.T., Flückiger A., Hauber L., Jordan P., Voegtli B., New Investigations On Swelling Rocks In the Belchen Tunnel, Switzerland, 8th ISRM Congress, Tokyo, Japan, September 1995.

[16] Bäumle M., Chienbergtunnel, Tunnel in anhydritführenden Gebirge, Kolloquium ETH Zürich, 2014