Longitudinal Flow of Water Along Hose Assemblies Through Inner Tunnel Linings

For the construction of inner tunnel linings exposed to water pressure, both constructions made of waterimpermeable concrete and waterproofing with plastic sealing membranes are used. Both sealing options have in common that hose lines are led through the inner tunnel shell along a tunnel block. The article published in tunnel 3/22 examines the question of the extent to which water can flow along the longitudinal direction of the hose lines as a construction-related effect.

1 | Arrangements of hose lines in the tunnel inner lining area
Credit/Quelle: FH Münster, IuB
1 | Arrangements of hose lines in the tunnel inner lining area
Credit/Quelle: FH Münster, IuB

1 Introduction

In the case of inner tunnel shells made of water-impermeable concrete, internal joint waterstops are usually integrated at the transition to the neighbouring blocks. These joint tapes are provided with injection hoses as a post-sealing option, which are led through the shell at various locations, for later correction.

For tunnel linings with plastic waterproofing membranes, hose lines are routed from the sealing through the tunnel lining for targeted subsequent injection in the event of damage. Furthermore, the ZTV-ING [1] stipulates that specific testing and injection installations are provided between the tunnel lining and the waterproofing. Figure 1 shows the arrangement of hose lines at different positions in the area of the inner tunnel lining.

Thus, in both cases, hose lines are systematically and regularly routed through the inner lining, which may be in contact with the accumulating water. Unambiguous technical specifications on how these hose systems are to be designed in detail are currently not available.

2 | Leakages at injection hoses
Credit/Quelle: FH Münster, IuB

2 | Leakages at injection hoses
Credit/Quelle: FH Münster, IuB

Figure 2 shows a water leakage at injection hoses, which were placed in the tunnel lining at internal joint waterstops. Such effects can also be observed with hose penetrations on the inside of tunnels, which are sealed with plastic waterproofing membranes. 

Due to such leakages, the required watertightness class according to ZTV-ING Part 5, Section 5, Chapter 2.2 [1] and Ril 853, Module 853.4101 [2] cannot be guaranteed. Therefore, remedial measures often have to be initiated in these areas. In addition, such leaks regularly lead to discussions on construction sites about the cause and the responsibility for remedying the defects. 

Furthermore, there is previous experience that water inflows were also observed while the groundwater level, determined by level measurements, was still below the tunnel floor. This seems to be a contradiction. However, since only a calculated water-cement ratio between 0.23 and 0.40 is necessary for complete hydration of the concrete and higher w/c ratios are usually chosen due to better workability [3], it is possible that either excess residual water remains in the shell or water rises during vault concreting due to sedimentation effects and accumulates on the outside of the shell.

At the Institute for Underground Construction (IuB) at Münster University of Applied Sciences, various investigations were therefore carried out to examine the extent to which construction-related water circulation along the longitudinal direction of the hose lines (hereinafter referred to as “longitudinal flow”) is possible. The focus was on investigations into the influence of the geometry, the material of different types of hoses and the shrinkage behaviour of the concrete.

2 Test Setup

2.1 Scope of Investigation

A total of seven different hoses were chosen for the tests, which differ in surface characteristics, material and diameter, but are currently all used in the production of tunnel liners. To simulate an inner tunnel lining, the hoses were embedded into 40 cm thick concrete test specimens. For this purpose, two different concrete mix designs with different shrinkage behaviour were used. After filling the hoses with polyurethane resin or a cement slurry, the longitudinal flow of the hoses was tested in accordance with DIN EN 12390-8 [4] (Fig. 3).

3 | Representation of the experimental set-up (actual and schematic)
Credit/Quelle: FH Münster, IuB

3 | Representation of the experimental set-up (actual and schematic)
Credit/Quelle: FH Münster, IuB

The test pressure was set at 5 bar, as pressurised water-retaining tunnels with such water pressure stresses have already been constructed and such stresses would be present in the event of faults, leaks, etc.

2.2 Utilised Materials

2.2.1 Hose Lines

The hoses used can be classified into four different types:

4 | PVC-fabric hose
Credit/Quelle: FH Münster, IuB

4 | PVC-fabric hose
Credit/Quelle: FH Münster, IuB
1. The first type of hose is a (soft) PVC fabric hose (Fig. 4) with a smooth surface without grooves or profiling. Three different sizes (outer diameter 12, 20 and 26 mm) were used in the tests. These hoses are generally used both as testing and injection systems and as feed lines for grouting hoses in the area of joint waterstops.

 

5 | Cable conduit
Credit/Quelle: FH Münster, IuB

5 | Cable conduit
Credit/Quelle: FH Münster, IuB

2. Figure 5 shows flexible PVC pipes that are often used as cable conduits or empty conduits. Here, hoses with outer diameters of 20 mm (black) and 25 mm (grey) were tested. The height of the ribs is about 3 mm for the larger tube and about 2.5 mm for the smaller tube; the clearance between them is about 1 mm for both. The inner surface is the negative of the outer surface.

 

6 | Spiral hose
Credit/Quelle: FH Münster, IuB

6 | Spiral hose
Credit/Quelle: FH Münster, IuB

3. Another type of hose used is made of plasticised PVC with a reinforcing spiral (Fig. 6). This hose is smooth on the inside and corrugated on the outside, and the outside diameter is 26.5 mm. The 2 mm wide reinforcement is spiralled at intervals of approx. 7.5 mm and the rib height is 3 mm.

 

7 | Injection hose
Credit/Quelle: FH Münster, IuB

7 | Injection hose
Credit/Quelle: FH Münster, IuB

4. The fourth type of hose is an injection hose (Fig. 7), which is usually used to seal construction joints between two concreting sections. This hose is provided with four rows of slits in the longitudinal direction, through which the injection material emerges during a grouting process. The outer diameter of the hose is 13 mm.

2.2.2 Concrete

Two different concrete mixes were prepared for the tests, which differed significantly regarding the parameters of water penetration resistance, strength and shrinkage behaviour. The strength class listed in Table 1 was determined according to DIN EN 206-1 [5]; the water penetration depth represents the mean maximum water penetration depth of the concrete (determined according to DIN EN 12390-8 [4]). A Portland cement CEM I 52.5 N and grading curve A/B were used for the tests.

Table 1 | Properties of the concrete mixes

Table 1 | Properties of the concrete mixes

The first concrete formulation was designed as a water-impermeable concrete [6]. However, due to the high cement paste content, a strong chemical and autogenous shrinkage was to be expected. This basic shrinkage was to be reduced in the second concrete mix design, which is why the w/c ratio was increased to 0.80 and thus the cement paste content was reduced by about one third. This water-cement ratio inevitably increases the proportion of capillary shrinkage (early shrinkage), which, however, should have no influence on the longitudinal flow of the hose lines. In the first concrete formulation, a superplasticiser was also added, as the concrete was otherwise not sufficiently workable due to the low w/c ratio.

2.3 Test Procedure

Using the described hoses, three test specimens with the dimensions 45 x 12.5 x 12.5 cm were created for each concrete mixture. The test specimens were stripped of their formwork after 24 hours and then stored in a water bath for another 27 days. After the water storage, the test specimens were cut to a length of 40 cm on one side, which ensured a comparable test surface for all test specimens. The tubes were then filled with either cement paste or polyurethane resin. For this purpose, the cut surface was sealed with adhesive tape and the injection material was then filled into upright test specimens.

After a further 24 hours, the specimens were placed in a test rig (cf. Fig. 3) and then subjected to 5 bar water pressure for 72 ± 2 h in accordance with DIN EN 12390-8 [4]. If the test specimens did not show any leaks during this three-day test period, they were then split to measure the water penetration depth into the concrete.

3 Examination Results

The tests showed that all smooth hoses (PVC fabric hoses as well as injection hose) exhibited longitudinal flow immediately after the water pressure was applied and before the test pressure of 5 bar was reached. This was characterised by leakage through the side facing away from the water. Some PVC fabric hoses were even partially pressed out of the concrete on the bottom side of the test specimen by the water pressure, resulting in a cylindrical hollow body on the upper side of the specimen (Fig. 8, right).

8 | Water penetration depth into the concrete, and hose
displaced by water pressure
Credit/Quelle: FH Münster, IuB

8 | Water penetration depth into the concrete, and hose
displaced by water pressure
Credit/Quelle: FH Münster, IuB

In contrast, the cable conduits did not show any leaks. After splitting the test specimens, it became apparent that the water penetration depth at the thicker cable protection conduit in the first concrete formulation did not exceed the water penetration depth of the concrete determined in the reference tests (cf. Table 1). The cable conduit with the smaller diameter had an average water penetration depth of 2.5 cm, which was approximately twice as high. In addition, the splitting of the test specimens showed that both concrete mixtures filled the fine interstices of the hose and thus formed a tight bond with the corrugated surface (Fig. 9).


9 | Water penetration depth at the top with corrugated hose
Credit/Quelle: FH Münster, IuB

9 | Water penetration depth at the top with corrugated hose
Credit/Quelle: FH Münster, IuB

Of the spiral hoses, longitudinal flow was already noticeable in two out of six test specimens during the three-day test period. After splitting the initially tight test specimens, it could be seen that the water penetration depth along the hose in the first concrete mix was about 27 cm (Fig. 10). The second concrete mix also showed comparable results, meaning that all test specimens that were leak-proof during the test period exhibited extremely high water penetration depths.

10 | Water penetration depth with spiral hose
Credit/Quelle: FH Münster, IuB

10 | Water penetration depth with spiral hose
Credit/Quelle: FH Münster, IuB

The results of the tests are summarised in Table 2. The yellow marked leakage of the injection materials inside of the hose only occurred when polyurethane resin was used. This could possibly be caused by water residues from sawing on the inner wall of the tube when the test specimens were filled, which reacted with the polyurethane resin to form urea or polyurea with the emission of CO2. So, in the reacted state, the tightness would be negatively affected by the pores formed [7]. However, in the case of the smooth hoses, the longitudinal flow on the outer surface of the hose was so distinct that it happened before any water was discharged from the inside of the hose.

Table 2 | Results overview

Table 2 | Results overview

3.1 Evaluation

The tests showed that the best hose types for impermeable constructions are the cable conduits under the specified boundary conditions. The surface structure, which extends the path of the water, is comparable to the labyrinth principle of waterstops [8] and ensures high impermeability to penetrating water. The tests showed that both concrete mixes formed a tight bond with the corrugated hose. A possible withdrawal from the individual grooves of the hose induced by the shrinkage of the concrete could not be detected during the tests carried out.

The comparison of the two cable protection conduits in use showed that the 20 mm diameter conduit exhibited a higher water penetration depth of 1.5 cm on average. This could be due to either a shorter leakage path or a reduced depth of fill in the grooves. However, since the test showed that the concrete could spread sufficiently between the grooves – regardless of the concrete formulation – the former reason is probably decisive. Nevertheless, the water penetration depth of both hoses is so low after three days that this type of hose can probably be used without the occurrence of leaks. Long-term studies and further experimental variations will be necessary, though, to substantiate this statement.

The results were identical for all PVC fabric hoses and the injection hose. Here, all hoses showed pronounced longitudinal flow, which can be attributed to the smooth hose surface and the associated insufficient bond with the concrete. This was also confirmed by the fact that some hoses were pressed out of the test specimens by the applied water pressure. Therefore, leaks could occur in practice when using these types of hoses.

The results of the spiral hoses, on the other hand, were diffuse. Longitudinal flow on the outside of the hose going through the 40 cm thick concrete test specimen occurred within the test period and water penetration depths of up to 27 cm could be observed after three days. Due to the spiral-shaped reinforcements of the hose, the water was prevented from direct longitudinal flow. However, since the water was able to penetrate the test specimens relatively quickly along the hose, only a conditional recommendation for the use of spiral hoses for penetrations of tunnel shells can be given on the basis of the results.

At this point it should be mentioned that the injection hoses were filled with injection material without pressure during the tests. The pressure exerted on the concrete, especially in the case of softer hoses such as spiral hoses, could reduce or prevent possible leakage. Furthermore, the investigations carried out were intended to show the basic problem; for well-founded conclusions, a significantly larger number of samples as well as varying test executions are required.

4 Summary

In practice, various damp spots could be located in the area of the hoses on the reveal surfaces of the tunnel shell. This construction-related possibility of longitudinal flow due to the hydrostatic load was also shown in the investigations carried out. It was shown that the type of tube has a decisive effect on the longitudinal flow.

In tunnelling, in many cases hose lines run through the inner tunnel lining. There are currently no clear specifications for these in the technical regulations. The fundamental problems with the use of unregulated and therefore different hose systems could be demonstrated here.

The tests showed that the cable conduits have a high resistance to water penetration due to the bond with the concrete. However, it is questionable whether the spreading of (especially cement-based) injection materials is possibly influenced by the likewise corrugated inside surface.

Moreover, modifications were made to hoses with longitudinal flow in initial investigations at Münster University of Applied Sciences to remedy the described leaks. The first tests on this showed positive results. Due to these still unanswered questions and for the purpose of drawing well-founded conclusions, this topic should be analysed in more detail in further tests (including long-term tests).

References/Literatur
[1] Bundesanstalt für Straßenwesen: ZTV-ING – Zusätzliche Technische Vertragsbedingungen und Richtlinien für Ingenieurbauten.
Fassung 01/2018.
[2] DB Netz AG: Richtlinie 853: Eisenbahntunnel planen, bauen und instand halten. Fassung 03/2011.
[3] Locher, F. W.: Zement: Grundlagen der Herstellung und Verwendung. Verlag Bau + Technik, Düsseldorf, 2000.
[4] DIN EN 12390-8: Prüfung von Festbeton – Teil 8: Wassereindringtiefe unter Druck. Beuth Verlag, Berlin, 2019-08.
[5] DIN EN 206: Beton - Festlegung, Eigenschaften, Herstellung und Konformität. Beuth Verlag, Berlin, 2021-06.
[6] Deutscher Ausschuss für Stahlbeton: Wasserundurchlässige Bauwerke aus Beton (WU-Richtlinie), Fassung 12/2017.
[7] STUVA (Hrsg.): Abdichten von Bauwerken durch Injektion (ABI-Merkblatt). 3. Auflage. Fraunhofer IRB-Verlag, Stuttgart, 2014.
[8] Hohmann, R.: Fugenabdichtung bei wasserundurchlässigen Bauwerken aus Beton. 2. Auflage. Fraunhofer IRB-Verlag, Stuttgart, 2009.
Acknowledgement/Danksagung
The authors would like to thank Marc Meissner from Naue Sealing GmbH & Co. KG for providing several hose systems.
 
Die Verfasser bedanken sich bei Marc Meissner von der Naue Sealing GmbH & Co. KG für die Bereitstellung einiger Schlauchsysteme.
x

Related articles:

Issue 2021-01

Safe Hose Lines for Modern High-Pressure Compressors

The use of air compressors with high operating pressures has become routine on many construction sites above and below ground, but this can easily lead to a dangerous situation or even accidents with...

more
Germany

Safe Hose Lines for Modern High-Pressure Compressors

The use of air compressors with high operating pressures has become routine on many construction sites above and below ground, but this can easily lead to a dangerous situation or even accidents with...

more
Issue 2013-03 Injection

Innovative Injection Hose protects the Nordhavnsvej Tunnel from Baltic Sea Water Pressure

The 4-lane Nordhavnsvej Tunnel is currently one of Copenhagen’s biggest construction projects and belongs to the Nordhavnsvej-Vej-og Tunnelenterprise major scheme. It underpasses the Danish capital...

more
Issue 2016-08 Sika Deutschland

Bonded Tape Sealing Solution for Plastic Membrane Transitions

The waterproofing of road or rail tunnels generally consists of waterproof concrete and/or concrete construction with a plastic waterproofing membrane. A change of the type of waterproofing within a...

more
Issue 2019-03 Sika Deutschland

Full-Area Grouting behind existing Shotcrete Layer

After more than 60 years of use, refurbishment of the road tunnel between Neuhaus and Spitzingsee bei Schliersee, Germany, was urgently necessary. The waterproofing directly behind the shotcrete inner...

more