The Mexico city – Toluca Interurban Rail project is a long-distance, medium-speed rail project that will connect the Valley of Toluca Metropolitan Area to the Valley of Mexico Metropolitan Area. It has a length of 57.8km, of which 4.76km will run through a twin tube tunnel. The project’s objective is to connect the municipalities and boroughs of both metropolitan areas in a way that enables the inhabitants of both of the communities to transit across the region quickly and safely, complementing the existing system road network.

Figure 1 illustrates the complete diagram of the railroad project.

Principal characteristics of the Tunnel

The tunnel starts at (p.k.) 36+172 of the rail line, avoiding the forested area of Sierra de las Cruces and the Desierto de los Leones National Park.

The tunnel finishes at p.k. 40+938, in Mexico City, next to the portal of the road tunnel that forms part of the Mexico City -Toluca tolled highway.

The Marquesa tunnel has a length of 4,766m. Both tubes have a segment lining, with an internal diameter of 7.5m and an external (excavated) diameter of 8.57m, resulting in an excavated section of 57.68m2. The separation between tubes is 24m between axes, which results in a little less than two tunnel diameters between the tunnel walls (15.43m). The segment rings have a thickness of 350mm, a length of 1,500mm. The rings have an external diameter of 8,200mm, each ring consisting of seven segments in a universal configuration of 6+1. Waterproofing is achieved with EPDM gaskets.

This universal ring allows for nineteen rotation positions, with the possibility of creating a turning radius of 1,500m. The segments have been manufactured with resistance values f’c of 30 and 40MPa.

Table 1, left, includes some of the main details of the project’s rings.

The tunnel has a maximum overburden of approximately 140m and a minimum of a little more than 14m. The total volume of excavated material is about 560,000m3.

For safety reasons, it was decided to adopt a one-directional, two-tube solution for the railway. A total of twenty-five crosspassages between the tubes are planned, which comprise twenty evacuation routes and five technical adits.

The alignment of the tunnel features a minimum curve radius of 1,500m, with a variable upward slope between 2% and 4% from the east portal (the Mexico City side).

In general, the tunnels have crossed two geologically distinct zones:

  • A section mainly in rock (representing 65% of the total length of the tunnel) with andesites and andesitic breccia.
  • A section in soils and/or soft rocks (representing 35% of the total length of the tunnel), and passes through a sub-vertical fault and tuff materials.

For these excavation conditions with heterogeneous topography, and also accounting for water loads expected throughout the entire route with minimal impact on aquifers, two multi-mode TBM will be used for the project.

These are capable of working in both open mode (for rocky conditions) and closed mode or EPB, to allow excavation in soils and/or rock with significant presence of water, and the capability of working up to a nominal pressure of 7 bar at the tunnel axis.

For this, the Section II Construction Consortium that carried out the tunnel construction work, which was made up of the companies: Ingenieros Civiles Asociados (ICA) and Construcciones y Trituraciones (Cotrisa) purchased two TBMs from the German manufacturer Herrenknecht, S-948 “La Marquesa” (right tunnel) and S-949 “La Mexiquense” (left tunnel).

During the design phase of the project, the team opted to design and execute the twin tube tunnel from the Mexico City side (the east portal).

As illustrated in Figures 2 and 3 opposite, a wide platform was prepared, enough for the facilities and assembly of the two machines. Meanwhile, Figure 4 shows an aerial view of the Toluca side west portal, or exit portal.

Obstacles encountered by the Tunnel

From the initial analysis of the geologic-geotechnical information of the project, and throughout the longitude of the tunnels, a series of problematic zones were identified.

Accounting for all zones, a priori, the most problematic zones or the ones with greater geologicalgeotechnical challenges, were those located between P.K. 36+850 and 37+200. High water loads of up to 7 bar of pressure were forecast for this section of the alignment.

The conditions and characteristics of this zone were among the most important limiting factors when writing the technical specifications for the TBMs.

TBMS during excavation

Table 2 on page 30 includes a summary of the principal characteristics of the two TBMs used for the project. 

I. Operation modes

During the construction phase, Sener decided that it needed variable TBM operation modes. On the one hand, the TBM would encounter zones that had a significant hydraulic load and/or an unstable face, which were susceptible to pressurisation.

In the other, the impermeable zones and that, because of the stability characteristics of the rock mass, could be undertaken without pressurising.

During the initial stages of excavation of the tunnels from the east portal, the TBMs were working in open mode and removing spoil via a conveyor belt, it was seen that the ground’s reaction to the applied thrust was insufficient.

The consequence of having this low reaction signified that it was impossible to keep the alignment on track, which at this point was the progression from an initial 2% upward gradient used for the initial metres of the tunnel, to the 4% required for the greater part of the drive.

It was decided to stop the TBM and change to EPB mode, using the screw conveyor to extract soil from the muck chamber. Working in EPB mode also allows for the pressurisation of the chamber and increases the reaction with the front of excavation, enhancing the ability to accurately direct the TBM.

Excavation using the screw conveyor configuration did not result in a significant increase in the wear of the cutter head tools.

Periodical inspections were held at atmospheric pressure, with efforts made to carry these out in zones of low water pressure, even though the equipment and procedures had been procured to enable hyperbaric inspections.

It was assumed that the ability to pressurise the front quickly and without changing the extraction configuration could be achieved; once the condition of having a possible increase in wear was discarded.

The screw conveyor configuration was eventually adopted to excavate the complete length of the tunnel, filling the chamber and pressurising it as needed. Figures 5, 6, 7, and 8 on pages 31 and 32 summarise the earth pressure data, the contact forces and penetration data of the TBMs.

The increase in pressure was observed in zones in which the water load levels necessitated it, mainly in the second half of the tunnel, stemming from the second excavated kilometre all the way along to a distance of 500m from the west portal.

Another observation was that the penetration was correctly maintained in the ranges of the 10-20mm/revolution throughout the tunnel, except at the beginning of the excavation in open mode.

This as a consequence of the lack of contact force and also as an attempt to keep the TBM on alignment. The results of attempts to work unpressurised can be observed in Figures 7 and 8, between rings 1870 and 1895 of the La Mexiquense drive. This methodology was eventually discarded upon noticing there was insufficient contact force.

Sener recommended the use of disc cutters with pressure compensators to help maintain the durability of the disc gaskets at earth pressures in excess of 3 bars.

II. Advancing through the problematic zone: A section of High Pressure

The advance of the tunnels in the areas that presented the greatest difficulty were primarily beset with risks deriving from geological changes, which resulted in excavation in mixed and unstable ground and an elevated water load of up to 7 bar at the tunnel crown.

For these high-pressure zones, any special consolidation and injection treatments from the surface were dismissed to avoid the risk of polluting shallower aquifers and springs. As a result, it was deemed adequate to traverse this entire section in EPB mode, adopting sufficient pressure to avoid loss of water flow through the tunnels.

The principal challenges that had to be overcome consisted of correctly conditioning the soil in the muck chamber and maintaining control over the status of the cutting wheel tooling.

As soon as excavation began in the high water load zone, spillage of material outside of the extraction system occurred. This happened because the water under pressure dragged disaggregated material through when opening the top gate of the screw conveyor. This phenomenon occurred principally in the first excavated metres, in this ‘Artesian Zone’.

Figures 9 and 10 on page 32 show the earth pressure values reached in TBM S-948 and TBM S-949, respectively, in the Artesian Zone, approximately between p.k. 36+900 and 37+200.

As a result of the high pressures reached, significant amounts of material fell to the bottom of the shield, which made it difficult to erect the segments until the machine was cleared.

The time spent waiting while performing manual cleaning could end up being quite long periods (in excess of four hours), which causes the material in the chamber to disaggregate.

Once the ring was built and when readying for excavation, the muck chamber was again working under heterogeneous conditions and caused new losses of material through the screw conveyor’s top gate, again creating new spills at the zone of the bottom shield.

It was possible to produce homogenous spoil after several tries, with less humidity in the chamber, through the intensive use of polymers and increasing the material density inside the chamber.

When the flow of material passing through the screw conveyor gate was stopped, excavation rates in the high water pressure zone was not significantly different to those carried out at lower pressures. This meant that the rates of advance were recovered and it was possible to get continuous excavation cycles.

Additionally, it was also planned having the option to inspect the tools in hyperbaric conditions but such an intervention, one that must be done at around 6-7 bars of air pressure, is thought to be inappropriate due to the limited time that would be avaialble to perform an intervention. The main difficulty lies in the need to perform a complete change of the disc cutters in the cutting wheel, which could extend the intervention by up to four weeks.

Therefore, it was considered adequate to continuously excavate, without any stops, while keeping a continuous surveillance of the principal excavation parameters to try to pass through the 300m of this Artesian Zone in one effort.

Some of the primary parameters that had to be continuously monitored were the contact force (kN), the torque of the head (kNm), the temperature of the bulkhead [ºC], and the penetration [mm/rev].

Figures 11, 12, 13, and 14 show the charts that represent the contact forces, as well as the torque of the cutter head, along with the penetration rates in the Artesian Zone obtained in each TBM respectively. These graphs show excavation with an elevated contact force relating to the earth pressure, a moderated torque, and a correct maintenance of the penetration throughout the section.

III. General Tunnel production rates

There was a prolonged learning curve, which aggravated a moderated production startup during the first months on the early stages of the work. From that moment on, production achieved the mean advances planned in the project of 12m/day, although it reached a maximum monthly excavation rate of 475m/month.

Figure 15 opposite illustrates the monthly production of each of the TBM compared with those values provided initially in the Project. In the graph, it can be observed that a decrease of production coincided in December 2017 and January 2018, due mainly to a program stop for preventative maintenance of the TBMs.

Sener undertook a complete analysis of the production times from the commencement of works all the way to the completion of excavation work for both tunnels.

Approximately 40% of the time was dedicated to the tasks of tunnel excavation and assembly of the segmental rings in both tunnels. Maintenance of the tunnelling machines represented percentages of 24 and 28%. Failures of the tunnelling machines, of the auxiliary equipment, and services extension procedures represented close to 14% of working time on both tunnels.

With regard to the consumption of tools, a total of 1,455 discs were changed between the two machines, with the with the distribution by lithology detailed in Table 3 below.