Case Study: Putra LRT Tunnel

The project owner is KL City Council (Dewan Bandaraya, Kuala Lumpur) and the LRT tunnel asset owner is Syarikat Prasarana National Berhad (SPNB). The project concerned the construction of a dual carriage underpass above a LRT tunnel. The clearance between the road underpass to the LRT tunnel crown is approximately 1.9 metres after considering the depth of the base slabs of the structure.

The main restrictions imposed on the underpass works in relation to safeguarding the LRT tunnel included a limited tolerable displacement of the tunnel from vibration impacts during installation works e.g. installation of the king post, sheetpiles, grouting etc. Other restrictions included the limited tolerable displacement associated with the removal of overburden and the minimal imposed loading on the tunnel due to the underpass, either during construction or in its permanent state.

The geology of the site consisted mainly of rock mass of meta sedimentary rock formation of interbedded sandstones and shale of Upper Silurian-Devonian age (Kenny Hill formation) overlaid by alluvial deposits. The water table fluctuated between 6 to 8 metres of depth from the surface depending on the season.

The road underpass construction

The methodology for the road underpass construction consisted of permanent sheet piling system and king post coupled with steel strutting and temporary removable ground anchors and top down excavation. Refer to figures 1 to 5.

Tunnel instrumentation

A detailed study was carried out on the instrumentations required to ensure adequate monitoring and response times. The objective was to install an instrumentation system that was able to measure movements of the tunnel and to provide warning on excessive displacements.

Several constraints were identified;

  • Access to the tunnels for manual readings was limited to four hours per day (i.e. during non-operational hours of the trains);
  • Train operations introduce dynamic forces onto the installations within the tunnel;
  • Construction works on the underpass had to be carried out during the day, hence the need to obtain information immediately rather than in non-operational hours.

Based on the above considerations, real time monitoring systems were chosen. These systems were backed up by the following manual systems;

  • The automatic and real time displacement measurement using electro level beams (ELB) was used to find a measurement of displacement measurement at the tunnel crown level. The ELB system consists of a horizontal electrolytic level sensor with measurement range of 40 arc minutes and resolution of 1 arc second (0.01 mm for a 2 metre beam). The beams were two metres long aluminium and had a large temperature operating range of -20 to 50 degrees Celsius.
  • The automatic and real time displacement measurement basset convergence system (BSC) was installed to measure the tunnel ring deformation within 20 metres of the centre zone of the monitoring zone. These readings provide direction of movements in 2D. They were installed at five ring intervals (5 metres). The BCS systems consist of double electrolytic tilt sensors to an accuracy of 0.02 mm and measure along two axes (vertical and horizontal). Low profile design enables the system to fit close to the tunnel wall. The brackets and steel arms were stainless steel.
  • The purpose of thee electro-optical survey systems (ATMS) instrument was to serve as backup and checking system for the real time electrical sensors of the ELB and BCS systems. The ATMS consists of an automatic target recognition (ATR) system, which is fast and accurate in referring to a fixed datum. It is essentially an automatic total station with 0.5 inch angle measurement accuracy (0.05 mm), mini glass reflective prisms, APS Win data acquisition software and data analysis and report generation software.
  • The purpose of the Manual Level Survey Markers was to derive the absolute movement of the track slab. The system consists of survey markers at two metre intervals along the track slab with a survey accuracy of 0.01 mm.


Tunnel displacement

In stage one of the works, excavation to five metres, the tunnels settled or shape deformed, to about 3 mm at the crown during the installation of the king posts, preparation of the steel decking, and minor excavation works to -1 metres. The settlement at track level was less than 1 mm. It was noted that this was due to the additional surcharge introduced by heavy excavation machinery at the surface.

On completion of these works, the excavation to the 5 metre level commenced and the tunnel began to heave with further shape deformation, however the rate of displacements was within the predicted rate and the absolute displacements was within the predictions at crown and track. As the overburden was reduced, the surrounding ground relaxed and the stresses within the tunnel linings begin to redistribute due to the loading relaxation, and the tunnels began to deform under the new stresses.

Jet grouting

The tunnels continued to heave during the jet grouting works. It was concluded that this was the short and medium term results of the ground relaxation effects. The rate of heave slowed down as expected and within the predicted displacement estimates. There was no reduction of overburden during this stage.

The grout cap improved the strength of the soil surrounding the tunnels. The triple phase jet grouting technique was selected. The grout forms a body of cemented soil above and alongside to maintain the shape profile of the ground in order to minimise ground relaxation surrounding the tunnels. Dewatering wells were installed during this stage. This had the effect of reducing the flotation pressure during the temporary works.

On completion of the grout cap, underpass excavation resumed to a depth of 7 metres. The rate of heave was much lower than the stage one rate, which again corresponded to the predicted rate at tunnel crown location, but was higher at the track level. The lower heave rate was due to the counteraction between the inverted U shaped jet grout cap and the settlements, resulting from the drop in the water table by the dewatering works.

The track level, absolute heave value was 60 per cent higher than predicted. This was assessed carefully and it was concluded that further mitigation works were needed. The excavation sequence was changed to reduce the amount of excavation by zoning and limiting the excavation works and introduction of permanent loadings within these zones.

In stage three of the project, excavation from -7 to -9 metres, the tunnel heave continued and the rate of the displacement at crown was slightly above the predicted value. This was at a crucial time since the excavation depth was just above the crown level. The frequency of the data gathering was increased and assessments were conducted on a daily basis.

The measured heave at the crown level was 10 per cent (11.8 mm) above the predicted value and the measured value at the track was 138 per cent (6.2 mm) above the predicted value. However, based on the design assessments, the out of balance radial stress was within acceptable limits. To mitigate the risk it was recommended that base slab construction of the underpass should be expedited. The tunnels were monitored continuously during this stage with the real time reading intervals set at one minute.

Conclusions

The following graphs illustrate the monitoring data over the project period at various stages of the works. They provide a comparison between the predicted displacement based on finite element analysis, as well as actual displacement at both the crown and invert levels of the tunnel against various stages of the excavations. It is observed that theoretical predicted displacements from the analytical work and actual monitored displacement showed trends in correspondence with variations to the absolute values at track level.

The deviation was a cause of concern and back analysis was carried out with additional adverse soil parameters in order to predict the revised displacements. The results indicated that the absolute maximum of 15 mm would not be breached. Probable cause for the differences was soil profile variation within short intervals, as was observed during the tunnel drives.

The real time monitoring provided the necessary information for immediate response to an event and the manual monitoring provided essential back-up checks. It is noted however that the absolute values comparison between the real time and prism monitoring at the crown level showed a lower absolute displacement value for the real time monitoring than the prisms.

Post completion monitoring continued over a period of six months until negligible displacement. Pre- and post dilapidation surveys of the affected length of tunnel were carried out and no structural deterioration of the tunnels was observed.