The Kuala Lumpur Light Rapid Transit System 2 Project will provide a new public transport service in the Malaysian capital. The route is 29km long and runs from the suburbs in the south west to those in the north east. The section through the city centre is in tunnel, passing close to the KLCC Petronas twin towers. The rest of the route is on viaducts. There are 24 stations, five of which are underground.
The client for the project was Projek Usahasama Transit Ringan Automatik Sdn Bhd (PUTRA). The Renong Group had set up PUTRA specifically to operate the LRT concession and had kept it as a wholly owned subsidiary. PUTRA established Pengurusan LRT (PLRT) as its project manager, and employed HHDC as its technical adviser. The latter consisted of HSSI (a Malaysian consulting engineer), Halcrow and De-Leuw Cather. Its role was to provide technical advice, co-ordination and construction supervision across the project.
The whole viaduct section and two underground stations were carried out under separate design and construction contracts involving local designers and contractors. HHDC provided technical guidance, co-ordination of the designers, and site supervision of the contractors for these contracts. The remaining three underground stations and all the tunnelling works were built under design+construct arrangements. In these cases, HHDC undertook several roles:
- Technical co-ordination
- Design review and approval to ensure the contractor’s proposals met the client’s design criteria. This included all drawings, calculations, specifications, method statements and quality procedures
- Site supervision (particularly regarding quality, safety and environmental issues).
Tunnel works
The total length of the underground section under the city centre (Fig 1) was 4.4km. Tunnelling works were divided into three separate design+construct contracts:
- Western cut+cover tunnel: 500m of cut+cover at the western portal to the underground section.
- Bored tunnel Contract 1: 1.6km of 4.9m i.d. twin bored running tunnel, Munshi Abdullah Intervention Shaft and a drainage sump
- Bored tunnel Contract 2: 1.8km of 4.9m i.d. twin bored running tunnel, Kewalram Intervention Shaft, a drainage sump and 100m of cut+cover at the eastern portal
Each of the design+construct contracts included a ‘performance specification’ which set out the design and construction requirements.
Ground conditions
The bedrock in Kuala Lumpur varies, with both intact schist and limestone being encountered. The limestone has karstic features and its top surface is extremely irregular. Above the rock are two soft deposits: the Kenny Hill Formation – the material remaining after erosion of the schist, which is light grey to green, stiff to hard clayey silt, with traces of gravel (rock fragments); and alluvium – loose clayey silt or silty sand. The groundwater level is close to the surface. An added complication is that there is, in places, artesian water pressure within the limestone.
The tunnels were driven at a 11-19m deep, mainly in the Kenny Hill Formation. However, the alignment passed through rock in places and alluvium was sometimes encountered in the crown of the tunnels. Under the design+construct contracts, the contractors carried all risks associated with the ground conditions.
Bored tunnels
Each of the two bored tunnel contracts used a different type of TBM. Three EPB machines designed by Hazama and manufactured in Japan were used on Contract 1, which was driven almost entirely under the Klang River. These machines had a full face cutterhead fitted with picks and disc cutters and were launched from diaphragm walled temporary shafts.
The three TBMs completed four drives. At the end of one of the drives, the machine entered a reception shaft and the complete machine was recovered. At the end of the other three drives, the tailskin was left buried and only the internal machinery was recovered. During the drives, the machines generally encountered the Kenny Hill Formation but also coped with some areas of intact rock.
In one location, abnormal surface settlement was experienced. The tunnel face was in stiff material (Kenny Hill Formation), but with little cover. The suspected cause of the settlement was that the cover was loosened, allowing movement of alluvium directly above the crown. The solution was to treat the crown with a chemical grout, which strengthened the soft ground. The grout was injected primarily from the TBM, but additional grouting was carried out from the surface to ensure that no voids remained. The TBM then proceeded without further problems.
The machines were articulated to negotiate100m radius curves. The alignment design was the responsibility of the contractor, based on the performance specification regarding train performance.
Compressed air machines
The two TBMs that were used for the excavation of the bored tunnels on Contract 2 were manufactured by Wirth–Howden and had a bulkhead to enclose the face chamber, which was supplied with compressed air to control water ingress. Air pressure during the drives was 0.7-1.1 bar. Breasting doors were also fitted to provide support to the face. Excavation was carried out by backhoe. The excavated material was extracted from the face chamber with a screw conveyor and transported out of the tunnel using Volvo rubber tyred dumptrucks. Control of the excavation took place from the machine from either inside or outside the face chamber. A laser guidance system was used to control the tunnel alignment. The TBMs on this contract were articulated and were required to negotiate curves of 130m radius.
This type of machine was selected because the alignment passed through a number of temporary ground anchors (from the basement construction for the KLCC Petronas twin towers) that had to be removed as the machine progressed. Such a machine allowed access to the face, from where the anchors could be cut using hydraulic shears.
As with the EPB machines, the compressed air machines were launched from temporary shafts. Ground treatment, mainly jet grouting, was used to stabilise the alluvium directly outside the shaft. The breakouts from the shafts were performed by hand with timber face support. A temporary seal was installed around the ‘eye’ for each breakout to allow the compressed air to be applied to the face before the first lining rings were erected and grouted.
These TBMs also encountered soft ground in the crown, which resulted in loss of material and surface settlement. Face stability was recovered by timbering and grouting from the face and surface. Modifications to the breasting doors were made, which improved the stability that they afforded to the ground.
In one area, the tunnel alignment passed through an area of intact limestone which the machines were not designed to deal with. This was overcome by drilling into the rock and splitting it using expanding grout to allow the machines to progress, but it had an adverse effect on the excavation programme.
Surface settlements were monitored along the bored tunnel route and building damage assessments prepared for adjacent structures in the zone of influence. Some slight building damage was reputed to be caused by the tunnelling and was repaired.
|
“It was decided to use loads based on observations in stiff clays elsewhere in the world” |
Emergency intervention shafts
The Munshi Abdullah emergency excavation shaft (Fig 2) provides access to the running tunnels for the emergency services and provides an escape route. At the top of the shaft, a building houses fans which provide positive air pressure in an emergency to prevent build-up of smoke in the shaft. The shaft is separated from the running tunnels by fire-rated doors which are designed to resist the air pressure created by passing trains. At low level, a drainage sump is incorporated into the shaft works. Infiltrating water is pumped up the shaft to discharge into surface drainage.
The permanent shaft was 6.3m i.d. and 18m deep. From the base of the shaft an access tunnel extended out 24m over the top of the inbound running tunnel and under the Klang River. This tunnel was horseshoe shaped, with an internal width of 3.4m. At the end of this access tunnel, a sloped stair shaft was mined down between the running tunnels and linked to them by a cross passage. The sump, 1.8m in i.d. and 8m deep was sunk from the invert of the cross passage.
The intended construction sequence for the shaft was as follows:
- Create a temporary shaft using secant piles
- Excavate to full depth
- Cast the permanent lining inside it
Thirty-eight 725mm diameter secant piles were used to form a temporary circular shaft of 7m i.d. The secant piles were designed to resist the ground loads by ‘hoop action’. The hoop action of the temporary piled shaft depended on the shaft circularity being maintained with depth. At the outset, there was concern that the pile vertical tolerance to achieve this (1 in 200) was too onerous.
It was agreed with the contractor that excavation within the temporary piles would start but that the shaft would be regularly surveyed to determine the actual circularity that had been achieved. When the excavated depth had reached 12m, it was apparent that the allowable tolerance on pile verticality had been exceeded and excavation ceased. A pile vertical tolerance of approximately 1 in 75 had been achieved in practice. The solution was to cast the shaft permanent lining in the section that had been excavated so providing sufficient additional support to allow excavation to continue to the base. Then the base slab was cast, with relief pipes installed to prevent water pressure building up under it. The final section of the shaft lining was then completed and the relief pipes sealed.
Tunnel excavation and temporary works
Once the permanent shaft lining was complete, excavation for the access tunnel could begin. The access tunnel was to be excavated under the Klang River, with 6m cover to the river bed. Excavation was mainly through the Kenny Hill Formation, which, in this location, was a hard clayey silt. However, for part of the drive there was alluvium present in the crown.
To facilitate excavation through this alluvium, jet grouting had been used to improve the soil in a zone 3m above the crown for the entire tunnel length. Where the tunnel passed under an existing retaining wall, chemical permeation grouting was also used to supplement the jet grouting.
Excavation was performed by mini-digger and by hand. Temporary support used steel arches at 1m centres with 150mm thick shotcrete and steel reinforcement mesh. The face was split into a crown and an invert, with 1m advance lengths; crown advance was 3m ahead of the invert. ‘Elephant’s feet’ were used to provide support to the crown when excavating the invert. When water ingress was encountered, relief pipes were installed through the shotcrete to prevent water pressure build-up.
Design of the temporary lining was initially based on ground loads using a Terzaghi formula. However, this produced design loads that seemed to be low by inspection. It was therefore agreed to use loads based on observations in stiff clays elsewhere in the world. The analysis of the temporary lining was based on a 2D plane frame model with soil springs included to model the soil-structure interaction.
Shotcrete was supplied as a pre-bagged mix by MBT Malaysia. Based on an estimation of the advance rate and the age of loading of the lining, the 24h strength of the mix was used for design of the temporary lining. Regular coring and strength testing confirmed the quality of the shotcrete mix was being maintained.
Convergence monitoring was carried out using a tape extensometer. The results showed that the tunnel was stable, with maximum deformations similar to the predicted values (approximately 15mm).
Throughout excavation, temporary support was maintained in the running tunnels, consisting of steel ring beams, packed against the lining using bagged dry mix shotcrete, which extended along the length of running tunnel that was influenced by the access tunnel excavation. Flood doors were installed to mitigate the risk of flooding during excavation under the Klang River. The consequence of such a flood would have been extremely serious, since the running tunnels were linked to the remainder of the underground section.
Tunnel permanent works
Once all the tunnel excavation was completed, the permanent lining was cast. The design of this lining, incorporating the complex junctions, was based on a 3D finite element analysis of the concrete structure. Spring supports were used to model the ground and the soil-structure interaction. The ABAQUS finite element analysis package was used. The permanent lining was designed to resist the full vertical soil overburden. Initial horizontal earth pressures were based on a range of ko values of 0.5-1.5. The final lining comprised 400mm thick grade 40 concrete with steel bar reinforcement of approximately 65 kg/m3. Links were used at areas of high shear, at the junctions and in the tunnel knee. Vent and grout pipes were cast into the crown and facilitated grouting after the concrete was cast. A hydrophilic strip and a grout injection hose were incorporated into all construction joints.
Kewalram Shaft
This shaft was sunk between the two running tunnels, with short cross passages linking with the running tunnels. The shaft is square, the walls constructed by diaphragm walling. Excavation proceeded inside the shaft, with steel temporary propping at several levels. Once the excavation was complete, concrete fillets were cast in the corners of the shaft, linking the four wall panels together. The reinforcement for these fillets had been cast in the diaphragm walls, and was then bent out into place. Shaft excavation could have been less costly and with less risk of working at height if the fillets had been cast sequentially with the shaft excavation, so removing the need for the steel temporary propping. However, the contractor needed to complete the shaft excavation and tunnel connections in time to hand the tunnels over to the trackwork contractor.
Owing to programme constraints, the cross passages had to be completed while the TBMs were still working on these drives. This required uninterrupted access to the TBMs and therefore minimal temporary works to support the running tunnel during this operation. A staged installation of a steel opening frame, together with a concrete surround behind it, made this possible.
Cut+cover tunnels
There were two cut+cover tunnels on this project: at the west portal and the east portal. The latter was a conventional reinforced concrete box structure built in open cut and then backfilled. The cut+cover tunnel at the west portal was, however, more complex (Fig 3). It was 500m long and was constructed adjacent to the Klang River. At the northern end, it was fully buried, whereas at the southern end, the tracks were elevated and the cut+cover tunnel provided an abutment for the first above ground viaduct span.
It was designed and constructed in three sections: a diaphragm wall, top down section, the deepest part of the tunnel at the northern end; a middle section, which provided a link between the diaphragm walled and open cut sections (this part of the tunnel was constructed under the Leboh Pasar Besar Road Bridge, which was kept in operation throughout construction);. an open cut section, through which the rail alignment transferred from below ground to above ground.
The rail alignment in this part of the route was very restricted, passing between the Klang River to the west, and ‘shop-house’ buildings to the east – approximately 100 years old and rather sensitive. To route the alignment in this location, the two tracks were placed one above the other at the northern end of the tunnel. At the southern end, the two lines are at the same level.
The first section of the LRT was put into operation in 1998; the remainder is due to be running later this year.
Related Files
Cross Section of Shaft
Kuala Lumpur LRT System
Sequence diagram
