In the past five years construction has begun on three separate mass transit systems to link key areas of Bangkok, the capital of Thailand. The initial two were planned as overhead systems. One of these, the Bangkok Transit System (BTS), is now complete and operational. The other, the Hopewell scheme, has stalled because of financing difficulties.

The third line, now under construction, lies underground for its entire 22km length. Tendered as the Initial System Blue Line, it is now officially known as the MRT Chaloem Ratchamongkhon Line. The civil works for the project started in October 1996 and it is expected to be ready for operation by the end of 2004. The MRT will join the south of the city to the north and will link with the BTS at two locations and with the major state railway termini to provide an integrated transportation network.

Construction of the MRT is under the direction of the Mass Rapid Transit Authority of Thailand (MRTAT). Major civil works are divided into three main contracts, the south and north contracts each have nine stations and 11km of twin-bore tunnel, while the third package is a maintenance depot. All have been undertaken as design and construct contracts.

The south section was awarded to a joint venture (JV) of Bilfinger + Berger, Ch.Karnchang, Kumagai Gumi & Tokyu Construction (BCKT). Maunsell acted as the tender designer for the JV from March to May of 1996, then provided tender services for the north and depot contracts. The Underground Structures South (UGS) contract was awarded in October 1996 at a tender value of US$960M. The contract includes a variety of structures in addition to the tunnel and stations, including the intervention shafts, access tunnels, vent shafts and vent buildings.

Prior to the construction of the metro, tunnelling in Bangkok had been limited to relatively small diameter works using precast concrete segments or steel rings. The construction of the underground represents the first use of large-diameter bored tunnelling in the area. The UGS tunnels comprise 5.3m internal diameter precast segmental construction and have been excavated from two drive shafts using four EPB machines, two supplied by Kawasaki and two by Herrenknecht. All nine stations were designed as deep boxes constructed within diaphragm walls.

The contractor expects to complete the construction works for the contract at the end of 2002, with the tunnels due to be completed at the beginning of March. Following the award of the design and build contract to the JV at the end of 1996, Maunsell was commissioned by BCKT to carry out the detailed design.

Intervention shafts

Temporary support for the ventilation and passenger adits for the intervention shafts (IV shafts) are designed for forced ventilation and to enable emergency access to the running tunnels.

Because of the constraints of constructing a metro system under a densely populated city and the varying alignment of the running tunnels, a consistent design was not possible at all shaft locations. The position of both the shafts and tunnel alignment have been dictated by available land provisions. The stations have markedly different configurations and the layout of the three IV shafts reflect this.

The design of the IV shafts at the first location (chainage 17+770) incorporates two separate shafts, one either side of the parallel running tunnels. Land use at the surface prohibits the adoption of a single central shaft. Each shaft is connected to a single running tunnel via two horizontal adits, one providing passenger access and the other for ventilation. Both IV shafts were constructed using diaphragm walls, providing an octagonal shape. Because of the ground restraints, the locations of the shafts relative to the running tunnels resulted in lengths of the adits varying between 6m and 10m.

The layout of the IV shaft at the second location is very different. Here (chainage 16+100), the running tunnels were spaced further apart to allow the use of a single IV shaft, positioned centrally between the tunnels. The rectangular shaft, constructed using diaphragm walls, provides around 1.5m clearance to the running tunnel lining on either side. Hence the ventilation and passenger adits on each side are relatively short in length.

The tunnels at the third IV shaft location (chainage 10+900) link into a stacked station, bringing added design complications. A single, octagonal IV shaft, constructed using diaphragm walls, is similar to those at chainage 17+770. The shaft is again connected to each of the running tunnels via a ventilation adit and a passenger adit. Those connecting to the upper northbound tunnel are 12m in length, while the lower southbound tunnel adits are approximately 27m long. The lower adits run directly below the upper ones and pass under the northbound tunnel.

Intervention shaft adits

The geology consists of layered stiff clays and dense sands overlain by compressible soft marine clays, the adits were generally constructed within Bangkok First Stiff Clays. The provision for ground improvement measures were deemed necessary to control potential sand runs and water inflow through the sand lenses. Jet grouting was used in advance of the adit construction, and further treatment applied at the shaft diaphragm walls to fill voids and increase the strength of the ground as required.

In all, there are a total of 12 adits. The original scope for the design only included permanent lining for the IV shaft adits. This was carried out by German consultant Philip Schulz & Partners and took the form of a 300mm thick in-situ reinforced concrete lining designed for full overburden pressure and was identical for all three locations. Following a review by the JV, Maunsell’s commission was extended to provide design of the temporary lining. This is the only instance where detailed design for temporary works was conducted by the designer.

A series of meetings were held between contractor, Philip Schulz & Ptrs and Maunsell, it was agreed that the permanent support would remain as an in-situ reinforced concrete and the temporary support would be a sprayed concrete lining. It should be noted that the construction of the IV shaft adits was the first use of sprayed concrete for tunnel support under Bangkok.

Construction sequence

The construction sequence would have a major impact on the design of the temporary lining. Therefore details of the proposed sequence were included in the risk assessment and drawings; provisions were made for the different ground conditions and scenarios that may be encountered on site. A degree of flexibility in the construction sequence was proposed, allowing the contractor to omit or scale down a number of the phases based on working practices, experience and judgement on site.

The following is an outline of the construction sequence:

Before commencing construction, the area for each of the adit openings within the IV shafts was probed to determine ground conditions and ensure that any ground treatment carried out was adequate. Once initial probing was carried out and any necessary ground treatment completed the construction of each adit opening began.

Each adit was excavated in a series of headings and benches. First, the top half of the opening was made by stitch drilling and breaking out the necessary area of diaphragm wall, exposing the ground. The temporary support for the shaft opening was then installed, before the top heading of the adit was excavated. The temporary support was then installed to this first stage of the excavation before the lower area of diaphragm wall was broken. This was done using the same method as before. The bench was then removed, advancing the lower section of the excavation towards the front face of the top heading. A sealing layer of sprayed concrete was then applied to maintain stability. In non-cohesive ground, such as sand, a wedge of ground at the front face was retained to increase stability.

Forward probing was essential to confirm the ground conditions ahead and the effectiveness of any ground treatment measures, particularly when the presence of groundwater was likely. Because of the variable nature of the ground, a large number of probe holes were specified to minimise the risk of collapse. The extent of the probing covered a minimum 1m envelope around the adit and at least 5m ahead of the face with a nominal 1m overlap. Probing was carried out in the presence of an experienced engineering geologist and full records maintained. This included assessments of the ground conditions to determine the support required for the next stage.

Advance parameters

Each advance was set at a maximum length of 1m. To optimise the construction programme, the adits were advanced to a distance of one tunnel diameter from the running tunnels prior to the TBM drive. Once the TBM had passed and the setting out details for the adit opening had been adjusted to suit the tunnel rings, temporary propping was installed within the tunnel and the adit construction completed.

At the connection to the running tunnels the adits are enlarged to provide the required size of opening and to accommodate the support to the segmental tunnel lining across the opening. The enlargement was constructed in a series of excavations up to the back of the running tunnel to allow for staged monitoring of the tunnel lining. The connection to the running tunnel lining would then be installed before the temporary propping was removed. Finally, the permanent reinforced insitu concrete lining and the connection to the diaphragm walls of the IV Shaft were cast to complete the support.

The excavation and support drawings for construction included suitable sequences and additional support measures should they be required. Particular attention was given to joint details between the headings and benches. Simple convergence monitoring sections were also specified to confirm the stability of the temporary support. Trigger levels were specified as an indication of excessive deflections. A warning level of 7.5mm and an action value of 10mm were defined. The movement of the running tunnels was measured locally using the same method during the construction of the openings. Additional temporary propping was provided for use within the tunnels should it be required.

The adit construction at all three IV shafts used this same basic technique. For the construction of the stacked adits at chainage 10+900 there were the additional complications of the two lower adits passing beneath the northbound running tunnel. In this case all four adits were constructed up to a specified horizontal distance from the northbound tunnel. Once the TBM had then passed through, additional propping and monitoring points were installed in the northbound tunnel before the lower adits were advanced beneath and connected to the southbound tunnel. The construction of the upper adits could then be completed.

Design and analysis of the adit temporary support

The temporary support design was initially sized using previous experience, taking into consideration the JV’s requirements. The lining was designed as 200mm thick sprayed concrete with steel arch ribs at a nominal 1m spacing, which were to be reduced on site in response to the ground conditions to provide adequate support depending on the ground stand up time. Similarly, the length of advance was also to be reduced depending on these details.

For the design of the primary adit support, both passenger and ventilation adits and enlargements were modelled using a simple two dimensional beam and spring model to assess the effects of the calculated loads. It was decided that a single design would be adopted for all three locations. Therefore, a number of load cases were adopted to ensure that the worst case scenario was considered.

For permanent works the full overburden pressure is generally taken into account with specified factors of safety. This was considered as overly conservative for the design of the adit temporary lining. The full overburden pressure is unlikely to build up on the lining in the long term, due to friction within the ground assisting in the soil arching above the adit. Also, the build up of overburden loads in these ground conditions is gradual with time. Hence, for temporary works, applying the full load is considered unrealistic.

To decide on the magnitude of the ground loading a number of options were considered:

  • loading-based pattern theories such as Terzaghi;

  • results taken from monitoring existing tunnels;

  • adopting standard support measures used extensively on previous projects;

  • finite element/difference modelling of tunnel construction using computer programs such as FLAC or CRISP.

At all three IV shaft locations the adits were expected to be constructed entirely within the Bangkok First Stiff Clay. There are a considerable number of published papers on the build-up in load on a tunnel in soft ground. Unfortunately, the information available for tunnels in Bangkok is limited. However, when the geotechnical parameters of London Clay are compared with that of the Bangkok First Stiff Clay they are seen to be very similar. The build up of loads on tunnels in London Clay is recorded in several papers, notably “Long Term Measurements of Loads on Tunnel Linings in Over-consolidated Clay” by Barrett, O’Reilly and Temporal, 1994.

For the adits it was estimated that the temporary linings would be in place for between 5 and 12 weeks. The predicted theoretical principle loads are estimated as 36% to 45% of overburden horizontally and 24% to 28% vertically. It should be noted that these values are for isolated tunnels away from other construction. These values are similar to the 37% of overburden pressure determined by Protodyakonov’s Theory. Therefore the primary lining design adopted the above range with an appropriate factor of safety.

It is assumed that the running tunnel enlargements would only be supported by the temporary lining for a short time, estimated at around two weeks. It could be argued that the design loads could be reduced further for the junction enlargements. However, the loading regime at the junction and the effect of the internal tunnel propping is more complex. Therefore, as an added factor of safety, it was decided to ignore any further reductions to the applied loads.

To account for any differences when comparing the Bangkok ground conditions with London Clay and to minimise construction risks an assumed value of 45% overburden was used for the design calculations.

Having established appropriate loading for an isolated adit other effects were considered:

  • load effects of the TBM drives on the adits;

  • effects of constructing parallel adits;

  • effects of constructing the stacked adits at chainage 10+900

  • effects of constructing adits below the running tunnel at chainage 10+900

Due to the close proximity of the adits it was necessary to assess the additional unbalanced loading from constructing two adits next to each other. This was done using the “theoretical redistribution of stresses after the excavation of a circular hole”. For these purposes, the passenger adit was considered the worst design case due to its height to width ratio. Using this approach the maximum increase in the horizontal loading was estimated as 30% of the overburden pressure. Combining this with the calculated 45% factor previously estimated for the temporary loads, resulted in a total overburden factor of 60%.

In the case of the stacked adits at chainage 10+900 there was also the additional effect of one adit on another vertically. It was deemed that the most severe loading condition would occur if the upper adits were constructed first. This would reduce the vertical loads on the lower adits and result in the maximum ovalisation of the lining. To assess these vertical effects, the passenger adit design considered a number of load cases. These combined the assumed 60% overburden pressure in addition to various degrees of relaxation to the vertical load. From experience, a 20% reduction to the vertical loads is considered realistic. However, there is no numerical means of accurately gauging the affects of one tunnel above another.

Before deciding on which vertical load factor to adopt, the calculations for the varying degrees of relaxation to the vertical load were compared. The comparison showed that differences between the results for the different degrees of relaxation were small. Therefore, in this case, the influence of one tunnel on another vertically has been ignored. Hence a total overburden factor of 60% was adopted.

At the location of the IV shaft at chainage 10+900 the lower adits also pass below the northbound running tunnel. In theory the upper tunnel increases the vertical load on the lower adits as it sheds a percentage of its load into the surrounding ground. The side load on the lower adits will also be reduced as the tunnel above shields the adits from the full overburden load. These are only local effects that increase the difference in magnitude between the horizontal and vertical pressures. The increase in ovalisation will be most exaggerated on the vent adit with its lower height to width ratio.

An additional load case considering an estimated 50% reduction to the horizontal loads to account for this effect was carried out. The results for this analysis indicated that additional propping was necessary for the vent adit locally below the running tunnel. This was achieved by reducing the spacing between steel arch ribs. It was also necessary to prop the upper running tunnel locally until the lower adits were excavated and permanent lining in place.

In addition to the ground loading, a surcharge loading of 50 kN/m2 was included to account for construction plant and working slabs at ground level. A factor of safety of 1.4 was adopted for the design in accordance with the British Standards.

These load cases were assessed using the beam and spring model. The analysis confirmed that the 200mm thick sprayed concrete lining with steel arch rib supports was adequate for the adit excavations in the temporary condition.

Conclusion

The intervention shaft adits are a relatively small part of the Bangkok Metro project. The design of the adits was developed with the contractor in order to consider alternative methods of construction and the construction sequence. Feedback from the contractor has indicated that the ground treatment works have proved excellent. Thanks to the design approach adopted, the skill and experience of the contractor and the close working relationship between designer and the contractor, the Bangkok Metro has seen successful completion of the design and construction of the IV shaft adits.

Related Files
Figure 3: The IV central shaft configuration
Figure 2: The first IV shaft location
Figure 4: The stacked configuration of tunnels
Figure 1: Tunnel alignment