TBM driving under low overburden by adopting innovative methodology

Debasis Barman, Deputy Consultancy Head, Amberg India
TBM mining in urban areas particularly under low overburden (less than 0.5D, D = tunnel Dia) is a big challenge to design team as well as execution team. While working in a Lucknow metro project, the metro authorities encountered a stretch where TBM had to cross a nallah crossing of 15m length under overburden of <1m. Not only the stability of TBM due to uplift was of concern but stability of both banks of nallah during TBM driving due to possible high-volume loss (%) as the banks consisted of mostly made ground material. Also, most structures on the nallah banks were non-engineering structures whose stability due to vibration of TBM was another concern. Overall looking at all these aspects it was proposed to drive TBM by controlling TBM mining parameters and also adopting an innovative technology to control all possible modes of risks. The proposed method was guaranteed in terms of providing stability to both ground and superstructures and was cost effective and feasible in terms of construction. The whole solution from design to execution was a collaborative approach between client, contractor, and designer. The total process took 15-18 months to take shape. I was working as a design representative to the project’s contractor.

The Lucknow Metro Rail project phase 1A (north-south corridor) comprises a total 22.8 km of metro network connecting 22 stations from CCS Airport to Munshipulia, out of which 3.67 km is an underground section (package LKCC-06) starting near Charbagh Metro station and ending near K. D. Singh Babu Stadium Metro station and remaining is an elevated corridor. The underground package consists of two ramps one at the beginning and other at the end of contract boundary, three underground stations namely Hussainganj, Sachivalaya and Hazratganj, twin bored tunnel of internal diameter 5.8m and several cross passages along the alignment. The respective stretch is between Hussainganj station and south ramp TBM drive.


Description of the area
Haider Canal is in Hussainganj area of Lucknow, Uttar Pradesh, India. The densely populated area is very congested with limited access to any investigation and construction activities. All buildings along the banks of the canal are mostly brickworks with no proper foundation for even G+2 storied buildings. Most superstructures constructed lack proper engineering design and execution, and the materials are of inappropriate quality. Both banks of the canal were developed by dumping locally available fill material and daily used garbage is dumped from time to time. As per the initial assessment, the site was found to be on the very unstable ground of the banks of the canal, which raised many doubts about the stability of the superstructures on the banks. Ground cover available from tunnel crown was <1m and high risk of flotation along with sinking of TBM at canal bed due to soft ground was a major concern.

Table 1: Ground characterisation & soil parameters
Thickness of soil layer (m)	Description of ground	SPT/blows	Soil stiffness (MPa)	Cohesion (kPa)	Friction Angle (degree)	Unit Weight (kN/m3)
3	Fill	-	3	0	26	17
4	CL	10	9	30	10	19
5	CI	24	22	100	13	20
3	CL	28	25	140	13	21

Review & Ground Characterization
During the tender stage, two boreholes had been performed in Haider canal banks of 25m depth each. Additional new boreholes of 15m depth on the canal bed were proposed by me to reveal the actual ground conditions through which TBM would be mined. The boreholes on the Haider canal bed were of major considerations in the present studies as they replicated real ground conditions. The results of lab tests and borehole log charts of boreholes in canal bed indicated that the strata at the site was of cohesive soil predominant. The cohesive type soil comprised of either silty clay soil of low to medium plasticity and compressibility or clayey silt soil of low plasticity and compressibility belonging to CL, CI, ML group having 62 to 99 percent material finer than 75 microns.

However, the only non-cohesive type of soil is found to comprise of sandy silt ML type soil having 67 percent fines. Standard Penetration blows as recorded in the filled-up soil zone presented up to 3.00 metre depth below ground level in both the boreholes were found to range from 6 to 7. Beyond 3m depth, blows increased from 8 to 31. No water table was encountered in any of the boreholes during investigation.

Assessment of Existing Building Condition
Most of the structures lied in severe (most critical category of building damage assessment) category and were required to be evacuated prior to site activities. Certain buildings were tilting towards Haider canal and cavities were located on the floor of the structure. Cracks had been developed on the walls in due course of time and might get widened due to construction activities and TBM driving. One of the structures had a column projecting towards the canal bank which had the chances of getting dislodged due to vibration of construction equipment’s and TBM movement. Structures on both banks were surveyed in detail and proposed remedial measures for implementation prior to execution were planned accordingly.


Risk Assessment
As per site visits and review of existing information, the following risks were identified which might be critical during execution:

• Damage to the buildings lying on both banks during TBM mining due to uncontrolled volume loss (%) and large extent of damage along with partial collapse of structures due to its own condition in service life.
• Significant movement of ground lying on banks due to volume loss during TBM driving and possible failure of the bank slopes as the ground is mostly dumped fill which has occurred in long time.
• Drop of face pressure during mining due to weak ground encountered and loss of ground at face along collapse of face and hindrance to TBM movement.
• Ingress of water from cutter chamber as mining done under canal and possible flooding inside the TBM & rings built.

TBM Safe Mining: Preliminary Option
During tender design studies, I interpreted from the ground investigation report that soil below the nallah bed level of thickness 2m-2.5m would be loose due to erosion of water, so the zone might need to be improved through which TBM would pass. Also, flotation of tunnels would occur due to low ground cover above tunnels. Looking at the scenario, it was proposed that TBM driving zone to be confined within a solid concrete mass block so that weight of the block resists the upward thrust due to flotation.

Both the TBMs were proposed to be confined within unreinforced concrete blocks of dimension 12m x 12m x 12m. The concrete block were prepared by interlocking bored piles of 1.2m dia. Piles were bored by leaving an overlap of 100mm between two piles. Each pile was proposed to be terminated at least 3m below tunnel invert level. All piles were of cast in-situ concrete grade M10 (maximum) so that TBM can cut without difficulties. The following risks were identified which might be critical during execution:

• Maintain stability of the concrete block as a whole during TBM cutting the block was a major concern, as a chance that the whole block might move ahead.
• Since the number of piles to be bored were ~300 and of big dia (1.2m), it was difficult to execute as heavy capacity pile driving rig was difficult to be placed on such soft ground.
• Efficiency of the whole scheme would depend on the tolerances of each pile bored and such efficient agency was not available locally in the area.

Though the solution was acceptable in design checks but above possibilities during execution and installation of 300 bored piles would take a significant amount of time and cost in terms of initial preparations. Being from the design side, I proposed that the client should explore alternative options in terms of cost feasibility and ease of execution.

Design Option
To counteract flotation issues, I proposed another option of dead weight increase on the ground surface above the TBM. It was found that a 375mm thick brick slab was lying over canal bed. As the cushion between brick slab and TBM was <1m, an additional thick concrete slab of 1400mm was proposed to be constructed over existing slab as additional dead weight to counteract the flotation during tunnel mining process.


Prior to construction of the slab, a number of 400mm dia. cast in-situ concrete piles were bored for a depth of 20m at a spacing of 3D (D = pile dia) - leaving a 2.5m clear space from TBM both sides. The piles were supporting the thick slab and transferring load to deep soil stratum. It was interpreted that dead load of slab over the ground would try to compress the ground above tunnel crown to some extent which would release the pore pressure in clays to some extent. This would reduce the chances of consolidation settlement of ground to some extent prior to the tunnel mining process. The ground beneath the canal bed is very soft cohesive soil and additional load like slab weight would impose increase in pore water pressure which were required to be drained prior TBM driving.


Design Studies
Check against Uplift: The bored tunnel checked for flotation possibilities considering the dead weight of slab would counteract the uplift forces. Both FOS in temporary and permanent conditions were respectively less than 1.1 & 1.5.

Bored Pile Vertical Capacity determination: The bored pile of 400mm dia. with depth 13m for vertical capacity as per available geotechnical information. The capacity was 40T for a single Pile.


FEM analysis to check Tunnel movements & Pile settlement: FEM modelling using Phase 2 software was prepared for checking the stability of composite system. The soil layers were modelled as per thickness encountered during investigation. Tunnel Liners & Piles were modelled as Plate elements with respective properties provided. The existing slab & thick slab were taken in considerations too. Soil zone around the tunnel between outer boundary piles were modelled with increasing stiffness taking in considerations of grouting.


FEM analysis predicted a max of ~60mm movement due to slab dead weight imposed on slab and a minimum 5-6 months period to dissipate 75%-80% of ground pore water pressure. Accordingly, the TBM duo driving was planned. During TBM driving 5-6 m movement were predicted at slab bottom which were very negligible considering slab weight. Forces imposed on concrete slab & pile were within allowable limits due to TBM movement.

Construction Scenario

Retrofitting of all superstructures within influence zone
To evaluate the impact on all superstructures lying within influence zone of both TBMs, building impact assessment was carried out and classified in terms of potential risk of damage based on categories proposed by Burlandet al. (1977). Boscardin and Cording (1989). A total of 27 buildings around Haider canal area was identified for analysis and almost all of them are small hutments in poor condition. Based on damage category buildings were classified for minor repair/major repair/partial demolition along with continuous monitoring by instruments. The following recommendations, was implemented in site on buildings depending on its existing health:

Table 2: Ground support parameters
Type of material	Stiffness (GPa)	Cohesion (MPa)	Friction angle	Unit Weight (kN/m3)
Grout	15	0.5	40	20 (assuming as lean concrete/backfill material)
Cast In-situ segments	35	-	-	25
Concrete Pile	30	-	-	20

• Major part of the buildings was temporarily supported using steel props. All the cavities as found in building floors were filled with cement slurry immediately.
• All backfilling works conducted by cement slurry were checked against water tightness and extent of fill along the voids in the soil mass. This were done by injecting water at high pressure through existing holes and possibilities of water leakage on the bank side or in other portions of buildings were explored.
• All ground lying on foundation of all buildings were properly regarded and compacted. Voids in the ground were filled with lean concrete.
• External edges of all buildings projected on the bank side were supported by propping as per availability of space.
• Total 1462 bags (73.1 ton) of cement were used to fill the cavity for both side of the houses along the nallah.
• Very critical category buildings were temporarily evacuated till TBM mining & ring building were not completed.
• Monitoring activities were continued as per design proposal and recordings interpreted for any type of movement in regular intervals.

Regrading slopes on both banks
Soil mass on both banks were regarded to stable slope and supported with shotcrete & wire mesh. A retaining wall was constructed along the banks till influence zone of TBM on both sides and backfill with lean concrete completed behind wall.


Execution of Cast In-situ piles along with thick concrete slab
The following steps were proposed to ensure that slab-pile monolithic system was built in canal bed without harassing the canal flows:

• After all building structures were supported both externally & internally and ground of the both banks were regarded, bored piling operations were initiated.
• To prepare a dry surface for working, canal were diverted on one side using temporary bunds.
• Holes were punctured slightly higher than pie dia and bored piles constructed using percusion drilling technique. Reinforcement from piles were projected to be connected with slab in future.
• All piling operations were completed, reinforcement placing & casting of the half slab commenced. The total thickness of 1400mm for slab was executed in two stages of 500+900mm.
• Completion of half slab, water was diverted on the slab and the other half was made dry to repeat the above steps.
• Finally the whole canal flow was restored to its original condition.

TBM driving parameters fixation
The following recommendations were implemented during TBM drive, as per discussions with client and contractor:

• While ring building wherever face pressure dropped down, bentonite slurry was proposed to be injected to maintain the face pressure.
• In case of water ingress from the screw conveyor from cutter chamber, the bulkhead gate was proposed to be closed to minimise the tunnel flooding from the nallah water.
• Heavy duty dewatering pumps to be mobilised to prevent flooding inside tunnel.
• Enzan system to be followed to monitor excavation volume and keeping volume Loss (%) at minimum.
• TBM Face Pressure was recommended for all 5 zones depending on position of TBM in the nallah area.

1. Zone 1 – 0.7 to 0.8 bar
2. Zone 2 – 0.7 to 0.8 bar
3. Zone 3 – 0.5 to 0.6 bar (Nallah crossing)
4. Zone 4 – 0.7 to 0.8 bar
5. Zone 5 – 0.7 to 0.8 bar

Monitoring of ground & structures
For building structures monitoring the following instruments were recommended as per results of building damage assessment:

1. Building settlement marker
2. Tilt plate
3. Crack meter
4. Bi Reflex target

For Ground Monitoring of bank slopes along tunnel alignment the following instruments were proposed:

5. Ground settlement marker
6. Multiple point borehole extensometer
7. Piezometer

Team communication arrangement
A dedicated Tunnelling Task Force Team was formed between the TBM Crew and respective senior personnel from contractor including consultant representative. The function of the team was to make a coordinated approach to the problem with inputs from different levels and to generate a common understanding. The team used to monitor all activities related to TBM movement on shift basis and provided their inputs based on TBM paramreters recorded along with ground & structures monitoring records. Overall the inputs of the team were very useful and valuable on successful completion of the operations. During real execution, designer representatives were mobilised in-site during both TBM drive to provide their inputs based on ground monitoring & TBM record obtained. As per designer observations, TBM speed, Face Pressure & Grout Pressure were modified to avoid blowing of ground & minimise TBM tail end volume loss.

TBM parameters monitoring during mining

The following TBM records were obtained which were quite close to recommended:

• Face Pressure – 0.3 bar (min) to 1.0 bar (max)
• Thrust Force – 18000 kN (max)
• Cutter Torque – 1500 kN-m (max)
• Grout Volume – 130% of excavation volume
• Grout Pressure – 2.0 to 3.0 bar

Conclusion
TBM driving under low overburden in urban areas are quite challenging task and can be implemented successfully by well coordination between client, contractor & designer. Early involvement of contractor is very necessary for inputs to designer on various execution options study. Also, TBM parameters monitoring along with surface and superstructures lying on tunnel alignment monitoring and understanding their co-relation are key to successful drive.

References

1. Loganathan, Nagen. 2011, An Innovative Method for Assessing Tunnelling-Induced Risks to Adjacent Structures: PB 2009 William Barclay Parsons Fellowship Monograph 25.
2. Jardine, F.M. 2003, Response of buildings to excavation induced ground movements, Proceedings of the international conference held at Imperial College, London, UK, on 17–18 July 2001, CIRIA
3. Poulos, H.G., Pile settlement zones above and around tunnelling operations, Australian Geomechanics Vol 41 No 1 March 2006, 81-90
4. Selemetas, D, Standing, J.R., Mair, R.J., The response of full-scale piles to tunnelling, 2006 Taylor & Francis Group plc, London, UK
5. Boscardin, Cording, Building Response to excavation induced settlement, Journal of Geotechnical Engineering, Vol 115, No 1, January 1989