Read Microsoft Word - S61-Pakianathan.doc text version

Design and Construction of Deep Shafts in Hong Kong Special Administrative Region (SAR), China

L. J. Pakianathan

Mott MacDonald Pte Ltd, Singapore

A. K. L. Kwong

University of Hong Kong, Hong Kong

D. D. McLearie

Montgomery Watson Harza, Hong Kong

W. K. Ng

Drainage Services Department, Government of the Hong Kong SAR, Hong Kong

ABSTRACT: Shafts play an essential part in the construction, operation and maintenance of tunnels and deep underground structures but are rarely given exclusive prominence in technical publications. The aim of this paper is to summarise the experiences gained in Hong Kong SAR during the construction of the Harbour Area Treatment Scheme Stage 1 where seventeen shafts were constructed. Their excavated diameters range between 2.5 m to 50 m and at a maximum depth over 150 m these are the deepest shafts below sea level in Hong Kong. All shafts were located in reclaimed land and in close proximity to the sea. The upper shafts in soils and weak rock were constructed by diaphragm walling method and the lower shafts in rock by mainly drilling and blasting. Raise boring and blind shaft drilling methods were also employed. The upper shafts and permanent shaft linings were designed using conventional methods and the primary support selection for the lower shaft was based on Barton's (1974) `Q' system. Settlement monitoring and inclinometer measurements were undertaken during excavation to confirm the design assumptions. During construction several difficulties were met that had to be overcome. All shafts with the exception of one were successfully excavated and completed. This paper addresses the key design and construction issues and the difficulties that were encountered which may be common for deep shafts constructed in an urban setting near a coastline. 1 INTRODUCTION

The Harbour Area Treatment Scheme (formerly known as Strategic Sewage Disposal Scheme) is an environmental improvement project aimed at cleaning up the waters in the Victoria Harbour. The first stage consists of transfer tunnels linking the primary treatment works located at the southern part of Kowloon and eastern part of Hong Kong Island to a centrally located chemically enhanced treatment facility at the Stonecutters Island. A network of 25 km long transfer tunnels were constructed in bedrock at depths varying between 75 m and 145 m below sea level making these the deepest tunnels to date below sea level in Hong Kong SAR. In order to construct the tunnels and to transfer the sewage from the coastal treatment works, 17 shafts were constructed. The excavated diameter of the shafts varies from 2.5 m to 50 m and they reach down to a maximum depth of over 150 m. The decision to locate the tunnels at a deep level in the rock well below toe levels of pile foundations, made it possible to construct the tunnels along a most direct as well as shortest route. It became necessary however to sink deep shafts to link the tunnels to the ground surface. The functions of the different types of shafts are summarised in Table 1.

Table 1 Function of different types of shafts Shaft Type Function Production shafts to excavate the tunnels and to construct the permanent lining Drop shafts to transfer the sewage from the terminal manholes to the tunnels Riser shafts to convey the sewage from the tunnels back to the surface installations Pumping station shafts to raise the hydraulic head of sewage using submersible pumps Figure 1 shows the location of the shafts and Table 2 shows their particulars. All deep shafts were excavated in two parts as upper and lower shaft to suit the operation and the differing ground conditions. The upper shafts were constructed by diaphragm walling or open cut methods through soil and weak rock and the lower shafts were excavated by drilling and blasting, raise boring or blind hole drilling methods in hard rock. Of these, diaphragm walling and drilling & blasting methods were predominantly used. The upper section the production shafts were typically 10 m in diameter and reduce to 8.0 m at the lower section by the installation of a 1 m thick toe level ring beam at the rock/soil interface, Figure 2. The drop shafts are of a larger diameter in the upper section to function as a chamber to remove air from the sewage and to accommodate a bell mouth and vortex drop pipe. These reduce in size to approximately 2.5 m excavated diameter in the lower section. The drop shafts incorporate a 4.0 m deep sump below the tunnel invert level to accommodate submersible pumps for emergency dewatering. The land based riser shafts were excavated at the same size as the production shafts to enable the removal of the tunnel boring machines. The permanent linings for the riser shafts are made of steel pipes or in-situ concrete. Their internal diameters are identical to those of the tunnels to maintain the same flow velocity so as to prevent any sedimentation at the shaft bottom. The pumping station shafts were sized on the basis of the required holding capacity and pumping arrangement. They are up to 38m deep and are founded in soil. The Stonecutters Island Main Pumping Station (SCIMPS) shaft at 50 m diameter is among the largest in Asia. Contractors Skanska-Shui On-Balfour Beatty Joint Venture excavated all the production shafts and Kwun Tong pumping station shaft under an advance works contract DC/93/10. The value of this contract was HK$226 million and the works commenced in August 1994. In parallel another advanced works contract for the construction of diaphragm walls and soft ground excavation of the SCIMPS and riser shaft was awarded to Leighton Contractors at HK$116 million. The remaining drop and riser shafts were excavated later on as part of the tunnelling works contracts as shown in Table 2. 2 GROUND CONDITIONS

The shafts were constructed through recent Fill, Marine Deposits, Alluvium, Completely to Highly Decomposed Rock and Bed Rock. The marine deposits are generally soft, greenish grey clays with variable amounts of silt, sand and shell fragments. The alluvium deposits are generally characterized by variable firm to stiff silts and silty clays. The completely decomposed rock is generally firm, clayey, sandy Silt with some angular to sub-angular fine to occasional coarse gravel sized rock and quartz fragments. The bedrock is made up of either Granite or Volcanic Tuffs. Three out of the seventeen shafts were excavated in volcanic tuffs and remainders were in granite. The ground water table was at sea level and the water met in the shafts was saline.

Table 2. The main features of the shafts

Shaft No.

1 2 3 4 5 6A & 6B 7 8 9 10 11 12 13 14 15 16 17



Kwai Chung PTW Tsing Yi PTW Stonecutters Island STW Stonecutters Island STW Stonecutters Island Undersea Outfall To Kwa Wan To Kwa Wan PTW Kwun Tong PTW Kwun Tong Pumping Station Kwun Tong Pumping Station Kwun Tong Pumping Station Tseung Kwan O PTW Shau Kei Wan PTW Shau Kei Wan PTW Chai Wan PTW Chai Wan


Drop shaft Production and drop shaft Riser shaft Pumping station shaft Outfall production and drop shaft Outfall riser shafts Production shaft Drop shaft Drop shaft Production and drop shaft Pumping station shaft Production and riser shaft Production and drop shaft Drop and riser shaft Diversion chamber shaft Production and drop shaft Production shaft Ø (m) 13.5 8 10 50 10 10 12 13 13 15 10 10 9 7.5 9 10

Upper Shaft Excavation

Depth (m) 33 3 63 38 10 60 32 32 37 25 33 32 25 26 26 21 Constructio n method Diaphragm wall Open cut Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Diaphragm wall Contract No. DC/93/14 DC/96/20 DC/93/10 DC/93/11 DC/93/11 DC/93/10 DC/93/10 DC/93/14 DC/93/14 DC/93/10 DC/93/10 DC/93/10 DC/93/10 DC/93/13 DC/96/17 DC/93/13 DC/93/10 5 8 8 8 4.5 8 2.5 2.5 8 50 63 105 107 75 (113)


Lower Shaft Excavation

Ø (m) 2.5 8 8 8 97 Depth (m) 107 137 68 Construction method Raise boring Drill and blast Hydraulic hammer; Drill and blast Drill and blast Blind hole drilling Drill and blast Raise boring Raise boring Drill and blast Drill and blast Drill and blast Drill and blast Drill and blast Drill and blast Contract No. DC/96/20 DC/93/10 DC/93/14 DC/96/20 DC/93/10 DC/93/18 DC/93/10 DC/96/18 DC/96/18 DC/93/10 DC/93/10 DC/93/10 DC/96/17 DC/93/13 DC/93/10 Geology Granite Granite Granite Granite Granite Granite Granite Granite Granite Granite Volcanic Tuffs Granite Volcanic Tuffs Volcanic Tuffs

Permanent Lining

Type Concrete Concrete Concrete Concrete Concrete Steel pipes Backfilled Concrete Concrete Concrete Concrete Steel pipes Steel pipes Concrete Concrete Concrete Backfilled Contract no. DC/96/20 DC/96/20 DC/96/20 DC/93/16 DC/93/18 DC/93/18 DC/96/18 DC/96/18 DC/96/18 DC/96/18 DC/93/14 DC/96/17 DC/96/17 DC/96/17 DC/96/17 DC/96/17 DC/96/17

83 109 116 114 -

Shaft excavation was discontinued before completion

3 DESIGN 3.1 General The design of the shafts was based on their function which was initially to provide temporary access for tunnel construction and then to transfer the sewage from the treatment works to the deep level tunnels followed by conveying to the central treatments works or outfall. The availability of suitable land space and the location and orientation of the terminal manholes at the treatment works were main factors in deciding the location of the shafts. The upper shaft situated within the soft ground was designed as an octagon suitable for construction by the diaphragm walling method. The lower part of the shaft was designed to take advantage of the inherent strength of the rock during the temporary stage and to with stand hydrostatic pressures during the permanent stage. 3.2 Upper Shaft The upper shaft was designed to withstand the loading from the ground and ground water pressure with an allowance made for surcharge and flooding of the surrounding area. The permissible deviation of the diaphragm wall panels from true verticality was 1:75. The thickness of the walls was chosen to maintain at least 300 mm contact between adjacent panels for the worst case scenario where their verticality is offset in the opposing directions. For a 30 m deep shaft this works out as 1100 mm. The typical thickness adopted for the diaphragm walls was either 1000 mm or 1200mm. A 150 mm construction tolerance was added to the required internal radius and the contractor proposed to trim back any excess concrete encroaching beyond this. Where the diaphragm walls were very deep and the resulting thickness is excessive specialist equipment was used to control verticality. The quasi-circular shafts were designed to carry the loads in hoop compression without any internal propping or strutting. At toe level a nominal 1 m x 1 m ring beam was designed to tie the individual panels together. Where the rock head variation was more than 1 m then deeper ring beams were designed and installed. Where it is not possible to install a toe level ring beam as in the case of the Shau Kei Wan diversion chamber, shear pins were drilled and grouted into the rock. The reinforcement for the panels was selected not only to carry the forces but also to make the cages sufficiently rigid for handlings purposes and to minimize the entrapment of bentonite mud during the concrete placing. Steel pipes and inclinometer tubes were incorporated into the rebar cages to facilitate the drilling of contact grouting holes and for monitoring respectively. 3.3 Lower Shaft Four types of primary support as shown in Figure 3 were specified. The primary support design for the lower shaft was based on the Barton's rock mass quality `Q' system, Barton et al, (1974). Using the information from the initial site investigation (boreholes drilled at the centre of the shafts) it was possible to estimate the corresponding `Q' numbers and select the appropriate support type at different depths. The Bills of Quantities were prepared using this method to quantify the extent of the different support type. As the work proceeded the exposed rock face was geologically mapped after each round of excavation and the `Q' value was re-calculated and agreed with the Engineer's Representatives on site prior to installation of the appropriate support type. The extent of estimated support type under contract DC/93/10 is compared with the actual in Table 4. Table 4. Comparison of estimated support type with actual Support type Estimate (m) Actual (m) 97 139 Type A 353 367 Type B 99 37 Type C 19 25 Type D

The permanent lining was designed to withstand full external hydrostatic pressure under flooding conditions assuming that the shaft was empty. The inner surfaces of the permanent linings that are exposed to condensation were protected by a High Density Polyethylene (HDPE) fully welded protective membrane. This is resistant to hydrogen sulphide attack from the sewage. The surface areas that are always fully submerged did require such protection. 4 CONSTRUCTION OF UPPER SHAFT

4.1 Guide Walls The diaphragm wall construction began with the construction of guide walls. These were temporary structures constructed along both faces of the diaphragm wall. The top of the guide walls was located approximately 0.5 m to 1.0 m above the surrounding ground level so that a positive head of bentonite slurry can be maintained in the excavation to control ground settlement. A sheet pile cofferdam was first erected before the excavation of the typically 1.0 m deep guide walls in view of the high ground water table. The guide walls were constructed of nominally reinforced concrete. 4.2 Diaphragm walls The diaphragm walls were excavated as eight separate panels generally using clamshell grabs suspended from a 50 Tonne crawler crane. The storage silos for the bentonite and plants for slurry separation and desanding were installed on site prior to the commencement of excavation. The panels were excavated in one to three bites. The operation of the grabs was stopped when a hard stratum was reached and it was no longer practical to use this method. Circular and rectangular chisels were employed to excavate through the hard stratum until the predefined toe level of the panel which is at least 500mm below the top of Grade III rock was reached. Following completion of excavation of a panel stop ends were installed and recirculation and pumping out of bentonite from the toe level was carried out for long periods of time (usually overnight) to remove all sediment deposits from the founding level which were mainly sand and rock chippings. When the trench is sufficiently clean the reinforcement cages were lowered in sections up to 12 m long and coupled up vertically using bulldog clips. A tremie pipe was positioned with its end at the bottom of the excavation to enable underwater concrete placing. A high slump Grade 35 concrete mix was delivered to the site and was discharged directly from the truck mixers to the hoppers fitted on top of a tremie pipe. During concrete placing the tremie pipe was carefully lifted up with the free end securely buried at least 1 to 2 m inside the fresh concrete to avoid contamination from bentonite. The displaced bentonite was returned to the storage silos after being cleaned in the separation plant. 4.3 Contact Grouting In general practice excavation inside the diaphragm walls rarely continues deep to expose the toe of the wall panels. In the case of HATS, shafts were sunk below the founding levels of the diaphragm walls and therefore some form of cut off against possible water ingress through the uneven joint at the wall/rock interface became necessary. This was achieved by drilling at least 5 m below the toe (of the deepest panel) through pre-installed pipes cast in the wall panels and injecting a stable cement grout via a single stage packer. This method proved to be effective in stemming any ingress at the wall/rock interface but despite this two shafts required additional treatment described in Section 9.7. 4.4 Excavation The excavation of the soft ground inside the diaphragm walls was carried out by a 0.25 to 0.3 m3 capacity backhoe type excavator and loaded into 4 m3 capacity muck skips. The filled skips were removed to the surface by a crawler crane. Any water that was trapped inside was removed by pumping into the muck skips as the excavation proceeded. The shaft walls were surveyed for each 1.0

m depth and any projection inside the required internal perimeter was removed by a hydraulic hammer mounted on the excavator. The reinforcement bars that became exposed during this operation were coated with anti corrosive paint and protected further by a layer of sprayed concrete. 4.5 Instrumentation and Monitoring Inclinometer readings to detect any horizontal movement of the wall panels were taken daily when the upper shaft excavation was in progress. This was necessary to verify that the design assumptions and confirm the stability. The readings were generally satisfactory overall but occasionally unreliable readings were detected. In the latter case extensometer pins were installed and additional convergence readings were taken. 4.6 Water Ingress The specifications stipulated that the upper shaft shall be watertight. However during excavation seepage was observed in a few locations, mainly through the wall panel joints and the isolated bentonite pockets. The leaks through the joints were repaired by drilling and injecting with a chemical grout. The trapped bentonite pockets were repaired by first removing the loose materials followed by scabbling back to sound concrete and then backfilling with a repair concrete mix. Any seepage water was first diverted using pipes during this operation and then grouted after the repair concrete has reached sufficient strength. 4.7 Progress rates The diaphragm wall construction and excavation have taken approximately four to six months. The delay at the Shau Kei Wan D/S can be attributed to the large variation in the rock head level requiring extensive chiselling. The durations of upper shaft construction activities are summarised in Table 5. Table 5. Duration of upper shaft construction in calendar days Shaft Guide walls Diaphragm Toe Soft walls grouting excavation SCO D/S 20 41 12 26 TKW P/S 20 66 16 47 KTPS D/S 18 46 17 31 KTPS R/S 23 58 14 48 TKO P/S 20 56 17 48 KTPS 21 69 16 60 CW P/S 22 51 13 24 SKW D/S 21 110 11 35 Ring beam 24 13 22 21 27 12 12 Total duration 123 162 134 164 168 166 122 189

5 CONSTRUCTION OF LOWER SHAFT The lower shafts were constructed by drilling and blasting and by raise boring methods. A typical cycle of advance for drilling and blasting consisted of cleaning the face, marking out and drilling shot holes, charging, blasting, fume clearance, mucking out and support installation. Probing and grouting was done at a certain frequency as described below. 5.1 Probing and grouting The ground conditions were explored by two methods during excavation. Advance probing where holes up to 30 m long were drilled at the four corners of the shaft was the preferred method. The other method was to drill an approximately 100 mm diameter central hole all the way down to the shaft bottom using a down the hole hammer.

Where significant water inflow was met in the probe holes further holes were drilled to inject cement grout. It was common practice to maintain approximately 5 m overlap between fans of probe holes. 5.2 Drilling and blasting The lower shaft excavation was carried out generally by drilling and blasting. Immediately below the toe of the diaphragm walls 1.5 m long blasting holes were drilled. At each shaft a trial blast was conducted to confirm the blast design, to demonstrate compliance with the Mines Department regulations and to prove that blasting induced vibrations were below the permissible limits. As the ground conditions improved with the depth of excavation the shot hole length was increased to 2.4 m. Two types of full face blasting patterns namely `wedge cut' and `parallel hole cut' were used. Where the water inflow was high the shaft blasting was done in two halves so that the lower half was used as a temporary sump while drilling was carried out in the upper half. A typical cycle began with the cleaning the rock face after mucking out and marking the centre of the shaft by lowering a plumb bob from a steel beam temporarily placed over the shaft top. The outer perimeter of the excavation was then marked out by spray paint taking account of the primary support thickness. The locations of individual blast holes were marked out as dots of spray paint. The holes for the wedge cut were drilled at an inclination dipping towards the shaft centre. The ring of holes immediately in front of the perimeter holes were drilled vertically down and the perimeter holes were drilled at a slight angle dipping away from the shaft center. With the parallel hole cut, relief holes approximately 100 mm in diameter were drilled near the shaft center and all blast holes were drilled vertically downward. It was important to drill the wedge cut holes accurately to maintain an even spacing of the rings at the toe of the holes. This became particularly important in massive granite with few joints. On occasions blast hole numbers were increased where such conditions were encountered. The blasting vibration can be estimated using the equation given in Geoguide 4, GEO Hong Kong (1992): A = KQdR-b (1)

where A= predicted particle velocity in mm/s; Q = maximum charge weight per delay in kilograms; R= distance between the blast and the measuring point in metres; K= rock constant; d= charge exponent; and b= attenuation exponent. However the Mine's Department equation (2) for calculating the peak particle velocity (PPV) was more widely used: PPV = K(R/Q0.5)B (2)

The site specific constants K=644 and B=-1.22 were derived from a regression curve representing a large number of measurements taken at various locations in Hong Kong. 5.3 Spoil Removal The spoil removal commenced soon after blasting and smoke clearance. The equipment used was the same as that used for the upper shaft with the exception of a 15 tonne Hagglund gantry crane replacing the crawler crane. The skips were only 75% loaded to avoid the risk of falling rock. 5.4 Primary Support The contractor proposed certain changes to the typical primary support types stated in the contract documents and these were accepted by the Engineer. The main changes are as follows: Type A ­ replacement of chain link mesh with 20 mm sprayed concrete since there was a risk of fly rock from blasting being temporarily caugt in the mesh.

Type D ­ replacement of the steel arch ribs with a mesh reinforced sprayed concrete beam. 5.5 Progress Rates The excavation was carried out in two 12 hour shifts. The planned and actual rates of progress for rock classification/ primary support types are compared in Table 6. Table 6. Average excavation progress per week Rock-Mass Support Planned Progress Classification "Q" Type Rate (m/week) >4 A 10.3 0.4 ­ 4 B 9.3 0.1 ­ 0.4 C 4.7 <0.1 D 3.9 Actual Progress Rate (m/week) 8.1 6.3 3.8 2.4



A number of difficulties were encountered during the construction of the shafts. These ranged from coordinating the construction activities with the 24 hour operation of the treatment works, dealing with landfill materials present in the soft ground, substantial variations in the rock head levels, ground water leakage and ground loss, constraints imposed by explosives delivery and limitations on blasting periods, environmental restrictions and the necessity to complete the works in a timely manner so as to avoid delays to the following on tunnelling works. Some such difficulties were not anticipated by the contractors and therefore different mitigation measures were tried starting with the simple and progressing to the more complex based on the success. As a result significant delays were encountered even though almost all the problems were resolved successfully. 6.1 Proximity to the sea All shafts were located close to the sea. Three were located within 10 m of the sea wall and two of the outfall risers were in fact constructed below the sea bed. During the excavation of the upper shaft of Chai Wan and Shau Kei Wan shafts which are located very close to the sea wall several large size (up to 1.5 m long) rock backfill that was placed previously to construct the sea wall were encountered. These could not be plucked out by the clam shell grab and therefore a cactus type grab was used instead. In addition substantial losses of bentonite slurry also occurred through this medium. This problem was overcome by backfilling the excavated trenches with completely decomposed granite and re-excavating through it. Furthermore only one panel was excavated at any one time. The surcharge loading within a 10 m zone behind the sea wall was restricted to 10 kN/m2. This limitation made it difficult to install bentonite slurry silos and gantry crane foundations near the shaft and special dispensation was sought for the Shau kei Wan diversion chamber shaft to install the slurry tanks immediately behind the sea wall because of limited land space. There was sufficient redundancy available in the sea wall designs to accommodate additional temporary loading but systematic settlement monitoring of the sea wall coping was undertaken to demonstrate that the excess surcharge loading did not have any adverse effect. 6.2 Working on Reclaimed Land The shafts were constructed on land that was reclaimed from sea. The fill material and placing methods in the case of old reclamations were not as strictly controlled as those new. At Kwun Tong shaft site there was evidence of household refuse in the land fill. In addition that site was previously used as a fuel farm for unloading fuel from ships and storing there temporarily. Upon close

investigation contamination from the fuel was found to be limited since additional fill material had been placed over the site and the surface runoff has washed away and diluted the fuel concentration in the ground. However during the excavation of the upper shaft through the land fill material bubbling of gases was observed. Gas monitoring was carried out to detect hydrogen sulphide and explosive gases and forced ventilation was set up. The natural ventilation was found to be sufficient to disperse the hazardous gases since they occurred at a relatively high elevation. At Tseung Kwan O the shaft was located in an old land fill and at the foot hills of a recent waste dumping site. During the excavation of the upper shaft leachate was encountered. This was carefully removed and disposed off site. Frequent gas monitoring readings were taken to check for explosive gas content. The gas concentration measured was sufficiently low after dispersal by forced ventilation. Since the diaphragm walls were watertight in the case of both shafts the hazardous gases problem was resolved once the layer containing decaying waste was removed. 6.3 Variations in rock head level and deep weathering One borehole was drilled at the centre of each shaft during pre-tender site investigation. Since the foot print of a shaft is relatively small significant variations in the rock head levels were not anticipated and a variation of less than 1m was expected. However during the detailed design of the upper shaft diaphragm walls three or more boreholes were drilled and substantial variation in the rock head levels between different panels was discovered for Stonecutters Island riser shaft, To Kwa Wan production shaft and Shau Kei Wan shafts. The largest variation of 7.5 m between the highest at and lowest toe level was met at the Shau Kei Wan Diversion chamber shaft. In the case of the Stonecutters Island Riser Shaft and To Kwa Wan production shaft deep weathering compounded the difficulties taking the diaphragm walls to some 50 to 60 m deep. A hydro-fraise type diaphragm wall trench cutter was used to ensure the verticality of the deep wall panels. In addition deep toe level ring beams were constructed to compensate for the variations in the rock head level. There was not sufficient space inside the Shau Kei Wan diversion chamber to install a ring beam and in any case this shaft was not required to be deep. The panels were dowelled into rock with a bundle of 4 x 50 mm cement grouted reinforcement bars. Inclinometer and convergence measurements were taken during the shaft excavation to verify the stability and no adverse trends were observed. 6.4 Marine mud ingress into Diaphragm wall excavation While the upper section of the Stonecutters Riser shaft was being excavated a sudden inrush of marine mud from the toe of a diaphragm wall panel occurred. This loss of ground through gaps in between undetected corestones caused a depression in the ground surface. Since the location of the ground loss was situated close to a 60 m diameter shaft and in an area planned for the construction of an adit linking the two shafts, serious concerns arose when this incident occurred. The contractor proposed to fill the shaft with water immediately to contain the mud flow by equalising the hydrostatic pressure, install another diaphragm wall panel behind that affected and then undertake jet grouting from the surface to strengthen the collapsed ground. These remedial works were successful but caused a major delay to the completion of the shaft excavation. In addition the method of excavation of the adit was changed to open cut from bored tunnelling by installing additional diaphragm wall panels between the jet grouted area and the 60 m shaft. When the shaft was re-excavated after the remedial works frequent inclinometer and extensometer readings were taken to monitor the convergence and verify the stability. 6.5 Flooding incidents in the upper shaft At the Stonecutters Island outfall shaft an unexpected inrush of water was met at the initial stages of rock excavation works immediately under the toe of the diaphragm walls. During the time of the incident an excavator was cleaning the shaft bottom making it ready for the next round of blasting.

There was sufficient time to rescue the excavator operator before the shaft completely flooded. According to his account the water entered through a clay seam in the rock. It was decided to undertake tube-a-manchette grouting from outside the shaft over the zone of the leak. After completion of grouting the shaft was pumped out dry and no further leakage was met. It was later discovered that the water leak occurred through the clay matrix in between two large corestones. This area was strengthened by applying sprayed concrete before undertaking further excavation. A similar incident occurred as well at the foot of the diaphragm wall of the Tseung Kwan O shaft. The pumping system was able to cope with the inflow and therefore it was decided to grout the leaks from inside the shaft. First deep holes were drilled to intercept the leaks some distance behind the shaft wall and mechanical packers were installed in these holes. After channeling the water to pipes set in rapid hardening cement, a layer of sprayed concrete was applied to the closely jointed shaft wall and the packers were grouted up under pressure as soon as the sprayed concrete has reached sufficient strength. Eventhough the remedial method was successful in staunching majority of the inflow there was residual leakage from this area which persisted until the permanent lining was installed. 6.6 Wide Clay filled sub-vertical joint in rock At the Stonecutters riser shaft an approximately 1 m wide clay filled hydro-thermically altered joint was met immediately below the diaphragm wall panels. The rock mass quality was substantially reduced by the presence of this weak material and it was decided to adopt Type D primary support which is based on steel arch ribs and sprayed concrete. The sub-vertical joint persisted for up to about 30 m deep before disappearing into the side wall and out side the foot print of the shaft. The steel arch ribs were continued until the influence of this joint on the shaft was no longer significant. 6.7 Plant breakdown The high humidity, constantly wet and salty environment in the shaft lead to higher than normal wear and tear of mechanical plant and equipment. The excavator, water pumps and shotcrete pumps suffered from frequent breakdown. On occasions the gantry crane also broke down contributing to the average lost time of approximately 20%. Even though the direct loss of time from plant break down can be averaged out as 5 hours per day its effect in terms of the 10 hour window available for blasting was very serious since a lost blast meant that the whole day's production was lost. 6.8 Delivery of explosives The storage, transportation and use of explosives were strictly controlled to prevent both misuse and mishaps. The proximity of the shaft site in relation to the Mine's Department magazine generally governed the time of arrival of explosives on site. The earliest delivery was received at the shafts nearest to the magazine at around 9 to 10 am and the furthest shafts where transportation also included use of a boat received deliveries by noon. This was satisfactory for undertaking one blast a day but not for two. As the excavation works were falling behind programme the contractor attempted to do two blasts a day without success. With help from the client special dispensation was obtained for delivery of explosives and blasting on Sundays and Public Holidays. This proved to be effective and resulted in a 15% improvement in production. 6.9 Shaft/tunnel junctions The forming of junctions between shafts and tunnels were undertaken without difficulties in the rock though several of these junctions incorporated a chamfer to permit lowering of the tunnel boring machine components. However this activity proved to be very difficult in the upper shaft through the soft ground. During the construction of the Kwun Tong pumping station shaft/adit tunnel junction a sudden inrush of sand and ground water occurred. The remedial works necessitated the use of liquid nitrogen freezing to form an impermeable plug in the soil while the tunnel eye opening was made in the diaphragm wall, Pakianathan et al (2002). At Shau Kei Wan drop shaft soil grouting followed by the installation of a circular fan of closely spaced horizontal grouted pipe piles was put in place before

making the tunnel eye opening in the diaphragm wall to construct an adit tunnel. The presence of a water retaining box culvert directly above the this junction made the task even more difficult. This junction opening was formed without incident or adverse settlement to the structure above. 6.10 Substantial water ingress

The Chai Wan production shaft was located in Volcanic Ash Tuffs within the zone of influence of the Chai Wan fault. The water ingress progressively increased from 80 l/min at 35 m depth to 1400 l/min at 96 m despite pregrouting. Majority of the grouting holes turned out to be dry and ineffective in intercepting water making joints which were tight and closely spaced. Initially an OPC cement grout mix at varying water cement ratio (from 5:1 to 1:1) with 2% bentonite was used at a maximum back pressure of 30 bar. When this was found not very effective Rheocem 650 microcement with 3% Rheobuild 1000 at a water cement ratio of 3:1 was tried at the same pressure. This resulted in higher grout take but the progress became very slow and water inflow continued. As the overall project time table could not be met at the rate of progress being acheived it was decided to reverse the direction of the TBM drive and abandon this shaft. Upon completion of the shafts the total inflow rates at the bottom of the shafts were measured before handing over to the tunnelling contractors. These rates are given in Table 7. Inflow rates from all shafts with the exception of the abandoned Chai Wan production shaft were well below the 300 l/min limit set for dewatering purposes. Table 7. Ground water ingress after completion of shaft excavation Shaft Depth of Total Inflow rate per Inflow rate per square shaft inflow rate metre of shaft metre of shaft (m)## (l/min) (l/min/m) (l/min/m2) 0.020 0.5 69 Tsing Yi 140( 129) 0.024 0.6 36 Tseung Kwan O 95 (63) 0.008 0.2 18 Stonecutters Outfall 107 (97) 0.040 1.0 54 Kwun Tong Riser 83 (50) 0.020 0.5 60 KwunTong Drop 151 (114) 0.076 1.9 160 To Kwa Wan 138 (83) 0.744 18.7 1400 Chai Wan# 96 (75) #Shaft excavation was discontinued before completion ## - Depth of shaft in rock is shown in brackets where nominal water seepage was permitted 7 GROUND SURFACE SETTLEMENT

The ground surface settlement from the excavation was monitored around the shafts to ensure that it did not exceed the specified maximum value of 25 mm including the measurements taken on seawall copings adjoining the shafts. In general this limit was not exceeded but at Chai Wan production shaft where there was a high ingress of water into the shaft the settlement values were more than 25 mm. This shaft was located on land recently reclaimed from the sea and the long term consolidation settlement was still in progress at the time of excavation. It was therefore difficult to conclude whether the contribution from shaft excavation alone was responsible for the excess settlement. In any event there was no sensitive structure around this shaft at the time apart from the gantry crane foundations which were underpinned to compensate for the subsidence. 8 CONCLUSIONS

The aim of this paper is to present the design and construction aspects of the deep shafts in Hong Kong SAR as a case study with particular emphasis on ground related difficulties and how these were dealt with in order to complete the works successfully. It is hoped that the problems highlighted and the

solutions described in this paper could provide valuable documented experience to the construction planning of similar shafts in the future. REFERENCES Barton N., Lien R. & Lunde J. (1974) Engineering classification of rock masses for the design of tunnel support. Rock Mechanics, Vol. 6, No. 4, pp. 183-236. GEO, Geoguide 4 (1992), Guide to Cavern Engineering, Geotechnical Engineering Office, Civil Engineering Department , Kong Kong, pp. 76-78. Pakianathan L. J., Kwong A. K. L., McLearie D.D., Chan W. L. (2002). Pipe Jacking: Case Study on Overcoming Ground Difficulties in the Hong Kong SAR Harbour Area Treatment Scheme. Trenchless Asia 2002, 12-14 November 2002, Hong Kong.


Microsoft Word - S61-Pakianathan.doc

14 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


Notice: fwrite(): send of 199 bytes failed with errno=104 Connection reset by peer in /home/ on line 531