Methods of Resisting Hydro Static Uplift in Substructures

Tunnelling and Underground Space Technology 16 Ž2001. 77᎐86

Methods of resisting hydrostatic uplift in substructures I.H. WongU Mitic Associates, 95 Cashew Road 噛03-03, Singapore 679666, Singapore Received 5 March 2001; received in revised form 8 May 2001; accepted 10 May 2001

Abstract Many underground structures are constructed for use as car parks and shops in basements of buildings and as mass rapid transit stations, depressed roadways and civil defense shelters in cities located in coastal areas where the ground is level and the elevations are low, with an attendant high groundwater table. This paper discusses the various methods of resisting hydrostatic uplift. These include the use of tension piles and the installation of a water pressure relief system under the base slab of the basement. A case history in Singapore employing a pressure relief system below a three-level basement is presented. 䊚 2001 Elsevier Science Ltd. All rights reserved. Keywords: Deep excavation; Groundwater pressure; Uplift; Stiff; Soft soils

1. Introduction A large amount of underground space is being constructed each year in Singapore and other cities. Basements in buildings mainly serve as car parks and shops. In Singapore, government regulations require the developers of shopping malls, offices, apartment complexes and hotels to provide on-site vehicle parking. The very high costs of land here have forced developers to house the car parks within the buildings, mostly in basements. Underground space is also used as mass rapid transit stations, depressed roadways and civil defense shelters. Many cities are located in coastal areas where the ground is level and the elevations are low, with an attendant high groundwater table. The underground structures in these cities thus have to be designed to resist high hydrostatic uplift loads. Many jurisdictions require that the design groundwater table for uplift be taken at the ground level. This paper discusses the various methods of resisting U

Corresponding author. Tel.: q65-7664307; fax: q65-7626924. E-mail address: [email protected] ŽI.H. Wong..

hydrostatic uplift. These include the use of tension piles and the installation of a water pressure relief system below the base slab of the basement structure. A case history in Singapore employing a pressure relief system below a three-level basement is presented.

2. Mechanism of flotation caused by hydraulic uplift The design of underground structures and basements of buildings requires checking for the possibility of flotation due to the forces of hydrostatic uplift. In the permanent condition, the minimum ground water level adjacent to the excavation is commonly assumed to be at the highest recorded flood level, or the finished ground level, whichever is higher. The mechanism of flotation caused by hydrostatic uplift is illustrated in Fig. 1 for the general cases of a building with a basement and an underground structure. The uplift force U acting on a structure of base width B is given by: Uplift force, Us ␥w z B

0886-7798r01r$ - see front matter 䊚 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 8 8 6 - 7 7 9 8 Ž 0 1 . 0 0 0 3 7 - 2

I.H. Wong r Tunnelling and Underground Space Technology 16 (2001) 77᎐86

78

Where: ␥w s z s B s

unit weight of water water table height Žoutside the excavation. relative to the base of structure width of the structure at its base

The resisting force R is offered by the self-weight of the structure W, weight of backfill or the wedge of soil sticking to the basement wall Ws and the shear resistance S of the soil along the planes a᎐a⬘ and b᎐b⬘. In computing the weight of backfill or the soil wedge, the saturated or the bulk unit weight of the soil is used below and above the water table, respectively. For Case A in Fig. 1, the submerged unit weight of the soil below the water table should be used in computing the weight of the soil wedges. Total resisting force, R s Wq ⌺Ws q ⌺S Factor of safety against flotation s RrU. For underground structures, the weight of soil back-

fill within the top 1.5᎐2.0 meters of the ground surface is normally ignored if the structure width is less than 15 m. However, if the structure width is more than 15 m, the backfill within the top 1.5 m is ignored for a half-width of the structure. Where a building basement is constructed using conventionally cast walls and waterproofing membrane inside a temporary cofferdam, the shear resistance between the walls of the structure and soil is generally ignored.

3. Conventional methods to resist uplift For tall buildings with basements, the weight of the completed structure is generally adequate to resist the uplift at the base. However, measures to counteract flotation are still provided to cater for the construction stage of the structure, and also to reduce the bending moments in the base slab. One of the following methods can be used to counteract uplift forces on the substructure. The method chosen depends on the subsurface conditions, the particulars of the project and the method of construction. 3.1. Toeing in of base slab into surrounding ground When a substructure is constructed inside a temporary cofferdam or open excavation, permanent resistance to uplift can be provided by extending the base slab beyond the perimeter wall. The weight of the backfill above the toed-in base slab adds to the weight of the structure in resisting uplift. This method is not feasible where a diaphragm or secant pile wall is used as a permanent retaining structure. 3.2. Increasing dead weight of structure The self-weight of the structure can be increased by thickening its structural members including the structural base slab. Increasing the base slab thickness is not very economical because only the submerged weight of the concrete gives additional resistance to uplift. This is because the contribution from the weight of any additional thickness of concrete should take into account the increased volume of water displaced. An increased base slab thickness requires a deeper excavation, resulting in larger ground movements and requiring a stronger temporary support system. In some projects, the dead weight of the low rise podium in a high rise complex is increased by incorporating a roof top garden with a thick soil fill. 3.3. Ground anchors

Fig. 1. Basement and underground structure subjected to hydrostatic uplift forces.

Prestressed anchors can be used as a temporary measure to counteract flotation forces. In many juris-


dictions, their application as a permanent measure to resist uplift is limited by concerns about their long-term performance with respect to corrosion. 3.4. Tension piles This is the most commonly used method of resisting uplift. The various types of tension piles include steel tension piles, micropiles and bored piles. Steel tension piles are discouraged in Singapore due to concerns about their corrosion. However, studies by Romanoff Ž1962, 1969. showed that corrosion of driven steel piles in undisturbed natural soils is very small. According to British Standards BS8004:1986 ŽBritish Standard Institution, 1986., for a steel pile driven into undisturbed natural ground below the permanent water table, the corrosion rate is negligible. Wong and Law Ž1999. reported a maximum corrosion rate of 0.015᎐0.018 mmryear and an average corrosion rate of 0.01 mmryear for steel H-piles driven into completely decomposed granite in Singapore. Micropiles and bored piles are commonly used for resisting uplift. The full-length reinforcement for these tension piles is required to be corrosion protected by epoxy coating or by hot dip galvanizing. Due to the inability of concrete to carry tension, the use of bored piles for resisting uplift is inefficient. Bored piles are commonly installed before the excavation for the main basement structure is carried out. Thus, the part of the bored piles through the basement space is usually left empty or is backfilled with soil. Such unproductive drilling increased the cost of bored piles. The presence of the struts, king posts and decking system normally precludes the deployment of bored piling machines at the base of the excavation. Micropiling machines require smaller headroom, and thus the installation of micropiles can await the completion of excavation of the basement. It is common for the base slab to be cast with pipe sleeves to be left in the base slab. Micropiles can then be installed through the pipe sleeves in the slab. The time required for the drilling of the bored piles or micropiles at the base of the excavation before the casting of the base slab could result in larger ground movements. If soft clay is present at the base of the excavation, trafficability of the piling machine on the soft clay could be difficult.

Various published tests have indicated that the skin frictional resistance of short piles to uplift load is lower than that mobilized in resistance to compression loading. Accordingly, higher factors of safety are generally used in the design of uplift piles. Pullout tests are also conducted for large projects to ensure that there is an adequate factor of safety against the pile pulling out of the ground. When a large number of tension piles are arranged in a closely spaced group below a substructure, the uplift resistance of the group may not be equal to the cumulative uplift resistance of all the piles in the group. This is because, at limiting equilibrium, the entire block of soil enclosed by the piles may be lifted. The load transfer mechanism between the piles and soil is complex and depends largely on the character of the ground and the method of pile installation. For tension pile groups in cohesionless soils, the volume of soil that can be lifted by the pile group is defined by a simplified spread of load from piles to soil as shown in Fig. 2. Tomlinson Ž1994. suggests using an empirical distribution of one horizontal to four vertical. In this method, the total weight to be lifted includes the combined weight of the piles and the surrounding soil, plus the weight of the structure. For tension pile groups in cohesive soils, the uplift resistance of the block of soil in the undrained loading case will be similar to that shown in Fig. 3. The total uplift resistance of the group Q u is given by: Q u s Ž 2 LH q 2 BH . c u q W Where L and B are the overall length and width of the pile group, respectively, H is the length of the pile, c u is the average undrained shear strength of the soil around the sides of the group and W is the combined weight of the block of soil and piles enclosed by the pile group plus the weight of the structure. For the cases shown in Figs. 2 and 3, if the design uplift force corresponds to that acting at the base of the substructure, then only the submerged weights of the part of the piles and the soil below the water table

4. Design of tension piles In tension piles, the resistance to uplift is provided by the friction or adhesion between the pile and the surrounding soil. The uplift resistance can be increased in the case of bored piles by under-reaming or belling out of the bottom of the piles.

79

Fig. 2. Tension pile group in cohesionless soils.

80


Fig. 3. Tension pile group in cohesive soils.

should be used in resisting the uplift. The allowable uplift resistance of the group must not be greater than the sum of the uplift resistance of the individual piles in the group.

5. Methods to reduce or eliminate uplift An attractive alternative to the use of tension piles or other uplift resisting measures is to provide permanent dewatering using an under-drain system. This technique has been successfully used on many projects. Significant savings can be realized. Apart from significant up-front cost savings, this technique helps to keep the basements dry. Lowering the water table by pumping from an under-drain system is one of the most effective methods of preventing water ingress into a basement and of reducing the uplift pressure on the base slab ŽCedegren, 1967.. Cedegren lists many common applications of dewatering for control of water ingress and for uplift resistance for road bases, airport runways, dry docks, dams and basements. Dewatering by pumping from sumps connected to an under-drain is widely practiced in a large number of buildings with basements. In the United States, for example, most single family or detached houses with basements in areas where the water table is high, or where it could rise during the wet weather, have pumps installed in sumps connected to an under-drain. The pumps would turn on automatically when the water level in the sump reaches a certain pre-set level and would switch off by themselves after the water level falls to another pre-set level after pumping. In Southeast Asia, permanent dewatering is employed or proposed on a number of buildings. In Jakarta, permanent dewatering is reportedly employed or planned for the proposed 62-story BDNI project with a 22-m deep basement, and the proposed Menara Jakarta with a height equivalent to a 74-story building with a three-level basement. In Hong Kong permanent

dewatering is utilized at the Jockey Club headquarters building at Happy Valley with a 16-story tower and a 4.5-m deep basement. Most permanent pressure relief or dewatering systems involve passive dewatering where a drainage blanket is installed under the raft or bottom slab of the basement and water entering the drainage blanket flows towards one or more sumps where it is removed by pumping. Active dewatering involves pumping from permanent well points or deep wells to lower the water level. Active dewatering is required where there is a relatively thin impervious soil zone below the base slab and this impervious soil is in turn underlain by a pervious deposit. In this case the base of the impervious zone is subject to the full hydrostatic pressure corresponding to the head difference between the original ground water table and the base of the impervious zone.

6. Design of under-drain system The first thing in the design of the under-drain is to estimate the quantity of seepage expected. The estimation can be done by the flow net method assuming two-dimensional condition or by numerical analysis such as the finite element method. Since the underdrain operation is long-term, steady state seepage analysis is appropriate. From an analysis of flow in an isotropic soil using a flow net, the quantity of flow q per meter run is equal to kHNfrNd where k is the permeability of the soil, H is the head using the base of the drainage blanket as datum, Nf is the number of flow channels and Nd is the number of head drops. For an anisotropic soil, the effective permeability is Ž k h k v . 0.5 where k h and k v are the horizontal and vertical permeability, respectively. It is conservative to multiply the value of q from the 2-D analysis by the total length of the perimeter of the basement to obtain the total flow to be discharged by the under-drain system. Finite element and finite difference analysis programs are now readily available for the analysis of seepage problems. Three-dimensional analyses are complicated and time consuming to perform. It is common practice to treat the problem as 2-D or axi-symmetrical. As in the method using flow nets, the flow obtained in the 2-D analysis can be multiplied by the total perimeter length to arrive at the total discharge. A substructure that is rectangular in plan can be replaced in the analysis by an equivalent circular structure with the same perimeter length and analyzed as an axi-symmetric case. The drainage blanket should be pervious and thick enough to discharge the quantity of water seeping into it from the underlying soil and also should be able to

I.H. Wong r Tunnelling and Underground Space Technology 16 (2001) 77᎐86 Table 1 Discharge capacities of pipes and aggregate drains Drain type

152-mm Diameter concrete pipe 2.8 m2 Aggregate drain 12 m2 Aggregate drain 370 m2 Aggregate drain

Aggregate size Žmm.

Permeability of aggregates Žmrday.

Gradient

0.01 19᎐25 6᎐19 Clean pea gravel

36 500

0.01

9100

0.01

300

0.01

resist the migration of fines from the underlying soil. Typically, sand or gravel is used. The thickness and the permeability of the material in the drainage blanket required depend on the quantity of the discharge. It is common practice to embed perforated or slit pipes in the drainage blanket to increase its discharge capacity. Cedegren Ž1967. has compared the sizes of different aggregate drains with the same discharge capacity as 152-mm diameter concrete. This comparison is presented in Table 1, and it follows that drain pipes would greatly increase the discharge capacities of aggregate drains in which they are embedded. For a certain project in Singapore involving the installation of an under-drain system, the expected seepage into the 100-m by 40-m blanket is 18.9 m3rh. There are four sumps placed at even intervals along the longitudinal axis of the blanket. The seepage toward each sump from each direction Žlongitudinal. can be taken to be approximately Ž18.9r8. s 2.36 m3rh. From the results of laboratory tests, the permeability of the granular material in the drainage blanket is 1 = 10y4 mrs. For a 0.3-m= 25-m strip, the discharge at a hydraulic gradient of 0.01 is: Q b s Akis 0.3= 25 = 1 = 10y4 = 0.01 s 7.5= 10y6 m3rs s 0.027 m3rh Flow through a pipe can be computed as follows: Q p s Ž A prn . r 2r3 s 0.5

81

where A p is the area of the pipe and r is the hydraulic radius of the pipe, s is the hydraulic gradient, and n is Manning’s coefficient. The calculations in Table 2 show that, without the pipes, the discharge capacity of the drainage blanket will be inadequate. They also show that the discharge capacity of the under-drain system is greatly increased if pipes are embedded in the blanket. In Singapore, drain pipes are embedded in the drainage blankets in the permanent dewatering system in use in the entire Raffles City Complex, and in the reconstructed Fullerton Square building. To resist the potential for piping or migration of fines from the underlying soil, a filter should be placed between the drainage blanket and the subgrade soil. The filter should be designed based on criteria such as the ones proposed by Terzaghi as shown below: w D 15 Ž of filter . rD 85 Ž of soil .x - 4 to 5 - w D 15 Ž of filter . rD 15 Ž of soil .x where D 15 and D8 5 are the sizes at which 15 or 85% by weight of the material is finer. Compliance with the criterion on the left side of the inequality generally will prevent piping. The criterion on the right side of the inequality will ensure sufficient permeability to prevent the buildup of large seepage forces and hydrostatic pressures in the filters. Filters can consist of granular materials or geofabrics. Where perforated drainpipes are embedded in the drainage blanket, it is prudent and a common practice to wrap the drain pipes in geofabrics. The pumps installed in the sumps should be able to handle the maximum estimated flow. It is common practice to have pumps that can discharge from five to 10 times the maximum estimated flow. Multiple sumps should be used, sited strategically at different parts of the substructure. There should be two pumps at each sump, the second pump serving as standby. A standby electricity generating set should be provided. 6.1. Operating costs The operating costs for an under-drain system in-

Table 2 Discharge capacities of drain pipes of various sizes Pipe radius R Žm.

Hydraulic radius r Žm.

Hydraulic gradient i

Discharge Qp Žm3rs.

Discharge Qp Žm3rh.

0.05 0.075 0.10 0.15

0.025 0.0375 0.05 0.075

0.01 0.01 0.01 0.01

0.0048 0.0141 0.030 0.089

17.28 50.76 108.0 320.5

Ratio Qp r Qb 640 1880 4000 11 870


82

clude maintenance of the pumps and the electric and water disposal charges. The power required in running the pumps can be estimated from the following equation: Ps ␳ gQHr Ž 1000n . where: P s unit power capacity in kW ␳ s mass density of water in kgrm3 g s acceleration due to gravity in mrs 2 Qs discharge in m3rs Hs effective head in m n s efficiency For the example given here, with a discharge of Qs 18.9 m3rh or 0.00525 m3rs and a head of 15 m, and an efficiency n s 0.5, the power Ps 1000 = 9.81= 0.00525 = 15rŽ1000 = 0.5. s 1.55 kW. The annual power consumption is 1.55= 24 = 365 s 13578 kW h. At a charge of Singapore $0.16 per kW h, the annual electricity charge is S $2172.

7. Effects of dewatering on adjacent ground and structures Continual pumping of water from a drainage blanket beneath a basement or underground structure may cause settlement of adjacent ground and structures. A proper knowledge of the soil and groundwater conditions is essential in deciding the provisions for permanent drainage beneath substructures. It is important to obtain all the necessary information on the geohydrological regime of the ground during the site investigation stages of the project. A perfectly watertight wall penetrating into an impermeable soil layer at the bottom of the excavation would preclude flow of water from the surrounding soil towards the drainage blanket. In practice however, this situation rarely exists. When water flows into the

drainage blanket from the adjacent ground, a decrease in ground water pressure will occur. This will cause an increase in effective stress and settlement of the soil surrounding the excavation. If a compressible clay layer exists above or below the water bearing layer from which pumping is being carried out, the increase in the effective stress causes the soft soils to consolidate, with accompanying settlements of the ground surface. The effects of groundwater lowering will be more severe in the case of clay and peat. The settlements may be significant even in the case of loose sands and silts when the water table fluctuates. Little or no settlements may occur in the case of dense sands, gravels and very stiff to hard clays. Due consideration must be given to settlements of piled foundations as the ground settlements may impose significant down-drag forces on the piles due to negative skin friction. Groundwater table lowering may expose and cause untreated timber piles to rot. Four general cases of substructures with permanent drainage provisions in different ground conditions are considered here to illustrate the suitability of different subsurface conditions for permanent under-drains. Fig. 4 illustrates subsurface conditions where permanent under-drains as a means of uplift control are suitable in Case A, the substructure is located in an impervious soil of large thickness. In this case, the seepage will be very small and the extent and effects of the lowering of the groundwater level will be very limited. In Case B, a cutoff wall penetrates past the pervious layer into the impervious zone below. The seepage amount will be small, as the cut-off wall would prevent the flow of water from the pervious soil toward the basement drainage system. The effects of the groundwater lowering on adjacent structures and ground will be very small. Subsurface conditions shown in Fig. 5 are not suitable for the use of permanent under-drains as a means of uplift control. In both cases shown in Fig. 5 a pervious water-bearing layer exists below a soft clayey soil. In Case A, the basement wall does not penetrate

Fig. 4. Two cases where a permanent under-drain system is feasible.


83

Fig. 5. Two cases where a permanent under-drain system is not feasible.

into the impervious soil below the pervious layer. When the pervious soil layer is thin, and as the water it contains is depleted, the lowering of the piezometric level could be large and wide. Consolidation settlement of the soft clayey layer would be large and the effects on adjacent structures and ground could be potentially severe. In Case B, when the pervious soil layer is thick and the cutoff wall does not penetrate fully and go past the thick pervious layer, the quantity of seepage will be very large and the cost of pumping may render a permanent under-drain scheme uneconomical. The ground settlements due to consolidation of the upper soft clayey layer may also be large. Permanent under-drain systems shown in Fig. 5 can be made feasible by constructing cut-off walls and socketting them into the underlying impervious layer. In practice, if the underlying material is not fully impervious, piezometric level lowering and ground settlements will likely occur. The consolidation settlement of a clay layer induced by a lowering of the groundwater or piezometric level can be calculated as follows:

S s w HCrr Ž 1 q eo .x x log wŽ pc . rpo x q w HCc r Ž 1 q eo .x x log wŽ poq ⌬ u . rpc x for a soil that is lightly or moderately over-consolidated. The pre-consolidation pressure pc may be exceeded if the pore water pressure is lowered. A lowering of the groundwater or piezometric level will cause large settlement if the pre-consolidation pressure is exceeded. In the preceding three equations, po pc ⌬u H Cr Cc

is the existing effective overburden pressure. is the pre-consolidation pressure. is the decrease in pore water pressure. is the thickness of the clay layer. is the recompression index in the pressure range below the pre-consolidation pressure. is the compression index in the pressure range beyond the pre-consolidation pressure.

for a soil that is heavily over-consolidated such that the pre-consolidation pressure pc is not likely to be exceeded when the pore water pressure is lowered, p 0 < pc . Settlements will generally be small .

In Singapore the maximum water draw down is generally limited to 3 m except for works in the vicinity of mass rapid transit system while the limit is only 1 m where soft clayey soils are present. Recharging of the surrounding ground with recharge wells is often used to reduce the ground and structure settlements caused by dewatering. While recharging has often been utilized for temporary excavation projects, its use in connection with a permanent under-drain system is uncommon and thus its effectiveness uncertain.

S s w HCcr Ž 1 q eo .x x log wŽ poq ⌬ u . rpo x

8. Case history Raffles City

for a soil that is normally consolidated. That is a soil that has not been subjected to a higher effective pressure than the present effective overburden pressure. po s pc . A lowering of the groundwater or piezometric level will generally cause large settlements.

A landmark project where permanent dewatering has been employed for control of water ingress and uplift is the Raffles City project ŽFig. 6. in downtown Singapore, located on the Bouldery Clay formation. The project comprises four high rise structures plus an eight-story

S s w HCrr Ž 1 q eo .x x log wŽ poq ⌬ u . rpo x


84

podium. The high rise structures are one 73-story tower, one 42-story tower and two 28-story towers, all of which are supported on rafts resting directly on the Bouldery Clay. The eight-story podium is supported on individual footings also resting on the Bouldery Clay. An under-drain system has been installed under the rafts of the towers and the base slab under the podium connected to nine sumps. The thickness of the drainage blanket is 600 mm. Pumping from the sumps keeps the basement dry and relieves the hydrostatic pressure that otherwise would build up under the rafts and the base slab. No tension piles or soil anchors are used. The top surface of the tower rafts is located 12.6 m below the ground surface. The settlements of the tower blocks and of the podium structure were monitored at regular intervals during the construction. The maximum settlements are 6᎐10 mm for the podium, 16᎐21 mm for the two 28-story towers, 27 mm for the 42-story tower and 48 mm for the 73-story tower. 8.1. Site conditions At the Raffles City site, the soil consists of fill and

Table 3 Index properties of clay matrix of bouldery clay at Raffles City Complex Index properties

Range of values

Liquid limit % Plasticity index % Percent fines % Percent clay %

37᎐54 21᎐36 64᎐100 38᎐54

the Bouldery Clay. The fill, a predominantly medium to coarse-grained silty sand, was encountered over the entire site and extended down to 14 m depth at the corner of Bras Basah and Beach Roads. The Bouldery Clay below the fill is very stiff to hard and is composed of predominantly clay. Gravel, cobbles and boulders are embedded in the Bouldery Clay. At the site, the thickness of this layer varies from 50 to 80 m. The index properties of the stiff clay matrix are given in Table 3. The properties of the Bouldery Clay have been reported by others ŽShirlaw et al., 1990; Wong et al., 1996.. The pressure meter modulus Žfrom tests done without unloading and reloading. was 40᎐150 MPa with an average value of 95 MPa. Laboratory permeability tests using triaxial machines on intact samples indicate that the permeability of the clayey silt and silty clay matrix material is very low, from approximately 5 = 10y1 0 to 8 = 10y1 1 mrs. Field variable head permeability tests done in bored holes indicate a higher permeability ranging from 2 = 10y6 to 2 = 10y8 mrs. The boulders encountered within the clayey silt and silty clay matrix are sandstone. In some of the boulders the sandstone has been weathered. However, in most boulders it is fresh with its unconfined compression strength ranging from 24 to 108 MPa. The bedrock beneath the Bouldery Clay consists of shale, siltstone and fine-grained sandstone, which are moderately to highly fractured and weathered. 8.2. Performance of under-drain system

Fig. 6. Raffles City complex.

The under-drain system in the three-level basement has performed very well since the buildings were put into service some 18 years ago. The flows into the sumps have been very small. Table 4 presents the quantity of flow into each of the sumps. No adverse effects on neighboring buildings have been reported as a result of the dewatering at Raffles City. Many of these neighboring buildings are or were pre-war structures supported on either footings or short piles and thus would be sensitive to settlements. In Fig. 7, photograph A shows the flow being measured at one of the sumps while photograph B shows another sump where the flow was also very small. As

I.H. Wong r Tunnelling and Underground Space Technology 16 (2001) 77᎐86 Table 4 Measured flow rates at sumps at Raffles City Complex

Sump 1 Sump 2 Sump 3 Sump 4 Sump 5 Sump 6 Sump 7 Sump 8 Sump 9 Total flow

Flow rate on 23 March 1995 Žlrmin.

Flow rate on 15 August 1996 Žlrmin.

Flow rate on 19 February 1997 Žlrmin.

Sunny day

Rainy day

Sunny day

0.165 0.097 0.143 0.106 0.085 nil 0.031 nil 0.025 0.652

0.515 0.113 0.355 0.269 0.097 0.061 0.098 0.016 0.075 1.599

0.151 0.068 0.145 0.089 0.046 Trickle 0.054 0.012 Trickle 0.565

noted in Table 4, on 2 of the 3 days when flow measurements were made, the weather was dry, and on the third day, it was raining. The total flow ranged from 0.6 to 1.6 lrmin, which is very small.

9. Beneficial effects of pressure relief system for waterproofing of basement A pressure relief system installed below the base slab of a basement has the added merit of helping to keep the base slab dry. Wong Ž1997. discusses the wetness problems of basement walls built of diaphragm walls. A survey of deep basements in Singapore by the author indicates that many of these basements have slabs that have leakage problems. Such leaking base slabs are unsightly, pose safety problems for tenants and members of the public, and generally downgrade the quality of life and enjoyment for those using the basement. By contrast Raffles City is very dry because of the pressure relief system installed. There are no evidences of any leaks through the base slabs. For the five-star

85

Fullerton Hotel, which was converted from the old Fullerton Building, the ballroom is located in the onelevel basement where an under-drain system has been installed to relieve the hydrostatic pressure.

10. Conclusions Methods of resisting hydrostatic uplift loads at the base of basements and other substructures include the use of tension piles, shear keys and under-drains. Tension piles that can be used to resist uplift loads include steel H piles, micropiles and bored piles. Published literature strongly suggests that steel piles installed in undisturbed, native soil undergo very little corrosion. Piles installed under a substructure would invariably be in natural soil since the overlying fill will mostly likely have been excavated. Micropiles similarly are suitable for resisting uplift loads. Bored piles are uneconomical structurally as tension piles because the concrete in the bored piles cannot carry tension loads. Installing shear keys into the soil beyond the perimeter of the basement is not practicable in situations where diaphragm walls, secant pile walls, or contiguous bored pile walls serve as permanent walls. Where ground conditions are suitable, under-drains are very effective for relieving the hydrostatic pressure acting at the base of substructures, particularly if the substructures are located in stiff and impermeable soils. For very stiff or hard soils, the lowering of the groundwater level induce little or no adverse effects on surrounding ground or adjacent structures For sites surrounded by soft soils or deep pervious layers permanent dewatering is not feasible because of the potential for large ground settlements and for the influence zone of the groundwater lowering to spread far from the substructure site.

Fig. 7. Photographs of sumps in under-drain system at Raffles City complex Žafter Anand, 1997..

86


References Anand, S., 1997. Design of basement slabs against hydraulic uplift. MSc Dissertation. Nanyang Technological University, Singapore. British Standard Institution, 1986. British Standard Code of Practice For Foundations, BS 8004:1986. London. Cedegren, H.R., 1967. Seepage, Drainage and Flow Nets. John Wiley, New York. Romanoff, M., 1962. Performance of steel pilings in soil. NBS Monograph 58, National Bureau of Standards. US Dept. of Commerce. Romanoff, M., 1969. Performance of Steel Pilings in Soil. Proc. 25th Conference. National Assoc. of Corrosion Engineers, USA. Shirlaw, J.N., Poh, K.B., Hwang, R.N., 1990. Properties and origins of

Singapore Boulder Bed. Proceedings 10th Southeast Asian Geotechnical Conference. Taipei, pp. 463᎐468. Tomlinson, M.J., 1994. Pile Design and Construction Practice, 4th edition Spon, London. Wong, I.H., 1997. Experience with waterproofness of basements constructed of concrete diaphragm walls in Singapore. Tunnelling Underground Space Technol. 12 Ž4., 491᎐495. Wong, I.H., Law, K.H., 1999. Corrosion of steel H piles in decomposed granite. J. Geotech. Geoenviron. Eng. ASCE 125, 529᎐533. Wong, I.H., Ooi, I.K., Broms, B.B., 1996. Performance of raft foundations for high-rise buildings on the Bouldery Clay in Singapore. Can. Geotech. J. 33, 219᎐236.

Methods of Resisting Hydro Static Uplift in Substructures

Recommend Documents