PROCEEDINGS OF THE 3RD INTERNATIONAL GEOTECHNICAL SEMINAR ON DEEP

FOUNDATIONS ON BORED AND AUGER PILES/ GHENT/ BELGIUM/19-21 OCTOBER 1998

Deep Foundations on Bored

and Auger Piles
BAP III
Edited by
W.F.VAN IMPE & W.HAEGEMAN
Laboratory of Soil Mechanics, Ghent University, Belgium

Effect of construction time and bentonite viscosity on shaft capacity of

bored piles
N. Thasnanipan, G. BAskaran & M. A. Anwar
Seafco Company Limited Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 1998

 Deep Foundation on Bored and Auger Piles, Avan Impe & Haegeman (eds) © 1998 Balkema, Rotterdam, ISBN 90 5809 022 1

Effect of construction time and bentonite viscosity on shaft capacity of bored

piles

N. Thasnanipan, G. Baskaran & M. A. Anwar

Seafco Co., Ltd., Bangkok, Thailand

ABSTRACT: Effect of construction time and slurry viscosity on the shaft friction capacity of cast in situ
bored piles have been studied by various investigators in the past but the extent to which these parameters
effect the shaft capacity is still not clear. Major obstacle in this regard is that it is hard to normalize the shaft
capacity degradation against other numerous parameters which are influencing simultaneously. Results from
eleven pile load tests, constructed with different slurry viscosities and construction times, in the layered
alluvial strata of Bangkok are presented in this paper.
It has been concluded that slurry viscosity, do not have significant effect on the shaft load transfer of these
piles but considerably reduced with increase in construction time. Trend of reduction in capacity seems to
follow an exponential decrease with increase in construction time with major part of degradation within first
24 hours of construction time.

1 INTRODUCTION 60 m from ground level are quite usual. Soil profile

of plain of Bangkok generally consists of Quaternary
Shaft friction capacity of bored piles, especially, in alluvial deposits of alternating sand and clay layers
granular soils is greatly affected by the installation as shown in Figure 1, down to the rock face which is
procedures, but the sole effect of bentonite slurry reported to be at least 550 m deep in the area
(slurry) properties and its exposure time is still (Balasubramanium, 1991).
unclear. Results from the previous researches on 2
Average Engineering P o re P re s sure (t/m )
deleterious effects of slurry properties and its contact Parameters

time on the shaft load transfer of wet process bored 0 20 40 60

0 0
piles are a bit anecdotal and to some extent Weathered Crust

contradictory. Major problem in this regard is that, Su = 10 - 20 KPa

γt = 14-16 KN/m3
the effect of either parameter can not be totally
Bangkok Soft Clay (BSC)
(CH)
10 -1 0

isolated from other influencing variables. Due to

this reason researches have been found to report
almost converse conclusions for the similar type of 20 Medium to Very Stiff Clay Su = 40 - 140 KPa
-2 0
Depth (m)

(CH) γt = 17-21.0 KN/m3

soils.
In order to assess the effect of slurry viscosity and
exposure time on the shaft load transfer, data from
Actu

30 -3 0
Hy

Dense to Very Dense SPT-N = 20 - 50

γt = 20.0 KN/m3
ro
al P

Sand (SM)

eleven instrumented bored pile load tests is

sta
ie

tic
zom

Lin

presented in this paper. All test piles were

etric

40 Su = 250 KPa -4 0
Dra

constructed using same type of equipment and Hard Clay (CL)

γt = 21.2 KN/m3
wD
o

procedures, in a similar type of soil conditions. So it

wn
Lin

is assumed that the influence of other variables

e

50 -5 0

would be minimized and the effect of slurry Dense to Very Dense

Sand (SM) SPT-N = 50 - 100
γt = 20.0 KN/m3
viscosity and exposure time, to some extent, could
be estimated. 60 -6 0

2 BORED PILES AND SOIL CONDITIONS

Figure 1. Typical soil profile of Bangkok plain with
Use of bored cast in situ piles in Bangkok, is very piezometric draw down conditions.
common. Diameters of these piles normally fall in
the range of 0.80 to 1.50 m and toe depths down to
171
Below the top few meters of weathered crust, a thick noted that major part of the reduction i. e. 43 percent
layer of well known “Bangkok Soft Clay” (BSC) took place in the first 8 hours of contact time and
changing to medium at about 15 to 18 m, is present. only 13 percent (56-43) reduction took place in the
First sand layer is usually 5 to 10 m thick and found rest of 89 hours (97-8), for the second barrette.
at 25 m to 30 m depth, below is a series of stiff to These findings support the concept that major part of
hard clay and medium to very dense silty sand the shaft friction capacity reduction with increase in
layers. Actual pore pressure conditions in the upper construction time, took place in first few hours of
BSC are hydrostatic from approximately 1 m below construction time and further increase in
ground level. Then the hydrostatic conditions construction time have minor contribution towards
changes to piezometric draw down near the bottom the reduction in shaft friction capacity.
level of BSC, Figure 1. Piezometric draw down
resulted in increased effective overburden pressure Corbette (1975), presented the results of load testing
of about 20 ton/m2 in the first sand layer and below. of a diaphragm wall panel excavated in conjunction
The depressurization of sand layers is attributed to with bentonite suspension, and construction time of
the deep well pumping in the area, Nutalaya (1981). approximately 48 hours. He inferred that the
presence of bentonite suspension even for such a
long time has not adversely affected the
3 REVIEW OF PREVIOUS WORK development of skin friction.

Degradation of perimeter load transfer in permeable Littlechild and Plumbridge (1998), concluded from
soil layers is attributed to the formation of filter cake the pile load tests of BERTS project Bangkok, that
which is left in place and get sandwiched at soil-pile shaft friction capacity of bored piles constructed
interface and is not scrapped of by fluid concrete. under slurry tends to decrease as the construction
Scouring capability of fluid concrete depends upon time and slurry viscosity increases.
its shear strength and that of the filter cake and it is
argued that if shear strength of filter cake is more Majano and O’Neil (1993) attempted to model the
than the fluid concrete, it can not be scoured and left formation of filter cake in the laboratory with
in place resulting in the degradation of load transfer different slurry dosages, differential pressures and
capacity of the shaft. In case of impermeable soils exposure times. They stated that perimeter load
reduction in the shaft load transfer is mainly due to transfer is a complex function of the physical and
the softening of the soil at the perimeter of the chemical characteristics of the slurry and the
borehole, because the formation of thick filter cake geomaterials, the roughness of the borehole, the
is not possible as slurry can not permeate in to the fluid pressures exerted by concrete, the shearing
soil, but some researchers e. g. Veder, C. (1963) properties of the soil, and possibly the chemistry of
observed a thin cake of few millimeters even in clay the fluid concrete. They attempted to correlate the
formations and argued that it may be the result of potential degradation of the soil-pile interface to the
electrical forces or chemical reaction of bentonite thickness and shear strength of the cake formed
suspension on the wall of the borehole. against the walls, and argued that the shear strength
and thickness can not be measured with only one or
O’Neil et al. (1992), reported the effect of slurry two parameters like slurry dosage and differential
properties on two drilled shafts constructed in pressure. With a bentonite concentration of 72
similar ground conditions (alternating layers of very kg/m3 and differential pressure of 0.5 psi they
stiff clay and dense sand). Shafts 1 and 2 were achieved a filter cake thickness of 3.14 and 4.5 mm
constructed with Marsh cone viscosity of 37 and 49 after a contact time of 4 and 24 hours. Authors also
sec. respectively, time between opening the compared the laboratory tests with the actual field
borehole and completion of concreting for both piles conditions and proved that due to the presence of
was 5 to 7 hours. Minor difference in load transfer enormously high differential pressures in the field in
of both shafts was found and authors concluded that some cases, thicknesses of filter cakes which would
the difference in slurry properties used for the be considered detrimental for shaft load transfer,
construction for both shafts had a little effect on require few hours to develop (They proved that a
shaft load transfer. filter cake of 6 mm thickness would require only 2.3
hours to develop in the field under a differential
Cernak (1976), performed full-scale load tests on pressure of 10 psi).
three barrettes in sandy gravels, two barrettes were
constructed using slurry with different exposure
times, 8 and 97 hours. Load test results indicated a 4 METHOD OF CONSTRUCTION
decrease of skin friction capacity of 43 percent and
56 percent, respectively, as compared to the barrette All the eleven piles reported in this paper were
excavated dry and concreted immediately. It must be constructed, following the internationally accepted
172
guidelines and specifications. Slurry in conjunction must also be noted that the actual contact time for
with rotary bucket is used for all piles. slurry starts, once it is fed to the borehole till the up
Temporary casing, having length depending upon flowing concrete reaches the bottom level of
the thickness of the BSC at the site is driven using a temporary casing. Since the accumulated time
high frequency vibro-hammer. Drilling operation is required for concreting above the bottom level of
commenced with an auger down to the base of the temporary casing and auger drilling inside the
temporary casing. Further drilling below the bottom temporary casing roughly required one hour,
of temporary casing is continued using a drilling construction times given in Table 2 are supposed to
bucket by first filling the borehole with slurry. be appropriate for comparison.
Bentonite, classified as Activated Sodium Bentonite
is usually used and slurry properties are controlled
within the limits given in Table 1. 5 SHAFT CAPACITY ESTIMATION

Table 1. Slurry specifications followed. A commonly accepted design approach which

Slurry Property Control Limits consists of a combination of total stress method for
Viscosity (Marsh cone) 30 - 60 clay layers and effective stress method for sand
layers is adopted to estimate the shaft friction
Density (g/cc) 1.04 - 1.15 capacity reported in Table 2, col. F. Skin friction
pH 7 - 11 capacity of clay layers is estimated as fs = α.Cu,
Sand Content <4%
with adhesion factor α. taken from the curve
proposed by Suchada (1989), Figure 2, for Bangkok
subsoils. It can be noted that value of α used is quite
comparable to the previous recommendations by
Outer diameter of the drilling bucket is slightly less Tomlinson (1957) and Kulhawy (1984).
than the nominal size of the pile to provide an
annular by-pass for the slurry during lowering and
hoisting operations. Required nominal size of the 1.20
Tomlinson,1957
Bored Cast Insitu
pile is achieved using a cutting edge with Piles
sidecutters. Slurry level in the borehole is 1.00
Stas and Kulhawy, 1984

maintained within 1 to 1.5 m from top of temporary

casing. After drilling down to the required tip level 0.80
α

of pile, special cleaning bucket and/or air lift

Adhesion Factor ,

Suchada,1989
technique is normally applied to clean the borehole 0.60 for bored piles in Bangkok soils

base of any congregated sediments before lowering

the reinforcement cage. Concreting is done using a 0.40

tremie pipe at the center of the borehole. Total time

to carry out this operation, starting from the casing 0.20

driving till the completion of concreting, usually fall

in the range of 10 to 20 hours excluding some 0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
accidental delays due to equipment break down, Undrained Shear Strength (ton/m )
2

delay in ready mixed concrete due to traffic

congestion or similar reasons. In most of the cases
whole operation of pile construction is accomplished
within 24 hours, but in some cases, due to the local Figure 2. Comparison of Adhesion factor α used for bored piles
authority regulations which do not permit to operate in Bangkok Subsoils.
during night time hours, borehole has to be left
exposed for 10 to 12 hours unagitated. In such
cases, drilling is intentionally curtailed above the Undrained shear strength Cu of shallow clay layers
final excavation level and resumes next day. We is normally determined in the laboratory with
assume that in such cases, during drilling after a unconfined compression test while for deeper stiff to
delayed period of 10 to 12 hours, any possible filter hard clay layers, Cu is indirectly estimated from
cake on the borehole walls is automatically scraped Standard Penetration Test (SPT-N), as Cu =
of. Construction times more than 24 hours reported C1.SPT-N (ton/m2), with C1 equal to 0.674 and
in Table 2 are examples of such cases. Construction 0.507 for high and low plasticity clays respectively.
time reported in Table 2, col. D is the time elapsed C1 values mentioned here are based on the statistical
between the start of auger drilling to the finishing of correlations suggested by Pitupakorn, 1985. Skin
concreting. Longer construction times for the test friction capacities for the sand layers are calculated
pile reported are usually due to the time consumed in as fs = σv’. Ks.tan.δ with coefficient of horizontal
instrumentation installation with the rebar cage. It earth pressure Ks equal to 0.7 and δ equal to 0.75 φ.
173
 Table 2. Construction parameters and shaft capacities of test piles.

Bentonite Pile Age at

Pile Slurry Construction the Time of Estimated Actual Shaft Actual/Estimated
Pile No. Dimensions Viscosity Time Load Testing Shaft Capacity Shaft Capacity
(m) (sec.) (hrs) (days) Capacity (ton)
(ton)
A B C D E F G H=(G/F)
TP-1 φ1.2x57.1 38 43.0 32 1230 1200 0.98
TP-2 φ1.2x46.3 38 13.1 29 930 1300 1.40
TP-3 φ1.0x46.5 37 9.8 25 800 1250 1.56
TP-4 φ1.0x49.5 37 38.7 15 930 750 0.81
TP-5 φ1.0x43.0 38 26.0 19 690 700 1.01
TP-6 φ1.0x41.0 45 11.8 32 620 800 1.29
TP-7 φ1.2x43.6 55 16.8 32 850 1100 1.29
TP-8 φ1.2x43.5 38 12.3 24 640 750 1.17
TP-9 φ1.0x43.5 37 11.3 18 530 700 1.32
TP-10 φ1.2x62.0 41 32.4 27 1534 2000 1.30
TP-11 φ1.2x54.2 38 30.0 39 1050 1300 1.24

for 12 hours followed by second cycle of loading up

Value of Ks used is in good agreement with the to 2 times of Qd and maintained for 24 hours. Third
recommended values by different researchers like cycle of loading is applied up to 2.5 times Qd or
Fleming (1977), who suggested a value not more maximum pre-decided test load. Occasionally, a
than 0.75 to account for the possible degradation due fourth quick loading cycle up to the maximum test
to the formation of filter cake. Angle of internal load is also applied and maintained for two hours.
friction φ is also estimated from SPT-N values by All test piles, except TP-2 and TP-10, were tested to
first correcting the N values for overburden well above the expected failure loads with the
correction (Bowels, 1988). Some designers also use concept of sufficiently mobilizing the end bearing.
an equivalent term of β instead of Ks.tan.δ It must Load displacement curves of the test piles are shown
be noted that the empirical parameters discussed in Figure 3 for comparison.
here are based on the effective overburden pressure
σv’ which is calculated with the assumption of
hydrostatic conditions from the ground level, 7 ACTUAL SHAFT FRICTION CAPACITIES
ignoring the piezometric draw down conditions
below BSC shown in Figure 1. Actually mobilized shaft friction capacities are
calculated from the VWSG load transfer results at a
level where load displacement curves show a
6 TEST PILES maximum curvature and are given in Table 2, col. G.

Test piles discussed in this paper were all constructed

following the procedure discussed in section 4. All 8 EFFECT OF SLURRY VISCOSITY
piles were instrumented with Vibrating Wire Strain
Gauges (VWSG) at five to seven different levels to Effect of slurry viscosity on the shaft load transfer
estimate the shaft friction transferred to different soil can best be seen from the plot between the ratio of
horizons. One set of telltale extensometer rods was Actual/Estimated shaft capacity versus slurry
also used near the tip of each test pile to measure the viscosity, Figure 4. Over all scatter of data do not
elastic shortening of the pile shaft and finally pile suggest any trend. Since the effect of slurry viscosity
base movement at different stages of load testing. is difficult to separate, without normalizing the
Static maintained load testing method was used for effect of construction time. Test piles TP-6, 7 and 9
all piles. Test loads were applied using a system of have a slight difference in construction time and pile
hydraulic jacks against the reaction frame of steel age at load testing but slurry viscosity is
girders fixed against anchored reaction piles. significantly different, 45, 55, and 37 sec.
Normally three cycles of loading are applied with respectively but the ratio of Actu./Esti. shaft
first cycle up to the design load Qd and maintained capacities for these piles are necessarily the same
174
 Applied Load (Tons)
5, plot between Actu./Esti. shaft capacity versus
0 500 1000 1500 2000 2500 construction time. Though the data is widely
0 scattered, best fit curve shows the trend of
exponential decrease of shaft load transfer with
20
TP-1
increase in construction time. Best fit curve for the
TP-2
TP-3
available data also reveals that the available shaft
40 TP-4 capacity is more than the estimated capacity for a
Movement (mm

TP-5
TP-6
construction time as long as 40 hours. This confirms
60 TP-7 the validity of recommendations made by some
TP-8
TP-9 researchers like Fleming (1977), who suggest to
80 TP-10
TP-11
complete the concreting within 24 hours after
finishing the drilling, and seems to be more
100
reasonable because slurry is kept continuously
120
agitated during drilling operation and actual
exposure time for slurry starts, once the drilling is
Figure 3. Load settlement curves of the test piles. completed. Test pile TP-4 which exhibited the
lowest capacity and has the minimum slurry
which suggest that the sole effect of slurry viscosity viscosity (37 sec.), second longest construction time
on shaft capacity reduction is insignificant. (38.7 hours) and the minimum age (15 days) at the
Additionally, TP-1, 4 and 5 which exhibit a time of load testing, in the group reveals that the
maximum reduction in shaft capacity have almost shaft capacity reduction is contributed by increased
same slurry viscosity but different construction time construction time as well as reduced pile age while
which confirms that slurry viscosity within the data slurry viscosity have minor effect.
set available do not directly have any effect on the It can also be noted that the major part of the shaft
shaft load transfer of the piles. capacity reduction is contributed by the first 24
hours of construction time which normally
considered feasible and agreeable by most of the
9 EFFECT OF CONSTRUCTION TIME piling contractors under normal circumstances, and
further delays beyond 24 hours have a minor
Since ten out of eleven bored piles were constructed addition to the shaft load capacity degradation.
with slurry viscosity varying within narrow range These findings are in good agreement with the
effect of construction time can best be seen in Figure previous observations discussed in section 3.

1.60
25

TP-1
1.40 29(Pile age at load testing, days) TP-2
Actual / Estimated Shaft Capacity

18 27 TP-3
32 32
39 TP-4
1.20
24
TP-5
TP-6
1.00 19
32 TP-7
TP-8
TP-9
0.80 15
TP-10
TP-11
0.60
35 40 45 50 55 60
Bentonite Viscosity (sec)

Figure 4. Effect of slurry viscosity on shaft capacity of bored piles.

175
 1.60
25
TP-1(Tip in SAND)

TP-2(Tip in SAND)
Actual / Estimated Shaft Capacity

1.40 29 (pile age at load testing, days)

TP-3(Tip in SAND)
18 27
32 32 TP-4(Tip in CLAY)
39
1.20 TP-5(Tip in SAND)
24
TP-6(Tip in SAND)

19 TP-7(Tip in CLAY)
1.00
32
TP-8(Tip in SAND)

TP-9(Tip in SAND)
0.80 15
TP-10(Tip in
SAND,Base Grouted)
TP-11(Tip in
SAND,Base Grouted)
0.60
5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0
Construction Time (hours)

Figure 5. Effect of construction time on shaft capacity of bored piles.

10 DISCUSSION or Maintained Load (ML). Reduction in construction

time, generally, lead to better performance of pile
From Figure 5 it can be noted that actually shafts but this reduced construction time could
mobilized shaft capacity is equal to the estimated never exclude the accidental occurrence of delays.
shaft capacity for a construction time of Any potential degradation in load carrying capacity
approximately 35 - 40 hours or in other words, the of such piles caused by uncontrollable delays must
empirical design parameters commonly used to be considered along with the conceivable trend in
estimate the shaft friction capacities discussed in reduction under these conditions.
section 5, are justifiable for a total construction time It must also be noted that because of high
of longer than 24 hours, especially, to account for differential pressures imposed due to depressurized
the unexpected delays during construction in the sand layers in Bangkok sub soils (differential
downtown areas of Bangkok which are notorious for pressures of the order of 30 psi will be imposed if
its traffic congestion and other constraints imposed slurry level is maintained near the ground level due
by local authorities. This also confirms that the to piezometric draw down of 20 m in the sand
allowable time of 24 hours after finishing the layers), formation of filter cakes which are
drilling operation (Fleming, 1977) is also justifiable. considered to degrade the shaft load transfer would
However, with no dispute, shorter the construction theoretically take fraction of seconds to develop, if
time better the shaft load transfer response. we calculate on the basis of procedure recommended
It must therefore be emphasized that the selection by Majano (1993). This means that the degradation
of empirical parameters like α, β, δ and Ks, to of pile shaft interface due to the formation of filter
estimate shaft load capacities at the design stages cake in permeable soil layers is unavoidable within
must also consider the practically possible the practical construction time limits.
construction times by suitably incorporating the
effects of any possible delays specific to the site and
working conditions. Empirical factors mentioned 11 CONCLUSIONS
above are not the fundamental properties of the soil-
pile system, and need to be accounted for the 1. Bentonite slurry viscosity within the data set
installation procedures or even the method of pile available do not have significant effect on the
load testing like Constant Rate of Penetration (CRP) shaft load transfer of bored cast in situ piles.
176
2. Longer construction time reduces the shaft Suchada, (1989), Bored Piles in Bangkok Subsoils, , M. Engg.
carrying capacity of bored piles constructed Thesis, AIT, Bangkok.
Tomlinson, M. J. (1957) The adhesion of Clay Piles Driven in
under slurry. However, the major portion of Clay Soils, Proc. 4th Intl. Conf. on Soil Mechanics and
degradation occurs within the first 24 hours Foundation Engineering, London.
period, which is unavoidable under normal O’ Neil et al.(1992), Behaviour of Axially Loaded Drilled
construction time. Shafts in Beaumont Clay, The University of Texas at
3. Degradation of shaft capacity with increase in Austin.
construction time beyond 24 hours is not Veder, C. (1963), An Investigation on The Electrical
Phenomena At The Area of Contact Between Bentonite
excessive and can be adjusted within the Mud And Cohesionless Material, 5th Intl. Conf. on Soil
empirical design parameters or factor of safeties Mechanics, Paris.
chosen at the design stages.
4. Piles constructed within 24 hours showed
generally higher shaft capacities than estimated
as per normal design practice adopting
empirical parameters suggested by different
researchers for Bangkok subsoils and can safely
be used for slurry viscosity up to 55 sec. and
total construction time up to 40 hours.

AKNOWLEDGEMENTS

Writers wish to acknowledge the co-operation of

Mr. Aung Win Maung and Mr. Pornpot Tanseng, in
the preparation of this paper.

REFERENCES

Baker, Jr. C. N. et al. (1970), Load Carrying Capacity

Characteristics of Drilled Shafts Constructed with the Aids
of Drilling Fluids, Research Report 89-6, Center for
Highway Research, The University of Texas at Austin, TX.
Balasubramanium, A. S. (1991), Inaugural Lecture on
Contributions in Geotechnical Engineering Soil Mechanics
and Foundation Engineering, AIT, Bangkok, March 14,
1991.
Bowles, J.E.(1988), Foundation Analysis and Design,
McGraw-Hill.
Cernak, B. (1976), The Time Effect of Suspension on the
Behaviour of Piers, 6th European Conf. On Soil Mechanics
and Foundation Engineering, Vienna.
Corbett, B. O. et al. (1975), A Load Bearing Diaphragm Wall
at Keinsington and Chelsa Town Hall, London, Conf. on
Diaphragm Walls and Anchorages, ICE, London.
Fleming, W. K. (1977), The Use and Influence of Bentonite in
Bored Pile Construction, CIRIA Report PG3.
Littlechild, B. and Plumbridge, G. (1998), Effects of
Construction Technique on the Behaviour of Plain Bored
Cast In Situ Piles Constructed Under Drilling Slurry, 7th
Intl. Conf. and Exhibition on Piling and Deep Foundations,
Vienna.
Majano, R. E. and O’Neil (1993), Effect of Slurry Dosage and
Exposure Time on Perimeter Load Transfer of Drilled
Shafts, ADSC Publication.
Nutalaya, P. (1981), Subsidence Studies of the Bangkok Plain,
AIT Research Reports.
Pitupakorn, W. (1985), Predictions of Prestressed concrete pile
capacity in Bangkok stiff clay and clayey sand, 8th
Southeast Asian Geotechnical conf., Kuala Lumpur.
Stas, C. V. and Kulhawy, F. H. (1984), Critical Evaluation of
Design Methods for Foundations Under Axial Uplift and
Compression Loading, Report EL-3771, EPRI, Palo Alto,
Calif.

177