Induced Pore Pressures during Pile-Driving Operations

P ressio n s in te rstitie lle s p ro d u ites d u ra n t le b a tta g e d e p ieu x

K. Y. LO, Supervising Foundation Engineer, Department of Highways, Ontario, Canada

A. G. STERMAC, Principal Foundation Engineer, Department of Highways, Ontario, Canada

SUMMARY SOMMAIRE
A theory for estimating the maximum pore pressures induced On présente une théorie pour évaluer les pressions inter­
by pile driving is presented. Theoretical results are compared stitielles maximales produites durant le battage de pieux. Les
with measurements carried out at three sites, and some published résultats théoriques sont comparés aux valeurs expérimentales
data. The maximum pore pressures may be estimated with mesurées en trois sites différents. La théorie proposée permet
reasonable accuracy by the method proposed. Its magnitude d’évaluer de façon assez précise la pression interstitielle maxima
depends on the stress history and pore-pressure characteristic dont la valeur est fonction des caractéristiques de l’argile et
of the clay, but is independent of pile dimensions. Within a radial est indépendante des dimensions du pieu. A l’intérieur de la
failure zone of the soil surrounding the pile, the pore pressures zone radiale de rupture, la pression interstitielle est constante
are constant; outside this zone they decrease rapidly with dis­ tandis qu’à l’extérieur de cette zone, elle décroit rapidement en
tance from the pile. The pore pressure generally increases with fonction de la distance du pieu. En général, la pression intersti­
depth resulting from increase in effective overburden stress. tielle augmente avec la profondeur à cause d’une augmentation
There is no direct summation of pore pressures within the de la contrainte effective. Enfin, on n’a observé aucune augmenta­
failure zone of a pile owing to the driving of adjacent piles. tion de la pression interstitielle, dans la zone de rupture, due au
battage de pieux adjacents.

THE PHENOMENON OF PORE WATER PRESSURES being Set Up

when piles are driven into the ground has long been recog­ Th
nized. It is known that the dissipation of these pore pressures
is a governing factor in the increase of bearing capacity of °il
piles. It is also possible that the induced pore pressures may M o
affect the stability of slopes into which piles are driven. In INITIAL STATE
order to explore these problems, a knowledge of the magni­
—t-
tude and distribution of the induced pore pressures with
respect to both depth and distance from the pile is essential. / < \\
Information on pore pressures set up because of pile
driving is rather scanty. Some field measurements have been FINAL STATE
reported by Bjerrum, et al. (1 9 5 8 ), Bjerrum and Johannes- f ig . 1. Stresses in the failure zone around a
sen ( I 9 6 0 ), and recently, by Milligan, et al. (1 9 6 2 ). Theo­ pile.
retical treatments of the problem have been attempted by
Soderberg (1962) and Ladanyi (1 9 6 3 ). The inadequacy of
Before the pile is driven, the initial stresses in the ground
some theoretical approaches used has been discussed (Lo
are the vertical effective stress cr'u and the horizontal effec­
and Stermac, 1963). In this paper, the behaviour of pore tive stress together with the pore pressure w0. The verti­
pressures caused by driving a single pile and a group of piles
cal and horizontal stresses are also the principal stresses.
is described. A tentative method of estimating the maximum Since the horizontal strain in a natural deposit is zero, cr'8i is
induced pore pressure is proposed, based on the pore equal to K 0 <t'u where K a is the coefficient of earth pressure
pressure-strain theory (Lo, 1961). Results of field measure­
at rest.
ments at three sites are compared with the predictions of the
When the pile is driven, the direction of maximum dis­
theory. The theoretical framework is based on the behaviour
placement is radial. It follows that within the failure zone,
of normally consolidated clays, however. The application of
the radial stress becomes the major principal stress.* The
the theory to overconsolidated clays therefore requires
resulting maximum excess pore pressure A « m will then be
modification and is presently under consideration. It is also
composed of two parts, one resulting from the change in
obvious that pore pressures close to the pile tip cannot be total ambient pressure Acr3 and the other from shearing as
predicted by the theory because of end effects.
follows:
INDUCED PORE PRESSURE DUE TO SOIL DISPLACEMENT Aua — (1 — KoWu, (1)
When a pile is being driven into the ground, the soil is
displaced radially by the pile body. An element of soil close A us (A h / 1i, (2 )
to the pile will be deformed as shown in Fig. 1. Within a *It may be shown that the tangential stress <st > a7, the vertical
distance rn from the centre of the pile, shearing strain will be stress by using the plastic solution of a thick-wall sylinder sub­
so large that complete failure of the soil within this radial jected to both radial and axial flow under internal and external
zone will occur. stresses.

285
where (A u/p)m is the maximum pore pressure ratio and p is The sensitivity of the clay varies from 4 to 6 . Consolida­
the consolidation pressure. Hence, tion tests showed that the clay is normally consolidated.
Four series of triaxial tests with pore-pressure measure­
Aum = Aua + A Mg ments were performed on samples isotropically consolidated,
and one series of tests was carried out on samples cut
= [(I - Ko) + ( A u / p ) my 1{ . (3 ) horizontally. The pore-pressure ratios (Aw//7) m are plotted in
It is seen, therefore, that the estimation of the pore Fig. 3 against consolidation pressure for both vertical and
pressure set up as a result of pile driving involves the
determination of the coefficient of earth pressure at rest and ^ ^ R ange of effective overburden pressure

the maximum pore pressure ratio.

The pore pressure ratio Au/p is measured in a conven­
tional consolidated-undrained triaxial test with pore-pressure
measurements. In this test, the pore-pressure ratio Au/p SYMBOL SPECIMEN ELEVATION ( feet)
• Normal 540
increases with the applied stress difference, then reaches its x 539
maximum (A u/p)m, and remains constant after a certain ▼ ■■ 532
a .. 530
strain is attained. It has been shown that the ratio Au/p is o Horizontal 536
independent of the direction of the stress path, time of
0 10 20 30 40 50
sustained loading, and the consolidation pressure for nor­
CONSOLIDATION PR ESSU RE p ( lb. / sq in )
mally consolidated clays (Lo, 1961; Bjerrum and Lo, 1963).
Experimental evidence also suggests that (A u/p)m is inde­ fig . 3. Relation of (Au/p)m and p for Wallaceburg clay.
pendent of the stress system employed (Lo, 1963). The
advantage of using the pore-pressure ratio (A u/p)m is horizontal samples. N o significant difference in the pore-
therefore apparent. In this paper, the prediction of excess pressure behaviour was found between the two types of tests.
pore pressure due to pile driving will be limited to its maxi­ A value for (A u/p)m of 0.57 is representative in the nor­
mum value as already described. mally consolidated range. The coefficient of earth pressure
at rest K0 from five tests was found to be 0.5.
CASE 1, WALLACEBURG
Instrumentation
A comprehensive instrumentation programme was carried
out at a site 3.5 miles west of Wallaceburg at the southwest­ Two model steel piles, 3.5 in. in diameter, with closed ends
ern tip of Ontario. The subsoil stratigraphy, together with were driven by a drop hammer weighing 350 lb falling 30 in.
the results of field and laboratory tests on 3-in. piston Eight piezometers were installed at different distances and
samples from a typical borehole, are shown in Fig. 2. The depths from the piles as shown in plan in Fig. 4a and
points for laboratory tests represent the average result of at Table I. The first pile (A ) was driven to 45 ft and the
least two experiments. In-situ shear tests were performed by second (B ) to 48 ft. The two piles were spaced as close
vanes manufactured both by Geonor and by the Ontario
Department of Highways. The line plotted represents the
statistical line for a large number of tests at close intervals
with the Norwegian vane.

S H E A R IN G STRENGTH M O IS T U R E
CO NTENT U
W E IG H T
»A M PLE

5. / CU. ft.

S O IL
X D. H.O. Vone
h- D E S C R IP T IO N
O - Un confined Comp. w W
,w i
O Trioxio 1 Comp. j_ -
lb ./ sq.ft.
EL. 1 *
573- 3 G .L 2 00 400 600 800 20 40 60
00 O R G A N IC
S IL T a CLAY
V e ry S o ft
b-B
SANDY S I L T &
C L A Y EY SILT
L o o se

190 _ 1 14 2
cf

1 10 5

— G.l. Vane
1 08-2
—
—
V —cx
1
~1T~ > 115 5
fig. 4. (a) locations of piezometers; (b) distribution of induced
7~ 04 \ __Q 1 1 3-7
pore pressures, Wallaceburg.
~8~ ¿i \ t— o- 1 14 7
S IL T Y CLAY
S o il to F ir m 1 15-7
together as possible (6.5 in. edge to edge) with one group of
To I 17 2
piezometers in between so as to study the effect of the second
TT
pile on the pore pressures set up by driving the first pile.
II 1 8
TT AC
Three piezometers were installed 20 ft away from the test
116 2
-
site to determine the groundwater conditions and possible
' c 115 1
fluctuations during the test.
~Ï4~

TT o 1 1 2-2 Results of Field Measurements

û] \
94 V
The results of pore-pressure measurements are plotted in
SA N D
V e ry Dense
Fig. 5. The pore pressures due to the driving of each pile
are quite easily discernible. Observations show that the pore
fig . 2. Results of a typical borehole, Wallaceburg. pressure starts to increase when the tip of the pile is about

286
 Existing Ground -
-870-- SAND-GRAVEL-TOPSOIL

VARVEDCLAY
TIME (hours) SOFT
f ig . 5. Results of pore-pressure measurements, Wallaceburg.
P3 P2U“PIi
3 ft from the piezometer tip, then increases very rapidly
when the pile passes by the tip. The maximum pore pressure
for piezometers close to the pile is registered some 2 0 min
after the pile reaches the piezometer tip. This delay is
probably due to the time response of the piezometers. For
¡¡P7
piezometers at farther distances from the pile, some small
effect of the redistribution of pore pressures from regions of SAND a GRAVEL
high pressures to regions of low pressures is present.
The pore pressures at different depths and distances
induced by driving each pile are presented in Table I. In
Fig. 4b, the results for piezometers 35 ft deep are plotted f ig . 6 . Locations of piezometers and pile-driving
against distance from the piles. The sharp drop in pore sequence, Ghost River.
pressures between 1.5 and 2.0 ft from the pile is evident,
suggesting the existence of a failure zone of high pore pres­ tario. The soil stratigraphy is shown in Fig. 6 , and some
sures around the pile as postulated in the theory. The pore relevant soil properties may be found in Fig. 7. Artesian
pressures set up by the second pile are only about 30 per pressures were measured in all strata. A series of triaxial
cent of those due to the first pile within the failure region. tests was carried out on samples taken from the varved clay
However, outside this zone the summation of pore pressures stratum. The pore pressure ratio ( Au/p)m was found to be
is 100 per cent. These results are of considerable theoretical 0.65. The varved clay is slightly overconsolidated.
Attempts to recover undisturbed samples from the silt
TABLE I. RESULTS OF PORE-PRESSURE MEASUREMENTS, stratum were unsuccessful. Consequently, no triaxial tests
WALLACEBURG
were performed. However, both the standard and cone
penetration tests showed that the silt is very loose except
Piezometers A n (lb/sq. in)
near the bottom of the stratum. The behaviour of the silt
Distance (in.) Due to driving during shear is probably similar therefore to that of the silt
Depth laminae in the varved clay stratum. Since the triaxial samples
No. (ft) Pile A Pile B Pile A Pile B Total Calculated
of varved clay are composed of more than 70 per cent of silt,
K -l 25 G.O G.O 7.2 3.5 it is not unreasonable to assume that the ratio (A u/p)m of
107 \ 0 Q
K-4 25 7.5 14.0 7.5 1.6 9.1 / 99 the silt is of the same order of magnitude.* The coefficient
K -2 35 5.0 5.0 12.6 5.G 18.2 ) of earth pressure at rest is taken as 0.4.
K -5 35 7.5 17.5 15.5 1.8 17.3 |_ 14.0
K -7 35 20.0 20.0 14.4 2.1 16.5 J Instrumentation
K-8 35 26.0 26.0 5.3 4.9 10.2
50.0 2.7 2.6 5.3
Piezometers were installed at a test pile outside the
K-9 35 50.0
K-10 35 98,0 98.0 0.2 0.2 0.4 excavation and at the excavation to observe the pore pres­
sures resulting from driving a cluster of piles. The locations
and practical significance. It immediately suggests that when of the piezometers are shown in Fig. 6 .
a cluster of piles is driven, the pore pressures due to each At the test pile (10 BP 42 steel H -pile), seven piezometers
individual pile will not sum up directly, but will remain were installed. Piezometers P -l, P-2, P-3, and P-4, were
constant at a maximum value. This fact will be illustrated placed at a depth of 20 ft at distances of 2, 4, 8 and 15 ft
in another case record subsequently. from the centre of the pile. P-5, P -l, P -6 and P-7 were put
The calculated maximum pore pressures at 25 and 35 ft down at 10, 20, 30, and 40 ft depths at equal distances of
are respectively, 9.9 and 14.0 lb/sq.ft. and are shown in 2 ft from the pile. At the pier excavation, six piezometers
Table I. The calculated values agree reasonably with the were placed at Elevation 839, 25 ft below the bottom of the
measured pore pressures induced by the first pile. excavation.

C A S E 2, G H O S T R IV E R
*Tests carried out on a fine very loose sand by Bjerrum, Kring-
The site at which measurements were carried out was stad, and Kummeneze (1961) showed similarly high pore pres­
approximately 20 miles east of Matheson in Northern On- sure ratios.

287
 EX C E S S PO RE W ATER PR ESSU RES Au ( lb / s q .in .) rated energy of 22,500 ft-lb, but no increase in pore pressure
was observed. These observations illustrate further that the
pore pressures are predominantly strain-controlled.
The predicted and measured induced pore pressures at
different depths are shown in Fig. 7 for piezometers 2 ft
away from the centre of the pile. Data from measurements
in the pier excavation are also plotted in the same figure.
The calculated and measured values are in reasonable agree­
ment. The measured pore pressure at 40-ft depth is too low
because of the elapsed time required for splicing of the pile
at 36 ft. Some dissipation of pore pressure must have
occurred during this period. The artesian pressures existing
at the site are also plotted in the same figure.
Fig. 8 shows typical results recorded in piezometers E-3,
E-4, and E -6 at the excavation during the driving of a cluster
of piles. The numerals indicate the sequence of pile driving
corresponding to those shown in Fig. 6 . The piles were
driven at a fairly rapid rate, but the figure shows that a
certain value of pore pressure was not exceeded. In fact the
maximum value attained is approximately the same as that
found in the driving of a single pile as shown in Fig. 7. It is
clear, therefore, that within the failure zone of an individual
pile there is no direct summation of pore pressures due to
adjacent piles in a piled foundation.

C A SE 3, W A B I R IV E R , N E W L IS K E A R D

The layout of the foundation piles (14 BP 73 steel H-piles)

fig . 7. Comparison of calculated and measured together with the piezometers, is shown in Fig. 9a.
excess pore water pressure, Ghost River. The soil conditions were determined by extensive borings
and in one borehole, continuous 3-in. piston samples were
Results of Field Measurements taken to a depth of 41 ft. Briefly, the soil profile consists of
The results of measurements for the test pile are shown in 9 ft of mottled grey and brown silty clay, followed by a
Fig. 7. It is evident that the magnitude of the induced pore laminated clay to a depth of 23 ft. A thick deposit of varved
pressures increases with depth. Induced pore pressures at clay follows and extends to bedrock at 150 ft.
20-ft depth and at 2, 4, 8 , and 15 ft away from the pile were In the varved clay stratum, the thickness of the silt layers
6.2, 5.1, 2.3, and 0.8 lb/sq.in., decreasing with distance is 0.5 in. while that of the clay layers varies from 0.5 to 1 in.
from the pile. It is of interest to note that the maximum pore The liquid and plastic limits of the silt laminae average 28
pressure is registered almost instantaneously when the pile and 20 per cent and those of the clay laminae 65 and 26 per
tip passes by the piezometer tips for the piezometers which cent. The moisture contents of the silt and clay layers are
are closest to the pile, probably as a result of the relatively respectively 27 and 68 per cent. The undrained shear strength
high permeability of the silt. When the pile reached bedrock, of the bulk sample is 700 lb,/sq.ft. and the sensitivity is
it was given 100 blows by a Delmag diesel hammer with a approximately 10. The unit weight averages 110 lb/cu.ft.

f ig . 8. Typical results of measured pore pressure in a piled foundation, Ghost River.

288
 The results of this study lead to the following conclusions,
applicable to normally consolidated and very slightly over­
consolidated clays or loose silts.
1. The maximum induced pore pressure by driving a
single pile may be estimated with reasonable accuracy by the
method described.
2. The magnitude of the maximum induced pore pressure
depends on the pore pressure ratio (A u/p)m, and the initial
state of stress in the ground, but is independent of the
dimension or type of pile.
3. As a result of the increase in effective overburden
pressure, the induced pore pressures increase with depth at
the same radial distance from the pile. For the case records
studied herein, the maximum induced pore pressure varies
from 1.0 to 1.3 times the initial effective overburden stress
in the ground.
4. Within the failure zone of the soil surrounding the pile,
the induced pore pressures are maximum and equal. Driving
of adjacent piles only increases the pore pressure slightly.
Outside this zone, the induced pore pressures by driving
adjacent piles sum up directly until the maximum value is
f i g . 9. (a) layout of piles and piezometers; (b) distribution of attained. Therefore, the maximum pore pressures induced by
pore pressures, Wabi River. driving a number of piles in a pile foundation may be
predicted.
The varved clay is slightly overconsolidated. The ratio 5. Outside the failure zone of a pile, the pore pressure
(Au/p)m determined from specimens cut horizontally from decreases rapidly with distance. At a distance approximately
tube samples is 0.42. The coefficient of earth pressure at rest 16 times the diameter of the pile, the pore pressure is prac­
was found to be 0.5. Details of results of triaxial tests will be tically negligible. This result has been reported by Bjerrum
reported elsewhere (Lo, 1964). and Johannessen (1 9 6 0 ), and Milligan, et al. (1 9 6 2 ).
Results of Field Measurements ACKNOW LEDGM ENTS
The increase in pore pressure due to pile driving is plotted The field work was carried out by Messrs. A. Barsvary,
against distance from the centre of the pile in Fig. 9b. It is B. Ghadiali, H. Szymanski. The authors are indebted to
clear from the figure that the induced pore pressures increase Mr. H. W. Adcock, Assistant Deputy Minister (Engineering)
with depth at the same radial distance from the pile. The for permission to publish this paper.
short dotted lines indicate the estimated maximum induced
pore pressures at depths of 29 ft and 37 ft. The agreement REFERENCES
between calculated and observed values is not unreasonable. B je rr u m , L., J. B r i n c h H a n s e n , and R. S e v a l d s o n (1958).
Similar results as those shown in Fig. 8 at the Ghost River Geotechnical investigations for a quay structure in Horten.
site were obtained when the foundation piles were driven. Norwegian Geotechnical Institute, Publ. No. 28, pp. 1-17.
These results are therefore not presented herein. B j e r r u m , L., S . K r i n g s t a d , and O . K u m m e n e z e (1961). The
shear strength of a fine sand. Proc. Fifth International
S O M E P U B L IS H E D C A S E R E C O R D S Conference on Soil Mechanics and Foundation Engineering,
Vol. 1, pp. 29-37.
1. Bjerrum, et al. (1958) observed that the induced pore B j e r r u m , L., and I. J . J o h a n n e s s e n (1960). Pore pressures
pressure within the foundation of 35-cm-square concrete resulting from driving piles in soft clay. Conference on Pore
piles in a soft clay was as high as the effective overburden. Pressure and Suction in Soil, pp. 14-17.
The soft clay has a sensitivity of approximately 3. (A u/p)m B j e r r u m , L., and K. Y. Lo (1963). Effect of aging on the shear
may therefore be assumed to be 0.6 and K0 is probably 0.5. strength properties of a normally consolidated clay. Géo­
From Eq 3, the estimated A « m is 1.1 times the effective technique, Vol. 13, pp. 147-57.
overburden. L a d a n y i , B. (1963). Expansion of a cavity in a saturated clay

2. Bjerrum and Johannessen (1960) recorded induced medium. Proc. American Society of Civil Engineers, Vol. 89,
pore pressures of 5 tons/sq.m . at depths of 7.5 m and 10 m SM4, pp. 127-61.
Lo, K. Y. (1961). Stress-strain relationship and pore-water
depth in a foundation of 2 0 -cm-square, hollow-box steel
pressure characteristics of a normally consolidated clay.
piles. The sensitivity of the soft clay is 6 . The (A u/p)m ratio Proc. Fifth International Conference on Soil Mechanics and
may be taken as 0.7 and K n assumed to be 0.6. The estimated Foundation Engineering, Vol. 1, pp. 219-24.
values of A«m are 7.5 and 9.6 tons/sq.m ., respectively. Since --------- (1963). Discussion on “Study of failure envelope of
the closest piezometers were 1 m away from the pile, the soils” by T. H. Wu, A. K. Loh, and L. E. Malvern. Proc.
maximum induced pore pressures were not measured. H ow ­ American Society of Civil Engineers, Vol. 89, SM4, pp.
ever, the observed values lie below the predicted pore 243-5.
pressures. --------- (1964). Geotechnical properties of a varved clay deposit
in Northern Ontario. (To be published.)
C O N C L U S IO N S
Lo, K. Y., and A. G. S t e r m a c (1963). Correspondence. Géo-
technique, Vol. 13, pp. 163-4.
A theory based on the pore pressure-strain relationship M i l l i g a n , V., L. S o d e r m a n , and A. R u t k a (1962). Experience
for estimating the maximum induced pore pressure due to with Canadian varved clays. Proc. American Society of Civil
pile driving has been developed. The predictions of the Engineers, Vol. 88 , SM4, pp. 32-67.
theory are compared with field measurements in three case S o d e r b e r g , L. O. (1962). Consolidation theory applied to foun­
records, and some available information in the literature. dation pile time effects. Géotechnique, Vol. 12, pp. 217-25.

289