Documente Academic
Documente Profesional
Documente Cultură
Evaluating Pile Drivability for Hard Clay, Very Dense Sand, and Rock
by Robert S. Stevens, McClelland Engineers, Inc.; Edward A. Wiltsie, Arabian American Oil Co.; and
Thomas H. Turton, McClelland-Suhaimi, Ltd.
8../8._
unc- = (OCR)0'85 (2) Rock
where: Driving piles into rock is anticipated to severe
= actual undrained shear strength of clay ly fracture the rock layers and reduce the rock to a
having a given PI, and granular material. Therefore, unit skin friction for
= undrained shear strength of the same piles driven in rock layers is computed assuming sand
clay if normally consolidated. parameters. Silty sand parameters are assigned to ro
layers that are interbedded with silt or clay seams a
According to a relationship described by Skempton, layers.
s = 5 (0.11 + 0.0037~PI) . . . ". .(3) For poor to fair quality rock, unit end bearing
unc v
where: is limited to values given for granular soils. For
5 = effective overburden pressure, and more competent rock, unit end bearing is computed usi
the following equation:
PI = plasticity index.
.(6)
Unit end bearing in clay is computed using the
following equation: where compressive strength of rock, and
= diaensionless bearing capacity factor.
q = s N .(4)
^ u c
A value of 3 is used for N .
where: u
N = dimensionless bearing capacity factor. Lower and upper bound values of soil resistance
A value of 9 is used for N . to driving are computed for both coring and plugged
c pile conditions. When a pile cores, relative movemer
between pile and soil occurs both on the outside and
Granular Soils
inside of the pile wall. Skin friction is, therefore
For piles driven in granular soils, the unit skin developed on both outside and inside pile wall. The
friction during continuous driving is computed using end bearing area is equal to the cross-sectional ares
static pile capacity procedures, and is based on the of steel at the pile tip. When a pile plugs, the soi
equation: plug moves with the pile during driving. Skin fric-
tion is mobilized only on the outer wall. The end
f = K 5 tan 6 (5) bearing area is the gross area of the pile. Unlike
static pile capacity computations, end bearing is not
466
limited to the frictional resistance developed by the counts are badly underestimated using either the static
soil plug. pile capacity, or the soil resistance to driving for
both the lower and upper bound coring cases. Good
The resistance to driving should be based on a agreement is obtained for the lower and upper bound
reasonable upper bound in contrast to static pile plugged cases. The driving records for all six piles
capacity, which is based on a reasonable lower bound. are shown on Figure 4. Blow counts were computed using
When a pile cores, a lower bound value is computed the measured hammer properties for each pile. The low-
assuming that the skin friction developed on the in- est blow counts predicted for the lower bound plugged
side of the pile is 50 percent of that on the outside case and the highest blow counts predicted for the up-
of the pile. For the upper bound coring case, the in- per bound plugged case are also plotted in Figure 4.
ternal skin friction is assumed equal to the external Our predicted range in blow counts is in good agreement
friction. with the scatter observed in the field. The plugged
cases give higher predicted blow counts than the coring
When a pile plugs, a lower bound value of the cases not only because the soil resistance to driving
resistance to driving is computed using the procedures is larger, but also because the percent tip resistance
presented in the preceding paragraphs to compute unit is larger.
skin friction and unit end bearing. For the upper
bound plugged case, values of unit skin friction and A comparison of observed and predicted driving
unit end bearing for granular soils are developed by records for a second site in the Safaniya field is
increasing the lower bound values of unit skin frictioi presented in Figures 5 and 6. A Vulcan 340 hammer
by 30 percent, and increasing unit end bearing by 50 with a steel plate and wire rope cushion was used to
-percent. A_corr espojnd ing_inereas_e_ ia_ 1 imit ing values drive two piles to jlesign^penetration,, and_tp^nitial-
for unit skin friction and unit end bearing is assumed ly drive the remaining four piles. Final driving of
For cohesive soils, end bearing is computed using N = these piles was accomplished using a Vulcan 560 ham-
15, an increase of 67 percent. Unit skin friction mer. As was done previously, we have plotted the pre-
is not increased. dicted blow counts determined from the hammer data
giving the minimum value for the lower bound plugged
CASE HISTORIES case, and the maximum value for the upper bound plugged
case. Good agreement is obtained assuming a. plugged
The purpose of the case histories was to confirm pile.
the soil quake and damping parameters used in our wave
equation analysis, and to determine the best proce- HARD CLAY PROFILES
dure for estimating the soil resistance to driving.
Our evaluation was made by comparing predicted and Our experience with pile driving in hard clays in
observed driving records. The estimated soil resis- the Arabian Gulf has shown that blow counts may be con-
tance to driving at a given penetration was used to stant with penetration, blow counts may be very erratic
determine the corresponding blow count from wave equa- due to seams and thin layers of rock, and driving de-
tion results obtained using the measured hammer per- lays will result in an increase in blow counts during
formance data. redriving.
Unless otherwise stated, each case history is the The soil stratigraphy and curves of estimated
installation of a six pile platform in the Arabian soil resistance to driving are shown in Figure 7 for
Gulf, consisting of 42-in.-diameter piles having a con- a site in the Zuluf Field. The pile capacity factor
stant 1.5-in. wall thickness, with a 5-ft long driving computed for this site is equivalent to a sensitivity
shoe having a wall thickness of 1.75 in. The case of 2.18. The piles were driven through a hard calca-
histories are divided into the following categories: reous clay stratum present from 26 to 69-ft penetra-
very dense sand, hard clay, mixed profile, and rock. tion. The clay contains gypsum seams and layers at 43
to 46-ft and 54 to 55-ft penetrations, and numerous
VERY DENSE SAND PROFILES fragments of gypsum and claystone. Final pile driving
was accomplished with a Vulcan 560 hammer with a cushioi
Blow counts generally increase with depth in most made up of alternating layers of 0.25-in.-thick steel
sand profiles encountered in the Arabian Gulf. Erratic plates and two layers of 0.25-in.-thick Ascon. Shown
blow counts with depth and between adjacent piles are in Figure 8 are observed hammer performance data for
partly a result of cemented sand seams and layers. pile B-l. The stiffness of the Ascon cushion has' been
found to be fairly constant until the Ascon discs dis-
The soil stratigraphy and curves of estimated soi3 integrate.
resistance to driving are shown in Figure 1 for a site
in the Safaniya Field. A Menck 1800 hammer with a The curves of soil resistance to driving plotted
Bongossi hardwood cushion was used to drive the piles in Figure 7 differ by only 33 percent between 42-ft
through the very dense fine sand layer from 21 to 58- and 64-ft penetration. Consequently, the range in pre-
ft penetration. Shown in Figure 2 are values of ham- dicted blow counts is a narrow band. Blow counts pre-
mer efficiency, cushion stiffness, and cushion coeffi- dicted for pile B-l for the upper bound plugged case
cient of restitution measured during the driving of are roughly two blows per foot less than observed to
pile B-l. A decreasing cushion stiffness with in- 48-ft penetration, as shown in Figure 9. The driving
creasing hammer blows is typical of a wood cushion. records for all six piles presented in Figure 10 show
Use of an average value of stiffness in a wave equation considerable scatter. During the first few feet of
analysis would underestimate blow counts and overesti- redriving, blow counts are higher than predicted due
mate driving stresses when a worn cushion is used. to setup occurring when pile driving was interrupted
to change hammers. Scattered high blow counts at
A comparison of observed and predicted blow counts deeper penetrations are due to the rock fragments,
for pile B-l is presented in Figure 3. Predicted blow seams, and layers in the clay.
467
The field blow counts during final driving of a histories presented in this paper. Although these
three-pile well protector in the Zuluf Field, plotted conclusions are directly applicable to the Arabian
in Figure 11, are constant with depth with little Gulf, they may be used at other locations with some
scatter. Final driving of the 36-in.-diameter piles discretion.
was accomplished with Vulcan 540 hammer and Ascon
cushion. Increased blow counts at the completion of 1. For very dense sand, hard clay, or mixed soil
driving are due to a gypsum stratum encountered 5 ft profiles, a reasonable prediction of field blow counts
higher than indicated by the boring. Good agreement is obtained using the lower and upper bound curves of
is obtained with blow counts computed for the lower soil resistance to driving for a plugged pile.
bound plugged case.
2. When computing the soil resistance to driving
MIXED PROFILE for a plugged pile, end bearing should not be limited
to the frictional resistance developed by the soil
A platform installation in the Abu Safah Field plug.
was selected for our mixed profile. The soil condi-
tions consist of a medium dense carbonate sand from 3. For a sand profile, blow counts predicted
35-ft to 79-ft penetration, a very stiff to hard car- using the static pile capacity or the resistance to
bonate clay from 79^-ft and 117-ft, dense carbonate driving computed for a coring pile will badly under-
silty sand from 117-ft to 131-ft, and a very stiff to estimate the field blow counts.
hard calcareous silty clay from 131-ft to 147-ft. Ex-
cept for the first pile section, piles were driven 4. For a clay profile, the range in blow counts
-wtth-a—Menek 3000-hammer -and- a- Bongossi-hardwood _. - -predicted-assuming- either a-coxing -pile on a_plugged _
cushion. pile may be fairly narrow. When the clay strata con-
tain numerous seams and thin layers of rock, field
Curves of estimated resistance to driving for blow counts will be erratic, and may exceed predic-
both lower and upper bound coring and plugged cases tions for the upper bound plugged case.
are presented in Figure 12. Blow counts were computed
at six penetrations. The coring and plugged cases 5. For a mixed soil profile, blow counts pre-
give similar predictions, as shown in Figures 13 and dicted assuming either a coring pile or a plugged pile
14. may be very similar.
468
American Petroleum Institute, Recommended Practice for
Planning, Designing and Constructing Offshore
Platforms, API RP 2A, 12th Edition, (1981) .
\ X\ \\
Coring - Lower Bound Very Dense
• Coring - Upper Bound Fine Sand
40 V, V \ V Plugged - Lower bound -
10
<L> Plugged - Upper Bound
01
\ \
VI \
•^ ~- \.^- .— '
i f
§ 60 v\ ^ Interbedded
•H \ NN
\^\\ Very Dense
Clayey Fine
Sand and
C
0)
cu
80
''\v 1
''•••AJ* ^
\
" ._
Sandstone
'I
— —^~
"\X \ \ \
Very Dense
\\\ \ I
\ Fine Sand
\
100 \\\ \ \
469
Hammer Efficiency, Cushion Stiffness, Coefficient of
% k/in. Restitution
40 60 80 0.40 , 0.60 Q.80
10
20,000 40,000 60,000
20
3°
50
60
Fig. 2 — Measured hammer performance data—Case 1
20
o
r-i
1- 40
\
60
LEGEND
80 - Static Pile Capacity
••• Coring - Lower Bound
— Coring - Upper Bound
— Plugged - Lower Bound
— Plugged - Upper Bound
100
20
•p
<D
<D
0>
CD
Lower Bound - Plugged'
c 60
O
•H
OJ
C
80
100
to
Very Dense
CD
CQ Fine Sand
c
o 60 _
•H LOWER 82/6"
4J
m BOUND
PLUGGED
(U
c
80 _
100
20
S-t
-o-
O
fO
(D
to
o Very Dense
Fine" Sand
§ 60
•H
JJ
4)
Q.
80
LOWER
BOUND
PLUGGED
74/6"
100
K
i
LEGEND \
Loos>e to
Very Dense
— — Corincj - Upper Ba jnd Claye y Fine
•^•^-—— Plugge;d - Lower B<Dund to M edium
c-
Ocmd
20 ——— nugg<;d - Upper B<jund
)\\/V2 ' r
,t
—_ _
M Heird
(0 Calcaireous
0)
CO Clay with
Gypjsum
Frag nents
0 ^^^
^
CD
•H
60 \r
4J
re
V_?^
^"^> \ ~~^> Poor Quality Gypsum
OJ t
(D \
\x *
Q.
He rd
\>
80 ~ Calca reous"
CJ ay
\ \"
\ N ^^ ir
30
—ttr
I
0>
to
CO
g 5.0
0)
60 \
70
20
(U
o
o
<D
I
60
03
c
o
•H
•P
re
0)
80
LEGEND
120
20
<D
0)
(107, 125, 60, 79)
-(45)
o
o
UPPER BOUND - PLUGGED
(0
o; (105)
CO
o
o>
CO
o 60
•H
(0
LOWER
<u
D. BOUND
PLUGGED
80
100
10
J-)
Very Dense
V
20 Fine to
Medium
~Or Sand
O
ro
<D
CO
O 30
.—I
0)
CO
UPPER BOUND - PLUGGED
O
•H
J-i
CD
50
LOWER
BOUND
PLUGGED Number in parentheses are delay time in hours
60
V •
—
*
£_
K — — — — - .... -
N
50 ' \ v ^s Med ium
-P
(D
03
^ >X\
•••-N*\ Derise
Carbcanate
N
Fine Sand
o m
\
r-i
«t-
75 — \r
fO
^— —• — i
0)
cn \ V%xfX"
\ .
0)
CD
100
\
*v- \"'-N \ X Very Stiff
to hiard
- Carb<jnate -
\\
•H
•P
(D
\\
\ -
::V- \.
Cl ay
\r
Dense
N\ ~~x^\
\L
125 Carbonate . SK
Silty Sand
Very : Stiff to
3 '
V" \s p
Hard Calcareous \
Silty Clay \
150 Silty Sand.) "**«t *^^_
—
LEGEND
ig - Lower Bound
— •— Coririg - Upper Bound
Plugcled - Lower Bound
Plugc ed - Upper Bound
17 S
.p
0)
CD Medium Dense
U.
Carbonate
S-i Fine Sana
O
O
ro 80
0)
in (94,97,13 and 12)Very Stiff
o to Hard
rH
(U Jpper Bound - Coring Carbonate Clay
GO
102
O
•H 120
-P Dense Carbonate
Silty Sand
Very Stiff to
Hard Calcareous
0> Silty Clay
QL Eiltv Sand
160 Lower Boun
Coring
200
Medium Dense
Carbonate
--o Eine Sand
o
80
<D
CO -(94,97,13 and 12) Very Stiff
to Hard
CD
Upper Bound - Coring Carbonate Clay
CD
102
12 (97,90) Dense Carbonate"
°
(0 Silty Sand
P Very Stiftto
-P
CO Hard Calcareous
CD Siltv Clav
Q. Lower :y Sanrl
Bound
160
Coring
200