Professional Documents
Culture Documents
Gbs Design
Gbs Design
Scholars' Mine
International Conference on Case Histories in (1984) - First International Conference on Case
Geotechnical Engineering Histories in Geotechnical Engineering
May 6th
Knut H. Andersen
Norwegian Geotechnical Institute, Oslo, Norway
Recommended Citation
Eide, Ove and Andersen, Knut H., "Guest Lecture – Foundation Engineering for Gravity Structures in the Northern North Sea"
(1984). International Conference on Case Histories in Geotechnical Engineering. 1.
http://scholarsmine.mst.edu/icchge/1icchge/1icchge-theme4/1
This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in International
Conference on Case Histories in Geotechnical Engineering by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright
Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact
scholarsmine@mst.edu.
Guest lecture
Foundation Engineering For Gravity Structures In
The Northern North Sea
Ove Eide
Chief Engineer, Norwegian Geotechnical Institute, Oslo, Norway
Knut H. Andersen
Head, Analysis Group, Norwegian Geotechnical Institute, Oslo, Norway
SYNOPSIS During the past 10 years, 15 gravity structures have been installed in the northern North
Sea. As new gravity structures are being designed for installation on softer soils and at greater
depths, they still pose a great challenge to soil mechanics and foundation engineering. Great
improvements have been made during the 10-year period. This applies to soil investigations, in-situ
measurements, undisturbed sampling, laboratory testing and design analyses. Compared to structures
on land, offshore gravity structures are characterised by large foundation areas, the installation
method, and the cyclic wave loading state. The paper reviews investigation methods, site and soil
conditions, construction principles, instrumentation and installation. The main emphasis , however,
is given to current foundation design practice and experiences from full scale measurements.
INTRODUCTION
11°
&0'
51'
62
~ ~
51'
~
.,,
"'
Fig. 1. Oil and gas fields in the northern F1g. 2. The different continental shelfs around
North Sea. Norway.
1627
Mrd.t.o.l
Expected tot•l ncovenllle reserves
0.21 Quantity
pred•ced by SECTION I-I
1.1.13
. 3. Reserves discovered and prod ucti~n in Fig. 4. The Ekofisk oil s t o rage t a nk •
t he Norwegian sector south of 62 •
1628
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
One reason for utilizing fixed platforms for oil Costs are certainly a major competitive aspect,
and gas production in the North Sea is the great and the first concrete platforms were less
depth to the reservoirs, usually 3000 to 5000 m. expensive than steel jackets. This has,
however, been evened out by improvements in the
By 2 diverting the wells, a very large area (many steel jackets, increased capacity of pile
km ) can be covered from one platform location, driving equipment and crane barges. The cost
as illustrated in Fig. 5. of the concrete structure or the jacket itself
is, however, less than 10% of the total cost of
The structure of a typical North Sea oil well is the platform investment, which may be of the
shown in Fig. 6. order of u.s. $ 2 billion.
The potential advantages of concrete gravity Up to 1983, 13 concrete and 1 steel gravity
platforms compared to traditional steel jackets drilling and production platform have been in-
may be listed as follows: stalled in addition to the Ekofisk tank, as
listed in Table I.
• The structure can be completed near shore in
calm waters and the deck and all fittings Figure 7 shows typical pictures of the different
installed. platforms. All the concrete platforms are
single base structures, whereas the Maureen
• There is a short installation period and steel gravity structure is a tripod (The Oilman,
limited risk during installation. 1983).
• The concrete will gain strength with time, and A major reason for Phillips Petroleum Company
has few corrosion and fatigue problems, and choosing the Tecnomare tripod platform for
consequently there will be less need for Maureen was that it provided an open space in
inspection. the center,and the platform was to be installed
over a pre-installed template. In fact, the
• Conductors and risers are protected in structure was placed within 50 mm of the ideal
concrete shafts. position.
• There is potential oil storage capacity with
small additional cost.
VERTICAL SECTION:
¢__
I Well
...
Seafloor
c J)J
Om
;;;
..
I :. :
SOIL
CROSS-SECTION
I . :.
30"x !"casing
I ...
90 m
I ...
I
.
..
20"x 0.635''casing
~
I "'300m
D Cement
l : -·:-13 3/8"x0.480"casing
-~. ~ (not
7',Q<S3' oosiog -
l~ cemented l
1 U ~ 3000m
1629
1630
14 Con- Stat- Mobil Norway Norway 1981 145 3.7 18,200 3.6 m steel 4 Stiff clay (sand cover 0.2 -
deep fjord B 0.9 m concrete 1.5 m)
15 Teena- Maureen Phillips Scotland U.K 1983 96 1.5 4,350 3.4 m steel Guide Stiff clay (sand cover 2 - 6 m)
mare piles
l6 Con- Stat- Mobil Norway Norway 1984 146 3.9 12,770 3.8 m steel 4 Stiff clay (sand cover 0 - 3 m)
deep fjord C
17 Con- Gull- Statoil Norway Norway 1986 133 3.9 11,000 0.4 m steel 4 3 m moraine material above stiff
deep faks A 0.4 m concrete clay
l8 Con- Gull- Statoil Norway Norway 1988 143 3.0 8,700 1.3 m concrete None Dense sand
deep faks B
59°
•
ss•
\
'I
~\
......
57•
1632
1633
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
SHETLAND In glaciated areas the changing sea level
creates a very complex geological situation.
Firstly, the world water balance is altered
NNSW during glaciated periods, and the frozen wat
held in glaciers on the land surface causes
100 eustatic lowering of the world-wide sea leve
E
It has been estimated that
::r:' the sea-level may have been lowered by 80 -
1--
...,
n. m during glaciations (West, 1968). Secondly
C) 200 weight of the ice, approximately 3000 m thic
central parts of Scandinavia during glaciati
maximum, caused isostatic depression of the
landmasses. During deglaciation and removal
300
the ice load, a slow recovery took place,
leading to upwarping of glaciated areas and
downwarping of the marginal areas.
400
The eustatic variations occur simultaneously'
the increase and decrease of the glaciers,
whereas the isostatic variations are slow, a1
Fig. 12. Schematic distribution of water masses depend on the elastic properties of the eartl
in the secti~n from Shetland to crust.
Norway, N 60 45' (Hackett, 1981).
The actual change in sea level is the net dil
ference between the eustatic and the isostatj
variation. In central glaciated areas the
isostatic upheaval after the glaciation was
larger than the eustatic rise, which resultec
Figure 12 shows the different water masses in a a net rise of land, such as most of the U.K.,
cross-sec~ion from Shetland to Norway at lati-
tudeN60 45' (Hackett, 1981). Norway, Sweden and the northern part of Denmc
The opposite situation is going on in areas
such as southern England, Denmark, northern
• NNSW, Northern North Sea Water on the plateau Germany and the Netherlands. This effect of
to the west of the Norwegian Trench. land uplift and down-warping is still proceed
today, as shown in Fig. 13.
• AIW, Atlantic Inflow Water comes through the
Faeroe - Shetland Channel, flowing
southward along the western slope of the
Norwegian Trench.
• NCW, Norwegian Coastal Water flows along the
Norwegian coast from the Kattegat to the
Barent Sea as a wedge of fresher water.
· NTW, Norwegian Trench Water flows northward
under the Coastal Current, restricted to
the west by the Shelf Edge Current.
• NTBW, Norwegian Trench Bottom Water along the
bottom on the eastern side of the
Norwegian Trench.
The current in the North Sea does not contribute
very much to the environmental loads on a gra-
vity platform, but it may be of great importance
with regard to the scour potential around a
platform. It is also of importance for towing
conditions.
GEOLOGICAL CONDITIONS
When evaluating the foundation conditions in the
North Sea, it is important to understand the
geological history of the area (L0ken, 1976 and
Heiberg et al., 1982). During the Quaternary
Period, i.e. the last 2 million years, the North
Sea area was exposed to major climatic changes.
These resulted in several cycles with drastic
changes in geological conditions, which are
reflected both in the bathymetry and in the Fig. 13. Present uplift and down-warping in
geotechnical properties of the sediments laid northwest Europe. Rate of change in
down during this period. mm • per year.
1634
LEGEND: c=J glacier (?:-:::::::::1 sea water [;::::~-;~:~ fresh water ~ dry Land
luring the last glaciation, at a maximum of As a result of the continuous improvement of the
tbout 20 000 years B.P., the ice sheet caused climate, the front of the melting ice sheets
1rosion and loading of the older sediments both retreated, producing large amounts of sediments
.n the plateau area and on the slopes of the suspended in meltwater rivers, which carried the
lorwegian Trench. Just after the last glacial material out into the Norwegian Trench. Here
:etreat, the plateau area and the channel beaches of gravel and sand were built up at the
ihoulder were dry land. The sea level rose temporary shoreline near the top of the Trench
·apidly, and these areas were exposed to heavy slopes, whereas the finer material of silt and
1rosion, which resulted in soft, loose sediments clay filled up the deeper western parts of the
>eing deposited as a beach on the upper part of Norwegian Trench.
:he slope. Only a very thin cover of sand, less
:han 0.5 m thick, is of recent age. About 11 000 years B.P. 1 the sea, encroaching
from the north, gradually submerged the northern
luring the last glaciation, the northern areas parts of the North Sea plateau. Most of the
>f the North Sea and the Norwegian Trench were Norwegian land area and probably large parts of
:overed by a grounded ice sheet, Fig. 14a. At the Scotland and Shetland areas were still
:his stage the southern areas were dry land with covered by ice. The major banks were small and
.ce-dammed lakes, exposed to permafrost and sur- large islands. The large shallow tidal flats in
:ace erosion by rivers and winds, in a similar the central North Sea areas suffered shoreline
tanner to the arctic areas today. erosion and reworking of the sediments under
conditions similar to those in the German and
.s a result of a gradually changing climate from Dutch coastal areas today.
1trongly arctic to somewhat milder, the
:hickness of the ice sheet decreased at the same During the following 2 000 years all the rest of
:ime as the sea water level rose. This resulted the glacier ice melted away from the land areas,
:irst in buoyancy of the ice within the and the sea encroached further southwards. By
lorwegian Trench, followed by the breaking up of about 8 500 years B.P. the shoreline was pro-
:he floating ice by a calving front. Some fluc- bably about 35 m lower than the present level.
:uations, both in the sea water level and in the A little later the northern part of the North
.ocation of the ice front, most probably Sea and the southern part of the British Channel
>ccurred due to short climatic oscillations. became connected.
'or example, a radio-carbon dating from a vibro-
:ore sample at 1. 5 m depth at a site located The result of this geological history is that
>etween the Statfjord and Gullfaks areas gave an large areas of the North Sea, which are at pre-
tge of 18 860 ± 260 years B. P. to a silty clay sent shallower than 150 m, have been dry land
lUSt below an upper layer of reworked till several times during the glacial periods. At
Rokoengen et al., 1982} •. the times when the glacial ice sheets were
retreating, erosion took place on the banks,
•bout 13 000 years B.P. the Norwegian Trench whereas re-sedimentation occurred in the
1ust have been a wide, open fjord with drifting depressions. This is the reason why erosion
.cebergs, from the calving ice front of the gla- relics from glacial moraines are found as sur-
:iated Norwegian landmass on one side and the face sand and gravel on the banks. Loose sand
lry land of the North Sea plateau, still more or and silt deposits are found along the slopes of
.ess covered with ice sheets, on the British the Norwegian Trench, and in the middle of the
iide (Fig. 14b}. trench more than 50 m of soft, normally con-
solidated clay occurs.
1635
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
The depressions on the shelf plateau are in SITE INVESTIGATIONS
general filled with soft and normally con-
solidated marine clay and silt. Most of the Site investigation for gravity platforms in
remaining areas have a cover of uniform top sand North Sea are discussed by: Bjerrum, 19731 E
over overconsolidated glacial clay. The 19741 McClelland, 19751 Ruiter, 1975: H0eg,
shallow areas are usually more overconsolidated 1982: Kjekstad et al., 1978; Andresen et al.
than the deeper areas. 1979; Schjetne and Brylawski, 19791 and Lunn,
and St.John, 1979. It involves the followin'
Some special features which may affect the foun- aspects:
dation of platforms adversely, should be men-
tioned. • Continuous geophysical profiling
• Bathymetry and sea-bed inspection
Slope instability and submarine slides are well • In situ soil testing
known from Norwegian fjords (Bjerrum, 1971) and • Undisturbed sampling
very large submarine slides have occurred at • Laboratory testing on board
Storegga, off M0re (Bugge et al., 1978). Op to • Laboratory testing on shore
now little data has been available on slope con- • Establishing engineering design parameters
ditions in the North Sea, but a submarine slope
stability study is running at NGI with par- Site investigations are usually carried out
ticular concentration on the Western slope of stages, starting with a preliminary investig
the Norwegian Channel (Karlsrud and Edgers, 1982 tion. The final investigation is carried ou
and Edgers and Karlsrud, 1982). when the platform location is fixed in relat
to the reservoir. In some cases the soil an,
Gasified sediments are found in the North Sea, foundation conditions will also influence th
and in most cases the gas is biogenic methane selection of the final location of the platf
produced within the sediments. Only small traces
of petrogenic gases, migrated upwards from The instruments used for continuous geophysi
deeper gas and oil reservoirs, have been found. arofiling are shown in Fig. 15. In order to
The presence of such shallow gas can be eterm1ne the type of soil material located
recognized in seismic records as acoustic between the different reflectors, the profil
blanking, and it is very important to locate has to be calibrated against borings. Boreh
such zones before drilling and soil sampling are should preferably be located at the crossing
started. points of profiles.
Pockmarks are shallow, more or less circular
depressions found in areas with soft and nor-
mally consolidated clays. The size of the pock-
marks depends on the thickness of the soft clay
layer, in such a way that in areas with a rela- Single group hydrophone cable
tively thin top layer there will be many small
pockmarks, whereas a thicker top layer results
in fewer, but larger pockmarks, up to 15 m deep
and 300 m wide. Size and distribution of pock-
marks in the western part of the Norwegian
Trench is given by Hovland (1981).
50m
The explanation for the pockmarks is not quite 160m
350m (variable)
clear, but they are probably caused by migrated 380m variable)
gases.
NOT TO SCALE I
Iceberg plough marks are found in areas of
morainic material at water depths between 120
and 350 m. The typical size is 3 - 5 m in depth
and 25 - 50 m in width and up to several km in Fig. 15. Analog instrumentation layout.
length, but maximum values of four times as
large have been measured. A special study of
the effects of iceberg plough marks on the At a 2 potential platform site one may cover a
engineering of pipelines and gravity platforms
is being undertaken by the Norwegian Continental 5-km area with a 200 to 1000 m grid of high
Shelf Institute (IKO). power sparker profiling. Detailed informati
close to the platform site is obtained using
Rock boulders of all sizes up to 1.5 min equipment giving less penetration and higher
diameter are found, mostly lying on top of the resolution, such as a multi-electrode sparke
stiff clay with little or no embedding, as found boomer. 2 The actual platform area, of the or
in the Statfjord area (L0ken, 1976). These of 1 km , may be covered with a 100 to 200 m
boulders are believed to have been rafted by ice grid, and the final area where detailed soil
and melted out of floating icebergs from investigations are carried out (250 x 250 m)
Norwegian fjord glaciers at a stage illustrated be surveyed with a grid spacing of approxima
approximately in Fig. 14c. These boulders were 25 m.
removed from the platform sites before the gra-
vity platforms were installed. Simultaneously with the sparker-boomer surve
one can carry out bathymetric mappinf with a
A more serious problem, related to skirt high precision echo-sounder and sea loor
penetration, is the rock boulders embedded at inspection with a side-scan sonar. The rela
some depth in the sediment. Such boulders can he1ghts of a number of points on the sea flo
be recognized in seismic records by their para- can be determined with a differential pressu
bolic reflections.
1636
t
North Sea.
Hydraulic fracturing tests are carried out in
145
~
' order to determine the required conductor depth
for mud circulation to deck level when drilling
wu
for the 20" casing.
Great improvements have been achieved in
undisturbed sampling in the North Sea. To start
with, percussion wire line sampling, as deve-
loped in the Gulf of Mexico, was the most common
practice. Thin wall cylinders with internal
~145.6 diameters of 54 - 75 mm, are driven into the
soil at the bottom of the drillstring by blows
145.7 of a wire line operated hammer. The quality of
0 the samples are usually poor as the driving
causes disturbance.
r
145.7
Push sampling, using the weight of the drill-
string to penetrate the tube,provides samples
of better quality. The method can, however,
only be used in fairly soft material due to the
limited penetration force.
Water depths in metres
The sea-bed jacking units shown in Fig. 17 make
LEGEND: it possible to transmit higher penetration for-
"' Seacalf (CPT! ces to the drill pipe. The sample can thus be
o Wison (CPTJ cut with a constant rate of penetration and with
• Borehole one stroke, even in dense material. Piston
samplers are now available for sampling in soft
materials offshore.
1637
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
A laboratory on board the vessel where samples d i sturbance and the effect of the release of l
can be x-rayed, extruded, classified, tested, h igh in-situ stresses on the soil specimens~
photographed and sealed, serves several pur- 'k
poses such as: When establishing design parameters and pr,e~·
for foundation design analyses, most emphasi$
• enabling continuous updating of the soil pro- placed on the test results from undisturbed
perties in order to modify the sampling or samples (Kjekstad and Lunne, 1979). In situ
boring program. test results give valuable additional infor-
mation about homogeneity and variation in soi.
checking the disturbance of each sample imme- conditions. Substantial efforts have been
diately after recovery. focused on correlating cone penetration
resistance and sleeve friction with strength ,
• obtaining shear strengths and water contents deformation characteristics of sands and cla~
as quickly as possible after sampling. and this is made use of in the evaluation of
design parameters. Lately piezocone test
The practice at NGI has been to extrude stiff results are also utilized.
clay samples as soon as possible to prevent
swelling by sucking water from cuttings or free
water in sand layers. The samples are then
waxed and sealed for transportation to labora-
tories on shore. Soft clay samples are, z
however, kept in the cylinders to avoid unne- WATER UN DRAINE!
Q
cessary handling and disturbance. 1- CONTENT, SHEAR
.....JO... STRENGTH
X-ray inspection of the samples in the tube on ocr
VlU
o/o
kN/m2
board the ships has proven very useful. It 1/)
gives immediate information about the quality of ~ 0 20 40 60 0 200 400
the sample and type of material regarding
layers, gas etc, Fig. 18 .
Onshore laboratory testing. Andresen et al. CLAY
(1979) describe the details of the static and
cyclic testing techniques followed by the 20.5
Norwegian Geotechnical Institute once the
samples have been brought ashore. The specified 10 1---::-.- - -- - ---l 21 .71---+--
laboratory procedures and data evaluation
attempt to account for offshore sampling
o'
UJ
(D 19.7
<(
UJ
1/)
•
FINE
50
SAND
I. · - - - - - - 1 17.7
·;~·~· ~: .I ..
SILTY CLAY
60 ~Jf~
.. · ----1
Ill
~ 21.0
Qj
E 20
21.0
o"
I.LJ
m 21.7
<
UJ
a) b) c) Ill 21.0
==tl~~
Fixed platforms are now being considered for the
Troll field in the middle of the Norwegian
Trench, with a water depth of 340 m and soft,
normally-consolidated clays to great depths
(Fig. 10; Dybwad et al., 1980; Schjetlein 1983). ..· ···'···.- ·---· ''
The soil conditions are illustrated by the bore- (a) Construction of ..... ·:· -'~:.;.M#'f. • ,_.+.... • ••· •• ,• ..!··
hole profile in Fig. 22. Last summer a very caisson base in dry lbl Slipforming of
comprehensive soil investigation was carried out dock caisson walls
at the site, utilizing piezocone, field vane and ~Steel deck
piston sampler for taking undisturbed samples
(Moeyes and Hackley, 1983). For the first time
in the North Sea, X-ray inspection of the
samples in the tube was done on board the ship.
This proved very useful in checking the quality
of the sample.
UNDRAINED SHEAR
\ol ATER CONTENT % STRENGTH kN/ml Fig. 23. Construction sequence for concrete
20 40 60 80
gravity structures (Mo, 1976).
.. ....
0 0 50 100
\
... • 7"". 15.0
•: I •
16.0
10
.. .
,;. -~;ive !compression}
Base section is constructed in dry dock
Floating out to deep water construction site
~· Direct simple shear
• Slipforming caisson and towers
\olp 1>, 17.0 Deck mating with the structure submerged
e 20
Towing to the site and installation
Ei •• .. ;:.
"'
~
V1
~ 30
>--
t' 17.3
19.6
20.0
\%•ssive
!extension} Main factors in the design are:
\~
~ Floating stability
~- Capacity of carrying deck load during towing
"'>-
:I:
fu
0
40
.. 20.5
out
Wave loading
so
'!
.
20.7 \ \\ Foundation
Load cases for structural design
The design of the base sections may vary with
60
. 20.8 \ \ regard to the following features:
•
•
Flat base slab contra spherical domes
Cantilevered slab or not
70
.• 21.0 \ \ •
•
Type of skirts if any
Dowels or not
Typical base sections and skirt geometry are
shown in Figs. 24 and 25.
1640
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
he purpose of the skirt system is to: no skirts. For the Norwegian Trench with its
soft clay conditions, skirts as deep as 20 - 30
Confine any soft top soil layers and bring the m are now being considered.
foundation level down to more competent soils.
The purpose of the dowels are to:
Improve hydraulic conditions at the edges, and
reduce the hazard of scour. • Keep the platform in position prior to skirt
penetration.
Facilitate conditions for grouting the base
area. • Prevent damage on steel skirts at touch down.
• Avoid skidding.
Centerline of Platforms On the Ekofisk tank and the CDP-1 platform
·' tI · •::.;·
u
:j. ·"
skidding during installation has been reported.
(ondeep:
.:.1::'1·' Beryl A
Beryl B and D
Doris:
Frigg Scotland Manifold
Frigg CDP-1
Sea Tank:
Frigg TP-1
Condeep Beryl A
Condeep Brent B Condeep Statf jord A
(ondeep: Condeep Brent 0
Statf jord A
I
I
·+····r .....,.r ··r·····-·r·T""1 Andoc:
Ounlin A
I Condeep:
Frigg TCP-2
Condeep Frigg TCP2
l
Andoc Ounlin A
All the Condeep platforms and the Howard-Doris Fig. 25. Geometry of steel skirts for some
Ninian Central platform are equipped with steel concrete gravity platforms.
skirts, mainly 3 - 4 m deep. The Sea Tank plat-
forms are equipped with wedge-shaped, 2 - 3 m
deep concrete skirts. The Ekofisk tank and the
two other Howard-Doris platforms have virtually
1641
LA], -_Sm
The building site and the towing route for the
first 13 gravity structures are illustrated in
lt.6m LAJ,cJ!l" Fig. 28, and in addition the Statfjord B and tl
Maureen platform are now in position. Of the 1
L.,Td~" structures, eight have been built in Norway,
five in Scotland, one in Holland and one in
Sweden. Two of the structures built in Scotlar
/. Ch•ck of and the one built in Holland were towed to
w~terftow Norway for deck mating.
Of the 15 structures 11 are located in the
British sector and four in the Norwegian sector
By 1988 three more concrete platforms will be
installed in the Norwegian sector.
The towing time to the sites has been of the
order of one week, depending on weather con-
ditions. Installation has to take place in the
summer season from May to September.
1.50
Suctionpressure= __.. _..-
1.40 200kN/m 2 "--y_,-'
_,/""
/
/
/
/
/
/
1.20 /
/
/
/
1.10 /
/
/
/
/.
1.00 ~--------~--------~--------~------~ LEGEND
1643
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
STRUMENTATION Inclinometers, to study long-term tilt
effects. In addition optical levelling on
strumentation is included for two reasons. deck is done twice a year.
e is to obtain correct and safe installation
the platform. The other is to check perfor- • Settlements, usually recorded relative to a
nce during operation (DiBiagio et al., 1976~ reference point at 50 to 60 m depth.
Biagio and H0eg, 1983).
• Pore water pressure in the soil at different
e sensors which have to be placed in the base- levels below the platform base.
ction during construction work in the dry
ck, will be submerged in sea water for two • Accelerations at caisson and deck level.
ars or more before installation of the plat-
rm. This makes special requirements of the Long term horizontal displacement due to pre-
nsors, and as they are not accessible for dominant storm direction.
intenance or replacement, they are usually
plicated. It is also very important that Possible scour of the sea bottom near the plat-
asurements during installation are available form is monitored by a submarine. Reference
such a form that the installation personnel sticks are set up in order to improve these
n make immediate decisions. This makes spe- inspections.
al requirements of the data acquisition
stem. Some of the instruments for the As an example case, the instrumentation of the
stallation phase may also continue to be used most recently installed concrete gravity plat-
the operational phase. form in the North Sea, the Statfjord B Condeep
platform is given in Tables III and IV and
e main control measurements for the illustrated in Fig. 29 (DiBiagio and H0eg,
stallation phase are as follows: 1983).
Position determined by acoustic transponders
or electronic distance measurements.
Draught determined from sea water pressure
measurements near the base.
Base clearance determined by echo sounders and
special mechanical devices.
Axial stresses and bending moments in dowels
determined by strain gauges.
Ballast water level in different cells
controlled by pressure transducers.
Tilt measurement controlled by biaxial
inclinometers.
Water pressure in skirt compartments
controlled by means of differential pressure
transducers.
Contact pressure against spherical domes
measured by earth pressure cells.
Strain in reinforcing steel in spherical domes Strain gauges
giving total loads on domes.
ort term settlements have been measured by
ans of a closed hydraulic system with reser-
ir on the platform and transducer on the sea
oor. During skirt penetration inspections
ong the periphery of the base have been
.rried out by a submarine taking video tapes
·r immediate checking on board the platform.
·r performance measurements during the opera-
onal phase, the following are recorded:
Skirt water pressure, in order to study the Fig. 29. Instrumentation for monitoring the
effect of the drainage system. In addition, long-term performance of Statfjord B
the amount of drainage water is measured. (DiBiagio, 1983).
1643
1644
~ 3000
a::.
I..LI
1-
<( Installation Grouting
3
1-
Vl
2000 Jm Ylfl/Tftt/W///TfoW//J
<(
...J
...J
<(
a:\
u... 1000
0
1-
:X:
l::l
ijj
3
0
Symbol:-& indicates depth when conc:rete skirts contact seabed 0 10 20
TIME IN DAYS
J. 30. Observed skirt penetration resistance
for platforms equipped with steel
skirts. Fig. 31. Time schedule for installation and
grouting of Statfjord B.
The first conductor was installed with the water 2000 4000
level in the drill shaft 20 m below mean sea
level. This was a temporary low water level as
the cell pressures had not yet been regulated to • Measured
their final value. Even with 3.5 m deep s~irts, Calculated assuming
piping took place, and approximately 400 m of ljl = 42• and q,... = 1.7
silt and fine sand was washed into the cell
through the 1" wide annulus around the conduc-
tor, and it all happened in a couple of minutes.
The additional settlement due to this event was
2 em, and regrouting had to be carried out. The
intact instrumentation proved very useful during
this event. .E
;r;'
. .-
0.2
I-
CL
w
Brent B 0
:z:
During grouting of the last compartment a !2
I-
serious leackage took place into the minicell, <
0:
I-
probably from a broken, embed~ed grout pipe. w
:z:
The leakage was 200 m /h at its maximum, w
CL
and the skirt water pressure dropped more than 0.4
40 m before it was regulated back to normal by
giving access to sea water. The leakage was STRESS DISTRIBUTION •
stopped by adding various sorts of fiber
material to the water which was sluiced into the
.
-~ "'(t
~y . .,.;. :..
··::~···
~r-_'11tffV ~
skirt compartment, and grouting was finally
completed.
q,.••
Frigg TP-1 O.Sr r O.Sr
0.6
The seafloor here slopes 0.6°, and the platform
was grounded against the slope. Narrow toleran-
ces were given due to a bridge connection to an
already existing platform. This made it
necessary to pull and rotate the platform with Fig. 32. Soil pressure on domes as a functi
the concrete skirt partly embedded Some sort of penetration depth. Comparison
of "bulldozing" took place (Foss and Warming, measured and calculated values. E
1979). TCP-2 Condeep platform (Kjekstad,
1978).
The capacity to apply eccentric ballasting was
not sufficient for vertical penetration and some
tilt of the platform had to be accepted.
During the gro~ting operation grout of the order Frigg TCP-2
of 600 - 700 m was lost out onto the sea-bed.
This was due to piping taking place during skirt The slope 0of the sea-bed here is the same a~
penetration and insufficient skirt depth. TP-1, 0.6 , corresponding to 1 m difference
elevation over the base diameter. Due to the
cylindrical cells the Condeeps have great ca
Cormorant A and Brent C cities for eccentric ballasting, and the don
at TCP-2 were made especially strong in orde
During grouting the same type of problem as was be able to penetrate. The platform could th
experienced with TP-1 was encountered. A step- be made to penetrate vertically. Actually,
by-step grouting procedure helped to solve the request from E&f, the platform was installed
problem, but on CorTorant A the total volume a slope of 0.1 to the slope of the sea-bed.
injected was 7500 m3 , compared to the theoreti- The maximum eccentricity during penetration
cal value of 4900 m • 25 m. Contact stress as a function of penet
tion depth is shown in Fig. 32 (Kjekstad and
Stub, 1978).
1646
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
mt D Ninian Central
>marine inspection during skirt penetration The Ninian Central Howard-Doris platform was
>wed that some local piping still took place ballasted centrically and allowed to tilt during
a skirt penetration depth of 1.5 - 2.0 m. the skirt penetration phase. When the final
full penetration depth of skirts here was penetration depth of 3.7 m was reached on the
m. one side, penetration was 2 m less on the other
side, corresponding to an inclination 51 '• With
to a local sand layer, high contact pressure further centric ballasting the structure rec-
; developed at one dome, and this determined tified itself, and at final ballasting the
~n ballasting should be stopped and grouting inclination was only 6', corresponding to 0.2 m
~rted. The platform continued to penetrate difference in elevation over the base area.
)ther 10 - 12 em after ballasting was stopped.
~ high contact pressure at the critical dome
)pped off due to punch-through failure in the Statfjord A
:al upper sand layer. The reduction in dome
1tact stresses with increased penetration At Statfjord A a still more pronounced delayed
?th took place during one day as shown in skirt penetration took place after ballasting
1· 33. was terminated than that experienced at Brent D.
The additional penetration after ballasting was
stopped, was of the order of 0.2 - 0.3 m. This
is assumed to be due to relaxation of the skirt
wall friction due to rate effects. This rate
NET CONTACT PRESSURE, kN/~ effect is estimated to have been of the order of
20%.
1000 1500 2000 2500
Statfjord B
Contrary to the other Condeep platforms
Statfjord B has a flat base slab. The reason is
100 that the base area had to be extended in
order to obtain adequate foundation stability.
The cantilevered slab is supported by inclined
struts (Fig. 34) and the basement is used for
storage of sand ballast. Statfjord B is the
200 largest gravity platform installe2 in the North
Sea, with a base afea of 18,200 m and submerged
weight of 3.7 • 10 kN.
Due to the flat base it was important to control
the skirt penetration, and to stop at a certain
depth. On the basis of previous experience this
was handled very efficiently. During the skirt
penetration phase water was sluiced both from
the skirt compartments and from open sea in such
a way that no differential skirt water pressure
developed.
The experience from Brent D and Statfjord A was
that skirt penetration continues after the
stopping of ballasting, and this was not accep-
table for Statfjord B. Instead of waiting for
long term penetration to occur, the principle
used was to unload, as illustrated in Fig. 35.
The most expedient penetration force to regulate
is the skirt water pressure. One meter of dif-
ferenti~ skirt water pressure corresponds to
1.8 • 10 kN penetration force. The penetration
force applied is shown in Fig. 36. lhe maximum
total penetration forfe was 1.8 • 10 kN, which
consiste~ of 1.3 • 10 kN ballast water and
0.5 • 10 kN suction force.
In order to demonstrate the efficiency of
unloading by reducing suction, this principle
was applied 20 em above the preset penetration
depth. During the eight-hour stop, no further
penetration took place. The same procedure was
employed at the final penetration depth (Fig.
NET CONTACT PRESSURE, kN/m2 37).
1647
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
PENETRATION RESISTANCE
,.--- ,.---
I
r-- ~
-- -- -
- - -
-
-
'---- 1-- :X:
)-
0.. Long term
w
0
z
1~6m 0
)-
< Delayed penetrat
cr:
)-
w
zw
0..
ni
u
TIME IN DAYS
PENETRATION RESISTANCE IN MN.
0 2
0 1000 2000
3.0
Ballasting
E 1 stopped and
r.· E suction release
t-
a... ::c
IJ.J t-
Cl a...
LLJ
z 0
Q 2
t- z
~ 0
0:: I-
t- <(
~m---
IJ.J 0::
z I-
IJ.J LLJ
a... 3 zLLJ
t-
0::
a... 4.0 --- --- ---
:>2
V1
SEA FLOOR
1648
The wave load period is typically of the order ACTIVE FLAT INCLINED PASSIVE
of 10 - 20 seconds. This may introduce dyna- f+---------;.!1+--------'"1"--·-·-----------1
mic (inertia) effects. Further, undrained
shear strength of soil may depend on time to
failure. Fig. 38. Principles of slip surface method
(Lauritzsen and Schjetne, 1976).
The soil foundation will consolidate under the
weight of the platform. The amount of con-
solidation prior to the arrival of the design
storm will influence the undrained shear The analysis is performed with the forces from
strength to be used in the analyses. the design wave. Even if the wave loads are of
a dynamic nature, the analysis is usually per-
The cyclic loading from the waves in the formed as a quasi-static analysis. The dynamic
design storm induces cyclic shear stresses in (inertia) effect is taken into account by
the soil which will generate pore pressures in multiplying the static wave forces with dynamic
the soil and reduce the effective normal amplification factors which depend upon the
stresses. The "cyclic strength" of the soil dynamic characteristics of the platform and the
may thus be smaller than the original soil. For current North Sea designs the dynamic
undrained static strength. The static amplification of the forces has been relatively
strength may also be reduced by the cyclic small (of the order of 10%). For platforms in
loading. greater depths of water or on other soil con-
ditions, the dynamic amplification may become
Depending upon the soil conditions, drainage larger.
may occur during the design wave and during
the design storm. This will influence whether The load duration is so short that even for sand
the design wave may be assumed to act under it may be realistic to assume that the soil is
drained or undrained conditions and the effect undrained during the action of the design wave
of cyclic loading on soil strength. (Andersen et al., 1982). Bowever, for sand one
should be careful in relying on high undrained
Smaller storms may also generate pore shear strengths which depend upon large negative
pressures which later drain away. This pore pressures. These may drain away locally.
"precycling/drainage" will influence the soil For clays, it is assumed that undrained con-
behaviour under subsequent cyclic loading. ditions prevail during the design wave.
' -, / pods.
DSS
Triu
I \ DSS
Trin
extension
(6) the structure will find the most efficient
way of distributing the loads among the
pods. This means searching for the load
distribution which gives the highest total
compression failure load.
Fig. 39. Typical elements along a potential The assumptions have been proposed by Lauritzsen
failure surface beneath an offshore (1983) who has also formalized a procedure to
gravity platform. perfom tripod stability analyses. Their vali-
dity is being checked by means of model tests
and finite element analyses.
Analyses of Tripod Platforms
The analyses described in the previous section
are valid for gravity platforms in general and
cover both single base and tripod type plat-
forms. The tripod type platforms, however,
require some additional considerations.
Figure 40 shows schematically a tripod platform
and the forces that it transfers to the soil.
v
!
h
,
/
1650
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
~other factor which has to be considered when Design Storm Loading
alculating the bearing capacity of a tripod
~undation, is interaction through the soil. The platforms have to be designed such that the
he failure surfaces from the three pods may so~l is a~le to withstand the design force after
~erlap, and the bearing capacity may be be1ng subJected to the cyclic loading from the
nfluenced. An example is shown in Fig. 41, other waves in the storm. In the North Sea it
hich shows the foundation beneath a tripod has often been assumed that the design storm
odel test after failure. How far apart the dur~tion is 6 hours and that the design wave
ods must be to avoid interaction depends upon arr1ves at the end of the storm when the effect
he failure mode of the pods. of cyclic loading on the soil strength is most
severe. A typical 6 hour design storm com-
position is shown in Table v. Laboratory tests
onsolidation Effects on Stability have shown that cyclic loading will tend to
generate e~cess pore pressures in the soil, and
hen designing the platform it is assumed the effect1ve normal stresses and the shear
~at the design storm may arrive during the strength of the soil will be reduced (e.g
Lrst autumn after the platform has been Andersen, 1983).
nstalled. If the platform is to be towed out
nd installed safely, this has to be done during TABLE V. Storm compositions for various storm
he summer, before severe storms may be durations (Hansteen, 1981)
xpected. Then open spaces between the base and
he soil will be grouted, and the platform will 3-hr. duration 6-hr. duration 24-hr. duration
e ballasted to its full weight. These opera- No. % of No. % of No. % of
ions should be completed before the design of max. of max. of max.
torm is expected. cycles force cycles force cycles force
1651
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
force). The storm has its maximum intensity generated the first winter, there was no ten-
when the difference between the maximum and the dency to pore pressure generation the second
minimum pore pressures in Fig. 42 are greatest. winter. The main reason is most probably that
the soil strength increased due to the con-
1.56 solidation which took place between the two
1 Max. storm periods. The soil was actually fully co
.1~·/
\ solidated before the second winter period {Fig
I . 78).
t\ j \ ., Mean
""e
1.54
.~\ I. \
ii .I. /fi· .r;\ ., \
I ,... 't/ .J ,, · f'/'\ j\r· ·-· J
i\ .t,
Another example of field observations which
shows that storm loading may also induce exces
pore pressures in sand foundations, is the
'J:
z
·J \L,
. I 1'1
I 1 1V: "-'v'
1\\. . •
,\I'\ oJ('I/ /"' ,., /\
observations on the Ekofisk tank. The Ekofisk
tank is located on a very dense, fine sand.
,,.,
~· 1.52 If• .,J I I
~
\'J~" " 1/
• I I ,, ,-... \.J -
i\ I I I Figures 43a and 43b show pore pressure measure
Bl I 1 I ~ ments in the sand beneath the platform during
VI
L.IJ ,J " 'v major storm 4.5 months after installation. T
IX
Q. significant wave height during this storm was
L.IJ 1.50 m. The maximum wave height was estimated to b
IX 21 m, corresponding to 90% of the 24 m design
0
Q.
wave. Figures 43a and 43b show that the avera
pore pressure increases by 10 - 20 kPa during
this storm. These observations indicate that
1.48 generation of excess pore pressures during a
storm must be considered and taken into accoun
in stability analyses even for structures plac
on very dense sand deposits.
1.46 ~__. _ _...._ __.,_ __.,_.......__ _.__.....__ _.__
5 6 7 8 9 10 11 12
DECEMBER, 1975
1652
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
0
- 0
z
..:.::
40
®
10 0 w' 30
(i)9 u
0::
0
u.. N=14
5 10 ~~
0 _J 20 20 .---"!!-- CYCLIC
<:( 20r-J
20 1- I
:z I
~ 110 kN
LEGEND:
-4th Nov. 13Qg-14~
---16th Nov.
(0 0
N
0::
10
~
5 ,!•20
ho
I
0.3m I ~
kZ:z;z;'22!
0 1m
30 :r:
1l 'lL
650 750 850 950 0 40 80
0
PORE WATER PRESSURE PORE WATER PRESS- 1 2 3 4 5
0
kN/m2 URE I.NCREASE, kN/m2
DURING 6. NOV. STORM HORIZONTAL DISPLACEMENT, mm
iJ_g-.£1~
t~/
leads to reduced effective normal stresses. A
soil element with a cyclic shear stress ampli-
tude of TBcy will thus move towards the failure
envelope as indicated in Fig. 46a. After a cer-
tain number of cycles, N1, the failure line is
reached. The soil element will then experience 0 tc
o~ime
-ra ·r:;:-
large shear strains and be in a state of cyclic time .itcy
failure. The failure mode may take the form
either of large cyclic shear strain amplitudes
or large permanent shear strains, depending on
t:bJLL'Jt~y t:h~Ttcy
the combination of cyclic and average shear
stresses that the soil element has been sub-
jected to.
If the cyclic shear stress amplitude is smaller, 0~ 0~
e.g. rAcy in Fig. 46a, a much smaller pore
pressure will be generated after the same number
of cycles, N1· However, if cycling is con-
tinued, this element will also reach the failure
line unless the cyclic shear stress is very
small. The "cyclic strength" of a soil element I Triaxial tests
is therefore not a material constant, but
depends upon the number of cycles that it is • Simple shear tests
subjected to.
If cycling is stopped before the element has Fig. 47. Example of loading of soil elements
r~ach~d the failure line, the cyclic loading
Wll~ :nfluence the static strength that can be
along a potential failure surface in
the foundation beneath an offshore
mob1~1zed under subsequent monotonic static
load7ng. This is illustrated in Fig. 46b. The gravity platform. Simplified.
stat1c stress path for the element which has
been subjected to cyclic loading will reach the
failure li~e at a lower shear stress than for an
elem~nt wh1ch has not been subjected to previous The data presented in the following are from
cycl1ng. The shape and the slope of the static laboratory t~sts on clay specimens with both Ata
stress paths may also change due to the cycling and the cycllc loading applied under undrained
and ~he reduction in static strength due to ' C?nditions. With sand it will as discussed pre-
cycl1ng may be smaller than the percentage Vlo~sly be more representative to apply at
~eduction in effective stresses. The reduction dra1ned, and it is also necessary to evalu~te
1n static strength will depend upon both the the 7ffect of partial drainage during cyclic
aver~ge shear.stress and the cyclic shear stress load1ng. The following section will first pre-
~pl1tude dur1ng cycling. Like the cyclic sent the results from tests in which the cyclic
~ength, it will also depend upon the number of shear stress is kept constant throughout the
'les. test. Afterwards it will be shown how the cycic
1654
.0
0 iA-" f'h!
~
I
-.:<D
b~fJ.-L~J-.:~y
"ta
1:~3
"ta 1-.:cy
0~ 0 time
1655
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
on so-called "strain contour diagrams" two storms the cyclic loading is determined by
constructed from cyclic tests with constant scaling the load levels for a 6-hour storm in
cyclic shear stress. Examples of contour Table v, such that the maximum cyclic shear
diagrams are shown in Fi~. 50_which is valid for stress corresponds to 100% in the table.
cyclic simple shear loadlng Wlth zero average
shear stress. Separate diagrams have to be The maximum cyclic shear strain for each element
established for other stress paths and other in Fig. 51 occurs during the cycle with the
combinations of cyclic and average shear maximum cyclic shear stress. This maximum
stresses. cyclic shear strain amplitude is plotted as a
function of the maximum cyclic shear stress
amplitude on the right hand graph in Fig. 51.
The right hand graph in Fig. 51 shows that if
the maximum cyclic shear stress in a storm
exceeds a certain critical value, the cyclic
shear strain amplitude becomes very large. This
critical value is the "cyclic strength" of the
clay. This cyclic strength is not a material
constant, but depends upon the storm composition
and the storm duration. It also depends upon
the overconsolidation ratio. Figure 52 shows an
example of the ratio between cyclic strength and
conventional undrained static strength for a
10 100 IDOO 10000
simple shear element. The cyclic strength is
valid for the 6-hour storm in Table V.
NUMBER OF CYCLES
1.0
~
Fig. 50. Countour diagrams for cyclic shear
strain and excess pore pressure.
Simple shear tests with symmetrical 0.8
cyclic loading on Drammen clay. thf
is the undrained static shear strength
"Cfc
Su
0.6 r- ~-----~------_-
for specimens sheared to failure in
approx. 2 hrs. (Andersen et al., 0.4
1980).
0.2 r- -
Figure 51 shows how the "strain accumulation"
procedure is used to determine the development 0.0
of the cyclic shear strain amplitude of a soil 1 4 10 40
element during a 6-hour storm. The assumed OVERCONSOLIDATION RATIO, OCR
cyclic shear stress history is given in Table V.
It is assumed that the small waves arrive first Fig. 52. Cyclic strength normalized with
and that the maximum wave arrives at the end of
the storm when the effect of cyclic loading is respect to the undrained static shear
most pronounced. The graphs on the left in Fig. strength. su is the undrained static
51 show examples of two storms with different shear strength for 2 hours to failure
maximum cyclic shear stresses. For each of the prior to cycling (Andersen et al.,
1982).
%60r-------~----~ Diagrams like the one in Fig. 52 may also be
%60
established for other stress paths and various
tcy !40 ® values of the average shear stress. It should
lkPal t 20 ~~~~;
%ItO ®.~·- be kept in mind that in the general case, the
cyclic strength to be used in a stability analy-
'[cy
sis is equal to the sum of Ta and tcy at
lkPal failure.
dO
0 ):::::*=*=*=:::: t20
The values in Figs. 51 and 52 are evaluated for
a 6-hour storm. As mentioned previously, the
tcy t40 clay foundation may be undrained for a longer
lkPal ! 20 0 '-----1-----' period than a 6-hour storm. The effect of storm
0 %10 t20 duration on the cyclic strength of plastic
0 t=:==::::... Drammen clay is presented in Fig. 53. The
results in Fig. 53 show that it is important to
1 10 100 1000 10000 base the analyses on a correct storm duration
figure.
NUMBER OF CYCLES. N Another procedure for predicting soil behaviour
under varying cyclic shear stress from tests
Fig. 51. Maximum cyclic shear strain amplitude with constant cyclic shear stresses, is to use
as a function of the maximum cyclic the cyclically induced excess pore pressure as a
shear stress amplitude in a 6-hour parameter instead of the cyclic shear strain.
design storm (right hand diagram). For sands where it is important to incorporate
The construction of points A and B are partial drainage, this pore pressure accumula-
shown in the diagrams on the left. tion procedure may be most suitable. For clays
1656
Vl
Vl
60
U.JN
~~
Vl ~ 40
a:
<
....
:E:
"~ 0 V1
-
(
.
.J
.: 0
c
- ....a: 60
:J
-5%
!:
u
t:
~ Ne 40
-
ll SYMBOL OCR ~ ~
0..¥
20
..Jx:
-C:: ....a: 0
.JO 0 1
-'t- 0 -20
:Vl
z:.C::
-10% •
)( 10
4 Q.
-::1
'"'o
:J:E: -60
z:.J. 0 5 10 15
c(
:J:O
,
..JI-
-15°/. SHEAR STRAIN, y (%)
3 10 30
Fig. 54. Results from undrained static simple
shear tests on specimens with and
STORM DURA liON, hours without previous undrained cyclic
loading. Drammen clay with OCR = 4
ig. 53. Effect of storm duration on cyclic (Andersen et al., 1980).
strength of Drammen clay. Based on
constant volume, simple shear tests .. 20 "/.
(Andersen et al., 1982).
I 0 •
tatic Shear Strength Reduced for Effect of t9 0"/o •
0--c;-..
~ ••
ndrained Cyclic Loading z 0
w ••
:eduction in undrained static strength due to ....n::
:yclic loading can also be determined from (/) -20"/o 0
~· i••
.aboratory tests. The laboratory tests are then ()
'irst subjected to undrained cyclic loading. ~
"""
,fterwards they are loaded to failure by
1ndrained monotonic static loading.
:xamples of results from static tests on clay
;pecimens with and without previous undrained
:yclic loading are shown in Fig. 54. Both the
w
1-
(/)
z -40 '/
0
z
0 "'
Symbols
• N < 1000
<t o N > 1000
1ndrained static shear strength and the
1ndrained static shear modulus are reduced by ~ -60"/o
:he cyclic loading. An example of reducti?n in 0 ±1 ±2 :t3 ±4 ±5
1ndrained static shear strength as a functLon of
:yclic shear strain which occurred ~uring cyclic CYCLIC SHEAR STRAIN DURING PREVIOUS CYCLIC LOADING,
loading, and the number of cycles, 1s presented Ycy( "!.)
ln Fig. 55.
~or design purposes, the reduced undrained sta- Fig. 55. Reduction in undrained static shear
tic shear strength can be evaluated by first strength due to undrained cyclic
ietermining the cyclic shear strain ~plitudes loading. Simple shear tests on
in the soil beneath the platform dur1ng the Drammen clay with OCR = 4 (Andersen,
iesign wave at the end of the design storm. 1976).
1657
1658
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
gure 57 shows that for normally consolidated sely linked with the selection of soil strength
ays the wave forces can be increased by 30% values. Figure 58 shows that for OCR= 1, the
fore the material coefficient drops to 1.0. foundation capacity increases with the same per-
r higher ocerconsolidation ratios, however, centage as the strength increase. However, for
ly small increases in wave forces lead to high OCRs the calculations show that the calcu-
ductions in the material coefficient from 1.3 lated foundation capacity is little influenced
1.0. For OCR= 10, for instance, this drop by the rate effect.
'Curs for a 5% increase of the wave forces in
e storm.
Stability Analyses Based on Cyclic Shear
tese results indicate that when making stabi- Strengths
ty analyses based on static strength reduced
> account for the effect of cyclic loading, the For the more recent platforms, stability analy-
trgin of safety against uncertainties in storm ses based on cyclic shear strengths have been
>ading depends dramatically upon the overcon- included in addition to analyses with static
>lidation ratio and this margin of safety is shear strengths (Foss et al., 1978; Andersen et
1desirably low for highly overconsolidated al., 1982). The choice of numerical values for
.ays. This is the case both when working with the material and load coefficients (or for the
1mp safety factors and when working with par- lump safety factor) is more open for discussion
cal safety coefficients. in this case. The material coefficient should
be applied to the cyclic shear strength, and it
will- be shown later, stability analyses based seems reasonable to apply the value of 1.3 as
1 cyclic shear strength will be better suited for the static strength. When deciding upon the
) account for the uncertainties in wave forces. numerical value of the load coefficient,
Lrst, however, the influence of rate effect on however, it must be kept in mind that the load
1e shear strength and on the calculated bearing coefficient in this case is applied to the for-
~pacity will be shown. ces from all the waves in the storm and not only
to the force from the single design wave.
s mentioned before, the static tests are often
rought to failure in about 2 hours with a Figure 59 shows the results of cyclic strength
8nstant rate of deformation. The actual wave analyses compared to analyses with reduced sta-
oading, however, is of a much shorter duration tic strengths according to current practice.
nd in cases where the wave loads are the main The results of the cyclic strength analyses are
riving forces, it would be more appropriate to shown for various values of the product Ym • Yf
un the static tests to failure in 5 to 10 (= SF in the case of horizontal sliding). The
econds. For plastic clays, the undrained sta- diagram shows that in the case with no rate
ic strength increases significantly with effect, and if a material coefficient of 1.3 and
ecreasing time to failure. For lean clays, the a load coefficient of 1.0 are applied, the
ffect is less pronounced, and for sands it is cyclic strength analyses will be governing for
egligible. The effect of rate of loading on overconsolidation ratios higher than 3.5.
he measured undrained soil strength is shown
or some clays in Fig. 58.
1.6
1.1.
0
I
......
(
)
..._ f::__CuRRtN I 1, 4 1--~....----+---
~-..!.. J:>
C>
~:I:
;.:
)-...
1.2 ...f!....::.c,.tc
I"'.Z·~ZRAT£
1-
):I:
u
3 (}_
u~
r--
<1 1.2
~t:
-u
1.0
v CURRENT PRACTICE -- ~~ u
z
3:<1
.J(L r----- -~--- 0
U<!
t::u ~
~ 1.0
" 10 40 :::>
OVERCONSOLIDATION RATIO, OCR fi
w
~ig. 58. Influence of 30% strain rate effect on ~
foundation capacity. H is the hori- 3 0.8
zontal design force that can be w
carried by the foundation (Andersen et 0::
al. , 19 82) •
1659
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
It is believed that the cyclic strength analysis corresponded roughly to a summer storm for the
will lead to a more consistent safety level sand beneath a typical North Sea gravity plat-
independent of the overconsolidation ratio, and form.
it is recommended that cyclic strength analyses
should be included in addition to or instead of This effect of precycling and drainage may have
the static strength analysis in cases where the important practical implications for gravity
wave forces are the main driving forces. The platforms on sand. If the platform is installei
numerical value of the load coefficient (or the early in the summer, it will most likely be sub-
lump safety coefficient, SF) must be decided jected to a small amount of wave loading which
upon by the regulatory authorities. will precycle the sand beneath the platform and,
since drainage is likely to occur fairly
rapidly, make it more resistant to subsequent
Effect of Storms Prior to the Design Storm undrained cyclic loading.
In design it is usually assumed that the design For a sand with a reasonably high permeability,
storm arrives early during the first major storm it is also likely that some drainage occurs
season before any other significant storm during the first, less critical part of the
loading has occurred. In reality, however, the design storm. The beneficial
platform will probably experience smaller storms precycling/drainage effect may thus actually
accompanied by drainage before the design storm occur during the first part of the design
arrives. This precycling/drainage will storm.
influence the soil behaviour under subsequent
undrained storm loading. The foundation The possibility of a beneficial precycling/
engineer must therefore investigate whether this drainage effect ought to be considered in the
may deteriorate the soil and make it more unfa- foundation design of gravity platforms on sand,
vourable if the design storm arrives at a later both for laboratory testing where some pre-
time than usually assumed. cycling and drainage might be applied prior to
the undrained testing, and when evaluating soil
With sands, laboratory tests have shown that properties for the calculations. Neglecting
repeated cyclic loading accompanied by drainage this effect may lead to unneccessary conser-
may reduce the tendency to cyclically induced vatism.
excess pore pressure considerably. Bjerrum
(1973} presented results from laboratory tests With clays, laboratory tests have shown that
on very dense sand which showed that the rise in precycling and drainage may be beneficial for
excess pore pressure per cycle decreased by a normally consolidated clays and lead to smaller
factor of 10 - 100 if the specimens were first pore pressure generation during subsequent
precycled with a low cyclic shear stress and cyclic loading (Fig. 61). For overconsolidated
then allowed to drain (Fig. 60}. The precycling clays, however, laboratory tests have shown that
precycling and drainage may be unfavourable and
20 lead to softening of the clay and greater
generation of pore pressure due to cyclic
10 loading (Fig. 62).
T
Test without preshuring 1)9(79·5)
s -....: 8(8H)
2
6
I
(75·07-7 (82'3)
" u
..2->
...
f b
..
o:'-
.. 0
~
~.! O·S
.o! SN)•-'
!3 (8H)
* u
f... ..,r;-
.. . I(88·?i
J•s (85-2}-
z 0 500
NUMBER OF CYCLES
1000
8.~
.. Q.
0·2 ~ ± 0.3 ,-,.--,---,----,---l-.::_--,--,---,--...,-__:..:..;
.2 (87·8)
~
O·l • Pi (89-7) Ui lr- ~
o:: ~ ± o.2 rr=¥---¥=--b--+-~-===1--===J...==::)>,.,,..-,:~=:::j
~
o-os
Test with pres oaring
I '-
/ <{!...
~~
Vl ± 0.1 1---1---+---+---l----1---1---
.t "':/ Pl (89·6)
u
~ 0.0~--~--~--~--~--~--~--~---L---L--~
002 J/
P2(8%)
Numbers In brackets indicate-
the relative density before
PI (SS·O) consolidation 'in %
001
o-1 0·2 0·3
Fig. 61. Cyclically induced excess pore
00
pressures and cyclic shear strains
Shear stress level: t cylcr~ during consecutive series of undrained
cyclic loading. Drainage between each
Fig. 60. Pore pressure rise per cycle observed seri~s. Simple shear tests with sym-
in undrained simple shear tests with metrlcal, constant cyclic loading on
cyclic loading on fine sand samples normally consolidated Drammen clay
prepared with relative densities of (Andersen et al., 1977}.
80% (Bjerrum, 1973).
1660
1661
1662
----1-
:!:600
:!:500
'
~
~
I
I
I
~
z!;. 1. Mode :.. 2. Mode j!- 3. Mode
I
I 'Ij.
X :!:400 I
-=t
:::E:
I
I I·'I
I 'I
I.
~- ±300 II
I I 'I
0:
0
I I I·
u. I I ·I
::;1 :!:200
I
.I·\
1- ' I
i
:z
0
~ :!'100 ·"'·
0
:I: 0 0.5 FREO.UENCY 1.0 Hz
O 0~------±~2~------~±~4--------±~6~----~~~B~
Fig. 67. Acceleraton spectra for the Brent B
HORIZONTAL DISPLACEMENT, MAX (em) Condeep platform as calculated from
measured acceleratons (Hansteen,
:!:2.0 1979). The two first modes of 1.78
and 1. 72 seconds represent bending in
the two horizontal directions. The
third mode of 1. 19 seconds is torsion
about the vertical axis.
~
1663
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
The normalized secant shear modulus measured in al SHEAR STRESS
laboratory tests at 50% of the undrained 60 1:h!kN/rrtl
strength, is in the range G5o/su = 10 - 25
depending upon type of test and overcon-
solidation ratio. The laboratory tests were
consolidated to the in situ stresses and sub-
jected to the same type of loading (stress
controlled) and the same time to failure as in
situ.
The laboratory tests thus give a soil stiffness
which at 50% of the failure value is 5 - 15
times smaller than the average stiffness back-
calculated from the plate loading tests. This
difference is probably due to the fact that the
major part of the soil beneath the plates is
subjected to a shear stress which is signifi-
cantly smaller than 50% of the shear strength
and that the shear modulus increases strongly
with decreasing shear stress. Simplified finite
element analyses showed that if the shear stress
distribution in the soil, the non-linearity of
the soil modulus and the influence of the stress
path were accounted for, reasonable agreement
between measured and calculated displacements
were achieved. In the case of the plate sub- b) SHEAR STRESS
jected to horizontal and moment loading, the 1:(kN/m21
results showed that it may be important to per-
form three-dimensional analyses. In this case
two-dimensional plane strain analyses overesti- 60 Load cycle no.:
mated the displacements and underestimated the 10 5000
stiffness by a factor of 1.67.
1664
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
hear apparatus causes relatively symmetrical !70
~yclic shear strains. The cyclic strain ampli- ( 1-
ude increases and the secant shear modulus N = 10
ecreases with the number of cycles. This is !60
lue to the pore pressure build-up which is
enerated by the cyclic loading.
t50 t- j: !§
----
'igure 68b shows that if the cyclic shear stress
s not symmetrical, the predominant behaviour Ne .... r--
!40 +
1ay be an increase in the permanent strain with
'-1:'
z ;;:t"'
0
tumber of cycles and a relatively small increase
0/<
.I<:
.n the cyclic shear strain amplitude. !30 r-
•
.....
'igure 68c shows that in the triaxial tests, 0)(,~
:here may be a permanent shear strain develop- !20
if
lent even if the shear stress is symmetrical
1bout zero. This is different from the simple
:hear test results which showed a symmetrical !10
:train response. The reason for the permanent
itrain in this triaxial specimen is that the
Indrained extension shear strength is lower than 0
01-
:he compression strength of the specimen. A
iymmetrical stress will therefore lead to a N = 100
1igher degree of strength mobilization on the !40 0 T(.-.
~xtension side than on the compression side. • 01-
~
0
0
+
~hethree examples in Fig. 68 show that it is
Lmportant to model the type of loading correctly
~ t30 r-
'-1:'
z ..;'(:
: .. +
v
.I<:
1hen trying to determine the stress strain beha- t20
riour or the modulus in situ from laboratory .....
:ests. It also shows that the stress strain
lehaviour is complex and complicated to for- !10
mulate in a general material model.
v
1re included. :!:20
C'
.....
~igure 69 represents a simplified picture of the
:yclic soil behaviour which needs further :!:10
investigation. Fqr instance, the data do not
:ontain results from tests with average shear
stresses in excess of 35% of the undrained shear 0
strength. Such tests would probably plot some- 0 :t 0.5 :t 1.0 :!: 1.5
N"hat less favourably in the diagram. There may Ycyl%l
~lso be soils with properties which do not plot LEGEND:
as favourably as the data in Fig. 69, even at x One-way simple shear
low values of 'a· However, until more infor- + Two-way simple shear
~ation becomes available, it seems reasonable to
assume that the cyclic secant shear stiffness of o Triax •cons = 0.35·Su, one-way
a soil element is governed only by the cyclic 0 Triax •cons = 0.35· 5u, two-way
shear stress amplitude and the number of cycles e Triax •cons = 0, two-way
and that it is independant of both stress path o Triax •cons = 0, one-way
and average shear stress. This simplifies the
soil modelling in the analysis considerably and
means that the shear modulus determined from Fig. 69. Relationship between cyclic shear
simple shear tests can be used for the whole stress and cyclic shear strain ampli-
foundation. tudes after 10, 100 and 1000 cycles.
The plots include triaxial and simple
shear tests with symmetrical and non-
symmetrical undrained cyclic loading.
Valid for Drammen clay with OCR = 4
(Andersen et al., 1978).
1665
1666
1667
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
TABLE VI. Vertical Settlement Components for an measured settlements of buildings on overcon-
Offshore Gravity Platform solidated London and Gault clay has been uti-
lized to supplement settlement calculations
~~=--Settlement Compone~n~t _____________________, based on parameters from oedometer tests. Fr
back-calculation, Butler (1975) found an equi
1a. Initial settlement. lent Young's modulus of E = 130 • su and a
(Shear strains under Poisson's ratio of v = 0.1 to give good corre
undrained conditions tion with measured settlements. The measured
due to application settlements included both settlement componen
of static load.) 1 and 2 (Table VI).
1b. Undrained creep
(shear strains under ~
1.·_ l -~l The vertical settlement due to cyclic loading
(components 4 and 5) is more difficult to cal
undrained conditions late. This is a load situation which seldom
s due to the sustained B llvol = 0 occurs to the same extent on land, and there
t load from the weight no generally accepted calculation methods to
a of the platform. predict these settlement components. In the
t (Continuation of
i component 1a) . qtot =480kN/m 2
c t-·
2. Consolidation settle- Q ~nload =210
~.t
DON CLAY:
''
;:]lQ!ili~
;:-r
-f 6.7m 8.2m
T
Mass concrete raft
..L ._-;.___ _ __
c
during cyclic load-
ing. (Undrained.) ~
0 = llvol 0
TIME IN YEARS
2 5 10 20 so
y 4b. Shear strains caused 500 1
c by cyclically indue-
r-'
--i~~.
1 ed excess pore pres- 1.00
i
c
sures and the corre-
spending reductions lj z
300
in effective stress llu > 0 ""0 f----1
and soil stiffness. 0 llvol = 0 "'(
200 ~
0
1 (Undrained.) .....
0 100
a
d 5. Volumetric strains -0
~ ::::;:::-- r-......
due to dissipation
of the cyclically 20 ......
induced excess pore
pressures. E3 llu+O
E
e 1.0 [\'\.\
llvol > 0 iiE
1- 60 \ ~"" r----- -....o
/
Pier 1
z
.....
:1::
..... 80 _\ \" ~
Analysis procedures .....
1- 100 \" r--...."""' ~ Pier 2
T
the cyclically induced pore pressure {component
5) may be calculated in the same way as conven-
1
174.0m equal to the cyclically induced excess pore
pressure.
The differential settlement associated with
static loading (i.e. settlement components 1, 2
I· 38.0m
and 3) may be evaluated by the same procedures
lltot =220kN/m2 as used to calculate the average vertical
~nload = 85 " • Observation point settlements provided data about lateral
'lnet =13SkN/m2 variation in soil properties and any unequal
PLAN distribution of the vertical load are available.
The differential settlement due to non-
+. m symmetrical cyclic base loads are more difficult
to calculate. Calculation procedures as
STIFF FISSURED CLAY described for calculating settlement due to
36.0m water content w =30-3S% cyclic loading (components 4 and 5, Table VI)
plastic limit,wp ::: 28% should be considered. Differential settlement
=
·--~~~~-_:7S%
~t_ ..sAHOY.JJ•.U =-:_-:_::_
SECTION
due to non-symmetrical loading due to preferred
wind, current and wave directions may be eva-
luated from simplified finite element analyses.
The permanent lateral displacements may also be
evaluated from simplified finite element analy-
TIME IN YEARS ses provided data about non-symmetrical horizon-
tal loads due to preferred wind, current and
02 OS 2 s 10 20 so 100 wave directions are available.
.....
so Settlement Analysis of Tripod Platforms
""'!-~ Tripod platforms are more susceptible to dif-
100
K ferential settlements due to their non-
150
200
!
I Monument
rfinished 1937
' \
early settle-
ment 8mm
1
symmetrical cyclic base loads, and this must be
given special consideration.
In the static load case, consideration also has
to be given to the interaction between the pods
through the soil (Fig. 75). This may cause a
i \\ tendency for some rotation of the pods. If the
platform is rigid, this will induce forces in
250
\ the structure. This effect may be evaluated by
300 : \ finite element analyses.
Initial settleme.
(sand and cia)
Consolidation
100 settlements (sa
I
\
E 200 \
E \
Theoretical
\~
....
V1 '\
'\,
z
lJ.J Measured
::;: 300
lJ.J
....J
Fig. 75. Illustration of interaction between ....
.... t.f)
the pods through the soil for a tripod lJ.J .....
platform during vertical settlement. V1 $!!
N
400
0
.
u - 100 "!. consolidation
(- 10 months)
Field Observations of V~~tical Settlement
Settlement has been measured for most of the
gravity platforms in the North Sea. A summary 500 ~------~--------~------~--------
of some of the settlement records is presented 0 5 10 15
in Fig. 76. These records are all for platforms MONTHS AFTER INSTALLATION
on stiff to hard overconsolidated clays and
dense sands. More details about the settlement
records may be found in Clausen et al. (1975), Fig. 77. Interpretation of the settlement
Andersen and Aas (1979), Eide et al. (1979), observations of the Brent B Condeep
Lunne et al. (1981) and Lunne and Kvalstad platform (from Andersen and Aas,
(1982). For most platforms the settlement 1980).
measurements did not start until some months
after the platform had been installed. The
settlement records have therefore been extrapo- Back-calculation of this settlement gives an
lated back to time equal to zero by means of estimated value of E/su in the range 100 to 1~
)ne-dimensional consolidation theory. The to be used together with v ~ 0.1 to calculate
settlement at the end of consolidation has been the sum of initial and consolidation settlemer
Ln the range of 80 to 230 mm. These numbers and (Lunne et al., 1981). As mentioned, Butler
:he settlement records in Fig. 76 do not include (1975) found E ~ 130 • su for London and Gault
:he initial settlement. Figure 77 shows a more clay. Lunne et al. (1981) also back-calculatE
letailed interpretation of the settlement of a constrained modulus, M ~ k • su, to calculat
:he Condeep Brent B platform. The initial the consolidation settlement and found values
settlement has in this case been estimated by kin the range 190 to 280, with an average of
means of theoretical calculations. 250. They also found that the agreement with
constrained moduli measured from the reloadin~
branch in oedometer tests was reasonably good.
The time required for consolidation is not
MONTHS AFTER PLATFORM INSTALLATION always well defined because the foundation soi
10 15 20 25 30 35 40 45 50 55 60 65 contain layers of both sand, silt and clay whi
consolidate at different rates. However, the
time required for consolidation has been esti-
mated from the time-settlement curves and whe~
available, from pore pressure measurements. P
Beryl 'A' example of settlement and pore pressure record
is given in Fig. 78. The equilibrium pore
i
pressure values are slightly below zero due to
'200
the influence of the underbase drainage system
250
The consolidation time for platforms on soil
consisting mainly of dense sand varies from
almost instantaneous consolidation for the
Ekofisk tank on a 26 m thick upper sand layer
8 to 10 months for platforms with 10 m thick
upper sand layers. For the Condeep Brent B
platform on 45 m thick layer of clay interbedd
g. 76. Summary of measured settlement for 5 with sand layers and with sand beneath 45 m, t
North Sea gravity platforms (excluding consolidation time is approx. 10 months. For
initial settlement). the Statfjord A platform on mainly soil con-
1670
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
....
~ sisting of clay, the observed consolidation time
-- ,---............._,., ... -
,..: 200
:r 0 ~-------- "'\ 300~ is approx. 40 months. The predicted con-
<!) ON solidation times depends upon the assumed
~ 150 0 UE
thickness of the layer and on drainage con-
a 200~~ ditions, but in general consolidation occurs
::g 1000 (!).
faster than predicted (Lunne et al., 1981).
0::
~ so 0
,00~~UJtfj
'----
~~ settlement distribution with depth depends
20 0 strongly on the soil layering.
1671
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
DISTRIBUTION OF LATERAL DISPLACEMENT The cyclic base contact stresses may be calc~
lated by finite element analyses of the type
WITH DEPTH used to calculate cyclic displacements. The
calculation of redistribution of static stree
x (mml y (mml with time, however, is more uncertain, since
-20 0 0 20 there are several effects occurring at the sc
Seabed time. Experience from performance observatic
I~ 1%
~ ~~~ ~~
is therefore very valuable.
.. v
1- 0 ~---~---+--~-
HORIZONTAL PLANE AT THE SEABED
~ • I
~
~-50~-__j/~~--4--4--~-4--4--+---
y (mml
20
Oct. 6, 76 ~ I I
~-100~~~-4--+1 -~-+--~-~-+-~-4--
June 15,76
10
-July5,76
,Feb. 4, 76 xmml
...J
-10 10 20 u
>
u -150 .L::---:l:-::-'=----:-~~-~--=-!:--:-l:--:-1:--l-_,..
50 40 30 20 10 0 10 20 30 40 5
l-10 DISTANCE FROM CENTER (m)
1672
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
rtion of the wave moment. The effect of these PIPING AND EROSION
irts is not fully accounted for in the calcu-
tions. Originally it was anticipated that the Waves and currents may cause erosion and piping
asured cyclic base contact stresses could be of the soil around and beneath a gravity plat-
ed to back-calculate the wave moment form. Such piping and erosion may damage the
perienced by the platform, thus providing a soil foundation and must be carefully considered
eck on the theoretically predicted wave in design.
ment. However, because stress concentrations
ar the periphery and forces on the skirts have Whether erosion of the soil surface around the
t been measured, these measurements are not platform will occur or not depends upon the
fficient to check the theoretical calculations water particle velocity, the soil grading and
the wave moment. the transient hydraulic gradients set up in the
soil by the cyclic wave moment.
e static base contact stress variation for the
ndeep Brent B platform is shown in Fig. 82. The cyclic wave moment causes cyclic pore
e figure compares the measured base contact pressure changes in the soil. These pore
resses on the 19 domes just after grouting pressures vary from point to point in the soil
tween the platform base and the soil with the and cause pore pressure gradients and a tendency
resses measured 2 years later. The high for flow of water in the soil and along the
resses developed against some domes during interface between soil and platform. Beneath
1stallation seem to remain for a long time. the uplift side of the platform,. the gradients
'wever, no tendencies to significant long-term cause a tendency to flow from outside the plat-
,creases which may locally overstress the base form in underneath the skirts. At the
1ve been recorded. The measurements indicate compression side, the flow tendency is in the
,ly modest changes in base contact stresses opposite direction.
th time. The stress changes which have
:curred, have mainly been due to special opera- If the gradients are too large they will also
.ons, such as changes in platform weight lead to piping and erosion along the skirts on
:hanges in deck load and oil storage), the uplift side where there is access to free
·outing, use of the underbase drainage system water from the outside. It is therefore
td installation of conductors through the plat- required that the platform does not lift off the
>rm base. ground at the uplift side when subjected to the
maximum wave moment. If a crack opens up, water
is sucked in and subsequently squeezed out,
leading to erosion and possibly serious con-
4000 sequences for the structure. For a platform
Maximum allowable value
without skirts, a positive contact pressure is
therefore required. Pockets of free water under
the structure should not be allowed to remain if
it is possible for this water to be squeezed out
3000 through the soil and thus lead to erosion in the
Expected : front of the platform. A structure equipped
1674
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
KNOWLEDGMENT Andresen, A., T. Berre, A. Kleven and T. Lunne
(1979), "Procedures used to obtain soil para-
e paper is based on accumulated knowledge and meters for foundation engineering in the North
perience gained at the Norwegian Geotechnical Sea", Marine Geotechnology, Vol. 3, No. 3, pp.
stitute over the years with offshore foun- 201-266, Also publ. in: Norwegian Geotechnical
tion activities related to gravity structures. Institute, Publication, 129.
e authors want to express thanks to their
lleagues at the Institute for valuable assist- Bjerrum, L. {1966), "Secondary settlements of
ce, and specially to K. H0eg and K. Schjetne structures subjected to large variations in
r careful review of the paper. live load", International Symposium on
Rheology and Soil Mechanics, Grenoble 1964,
pp. 460-471, Berlin, Springer, Also publ. in:
FERENCES Norwegian Geotechnical Institute, Publication,
73.
dersen, K.H. (1976), "Behaviour of clay sub-
jected to undrained cyclic loading", Bjerrum, L. {1971), "Subaqueous slope failures
International Conference on the Behaviour of in Norwegian fjords", Norwegian Geotechnical
Off-shore Structures, 1. BOSS'76, Trondheim Institute, Publication, 88, Also publ. in:
1976, Proceedings, Vol. 1, pp. 392-403, Also International Conference on Port and Ocean
publ. in: Norwegian Geotechnical Institute, Engineering under Arctic Conditions, 1,
Publication, 114. Trondheim 1971, Proceedings, Vol. 1, pp.
24-47.
dersen, K.H., S.F. Brown, I. Foss, J.H. Pool
and W.F. Rosenbrand (1977), "Effect of cyclic Bjerrum, L. {1973), "Geotechnical problems
loading on clay behaviour", Conference (on) involved in foundations of structures in the
Design and Construction of Offshore North Sea", Geotechnique, Vol. 23, No. 3, pp.
Structures, Institution of Civil Engineers, 319-358, Also publ. in: Norwegian Geotechnical
London 1976, Proceedings, pp. 75-79, Also Institute, Publication, 100.
publ. in: Norwegian Geotechnical Institute,
Publication, 113. Boon, c., D. Gouvenot, M. Gau and J.P. Geffriaud
{1977), "Stability of gravity-type platforms
dersen, K.H., O.E. Hansteen, K. H0eg and by filling under the raft", Offshore
J.H. Prevost (1978), "Soil deformations due to Technology Conference, 9. Houston, Texas 1977,
cyclic loads on offshore structures", Proceedings, Vol. 4, pp. 39-44.
Numerical Methods in Offshore Engineering, Ed.
by o.c. Zienkiewicz, R.w. Lewis and Bugge, T., R.L. Lien and K. Rokoengen {1978),
K.G. Stagg, Chichester, Wiley, pp. 413-452, "Kartlegging av l0smassene pa kontinental-
Also publ. in: Norwegian Geotechnical sokkelen utenfor M0re og Tr0ndelag; seismisk
Institute, Publication, 120. profilering", 55 p, Continental Shelf
Institute, Norway, Publication, 99.
dersen, K.H. and P.M. Aas (1980), "Foundation
performance", Shell Brent B Instrumentation Butler, F.G. (1975), "Heavily over-consolidated
Project; Seminar, London 1979, Proceedings, clays", Review paper: Session III, Settlement
London, Society for Underwater Technology, pp. of Structures; Conference org. by the British
57-77, Also publ. in: Norwegian Geotechnical Geotechnical Society, Cambridge 1974, London,
Institute, Publication, 137. Pentech Press, pp. 531-578.
dersen, K.H., J.H. Pool, S.F. Brown. and Clausen, C.J.F., E. DiBiagio, J.M. Duncan and
W.F. Rosenbrand (1980), "Cyclic and static K. Andersen (1975), "Observed behaviour of the
laboratory tests on Drammen clay", American Ekofisk oil storage tank foundation", Offshore
Society of Civil Engineers, Proceedings, Vol. Technology Conference, 7. Houston 1975,
106, No. GT 5, pp. 499-529, Also publ. in: Proceedings, Vol. 3, pp. 399-413, Also publ.
Norwegian Geotechnical Institute, in: Norwegian Geotechnical Institute,
Publication, 131. Publication, 108 and in: Journal of Petroleum
Technology, 1976, March, pp. 329-336.
dersen, K.H., s. Lacasse, P.M. Aas and
E. Anden~s (1982), "Review of foundation Cooling, L.F. and R.E. Gibson (1955),
design principles for offshore gravity plat- "Settlement studies on structures in England",
forms", International Conference on the Institution of Civil Engineers, London,
Behaviour of Off-Shore Structures, 3. Conference on the Correlation between
:ambridge, Mass. 1982, Proceedings, Hemisphere Calculated and Observed Stresses and
Publ. Wash. D.C. I McGraw-Hill, London Displacements in Structures, Papers, Vol. 1,
a.o.pl., Vol. 1, pp. 243-261, Also publ. in: pp. 295-3170
1675
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
Derrington, J.A. (1977), "Construction of Eide, o. (1983), "Case studies of planning a
McAlpine, sea tank gravity platforms at Ardyne carrying out site investigations for offsh
Point, Argyll", Conference on Design and structure foundations", International Soci
construction of Offshore Structures, London for Soil Mechanics and Foundation Engineer
1976, Proceedings, pp. 121-130. Sub-Committee on Site Investigation, Manua
Part II.
DiBiagio, E., F. Myrvoll and s. B. Hansen
(1976), "Instrumentation of gravity platforms Foss, I., R. Dahlberg, T. Kvalstad (1978),
for performance observations", International "Design of foundation of gravity structure
Conference on the Behaviour of Off-shore against failure in cyclic loading", Offsho
Structures, 1. BOSS'76, Trondheim 1976, Technology Conference, 10. Houston, Texas
Proceedings, Vol. 1, pp. 516-527, Also publ. 1978, Preprint, Vol. 1, pp. 535-545.
in: Norwegian Geotechnical Institute,
Publication, 114. Foss, I. and J. Warming (1979), "Three gravi
platform foundations", International
DiBiagio, E. and K. H0eg (1983), Conference on the Behaviour of Off-Shore
"Instrumentation and performance observations Structures, 2. BOSS'79, London 1979,
offshore", American Society of Civil Proceedings, Vol. 2, pp. 239-256.
Engineers, Conference on Geotechnical Practice
in Offshore Engineering, Austin, Texas 1983, Gazetas, G. (1983), "Analysis of machine fou
Proceedings, pp. 604-625. dation vibrations: state of ~he art",
International Journal of Soil Dyanmics and
Dybwad, K., E. Aldstedt and R. Lauritzsen Earthquake Engineering, Vol. 2, 1983, No.
(1980), "Feasibility of gravity platforms for pp. 2-42.
300m water depth", 27 p., Offshore North Sea
Technology Conference and Exhibition, Gerwick, B.C. ( 1975), "Design and constructi
Stavanger 1980, Offshore Field Development. for of concrete structures in the North Sea",
the 1980's, paper, T-II/8. Boston Society of Civil Engineers, Journal
Vol. 62, No. 1, pp. 20-24.
Edgers, L. and K. Karlsrud (1982), "Soil flows
generated by submarine slides. Case studies Grouting the Ninian Central Platform's Under
and consequences", International Conference on (1978), Ground Engineering, Vol. 11, No. S
the Behaviour of Off-Shore Structures, 3. pp. 17-24.
BOSS'B2, Cambridge, Mass. 1982, Proceedings,
Vol. 2, pp. 425-437. Hackett, B. (1981), "The Feie-Shetland secti
A Hydrographic Atlas", Joint project "Den
Eide, o. (1974), "Marine soil mechanics; appli- Norske Kyststr0m", University of Bergen,
cations to North Sea offshore structures", 44 Institute of Geophysics.
p., Offshore North Sea Technology Conference
and Exhibition, Stavanger 1974, Proceedings, Hansen, J.B. (1980), "A revised and extendec
Technology Volume, Also publ. in: Norwegian formula for bearing capacity", Geoteknisk
Geotechnical Institute, Publication 103, 20 p. institut, Copenhagen, Bulletin, 28, pp. 5-
Eide, 0., L.G. Larsen and 0. Mo (1976), Hansteen, H. (1983), "Personal communicatior
"Installation of the Shell/Esso Brent B
Condeep production platform", Offshore Hansteen, O.E. (1980), "Dynamic performance'
Technology Conference, B. Houston 1976, Shell Brent B Instrumentation Project,
~roceedings, Vol. 1, pp. 101-114, Also publ. Seminar, London 1979, Proceedings, London,
1n: Norwegian Geotechnical Institute, Society for Underwater Technology, pp. 89-
Publication, 113, and in: Journal of Petroleum Also pub!. in: Norwegian Geotechnical
Technology, Vol. 29, 1977, March, pp. 231-238. Institute, Publication, 137.
Eide, 0., K.H. Andersen and T. Lunne (1979), Hansteen, O.E. (1981), "Equivalent geotechnl
"Observed foundation behaviour of concrete design storm", 74 p., Norwegian Geotechni<
gravity platforms installed in the North Sea Institute, Internal report, 40007-17.
1973-1978", International Conference on the
Behaviour of Off-Shore Structures, 2. BOSS'79, Heiberg, s., T. L0ken and K. Torsethaugen
London 1979, Proceedings, Vol. 2, pp. 435-456, (1982), "Natural conditions influencing pE
Also pub!. in: Norwegian Geotechnical leum recovery and field development in
Institute, Publication, 127. Norwegian waters", Offshore Northern Seas
Conference and Exhibition, Stavanger 1982.
Ei~e, 0., o. Kjekstad and E. Brylawski (1979), Proceedings paper P/4, 84 p •
. Installation of concrete gravity structures
1n the North Sea", Marine Geotechnology, Vol. Hovland, M. (1981), "A classification of po•
3, No. 4! PP 315-368, Also publ. in: Norwegian mark related featuers in the Norwegian
Geotechnlcal Institute, Publication, 129. Trench", 28 p., Continental Shelf Institu1
Norway, Publication, 106.
Eide, o., A. Andresen, R. Jonsrud and E. Anden~s
(1982), "Reduction of pore water pressure Husl id, J. M., o. T. Gudmestad and A. A. -Paul:
beneath concrete gravity platforms" (1982), "Alternate deep water concepts fo~
International Conference on the Beh~viour of Northern North Sea extreme conditions",
Offshore Structures, 3. BOSS'82, cambridge, International Conference on the Behaviour
Mass. 1982,.Proceedings, Vol., 2, pp. 373-382, Off-Shore Structures, 3. BOSS'82, Cambrid•
Also publ. 1n: Norwegian Geotechnical Mass. 1982, Procee<:lings, Vol. 1, pp. 18-4~
Institute, Publication, 143.
1676
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
:?Jeg, K. (1976), "Foundation engineering for Lunne, T. and H.D. St.John (1979), "The use of
fixed offshore sturctures: state of the art", cone penetrometer tests to compute penetration
International Conference on the Behaviour of resistance of steel skirts underneath North
Off-shore Structures, 1. BOSS'76, Trondheim Sea gravity platforms", European Conference on
1976, Proceedings, Vol. 1, pp. 39-69, Also Soil Mechanics and Foundation Engineering, 7.
publ. in: Norwegian Geotechnical Institute, Brighton, England 1977, Proceedings, Vol. 2,
Publication, 114. pp. 233-238, Also publ. in: Norwegian
Geotechnical Institute, Publication, 128.
21eg, K. ( 1982), "Geotechnical issues in
offshore engineering", State-of-the-art Lunne, T., F. Myrvoll and 0. Kjekstad (1981),
report, International Conference on the "Observed settlements of five North Sea gra-
Behaviour of Off-Shore Structures, 3. BOSS'82, vity platforms", Offshore Technology
Cambridge, Mass. 1982, Also publ. in: Norwegian Conference, 13. Houston 1981, Proceedings,
Geotechnical Institute, Publication 144. Vol. 4, pp. 305-317, Also publ. in: Norwegian
Geotechnical Institute, Report, 52410/S-6,
!nbu, N. (1973), "Slope stability Also publ. in: Norwegian Geotechnical
computations", Embankment-dam engineering, Institute, Publication, 139.
Casagrande volume, pp. 47-86, New York, Wiley.
Lunne, T. and T.J. Kvalstad (1982}, "Analysis of
~rlsrud, K. and L. Edgers (1982), "Some aspects full scale measurements on gr~vity platforms:
of submarine slope stability", NATO Workshop Final report", Foundation performance during
on Marine Slides and Other Mass Movements, installation and operation of North Sea
Algarve, Portugal 1980, Proceedings, New York, concrete gravity platforms, Oslo, Norwegian
Plenum, pp. 61-81. Geotechnical Institute and Det Norske Veritas,
92 p.
jekstad, 0., T. Lunne and C.J.F. Clausen
(1978), "Comparison between in situ cone L0ken, T. (1976), "Geology of superficial sedi-
resistance and laboratory strength for over- ments in the northern North Sea",
consolidated North Sea clays", Marine International Conference on the Behaviour of
Geotechnology, Vol. 3, No. 1, pp. 23-36, Also Off-Shore Structures, 1. BOSS'76, Trondheim
publ. in: Norwegian Geotechnical Institute, 1976, Proceedings, Vol. 1, pp. 501-515, Also
Publication, 124. publ. in: Norwegian Geotechnical Institute,
Publication, 114, pp. 45-59.
iekstad, 0. and F. Stub {1978), "Installation
of the Elf TCP-2 Condeep platform at the Frigg Marion, H.A. {1974), "Ekofisk oil storage tank",
field", European Offshore Petroleum Conference Symposium on Ocean Engineering, London 1974,
& Exhibition, London 1978, Proceedings, Vol. Proceedings, pp. 83-94, London, The Royal
1, pp. 121-130, Also publ. in: Norwegian Institution of Naval Architects.
Geotechnical Institute, Publication, 124.
McClelland, B. (1975}, "Trends in marine site
'ekstad, 0. and T. Lunne (1979), "Soil parame- investigations: a perspective", 9 p., Offshore
ters used for design of gravity platforms in Europe 75 Conference, Aberdeen 1975,
the North Sea", International Conference on Proceedings, Paper OE-75220, London, Spearhead
the Behaviour of Off-Shore Structures, 2. Publications.
BOSS'79, London 1979, Proceedings, Vol. 1, pp.
175-192, Also publ. in: Norwegian Geotechnical Meyerhof, G.G. {1953}, "The bearing capacity of
Institute, Publication, 127. foundations under eccentric and inclined
loads", International Conference on Soil
Lcasse, s. and T. Lunne ( 1982), "In situ hori- Mechanics and Foundaton Engineering, 3. zurich
zontal stress from pressurerneter tests", 1953, Proceedings, Vol. 1, pp. 440-445.
Symposium on the Pressuremeter and its Marine
Applications, Paris 1982, Paris, Editions Mizikos, J.P. and P.Y. Hieber {1982), "Synthesis
Technip, pp. 187-208. of results from geotechnical instrumentation
of two different Frigg field gravity struc-
.casse, s. and T. Lunne {1982), "Piezocone tures (North Sea) as recorded from 1978 to
tests in two soft marine clays", Presented at: 1981", International Conference on the
Conference on Updating Subsurface Sampling of Behaviour of Offshore Structures, 3. BOSS'82,
Soils and Rocks and their In-Situ Testing, Cambridge, Mass. 1982, Proceedings, Vol. 1,
Santa Barbara, Cal. 1982. pp. 262-280.
uritzsen, R. and K. Schjetne {1976), Mo, o.
(1976), "Concrete drilling and production
"Stability calculations for offshore gravity platforms~ review of construction, installa-
structures", Offshore Technology Conference, tion and commisioning", 14 p, Offshore North
B. Houston 1976, Proceedings, Vol. 1, pp. Sea Technology Conference and Exhibition,
75-82, Also publ. in: Norwegian Geotechnical Stavanger 1974, Proceedings, technology
Institute, Publication, 113. volume.
uritzsen, R.A. (1983), "Personal Moeyes, B. and M. Hackley (1983}, "Soil investi-
communication". gations in the Troll area", Offshore Northern
Seas: Advanced Projects Conference, Stavanger
1983, Proceedings, T6, 37 p.
1677
1678