Professional Documents
Culture Documents
UNDERWATER TECHNOLOGY
AND OFFSHORE ENGINEERING
Volume 3
ORGANISING COMMITTEE
Volume 3
Published by
Graham & Trotman Ltd.
First published in 1985 by
Graham & Trotman Limited
Sterling House
66 Wilton Road
London SWIV IDE
PART I
GEOLOGICAL AND GEOPHYSICAL SURVEYS
PART III
EVALUATION OF DESIGN PARAMETERS
Summary
C. P. Wroth 307
1
Offshore Positioning
Requirements, Systems and
Quality Control: A Review
J. G. Riemersma, Shell Internationale Petroleum Maatschappij, The
Hague, The Netherlands
1
2 ADVANCES IN UNDERWATER TECHNOLOGY
their maximum range (up to 5 km); only the are mostly non-ambiguous (Syledis has an
Artemis system has a range of about 24 km. ambiguity of 10 km), and display their read-
This, along with the limitation of two lines of ing directly in 'metres'. They are not
position, i.e. no redundancy, are the disad- affected by skywave effects and have a high
vantages in using the system. Furthermore, accuracy (better than 3 m). The disadvan-
the system suffers from loss of signal at cer- tage is that their maximum range is nor-
tain distances due to mUltipath effects which mally limited to one to two times the line of
cause signal cancellation. The area where sight.
these occur can be calculated as they are a Examples of pulse measurement systems
function of the antennae heights. The sys- are Syledis, Maxiran, Trisponder and
tem is often used as a back-up or in dynamic Motorola. The last two systems are limited
positioning of vessels near platforms. to a maximum range normally of the order of
40 km due to their high frequency (9 GHz).
Circular Systems Both systems are often used for calibrating
other radio locationing systems.
This group of positioning systems is the Maxiran, operating in the 450 MHz band,
most common one used in site surveys. The is presently not much in use for site surveys.
mobile transmits a signal to the shore sta- One of the main problems is obtaining an
tions (three or more), and on receipt of the operating licence for this system, as the
return signals the distances to each of the bandwidth is 5.7 MHz and the maximum
shore stations are displayed. Some systems transmitting power is 20 kW. The 450 MHz
allow only the use of one mobile, while lies in a very congested frequency spectrum.
others have time-sharing facilities for several Another problem is the heavy onshore aerial
mobiles. Distances are determined either by tower and the need for a second aerial on
measuring travel-times of pulses or by board the mobile when working with three
phase comparison. ranges, as the mobile aerial is directional as
Phase comparison systems have the dis- opposed to omnidirectional. On the other
advange of ambiguity, equal to half the hand, ranges of over 500 km have been
wavelength of the lowest modulation fre- reached.
quency (lane). This requires a continuous Syledis, operating in the same frequency
check on the lane-se~tings and lane-counts; band of 420-450 MHz, with a bandwidth of
it also requires avoiding situations leading to 2 MHz and a transmitting power of 20 W, is a
loss of signal, causing loss of lanes. Phase more appropriate system for site surveys.
comparison systems, especially in the 2 MHz The maximum range of Syledis is 80 km, but
band, have the advantage of long ranges, up occasionally ranges of up to 150 km have
to 500 km, and have a high accuracy, in the been recorded with good accuracies (5 m).
order of 5 m. They suffer from night effect Syledis uses an omnidirectional antenna on
(skywave) causing very unstable readings board the mobile and has the option to use
and have therefore an increased chance of space-diversity antennae onshore to limit
losing lanes. Argo, Hifix/6 and Hyperfix are signal cancellation. Time-sharing with other
typical examples of phase comparison sys- mobiles using the same shore transponders
tems in the 2 MHz band. Their lane-width is is possible. The system has also an option to
approximately 80 m, allowing for a high use a passive mobile receiver, the SR3,
accuracy - 0.02 of a lane - but making it which uses the principle of 'pseudo' ranges.
more difficult to check lost lanes (80 m Up to now, there has been a reluctance to
ambiguity). Present techniques to avoid use this equipment for site surveys as some
lane losses have improved considerably, but problems with this receiver have not yet
occasionally lane-slips still occur. been solved (van Kuijk, 1984). When these
Pulse measurement systems, including problems have been solved, the use of SR3
the long coded pulse measurement systems, will become a possibility, especially in areas
OFFSHORE POSITIONING REQUIREMENTS, SYSTEMS AND QUALITY CONTROL: A REVIEW 5
where a Syledis chain is intensively used (50-100 kHz), a maximum range of 700 m.
(e.g. the North Sea). There is no difference, however, in the
method of deploying and operating them.
Hyperbolic Systems The majority can be switched on or off and
can be automatically released from the sea-
A hyperbolic system requires, per position- bed. There are three different types of
ing pattern, two shore stations, one master transponders: common or dual-mode, relay
and one slave transponder. All shore sta- and intelligent.
tions are synchronized with the signal The common transponder listens on a
transmitted from the so-called master sta- common frequency and, upon receipt of a
tion. The mobile consists of a passive signal from the acoustic control unit, re-
receiver measuring differences in the arrival transmits the signal, on a selected fre-
time of signals from two stations, either by quency, back to the transducer of the con-
phase comparison or by timing pulse arri- trol unit. Knowing the velocity of sound in
vals. A disadvantage of the hyperbolic sys- seawater, the slant range between trans-
tem is the wide lane expansion when moving ducer and transponder can be calculated
away from the master-slave baseline. The from the travel-time of the signal. Systems
lane-width on the baseline is approx- using this type of transponder are mainly for
imately 80 m for the 2 MHz systems, while the direct positioning of survey vessels or
at the edges of the coverage the lane-width manned submersibles.
can expand to 500 m. All the 2 MHz equip- The procedure, in which the relay
ment mentioned above under 'Circular Sys- transponder is used, is often called the
tems' have a hyperbolic option, as does 'sing-around' method. The control unit
Syledis. transmits a signal on the relay transponder's
frequency, and upon receipt the relay
transponder re-transmits this signal to the
UNDERWATER ACOUSTIC transducer of the control unit and to the
POSITIONING SYSTEMS transponders in the array. The control unit
then receives the direct signal from the
Underwater acoustic systems can be divided relay transponder and the array
into long, medium, short and ultra-short transponder return signals, which have
baseline categories. The purpose of the long travelled via the relay transponder. From
and medium baseline systems is the direct the observed travel-times the distances
positioning of a survey vessel or an under- from the array transponders to the vessel's
water vehicle. The short and ultra-short transducer and the relay transponder can be
baseline systems are used from a survey calculated. The relay transponder is used for
vessel for vehicle and towed-body tracking, remote positioning of underwater vehicles
positioning a guide-base while drilling soil- and other objects.
borings, and for dynamic positioning. The latest type is the intelligent
transponder, which, in addition to its normal
Long and Medium Baseline Systems functions, can also receive, gather, process,
store and transmit data via telemetry.
The difference between medium and long These transponders also often carry other
baseline systems lies in their frequency sensors such as temperature and salinity
band, which affects the maximum range of probes, depth sensors and inclinometers.
the system. The low-frequency transpon- With transponders of this kind it is possible
ders (8-16 kHz) have a maximum range of 5 to measure directly the baseline distances
to 10 km, the medium-frequency transpon- between transponders in an array as well as
ders (24-36 kHz), a maximum range of 1 to 3 measuring the distance to the survey
km, and the high-frequency transponders vessel's transducer (Kelland et al., 1982).
6 ADVANCES IN UNDERWATER TECHNOLOGY
One of the major problems in Examples ar,e the Ferranti ORE and Oceano
underwater acoustic positioning is the systems.
propagation of the signals through
seawater. Due to the difference in Ultra-short Baseline Systems
temperature and salinity, the water column
is layered, causing deflections of the signal Ultra-short baseline systems work with an
path (Snell's law). Another major problem is acoustic transmit/receive array element
noise, caused for instance by propellers and housed in one single transducer unit of 20 to
engines, which results in distortion of the 30 cm width. The position calculation is
acoustic signal. All other sources, such as based on range and direction measure-
pingers operating in the same frequency ments, giving the relative position between
band, can interfere and even cause complete the transducer and the transponder.
disruption of the survey. This is especially The distance is determined by measuring
the case with the intelligent transponders, the round trip of the acoustic pulse.
as interference on a command frequel}Cy can The horizontal and vertical direction is
block the transmission of data. It is determined from differences in pulse
therefore essential that during the array arrival times at several acoustic sensor
calibration the frequency band is monitored elements distributed over the transducer
for unwanted signals. Examples of long and head. The transducer is usually mounted in
medium baseline equipment are the the vessel's hull and can be lowered,
Sonardyne PAN II and Micronav, Racal's through a gate-valve, to such a depth that it
Aqua-fix 4 and Wimpol's Marax systems. is free of propeller and thruster noise. The
system also includes pitch and roll sensors
Short Baseline Systems (Vertical Reference Unit) for the vertical
control, and a gyro for the horizontal
Two transducers are mounted on the orientation. Standard transducers remain in
vessel's hull, usuallyonabaselineoflOto20 m a fixed position whereas tracking trans-
perpendicular to the ship's axis. They ducers, which are automatically directed
transmit acoustic pulses which are received to the transponder by turning the trans-
and returned by a transponder on the ducer head horizontally and vertically, work
target. The distances from the two with a wide beam for search and a narrow
transducers to the transponder are beam for tracking. It is possible to track
determined by measuring the round-trip several transponders simultaneously, e.g.
travel time of the acoustic pulses. Using the up to 9 with the Simrad HPR 209. The
known length of the baseline and its system's accuracy is of the order of 0.5% of
bearing, determined by the vessel's gyro, the observed slant ranges and th~maximum
the position of the target can be computed. range is 1000 to 1500 m. Two frequently
Low-frequency systems (8-16 kHz) have a used systems are the Simrad HPR and the
maximum range of 2 to 3 km, a range Honeywell RS-7.
accuracy of 3 to 5 m and a bearing accuracy
of 2 to 3 degrees. High-frequency systems
(50-100 kHz) have a maximum range of CALIBRATION OF OFFSHORE RADIO
1 km, a range accuracy of 30 to 50 cm and a LOCATIONING SYSTEMS
relative bearing accuracy of 0.5 of a degree.
The systems are able to position several Calibration of a radio locationing system is
transponders simUltaneously, e.g. up to 16 needed to determine the systematic err,Or
with the Oceano system. The short baseline sources inherent in the system and to ch~ck
system has mainly been used with that tlk equipment is functioning properly
submersibles but, with the development of before the start of the survey. The nfctin
new systems, it is now hardly used. error sources are the propagation velocity of
OFFSHORE POSITIONING REQUIREMENTS, SYSTEMS AND QUALITY CONTROL: A REVIEW 7
the signal over seawater, and the zero or often used. A highly accurate Electronic
index error, caused by delays in equipment Distance Measuring (EDM) instrument,
and cabling. The two methods used to assess such as a tellurometer, is used for compari-
these errors are a long and a short baseline son. The delay value is derived from the dif-
calibration respectively. There is a slight ference in distance measured with the radio
difference in calibrating phase comparison locationing system and the EDM equip-
and pulse measurement systems. Differ- ment, taking into account the offset distance
ences also occur between systems producing between the two antennae. At least 20
hyperbolic, circular and linear patterns observations should be taken at regular time
(Riemersma, 1979). In this paper the calib- intervals.
ration methods are divided according to the Pulse measuring systems can be cali-
latter distinction. brated prior to their installation on a shore
baseline, the distance of which is known to
Calibration of Linear Systems an accuracy better than 0.1 of a metre. In
this case it has to be ensured that all equip-
The problem of propagation speed is ment is installed at the shore station and
irrelevant to this system. The position of the on the vessel in the same manner as it was
orientation reference marker should follow calibrated.
the same criteria as for shore stations. The The shore baseline should have a length of
calculation of the reference orientation (RO) not more than 5 km to avoid the effect of
value should be checked and it should be propagation speed errors and should be
very clear whether this value refers to a longer than the minimum distance recom-
True azimuth or to a Grid azimuth. This mended by the manufacturer in order to
value should be entered in the on-board avoid distortion of the signal (although with
computer to avoid mistakes and to ensure some equipment it is possible to use an
that it is logged in the computer. The RO attenuator). At least 20 observations, taken
value in the transponder should be set to at regular time intervals of about 10 sec-
zero when aiming at the orientation refer- onds, should be made for each set of equip-
ence marker. During the calibration, the ment. Experience has taught that it is best
orientation value should be checked against not to touch the setting of the mobile equip-
theodolite observations. During the survey, ment at all during the survey but to enter
the aiming at the orientation reference the delay values in the computer. It will
marker should be checked regularly. provide the user with a record - computer
printout and tape-log - of delays used, and
Calibration of Circular Systems it facilitates easy switching from one pattern
to another.
As mentioned before, circular patterns can The long baseline calibration is identical
be generated by both pulse and phase com- for both the pulse measurement and the
parison systems. In the latter case the sys- phase comparison systems. This calibration
tems have to be installed at the station site can be done either on a known baseline over
and on board the survey vessel before seawater or by crossing the line between
calibration because of the effect of the two shore stations. Where feasible, more
ground conductivity on the signal transmis- than one crossing should be done to ascer-
sion. The short baseline calibration takes tain redundancy in data.
place on board the survey vessel situated a
few kilometres offshore from the station Calibration of Hyperbolic Systems
site. It is essential with 2 MHz systems that
the minimum distance between mobile and The calibration of hyperbolic systems, using
transponder is not below that recommended either pulse measurement or phase com-
by the manufacturer. A distance of 10 km is parison techniques, is carried out after the
8 ADVANCES IN UNDERWATER TECHNOLOGY
system has been installed on the shore sta- Calibration of Long and Medium Baseline
tion sites and on board the survey vessel. (LBL) System
Observations of all position lines are taken
near the shore station sites, taking into con- The purpose of the calibration is to obtain,
sideration the minimum distance as without ambiguity, the co-ordinates of the
specified by the manufacturer. The co- transponders in an array and to determine
ordinates of these calibration positions have possible systematic error sources. For
to be known with an accuracy better than deploying transponders in an array, the fol-
that supplied by the system. Alternatively, lowing criteria are to be applied:
the calibration can be done by crossing
baseline extensions and by checking against (1) The distance between low-frequency
known offshore platforms (transit fixes). transponders should be no more than
The delay values for each pattern and the 5000 m, for medium-frequency trans-
propagation speed over seawater can be cal- ponders no more than 1500 m, and for
culated from the least squares solution of high-frequency transponders no more
the observations. The delay value includes than 700 m. However, the maximum
the error caused by a baseline crossing ranges depend very much on the water
totally or partly over land. At least 20 depth.
(2) The maximum number of transponders
observations should be taken at each posi-
tion and should be logged on a computer. in one array should be no more than 6.
(3) Transponders should be deployed so as
It is essential that during the calibration
the stability of the patterns is closely moni- to avoid impeding their acoustic line of
sight.
tored, for example, pattern values for
Pulse-8 should not vary more than 0.05 milli- The method of calibration depends on the
seconds during any I-second interval. type of transponder used. For intelligent
transponders the 'boxing-in' method, in a
dynamic calibration mode, is used. The sur-
CALIBRATION OF UNDERWATER vey vessel circumnavigates at least two of
ACOUSTIC POSITIONING SYSTEMS the array's transponders in a circle, the
radius of which should be no more than
One of the parameters required in an twice the water depth. At least 100 data sets
underwater acoustic calibration is the speed of acceptable quality, evenly distributed
of sound in seawater. This can be measured around the sailed figure, should be collected.
either directly or indirectly, and should have In addition, a maximum of 20 direct trans-
an accuracy of better than 2 m/s. Sufficient ponder baseline measurements should be
redundancy should be provided so as to made between all transponders, especially
ensure that there is no ambiguity in the between the calibrated transponders.
resulting velocity profile determined by the If non-intelligent transponders are used,
temperature/salinity bridge or seawater the survey vessel should sail around all the
sound velocity meter. The equipment used transponders in a cloverleaf type of figure.
must be properly calibrated against an This method is discussed in detail in
industrial standard high-precision ther- Riemersma (1977). To verify the transpon-
mometer and with a known saline solution. der calibration, the 'boxing-in' method
Calibration of underwater acoustic trans- should be carried out over at least two of the
ponders should always be carried out using a transponders. At least a 100 data sets
high-precision surface positioning system, should be observed and recorded.
which itself should be properly calibrated. All equipment must be interfaced to a
minicomputer to allow real-time calculation,
display and logging of data. Sufficient
redundancy in the positioning data should be
OFFSHORE POSITIONING REQUIREMENTS, SYSTEMS AND QUALITY CONTROL: A REVIEW 9
acquired for the computation of a mathemat- acquired for the computation of the size and
ical and statistical solution in order to derive sign of each of the following error sources:
the size and sign of the velocity error, the
(1) scaling error:
range index error and the offset error. The (a) velocity error
vertical separation between the vessel's (b) range index error
transducer and the transponder on the sea- (c) offset error
bed should be determined from depth sen-
(2) horizontal alignment:
sors in the transponder or from echo- (a) gyro error
sounder readings, together with the fixed
(b) transducer alignment
offsets for the transducer depth and the (3) vertical alignment:
height of the transponder above the seabed. (a) pitch err:or
For each fix the following data should be (b) roll error
printed and logged on cartridge or floppy (c) transducer vertical alignment
disc:
The gyro error is determined in the gyro
• time of fix; calibration. To determine the other errors,
• surface positioning raw ranges (unfil-
two calibration procedures are executed, a
tered);
dynamic and a static calibration.
• antenna northing and easting;
• standard deviation of the single observa- Dynamic Calibration
tion;
• gyro compass reading; Deploy a transponder ensuring that it is
• ranges to the transponders; clear of all structures and pipelines. The
• transducer northing and easting; survey vessel should sail around this trans-
• standard deviation of the single observa- ponder maintaining a constant heading,
tion;
steering either a triangle or a rectangular
• baseline ranges between transponders figure. The range of the transponder should
(intelligent mode).
vary from one to three times the water
Before the start of the survey, the cali- depth. At least 100 data sets of acceptable
brated array should be verified by simul- quality, evenly distributed around the sailed
taneously observing the ranges to the figure, should be collected. All position fixes
array's transponders and the surface posi- should be computed and displayed in real
tioning system's pattern data. The vessel's time to verify the systematic distribution of
position obtained from these data should be fixes around the transponder. For each fix,
compared, taking into account the offset the following data should be printed and
between aerial and transducer. logged on cartridge or floppy disc:
solely a measure of the stability of the may not have been originally possible due to
positioning system and will give an indi- operational pressures; and it provides a
cation on the presence of systematic necessary independent check on the work
errors. It should be noted, however, performed during the survey.
that these errors will not be noticeable if The procedures for this quality control
the vessel stays in one area and some potential pitfalls are set out in
(Riemersma, 1979). detail in McLintock (1984). Further checks
(2) The standard deviation in the position, especially suited to the densely surveyed
with parameters defining the standard grids used for site surveys are comparisons
error ellipse. These values are corre- of water depth at line intersections and the
lated with the configuration of the posi- positioning of either co-ordinated or non-
tion lines and give, in addition to the coordinated seabed features, e.g. wellheads,
performance of the positioning system pipelines, rocks.
and possible systematic errors, a meas- As was discussed for the reporting of
ure of the accuracy of the fix (Houten- shore control data, reporting the positioning
bos, 1981). aspects of a survey is fundamental to its suc-
(3) The residuals in the raw data after a cessful completion. Such reports will always
least squares position calculation. differ according to the type and area of the
survey but they should:
Upon completion of each survey line and (1) allow easy access to the final results of
after specific parts of the survey a statistical the survey;
summary in the form of histograms is (2) give recommendations that could
required. The histograms should cover the improve the execution of similar future
standard deviations in the single observa- jobs;
tions, the standard deviations in the posi- (3) be clearly marked with project title,
tion, the offsets from the planned survey contract number and start and comple-
lines and, where required, the interval bet- tion date of survey.
ween the fixes.
Also they should contain the following
information:
POST·PROCESSING AND REPORTING
(4) summarized statement of the con-
Irrespective of the quality of the position- tracted survey requirements, together
ing data collected and computed in real time with the contractor's assessment of
during the survey, all data should be post- results against objectives;
processed after the completion of the sur- (5) chronological report of major events;
vey. In the case of poor real-time data qual- (6) time summary, with total time allo-
ity, post-processing takes advantage of cated to port or transit time, weather
two-sided smoothing techniques to improve downtime, equipment downtime or
the reliability of the co-ordinates. If the surveying time;
original data are good, they act as a quality (7) list of survey personnel involved in the
control check on the real-time acquisition. survey;
Furthermore, this post-processing quality (8) report of equipment performance stat-
control should not be treated as an auto- ing usage, basic parameters such as
mated batch process. Unlike real-time data, station coordinates, survey accuracy,
it provides information on potential sys- breakdown, and including any remarks
tematic positioning errors affecting the or recommendations;
whole survey; it allows for more detailed (9) dated tabulated summary listings of
investigation into effects which may have results, for example, northing, easting,
degraded real-time data quality and that depth, etc.;
12 ADVANCES IN UNDERWATER TECHNOLOGY
PTSIISHELL
>-
~
SAT . NAV WITNESS POI NT 1 BEHIND STATION MARKER AT SA TNAV 6 I:.:r:j
0" .
I e
6
N
... ... ~~4 <:
19 83 ......
.. , REMARKS : f: A~nRS TAll()~ ,OR '1'R>IN';WCATIO ,I: SATNAV 1 (lEBO) I:.:r:j
4 {I. COORDI NA'I'&S: "GS 72: LAT . 01- 2 9 ' 09.4 56 " S
P.T. KOMARITIM LONG . 102 " 26 ' 29 . 4:5(>" E
~
1 k--D-• • - - - - I NUMe r W OF f' AS~E5 SATEL L IT~ kE'; ULT eASED ON: }~
.....
Co:>
Jl
2
Oahu OTEC Preliminary
Design: Sea-Floor Surveys
A. W. Niedoroda, A. C. Palmer, J. R. Pittman, R. J. Brown and
Associates, J. P. Vandermeulen, NOAA, US Department of Commerce,
and J. F. Campbell, Institute of Geophysics, University of Hawaii, USA
15
16 ADVANCES IN UNDERWATER TECHNOLOGY
. o,~-h
0
1800fl
S . 7 30 1t
20 0
,uo PROFILE AT KAHE POINT ,. -
~ :
~ .. _.-'
J sE~bnAIL
.CAlf: I : 1100
400
'O.~
CUIO
eEL OW
100
C.lll
UIIO
100
., 6000 4000
(uno',
3000
,. . . 101
2000
C. . . IO "
1000
UIIO"
o
c;:oo~--f---D-ET-A-IL-P-RO-F-IL-E-A-T-SH-E-LF--BR-E-AK-+r-.:J~>
=-.-... . ~ -=
. :::=
; -•• ::::!::;:;.. ::
.:, - •• "• "
• -. :::
. ::: :.:1
. .=.::;:-.; :-:;.
aCALE : \ :200 .~:-
J. "'':.4.~
.. or-- - r- -- -- - - -----+I-£;- · I I----------l
1'10 'IJ
_ I .. OOT"- + -_ _---;; .ouLor ... , "AROI ItOC. / HAltO IIIIOCIC,
IAMDY OUTC"O'P ..... 1 -10 n HIIO"' / LIGHT IAIID c o ve,.
. 00t---r------------:7L------------~
"Ia /
uor---t--~
/_--------I
'~:.
,uo /
~
:-
~.o~:-
' I I'O '~OOO 1100 11 0 0 1100 1100
t . . . o· ) teno' ) 'flOO " " ..0 " CUU"
Reports from bottom observations and relatively shallow shelf area. The sub-
depth meausrements made by the research marine escarpment section was to be crossed
submarines Makali'i and Turtle (Coles, with a double steel wall laminated pipe
1982; Offshore Investigations Limited, 1983) section which would be connected to a
were also used. fibre-reinforced plastic (FRP) pipe that was
The cold-water pipe route crosses a nar- to extend to a depth of 670 m. The 1830 m
row (approximately l.5-km) wide shelf long FRP section of the cold-water pipe was
which extends to a water depth of 115 m. to be installed using a surface-tow tech-
Beyond this depth the bottom slope nique. The steel escarpment section was to be
increases rapidly along a near vertical installed as a monolithic piece and anchored
escarpment to a water depth of about 200 m. to the bottom with piles.
Beyond 200 m the bottom slope decreases to Discussions at the beginning of the pre-
an average slope of 8° at a depth of 500 m. liminary design phase centered about
The presence of the near vertical sub- improving our knowledge of the shape and
marine escarpment along the conceptual composition of the sea floor in the project
design project depth profIle strongly influ- area so that an adequate preliminary design
enced the selection of the cold-water pipe could be accomplished. It was also recog-
design and the installation methods. The nized that additional survey data could
conceptual design was based on a three-part sharply alter the definition of the sea floor
CWP system. A concrete pipe would cross the and thus affect the overall design concept.
OAHU OTEC PRELIMINARY DESIGN: SEA-FLOOR SURVEYS 17
100m
Top of Scar
DIVE 423-20 " ........
/'
/
PROFILE FROM BATHYMETRIC MAP~/
150m
/ Depth
/
/'
/'
Ba 18 0 f Sea rp 'f"-r--.-.---.--,--,-.--,.--,---,--J 200m
100m 60m Om
Horizontal Distance
'lbp of Scarp
100m
........
PROFILE FROM BATHYMETRIC MAP \ ........ -
" ....
"'''',,'' 160m
DIVE 422-19
/'
'"
'" Depth
/
/
/
---_ ..... / 200m
;
tional oceanographic sampling equipment to the lines of position (LOP); and the observed
evaluate the distribution and mechanical ranges did not exceed 7000 meters, so that
properties of the bottom sediments. the system was operated well within its
design capability. The system had been cali-
brated over a fixed baseline ashore, and this
FIELD METHODS calibration was checked before and after
survey operations by comparison with sex-
The field survey was conducted in two tant fixes to geodetic control. A check angle
phases. was measured in addition to the basic two
The first phase utilized the NOAA Survey angles to ensure that the correct stations
Ship, the Rainier, to obtain precision bathy- were being observed.
metric soundings on a tight grid over the Digital data consisting of one depth input,
entire project area. This vessel was also two navigation LOPs, ship's heading, and
used as a platform for sub-bottom profiling time were recorded in real time using the
using a 10 in 3 airgun, an 800 J sparker, and a ship's HYDROPLOT system. A sample
3.5 kHz profiler (supplied and operated by interval of approximately 6 seconds was
NORTEC of Seattle, Washington). The selected as suitable for the survey scale and
primary tracklines were to be oriented per- at the ship's speed being used. Depths were
pendicular to the general bottom contours recorded in fathoms and tenths, ranges in
and spaced at 50 m intervals so that a meters. In addition to the digital data
1 : 3600 scale map could be prepared. All recording, the HYDROPLOT system gen-
navigation was based on a Motorola erated a real-time plot so that progress and
MiniRanger III electronic navigation sys- preliminary results could be evaluated. This
tem. The geophysical and bathymetric data also provided steering information to the
were collected along the same tracklines. helmsman to guide him along the intended
Five crosslines were surveyed. survey lines. Seismic data were recorded
The bathymetric data were collected with independently, except that time and event
a Raytheon DSF 6000 dual-beam precision marks were kept synchronized with bathy-
fathometer. This instrument operates on metric data recordings.
frequencies of 24 and 100 kHz. The 100 kHz These operations were, for the most part,
acoustic beam has a beamwidth of 7.5 be- 0
the same as those which the ship normally
tween its half power points and is intended conducts as part of its hydrographic survey
for precise depth measurements in rela- mission. Since this was the first use of the
tively shallow water. The 24 kHz beam is DSF 6000 echo-sounder aboard the Rainier,
much wider (27 in the fore-and-aft direction
0
some problems were encountered in its
and 47 athwartship). It increases the sys-
0
operation, particularly in controlling the
tem's ability to detect shoaling bottom fea- system's gain when the depth changed
tures to the side of the vessel track, and it rapidly. Steering the ship at slow speed (4
extends the system's overall depth range knots) accurately down the rather short and
capability. Normally, the returns from both closely spaced sounding lines in a tidal
beams are displayed on the sounder's cross-current was also a challenge. The slow
graphic record while the return from either speed was necessitated by the seismic sen-
the high- or low-frequency beam is selected sors being towed astern and by the need to
for digitization and computer recording. resolve the very steeply sloping bottom as
The two-range method of navigation was sounding lines were run nearly perpendicu-
used with the MiniRangei III line of sight lar to the bottom contours. Except for some
positioning system. Shore transponders were caution in turning from one line to another,
installed over existing geodetic control sta- no maneuvering interference was experi-
tions along the coast. The selected sites pro- enced from the towed seismic gear. Other
vided excellent intersection angles between vessel traffic was, fortunately, very limited
OAHU OTEC PRELIMINARY DESIGN: SEA-FLOOR SURVEYS 19
in the area, but the very short survey period ger the bathymetric digitizing system, could
involved did have to be timed to the last be maintained to a water depth of approxi-
moment to avoid interfering with routine mately 330 m. This coverage mapped most of
Navy operations in the same area. the steeply sloped portion of the project area.
Several days after the ship's bathymet- However, it did necessitate manual switch-
ric and seismic survey had been completed, ing to the lower frequency acoustic beam in
one of the Rainier's survey launches was used deeper water. When this was done over a
to tow a Klein 100 kHz short-range side-scan sloped bottom, a distortion in the depth
sonar in the area of the steep escarpment. measurements was noted on the records.
Several lines were run generally parallel to This was probably due to the different
the bottom contours to give full bottom beamwidths associated with the two fre-
coverage. quenices. Figure 3 shows the trackline loca-
During the survey it was discovered that tions for the measurements taken aboard
a strong acoustic return, sufficient to trig- the Rainier. The three potential alternative
400
I
600 800
data analysis were used for the final plan- HORIZONT AL DISTANCE
ning of the Kana Keoki cruise. The objec- 24kHz
tives of this cruise were to image the entire
project area sea floor with the SeaMARC II
wide-area side-scan system and obtain grav- ~
1000
900
r 100kHz
/
//'
a: 800 ,::
UJ
...>-
;: 700 7
600
DATA REDUCTION
o 200 400 600 800
HORIZONT AL DISTANCE
Much of the data collected during the pre-
liminary design survey required only stan- Fig. 4 Horizontal distance effects of chang-
dard data reduction procedures. However, ing between the 24 kHz and 100 kHz beams
several unique problems were identified and
solved. These problems included adequate
removal of artificial features caused by
switching between the 100 kHz and 24 kHz
precision fathometer beams, establishing icant peaks, and deeps not previously
the relative relief offeatures seen on the Sea- selected. The editing tape was then merged
MARC II images, and combining data from with the raw data tape and other corrected
the precision bathymetric survey data and information in the HYDROPLOT system
SeaMARC II images with information from aboard the Rainier to generate the required
the research submarine dives. Each of these position and sounding plots. Position compu-
procedures is discussed briefly below. tations were corrected for any range calib-
ration correctors and for the offset between
Bathymetric Data the echo-sounding transducer and
MiniRanger antenna locations on the ship.
Bathymetric data were corrected for the Depth values were corrected for vessel
sound velocity profile based on a sound vel- draft, predicted tide, and sound velocity.
ocity cast taken at the conclusion of the Plots were made in fathoms and in feet on
survey work. These data were also cor- the modified transverse Mercator projection
rected for the small tide variation based on used by the NOAA Charting and Geodetic
predicted tidal elevations at Honolulu Har- Services.
bour, which have been shown to be nearly Figure 4 shows a comparison of the rela-
identical in phase and amplitude to the Kahe tive depth changes associated with switch-
Point project area. ing between the 100 kHz and 24 kHz
Digital depth values were edited manually fathometer beams. These effects were sub-
by scanning the analogue depth-sounding jectively removed while preparing the final
record and picking up missed depths, signif- bathymetric contour map.
OAHU OTEC PRELIMINARY DESIGN: SEA-FLOOR SURVEYS 21
feature in the project area from the several the slope was proportional to the hachure
'illumination' orientations. This, in turn, density. Tapered hachures were used so
made it possible to distinguish changes in that the apparent upslope direction corres-
the acoustic returns due to small bottom ponded to the thicker end of the lines.
slopes from those due to differences in bot- Broad, relatively flat and featureless areas
tom material or texture. The bottom slopes were shaded.
were marked with a slope hachure notation. The results of mapping the SeaMARC II
The length of the slope was equal to the data are shown in Fig. 6. Most of the down-
hachure length. The apparent steepness of slope trending linear features which appear
in Fig. 5 occur as long slope facets oriented images. Careful examination of many of the
across the large-scale bottom slopes. linear features which appear in Figs 5 and 6
Most of these small linear slope facets occur show that they are long, low and positive
in pairs and many define long, narrow sea- relief features, not negative ones. These low
floor valleys whose axes parallel the sea-floor ridges are quite different from
downslope direction of the larger-scale anything reported in similar environments.
bathymetry. These features are particularly Unlike the narrow sea-floor valleys, these
noticeable in the northern and southern por- ridges seldom branch. However, they are
tions of the deep offshore area (see Fig. 6). discontinuous. The best terrestrial analogue
In some places the narrow sea-floor val- to them appears to be the natural levees
leys branch so as to resemble a dendritic which occur along alluvial rivers.
river pattern. In all cases these long, nar- The hachure-pattern mapping of the fine-
row valleys are discontinuous. That is, scale relief shown by the SeaMARC II
unlike most terrestrial stream valleys, these images does not permit evaluation of the
eventually peter out and end. vertical scale of the features. Examination
Even more striking features were of the fathograms showed that these fea-
detected when mapping the SeaMARC II tures were too subtle to show up on the 24
kHz or 100 kHz beams. Examination of the
records of observations made during recon-
naissance dives of the Research Submarine
Turtle (N oda, 1983) showed that small
sedimentary scarps were reported at the
exact points where the dive-track crossed
some of the long, low and linear sea-floor
ridges. Figure 7 shows examples. The relief
of most of the linear features which show on
the SeaMARC II images, but not on the
fathograms, is taken to be in the range of 1
to 4 m.
Bathymetry
smooth slope gradients. Because the slopes to the low, long ridges, and their associa-
along these routes are not too steep it has tion with these features suggests that they
been possible to alter substantially the con- have an erosional origin.
cept for installing the CWP and, in turn, The origin of the prevalent small-scale
change its design. The steel composite pipe relief features is not immediately apparent.
section which was thought necessary for It has been postulated that rapid sediment
spanning the escarpment zone has been transport events have occurred within the
eliminated. It is now envisaged that the por- project area since November 1982 when a
tion of the CWP running from the top of the series of current meter moorings in the deep
escarpment to its seaward end will consist of offshore area were episodically disrupted
fibre-reinforced plastic (FRP). It will be and some were moved rapidly downslope by
installed using a bottom-tow method. a hurricane (Noda 1983). Dengler (1984a,b)
The shape of the sea floor in the deep off- used the pressure records from the current
shore zone has been well defined by the pre- meters which were moved and subsequently
liminary design survey. The most significant recovered to study the temporal pattern of
features include broad undulations of the the disrupting events. He concluded that
bottom with axes parallel to the area-wide relatively low speed turbidity currents,
slopes. One such undulation provides a gen- triggered during the near passage of Hur-
tle, subtle high extending offshore along the ricane I wa, were the most likely cause of the
centre of the project area. The survey has episodic mooring movements. Other proces-
also located at least two mound-like features ses such as submarine debris flows caused
with 6 to 20 m of relief (one at a water depth by breaking internal waves have been con-
of 630 m and the other at 650 m), whose sidered. However, the geometry of the
origins are unknown. The morphology could small-scale sea-floor relief and the relation-
indicate that they formed as patch reefs in ship between the probable path of the tur-
much shallower water. Bottom photographs bidity currents and current meter move-
taken by the Turtle of a similar feature ments appear to support the mechanisms
located north of the OTEC survey area more proposed by Dengler.
clearly suggest that these features are The elongated small-scale relief features
drowned patch reefs. Gregory and Kroenke are closely aligned with the downslope
(1982) found similar features at similar direction of the next larger scale of bathy-
depths on a broad terrace south of Oahu. metric features. This further implies that
Drowned reefs have also been discovered off they originated through gravity-driven sed-
the Island of Hawaii (Campbell, 1984; Moore iment transport processes. Furthermore, it
and Fornari, 1984). permits the selection of a CWP route which
The greatly improved resolution of the minimizes the chance of impact by turbidity
bathymetric maps which resulted from com- currents and the angle of collision should
bining the high-resolution depth measure- such an impact occur.
ments with the detailed image resulting
from the SeaMARC II has focused atten- Submarine Observations
tion on the small-scale relief in the project
area. It is recognized that, although some of The use of submarine observations to esti-
the elongated features seen in the Sea- mate the relative relief of the elongated
MARC II images result from differences in small-scale features has already been dis-
bottom composition and texture, most of cussed. The submarine observations also
these features represent low-relief valleys recorded occasional coralline limestone cob-
or ridges. The submarine observations have bles in the deep offshore zone and numerous
shown that some ofthese features have near- large limestone blocks in a submarine talus
vertical relief from 2 to 4 m high. The orien- area at the base of the escarpment zone
tation of these low sediment scarps, parallel (N oda, 1983).
><
C,!)
o
....:I
o
Z
:I::
u
J;il
E-<
~
J;il
~
~
J;il
Ci
Z
~
.....
Z
...... :
U1
o
...
J;il
U a:
III
Z
;: ~ ..
-'. -' ~
";"~ ~ ...::.- :.
Ci :.:::::.-- ~
<
"...,
END
"000 .0000 ...., 1000 7000 ...., 6000
HORIZONT Al OISTANCE FROM CENTER OF lecs, (It)
- :1000 2000 .000 .000
LEGEND: ~ MEDIUM SHELL HASH
~
f:·;::·:.:] SAND . '~ '.
~
".:.,\" DENSE SHELL HASH
• SilT
~ CARBON"TE ROCK
I:-·.-·; :j LIGHT SHELL HASH ~%1 AND
INTERBEDDED SAND.
SilT. SilTY CLAY
LIMESTONE
Fig. 8 Shelf composite
co
C\I
OAHU OTEC PRELIMINARY DESIGN: SEA-FLOOR SURVEYS 27
1I
' 0'
I
Ii
·l
r..,'t ._-, ...
I
....~ II
O'?- I
,. t 1
I
KTF-D [SIB 1
1
I NBR ' I
rcftJU-SSA
I
___ J.
I
SHAP.~I
) 0'
I
-, I
~ ':.' .
', . Q
0NNAP
.
I -CV1
J'" SNAP I
---1--612 -~
".
O -GAP
.'
SHAP Soulhern HolI~ros Abyssal Pia", NNAP Norlhern Nores Abyssal Plom MADCAP ModMO Cap. Vorde AOyssol Pia",
SBR So<... thern Ber_da Rise KTF /("'9S Traugh Flank CV1 Cope Verde I
NBR Norlhern Ber_do Rise IB 1 Ibe"o AOyssal Plom CV2 COpe Verde}
SSAP Southern 50hm Abyssal Plain MAP "'ade"a Abyssal Plam/OM£} CV3 Cop. Verde]
SNAP Southern Nares Abyssal PIa", GMW Great ",.Ieor West GAP GamolO A llyssal Plam
Fig. 1 Areas in the Atlantic Ocean studied by the Seabed Working Group. The Cape Verde sites 1
and 3 are part of the Cape Verde Abyssal Plain
ATLANTIC ABYSSAL PLAINS AND THE ENGINEERING IMPLICATIONS 31
~ ~ ~
JJNIr---~----~----~~~--'-----r----.-----r----.----i~L----r----'-----r----.----'-----~~----'-----r----.-----r----r----1~
IV
I )IN
31N
liN 31
Fig.4 The sediment surface in the central part of the Madeira Abyssal Plain with large animal
tracks
with maxima of more than 1750 mover faults with a small vertical offset, especially
fracture zones. Sub-bottom reflectors are in the central and central-northern areas of
continuous in the western part of the plain the plain.
(west of 24 OW). To the east a lack of con- The faults have no surface expression but
tinuity and additional reflectors complicates some remain visible close to the surface.
seismostratigraphic interpretations. The On 3.5 kHz records the offset generally
sedimentary section can be subdivided into increases with depth. Some continue down-
three general intervals: The uppermost 120 m ward in the lower sedimentary units as
interpreted as alternating turbiditic and observed on watergun records. A fully con-
pelagic sediments; an intermediate interval vincing explanation of the origin of these
of 200 to 400 m of inferred pelagic with some faults has not yet been presented (Duin et
turbiditic sediments; a lower interval com- al., 1984).
prising very strong but less continuous
reflectors (volcanoclastics? limestones?) Stratigraphy
with varying thickness and a basal pelagic
drape. Penetration on the 3.5 kHz records Some 60 piston cores up to 22 m in length,
varies across the abyssal plain, depending obtained in 1980-1984 during various
on sediment texture. In predominantly cruises, consist primarily of correlatable
clayey sediments, penetration generally layers of turbiditic marly calcareous ooze
exceeds 80 m. However, in areas with (0.2-5 m) separated by thin but continuous
near-surface sands penetration is much less. beds (c. 0.1 m) of hemipelagic ooze to clay
Slight undulations in the shallow, parallel- (Fig. 5). Pelagic sediments have accumu-
bedded sub-bottom generally reflect the lated around the carbonate compensation
presence of basement culminations deeper depth (CCD), resulting in highly variable
in the section. Seismic and 3.5 kHz data CaC0 3 contents. Fine-grained turbidites
have revealed the presence of discontinuous typically have a 45-55% carbonate content,
34 ADVANCES IN UNDERWATER TECHNOLOGY
82 PCM 29
e
,~
lc:. a 20 40 50 80 100
Water Content (dry) %
120 140 150 180 200
a
0,50 1,00 1.50 200
Densi ty (Mg/m 3 )
_ =bulk density
___ = dry dMS ily
2
-- .-- ~ wale" conten'
Fig. 6 Isopach pattern of Madeira Abyssal Plain turbidite III deposited between 73 000 and
60000 years BP. The map illustrates the gradually increasing thickness of an individual fine-grained
turbidite to the west, which is a characteristic of all major MAP turbidites. (From Kuijpers et al.,
1984)
dites, 68% for sandy and silty turbidites and with a motor vane and a fall cone device
71 % and 77% for pelagic oozes and clays. were made in core 82PCM29 perpendicular
Specific gravity varies between 2.46 and to bedding planes at intervals of 10 cm. The
2.73 g/cm 3 with an average of 2.64 g/cm 3• fall cone measurements included one obser-
Water content (dry) is about 110% by vation made in the geometric middle of the
weight for fine-grained turbidites at depth, subs ample with nine other measurements
a maximum of 82% for sandy and silty turbi- that were averaged afterwards. Shear
dites and around 95% and 129% for pelagic strength plotted against burial depth gives
oozes and clays respectively. Wet bulk curves in which fluctuations measured with
density averages 1.43 g/cm 3 for fine-grained the two different techniques correspond
turbidites at depth, 1.58 g/cm 3 for sandy and remarkably well (Fig. 8).
silty turbidites and 1.5 to 1.4 g/cm 3 for The lowest shear strength with values
pelagic ooze and clays (Fig. 7). below 2 kPa in core 82PCM29 was measured
Permeability measurements on core sam- at the base of the water-rich surficial sec-
ples indicate values averaging 6 x 10- 5 cm/s tion. Distinct shear-strength peaks with
for fine-grained turbidites, values up to 10- 3 motor-vane values exceeding 25 kPa can be
cm/s for sandy and silty turbidites, 10- 6 cm/s related to small intervals with pelagic sedi-
for pelagic ooze and 5 x 10- 7 cm/s for ment. Turbiditic intervals at depth reach
pelagic clay. Evidence for open burrows shear strengths of 5-14 kPa. Sensitivities in
found in a number of cores even at greater turbiditic sediments average around 4.
depth (12 m) might markedly increase bulk Temperature gradients of a few heat flow
permeabilities. stations appear linear and suggest normal
Shear strength measurements performed heat flow fluxes (Noel, 1984).
36 ADVANCES IN UNDERWATER TECHNOLOGY
!(l1Q
tf
--~ eSZ8
~
~
'-~
n
**
~
~
~ \!!
'-
V;
'-
!jl-
...
Vi ~
Sl
'"
<0
eSi8
0>
N
~
U
~ C+---r--.--~--.---~-.--~--~~
a:> w u! '1ldllO C
Fig. 7 Madeira Abyssal Plain: bulk density, dry density and water content (dry) of core
82PCM29. (From Kuijpers et al., 1984)
82 pes 20
8
~ turbidite of which Isopachs
are shown 'J£ll'
•
O~ CJ •
$~
0 I
32N 10 D
• JZN
0, • /
I< /
/
D
D • 0
• cs/\o
. vD JIN
~
~~--~--~--~--~--~~--~--~--~--~----~-~~~N
'"
;to ;t
::::
D
•
Abyssal hills and other elevated areas
not covered by turbidites 13.5 kHz dala} D Area of turbiditic sedimentation
Turbiditem not identified
Fig.8 Madeira Abyssal Plain: shear strength of core 82PVS29 determined with motor vane (left)
and fall cone (right). (From Kuijpers et al., 1984)
38 ADVANCES IN UNDERWATER TECHNOLOGY
• Piston eor.
Fig. 9 Nares Abyssal Plain: bathymetry and Dutch sample stations. More than a dozen Canadian
cores (not indicated) are concentrated between cores 84PCM36 and 84PCM19
The water depth in the area is generally 0.1-0.2 m deep and 0.2-0.4 m wide, with
between 5800 and 5900 m (Fig. 9). Apart spacings in excess of 0.8 m. Furrows are
from a few scattered abyssal hills, most of about perpendicular to ripple crests. The
the sea floor is almost flat with an overall mud clouds stirred up by the camera were
low-angle (0.01°) dip to north-east. South of moving in the down stream direction as indi-
22°30'N the seabed rises towards the Grea- cated by the ripple asymmetry. Rather
ter Antilles Outer Ridge, a broad sedimen- strong near-bottom currents appear not to
tary swell north of the Puerto Rico Trench. be uncommon on the west side of the West-
ern North Atlantic Ocean.
Sea-floor Imagery
Sediment Thickness and Structure
Camera stations from depths between 5850
and 5900 m show biogenic trails and mounds The sedimentary cover over an Upper Cre-
without any evidence of recent current taceous basement generally ranges from 300
activity. The seabed at a depth between to 700 m with values of up to 1000 m in
5700 and 5750 m near the southern margin of basement valleys and buried fracture zones.
the plain appears to be largely free of The lowermost transparent unit consists of a
macro-biogenic structures and shows vari- pelagic clay drape. This unit is at the seabed
ous types of current-induced bedforms (Kui- on small abyssal hills. On top is a unit with
jpers, 1985; Kuijpers and Duin, 1985). These acoustically laminated sediments that may
bedforms are asymmetric ripples with a consist of cherts, limestones, silts and clays
wavelength of about 0.15 m and furrows (Tucholke and Ewing, 1974). An overlying
ATLANTIC ABYSSAL PLAINS AND THE ENGINEERING IMPLICATIONS 39
unit consists of distal turbidites; the upper from scattering and attenuation of signal
part of this unit decreases in thickness from energy due to short-distance changes of sed-
west to east and north-east. iment properties and bedforms as a result of
On 3.5 kHz records five acoustic sediment braided and narrow silt layers in the plain.
types with transitional boundaries are rec- This echo-type occurs in the generally flat
ognized which show the following charac- central part of the area and also has exten-
teristics (Duin, 1985): sions in an ENE direction.
(3) Acoustic sediment type IIIa is charac-
(1) Acoustic sediment type I represents terized by a 'transparent', low-reflective
predominantly continuous sub-bottom seabed with parallel and continuous regu-
reflectors below a relatively flat seabed. larly laminated reflectors with good resolu-
Penetration is moderate (40-70 ms). This tion (Fig. 11). Signal penetration is good
acoustic type is typical for the turbiditic (70-110 ms). Acoustic sediment type IlIa
area west of approximately 64 oW and north interpreted as pelagics with turbiditic over-
of 23 °45'N. bank deposits has a patchy occurrence in the
(2) Acoustic sediment type II is representa- central part of the area, where the sea floor
tive of irregular and discontinuous sub- shows minor undulations with vertical amp-
bottom reflectors that can be traced later- litudes of up to a few metres.
ally over only a few hundred metres (Fig. (4) Acoustic sediment type IIIb represents
10). These features and the low signal a transparent upper layer with generally
penetration (30-50 ms) probably originate continuous sub-bottom reflectors. Thickness
- ,
! \'! ! , '\ I ! i
, ' . I
, ,i ' I
II
I I
i ' I ; ! I ~ . I: ' !. '- 1
1 : \ ,. .. ' •• : :.
IIt
I
I • , . I I ! ,: ' . , I I • .
,, I II I ! -I .:! '
1
J ' I ' I
' ! :I
, i l l
. ,l ! , ~
1 .
! I! i . ;~ I' \ ! \.,' ! 1
,-~
• :1 I
! I ' ! I I' I :
I' , '
: :.
: ~ I : .
I '1
.!
.i ,. 'I'i ! .
I
~
l! 1\ : ,
i
! I :~ " .
:, .' ,
..
I I . ! _
"
I' i I
. :
, ·1
,.
--
,',
,;
• • I I
. I 'I
, .1..
It ! ·' :" .
' , '. ..I
Fig. 10 Example of NAP acoustic sediment type II with characteristic irregular and discontinu-
ous reflectors and with focusing features. Distance between horizontal lines is 100 ms. Profile length
is c. 6.5 km
40 ADVANCES IN UNDERWATER TECHNOLOGY
...
Fig. 11 Example of NAP acoustic sediment type lila with parallel and continuous retlectors and
some apparent faults and focusing features. Distance between horizontal lines is 100 ms. Profile
length is c. 6.5 kIn
of the upper layer varies between 10 and 20 (Fig. 11). These faults are less obvious than
ms; signal penetration is moderate to good those observed in the Madeira Abyssal
(60-90 ms). This clay layer is found in an Plain; but the difference between echo-
area with a slightly deeper sea floor, cen- focusing features and faults is not easy to
tered near 23°20'N, 63°W and orientated in distinguish (Duin, 1985). It could be that
a WSW to ENE direction. small faults will turn out to be a normal fea-
(5) Acoustic sediment type IIlc is represen- ture of abyssal plains with a turbidite fill, if
tative of continuous sub-bottom reflectors these abyssal plains receive the same atten-
(thickness c. 30 m), with a transparent tion as the Madeira and Nares Abyssal
pelagic unit below, in which two distinct Plains.
reflectors conformable to the oceanic base-
ment occur. Acoustic sediment type IIIc is
Stratigraphy
found in the southeastern sector with its
irregular morphology and is most directly
affected by the deep contour current. Some 50 piston cores up to 13 m taken from
1981 onwards have shown that the superfi-
Not fully understood small faults compar- cial sediments consist of grey clays and silts
able to those observed in the Madeira and brown clays. CaC03 content in clays is
Abyssal Plain seem to occur, especially in low (5%).
the central part of the Nares Abyssal Plain, The frequency, thickness and burial depth
among numerous echo-focusing features of silt layers are variable from core to core
ATLANTIC ABYSSAL PLAINS AND THE ENGINEERING IMPLICATIONS 41
Cumulated
thickness Imm
of sill
(laminated)
int ervo Is in
upper 030m
of sediment
~>5O"""
~ 25-50 ....
~~~ <25mm
"" .......
ill lll~ i':'r-..gorl_
a Abyt.~1 hllb
DE""'".. .
DEtb'.... u
EIJ E.dIoOI~mA
D w.oty•• me
Ech~lPl' me
Fig. 12 Nares Abyssal Plain: silt dispersal pattern in uppermost 0.3 m with acoustic sediment
types (echotypes). (From Kuijpers and Duin, 1985)
Cumulaled
thici<""sslcm)
of silt
llaminated)
intervals in
upper 6.50m
of sediment
~ >7Sem
~ 2S-7Scm
~.-;-:.. ---2Setn
D"'~5OII\i.ll!.
DE-!>.. ,
o Eeh • ."..n
mIl EmOIyp. alA
0......,..·,.8
_[<too". .. ,.C
Fig. 13 Nares Abyssal Plain: silt dispersal pattern in uppermost 6.5 m with acoustic sediment
types (echotypes). (From Kuijpers and Duin, 1985)
increase of 2 kPa near the surface to 9 kPa at sisting of gravity to turbidity flows; the sed-
8 m down. iment transport mechanisms in the Nares
Permeability values of the clay layers are Abyssal Plain appear to be much more con-
of the order of 10- 7 to 10-8 cm/s, and for silts tinuous, consisting of the interaction of tur-
they are thought to be 10-4 to 10- 5 cm/s. bidity flows and contour-following currents
(Kuijpers and Duin, 1985). Maximum grain
sizes are about 450 m (MAP) and 50 m
COMPARISON OF MAP AND NAP (NAP), and maximum CaC0 3 contents are
c. 99%andc. 17%. The coarse sediment of the
As outlined above, recent studies have MAP consists of angular volcanic mineral
shown marked differences in transport grains; however, that of the NAP consists
mechanisms and sedimentary patterns in mainly of well-packed micas, which might
the abyssal environments of the Eastern explain the differences in water content.
and the Western North Atlantic Basin, This appears to influence the performance of
which significantly affect sediment proper- conventional sampling equipment like piston
ties of the abyssal plain sediments involved. cores. The great areal sedimentary variabil-
A summary of these differences was based ity in NAP (Figs 12 and 13) is probably
on data from the Madeira Abyssal Plain in caused by topographic effects on the flow
the eastern basin and the Nares Abyssal regime, and by the fluctuating bottom cur-
Plain in the western basin. rent strength. In the gravity-flow environ-
The Madeira Abyssal Plain shows a highly ment of MAP variability is much lower (Fig.
episodic sediment transport mechanism con- 6), which means that the predictability of a
ATLANTIC ABYSSAL PLAINS AND THE ENGINEERING IMPLICATIONS 43
site in an abyssal plain like the MAP is much (Eds S. Saxov and J.K. Nieuwenhuis).
higher than in a NAP-type abyssal plain. A Plenum Press, New York-London, pp.
sampling station grid of, say, 10/10 miles in 189-213.
MAP would give a degree of reliability simi- 5. Kuijpers, A. 1982. Sediment studies in the
lar to or even higher than a 1/10 mile station western Madeira Abyssal Plain. Geological
Studies on Abyssal Plains in the North
grid in NAP. Atlantic, Part II. Internal Report, Rijks
Geol. Dienst, Haarlem, pp. 41-107.
6. Kuijpers, A. 1985. Sediments of the South-
ACKNOWLEDGEMENTS ern Nares Abyssal Plain, Western North
Atlantic. Geological studies in the Western
The present knowledge about the Madeira North Atlantic. Internal Report, Rijks
and Nares Abyssal Plains is to a large Geol. Dienst, Haarlem (in press).
extent due to the activities of the national 7. Kuijpers, A. and Duin, E. J. Th. 1985.
site assessment programmes of Canada, The Boundary current-controlled turbidite
Netherlands, the United Kingdom and the deposition: a sedimentation model for the
United States. Permission to publish this southern Nares Abyssal Plain, western
paper was granted by the Director of the North Atlantic. Geomar. Lett. (submitted).
8. Kuijpers, A., Rispens, F. B. and Burger, A.
Geological Survey of the Netherlands. Mr J. W. 1984. Late Quaternary sedimentation
Bruinenberg supervised the drawings and and sedimentary processes on the Madeira
Mrs N. Bauer typed the manuscript. Abyssal Plain, Eastern North Atlantic.
Meded. Rijks Geol. Dienst, 38(2), 91-118.
9. Noel, M. 1984. Measurements of sediment
REFERENCES temperatures, conductivity and heat flow in
the North Atlantic and their relevance to
1. Auffret, G., Buckley, D., Laine, E. P., radioactive waste disposal. lOS Report 172,
Schiittenhelm, R. T. E., Searle, R. C. and Institute of Oceanographic Sciences.
Shephard, L. E. 1984. NEA Seabed Work- 10. Searle, R. C., Weaver, P. P. E., Schul-
ing Group: status on site qualification for theiss, P. J., Noel, M., Jacobs, C. L., Hug-
nuclear waste disposal within deep sea sed- gett, G. J. and McGiveron, S. 1984. Great
iments. SAND 83-2037, Sandia National Meteor East (Distal Madeira Abyssal Plain):
Laboratories, Albuquerque, USA, pp. 1-64. geological studies of its suitability for dis-
2. Duin, E. J. Th. 1985. Geophysics of the posal of heat-emitting radioactive wastes.
Southern Nares Abyssal Plain, Western Report Institute of Oceanographic Sciences.
North-Atlantic. Geological Studies in the 11. Shephard, L. E. 1985. Geotechnical prop-
Western North Atlantic. Internal Report, erty characteristics of Nares Abyssal Plain
Rijks Geol. Dienst, Haarlem (in press). sediments: a summary report. Geological
3. Duin, E. J. Th., Mesdag, C. S., and Kok, P. Studies in the Western North Atlantic.
T. J. 1984. Faulting in Madeira Abyssal Internal Report, Rijks Geol. Dienst, Haar-
Plain sediments. Mar. Geol. 56, 299-308. lem (in press).
4. Embley, R. W. 1982. Anatomy of some 12. Tucholke, B. E. and Ewing, J. I. 1974.
Atlantic margin sediment slides and some Bathymetry and sediment geometry of the
comments on ages and mechanisms. In Greater Antilles Outer Ridge and vicinity.
Marine Slides and other Mass Movements Bull. Geol. Soc. Am. 85, 1789-1802.
4
The Quaternary Succession on
the Northern United Kingdom
Continental Shelf and Slope:
I mplications for Regional
Geotechnical Investigations
M. S. Stoker, D. Long, A. C. Skinner and D. Evans, British Geological
Survey, Edinburgh, Scotland
45
46 ADVANCES IN UNDERWATER TECHNOLOGY
technical map indicates regional variation Quaternary maps which present an insight
in sediment strength, and this is related to into the geometry of the Quaternary succes-
the geometry of the Quaternary sequence. sion and its geotechnical variations, thus
providing an essential element of any site
investigation project.
INTRODUCTION This paper presents an overview of the
Quaternary succession on the Northern UK
The area of study extends from 56 oN to Continental Shelf and Slope. It includes a
62°N and covers the Northern UK Conti- brief reference to the pre-Quaternary
nental Shelf and Slope down to about 1100 m geological framework as this has influenced,
water depth (Fig. 1). The British Geological to some extent, the position of Quaternary
Survey (BGS) has been actively surveying basins. Most of the early stratigraphic work
this area since 1968 as part of its regional in the North Sea and Sea of the Hebrides
offshore mapping programme. Conse- has been revised since 1980; therefore this
quently an extensive geological and study presents an update on the Quaternary
geophysical database has been developed for geology of these areas. Although work on
the offshore area. These data, which total the West Hebridean and North-West Scott-
246 boreholes, about 13 500 shallow sample ish Shelf is still at an early stage, a brief
stations, and in excess of 75 000 km of summary of the available data is included. A
seismic traverse, have been used to compile large amount of geotechnical data exist for
a variety of offshore geological maps at the Quaternary sequence, and it has been
1 : 100 000 and 1 : 250 000 scale (see Fan- possible to produce a summary geotechnical
nin, 1980). These maps are especially valu- map for most of the North Sea and for local
able to the offshore industry as they contain areas in the north-west. It is shown that a
regional background data for more localized knowledge of the Quaternary geology is cru-
site studies. Of particular interest are the cial to the understanding of regional
AREA OF STLO'f
SEA
CENTRAL
NORTH
SEA
geotechnical variation on the shelf and towards the graben system that has been
slope. the focus of deposition in the area since
Permian times (Glennie, 1984). This graben
system, which was formed by extensional
faulting during the Mesozoic, consists of the
PRE-QUATERNARY GEOLOGICAL Central and Viking Grabens as well as the
FRAMEWORK westerly branch which forms the Moray
Firth Basin. The Mesozoic rocks are over-
The area may be considered as two geologi- lain by softer Tertiary sediments which
cal provinces: the North Sea region and the again thicken towards a subsiding deposi-
area to the west and north-west of Scotland. tional centre overlying the grabens,
The provinces are separated by the predo- although their deposition was not accom-
minantly Precambrian basement massif of panied by faulting.
the Scottish Mainland and its north- To the west and north-west of Scotland
northeastward extension to Shetland, which (Naylor and Mounteney, 1975; Stoker and
includes Upper Palaeozoic Devonian sedi- Fannin, 1984) Permian and Mesozoic rocks
ments (Fig. 2.). Mesozoic sediments occur are again common offshore, although they
only on the periphery of the landmass on the generally occur in smaller faulted basins
west coast and around the Moray Firth. separated by Precambrian basement ridges
These exposures provide a hint of the great which often form rugged topography.
change that takes place in the geology off- To the west of Scotland, early Tertiary
shore. igneous activity has produced rocks which
In the North Sea, Permian and Mesozoic tend to form irregular topography, whereas
rocks occur within a few kilometres of the Tertiary sediments on the Atlantic side of
Mainland coast (Evans et al., 1981), except Scotland are almost entirely restricted to
east of Shetland, and thicken eastwards the outer shelf and slope.
3· 2·
,. 2·
rately defined. West of Shetland and may represent several individual Quater-
north-west of Orkney the Quaternary suc- nary formations.
cession appears to thicken towards the shelf Mapping in the North Sea is almost com-
break. However, to the south-west of the plete; the blank areas on the map are cur-
Wyville-Thomson Ridge the sequence rently being worked upon. West of Scot-
appears much thinner at the shelf break, land, regional mapping has only been com-
and the Quaternary cover on the ridge itself pleted in the Sea of the Hebrides, although
is generally thin, and locally absent. The areas on the north-west shelf are being
sequence may be thicker again to the south actively interpreted. Information therefore
of St Kilda. Down the continental slope the exists for most of the shelf and slope and this
thickness of the Quaternary is presently is summarized below.
unknown.
Figure 4 depicts areas of Lower, Middle
and Upper Pleistocene as well as Holocene North Sea
sediments at or near the seabed. It does not
include the uppermost 'mobile layer' which Quaternary stratigraphic sequences have
reflects present-day seabed processes, been established in the Central and N orth-
although a brief summary of the seabed sed- ern North Sea. Diagrammatic sections illus-
iments is included at the end of this section. trating the stratigraphic successions are
This Quaternary map is simplified from the shown in Figs 5 and 6, and a basic chronos-
1 : 250 000 map series, although the latter tratigraphic framework which correlates the
tend to show individual Quaternary forma- two areas is present in Fig. 7. This stratig-
tions as the primary mappable unit. Figure raphic framework is based on the Dutch
4 is therefore an amalgamation of the stratigraphic classification (Zagwijn, 1979)
stratigraphic data, and each unit on the map which takes the base of the Quaternary to be
.KE.!..
aUATERNARY RlRMAnONS
FH - f'o,lh Fm weA -w... Bene i. Fm
woo -W;tch Gro ...,Hin alP -Cool P; I Fm
see-51 AtIC. Fm FIS-Fishe< Fm
SWAT-Swolchwoy Fm LBK - ling 9ori< Fm
MAR- Mar Bene Fm ANG - AberdHn Clrotr<l Fm
Fig. 5 Cross-section showing the geometry and general seismostratigraphic relationships of the
Quaternary formations in the Central North Sea. (Vertical exaggeration x234)
at about 2.5 million years BP. The Lower/ Middle and Upper Pleistocene sediments in
Middle Pleistocene boundary is generally the area. This clearly refutes the earlier
taken at the Brunhes/Matuyama palaeo- hypothesis that the greater part of the
magnetic boundary (Butzer and Isaac, Quaternary succession is of Upper Pleis-
1975), which has been variably dated at tocene age (Holmes, 1977; McCave et al.,
730 000 years BP (Mankinen and Dalrym- 1977; Eden et al., 1978).
ple, 1979) and 790 000 years BP (Johnson, Ten Quaternary formations have been
1982), and is placed towards the base of the established and these can be mapped over
'Cromerian Complex' (Zagwijn, 1979). The the entire area of the Central North Sea. A
base of the Upper Pleistocene is taken at the detailed account of the stratigraphy has
beginning of the Eemian (Butzer and Isaac, been presented elsewhere (Stoker et al., in
1975), which began at about 128 000 years press, a) and this is summarized below.
BP (Bowen, 1978). The Upper Pleistocene/ The geometry of the sequence is illus-
Holocene boundary is taken at about 10 000 trated in Fig. 5, which shows the succession
years BP (Mangerud et al., 1982). increasing in thickness towards the east,
and it is in this area that the most complete
Central North Sea sequence is preserved. In the west, the base
of the succession rests with angular uncon-
The Quaternary succession in the Central formity on rocks of Palaeozoic, Mesozoic and
North Sea has recently been revised (Stoker Tertiary age. To the east, basal Pleistocene
et al., in press, a and b), and a more com- has nowhere been proved and the base of
plete sequence of Pleistocene sediments has the succession has still to be defined. The
now been recognized. Micropalaeontological Aberdeen Ground Formation is the oldest
data (Stoker et al., in press, a and b) and proven Quaternary unit and ranges from
amino acid geochronology (Brigham-Grette Lower to Middle Pleistocene (?Tiglian to
and Sejrup, 1984) support an earlier Elsterian) in age. With the exception of
palaeomagnetic study (Stoker et al., 1983) minor isolated outliers in the west of the
which identified the existence of Lower, area (too small to be shown on Fig. 4), this
THE NORTHERN UNITED KINGDOM CONTINENTAL SHELF AND SLOPE 51
formation occurs wholly at subcrop beneath deglaciation. The Wee Bankie Formation is
younger sediments. The Lower Pleistocene a subglacial till, whereas the Marr Bank,
part of the succession consists predomin- Swatchway and St Abbs formations com-
antly of bioturbated marine muds with occa- prise glaciomarine sands and muds. The
sional interbedded sands which were depo- Forth and Witch Ground formations trans-
sited in a warm, temperate, inner to middle gress the Upper Pleistocene/Holocene
shelf environment. boundary and are shown on Fig. 4 as the
The Middle Pleistocene succession com- Holocene to late Weichselian subdivision of
prises the upper part of the Aberdeen the Holocene/Upper Pleistocene unit. The
Ground Formation, the Ling Bank and Forth and Witch Ground Formations trans-
Fisher formations and the basal part of the glacial estuarine and fluviomarine sands and
Coal Pit Formation. On seismic records the muddy sands overlying late-glacial gravelly
sequence is generally much more complex muds. The Witch Ground Formation is a
than the Lower Pleistocene and distinct predominantly argillaceous, basinal marine
irregular, interformational erosional sur- sequence with some minor glacial input (e.g.
faces characterize the sequence (Fig. 5). A dropstones) only in the lower part of the
change in environmental conditions is evi- sequence. Pockmarks are abundant in the
dent since the bulk of the sequence com- Witch Ground Formation. Radiocarbon dat-
prises sediments of a predominantly distal ing (Holmes, 1977) and pollen analysis (Jan-
glaciomarine nature. These are generally sen et al., 1979) of the Holocene sequence
argillaceous sediments with abundant scat- suggest that most of it was deposited in the
tered, matrix-supported clasts (dropstones) early Holocene between 7000 and 10 000
and shell fragments. Locally in the west, years BP.
subglacial and proximal glaciomarine sedi- Extension of the Central North Sea
ments have also been identified (Stoker and stratigraphy into the Moray Firth casts
Bent, in press). One thin (2 m) interglacial doubt on the suggestion by Chesher and
horizon has been identified towards the base Lawson (1983) that the sequence in this area
of the Ling Bank Formation and this has is strictly Upper Pleistocene to Holocene
been tentatively assigned a Holsteinian age. in age. This assumption is based solely on a
Reference to Figs 4 and 5 indicates that radiocarbon age date of 43 500 years BP
Middle Pleistocene strata are at, or near, obtained from shell material in their lower-
the surface in the east and north-east of the most unit. However, this basal unit can be
area. correlated seismically with the Aberdeen
Upper Pleistocene and Holocene sedi- Ground Formation, which implies that it is
ments crop out over a wide area of the Cen- Lower or Middle Pleistocene in age.
tral North Sea and comprise most of the
Coal Pit Formation, the Wee Bankie, Marr N orthern North Sea
Bank, Swatchway, St Abbs, Forth and
Witch Ground Formations. Eemian intergla- Aspects of the Quaternary sequence in the
cial strata have been identified near the base Northern North Sea have recently been
of the Coal Pit Formation and up to 30 m of presented by Skinner and Gregory (1983),
bioturbated marine muds and shelly sands Stoker et al. (1983) and Rise et al. (1984).
have been recovered locally. The upper part Eleven Quaternary formations have been
of the Coal Pit Formation consists mainly of mapped throughout the area, and their
proximal and distal glaciomarine sands and generalized stratigraphic relationships are
muds of uncertain Lower/Middle Weich- illustrated in Fig. 6.
selian age. The remainder of the Upper The basic geometry of the sequence
Pleistocene sequence is composed of Upper reflects a northeasterly offlapping succes-
Weichselian sediments, which reflect the sion which is thickest in the area of the Vik-
last regional glaciation and subsequent ing Graben. It is apparent from cross-
52 ADVANCES IN UNDERWATER TECHNOLOGY
A e
W
OUAlERNARY FORMATIONS
SP -9'@t land P!ot kw'm secknents
FLG-Flogs Fm
VKB-Vlklng BorV< Kn
KSE -Kleppe 5('nlo( Fm
NR - Norwegian Trench F"m
TAM - Tomppn Fm
SoPE -~erus Fm
C50 -Cope Storq Fm
f D - Fprde-r Fm
MRN - Mar.ner Fm
SHN - Shackle ton F'm
PRE-QUA1ERNARY
GN - RJIooog<!oo oro Noogone
o- Qe·,,,omon
"" ....'
4'E
Fig. 6 Cross-section showing the geometry and general seismostratigraphic relationships of the
Quaternary formations in the Northern North Sea (Vertical exaggeration x234)
section A-B in Fig. 6 that although the con- muds, which are very gravelly at the base of
tinental margin is pre-Quaternary in age the the succession in the west although the
present-day profile is clearly due to deposi- gravel content decreases to the north and
tion and selective erosion of Quaternary east. These sediments crop out in a broad
strata. band to the north and east of the Lower
In the extreme west of the area the base Pleistocene outcrop (Fig. 4).
of the sequence rests with angular uncon- The Upper Pleistocene/Holocene succes-
formity on Tertiary and older strata. To the sion includes the bulk of the Ferder Forma-
east, the base appears to rest conformably tion, the Cape Shore, Sperus, Tampen and
on Upper Pliocene sediments. The lowest Norwegian Trench formations. These are of
Quaternary unit is the Shackleton Forma- Eemian to late Weichselian age, while the
tion, which comprises Lower Pleistocene Shetland Platform sediments and the Flags,
marine sands and clays that were deposited Viking Bank and Kleppe Senior formations
in a boreal, estuarine-to-offshore basinal are late Weichselian to Holocene in age.
environment. These sediments crop out as a Shelly, sandy, silty clays of Eemian inter-
narrow strip in the central part of the area glacial age have been identified (Skinner and
(Fig. 4). Gregory, 1983) near the base of the Ferder
The Middle Pleistocene sequence rests Formation. However, most of the sequence
unconformably on the Lower Pleistocene consists of Weichselian sediments, which
sediments and consists of the Mariner For- include glacial gravelly sandy clays, as well
mation and the basal part of the Ferder as glaciomarine sandy and silty clays, and
Formation. Interformational boundaries are muddy sands. The late Weichselian to
markedly irregular. The sediments consist Holocene units are mostly lateral equival-
dominantly of glacial or glaciomarine sandy ents. The Shetland Platform sediments con-
THE NORTHERN UNITED KINGDOM CONTINENTAL SHELF AND SLOPE 53
sist of glacial till overlain by late- to post- stratigraphic sequences have many
glacial clays which form a veneer on the similarities and both are correlatable with
pre-Quaternary bedrock in the area the Quaternary succession established in the
immediately east of Shetland. The Flags Southern North Sea (Cameron et al., 1984).
Formation and lower part of the Viking Although much more detailed work remains
Bank Formation are dominantly argillace- to be done to better define the individual
ous and tend to occupy hollows. Pockmarks stratigraphic units, a consistent picture is
occur locally at seabed in the Flags Forma- emerging throughout the North Sea.
tion. The upper part of the Viking Bank
Formation includes sands and gravels depo- West and North-West of Scotland
sited in an environment characterized by
shallow mobile barrier islands. The Kleppe Regional Quaternary mapping in this area is
Senior Formation crops out in the N or- still at a preliminary stage; however, a great
wegian Trench and is characterized by late deal of basic data exists and these are briefly
to post-glacial clays with surface pockmarks summarized below.
and iceberg scour marks in some areas. The
Upper Pleistocene and Holocene sediments Sea of the Hebrides
occur widely at outcrop over most of the
Northern North Sea (Fig. 4). The Quaternary stratigraphy in this area
was originally described by Binns et al.
Stratigraphic Correlation (1974) but has recently been revised by
Davies et al. (1984) whose work is summar-
A tentative correlation between the ized below.
Quaternary formations in the Central and The succession has been divided into eight
Northern North Sea is shown in Fig. 7. The formations whose relationships are shown
schematically in Fig. 8. The pre-Quaternary
bedrock surface is characterized by consid-
erable variations in relief in this area and
the bulk of the Quaternary sediments is pre-
served in the basins. There are no direct
chronological controls on the Quaternary
sediments and their age remains unproven.
Saolian The basal three units, the Skerryvore,
Malin and Canna formations, are considered
Holsteinian
:a
.JII-------l
~ Elste!'ian
E
~ Cromerian
~ -~
~ Menapial
WJalian
::::}:::::~
1------;- - - ::: : :::::: : : ::::: .~n
::::::::::::::::::::::::::::
.............................. . . . . . . ...
... .... .,........
..........................' .... ':-:':':-:':':-:':':-:':':-:"
Fig. 7 Correlation of the Quaternary forma-
tions in the Central and Northern North Sea Fig. 8 Schematic cross-section showing the
with the Pleistocene stages of N orth-West general relationships of the Quaternary forma-
Europe tions in the Sea of the Hebrides
54 ADVANCES IN UNDERWATER TECHNOLOGY
B
unit is absent and the argillaceous sediments
0
Soa_
0 crop out at the seabed. At present there
\00
are no chronological data available for this
part of the sequence. Micropalaeontological
200
data have indicated the presence of temper-
XX)
ate horizons in several boreholes which sug-
IIJO gests the possibility that pre-Upper Pleis-
~ tocene sediments may be present. However,
!IlO
the subdivision of Quaternary units is still
C uncertain and reference to Figs 9 and 10
i'>JJ illustrates the lateral and vertical stratig-
!IlO
raphic changes. Work in the North Sea has
9:1) amply demonstrated the variable distribu-
an tion of Pleistocene and Holocene sediments,
1\00
and the sequence can vary a great deal in
age between adjacent boreholes. Despite
1200
uncertainties regarding the dating of the
llOO
sequence, micropalaeontological data sug-
1<00 gest that the bulk of it was deposited under
0 km 20
'--------' harsh climatic conditions, mainly in a
glaciomarine environment. Closer to land
Fig. 10 Cross-section illustrating the nature areas such as Shetland (Long and Skinner,
of the Quaternary succession on the N orth-West in press), subglacial sediments have been
Scottish Shelf. (Vertical exaggeration x 100) identified.
Towards the shelfbreak at about 200 m
corer. A cross-section across part of the water depth, the top sand unit becomes
shelf and slope is shown in Fig. 10. The sec- generally finer, although gravelly sands
tion is located in an area of the shelf where have been recovered down to 90 m water
the base of the Quaternary has been iden- depth in the Wyville-Thomson Ridge area.
tified, although base Quaternary remains Down the slope the sand cover becomes
undetermined on the slope. The variable very fine and silty and is generally much
thickness of the sequence and its relation- thinner (~20 cm) than on the shelf. As on
ship to the underlying pre-Quaternary geol- the shelf, the underlying sediments are
ogy is well illustrated in the section. argillaceous, while the contact between the
On the shelf, the sequence comprises a two facies varies from sharp to transitional.
cover of shelly sand and gravelly sand, West of Sula Sgeir, the argillaceous sedi-
mostly less than 2 m thick but locally up to ments on the upper slope, between 200 m
20 m thick, which overlies predominantly and 600 m water depth, have a distinct
argillaceous sediments characterized by gravel fraction and resemble the
gravelly silty muds with subordinate sandy glaciomarine sediments on the shelf. Below
horizons. Micropalaeontological data sug- 600 m water depth the argillaceous sedi-
gest that the top sandy unit is Holocene in ments in this area are gravel free. However,
age and on parts of the shelf is still subject this distribution is not ubiquitous as there is
to reworking. The contact between the sand a widespread occurrence of a gravel fraction
and the underlying argillaceous sediments is in the argillaceous facies down to at least
usually very sharp and erosional. Locally 800 m on the south-west flank of the
the top of the argillaceous sequence appears Wyville-Thomson Ridge, and down to 110 m
to be weathered, and this weathered zone in the area north-west of Shetland.
can occur in present-day water depths of It is probable that the sediments pre-
down to 160 m. In some areas the top sand served on the slope represent a more con-.
56 ADVANCES IN UNDERWATER TECHNOLOGY
densed and more complete Quaternary their thickness is variable and their age
sequence than those preserved on the shelf. often difficult to determine. In the northern
Correlation between the shelf and slope has part of the Central North Sea early
not yet been attempted. Holocene sediments occur very close to the
seabed and only the top 0.15 m have been
Seabed Sediments subjected to minor reworking (Erlenkeuser
1978). The essentially relict nature of the
The distribution of seabed sediments shown seabed in this area is confirmed by the local-
in Fig. 11 has been simplified from the BGS ized occurrence on the sea floor of sea-ice
1 : 250 000 map series. These maps are scour marks formed about 18 000 years BP
based on grain-size analyses of seabed (Stoker and Long, 1984). Areas of relict sed-
deposits and are classified after Folk (1954). iment also occur on the Hebridean and
In the North Sea sand is ubiquitous with N orth-West Scottish Shelf and in the
gravelly sediments occurring mainly on Northern North Sea where late Weichselian
topographic highs in shallower water, and iceberg scour marks have been observed at
muddy deposits generally occur in water the seabed (Belderson et al" 1973; Skinner,
depths below 120 m (Owens, 1981). On the 1983).
N orth-West Scottish Shelf sandy and Elsewhere, areas of the shelf are covered
gravelly sediments are predominant, by a variety of modern bed forms formed as
whereas the sediment distribution in the a result of sand and gravel on the shelf and
Hebrides and Minches is highly variable but in the nearshore zone, being transported by
tends to be depth dependent. storm and tidal currents. In the area around
The seabed sediments form the upper- north-east Orkney, for example, mobile
most layer of the Quaternary sequence but sandwaves with a thickness of up to 20 m
51'
m
,: ; ·:··~
..... ~y r<>rdIscndy 9"M4
Aoc~ outaClP
Fig. 11 Simplified map showing the distribution and lithology of the seabed sediments. (Com-
piled by Colin Graham, BGS)
THE NORTHERN UNITED KINGDOM CONTINENTAL SHELF AND SLOPE 57
have recently been described (Farrow et al., recording the undrained shear-strength
1984). The 'mobile layer' is, however, more value at, or near, one metre below the sea-
related to present-day conditions than to the bed and noting the lithology of the sample.
Quaternary history of the area. The values were divided into four groups to
indicate the undrained shear strength as
measured by a hand vane at 1 m depth. One
GEOTECHNICAL PROPERTIES metre was chosen as a useful horizon as it
was reached by many of the gravity cores as
As an integral part of examining cores col- well as the vibrocores collected, and it is also
lected on the UK shelf and slope, simple an important level in studies of a regional
geotechnical measurements are made using nature, such as pipelines. Below 1 m there is
pocket penetrometers and handvanes. Sub- a marked reduction in the frequency of read-
samples may also be taken for moisture ings and their distribution is too spasmodic
content and density measurements. Occa- to attempt to produce a geotechnical map at
sionally other geotechnical tests are under- a lower depth. However, values and com-
taken, including triaxial studies, consolida- ments are recorded on the 1: 250 000
tion tests, particle-size analyses and Atter- Quaternary Geology map series.
berg limits.
Together with data collected by commer- Geotechnical Map (1 m depth)
cial agencies, an attempt has been made to
provide ranges of values for various The simplified map (Fig. 12) indicates sev-
geotechnical parameters for the different eral features which are related to the
seismostratigraphic formations identified in geological history of the area. The geotech-
order to produce a simplified geotechnical nical properties of the sediments are very
map of parts of the UK Continental Shelf variable over this wide area, since the range
and Slope (Fig. 12). This was made by is dependent on many factors, such as lithol-
6~~J:~~~~I;~~~~~TI--~~~r-~~
GEOTECHNICAL MAP dep.h) (rn
Undro,nfd sneer strength
<20kA:l
2O-501<Po
ogy, clay mineralogy and the stress history Shetland the shear strengths are in distinct
of the site. The latter is extremely impor- zones. The ridge at the plateau edge formed
tant when it is considered that a site may by the Tampen Formation (see Fig. 6) forms
have had large thicknesses of ice resting on a well-pronounced area of stiff gravelly and
it, or sediment removed from it. Moreover, sandy clays thought to represent a lateral
large areas of the North Sea were dry land moraine of late Weichselian age, while very
during the Quaternary and dessication and soft sediments occur in the Norwegian
periglacial processes may have affected the Trench and on the continental slope. The
sample. Quaternary geology west of Scotland is
In an arc across the Forth Approaches is complex and no attempt has been made to
an area with shear strengths typically in portray geotechnical data at 1 m depth on
excess of 100 kPa. This corresponds with the the small-scale map. The Sea of the Heb-
Wee Bankie Formation, believed to repres- rides has widespread areas of very soft
ent an end moraine of Late Weichselian age recent clays close to tills and rock outcrops.
(Stokeret al., in press, a). To the east of this The area west of Orkney and Shetland has
is a large area of sand, which equates with only recently been examined and so the map
the Marr Bank and Forth formations. In the only indicates a limited area. However,
extreme east lies an area of highly variable certain features do show up, such as the
shear strength, which reflects the complex sharp break at the shelf edge where stiff
distribution of Quaternary units (see Figs 4 glaciomarine clays occur to a water depth of
and 5) which have been subject to varied 200 m beyond which very soft muds are
stress histories. The pattern is further com- recovered on the continental slope. There
plicated by the localized occurrence of are two areas of exception, the northern end
fluviomarine sediments of Holocene age. of the Tampen Formation referred to above
In the northern part of the Central North and the Wyville-Thomson Ridge where stiff
Sea is a large area of soft sediments, repres- clays are recovered at 1 m depth in water
enting a distinct basinal area of Upper Pleis- depths up to 350 m beyond which stiff clays
tocene to Holocene marine sedimentation. It are recovered below a slightly thicker (1-2.5
is known as the Witch Ground Basin. On the m) covering of very soft muds. It is thought
southern edge of this basin are areas of sand that the ridge may have caused the ground-
corresponding to the Swatchway Forma- ing of large icebergs drifting down the
tion, which lie adjacent to areas of stiff to Faeroe-Shetland Channel during periods of
very hard clays formed by the Fisher For- lower sea-level, resulting in the overconsoli-
mation. East of Caithness and Orkney is an dation of the seabed.
area of sand and rock outcrop with small On the north-west shelf the shear
patches of mud of various strengths (not strengths are generally high as the near-
shown on the map). seabed sediments are possibly tills or
North of the Witch Ground Basin, in the glaciomarine sediments which have been
Northern North Sea, is an area of variable subsequently loaded by ice or affected by
lithology and shear strengths similar to the dessication or periglacial processes. There
eastern part of the Central North Sea. In are small areas of recent soft marine clays
the east is a large sandy area with several and sands on top of these stiff clays. Fre-
banks (Rise et al., 1984), which correlates quently rock occurs at or close to the surface
with the Viking Bank Formation and with near the various islands and banks of this
the sandy southern outcrop of the Cape area.
Shore Formation. East of the Shetlands are
several depressions filled with very soft sed- Below 1 m depth
iments of late Weichselian age (the Flags The sediments below 1 m depth are
Formation), overlying Lower to Middle extremely variable both vertically and lat-
Pleistocene firm to hard clays. North-east of erally and their shear strength depends on
THE NORTHERN UNITED KINGDOM CONTINENTAL SHELF AND SLOPE 59
their geological history. The Lower and Lithological variation is also important,
Middle Pleistocene sediments (the Shackle- particularly the recognition of buried
ton, Mariner and Aberdeen Ground forma- gravelly sequences which often cause prob-
tions of the North Sea and the Skerryvore lems for offshore drilling companies when
and Malin formations of the Sea of the Heb- they attempt to get holes 'spudded-in'.
rides) vary from normally consolidated to
heavily overconsolidated with undrained
shear strengths locally in excess of 1500 CONCLUSIONS
kPa, although figures of 500-800 kPa are
typical. Regional offshore mapping has provided a
Some Upper Pleistocene sediments basis for detailed geological studies. The
exhibit high strengths near the seabed development of a regional Quaternary
(approximately 5 m depth) for various framework has led to a greater understand-
reasons. In the Northern North Sea, where ing of how and why the geotechnical proper-
strengths of 800 kPa have been recorded in ties of sediments vary on the shelf and slope.
the upper 10 m, some cementation has been The occurrence of Quaternary units of dif-
noted (M. J. Edge, 1984, personal communi- ferent ages and with different stress his-
cation). In the UK block 16 similar high val- tories, at or near the seabed, is a major
ues have been recorded (Marsland et al., factor governing the regional geotechnical
1982), probably due to permafrost, evidence variations. Therefore an appreciation of the
for which has been identified in several regional geological setting is a vital part of
parts of the North Sea (Derbyshire et al., in any site-investigation project.
press).
Elsewhere low shear strengths have been
recorded well below seabed from within ACKNOWLEDGEMENTS
channels or troughs extending to 150 m
below the shoulder of the channel. In the We would like to thank all of our colleagues
North Sea these channels are recorded at in the Marine Geology and Marine Geophys-
several horizons within the Quaternary ics Research Programmes who have assist-
sequence and they are an important aspect ted in the collection of the data. This paper
in the evaluation of a site for foundation is published with the permission of the
conditions. Other phenomena, which can be Director, British Geological Survey, NERC.
assessed on a regional scale, have been
described in detail elsewhere (Fannin,
1980). They include gassified sediments, REFERENCES
pockmarks and slope stability. Pockmarks,
for example, occur predominantly in the 1. Belderson, R. H., Kenyon, N. H. and Wil-
Witch Ground, Flags and Kleppe Senior son, J. B. 1973. Iceberg plough marks in the
formations; therefore their distribution can Northeast Atlantic. Palaeogeog., Palaeoc-
be mapped over a wide area of the shelf and lim., Palaeoecol. 13, 215-224.
slope. Areas of slumping and mass move- 2. Binns, P. E., Harland, R. and Hughes, M. J.
ment have been identified over most of the 1974. Glacial and post-glacial sedimentation
Northern UK Continental Slope, although, in the Sea of the Hebrides. Nature 248,
in contrast to similar features recorded 751-754.
within the Norwegian Sector (Fannin, 3. Bowen, D. Q. 1978. Quaternary Geology: a
stratigraphical framework for multi-
1980), they are relatively small-scale struc- disciplinary work. Pergamon Press,
tures up to 25 m depth and 2-5 km in length. Oxford. 221 pp.
Areas of slumped sediment have also been 4. Brigham-Grette, J. and Sejrup, H. P. 1984.
identified within open channels in the North Stratigraphic resolution of amino-acid geo-
Sea. chronology in North Sea Quaternary sedi-
60 ADVANCES IN UNDERWATER TECHNOLOGY
ments (Abstract). In Quaternary Stratigra- 16. Fannin, N. G. T. 1980. The use of regional
phy of the North Sea (Eds 1. Aarseth and H. geological surveys in the North Sea and adj-
P. Sejrup). Symp. Univ. of Bergen, pp. acent areas in the recognition of offshore
30-32. hazards. In Offshore Site Investigation (Ed.
5. Butzer, K. W. and Isaac, G. L. 1975. After D. Ardus). Graham & Trotman, London, pp.
the Australopithecines. The Hague. 5-22.
6. Cameron, T. D. J., Laban, C. and Schiit- 17. Farrow, G. E., Allen, N. H. and Akpan, E.
tenhelm, R. T. E. 1984. Flemish Bight: B. 1984. Bioclastic carbonate sedimentation
Sheet 52°N 02°E. Institute of Geological on a high-latitude, tide-dominated shelf:
Sciences and Rijks Geologische Dienst. Northeast Orkney Islands, Scotland. J. Sed.
1 : 250 000 Series: Quaternary Geology. Pet. 54, 373-393.
7. Caston, V. N. D. 1977. A new isopachyte 18. Folk, R. L. 1954. Sedimentary rock nomenc-
map of the Quaternary of the North Sea, 1. lature. J. Geol. 62, 345-351.
Rep. Inst. Geol. Sci. No. 77/11, 1-8. 19. Glennie, K. W. 1984. The structural
8. Caston, V. N. D. 1979. A new isopachyte framework and the pre-Permian history of
map of the Quaternary of the North Sea. In the North Sea area. In Introduction to the
The Quaternary History of the North Sea Petroleum Geology of the North Sea (Ed. K.
(Eds E. Oele, R. T. E. Schiittenhelm and A. W. Glennie). Blackwell Scientific Publica-
J. Wiggers). Symp. Univ. Uppsala, pp. tions, Oxford, pp. 17-39.
23-28. 20. Holmes, R. 1977. Quaternary deposits of the
9. Chesher, J. A. and Lawson, D. 1983. The central North Sea. 5. The Quaternary geol-
geology of the Moray Firth. Rep. Inst. Geol. ogy of the UK sector of the North Sea bet-
Sci. No. 83/5. ween 56°N and 58°N. Rep. Inst. Geol. Sci.
10. Chesher, J. A., Smythe, D. K. and Bishop, No. 77/14.
P. 1981. The geology of the Minches, Inner 21. Jansen, J. H. F., Doppert, J. W. C.,
Sound and Sound of Raasay. Rep. Inst. Hoogendoorn-Toering, K., De Jong, J. and
Geol. Sci. No. 83/6. Spaink, G. 1979. Later Pleistocene and
11. Davies, H. C., Dobson, M. T. and Whittingh- Holocene deposits in the Witch and Fladen
ton, R. J. 1984. A revised seismic strati- Ground area, northern North Sea. N eth. J.
graphy for Quaternary deposits on the inner Sea Res. 13, 1-39.
continental shelf west of Scotland between 22. Johnson, R. G. 1982. Brunhes-Matuyama
55°30'N and 57°30'N. Boreas 13, 49-66. magnetic reversal dated at 790,000 yr BP by
12. Derbyshire, E., Love, M. A. and Edge, M. marine-astronomical correlations. Quat.
J. In press. Fabrics of probable segregated Res. 17, 135-147.
ground-ice origin in some sediment cores 23. Long, D. and Skinner, A. C. In press. Gla-
from the North Sea basin. In Soils and cial meltwater channels in the northern isles
Quaternary Landscape Evolution (Ed. J. of Shetland. Scott. J. Geol.
Boardman). John Wiley. 24. Mangerud, J., Birks, H. J. B. and Jiiger, K.
13. Eden, R. A., Holmes, R. and Fannin, N. G. D. 1982. Chronostratigraphical subdivisions
T. 1978. Quaternary deposits of the central of the Holocene: a review. Striae 16, 1-6.
North Sea. 6. Depositional environment of 25. Mankinen, E. A. and Dalrymple, G. B. 1979.
offshore Quaternary deposits of the Conti- Revised geomagnetic timescale for interval
nental Shelf around Scotland. Rep. Inst. ~5 my BP. J. Geophys. Res. 84, 615-626.
Geol. Sci. No. 77/15. 26. Marsland, A., Prince, A. and Love, M. A.
14. Erlenkeuser, H. 1978. The use of radiocar- 1982. Role of soil fabric studies in the evalu-
bon in estuarine research. In Biogeochemis- ation of the engineering parameters of off-
try of Estuarine Sediments. Proc. UNES- shore sediments. In Proc. 3rd Int. Con! on
CO/SCOR Workshop, Melreux, Belgium, Behaviour of Offshore Structures. MIT,
pp. 14~153. Boston, Vol. 1, pp. 181-202.
15. Evans, D., Chesher, J. A., Deegan, C. E. 27. McCave, 1. N., Caston, V. N. D. and Fan-
and Fannin, N. G. T. 1981. The offshore nin, N. G. T. 1977. The Quaternary of the
geology of Scotland in relation to the IGS North Sea. In British Quaternary Studies
shallow drilling programme, 197~1978. - Recent Advances (Ed. F. W. Shotton).
Rep. Inst. Geol. Sci. No. 81/12. Clarendon Press, Oxford, pp. 187-204.
THE NORTHERN UNITED KINGDOM CONTINENTAL SHELF AND SLOPE 61
28. Naylor, D. and Mounteney, S. N. 1975. geological framework for the North West
Geology of the Northwest European Conti- United Kingdom continental shelf and slope.
nental Shelf, Vol. 1. Graham Trotman Dud- Internal Rep. No 84/21, BGS Marine Geol-
ley, London. ogy Research Programme.
29. Owens, R. 1981. Holocene sedimentation in 35. Stoker, M. S. and Long, D. 1984. A relict
the North-western North Sea. Spec. Publs ice-scoured erosion surface in the central
Int. Ass. Sediment. 5, 303-322. North Sea. Mar. Geol. 61,85-93.
30. Rise, L., Rokoengen, K., Skinner, A. C. and 36. Stoker, M. S., Long, D. and Fyfe, J. A. In
Long, D. 1984. Norlige Nordsj~. Kvar- press (a). A revised Quaternary stratigra-
taergeologisk kart mellom 60°30' og 62°N, phy for the central North Sea. Rep. Brit.
og ~st for 1°0. (Northern North Sea. Geol. Surv.
Quaternary geology map between 60°30' 37. Stoker, M: S., Long, D. and Fyfe, J. A. In
and 62°N, and east of l°E.) M. 1: 500000. press (b). The Quaternary succession in the
Institutt for Kontinentalsokkelun- central North Sea. Newsletters on Stratig-
ders~kelser (IKU), Norway. raphy.
31. Skinner, A. C. 1983. Cormorant: Sheet 38. Stoker, M. S., Skinner, A. C., Fyfe, J. A.
61°N-O°. Institute of Geological Sciences and Long, D. 1983. Palaeomagnetic evi-
and Continental Shelf Institute, Norway. dence for early Pleistocene in the central
1 : 250 000 Series: Sea Bed sediments. and northern North Sea. Nature 304,
32. Skinner, A. C. and Gregory, D. M. 1983. 332-334.
Quaternary stratigraphy in the northern 39. Zagwijn, W. H. 1979. Early and Middle
North Sea. Boreas 12, 145-152. Pleistocene coastlines in the Southern
33. Stoker, M. S. and Bent, A. In press. Middle North Sea basin. In The Quaternary His-
Pleistocene glacial and glaciomarine sedi- tory of the North Sea (Eds E. Oele, R. T. E.
ments in the west central North Sea. Schiittenhelm and A. J. Wiggers). Symp.
Boreas. Univ. Uppsala. pp. 31-42.
34. Stoker, M. S. and Fannin, N. G. T. 1984. A
5
The Benigraph: A Sea-floor
Mapping System
A. L~vik, Bentech A/S, Tromso, Norway
63
64 ADVANCES IN UNDERWATER TECHNOLOGY
data in all the other subsystems. The hard- hull-mounted or towed transducer. The sys-
ware in the DASAP consists of an HP1000 tem may use either or both modes.
A900 computer, an leAN graphic terminal Figure 2 shows a block diagram of the
and the necessary peripheral equipment to signal processing in the APR. As can be
store and present the data. Among the seen from this figure, the measurement of
many software modules in the DASAP are: the two angles is carried out by a corrella-
tion method.
• data reduction algorithms; The required time difference, and thus
• survey journal; the angle, is found by cross-correllating the
• survey statistics; hydrophone signals. This is accomplished in
• redisplay of covered area; the frequency domain by the use of an array
• graphics. processor, which performs the Fourier
A description of two of the subsystems, the transforms needed. This processor is a co-
sonar and tow fish positioning system, is processor with a VME bus to a Me 68000
given in the following section. processor. By also measuring temperature
and depth the APR gives an angular resolu-
tion of 0.1 ° and a slant range resolution of
0.1°. The position update rate is 2 Hz.
THE TOW-FISH POSITIONING In order to estimate positions, a reference
SYSTEM frame has to be established in the tow fish.
This is done by using the inertial system,
The tow-fish positioning system is composed the IPR. The IPR system also combines the
of two units, an inertial and an acoustic posi- surface position with the relative position
tioning system (IPR and APR). and estimates the errors in the inertial plat-
The APR is designed to measure the posi- form by the use of a Kalman filter. Figure 3
tion of the vessel in relation to the tow fish shows the functions of the system.
with either no installations on the ship or Both the heading reference unit (HRU)
only minor ones. The system is thus an and the vertical reference unit (VR U) are
inverted one, having the receiving array gimballed systems. The total IPR perfor-
mounted on the tow fish. There are two mance gives a heading accuracy of 0.5° at
modes of operation, passive and active. In 60 N and an accuracy in roll and pitch of
0
the passive mode the position is measured 0.05°. The total MINAV package is placed
by locking onto the noise generated by the in the tow fish and the IPR on-board compu-
propeller. The active mode uses a ter is a VME based Me 68000 microcompu-
frequency-coded signal transmitted from a ter.
-'-EXX·
N
~~~ ~T
~
-'--EVy·
N
I I
I I
(J¥-'.J. I
N.E.O I
'1"' . ... ( .'10 I
I
I
Relat i ve po silio n I
I 00 s il lon :
IL ____ _____ ____ _____ ~ I
/;
measure the sea floor with the spatial resol-
ution necessary to allow a quantitative
20m assessment of the condition of the pipeline.
To make this possible, the angular resolu-
I
tion was set at 0.5° and the range at less
I than 10 cm. The cross-track coverage is of
30m the order of 100°. This gives the operation
-------
shown in Fig. 4. The fish is towed approxi-
mately 20 m above the seabed, giving a
coverage of 60 m cross-track on the bottom.
The horizontal ensonified area in a beam
underneath the sonar is 0.20 x 0.20 m. The
sonar uses a separate transmitter.
When the fish is towed close to the bottom
the sonar may operate at a fairly high fre-
quency, thus making it possible to use smal-
ler transducers than would otherwise be
necessary. In order to operate at greater
ranges the sonar is designed to work at 500
kHz and 740 kHz, in addition to the nominal
Fig. 4. Sonar operation frequency of 1 MHZ.
Figure 5 shows a block diagram of the
Transducer
array
B-L-
THE BENIGRAPH: A SEA-FLOOR MAPPING SYSTEM 67
sonar. Each of the 256 elements has a sepa- read and store the parameters and master
rate channel receiver with a digitally con- the operation.
trolled TVG and a quadrature detector. The The survey set-up done by a surveyor is
outputs are multiplexed into the beam- completed on the ship's positioning system
former. In order to form all the 200 beams and read and stored by the DASAP. Thus,
within a range cell of less than 5 cm, the during normal operation, the Benigraph
beam-former is realized as a Chirp-Fourier operator has merely to use the topographic
transformer. The technology chosen is one display. All operator interactions in the
which uses surface acoustic waves (SAW). DASAP are menu driven. Figure 6 shows an
The output is digitized and a preliminary example of the simulated display when the
data reduction is performed before the data Benigraph is towed over a pipeline. The
are transmitted on a 3.2 Mbit/s PCM link to topographic information is shown in the
the surface processor. Here bottom detec- largest square in Fig. 6. Below this picture
tion is performed in a specially built bit-slice information about position and time are
processor. This processor works as a co- given. To the right a selection of input
processor on a VME bus to a MC 68000 pro- modes are shown. Other areas are for sys-
cessor. The MC 68000 performs geometrical tem warnings and the listing of new and cur-
corrections and tags the sonar hit-points rent input.
with absolute position and mean reflectivity.
These data are sent to the DASAP for stor-
age and presentation. CONCLUDING REMARKS
BENIGRAPH
ver sio n 1,0
BENTECH A/S
Main menu :
Inil su rvey
ISe I pa ra ml
DIsplay sea bed
Hard -:"0 PY
Covera ge
Te sts
Survey journal
Reset
EXIT
69
70 ADVANCES IN UNDERWATER TECHNOLOGY
of the sea floor in water depths beyond 500 m A knowledge of the bottom geology, there-
Additional sensors for determining atti- fore, is necessary for the prediction of
tude, geographic position, and the velo- acoustic propagation in the ocean.
city vectors of the sensor platform provide In more recent times, geologists and
the information required to produce scale- geophysicists have been using acoustic
rectified motion-compensated geographi- reflectivity measurements to characterize
cally positioned and oriented mosaics of the geology, and this is one of the principal mis-
sea floor, and corresponding seismic profiles sions of 'Seamor'. Quantitative estimates of
and 'feature metrics' of the scattered and the acoustic reflectivity of the sea floor
reflected energy, which are used for charac- using a shipboard echo-sounder were car-
terizing the bottom and sub-bottom units ried out by Breslau (1967). King (1967) used
(Hutchins et al., 1982). textural analysis for systematic regional
Before describing the system, it will be mapping of the Canadian shelf. Real-time
useful to review briefly the evolution of quantitative estimates of the acoustic reflec-
developments in quantitative acoustic sur- tivity of the deep sea floor using a near sea-
vey methods as applied to systematic litho- floor deployed 4 kHz seismic profiler have
stratigraphic mapping. The precision of these been carried out on a systematic basis by the
methods has steadily improved over the Scripps Institution of Oceanography since
years and they now provide quantitative 1974 (Tyce, 1976). Prediction of attenuation
estimates of the surface roughness, the cor- from a wedge of sediments using a sparker
relation area, and the hardness and variabil- sound source deployed on the surface of the
ity of the sea floor, in addition to stratig- sea was reported by Li and Taylor Smith
raphic and bathymetric profiles of greatly (1969) but met with little success due to
improved image quality. This information is scatter in the data. Tyce (1976, pp.
relevant to sea-floor engineering studies and 692-695), using this approach and near sea
is particularly useful in the design of cost- floor measurements of reflectivity from the
effective sampling and testing programmes Carnegie Ridge, obtained attenuation
(Hutchins, 1980, pp. 92-97; King, 1980). measurements in general agreement with
Improved sound sources with increased those obtained from core samples taken in
frequency bandwidth and multichannel the same area.
receivers now provide the means for This early work was frustrated by high
measuring acoustic attenuation, acoustic variability in the acoustic reflectivity meas-
impedance and sound speed as a function of urements. Although much of this variation
depth at a horizontal scale of a few metres may be due to lithologic changes, a great
and a vertical scale of better than a metre deal is due to motion of the sensor plat-
(Dodds, 1980; Fagot, 1983). forms, variation in elevation and orientation
of the source and receiver, focusing and
defocusing caused by large-scale topog-
BACKGROUND raphic features and scattering by small-scale
surface roughness.
A great deal of work has been done since the In general, we may consider the horizon-
early 1940s concerning the relationships tal resolution of a normally incident acoustic
between the acoustic and physical proper- pulse as being equivalent to the diameter of
ties of the sea floor. The propagation of the first Fresnel zone, the area which lies
sound in the ocean, of interest to the milit- within one quarter of a wavelength of the
ary sonar systems specialist, is strongly minimum distance of the reflecting surface
dependent upon the reflectivity, scattering to the source-receiver. High-resolution sur-
and absorption properties of the bottom veys, therefore, require that the source-
and sub-bottom to depths of several receiver system be deployed close to the sea
wavelengths at the frequencies of interest. floor (Fagot, 1983).
HIGH-RESOLUTION ACOUSTIC SEABED SURVEYS OF THE DEEP OCEAN 71
-Fo?1
.~ terminal
TTRS and routinely executes internal performance
TTRS verification checks. Diagnostic modes of
~PrinMr
operation are provided for identification and
localization of faults.
S). The electrical core of the main tow cable is
'I~t(~~~ms a single, RG8U 50 Q coaxial cable which
System
terminal __ _
[PDS
carries both power and information between
~ ~ Sub bottom the ship and towed body. Power for the
+ profIles
High , Reflectivity system elements located in the towed body
density
digital
recorder
8 DACS
(shipboard
...;,)
....IrICS
_Bathymetry is provided by a 600 V DC supply on the
ship. The Data Acquisition and Control
System (DACS) receives commands
through the IPDS and transmits them via
the downlink to the towed body where they
Ship systems
are used for control of the sensors and towed
•
In_ryst_
body systems. Data from the towed body
sensors is transmitted via the DACS uplink
where they are processed by the IPDS and
TTRS to produce the required outputs.
Analogue outputs of all tow fish sensor data
are available at the DACS analogue output
interface.
Fig. 1 System block diagram
~~~\P--f,
5'6'
Ref
\ Depressor
Instrument Tether Latch probe subsystem
compensation profiler electronics package sonnar management coble and r ing
subsystem subsystem package subsystem subsystem assembly (engaged)
package
Fig. 2 Towed body arrangement showing the sensor vehicle and depressor in their coupled
configuration
HIGH-RESOLUTION ACOUSTIC SEABED SURVEYS OF THE DEEP OCEAN 73
Fig. 3 Operating principle of the two-stage towed body. Note the hydrophone streamer for the
seismic protiler trailing behind the sensor vehicle
depressor is attached to the free end of 5000 m To recover the towed body, the main
of the main tow cable loaded onto the winch on the ship is operated to bring the
drum of the main winch, which is located on towed body assembly to the surface. The
the ship. The sensor vehicle is fitted with a tether winch now operates to heave in the
tether winch on which is spooled 100 m of tether cable, bringing the sensor vehicle up
neutrally buoyant cable, the tether cable. to the depressor where the latches automat-
The free end of the kevlar stress member of ically engage, locking the two assemblies
the tether cable is secured to the depressor. together. It is then recovered by the crane
The sensor vehicle is firmly secured to the as a single unit. The towed body suspended
depressor by a remotely operated latching by the main tow cable is shown in Fig. 4 and
arrangement so that the towed body may in its stowage position on the deck of CSS
be handled during launch and recovery Hudson in Fig. 5. General specifications for
operations as a single unit. To launch the the handling system are shown in Table 1.
towed body, it is hoisted off the deck by a
crane and lowered over the stern of the ship
into the water. Following release of the
TOWED BODY SYSTEMS
latches which secure the sensor vehicle to
the depressor, the winch operates to payout
The sensor vehicle provides a stable plat-
the tether cable until the sensor vehicle is
form for the following system elements
approximately 100 m behind the depressor,
identified in Fig. 2:
its normal position for acquiring data. With
the ship underway the main tow cable is • The sub-bottom profiler sensor systems
paid out until the towed body has reached its which include the sound-source trans-
desired operating depth. The sensor vehicle ducer; the pressure compensator; air
has slight positive buoyancy and tracks bottles for the pressure compensator
behind the depressor, effectively decoupled (not labelled); energy storage capacitors
from heave and other motions. The depres- in an oil-filled pressure-balanced pack-
sor is of the dead-weight type, and its age; high-voltage DC-DC inverter supply
depth is determined by the cable scope and for capacitor charging included in the
ship's speed. power electronics package; hydrophone
74 ADVANCES IN UNDERWATER TECHNOLOGY
TABLE 1
General specifications-handling systems
Winch: electrohydraulic
Length 3.5 m
Height 2m
Width 2.54 m
Line speed at 44 kN 1.5 mls max.
Weight with 5000 m cable 95 kN
Horsepower 150 max.
Fig. 4 Towed body being deployed from CSS
Hudson Winch power pack
Length 1.83 m
streamer (not shown) and a near-field Height 1.5 m
calibration hydrophone (not labelled). Width 2.3 m
Weight 2245 kg
• The side-scan sonar sensor system which
includes the port and starboard tranduc-
ers in oil-filled pressure-balanced enclos- * May be extended to 6000 m with change of syntatic
ures and the power amplifiers and trans- foam and additional cable
The DACS and navigation sensor pack- profiler sensor signals have dynamic ranges
ages are housed in a single pressure case. in excess of 100 dB and frequency band-
• The power electronics package which width for the seismic profiler hydrophone
consists of DC-DC inverters which pro- exceeds 8 kHz. Low noise, wide dynamic
vide regulated output voltages for the range, freedom from cross-talk, adequate
various towed-body systems, using power bandwidth, and linearity and transmission
from the 600 V DC supply on the ship over long cables without amplification are
transmitted through the same RG8U the requirements which must be met.
cable which serves the DACS. These requirements favour digital time
• The Fish Positioning Subsystem, a division multiplexing for the DACS. The
pinger, or ultra-short baseline receiver, uplink uses alternate marks inverted at 10 6
depending upon the particular configura- bits per second and achieves a bit error rate
tion, for determining the position of the of one part in a hundred million. The uplink
sensor vehicle relative to the ship or, in modems can easily be converted for future
the inverted configuration, the ship rela- use with a fibre optic cable core. The down-
tive to the sensor vehicle. link uses frequency shift keying at 13 and 18
• The Tether Management Subsystem kHz. Specifications for the DACS are given
described earlier, which consists of a in Table 2.
submersible hydraulic winch, slip rings
for electrical interfacing between the tow Tow Track Recovery System (TTRS)
vehicle systems and the tether and main
tow cables and the latching mechanism The purpose of the TTRS is to provide data
for locking the tow vehicle and depressor for mapping the returns from the side-scan
together into a single unit. sonar and sub-bottom profiler, and for com-
pensation of these returns for tow fish
motion. To accomplish this purpose, the
DESIGN CONSIDERATIONS TTRS has two functions - absolute and
relative positioning. In absolute position-
Some of the considerations involved in the ing, it outputs the tow fish position and
design of the system are discussed in the
following subsections.
TABLE 2
Data and Power Transmission Specifications for digital acquisition and
control ?ystem
Tow speed and depth performance are
major considerations. Because the size and Uplink
power of the handling system depends upon SDLC protocol 10 6 bits per second
the volume of the cable, it is necessary to Alternate marks inverted
minimize the electrical cross-section of the Bit error rate 1 in 3 x 108 bits
Word length 4 gain bits plus 12 data bits
cable since it decreases the cable density,
does not contribute to strength and adds to Side-scan sonar and seismic profiling
the cable volume. The most efficient type of 2 channels each, 16 kHz sampling rate, binary gain
cable from the point of view of information ranging for 132 dB dynamic range
carrying capacity is a coaxial cable. The
Auxiliary sensors
information carrying capacity of a particular 18 Channels, 80 Hz sampling rate, binary gain
cable varies inversely with its length ranging for 108 dB dynamic range
squared and directly with its diameter
squared for constant input signal power. Downlink
Some form of data mUltiplexing is therefore 13-18 kHZ frequency shift keyed for command and
control of tow-fish systems
required. The side-scan sonar and seismic
76 ADVANCES IN UNDERWATER TECHNOLOGY
motion dynamics information, which is pro- The shipboard components of the TTRS
cessed in the IPDS to position patches on processes the data in real time and outputs
the seabed with respect to some local or the tow fish position, attitude and motion
global datum to an accuracy consistent with dynamics to the IPDS. TTRS also displays
the position of the ship. In relative position- the data on a CRT and printer. Any of the
ing, it outputs tow fish attitude and motion data (statistical summaries where the data
dynamics information, which is processed in are dynamic) can be displayed to the
the IPDS to position successive acoustic operator on a CRT at a rate selected by the
returns with respect to each other to an operator (with a maximum 0.1 Hz update
accuracy and repeatability consistent with rate). Any of the data can also be logged on a
the acoustic sensor resolution and the printer at an operator specified rate (with a
required image quality. maximum 0.01 Hz update rate). Lastly, the
To determine the attitude and motion operator may assign any of the data, with
dynamics of the tow fish, components of the appropriate upper and lower bounds, to 8
TTRS in the tow fish measure the water analogue (±10 V) output channels. These
pressure (tow fish depth), acceleration,
angular rates and local magnetic field vec-
tors. All this data is output in the form of
analogue signals to the DACS uplink, since
the TTRS processing components are
located on the ship rather than in the tow
fish with the sensors.
To determine the position of the tow fish,
an appropriate fish positioning system and
ship navigation system are used. The fish
positioning system determines the position
of the tow fish with respect to the ship. The
ship navigation system determines the posi-
tion of the ship. TTRS uses this information,
together with information about the velocity
of the tow fish, to estimate the position of
the tow fish. The structure of the TTRS is
shown in Fig. 6.
To [PDS
and is used to calibrate each shot. The other angle of the transducers to accommodate
hydrophone is a three-meter long array of different operational conditions. The double
ten elements connected series parallel, bank of elements allows for measurement of
streamed behind the sensor vehicle. The the arrival angle of the back-scattered
analogue signals from each hydrophone are signal, if means are provided for measure-
fed into independent input channels of the ment of phase in the sensor vehicle electron-
HIGH-RESOLUTION ACOUSTIC SEABED SURVEYS OF THE DEEP OCEAN 79
47°
t 47°
46° ~ 46°
Fig. 8 Location map showing the track of CSS Hudson near the Verrill Canyon area during the
test cruise in July 1984
80 ADVANCES IN UNDERWATER TECHNOLOGY
\'l,
1.5 • ~ I • • •
~ \.
'\
E
:.r.
0
W
<!l
.. .
Z
c(
a::
. \ .I
o 1 2 3 4 5 6 7 8 kr
Fig.9 Sidescan sonogram* from the upper continental slope off Nova Scotia, near Verrill Canyon,
in 100 m water depth, collected with the Seabed II (Seamor) system. The aspect ratio is
approximately 1 : 1 and the range is 1.5 km across track. From the 0-3 km marks across the profile,
the seabed is covered with pockmarks (gas escape craters). Extending from the 3-8 km marks is a
large zone of disturbed sediments with a subparallel hummocky morphology resulting from local
slumping. The curvilinear features across the disturbed zone are interpreted as 'slide block scars' or
impressions
Analysis of these side-scan sonograms and were collected concurrently from the same
high-resolution seismic reflection profiles area of the upper slope extending from 500
showed previously unknown geological fea- to 1300 m water depth.
tures at the seabed and in the subsurface. From 200 to 600 m water depth over the
Essential to the interpretation of seabed edge of the shelf the seabed appears to be
geology is the need to obtain both side-scan covered with iceberg furrows which are
sonar and seismic reflection data from a both degraded and modified through ero-
study area, in order to understand a three- sion, and partially buried, as evidenced by
dimensional configuration of structure and the occurrence of discontinuous iceberg fur-
morphology. To illustrate this system inter- row rims. These iceberg furrows were pre-
dependence, Fig. 9 (side-scan sonogram) viously unknown from the upper slope of
and Fig. 10 (high-resolution seismic reflec- this area. Downslope, from 500 to 1100 m
tion boomer profIle) are presented. They water depth, the seabed is covered with
• This sonogram has not been processed, scale rectified or pockmarks (gas escape craters) which range
geographically positioned. in depth to 4 m and cover 5-10% of the sea-
HIGH-RESOLUTION ACOUSTIC SEABED SURVEYS OF THE DEEP OCEAN 81
ZONE -+------~------~
o
>
!::
(J
o
jj 50
>
100
. .....
. .: ,-- . ~ .-: :.;:
,
'
I I I 5 km
o 2 4
Fig. 10 High-resolution seismic reflection boomer profile* from the upper continental slope off
Nova Scotia, Canada, near Verrill Canyon, in 1000 m water depth, collected with the Seabed II
(Seamor) system. Over 200 m penetration is obtained with a layer resolution of approximately 2 m.
The seabed, to a depth of 20 m, consists of a zone of sediments characterized by incoherent
reflections. These sediments have pockmarks developed on their surface from the 0-2 km marks,
and are hummocky, reSUlting from local slumping and 'slide block scarring' from the 2-5 km marks.
The continuous coherent reflections are interpreted as representing glaciomarine sediments. The
presence of a zone of buried incoherent reflections at a sediment depth of 50 m may represent an
instability event which occurred at 50 000 years BP. These profiles will enable the time intervals for
major sedimentational and erosional events to be determined, which is of prime importance for the
safety of hydrocarbon drilling operations
bed. Pockmarks also had not been previ- sediments. On the seismic reflection
ously detected in these water depths or profiles, the disturbed sediment is up to 20 m
environments on the Canadian continental in thickness and internally characterized by
slope. The survey continued downslope incoherent reflections. It would have been
across a zone of locally slumped sediments difficult to penetrate and resolve the
(Fig. 9) from 2 to 5 km marks. These sedi- subsurface stratigraphy of these sediments
ments exhibit a subparallel hummocky with conventional 3.5 kHz profilers.
morphology with relief of up to 5 m and Of significant importance is the occurr-
occasional pockmarks developed on their ence of a buried zone of incoherent reflec-
surface. Long curvillinear depressions tions at a depth of approximately 50 m (Fig.
interpreted as 'slide block scars' occur 10), similar to the zone of disturbed, locally
across the surface of these disturbed slumped sediments at the seabed. This
• This recording has not been processed, scaled or motion buried zone is of regional occurrence and,
compensated. from information based on limited bio-
82 ADVANCES IN UNDERWATER TECHNOLOGY
cial geology related to site investigation 18. McKeown, D. L. 1975. Evaluation of the
requirements on eastern Canadian shelf. In Huntec (70) Hydrosonde Deep Tow Seismic
Offshore Site Investigation (Ed. D. A. System. Report B1-R-75-4, Bedford Insti-
Ardus). Graham & Trotman, London, pp. tute of Oceanography, Dartmouth, Nova
37--59. Scotia, Canada.
15. King, L. H. and Fader, G. B. 1976. Applica- 19. Parrott, D. R., Dodds, D. J., King, L. H.
tion of Huntec Deep Tow high resolution and Simpkin, P. G. 1980. Measurement and
seismic system to surficial and bedrock evaluation of the acoustic reflectivity of the
studies - Grand Banks of Newfoundland. seafloor. Can. J. Earth Sci. 17(6) 722-737.
Paper 76-1C, Report of Activities Part C, 20. Simpkin, P. G. .1978. Evaluation of broad-
Geological Survey of Canada. band high resolution seismic data for sea-
16. King, L. H. and Fader, G. B. In press. Wis- floor sediment classification. In Proc.
consinan glaciation of the Continental shelf Oceanology 78, Brighton, England. Society
- southeast Atlantic Canada. Geological for Underwater Technology, Birdcage
Survey of Canada. Walk, London, England.
17. Li, W. N. and Taylor Smith, D. 1969. Iden- 21. Tyee, R. C. 1976. Near bottom observations
tification of seafloor sediments using of 4 kHz acoustic reflectivity and attenua-
underway acoustics. Geophys. Prospect. 17, tion. Geophysics 41(4), 673-699.
231-247.
7
High-resolution Geophysical
Surveys for Engineering Purposes
K. P. Games, Gardline Survey Ltd, UK
85
86 ADVANCES IN UNDERWATER TECHNOLOGY
•• I I I I I I
I
.,'t
<.,,' IT
", ., J~LJ
,
\.,
JIll 1\\ ,1.1\,
I
~
!
".
3
I I I
I I
( l.
ID H I"
Z0 48 " ' VPK
AVG.I CAL ICURSOR T: \
I
SUijZ~6 YOLTS AI
\
Fig. 1 A record of the direct arrival at the first
trace of a standard streamer. Note the response I 1 1
of each of the 15 hydrophones iN :I ~ .1
~~ .;a~ l '~ I' ,lvl'\\ VG
SU",25b
CAL
VOLTS
I CURSOR T: . a4I72
Ii: 1.0iE-3
10 .0
15.0
""
Fig. 3 A geophysical representation of the difference in time between a signal reaching the start of
an array and the end of the array for various array lengths and for different target depths
o-,--_ =-
10 10
·20 ' 20
-
. ", . ",
~ B ~ ~ =
wlt\l[Nu ,",8[~
~
11;, (:111",,1
~ _ ~ ~
~ ~ ~ ~ ~ ~
,*,AV[HUM8£R (c'tc/ ~"'"
~ ~ ~
Fig. 4 Pattern response (attenuation versus Fig. 5 Pattern response as in Fig. 4 - this
wave number) for a standard streamer of group time for the mini-streamer with group length
length 12.5 m 2.5m
HIGH-RESOLUTION GEOPHYSICAL SURVEYS FOR ENGINEERING PURPOSES 89
C) < -0.
€:> <-40 ..
Fig. 6 Combined response diagram for a source at 0.6 m, and standard 12.5 m group length
streamer at 3.0 m depth. Zero reflection time is 80 ms, stacking velocity 1500 mls
complete system used both in these trials oscillation 20 ms later. The Fourier trans-
and in the site survey work to date. The only form of the pulse reveals a flat spectrum
source we have used has been a mini-sleeve from 60 to 1000 Hz.
exploder - by now well known and In principle, of course, any source can be
accepted in the industry as a reliable high- used which provides a broad-band acoustic
frequency source. It produces alms wide pulse with a very narrow pulse width.
initial pulse, followed by a very small bubble The recorder used was a modified Texas
!EX
<:::) > -I.
(2) < -I.
e:::::::> <-10.
<..,'"
Fig. 7 Combined response diagram for a source at 0.6 m, and the mini-streamer at 3.0 m depth.
Zero reflection time is 80 ms, stacking velocity 1500 mls
90 ADVANCES IN UNDERWATER TECHNOLOGY
Instruments DFS V recorder, which could below seabed), which we believe to relate to
sample simultaneously 6 channels at 0.25 m real changes in sediment type; the bubble
and 24 channels at 1.0 m. effect (seen at 100 m is not too serious and
has been fairly well removed by the proces-
sing; fInally, the pinchout seen towards the
SEA TRIALS left-hand side of the section is extremely
well resolved. The lower reflector is so
In November 1983 the fIrst mini-streamer strong that it almost certainly represents a
was ready for testing, and so a test-line was hard layer, and we can follow it right up to
run in the Central North Sea. Figure 8 the pinchout at which there is a vertical sep-
shows a section of this test-line recorded at aration between the layers of 3.0 m.
a sampling rate of 0.25 m. The recorder Moreover, the vertical displacement of
fIlter settings were 27-1024 Hz, while at the about 2 m at the left of the pinchout, which
processing stage the high-cut fIlter was set would not have been apparent if we had
at 800 Hz, as there was information present used a standard streamer, is not a fault but a
up to this frequency. There are several buried relief feature, believed to have been
interesting features which were encourag- caused by the sub-cropping hard layer. Such
ing on this fIrst trial: the sharpness of the an interpretation on seismic data alone owes
reflectors, especially at around 890 m (13 m much to the very high resolution achieved.
SPS
Fig. 8 A section from the Central North Sea using the mini-streamer with a sampling rate of
0.25 m. The high-cut filter at processing was set at 800 Hz to include all significant frequencies
present
HIGH-RESOLUTION GEOPHYSICAL SURVEYS FOR ENGINEERING PURPOSES 91
The system was then used for a more been well suppressed by the processing)
extensive series of trials in the sheltered compared with analogue data, but the main
waters of Spey Bay. In these trials we set point of interest is that we were able to
out to obtain direct comparisons between obtain such good correlation between
ultra-high resolution analogue data, very analogue and digital data.
high resolution 0.25 m mini-streamer data, This gave us confidence in identifying real
and high resolution 24 channel, 600 m events on the mini-streamer data, and sug-
streamer data. To do this the same 5 km gested that in certain cases the analogue
stretch was used for running separate lines data may have been superfluous, though in
with a sub-tow boomer, the mini-streamer general they served as both a profiling tech-
and a 600 m streamer. nique and a sort of calibration tool for the
Figure 9 shows a section taken from the digital data. Note also that the bubble effect
analogue data. Three reflectors, A, Band C, of the mini-sleeve has been effectively
can be clearly identified at depths of 8 m, removed by the processing.
18 m and 24 m below seabed respectively, Figures 11 and 12 show the same sections
before the multiple obscures any further on the mini-streamer and standard streamer
penetration. data. The horizontal scales are the same, but
Figure 10 shows the same section at note the vertical scales. There was little
approximately the same horizontal scale penetration achieved below 0.5 s, even with
taken from the mini-streamer data. Again, the 600 m streamer, but the data present
the same three reflectors are clearly identi- offer a good comparison between the two
fiable. As would be expected, we obtained systems. In particular, the reflectors A, B
penetration below the multiple (which has and C cannot be identified on the 600 m
Fig. 9 Section of a sub-tow boomer record from trials in Spey Bay. Note reflectors A, Band C,
which occur at depths of 8, 18 and 24 m below seabed
92 ADVANCES IN UNDERWATER TECHNOLOGY
w w w w W
CD CD iii iii 51
SPS
I - - 100m----i
-ri+-,Ilm,IllllJ.'r------- - - Z. Z.
~--~-Z . l 5
*-__---1._ • .••
.------z.2'
ii'--- - - - z.u
1';-- -- - z...
0 - - - -- - ".:
\--- - - - Z.o:
~---- O. I'
; -- - - - 0.13
Fig. 10 The same section as Fig. 9 but recorded on mini-streamer, sampled at 0.25 m. The
same three reflectors, A, Band C, can be clearly identified showing the very high resolution
obtainable
.. . ;;
Fig. 11 The same line as Fig. 10 obtained using the mini-streamer, but reproduced at a scale chosen
to display a larger section with penetration of 250 m. Note the two faults, and the high-amplitude
reflector which probably represents a deposit of coarse material
HIGH-RESOLUTION GEOPHYSICAL SURVEYS FOR ENGINEERING PURPOSES 93
.. ..
;;; 0;
..
;;;
Fig. 12 A section covering the same line as Fig. 11, but this time obtained using a standard
24-channel 600 m streamer. Note the same faults identifiable on both sections, but the difference in
resolution as highlighted by the absence of any real detail, particularly in the top 100 ms
streamer, and the two faults, although has justified the choice of a 12.5 m group
clearly visible on the 600 m streamer data, length for the mini-streamer. To limit the
appear in much greater detail on the mini- number of illustrations, only two of these
streamer section. sites will be shown here, but the same gen-
One more line trial was run in a totally eral conclusions and quality of results apply
different area - the results again being to all five sites. The other four sites were
very satisfactory - before it was felt that platform site surveys and ideally suited to
the system was sufficiently proven to be the resolution capability of the mini-
offered to clients on a commercial basis. streamer. These four sites were surveyed
using the mini-streamer as the sole digital
acquisition system. Again for brevity, only
SITE SURVEYS PERFORMED WITH one of these sites is described in detail.
THE MINI-STREAMER
the multiple, probably mainly due to the below seabed, which probably corres-
0.25 m sampling rate. ponded to the clay layers referred to above.
In one of the Southern North Sea sites In fact the reflector at c. 20 m could be
surveyed - which was otherwise not spec- identified on the analogue records. There
tacular geologically - the most prominent was useful data present on this section down
event was a fault at the top of the Upper to about 600 m, which is more than adequ-
Cretaceous chalk with a diffraction pattern ate for engineering investigations. By com-
at c. 700 m. At this horizon the diffraction parison, on a conventional streamer record
patterns from the edge of the fault plane of the same section the shallow reflectors
could be seen on the 1.0 m (standard could not be resolved, and the erosion sur-
streamer) section. However, on the 0.25 ms face was not as clearly delineated. However,
data these diffractions did not appear - we did achieve better penetration on this
almost certainly due to the short group section, so if we had been interested in both
length of the mini-streamer - thus the the shallow features and the deeper struc-
sharpness of the reflectors at this level was tures the ideal section would have been a
increased. Another interesting feature was combination of previously recorded data.
the degree of penetration achieved with the Such a section was produced by merging
six-fold 0.25 m data - down to 700 m, and 0.25 m data down to 500 m with the 1.0 m
in fact with some data present down to the data below 500 m. The merging parameters
end of the record at 1.0 s. In the region were determined by routine testing during
0-400 m, the mini-streamer section was processing. Hence, if dual sampling is per-
clearly superior in terms of resolution and de- formed, a final section can be produced to
tailed information obtainable. The top 300 m order with the proportion of very high
of this consisted of flat-lying Quaternary resolution data varied to suit the geology of
sands and clays with some very minor chan- the area.
nelling. Immediately below this a sequence
of Tertiary shales and sandstone was cut by
channelling which could be more clearly MINI-STREAMER ONLY SURVEYS
defined on the mini-streamer section. How-
ever, below this channelling and down to the Four platform site surveys were success-
top of the chalk very little could be seen on fully completed using the mini-streamer as
the 0.25 m data compared with the 1.0 m the sole digital system; one of these carried
data. Hence for this site the ideal use of the out in the Southern North Sea will now be
two systems would be to use the mini- presented in detail.
streamer data down to 400 m and the stan- Figure 13 is a typical east-west section
dard streamer data for the rest of the across this site, and the record highlights
record, while also using the better defi- the fact that the geology of the area is domi-
nition present around the top of the chalk at nated by a graben. This feature trends
c. 700 m from the mini-streamer section. broadly north-south through the site and at
Another rather geologically uninteresting its widest part is over 1500 m across. It is
site surveyed in the Southern North Sea interpreted as a typical example of a col-
consisted of dense sands with occasional thin lapse structure caused by salt dissolution
clay layers down to an erosion surface at (JenyonI984). Thestrongreflectoratc. 350m
about 180 m which was possibly base Terti- is the top of the Upper Cretaceous chalk.
ary. Beneath this our records showed a This chalk, particularly on its eastern side,
thick sequence of mainly discontinuous hori- dips towards the graben, but then shallows
zontal reflectors with a few depositional fea- along the edge of the graben to produce a
tures such as onlap being visible. rim syncline on many lines. On the western
On the mini-streamer record there were side of the graben the chalk is shallower
some minor reflectors, e.g. at 20 m and 70 m than to the east, due to the irregular move-
HIGH-RESOLUTION GEOPHYSICAL SURVEYS FOR ENGINEERING PURPOSES 95
Fig. 13 A section from a platform site survey in the Southern North Sea performed using the
mini-streamer as the only digital system. See text for explanation of the geology
ments of the underlying salt. This has been important erosive phase developed. This is
called the 'trapdoor effect' (Trusheim, 1960) seen as a strong sub-horizontal reflector
and is clearly demonstrated in Fig. 13. at c. 150 m cutting across underlying
The origin of the graben lies in the history deposits.
of the underlying salt movement in the area. Figures 14 to 16 are also taken from this
It is believed that, under the effect of a con- site, and exemplify some of the more inter-
siderable overburden, salt beds may be esting features seen on the mini-streamer
triggered into movement, probably by fault sections.
activity. Various stages in the growth of the Figure 14 shows a direct comparison bet-
resultant diapirs are recognized. The mor- ween the analogue and digital sections for
phology of the salt on this site is that of a the same line. It is interesting to see how
'pillow'. The updoming of the salt produces easily the sand waves can be identified on
extensional faulting in the overlying strata, the digital data: the maximum height of
particularly the brittle chalk. According to these sand waves is about 4 m.
J enyon, the resulting fracture system will Below the seabed on this site we see some
allow access of ground water to the salt, strongly channelled sequences to depths of
thus leading to its solution. This in turn 30-40 m, and in places up to 60 m. The base
causes collapse of the overlying strata into of such channels is well defined and consis-
the incipient void, thus forming the graben tently outlined, even where it passes below
seen in Fig. 13. The graben remained active the water bottom multiple. This is good evi-
well into Tertiary times. dence that the weathering has been essen-
Sometime during the mid-Tertiary, an tially mechanical rather than chemical.
96 ADVANCES IN UNDERWATER TECHNOLOGY
Fig. 14 A comparison of analogue and digital sections from the same site as Fig. 13. Note the
definition of the sand waves on the digital section, and the channelling in the centre and on the
right-hand side of the sections just below seabed.
Furthermore, the infill is structureless with amplitude reflectors which are clearly phase
no evidence of slumping. The inference is a reversed compared with the seabed pulse:
homogeneous sand-fIlled channel, cut into an example is given in Fig. 15 where the
firm clay or perhaps a dense granular sedi- bright spot can be identified on both
ment. It can be seen from Fig. 14 that such analogue and digital lines. In the environ-
features can be clearly followed on both ment visualized they are almost certainly
analogue and digital lines. The latter often caused by organic-rich deposits. This may
shows the channels to be associated with mean anything from a small organic frac-
very strong energy scattering and a velocity tion, through peat beds, to tiny gas pockets,
anomaly indicating lower velocities within although peaty clay or peat is the most
the channels. Since there is little evidence to likely.
suggest organic deposits, the alternative is There is a strong correlation between
an interpretation of loose gravel and these peat deposits and acoustic masking
pebble-bearing granular sediment. This associated with the graben axis. This mask-
would then contrast with a more compact ing appears to originate at very shallow
clay or dense sand in the channel sides. levels, and it may be that continued subsi-
Along the axis of the graben is a zone of dence of the graben could have determined
'bright' spots. These are local high- the site of low-lying marshy ground suitable
HIGH-RESOLUTION GEOPHYSICAL SURVEYS FOR ENGINEERING PURPOSES 97
Fig. 16 A section from the same site as Fig. 13 showing the presence of a Danian outlier in the rim
syncline of the chalk
98 ADVANCES IN UNDERWATER TECHNOLOGY
SUMMARY INTRODUCTION
1. Geophysical, geotechnical and petrophys- The development of the Troll Field in the
ical data have been synthesized into a Norwegian Trench, Block 31/2 (see Fig. 1),
workable geological history of the Troll would require the installation of a produc-
Field, Block 31/2 area. tion platform at a water depth in excess of
2. Well-marked soil units exist within the 300 m with associated templates and flow-
study area with no large differences in lines. Such a development would require
lithology or soil properties. The integra- extensive knowledge of the sub-seabed soils
tion of geophysical and geotechnical data in order to provide reliable predictions of
permits lateral interpolation of soil type. geotechnical parameters for design and
This enables geotechnical predictions to development.
be made within the resolution of the With this objective, a study integrating
database. high-resolution seismic soil investigations
3. Seabed pockmarks have been mapped and geological data was carried out. Several
and analysed and a procedure to quantify sources of geophysical and geotechnical data
their importance in engineering terms are available in the area, but this paper
has been proposed. This procedure will deals primarily with data obtained from the
express 'pockmarks' in terms of intensity seismic survey and soil investigations
versus age and depth, and pockmark undertaken for the purposes of this study.
return period. This will express pock- By using engineering seismic and soil bor-
mark risk as design environmental forces ing correlation techniques the integration of
for incorporation into design procedures. both geophysical data and soil investigation
99
100 ADVANCES IN UNDERWATER TECHNOLOGY
,w,
,,,,
6tO
f-t
)0 6°
NORTH
SO .0
----r-t-- j-- j --
SEA
3° 2
~oe
,0 0
~T"o
,0 JO .0 5° 6 1° SO 9 '0 0 'e'
59° p
•• " 1$
.
57° A be rdeen
It ,. "
SCOTlANO
,.
" .. .
"
WEST
GERMANY
I
, \
'."veSTIGATED BLOCK
Fig. 1
AN INTEGRATED APPROACH TO THE INVESTIGATION OF NEW DEVELOPMENT AREAS 101
results has been accomplished. Further channels of data were recorded through a
integration with available geological data 10 m group length/interval streamer onto a
and results from earlier seismic and soil sur- DFS-V recorder with a sampling rate of 0.25
veys has resulted in a workable geological m/s and analogue peripherals (see data sam-
model for the identification of seabed and ple, Fig. 2). In addition, a second pass over
sub-seabed soil lithologies, together with the grid was performed using a Huntec deep
the soil properties required for general tow boomer (see data sample, Fig. 7) and a
foundation engineering. In this paper a side-scan sonar.
description of soil conditions over the area Horizontal control was achieved with a
surveyed is presented. Furthermore, zones Syledis integrated with a Simrad HPR
with potential problems are identified, map- acoustic system for both the side-scan sonar
ped and discussed. and Huntec tow fishes. Data obtained from
The objectives of the study were to inte- rig-site surveys and a 1981 survey by Gard-
grate the available geophysical, geotechni- line Surveys (MSE) have also been included
cal and geological data in order to determine in the study.
the composition, thickness, lateral extent The objectives of the high-resolution
and geotechnical conditions of the sub- engineering seismic survey were:
seabed soils, and therefore to aid in the
selection of optimum foundation parameters (a) to provide site-specific geophysical
by the accurate prediction of soil proflles. information at the selected platform
Such predictions should be possible not only location;
at the specific location of the platform site (b) to provide a regional reconnaissance and
but also at any development locations within appraisal of seabed and sub-seabed soil
the Troll Field covered by the data grid. distribution over the likely development
area in the Troll field;
(c) to provide an optimum selection of site
GEOPHYSICAL, GEOTECHNICAL investigations for soil calibration;
AND GEOLOGICAL DATA (d) to provide information on problems and
obstructions within the area;
Geophysical Data (e) to provide a geophysical data-base out-
lining major physical features and envi-
Geophysical data were obtained using the ronments, major soil types and their
mini-sleeve exploder seismic source. Eight distribution;
Fig. 2
lO2 ADVANCES IN UNDERWATER TECHNOLOGY
(f) subsequently to integrate the results and in situ tests in order to determine
of the soil investigation so that the the engineering properties of the soils;
data-base would provide detailed infor- (d) to use the soil data obtained to calibrate
mation for use in br·th the conceptual the detailed engineering seismic data;
and design phases of seabed production (e) to investigate the features of geotechni-
facilities and flowlines. cal and geological significance identified
from the seismic survey, such as pock-
marks and seismic 'masking';
Geotechnical Data (f) to compute all the properties and pro-
vide reliable characteristic geotechnical
The 'primary' soil investigation was under- properties for the soil units identified.
taken by Fugro in the summer of 1983 from
the drilling vessel MV Bucentaur. Data from previous exploratory soil borings
The investigation comprised five borings completed in 1981 were also included in the
in the site-specific area and a further four study.
borings in the general area at sites selected
from the seismic data. The depth attained
varied from 29 to 149 m below the mudline. Geological Data
Samples were recovered after each boring,
and restricted laboratory testing, X-ray The available petrophysical log data can be
analysis and geological logging was carried divided into:
out on board. Extensive laboratory work
(1) data obtained in two soil borings during
was performed later by NGI and Fugro on the primary geotechnical survey;
land. At selected intervals within the
(2) data obtained when exploration wells
boreholes, quasi-static cone penetration
tests were undertaken, along with borehole were drilled in the area.
vane tests, pressuremeter tests and In two of the soil borings, SB3 and GBA,
ambient pressure sampling. In addition, downhole logging was undertaken. The
continuous quasi-static cone penetration items measured were:
tests and vane tests were made from the
mud line in the area of the platform location, (a) gamma-radiation (gamma-ray log);
(b) hole diameter (caliper log);
and a separate borehole was used for hyd-
(c) acoustic velocity in the nearby soil
raulic fracture testing. (velocity log).
In addition, 15 continuous seabed cone
penetration tests from the mudline to All logs were run from the mudline to the
depths of up to 40 m below the seabed and maximum depth of each hole. Downhole log-
two seabed vane tests from the mudline to ging was performed in all the exploration
depths of 22 m were carried out as the site- wells. The items measured were:
specific area. (a) gamma-radiation;
The objectives of the final 'primary' soil (b) spontaneous potential;
investigations can be summarized as follows: (c) resistivity;
(d) velocity.
(a) to obtain site-specific geotechnical
information at the platform location; The exploration well logs commenced at the
(b) to obtain regional information on soil end of the 30 in casing; therefore they were
conditions at selected locations before only useful for the correlation of lower soil
exaIDlmng the stratigraphy and units in the area.
geotechnical properties of the soils in The geological data consisted of analyses
these locations; undertaken on samples obtained from the
(c) to analyse the results from laboratory soil borings at the site-specific platform
AN INTEGRATED APPROACH TO THE INVESTIGATION OF NEW DEVELOPMENT AREAS 103
location. Unfortunately, the testing prog- maximum depth below sea-level of 700 m in
ramme was restricted to those samples not the Skagerrak.
required for further geotechnical analysis. The origin of the Norwegian Trench has
This imposed constraints on both sample been a subject for much discussion. Accord-
size and sub-seabed depth. ing to Rokoengen and R~nningsland (1983)
The samples were analysed for: the formation of the Norwegian Trench was
(1) dinoflagellate cysts and foraminifera by probably initiated by tectonic movements
the British Geological Survey, combined with fluvial errosion in the
Keyworth, United Kingdom; Pliocene. During the Quarternary period,
(2) carbon-14 dating by the Low Level glacial processes modified and accentuated
Measurement Laboratory, Harwell, the channel-}ike morphology. However,
United Kingdom; according to van Weering (1983), Sellevol
(3) amino acid dating by the University Col- and Sundvor (1973) and Floden (1972) the
lege of Wales, Aberystwyth, United Trench was primarily formed by glacial ero-
Kingdom. sion during the Quaternary ice ages and tec-
tonics were of little importance.
Geological input from earlier investigations Within the near-seabed strata of the
has also been included in the stUdy. Norwegian Trench area described in the lit-
erature there is a well-marked angular
unconformity. There is, however, some con-
flict within the literature concerning the age
GEOLOGICAL REGIONAL OVERVIEW of this unconformity. According to van
Weering (1983), Floden (1972) and others it
Introduction marks the base of the Quaternary sequence,
but Rokoengen and R~nningsland (1983)
In this section a brief overview of the have described it as being of mid-Pliocene
geological history and Quaternary sedimen- age.
tation of the near-seabed strata in the Nor- Since this paper primarily concerns near-
wegian Trench is presented. There are some seabed soils, this unconformity marks the
discrepancies within the literature concern- depth at which the strata are discussed in
ing these topics and these will be discussed detail. The depth of the unconformity varies
in the following paragraphs. from a few tens of metres sub-seabed near
Emphasis has been placed on the litera- the coast and western flank of the trench to
ture available on the Norwegian Trench 220--240 m in the centre of the Norwegian
area. However, considerably more data are Trench off Bergen and about 500 m at the
available on the English part of the North shelf edge (Sellevol and Sundvor, 1973; van
Sea (see Caston, 1977; Eden et al., 1978; Weering, 1983). Below the unconformity the
Fannin, 1979, etc.). It is of interest that strata dips gently towards the west with an
good correlation is possible between the angle of dip of less than 50. The stratification
English and Norwegian stratigraphy, of the Quaternary sediments clearly indi-
despite the considerable distances separat- cates two main directions of transport, one
ing the areas. The reader's attention is across the trench and another parallel to the
drawn to Fig. 3 which is a time stratigraphic Trench. The transport across the Trench
table including the major dates and geologi- was glacial, while the transport parallel to it
cal periods and stratigraphy discussed was more likely fluvial or glaciofluvial.
below. The greater the depth below the seabed,
The Norwegian Trench forms a prominent the less information there is in the literature
morphological feature in the North Sea. on the seabed strata.
Including the extension of the Skagerrak, it In general, most authors agree on the
is about 900 km long, 80--90 km wide, with a stratigraphy of the youngest sediments;
....
~
Coml><ion
~
t.%j
~
570 ~
t.%j
Precambtian l.l
- ::q
Fig. 3 Z
o
t'"
o
Q
><!
AN INTEGRATED APPROACH TO THE INVESTIGATION OF NEW DEVELOPMENT AREAS 105
silty sandy clay, which is well laminated and Rokoengen and R0nningsland Model
was deposited around 11 000 years BP; and
(b) a lower, thicker laminated clay. An alternative interpretation of the stratig-
According to van Weering (1983), units 1 raphy is provided by Rokoengen and
and 2 cover the deepest part of the Trench, R0nningsland (1983). Their interpretation of
thinning and eventually pinching out the soils will be outlined below and differ-
towards the sides. Hovland (1981) describes ences from van Weering's interpretation
the two units, units 1 and 2, as a single unit will be mentioned. Rokoengen and
(unit 1) which he divides into sub-units, lA, R0nningsland (1983) have primarily divided
1B and Ie. Hovland's 1A corresponds to the sediments into three major units.
unit 1, and 1B and Ie correspond to unit 2 of
van Weering (1983). According to Hovland, (1) The upper unit consists of soft silty clay
unit 1 (lA + 1B + Ie) is a marine deposit overlying till-like material consisting of
consisting of soft, silty dark clays with a stiff silty clay with a variable content of
high water and organic content. The fine sand, gravel and stones. The base of this
laminae in the unit are believed to reflect unit, which the authors call level B, is a
variations in sedimentation rate and mineral prominent reflector which in some
composition. The material was supplied by places can be seen to represent a slight
winnowing from glacial drift deposits in angular unconformity between the
more shallow water and by melting of the ice well-layered sediments below and the
sheet (Holtedahl and Bjerkli, 1975). till-like deposits above. The authors
The base of unit 2 is an irregular horizon. explain this erosional character by gla-
The irregularities of this horizon suggest cial erosion and think that the horizon
that it is an erosion surface and that the fur- marks the base of the Quaternary sequ-
rows are the result of icebergs ploughing ence. This is based on biostratigraphical
into the sediments as they touched bottom. evidence from the Mal0Y plateau and on
These large icebergs were present during seismic character.
the deglaciation. This interpretation is The till-like deposit of Rokoengen and
agreed upon by all the authors (Belderson R0nningsland corresponds in descrip-
and Wilson, 1973; Hovland, 1981; Rise and tion to van Weering's unit 3. The overly-
Rokoengen, 1983; Rokoengen and R0nnings- ing soft silty clay unit corresponds to
land, 1983; van Weering, 1983; etc.). units 1 and 2. As mentioned below
under 'Depositional History', the base is
not thought to be of Quaternary age by
Unit 3
other authors.
(2) Their middle unit is subdivided into:
Unit 3 is characterized by the absence of (a) An upper sub-unit of sub-horizontal
internal stratification and by hyperbolic layered sediments. The individual
internal reflections (thought to be boulders). layers can be followed over large
The base of this unit is marked by a strong areas. This unit is considered as
horizontal reflector. Most authors describe having been deposited primarily in a
this unit as a glacial till consisting of low-energy environment below
homogeneous stiff grey clay with some chalk wave base and is predominantly
and flint fragments (Hovland, 1981). clay. The well-developed layering,
however, is thought to indicate
Unit .4 sandy layers.
(b) A basal sand layer. The top of the
Unit 4 has a similar seismic character to unit unit is often marked by a very
3. No further information about this unit is strong reflector of varying inten-
available. sity, which Rokoengen and
AN INTEGRATED APPROACH TO THE INVESTIGATION OF NEW DEVELOPMENT AREAS 107
• Very shallow marine (?), cold environ- is likely that channel formation would occur
ment deposition. under such conditions.
The lack of strong reflectors in the lower
part of the unit is likely to be the result of
higher sedimentation rates during and
GEOLOGICAL HISTORY immediately after the retreat of the ice-
sheet. At this time much of the North Sea
The available data are here developed into a was still dry land and meltwater flowed into
tentative geological depositional history on a the deeper Norwegian Trench carrying
unit-by-unit basis. large amounts of sediment. In addition,
material was also provided by winnowing of
Unit I deposits by marine erosion reSUlting from
the ever-increasing sea-level.
I t is suggested that deposition of this unit of I t is difficult to date precisely the onset of
marine sediments commenced in Late the glacial retreat due to the large geog-
Weichselian times, immediately after the raphical area which the ice covered. The lit-
deglaciation of the area. During deglacia- erature suggests it is likely to have started
tion, the ice front retreated eastward and as around 18-15 000 years BP on the western
it did so the water depths in the area gradu- side of the Trench and so it must be assumed
ally increased. As it melted, the ice appar- that unit I sedimentation commenced in the
ently did not form large buoyant ice-sheets. Troll Field after this date. In conflict, how-
Sediments deposited from buoyant ice- ever, amino acid geochronology provides a
sheets are characterized by a fining upwards slightly older date for the lower sediments
grain-size sequence, with an abrupt change of the unit of 30000 years BP. However,
to predominantly fine material only, when carbon-14 dating resulted in a date of around
all the ice in the area had melted. Since no 40 000 years BP, but this is most likely due
characteristic grain-size changes have been to reworking of the samples.
noted in unit I, it is thought that the ice- The two internal reflectors within the unit
sheet lay directly on unit II as it retreated, (IA and IB) which correlate with geotechni-
i.e. a tidewater front, bordering an expand- cal variations are tentatively attributed to
ing sea. the younger and older Dryas stadials
Analyses of dinoflagellates and forams (periods of colder climate). Using this agel
within unit I shows that the unit was depo- depth of burial relationship and others from
sited in a climate which was becoming the literature and previous surveys, a rela-
warmer, as would be expected after the last tively constant rate of sedimentation is pre-
glaciation. Sedimentation was relatively dicated throughout unit I to approximately
constant up to 8000 years BP. 1.5 m below seabed. A slight increase is
Such a depositional model would suggest apparent in the lower part of the unit, which
that the sedimentary layers of unit I should agrees with the suggested higher sedimen-
progressively onlap onto the top of the unit tation rate immediately after ice retreat.
II as the ice retreated eastwards. There is The 1.5 m below seabed depth is thought
only weak evidence of onlap from the data to represent c. 8000 years BP when the Engl-
but the E-W lines which would most clearly ish Channel opened and the onset of signific-
show such features are of limited lateral ant Atlantic influences began. Sedimenta-
extent. A very rapid ice retreat would obs- tion rate was then severely reduced, which
cure onlap features, but this is unlikely since may have produced higher compaction of the
this would produce large amounts of melt- sediments and thus the higher measured in
water and thus erosion and there is no evi- situ CPT values.
dence of channels in the underlying unit. It should be noted that van Weering
Even if the underlying soils were frozen, it (1977) suggests that the internal layering
AN INTEGRATED APPROACH TO THE INVESTIGATION OF NEW DE~vELOPMENT AREAS 111
within the unit was caused by layers of character by other processes, such as flu-
interbedded gravel and pebbles. This is, vial-marine or wind erosion. Wave-base
however, unlikely since such variations erosion in a marine environment is unlikely
should have been detected by the site inves- to create such a flat surface and would imply
tigations. that the ice-sheet was grounded and eroded
A more plausible alternative explanation the underlying sediments. The amount of
of the layering relies upon minor differences sediment removed by this erosion is not
in consolidation, which could either be known since reliable dates are not available
related to times of non-deposition (unlikely for the underlying soil units. Van Weering
in the geological framework) or to differ- et al. (1973) suggest the erosion was by the
ences in mineral content. It is very likely ice-sheet which also was responsible for the
that the layering is the results of a combina- deposition of the underlying unit III. How-
tion of these event. ever, this study cannot confirm or refute
The boulders within the unit were most this hypothesis.
likely deposited from drifting icebergs. It is Examination of dinoflagellates and forams
of interest to note that the quantity of boul- within this unit suggests a harsh, arctic-like
ders does not, as would be expected, environment. Poor floras are found which
decrease from the bottom of the unit are due to unfavourable environmental con-
upwards. The tentative distribution reveals ditions such as might be expected during
a maximum at around 7 m depth and a periods of glaciation. Periods of some slight
sub-maximum at 18 m depth. These varia- climate amelioration are indicated which are
tions in boulder distribution reflect glacial thought to represent glacial interstadials.
travel paths. Long glacial travel paths Unit II is thought to consist of a series of
abrade material into clay-size particles tills deposited by the same ice-sheet as that
(Boulton, 1972). Concomitantly, the travel responsible for the flat erosional base. The
path of ice-sheets during both the Dryas study suggests this was most likely to be the
stadials (colder periods) was far less, result- last major Weichselian glaciaton whose max-
ing in more boulder-grade material in the imum occurred c. 20 000 years BP. The flat
icebergs. The depths of boulder distribution base of unit II is thought to be a combination
maxima correlate with these stadials. It of the effects of compaction and erosion,
should be noted, however, that not all the probably from a dry-based glacier. The
hyperbolae on the seismic records are areas which show the absence of the reflec-
necessarily caused by boulders but may be tor at the base can be explained by freezing
due to other strong point reflectors, such as of the ground below the glacier to the ice-
patches of organic-rich material. sheet while it was moving.
U nit I is thought to represent glaciomarine Although the outer edge of the ice-sheet is
sediments deposited after the last glacial thought to have been dry based, Boulton
maximum. A low-energy marine environ- (1972) suggested that Quaternary ice-sheets
ment is likely with progressively increasing were characterized by being dry based at
water depths and temperatures. their edges but wet based in the internal
zone. Powell (1984) states that internal and
Unit II basal ice states can vary spatially and tem-
porally in glaciers.
The geological history of this unit is more As the ice-sheet moved across the area,
complex than that of unit I. The internal it subjected the soils to dry-based erosion,
reflectors are difficult to fit into a definitive followed by wet-based till deposition. It is
geological depositional framework. likely that the lowest sub-unit of unit II is a
The very flat character of the base of this lodgement till, although its compaction and
unit is likely to be the result of glacial ero- strength is lower than would be expected.
sion, since it is difficult to explain this This can be explained by high pore-pressure
WEICHSELIAN? OR SAALlAN?
liCE RETREAT r GLACIAL ADVANCE
5. X K • x • • • *
• •M •• .... * • • •* x• •• •• •• •~ •~ • • • •• • x x
... • k • • • • • EROSION OF TOP UN I T rsr
DEPO'SITION BASE UNIT m
AS LODGEMENT TILL
CHANNELS IN TOP UNIT 17
FROM MELT WATERS
FAULTING IN UNIT nz
DUE TO OVERBURDEN
1SEALEVEL INCREASE
3. .. , - .. ..
TERRESTRIAL OR SHALLOW
MARINE EROSION OF
UNIT '2:
SEALEVEL DECREASEl
2. QUATERNARY (OR MIDPLIOCENE
MSL ~ TRANGRESSION)
DEPOSITION OF UNIT ll:
- SANDS INSHORE
- LAYERED SANDS/CLAYS
OFFSHORE
i SEALEVEL INCREASE 1
1. ~ EROSION OF FOLDED/
FAULTED TERTIARY (?)
STRATA
Fig. 6
9
DEPOSITION OF UNIT I
UNIT I <> GLACIOMARINE SEDIMENTS
UNIT n C.A. 8000 YEARS 8 P
UN IT m - -- - - - - - - - - - - - - - - - - -- -- - SEDIMENTATION REDUCED
UPON ESTA8L1SHMENT
UNITTIl - - - - - -- - -- - - -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - OF CONDITIONS SIMILAR
TO MODERN DAY
7.
,. :I( )( x )I. :III: x=-R x )C * * $I )(
x x: GLACIAL ADVANCE
C.A 20000 YEARS B.P
DRY BASED MARGINAL
UNITm ZONE RESULTS IN
PLANAR EROSION
~~~~~~~~~~~~~~~ REMOVAL OF UNKNOWN
VOLUME OF OLDER
SEDIMENTS DEPOSITION
BASE UNIT n AS
LODGEMENT TILL
6.
UNITm GLACIAL RETREAT
DEPOSITION OF UPPER
...... _' .... -"~ TILLS OF UNIT m
.:.;.:.:.:..:. . :. .:.... .... . - :;7--- (AND POSSIBLY OTHER
SEDIMENTS)
114 ADVANCES IN UNDERWATER TECHNOLOGY
values within the till which would support undifferentiated Middle to Late Pleistocene
part of the overburden ice weight. The age is likely.
irregular top relief of this sub-unit can be As with unit II, examination of dinoflagel-
explained by irregularities at the base of the lates and forams within this unit suggest a
glacier and limited erosion by meltwaters harsh, arctic-like environment with poor
running down the glacier. floras due to unfavourable environmental
The upper two sub-units may both have conditions. There are indications of periods
been deposited during the subsequent abla- of some slight climatic amelioration which
tion of the ice-sheet. The middle sub-unit is are likely to represent minor glacial inters-
thought to represent a basal melt-out till tadials.
incorporating englacial sediments which The base of this unit is not marked by a
were deposited below the ice-sheet. flat sub-horizontal reflector but is an irregu-
Limited erosion by meltwaters must have lar erosional surface. It forms a well-marked
occurred and may be reflected by the angular unconformity with the underlying
irregularities at the top of the sub-unit. The sub-horizontal layered unit IV. The base of
upper sub-unit was then deposited as the ice the unit also shows evidence of channel
front retreated and is thought to indicate a forms cutting into the underlying unit.
dropstone diamicton made up of ice-rafted There is very little evidence of disturbance
and suspended sediments. Boulders in the of the sediments at the top of the underlying
unit have been mapped. These are likely to unit, suggesting that the unconformity best
be glacial erratics which are often found in characterizes erosion from a wet-base ice-
subaquatic dropstone diamicton and melt- sheet with channels caused by meltwater
out tills. flowing below the glacier. Unit III then rep-
The depositional model as outlined above resents tills deposited by this ice-sheet.
is illustrated in Fig. 6, phases 7 and 8. The lower sub-unit was probably depo-
It is emphasized again that with the avail- sited as a lodgement till. Many such tills are
able data exact processes and a definite characterized by a finer grained character
geological history of the unit are difficult to than other tills, which is true of the lower
establish. The history and processes sub-unit of unit III. Due to the similarity in
described above are a 'best fit' model based seismic character between units II and III,
on the data sources available. However, it is possible that the depositional history of
from the results of this study unit II is the other two sub-units was the same as that
thought to represent a series of glacial tills for unit II, e.g. a melt-out till followed by a
deposited as a result of the last major dropstone diamicton. However, there is
Weichselian glaciation. insufficient data for a confirmation of this
hypothesis.
The depositional model of this unit is illus-
Unit III trated in Fig. 6, phases, 5 and 6.
From the data available it is thought that
Although the threefold division of the unit unit III represents glacial tills of either the
based on seismic character is similar to that Weichselian or Saalian glaciation.
of the overlying unit II, due to the complex-
ity of glacial deposits, the depositional his-
tory is not necessarily the same. Unit IV
The unit is thought to be a series of glacial
tills which may have been deposited by the As there are very few geological test data
same ice-sheet as that responsible for unit II available on these deeper soils, there
or by one from a previous glaciation. Unfor- remains considerable dispute as to the age
tunately, the limited geological testing and origin of the soils. According to
would not permit dating of this unit, but an Rokoengen and R~nningsland (1983), their
AN INTEGRATED APPROACH TO THE INVESTIGATION OF NEW DEVELOPMENT AREAS 115
level B, i.e. base unit II, marks the base of ressively deepening water. These sediments
the Quaternary and the older units III, IV are probably very similar to unit I soils (see
and V are not glacial sediments but marine Fig. 6, phase 4). They may represent off-
deposits of Late Pliocene age. This contrasts shore deposits resulting from one of the
with the interpretation of van Weering Early Pleistocene glaciations.
(1983) and others, who suggests these older
units are still part of the Quaternary sequ- Unit V
ence and are potentially glacial sediments.
This is confIrmed by the geological test data As with unit IV, there are few geological
where Dinoflagellates of Early Pleistocene test data available on these soils, and con-
age and forams likely to be of Early Pleis- siderable dispute remains regarding their
tocene age have been recorded. age and origin. However, as with unit IV,
The base of unit IV is marked by an the geological analyses suggest that these
irregular, apparently erosional surface. The soils are of Early Pleistocene age.
origin of this erosion is considered by The base of the unit is a very well marked
Rokoengen and R~nningsland to be the angular unconformity which represents
result of marine erosion and reworking due either the base of the Quaternary soils or
to the marine transgression. The further (according to Rokoengen and R~nningsland,
transgression then caused a gradual infIlling 1983) the level A unconformity of mid-
of depressions by the well-layered sedi- Pliocene age. The sub-seabed depth of
ments of unit IV. These sediments are Rokoengen and R~nningsland level A corre-
thought to represent a period of relatively lates very well with the depths as deter-
quiet marine sedimentation in the N or- mined in the primary survey area. The
wegian Trench area, while deltaic sedimen- unconformity is thought to be formed by
tation was restricted to areas close to the fluvial and marine erosion of Tertiary rocks
Norwegian coast. These authors assume the followed by differential structural subsi-
sea-level rise was due to tectonic move- dence along the coast.
ments. However, if the soils are of Quater- The sediments immediately above the
nary age it may be due to melting ice-sheets. unconformity have been identifIed by
It is of interest to note that the dinoflagel- Rokoengen and R~nningsland as a basal
late and foram analyses indicate an amelior- sand unit. This agrees very well with the
ated climate, warmer than the present day seismic character of the unit and the
within this unit but becoming colder down increase in sand content apparent from the
the unit. field grain-size distribution results from the
The soils of unit IV contain a high percen- primary survey. These sandy soils are
tage of clay, indicative of a quiet marine thought to have been deposited in either
environment. According to palynological fluvial or shallow marine environments.
data, the maximum depth below sea-level Subsequently, an increase of sea-level,
was of the order of 50 m. The transgressive either due to tectonics or to glacial fluctua-
nature of the deposition is confIrmed by tions, resulted in the deposition of the
onlap structures within the unit. Apparently remainder of the unit, which shows an
the sea-level rise was not continuous but apparent fining upwards, becoming a sandy
fluctuated, resulting in minor onlap struc- silty clay as the water depth increases. This
tures and channel forms. is confIrmed by seismic data which show a
Minor faulting within the unit is most more continuous layering in the upper part
likely ice loading of the unit by the glaciers of the unit, indicating a lower-energy en-
responsible for the deposition of units II and vironment. Since the depositional environ-
III. ment, as indicated by geological tests, is
U nit IV is therefore thought to represent thought to be very cold, it is suggested that
glaciomarine sediments deposited in prog- this transgression was, in fact, due to sea-
116 ADVANCES IN UNDERWATER TECHNOLOGY
Geological Sciences (McQuillin and Fannin, seabed pockmarks in profile.) The average
1979; and various internal reports) and density of seabed pockmarks is 20 per km 2•
Statoil (Hovland, 1980, 1982). These inves- Apart from pockmarks at the seabed, a
tigations have been concerned with the Sco- number of buried pockmarks have also been
tian Shelf, the South Fladen area, UK sec- noted and mapped. These generally occur at
tor, North Sea, and the Norwegian Trench, the level of reflector IA. The frequency of
respectively. occurrence of these buried pockmarks is less
In the 31/2 area one of the most striking than those on the seabed. This is partly due
features is the lineation of seabed scars to a 100% coverage of layers underneath the
associated with the pockmarks. Hovland seabed using shallow seismic data. Allowing
(1981) in neighbouring areas described for this, results still differ in frequency since
orientations of NW-SE (42%), NNW-SSE on seismic sections the number of seabed
(30%) and N-S (24%). Similar percentages pockmarks is far larger than the number of
have been found valid for 31/2 pockmarks. buried pockmarks (approximately 10 : 1).
Individual pockmarks comprising these It should not be concluded from these fig-
lineations may be 8-60 m in diameter but the ures that the frequency of pockmark forma-
total length can commonly attain hundreds tion is greater today than in the past.
of metres. - Geological evidence suggests little sedimen-
The largest seabed pockmarks in the Troll tation has occurred in the area over the last
Field study area attain diameters of 100 m 8000 years, so the pockmarks seen today
and depths of 15 m. The most common depth represent many years of formation. It is
of occurrence is in the 3-5 m region with likely that pockmark formation would be
little evidence of border rims or terraces. suppressed during periods of rapid sedimen-
(Figure 7 is a boomer record showing these tation, and thus they are not found through-
TlKE 0...,.J~-
!.,IilIo...:2;;,;
OO.:..
O '""""T_""T'""--:_~_200~S_r--._r--,-_2~TIO_-r-----r_-;----r_2-0jIr-S--,
( • .sec s)
+
" ."",,: .~ .
Fig. 7
118 ADVANCES IN UNDERWATER TECHNOLOGY
out unit I in high concentrations. The levels reservoir, the implication being that the gas
at which buried pockmarks are found are has migrated through the overlying strata.
likely to represent periods of reduced The simplest case is to assume that the mig-
sedimentation. rating gas tends to follow paths of least
The erosive nature of some pockmarks is resistance upwards, i.e. faults and bedding
demonstrated by the truncation of acoustic planes.
layers on high-resolution seismic reflection A relationship between mapped pockmark
profiles. Numerous mechanisms have been lineations and faults would provide useful
suggested to account for the removal of evidence to support the above suggestion. It
what in many cases is a considerable volume is to be expected that the strike of the
of material. These include suggestions that youngest fault pattern would have the
they are caused by action from above the strongest influence on a possible lineation.
seabed (meteor impact crater, scour hollows The youngest major fault pattern has there-
around boulders, iceberg drop-stones, and fore been mapped in the Tertiary sequence
man-made artifacts such as wrecks, bombs, at the depth of a prominent reflector called
etc.), by biological activity (e.g. the action of the 'Purple Horizon'. The pattern is shown
sub-bottom or bottom-drilling creatures), on Fig. 8.
and by mechanisms which are essentially As is to be seen, the major fault plane
related to the last ice age (e.g. strudel scour, lineation is ENE to WSW. This does not
kettlemarks, both of which imply the previ- correlate with the pockmark lineation
ous existence of large bodies of ice trapped (NNW-SSE). This does not necessarily
within the sediment and collapse when the exclude fault-plane migration of petrogene
ice melted leaving a relatively easily erod- gas. However, a simple migration mechan-
able sediment and subsequently a depres-
sion). Of these mechanisms none has been
given much credence. The ice age related + 60· 54'
ism through the Quaternary sequence is not determining the distribution of annual
apparent. intensity, consideration has to be given to
The base (worst) case assumed is that the likelihood of occurrence (a) along a
pockmarks may create risks for: neighbouring fault; (b) along a given plane
(a) short-term settlement of a gravity- (or planes) of known structural weakness;
or, equally likely, (c) anywhere over the
based production platform;
(b) spanning of pipelines; area due to an apparent lack of correlation
between the geologic structure and the
(c) long-term risks (a) and (b) above due to
growth of new or enlarged existing pockmark position formation.
The data required to complete the study
pockmarks.
are:
Owing to the uncertainty of numbers,
(1) intensive geologic dating of unit I soil by
sizes and location of future pockmarks gen-
erated within the Troll Field development pollen, foraminiferal, dinoflagellate and
carbon-14 methods;
area, it is appropriate that engineers should
(2) input of field data required over a
express pockmark risk in terms of 'return
'pockmark study area', providing
periods', just as design environmental forces
further data on seabed and sub-seabed
are.
pockmarks;
It is desirable for the engineer to have
(3) the subsequent integration of data into a
available all the pertinent data and profes-
report describing average return
sional judgement of those trained in
periods.
geophysics and geology in a form most suit-
able for making a decision. Therefore it is Such studies are presently being completed
most useful to have information available as from available data for the Troll Field area.
cross-plots of pockmark intensity versus age
and depth, and pockmark intensity versus
ACKNOWLEDGEMENTS
average return period. This information can
then be used systematically in engineering The authors wish to express their apprecia-
design procedures. tion to the management of N orske Shell and
The methodology to provide an assess- to the other shareholders of Troll West,
ment of the risk from pockmarks at a site in namely Statoil, N orsk Hydro Production
these terms has to be considered. Historical A/s, Norsk Conoco A/s and Superior Oil
geological data must be synthesized. The Norge A/s for granting permission to submit
locations, sizes and activity types may be this paper.
many and different in kind or may not be
even well known. REFERENCES
A method has to be utilized to produce for
the engineer the desired relationships bet- 1. Belderson, R. H. and Wilson, J. B. 1973.
ween process, geologic history, pockmark Iceberg ploughmarks in the vicinity of the
age, size and intensity and their average Norwegian Trough. Norsk. Geol. Tidsskr.
return period for the Troll Field site. 53, 323-328.
The minimum data needed are the geolog- 2. Boulton, G. S. 1972. Modern arctic glaciers
ist's best estimates of average pockmark as depositional models for former ice sheets.
numbers with depth on a time-based plot. J. Geol. Soc.
The technique has to allow for integrating 3. Caston, V. N. D. 1977. Quaternary deposits
of the Central North Sea. 1: A new
the individual influences of pockmark isopachyte map of the Quaternary of the
sources near and far, and more or less North Sea. 2: The Quaternary deposits of
active, into a probability distribution of the Forties field, northern North Sea.
maximum annual geographic intensity. The Report No. 77/11, Natural Environment
average return period follows directly. In Research Council.
120 ADVANCES IN UNDERWATER TECHNOLOGY
4. Eden, R. A., Holmes, R. and Fannin, N. G. 17. Josenhans, H. W., King, F. and Fader G. B.
T. 1978. Quaternary deposits of the Central 1978. A side-scan sonar mosaic of pockmarks
North Sea - depositional environment of on the Scotian Shelf. Can. J. Earth Sci. 15,
offshore Quaternary deposits of the Conti- 831-840.
nental Shelf around Scotland. Report 77/15, 18. King, L. H. and MacLean, B. 1970. Pock-
Natural Environment Research Council. marks on the Scotian Shelf. Geol. Soc. Am.
5. Fannin, N. G. T. 1979. The use of regional Bull. 81, 3141-3148.
geological surveys in the North Sea and adj- 19. Mangerud, J. 1983. The glacial history of
acent areas in the recognition of offshore Norway. In Glacial Deposits in North-west
hazards. International Conference Papers. Europe (Ed. J. Ehlers). A. A. Balkema,
6. Floden, T. 1972. Notes on the bedrock of 1983.
eastern Skagerfrak with remarks on the 20. McQuillin, R., Fannin, J. and Judd, A. G.
Pleistocene deposits. In Stockholm (1979). IGS pockmark investigations
Contributions in Geology, Vol. XXIV, 1974-1978. Report 98, Natural Environ-
1971-1973. ment Research Council.
7. Floden, T. and Sellevol, M. A. 1972. Two 21. Newton, R. S., and Cunningham, S. 1980.
seismic profiles across the Norwegian D' Appolonia Consulting Engineers Inc. Mud
Channel west of Bergben. Stockholm Con- volcanoes and pockmarks: seafloor engineer-
tributions in Geology, Vol. XXIV, ing hazards or geological curiosities? Off-
1971-1973. shore Technology Conference.
8. Holtedahl, H. and Bjerkli, K. 1975. Pleis- 22. Platt, J. 1977. Significance of pockmarks for
tocene and recent sediments of the N or- engineers. Offshore Engineer.
wegian continental shelf (62°N-71 °Z) and 23. Powell, R. D. 1984. Glacimarine processes
the Norwegian channel area. Norges. Geol. and inductive lithofacies modelling of ice
Unders. 316, 241-252. shelf and tidewater glacier sediments based
9. Holtedahl, H. and Bjerkli, K. 1982. Late on Quaternary examples. Mar. Geol. 57,
Quaternary sediments and stratigraphy of 1-52.
the continental shelf off Trjlfndelag, W. 24. Rise, L. and Rokoengen, K. 1983. Surficial
Norway. Mar. Geol. 45, 179-226. sediments in the Norwegian sector of the
10. Hovland, M. 1980. Detailed seabed mapping North Sea between 60 30'N and 62°N. Mar.
0
expansion in the number of contractors vey works which can accompany oil produc-
undertaking marine geophysical site inves- tion construction and development.
tigation surveys, and in the range and com- It will be helpful first to state the two
plexity of instrumentation and hardware aims of the geophysical site investigation
available. The period has also seen a sub- survey for exploration rigs:
stantial growth in continental shelf explora-
(1) To define and delineate those features
tion, including movement into deeper water
which are a hazard to the safety and
and away from the established areas of the
integrity of any drilling exploration
USA, the North Sea, the Arabian Gulf, and
structure or vessel within a specified
South-East Asia. The commercial pressures area of operation.
on the contractor to exploit the market and
(2) To assemble and suitably present as
keep pace with the technology have proved
much geological, geotechnical, hydrog-
challenging, 'out the financial return on
raphic and environmental information as
investment js highly unpredictable, though
is practical within a certain budget, and
traditionally low. The number of survey
agreed scope; and to aid the timely, effi-
companies which have been closed down,
cient, safe siting of the rig and its subse-
endured long periods of severe financial
quent spudding and early drilling
restriction or been subject to takeover and
phases.
major restructuring, is perhaps too high to
be regarded as healthy. If the survey con-
tractor cannot expect an adequate financial
return on his investment, then the interests COMMERCIAL CONSIDERATIONS
of the equipment supplier/manufacturer and
the oil exploration company/drilling contrac- In 1970-1975 the standard survey package
tor and civil engineering contractor must for a rig-site investigation in the North Sea
inevitably also suffer when out-of-date (Fig. 1) comprised:
equipment becomes the norm and profes-
• Navigation receiver (Hifix or Mainchain).
sional practices in acquisition and interpre-
tation are limited by the contractor's own • Hydrographic echo-sounder.
commercial considerations. • Side-scan sonar.
This paper describes many of the com- • Boomer/multi-electrode sparker seismic
recording package.
mercial and technical factors which are con-
sidered by the contractor when preparing • One or two hydrographic surveyors, a
geophysicist and an electronics engineer
bids for surveys. The paper also updates and
to operate this equipment.
extends some of the points of discussion
which closed the geophysical survey section • A fairly basic vessel of 30-45 m length.
of the 1979 Site Investigation Conference. The survey contract was negotiated on a
Some comparisons are drawn between sur- daily charter with the expectation of work-
vey specifications and contract negotiation ing only during daylight. Furthermore, the
in Northern Europe and other parts of the contractor was frequently required to stand
world. The net result is an argument for a by to guide the rig onto its surveyed location
more rational and constructive approach by on completion of the investigation. This,
both the geophysical contractor and by the together with weather limitations on acquis-
oil and engineering companies and govern- ition, daylight working, plus standard day
mental institutions which direct the scope of rates recognized by the industry, gave the
work issued. contractor a respectable income. The 1985
The paper is structured around the indus- situation shows a dramatic change in hard-
try of providing geophysical site investiga- ware and operating procedures. Figure 2
tion studies for marine exploration drilling illustrates the development in hardware to
rigs. No accounting is made of similar sur- meet industry requirements. Improved
MARINE GEOPHYSICAL SITE INVESTIGATION FOR EXPLORATION DRILLING RIGS 123
BAND PASS
------,
~
FILTER
EG&G 259B
ATLAS SIDESCAN
OESO 10 SONAR
0:
SURFACE TQVoI MULTI·ELECTRODE
BOOMER SPARKER
HAND ELEMENT HYDROPHONE
WINCH
HULL OR OVERSIDE
MOUNTED
33+210KHz 200 METER CABLE SIDESCAN TQVoIFISH
TRANSDUCERS
navigation aids and larger vessels mean of short-duration site-specific surveys. The
24-hour operation with a higher threshold of fact that the oil industry in this area is
weather and sea state unsuitability. The geared to the issue of site-specific studies
survey crew required also shows an increase requiring fast turnaround does mean that it
to 9 or 10. Table 1 shows a guideline com- is also dependent on this surplus of survey
parison of capital costs for the 1975 and the vessels being maintained. As long as work-
1985 packages as illustrated, based on 1985 load remains reasonable then the contrac-
purchase prices in US dollars. The table also tors will stand by waiting for numbers to
indicates the required annual revenue which diminish or workload to increase. The end
each of the two investments should recover result of a temporary reduction in workload
to be commercially viable. From this, a must inevitably be a significant reduction in
range of percentage annual utilization fig- the number of contractors and a lack of pre-
ures are quoted and the day rate required to paredness when workload improves. The
achieve the viable revenue is calculated. ongoing effect of the way the North Sea
Below this the argument is progressed by market is structured is a severe limitation
estimates of other expenditure involved in on equipment investment and development.
the survey operation. Bid preparation, in particular the prep-
The fairly obvious end result of this finan- aration of fixed-price bids, is very demand-
cial analysis is that the contractor must ing on the contractor, requiring extensive
gamble on high utilization figures if he is to amounts of time by experienced people
compete at the rates currently prevalent in researching logistic and operating costs.
the established marine exploration areas, or The process of decision on award of contract
he must seek strategic advantages in new by the oil company should in theory be based
areas of interest. The situation in north- upon:
west Europe is of particular interest in this
context because there is perhaps a surplus (1) Price quoted.
of large, comprehensively equipped survey (2) Availability.
vessels competing for a share in a very (3) Suitability of vessel and equipment
limited market, which is composed primarily quoted.
....
~
Nov Compuler --jF~iire;-;Con=:;.lro=-1-+-___ ____--,
Fi lenumber ...
Textronill: Tope Transporls
254 ID
Oscilloscope
o
o
48ChonneiCopabilily 01 I Millisec
Ronqe of Filter.
Quol $ompli"9 Facility
Extended Header
Hull mounted
33+210KHz
Transducer Powered Kronhite
(Oplional) Coble Winch Filter lV.G
L---j--II[O.75mMoonpool I"""",,,,,,,,,
EPC3200 EPC3200
or
~ 1"'"""'''' ' Roytheon EPt4loo Generat~
FibreOplicLink 1807M
>
t:l
;;
Z
o
Marine
t.:r:l
Preformed r:n
Electri Terminotion EOO 604
0
withS~I~.Powered Winch
a Sidescan ~~egrOI~d ....
1000 met 'ng Assembly prof,ler Z
re coble c::::
Bu,It in Hydro, Z
Generotor WolerGun Sparkarroy t:l
t.:r:l
::0
TABLE 1
1975 Package (using approximate 1985 purchase values and costs)
Minimum
viable Minimum day rate revenue required based on
New annual percentage annual utilization
purchase revenue
value (note 2) 30% 40% 50% 60% 70% 80% 90%
Survey equipment as 115000 45430 415 311 249 207 178 156 138
illustrated in Fig. 1
Survey personnel and note 3 712 628 577 534 511 494 475
administrative overheads
1985 Package
Survey equipment as 1100000 434500 3968 2976 2380 1984 1700 1488 1323
illustrated in Fig. 2
Survey personnel and note 3 1603 1522 1368 1347 1175 1122 1068
administrative overheads
Notes
1. All costs and prices quoted are in US dollars
2. Minimum annual revenue calculation based on: 2.5% insurance; 12% minimum borrowing rate; 5%
net service costs: depreciation over 5 years
3. Simple calculation of personnel and administrative costs assumes 13 000 US dollars. cost per average
survey person, 50% administrative charge, 100% backup on survey personnel at 90% utilization reducing
to zero backup at 30% utilization
4. The total calculations are exclusive of vessel charges and navigation hardware and chain rental charges
(4) The quality and experience of the per- bidder. The unhealthy aspect of the busi-
sonnel offered to l.lndertake the work. ness is that the contract award may well go
(4) The previous performances of specific to the survey company which makes an
contractors. error in its costing calculations and is
awarded a contract which can only be made
When dealing with small site-specific profitable if significant shortcuts are made
studies the very high ratio of preparation in the execution of the contract.
time to acquisition time strongly favours
established vessels operating or on standby
close to the site; and this is normally SITE-SPECIFIC SURVEYS
reflected in the price quoted. An oil com-
pany requesting fixed-price bids, particu- At this stage it is worth considering
larly in an active area like the North Sea, whether the needs of the oil industry, the
must regularly be presented with situa- marine geophysical contractors and other
tions where return bids show a variation interested bodies are best served by the
between low and high bids of several present generally accepted system of con-
hundred per cent. In this situation it must tracting on a site-specific basis. The present
be very difficult not to ignore points 3, 4, 5 system requires the oil company to commit
and possibly also 2, and award to the low itself to a specific drilling location and use
126 ADVANCES IN UNDERWATER TECHNOLOGY
this as the centre of a grid survey with area on the seismic sections. It is important
and grid parameters determined by water to point out that a complete site-specific
depth and the type of drilling rig to be used. survey in parts of the North Sea could
The disadvantages of this system are be contained within the margins of a
numerous and, we think, highly significant. large infilled glacial channel and, treated
in isolation, the interpreting geophysi-
(1) The commitment of the oil company to a cist may be totally una ware of the
specific drilling site substantially in unusual nature of the site and the poten-
advance of the rig becoming available tial problems the drilling contractor may
creates problems in any exploration encounter with unusual sediment infill.
department. Normally target evaluation Similarly, site-specific analysis may
will continue right up to the arrival of identify a relatively strong seismic
the rig, as evidenced by the occurrence reflector as a useful but inconsequential
- which all survey contractors have lithostratigraphic marker horizon,
experienced - of changes of location whereas, if part of a more extensive
part way through a surveyor interpre- survey, the horizon may only demons-
tation. The normal consequence of this is trate this 'brightness' locally and the
that the oil company will delay the interpreter would then have to seek an
award of the site survey contract as long explanation which may be the presence
as it thinks possible and place the bur- of cemented material or high-pressure
den on the survey contractor, data pro- gas.
cessing centres and the interpreting (3) intensive geophysical investigation of a
geophysical team to complete every 3 km by 3 km site may only take 24-36
phase in as little time as possible. Little, hours of survey work. However, it is
if any, allowance is made for bad usually accompanied by mobilization and
weather, delays or equipment failure. transit to and from the site; calibration
The end results are frequently a limita- and perhaps mobilization of area-specific
tion of choice of survey equipment and navigation systems; general geophysical
equipment to those nearest to the loca- system testing, including such time-
tion; limitations on quality control in consuming duties as seismic-streamer
data acquisition and processing; and the cable balancing and satisfying the client
issue of a rushed report which may be company or its representatives as to the
sufficient to allay any fears held by the performance of the survey package. All
drilling rig operator or the insurance this frequently results in a preparation
company involved but provides little period which grossly exceeds the actual
constructive assistance for the spudding survey period. Whether the contract has
and early drilling phase. been awarded on a day rate or a fixed
(2) The interpretation of data from a small sum, these preparation charges - or at
area in isolation invariably restricts the least an estimate of their potential cost
geophysicist to discussion and illustra- - are included in charges made to the
tion of the obvious features of the area, client oil company.
such as seabed topography and the pres-
ence of any apparent geological uncon-
formities near to the seabed. Seismic REGIONAL SURVEYS OR SURVEYS
horizons are frequently selected and OF COMPLETE LEASE AREAS
contoured only to demonstrate the pres-
ence or lack of any structural features. The alternative to site-specific surveys is a
The analysis for drilling hazards or con- system of more extensive surveys covering
straints is primarily limited to a qualita- a complete leased area - or at least the
tive assessment of anomalous features whole of the area of potential interest - in
MARINE GEOPHYSICAL SITE INVESTIGATION FOR EXPLORATION DRILLING RIGS 127
the softer soils - are not discussed in this gin and one characterized by water depths
paper. greater than about 4000 m (Fig. 2). The
The words 'deep water' and 'deep sea' are specific deep sea floor bathymetric pro-
almost meaningless because of widely differ- vinces adjacent to the continents usually are
ing usage in oceanography and geotechnical an abyssal plain next to a passive margin
ocean engineering. In the offshore oil and and a trench next to an active margin. The
gas industry the words 'deep water' usu- industrial use of the name of the provinces
ally refer to the water depth at which explo- shown in Fig. 2 is strongly recommended
ration and exploitation are occurring within because each has a universally understood
a particular geographic region at a given water-depth range associated with it.
time. Figure 1 shows that the maximum I t is recommended that the general
deep-water depth has been generally expression 'deep water' be abandoned
increasing during the P9"t twenty years or whenever possible in favour of citing a
so. In 1965 the drilling record was almost specific water depth or depth range. Deep
200 m. In early 1985 the record is more than water in this paper signifies water depths in
ten times deeper. This depth trend is likely excess of 300 m. This is about the minimum
to increase in the future with water depths limit of the transition range of depths for
of 3000 m being predicted by the year 2000. using most conventional industrial offshore
'Deep water' in the oil and gas industry in geotechnical equipment (Table 1).
1985 generally refers to a maximum water The selection of 300 m precludes a
depth of about 300 to 1000 m, depending on description or discussion of many new and
the region. People in the nascent fascinating instrumentation designed to
polymetallic ('manganese') nodule and operate at lesser depths, such as the Dutch
poly sulphide mining industry use the term and Brazilian geotechnical diving bells
'deep water' to refer to operational water (Graaf and Smits, 1983; Hunt, 1984), new
depths of about 3000 to 5000 m. drilling, sampling and testing systems (e.g.
Academic oceanographers and ocean Okumura and Matsumato, 1981; Ardus et
engineers have very different requirements al., 1982; de Haas, 1983; Beard and Johnson,
in the 'deep sea' compared to industry. They 1984; Pheasant, 1984), mobile geotechnically
usually refer to the deep sea floor as being a instrumented crawlers (Benoit, 1983), most
province separate from the continental mar- pressuremeters, and the dilatometers. On
the other hand, innovative equipment pres-
ently operated in shallow water will be
CONTINENTAL MARGIN
briefly included if it serves to illustrate what
DEEP SEA FLOOR
,
CONTINENTAL I CONTINENTAL:
,
CONTINENTAL can or should be done in water depths grea-
ABYSSAL I OTHER
r-====~~'----~'------~~--~:
,
Emphasis has been placed on new
developments in water depths greater than
300 m in this paper. But it is very important
to keep firmly in mind that there are a sub-
stantial number of older tools and tech-
niques that are used offshore in water
"--------' ,'- - -
I depths greater than 300 m by the geotechni-
I
cal consulting companies. This use will be
TRENCH
I
I
I
discussed later.
I
A number of useful reviews of offshore
sampling and in situ testing equipment have
Fig. 2 Bathymetric provinces of the continen- been made during the past few years
tal margin and adjacent deep sea floor. The (Andresen, 1981; The Sub-Committee on
ater depths given are the usual averages Soil Sampling, 1981; Vollset and Gunleik-
132 ADVANCES IN UNDERWATER TECHNOLOGY
GUI D E:
ROD
~
DRIVE CHAIN
A B
Fig. 3 Two basic types of wheel-drives to obtain continuous penetration of test rods in a seabed
penetration apparatus. A, friction generated by elastic deformation of the material pressed against
the rods. B, friction generated by slight plastic deformation of the test rods driven by smooth or
serrated drive wheels
fully adapted by McClelland for offshore rod replacement if substantial drive forces
operation. The second type of drive-wheel are to be developed. The drive-wheel can be
mechanism drives the test rods by the fric- smooth or serrated. This type of drive sys-
tion of continuously moving pads pressed tem is used by Borros for land penetration
against the test rods (Fig. 3A). The pads are equipment, for the offshore Van den Berg
composed of an elastic material such as ROSON, and for Fugro's offshore DORA
neoprene or synthetic rubber. A 'Catapuller' and Sea Lion that will be subsequently
drive system of this type was devised by N. described. There are now so many different
D. Remy & Associates for the Lehigh U ni- types of deployment methods used for in
versity Geotechnical In situ Probe System situ testing offshore in water depths greater
(GIPSY) that was designed and constructed than 300 m that it was deemed useful to
in the 1970s and early 1980s. Laboratory summarize most of them (Table 2).
tests showed that the type of neoprene used People who are not intimately connected
in the pressure pads, which in this machine with the industrial offshore geotechnical
pressed against a hollow square test rod, profession frequently lose sight of the fact
was critical. Material that was too soft was that sampling and in situ testing equipment
deformed plastically or stripped off the is usually developed to meet the specific
pads. This type of drive does not mark the needs of clients, particularly in the oil and
test rod. The third type depends upon gas industry. It is not generally appreciated
developing friction by a very small plastic that the oil and gas industry will use almost
deformation caused by the wheel pushing any new piece of equipment once, but if it
against the test rod (Fig. 3B), which tends does not provide useful data or is difficult to
to mark the rod with flat faces or teeth- use at sea it may rarely be used again. This
marks and generally requires more frequent provides a useful filter for the retention of
134 ADVANCES IN UNDERWATER TECHNOLOGY
TABLE 2
Examples of deployment methods used for in situ testing
System and subsystems Example References
Downhole
Carrier tool, probe, sensor(s) Various Richards and Zuidberg (in press, b);
Zuidberg et al. (in press); this paper
Seabed
Penetration apparatus
Cable type
mUltiple Seacalf Zuidberg (1975)
Stingray Ferguson et al. (1977)
single electromechanical Tower Richards et al. (1972)
single mechanical; acoustic telemetry DORA This paper
single mechanical; seabed recording Sea Lion This paper
no cable, seabed recording ISV Babb and Silva (1983)
ISHTE Bennett et al. (1983a)
Submersible
Manned Inderbitzen and Simpson (1972); Richards
(1972); Lambert (1982); Geise and Kolk
(1983)
aNot designed for water depths greater than 300 m, but probably capable of operation at such water
depths with modifications
IN SITU TESTING AND SAMPLING OFFSHORE 135
TABLE 3
Sampling and in situ testing equipment most commonly used by the
offshore geotechnical industry
Examples a
Fugro McClelland
1. Sampling equipment
A. Downhole
1. Hammer or percussion sampler x x
2. Push sampler Wipsampler Stingray
3. Piston sampler Hydraulic Dolphin
piston sampler
B. Seabed
1. Drop or gravity corer x x
2. Piston corer x x
3. Vibrocorer and vibrohammer corer x x
B. Seabed
1. Cone penetrometer Seacalf Stingray
2. Piezocone penetrometer Seacalf Stingray
"Water and penetration depth compilations of various offshore equipment have been
published by Richards (1982), Vollset and Gunleiksrud (1982) and Briaud and Meyer
(1983)
x, unnamed equipment in common use
useful, easy to use, economical and success- son et al., 1980; Fyffe et al., 1982; and
ful equipment and the rejection of equip- others), as do self-boring machines such as
ment which is not useful, difficult to use, the PAM, which will be mentioned later.
uneconomical (prone to failures) and unsuc- Many shallow-water self-boring pres-
cessful. The equipment listed in Table 3 has suremeters, of course, are in use by indus-
survived the test of time and has gained the try (e.g. Hughes et al., 1984).
acceptance by the oil and gas industry for In Table 3 and elsewhere in this paper,
common use offshore. It will be noted that the word 'downhole' refers to sampling or in
this list is purposely restrictive and does not situ testing equipment that is lowered by
include the pressuremeter, which is used wireline or umbilical through the drill pipe to
only 2-3% of the available time by industry the bottom of the drill string. The word 'sea-
in water depths greater than about 30 m bed' refers to any equipment lowered
(Richards and Zuidberg, in press, a), or the through the water column to rest on or in
seabed vane. But push-in downhole pres- the seabed. Examples of the latter include
suremeters capable of operation in water seabed jacking or penetration machines,
depths greater than 300 m do exist (Hender- conventional piston corers and so forth.
136 ADVANCES IN UNDERWATER TECHNOLOGY
6::-'
1200 TONS
111<1
30 m WATER
2.30 TONS 61 mWATER
1955 1520 TONS
1959 87 m WATER
5000 TONS
1965
lOt mWATER
6510 TONS 114. m WATER
1976 1000 TONS
1970
259 mWATEA
19000 TONS 285 m WATER
1916 312 m WATER 39100 TONS
305 mWATER
HONDO 59000 TONS 1981
1978 08000 TONS
CEAVEZA
COGNAC 1983
Fig. 4 Significant offshore platforms considered to be 'deep water' in the year listed that have been
installed in the Gulf of Mexico and offshore of California. (Modified from a Shell Oil Company
illustration)
Details are given by Richards and Zuidberg over almost forty years in the Gulf of Mexico
(in press, b). and offshore of California is shown in Fig. 4.
The reason for the restricted number of These structures require piles. The
samplers, probes and sensors used in the geotechnical site investigations to deter-
offshore industry is related to the nature of mine vertical and lateral pile capacities are
the foundations to be investigated. The made by drilling and downhole sampling,
development of significant jacket structures which has been augmented by cone penet-
MAGNUS
NINIAN CENTRAL
!~.
ITEMPLA TE)
12 PILE
it!I:
2500 TONS 67 m WATER 67 m WAT~R
(1966) (TEMPLATE) (GRAVITY) _
12 PILE STORAGE '
6000 TONS 225,000 TONS I
(1972) (1973)
Fig. 5 Significant offshore platforms considered to be 'deep water' in the year listed that have been
installed in the North Sea. (Modified from a Shell Oil Company illustration)
IN SITU TESTING AND SAMPLING OFFSHORE 137
daries, there has been a spate of papers forces of concern to the geotechnical
dealing with the geotechnical problems of engineer. A later section will discuss new
the active or convergent continental mar- instrumentation to investigate shear waves.
gins or trenches (Von Huene, 1984). These The general area of activity concerns the
problems include consolidation and deforma- formation of a (Marine) Geotechnical Con-
tion (Carson et al., 1982), venting of pore sortium to investigate in detail selected
waters from subducted soils with attendant International Project of Ocean Drilling
methane-derived carbonate cementation (IPOD) cores up to 200 m long from the deep
(Carson et al., 1984; Kulm et al., 1984; Suess sea floor in different oceans. To date, the
and Massoth, 1984) and relating the state of Walvis Ridge in the South Atlantic Ocean
overconsolidation to the rate of convergence has been studied (Geotechnical Consortium,
(Shepard and Bryant, 1983). 1984), the investigations of the North-West
The second set of problems has to do with Pacific core have been completed (The
the recent and unexpected discovery of Marine Geotechnical Consortium, in press),
extensive and, presumably, major faulting and the cores collected from the Mississippi
of soils that does not appear to be related to Fan in the Gulf of Mexico are presently
basement topography in two abyssal plains being analysed. These studies in particular
of the North Atlantic Ocean (Duin et al., have yielded much information about soils
1984; Schiittenhelm et al., in press). It is that are apparently 'over consolidated' near
likely that a geotechnical process has caused the seabed and 'underconsolidated' at grea-
the normal faulting, although it is not yet ter depths. A recent discussion (Richards,
known which one. The proven existence of 1984) speculated that this relationship was
normal faulting implies the existence of real. Unpublished information, on the other
forces capable of generating excess pore hand, suggests that for the deep sea floor
pressures and consequent hydraulic flow of (Chaney, 1985, personal communication)
pore fluids, which in turn is very relevant to and for nearshore soils (Sills, 1985, personal
the high-level radioactive waste disposal communication) the apparent over and
investigations (NEA, 1984). This would under consolidation states may be artifacts
tend to corroborate a hypothesis of excess of the test method in a laboratory con-
pore pressures in Indian Ocean soils (Abbott solidometer or oedometer. In addition,
et al., 1981) and the proposed paradigm that other scientists and engineers have studied
all cohesive marine soils deeper than a few the geotechnical properties of various Deep
metres from the seabed may exhibit excess Sea Drilling Project (DSDP) and IPOD
pore pressures in situ (Richards, 1984) cores. Bennett et al. (1984) summarize
The third set of problems concerns the strength measurements on only the DSDP
venting of hot, mineral-rich pore fluids at cores.
divergent plate boundaries, such as the hyd- These four sets of problems and one gen-
rothermal vents in the middle of the Red eral area of activity have already generated
Sea (Guney et al., 1984), in the East Pacific new types of in situ testing equipment for
Rise in the Pacific Ocean and elsewhere site investigations, which will be described
(Rona et al., 1984; Rona, 1984). This type of subsequently. When earth scientists better
venting may have produced the mineraliza- understand that marine geotechnology is a
tion causing the highly unusual brittle soils means of quantifying many seabed geologi-
having very high water contents and also cal, geophysical and geochemical processes
high shear strengths in the Pacific Ocean it is likely that additional equipment will be
nodule province (Richards and Parks, 1977). constructed for these investigations on the
The fourth specific set of problems is con- continental slope and rise and on the deep
cerned with the in situ measurement of sea floor, which will complement the
shear waves to determine the shear modulus ongoing geotechnical engineering studies.
and other parameters reflecting dynamic
IN SITU TESTING AND SAMPLING OFFSHORE 139
TABLE 5
Summary of new developments: seabed in situ testing equipment a
Penetration
Water depth (m) depth (m)
Name Used Designed Used Designed References
P enetrators
(US Navy) Shear strength 4908 6000 12? ? Beard (in press)
(Sandia) Shear strength 5400 6000? >30 <50? Talbert (1984a)
(BREfJRC) Shear strength 5400 6000? >30 <50? Freeman et al. (1984a)
aThe water depths and penetration depths have been obtained fron manufacturers or operators, and may
be optimistic in some instances because 'used' does not necessarily mean used successfully
bPCPT, piezocone penetrometer
cSucessfully tested in a hyperbaric chamber at a pressure of 55 MPa (Bennett et al. in press, a)
dUP, underwater penetrometer
eThe GIPSY is unlikely ever to be used because of a terminated project
test rods. The ROSON has worked well in North Sea in 1984 from the Stingray
water depths of less than 300 m, but not (Amundsen et al., in press).
very well at greater depths. A specially developed miniaturized geo-
McClelland recently adapted the Borros technical system was devised by National
(Sweden) land 20-ton Static Sounding Oceanic and Atmospheric Administration
Push-and-Puller Apparatus for use offshore. (NOAA personnel for use on the Woods
This equipment continuously penetrates and Hole Oceanographic Institution's submer-
retracts test rods using four hydraulic cylin- sible Alvin (Lambert, 1982). The system in-
ders operating two and two in pairs con- cluded a coring machine (described later),
nected to a single automatic rod clamp under a cone penetrometer, a resistivity-
each pair. The equipment, called a Continu- conductivity probe, the In Situ Heat Trans-
ous Penetration Machine, or CPM, by fer Experiment, or ISHTE (Percival et al.,
McClelland was successfully operated in the 1984), a miniature piezometer and a
IN SITU TESTING AND SAMPLING OFFSHORE 141
The ISHTE piezometer design and con- lute pressure, quartz-crystal, oscillating-
struction was started at NOAA. Later, all beam pressure transducers, having an
geotechnical operations were transferred to operating range of 0 to 6.2 MPa, sense
the Naval Ocean Research and Develop- pore pressures at three locations 3.3, 6.4
ment Activity, as previously mentioned. A and 9.5 m below the mudline (Prindle and
differential semiconductor transducer was Lopez, 1983). A hydrostatic pressure trans-
used on the Alvin dives (Table 5) to measure ducer is mounted just above the sea floor on
differential pore pressure. The sensor mal- the Data Gathering Subsystem (DAGS).
functioned and then failed in the laboratory at This unit acoustically telemeters data to a
a pressure of 68.9 MPa. This type of trans- Command and Recording Subsystem
ducer was subsequently abandoned and a (CARS) on the surface. The DAGS, emp-
Validyne variable reluctance differential laced using heavy weights, is designed to
pressure transducer was selected that had a remain on the sea floor for an extended
precision of 0.34 to 0.69 kPa over the same period of time.
69 MPa range (Bennett et al., in press, a). Late in 1984, Ismes (Bergamo, Italy) per-
Laboratory test results indicated 'better sonnel successfully operated their new
than expected performance'. There have Underwater Penetrometer offshore in shal-
been zero-shift problems, but these are not low water. Bruzzi (1983) has described an
considered detrimental to the measurement earlier version of this machine. The new
of ambient pore pressures (Bennett et al., in equipment automatically levelled on a 25°
press, a). The present unit is shown in Fig. slope and no problems were experienced.
7. The probe part of the unit, only 8 mm in The current cable and handling system is
diameter, is made of a titanium tube. limited to a water depth of 350 m, but the
Recently, Bennett et al. (in press, b) have unit is designed to operate down to 600 m
shown how to derive shear strength and (Bruzzi, 1985, personal communication).
permeabilities from pore-pressure data. A differential piezometer probe having six
Both the PUPPI and the ISHTE levels of pore pressure ports, usually
piezometers have a single pore pressure arranged with a pair on opposite sides, was
port. The earlier NOAA piezometer system, tested at sea in 1983 by the US Geological
designed for water depths of less than 200 m, Survey in Woods Hole, Massachusetts
had multiple ports (Bennett et al., 1982). (Booth, 1985, personal communication). The
Another new mUltiple-port piezometer sys- probe was lowered on a mechanical cable
tem is the Sandia National Laboratories' and used acoustic telemetry to communicate
Geotechnically Instrumented Sea-floor data to the ship. On the sea trial the acoustic
Probe, or GISP (Reece et al., 1979). Abso- link performed well until near the seabed
when the signals faded out. The problem is
believed to have been solved, according to
Booth, but the machine has not yet been
tested again. Eventually the plan is to have
the equipment free-fall to the seabed.
This concludes the descriptions of new
seabed equipment that has actually oper-
ated offshore in water depths greater than
300 m. Three units will now be discussed
briefly that have yet to be tested at sea.
o Each has some interesting design features.
1--1'6, .. --~
TABLE 6
Geotechnical In situ Probe System (GIPSY) specifications a
tern called the Geotechnical In situ Probe Deep Ocean Research Apparatus, or
System or GIPSY was conceptuallized at DORA. An artist's impression of the DORA
Lehigh University. The GIPSY was is shown in Fig. 8. Table 7 lists specifica-
designed and constructed with one probe tions. The DORA is designed to use a
module in the late 1970s and very early
1980s. Table 6 lists specifications. The four
probes were each designed to operated with
the 'Catapuller' continuous drive system
(Fig. 3A) and hollow, square test rods. Each
of the four probes and the fixed piston corer
modules were designed to be simply plugged
in electrically and mechanically to the main
frame containing the battery pack and elec-
tronics. Data acquisition and control was to
be by digital telemetry over an elec-
tromechanical cable. A novel feature was
the utilization of a microprocessor to provide
positive feedback to the ship that any com-
mand sent to the machine on the seabed was
actually carried out, rather than only an
electrical circuit being activated. The entire
machine was essentially completed, includ-
ing one probe and the entire GIPSY data
acquisition and control system, when the
project was abandoned at the time the
Marine Geotechnical Laboratory was dis-
banded in 1982.
A consortium consisting of Fugro,
Laboratorium voor Grondmechanica (Delft Fig. 8 An artist's impression of the Deep
Soil Mechanics Laboratory), Marine Struc- Ocean Research Apparatus (DORA) designed to
tures Consultants and Ismes (Italy) has push test rods 50 m into the seabed in a water
:ompleted a detailed feasibility study for a depth of 6000 m
144 ADVANCES IN UNDERWATER TECHNOLOGY
TABLE 7
DORA specifications
I. General
Maximum water depth 6000m
Penetration depth 50m
Foundation loading, maximum 50 kN
Penetration method (speed) Drive wheels, continuous (2 cm/s)
Data acquisition and control Microprocessor, acoustic telemetry to vessel
Power supply Lead-acid storage batteries
Power pack Hydraulic, pressure compensated
II. Piezocone
Cone-resistance range 7.5kN
Maximum shear strength of soil 140 kPa
Pore pressure range 1 MPa
Diameter of cone 46mm
Cone base area 15 cm 2
TABLE 8
Sea Lion specifications
I. General
Maximum water depth 6000 m
Penetration depth 5m
Maximum weight 1000 kg in air
1500 kg in airwith ballast
Dimensions: base plate 2.4 x 2.4 m square
height 6.8 m without skirt
Additional bearing plates 4 plates (2.7 x 0.3 m)
Skirt 0.5 m (optional use)
Penetration method (speed) Drive wheels, continuous (about 4 cm/s)
Operation time About 10 penetration-retractions per lowering
Control after touchdown on seabed Automatic (programmable logic controller),
penetration-1r"etraction sequence initiated after
a trigger touches the sea floor
Data storage Multichannel, solid-state recorder on Sea Lion
Pinger Detects sea floor, advises go-no go on inclination
and when piezocone not in full up position using
different repetition rates
Power supply Lead-acid storage batteries
Power pack Hydraulic, pressure compensated
II. Piezocone
Cone resistance range 0-2.3 kN
Maximum shear strength 50 kPa
Pore pressure range 0-600 kPa
Diameter of cone 44mm
Cone base area 15 cm 2
IN SITU TESTING AND SAMPLING OFFSHORE 145
used for assessing the liquefaction potential VIII JADSS & EATS Steel
~
+ ~
of soils onshore. It is not known if the cone
can be used offshore.
VII GlADSS & EATS Steel ::::::>
Seabed Penetrators
At present, three organizations or groups of
organizations are using penetrators in
VI
~
ocean-basin water depths. A penetrator,
which is not a penetrometer -- although a V SIADSS Steel + concre0
penetrometer may be attached to a penet-
rator -- is a free-fall from a ship or fully
dynamic instrumentated device. The mod-
IV [~J ADSS Steel + Leav
.
"'-
0
'"....
"i
:>
u
..J
<{
U
40
'"u
<{
!!!
...."
C>
Z
'"0::
....
'"
....
zw
:Ii
8
<I)
20
O~~-LL-------l~O------------'20-------------~
+-----------~~
Fig. 11 Soil strength (su) profiles back-figured from penetrator data (Freeman et al., in
preparation, a, b) using the BRE/Fugro time-stepping computer programme 'PREPEN' (Freeman
et al., in preparation, b)
TABLE 9
Summary of new developments: Samplers
Maximum core
Water depth (m) length (m)
Name! Sampler
(Organization) Type Downhole Seabed Used Design Obtained Design References
(Storms et al., 1983). The VLHPC (Fig. gas-containing soils from boreholes at their
12) has been in operation since 1981. ambient pressure (Denk et al., 1981; Johns
The Fugro hydraulic piston sampler is an et al., 1982) in shallow water. In 1983 Fugro
adaptation of the Wipsampler (wireline push successfully designed and used a new and
sampler) to use a fixed piston (Toolan, 1983). simpler sampler, the Ambient Pressure
It has been in routine operation since 1982. Sampler (APS), in the Troll Field offshore of
It is likely that the improvement in soft soil Norway (Zuidberg et al., 1984). Consulta-
sample quality using this sampler may be as tion with geochemical engineers led to the
great, compared to the push sampler, as the conclusion that a sample volume of 30 to 50
change from using the percussion sampler to cm 3 was adequate for soil gas analysis. This
the push sampler reported by Young et al. led to a conservative design of the APS to
(1983). sample 65 cm 3 of soil with as little distur-
McClelland also have a new piston sam- bance as possible using a single ball-valve
pler for downhole use with their Swordfish adapted for actuation by the Wipsampler
or Dolphin, which has been routinely used (Fig. 13). A feature of the APS is that it
since 1984. collects a non-pressurized sample 40-cm long
Texas A & M University personnel adjacent to the pressurized part of the sam-
pioneered the design and use of the wireline pler contained in the ball-valve.
geotechnical sampler to obtain samples of
IN SITU TESTING AND SAMPLING OFFSHORE 149
COllAPSEO EXTENC[O
oral communication). No other details are behaviour. For these reasons, examples of
available. four different recent developments (Table
The proposed Lehigh University fixed- 10) of shear-wave measurements in water
piston corer module for the Geotechnical In depths greater than 300 m - or applicable
Situ Probe System (GIPSY) was designed to these water depths - are considered
by N. D. Remy Associates but not con- relevant to this review.
structed. The method of core-tube penetra- Ohya et al. (1984) of the Oyo Corporation
tion and retraction was by a chain drive, in Japan have successfully used a downhole
similar to the Remy design for the XSP cone logging tool that can propagate and receive
penetrometer equipment (Beard and Lee, both P and S waves. This tool provides
1982). It is unlikely that the GIPSY corer improved resolution more appropriate for
will be built in the future for the reasons geotechnical purposes than many of the con-
previously given. ventional logging tools used for oil and gas
exploration. The vibration source is an elec-
Offshore Shear Wave Measurements tromagnetic solenoid hammer contained
within the probe. Results are said to com-
Geotechnical engineers tend to think only of pare favourably with downhole and crosshole
in situ testing and sampling equipment for methods (Ohya et al., 1984). The tool has
offshore site investigations. Wave, earth- been used offshore, but details are not
quake, ice and other environmental forces gIVen.
on offshore structures are cyclic and, in con- The US Defense Advanced Research Pro-
sequence, knowledge of the shear modulus jects Agency has developed a marine
is important in assessing soil-structure seismic system and deployed it in a dedi-
interaction under cyclic loading conditions. cated borehole in collaboration with a
In addition, Taylor Smith (1983) has sum- number of other organizations (Foss and
marized how seismo-acoustic information Wallerstedt, 1984). The seismometers will
can be used to predict soil engineering prop- be used to listen for natural seismic events.
erties and to estimate consolidation A somewhat comparable method that has
TABLE 10
Summary of new developments: offshore shear-wave measurements
been used in the shallow water of the excellent example of how to combine many
Beaufort Sea (Justice et al., 1984) may also tools for site-specific surveys in water
be applicable to the water depths of concern depths of 2000 m.
in this paper. Studies by Kirkpatrick and Khan (1984),
Ocean-bottom seismometers and seismo- as well as others, on the stress relief caused
graphs have been used for some time by sampling appear to show effects that may
(Prothero, 1984). Whitmarsh and Lilwall be greater on some soils than previously
(1984) of the Institute of Oceanographic Sci- believed. This is additional evidence to bols-
ences (lOS) in England describe the success- ter the conclusion by Richards and Zuidberg
ful use of the lOS pop-up bottom seismo- (in press, a) that in situ strength testing
graph to record S waves generated when a offshore involves fewer significant problems
heavy weight is dropped to the seabed than sample collection and subsequent
within 1000 m of the seismograph. laboratory testing. Stress relief, soil degas-
Campanella and Robertson (1984) meas- sing and other deleterious phenomena are
ure the dynamic modulus of soil in situ on likely to increase with increasing water
land using a University of British Columbia depths. This is ample reason why in situ
cone penetrometer. Small seismometers are testing may gradually replace many sampl-
installed triaxially in the same probe. Shear ing applications in the years to come. For
waves are generated at the surface and are some problems sampling under ambient
received by the seismic cone penetrometer pressure followed by geotechnical and
everyone metre after penetration has been geochemical testing under the same pres-
stopped for the measurement. Excellent sure, without having to resort to personnel
agreement is claimed with conventional working in hyperbaric chambers, is likely to
crosshole techniques (Robertson et al. , increase.
1985). While the technique has not been Site investigations made by the geotech-
used offshore, there is no apparent reason nical consulting industry show clearly two
why it could not be used. different methods of testing and sampling:
(1) by work downhole and (2) by work on
the seabed. For the first, greater electro-
THE FUTURE mechanical sophistication can be expected
utilizing the mighty microprocessor and
After reviewing deep-water problems and digital telemetry. For the second, we
equipment one might well wonder what it can expect more automated machines
means and what the future portends. It has operating from a single electromechanical
been established that science fiction writers cable or perhaps utilizing only a mechanical
generally do a much better job at divining cable and communicating data and control
the future than scientists and engineers. signals by acoustic telemetry. This may be
Nevertheless, some trends appear for through the test rods, as has been done by
industry, working at depths less than 1000 Jefferies and Funegard (1983) in shallow
to 2000 m, and for many university water, or through the entire water column.
and governmental investigators, working I t is likely that the piezocone penetrome-
in water depths between about 4000 and ter will soon replace the cone penetrometer
6000 m. offshore because the former provides much
In water depths greater than 1000 m, the more useful information at little additional
development of the GLORIA, SeaMARC II, cost. Figure 14 shows the results from a
Sea Beam and the Bathymetric Swath Sur- Fugro piezocone used from the Seacalf in
vey Systems (see Richards and Hartevelt, the North Sea in which there is a moderate
1984) have very great potential or delineate amount of variability in the soil. Both Fugro
clearly areas for geotechnical site investiga- and McClelland have used their piezocones
tions. Prior and Doyle (1984) present an in water depths greater than 300 m. It is
IN SITU TESTING AND SAMPLING OFFSHORE 153
~ -- ,--, -1---,
I
jMN 'm~
PROF ILE . . . UNIT Sl.EevE FR ICTI ON I
' . 1 . -~
'I ' I
I
-r -r
PORE PRESSURE u 4M N m2)
-
~ listed by Campanella and Robertson (1982)
have not been included in Table 11 because
W-~. ! III
of a lack of suitability perceived by the
author for one reason or another. Neverthe-
less, enough examples remain to provide a
fruitful field for the solution of a large vari-
ety of industrial and governmental-academic
problems. It is highly probable that the
trend of using multiple sensors in one probe
system will continue.
Site investigations made by government
20 ~ ~ and university researchers at present lack
I --:Er~=-.?;:.~ downhole geotechnical testing tools to com-
~~~~~~i
: --- i
~ - ~ .~
plement the HPC sampler for deep-sea
drilling under the new Ocean Drilling Prog-
.t::.. _
ramme. A 1980 report or White Paper,
~ {J:.~ 'Report by the Ad Hoc Working Group on
~ ~ ~~~~ -==~~~---- Long-Range Plans, Sedimentary Petrology
~~ .. ~
.' I=. i
f--\t-------- and Physical Properties Panel', Joint Oceano-
graphic Institutions Deep Earth Sampling
~ ,. ) II
(JOIDES), recommended the technical
development of the following downhole
Fig. 14 Results from a Fugro piezocone equipment:
penetrometer used from the Seacalf in the
North Sea in 1984. The water depth was (1) sand-core catcher for the Hydraulic Pis-
217 m ton Corer;
(2) in situ vane shear meter;
also likely that the pressuremeter may gain (3) in situ pore-pressure measuring device;
more widespread use offshore in deep water (4) in situ hydrofracturing device; and
than the dilatometer because the former (5) in situ stress-strain meter.
yields more information. A Fugro full- Interestingly, in just five years all equip-
displacement pressuremeter, operating ment in this list are in use offshore - except
behind a cone or piezocone penetrometer, perhaps the first - in water depths of a few
has been designed and built for initial tests hundred meters or so. There is no technical
on land. It is the intention to adapt such a reason why any of these tools could not be
tool for offshore work in shallow and deep adapted to work downhole in 6000 m at rela-
water. The costs of working in water depths tively modest cost.
greater than 300 m are already so great that Seabed site investigations by the gov-
the cost differential between the pres- ernmental and academic groups are likely to
suremeter and the dilatometer is likely to be be orientated towards studies on trenches,
negligible. the abyssal plain and hydrothermal vents,
There are a variety of probes and sensors all of which involve measurements of in situ
that have been used offshore in shallow or pore pressures to determine hydraulic
deep water as well as some that probably gradients for the study of vertical and
could be used offshore (Table 11). In this horizontgal pore-water fluid movement and
table a combination acoustic cone incor- to elucidate the consolidation and eventual
porating both transmitting and listening lithification of deep sea floor soils.
154 ADVANCES IN UNDERWATER TECHNOLOGY
TABLE 11
Geotechnical probes, sensors and relevant tests which are used or could be used offshore
Water
Parameters measured Used depth
Probes and sensors or (derived) b offshore >300m Selected example references c
Acousticd
Passive compo wave Acoustic emisions x ? Tringale and Mitchell (1982);
Schaap and Hoogendoorn (1984)
Active compo wave Acoustic emissions This paper
Shear wave ('seismic') Shear modulus x x? Reece et al. (1979); Campanella
and Robertson (1984)
Densitometer, nuclear Bulk density
Backscatter x Keller (1965);
x x Rose and Roney (1971)
x x Tjelta et al. (in press)
Transmission Bulk density x x Hirst et al. (1975)
Dielectric cone Capacitance Singh and Chomg (1984)
Dilatometer Stress-strain x Burgess et al. (1983)
Electrical resistivity Resistivity/conductivity x Rietsema and Viergever (1979)
x x Hulbert et al. (1982);
Bennet et al. (1983b)
Hydraulic fracture test Fracture pressure x x Toolan (1983)
Penetrator, instrumented Shear strength x x Beard (in press); Freeman
et al. (1984)
Penetrometer, 'static'
Cone/friction sleeve Tip/sleeve resistance x x de Ruiter (1982)
Piezocone 'Dynamic' pore pressure x x de Ruiter (1982)
Penetrometer, 'dynamic'e Tip/sleeve resistance x Chari et al. (1979)
Permeameter Permeability Capelle (1984)
Piezometer Pore pressure
Differential pressure x x Richards et al (1975)
Total pressure x x Presland and Babb (1979)
Bennet et al. (in press, a)
Pressuremeter Stress-strain
Push-in x x Reid et al. (1982)
Self-boring x x Fay et al. (in press)
x Hughes et al. (1984)
Full displacement This paper
Temperature cone Temperature x Schaap and Hoogendoorn (1984)
Thermal conductivity cone Thermal conductivity Schaap and Hoogendoorn (1984)
Vane Shear strength
Downhole x x Kraft et al. (1976)
Seabed x x Demars and Taylor (1971)
x x Richards et al. (1972)
x Lewis et al. (1970)
x x Silva (in press)
x x This paper
aMany sensors can be used in combination. Downhole and seabed use is not differentiated
bOnly the principal parameter is listed
cGenerally the most recent general description is given, or references relating to work by different
organizations. The general reference cited may not relate to offshore use
d'Passive' refers to listening only and 'active' refers to transmitting and listening in the same probe
eThe free-fall type from the seabed, not the type used on land
IN SITU TESTING AND SAMPLING OFFSHORE 155
At these great depths, the trend is likely depths exceeding 300 m. A few, how-
to include more acoustic telemetry through ever, have been rapidly developed speci-
the water column to avoid the expense and fically for these water depths.
complexity of electromechanical cables; also 3. Industrial-type problems for site inves-
the use of fully autonomous free-fall (from tigations in water depths greater than
the ship) and 'pop-up' (to the ship) equip- 300 m tend to be different from site
ment having self-contained recorders or investigations conducted by governmen-
using acoustic telemetry is likely to tal-academic organizations in response
increase. to their particular problems.
A new generation of manned and unman- 4. Many more seabed penetration systems,
ned submersibles for water depths of 6000 m probes and sensors have been improved
is near at hand. The PLA2, developed by or developed during the past three to five
a number of French organizations, is a years than comparable downhole equip-
fully pre-programmed autonomous ROV ment. The reason for this probably is that
designed to descend to 6000 m and automat- downhole geotechnical investigations are
ically harvest a small load of polymetallic not usually available to governmen-
nodules. The prototype has been built and is tal-academic groups because of the cost.
expected to be ready for testing at sea very Three notable exceptions, the DSDP,
soon. The INFREMER (ex. CNEXO) IPOD and presently the ODP, are in
organization soon plans to operate its man- general under the control of scientists
ned submersible SM79, which is designed who are only now beginning to under-
for 6000 m, a depth at which the autonomous stand what geotechnical engineering can
Epaulard ROV already operates. The contribute to science.
larger submersibles could easily carry 5. The seabed testing systems utilize a var-
geotechnical instrumentation, some of which iety of different deployment methods,
is in an early stage of development. * This which are very briefly described in Table
activity in Europe is expected to be an 2. Particularly rapid development is
interesting addition to the Atlantis II and occurring in the continuous drive-wheel
Alvin submersible operations in the United devices and in the penetrators. The latter
States. have the potential of revolutionizing
ocean basin soil investigations in the
geoscience and geotechnical fields. This
SUMMARY AND CONCLUSIONS is because the penetrator is a means of
emplacing, relatively cheaply, a seabed
1. There are two moderately distinct test- laboratory deeply into the seabed to
ing and sampling activities in water receive data at the surface by acoustic
depths exceeding about 300 m: (1) indus- telemetry either in real time or until
trial, in water depths usually less than it is recalled from a memory system.
about 1000 m, and (2) governmental- The receiver may be aboard a ship, or
academic, in water depths generally on a buoy for transmission ashore via a
greater than about 4000 m. Some satellite.
equipment is shared between the two 6. Seabed samplers currently of great
activities and some is distinctly different. interest are the various devices to obtain
2. Most systems, probes and sensors have high-quality cores about 30 m long. Three
been gradually developed over long time different organizations have active
periods before being used in water development plans. At least two groups
will test new or improved equipment at
'From presentations at the Society for Underwater
abyssal plain depths in 1985, if present
Technology's meeting in December 1984 on objectives and plans come to fruition.
technology for ocean instrumentation deeper than 200 m. 7. One downhole and one seabed develop-
156 ADVANCES IN UNDERWATER TECHNOLOGY
ment are described to measure seismic pare this paper, for kindly making available
events or shear waves in ocean-basin Figs 10 and 11 in advance of formal publica-
water depths. In addition, one down- tion and for valued discussions on various
hole and one seabed-type development topics. Drs P. J. Schultheiss, Institute of
exists actually or potentially to measure Oceanographic Sciences, and R. H. Bennett,
shear waves in situ. These developments Naval Ocean Research and Development
are the likely precursors of routine Activity, made available unpublished man-
shear-wave measurements leading to an uscripts and permitted publication of Figs 6
improved understanding of the range of and 7, respectively. Mr R. L. Geer kindly
shear moduli to be found in soils of the made available Figs 1, 3 and 4 from the Shell
deep sea floor. Oil Company. Mr D. Cover, Scripps Institu-
8. Many probes and sensors exists for in tion of Oceanography, Deep Sea Drilling
situ testing in water depths exceeding Project, kindly made available Fig. 12. Mr
300 m. Many others, as well as certain A. W. W. Tieges, Delft Soil Mechanics
tests - such as the hydraulic fracture Laboratory, kindly permitted the use of
test - are applicable for use offshore in Fig. 9. Mr J. M. Bayne, McClelland; Dr J.
these water depths. Table 11 summarizes Booth, US Geological Survey; Dott. D.
both. It is concluded that both industrial Bruzzi, Ismes; Dr J. Burdett, Building
and governmental-academic organiza- Research Establishment; Drs A. Driscoll
tions will be combining more and more and A. Silva, University of Rhode Island;
sensors into a single probe system for use Dr I. Noorany, San Diego State University;
either downhole or from seabed equip- Dr G. Singh, Leeds University; Mr P. Le
ment. Tirant, French Institute of Petroleum; and
9. The future is examined in light of the var- others, as well as colleagues at Fugro,
ious developments reported. It is con- graciously provided much useful informa-
cluded that because of stress relief, soil tion. Mrs E. A. Richards assisted with the
degassing and so forth, as samples are references and illustrations. Mr D. W. Eij-
brought from water depths exceeding maal modified several illustrations and
300 m to the surface, in situ testing may drafted two. Messrs J. M. Geise, H. J. Kolk,
replace many requirements for sampling J. de Ruiter, J. van der Wal and H. M.
in the future. Industrial testing is likely Zuidberg reviewed the manuscript and
to be characterized by increased use made helpful comments. The opinions in this
of the piezocone penetrometer down- paper are those of the author and do not
hole and from continuous-drive seabed necessarily represent those of Fugro BV or
devices. Governmental-academic testing any other organization or person.
is likely to increase with the development
of downhole and seabed equipment for
use in water depths exceeding 4000 m.
The measurements in situ of true REFERENCES
ambient pore pressures to calculate the
existence of any hydraulic gradients in 1. Aas., J., Lacasse, S., Lunne, T. and Mad-
ocean basin soils is likely to receive shus, C. 1984. In situ testing: new deve-
emphasis during at least the next few lopments. N ordiska Greteknikermfjtet.
Vol. 2: Linkaping, pp. 705-716 (Repro-
years. duced in Norwegian Geotechnical Institute
Publication Nr. 153, 1984).
2. Abbott, D., Menke, W., Hobart, M. and
ACKNOWLEDGEMENTS Anderson, R. 1981. Evidence for excess
pore pressures in southwest Indian Ocean
Mr T. Freeman, Building Research Establ- sediments. J. Geophys. Res. 86 (B3),
ishment, is thanked for inviting me to pre- 1813-1827.
IN SITU TESTING AND SAMPLING OFFSHORE 157
tests and their application in offshore nental slope. Abstract, EOS Trans Am.
design. Geotechnical Practice in Offshore Geophys. Union 65(45), 1089-1090.
Engineering (Ed. S. G. Wright). American 31. Chaney, R. C. and Fang, H.-Y. In press.
Society of Civil Engineers, New York, pp. Static and dynamic properties of marine
244-266. sediments. First Shanghai Symposium on
21. Brooks, J. M., Kennicutt, M. C., II, Fay, Marine Geotechnology and Nearshore/Off-
R. R. and McDonald, T. J. 1984. Ther- shore Structures (Eds R. C. Chaney and
mogenic gas hydrates in the Gulf of Mexico. H.-Y. Fang). Special Technical Publica-
Science 225 (4660), 409-411. tion, American Society for Testing and
22. Brucy, F., Fay, J. B. and Le Tirant, P. Materials, Philadelphia.
1984. Three years' experience with the off- 32. Chari, T. R., Abdel-Gawad, S. M. and
shore self-boring pressuremeter 'PAM' Chaudhuri, S. N. 1979. Geotechnical sur-
Offshore Technology Conference Proceed- vey of the seafloor with a free fall penet-
ings, Vol. 1, Paper 4677, pp. 265-274. rometer. POAC 79, Fifth International
23. Bruzzi, D. 1983. Underwater static penet- Conference on Port and Ocean Engineer-
rometer. International Symposium Soil ing Under Arctic Conditions Proceedings,
and Rock Investigations by In-Situ Test- Vol. 2. The Norwegian Institute of Tech-
ing, Vol. 2, pp. 223-226. nology, Trondheim, pp. 833-843.
24. Burgess, N. C., Hughes, J. M. O.,Innes, 33. Colp, J. L., Caudle, W. N. and Schuster,
R. and Gleadowe, J. 1983. Site investiga- C. L. 1975. Penetrometer system for
tion and in-situ testing techniques in Arctic measuring in-situ properties of marine sed-
seabed sediments. Offshore Technology iments. In Oceans 75 Conference Records,
Conference Proceedings, Vol. 3, Paper Institute of Electrical and Electronics
4583, pp. 27-34. Engineers, and Marine Technology Soci-
25. Campanella, R. G. and Robertson, P. K. ety, pp. 405-411.
1982. State of the art in in-situ testing of 34. Cyr, R. J. 1984. Sea-bed surveys by acous-
soils: developments since 1978. Soil tic penetrometer. Oceanology Interna-
Mechanics Series No. 56, The University of tional, Paper 1. 7.
British Columbia, Vancouver. 35. Dahlberg, R. 1974. Penetration testing in
26. Campanella, R. G. and Robertson, P. K. Sweden. In Proceedings of the European
1984. A seismic cone penetrometer to Symposium on Penetration Testing, Vol.
measure engineering properties of soil. 1, pp. 115-131.
Paper presented at the Society of Explora- 36. Demars, K. R. and Taylor, R. J. 1971.
tion Geophysicists' 54th Annual Meeting, Naval seafloor soil sampling and in-place
Atlanta. test equipment: a performance evaluation.
27. Capelle, J.-F. 1984. A new instrument for Report No. 730, US Navy Civil Engineer-
the in situ measurement of the permeabil- ing Laboratory.
ity of clays: the self boring non-clogging 37. Denk, E. W., Dunlap, W. A., Bryant, W.
permeater. Field Measurements in R., Milberger, L. J. and Whelan, T. J., III,
Geomechanics, Vol. 1 (Ed. K. Kovari). 1981. A pressurized core barrel for sampl-
A.A. Balkema, Rotterdam, pp. 49-57. ing gas-charged marine sediments. In Off-
28. Carson, B. 1977. Tectonically induced shore Technology Conference Proceedings,
deformation of deep-sea sediments off Vol. 4, pp. 43-52.
Washington and northern Oregon: mechan- 38. Driscoll, A. H. 1981. The long coring facil-
ical consolidation. Mar. Geol. 24, 289-307. ity, new techniques in deep ocean coring.
29. Carson, B., von Huene, R. and Arthur, M. In Oceans 81 Conference Records, Vol. 1.
1982. Small scale deformation structures Institute of Electrical and Electronics
and physical properties related to con- Engineers, and Marine Technology Soci-
vergence in Japan Trench slope sediments. ety, pp. 404-410.
Tectonics 1 (3), 277-302. 39. Duin, E. J. Th., Mesdag, C. S. and Kok, P.
30. Carson, B., Ritger, S. D. and Suess, E. T. J. 1984. Faulting in Madeira Abyssal
1984. Precipitation of carbonate crust Plain sediments. Mar. Geol. 56, 299-308.
associated with subduction-induced pore- 40. Fay, J. B., Montarges, R., Le Tirant, P.
water expUlsion: Washington-Oregon conti- and Brucy, F. 1985. Use of the PAM self-
IN SITU TESTING AND SAMPLING OFFSHORE 159
boring pressuremeter and the STACOR Hole 532A, Initial Report of the Deep Sea
large-size fixed-piston corer for deep sea- Drilling Project, Vol. 75 (Eds W. W. Hay,
bed surveying. Paper 12, this volume. J.-C. Sibuet et al.). US Government Print-
41. Ferguson, G. H., McClelland, B. and Bell, ing Office, Washington, pp. 1109-1127.
W. D. 1977. Seafloor cone penetrometer for 51. Graaf, H. C. van de and Smits, A. P. 1983.
deep penetration measurements of ocean Offshore site investigations by rotary dril-
sediment strength. Offshore Technology ling from a diving bell. Ground Engng
Conference Proceedings, Vol. 1, Paper 16(1), 18-19.
2787, pp. 471-478. 52. Guney, M., Nawab, Z. and Marhoun, M. A.
42. Flodin, N. and Broms, B. 1981. History of 1984. Atlantis-II-Deep's metal reserves
civil engineering in soft clay. Soft Clay and their evaluation. Offshore Technology
Engineering (Eds E. W. Brand and R. P. Conference Proceedings, Vol. 3, Paper
Brenner). Elsevier, Amsterdam, pp. 4780, pp. 33-44.
27-156. 53. Haas, W. J. M. de, 1983. A seabed sampler
43. Foss, G. N. and Wallerstedt, R. L. 1984. for the dredging industry. World Dredging
Operational aspects of borehole deploy- Congress 1983, Paper E3, pp. 237-249.
ment of a marine seismic system in deep 54. Hansbo, S. 1957. A new approach to the
water. Offshore Technology Conference determination of the shear strength of clay
Proceedings, Vol. 2, Paper 4757, pp. by the fall-cone test. R. Swed. Geotech.
397-406. Inst. Proc. 14, 1-47.
44. Freeman, T. J. and Burdett, J. R. F. In 55. Henderson, G., Smith, P. D. K. and St.
preparation, a. Deep ocean penetrator John, H. D. 1980. The development of the
experiments. Final report to the Commis- push-in pressuremeter for offshore site
sion of the European Communities, Con- investigation. Offshore Site Investigation
tract 392-83-7 WAS, UK. (Ed. D. A. Ardus). Graham & Trotman,
45. Freeman, T. J., Carlyle, S. G., Francis, T. London, pp. 159-167.
J. G. and Murray, C. N. 1984a. The use of 56. Hirst, T. J., Perlow Jr., M. and Richards,
large-scale penetrators for the measure- A. F. 1975. Improved in situ gamma-ray
ment of deep-ocean sediment properties. transmission densitometer for marine sed-
Oceanology International, Paper 1.8. iments. Ocean Engng 3(1), 17-27.
46. Freeman, T. J., Murray, C. N., Francis, T. 56a. Hollister, C. D., Silva, A. J. and Driscoll,
J. G., McPhail, S. D. and Schultheiss, P. J. A. 1973. A giant piston-corer. Ocean
1984b. Modelling radioactive waste dis- Engng 2, 159-168.
posal by penetrator experiments in the 57. Hughes, J. M. 0., Jefferies, M. G. and
abyssal Atlantic Ocean. Nature 310(5973), Morris, D. L. 1984. Self-bored pressureme-
130-133. ter testing in the Arctic offshore. Offshore
47. Freeman, T. J., Murray, C. N. and Tal- Technology Conference Proceedings, Vol.
bert, D. M. In preparation, b. Penetrator 1, Paper 4676, pp. 255-264.
experiments in the Nares Abyssal Plain of 58. Hulbert, M. H., Bennett, R. H. and Lam-
the Atlantic Ocean. bert, D. N. 1982. Seabed geotechnical
48. Fyffe, S., Reis, W. M. and St. John, H. D. parameters from electrical conductivity
1982. The use of the push-in pressuremeter measurements. Geo. Mar. Lett. 2,219-222.
in offshore site investigation. Oceanology 59. Hunt, R. E. 1984. Geotechnical Engineer-
International Exhibition and Conference, ing Investigation Manual. McGraw-Hill,
Vol. 1, Paper 4.5. New York, pp. 103-104.
49. Geise, J. M. and Kolk, H. J. 1983. The use 60. Inderbitzen, A. L. and Simpson, F. 1972. A
of submersibles for geotechnical investiga- study of the strength characteristics of min-
tions. The Design and Operation of ing sediments utilizing a submersible.
Underwater Vehicles, Subtech '83. Paper Underwater Soil Sampling, Testing, and
7.3, Society of Underwater Technology, Construction Control. Special Technical
London. Publication 501, American Society for Test-
50. Geotechnical Consortium, 1984. Geotechni- ing and Materials, Philadelphia, pp.
cal properties of sediments from Walvis 204-215.
Ridge. Deep Sea Drilling Project, Leg 75, 61. Ingram, C. 1982. Expendable penetrome-
160 ADVANCES IN UNDERWATER TECHNOLOGY
ter for seafloor classification. Geo. Mar. Technology Conference Proceedings, Vol.
Lett. 2, 239-241. 3, Paper 2626, pp. 75-96.
62. James, L. T. and Calloway, T. M. 1983. 73. Kullenberg, B. 1947. The piston core sam-
Initial field tests of ISP-1 and ISP-2 - sys- pler. Svenska Hydrografisk - Biologiska
tems that extend the technology of Kommissionens Skrifter, Tredje Serien,
instrumented seabed penetrators. Oceans Hydrografi, Band 1, Hiifte 2.
83 Proceedings, Vol. 1. Institute of Electri- 74. Kullenberg, B. 1955. Deep-sea coring.
cal and Electronics Engineers, and Marine Reports of the Swedish Deep-Sea Expedi-
Technology Society, pp. 608-612. tion, 4(2), Fasc. 1, pp. 35-96.
63. James, L. T., Edrington, T. S., Reis, G. E. 75. Kulm, L. D., Thornburg, T. M. and Car-
and Suazo, J. E. 1981. Development of a son, B. 1984. Distribution of lithologies and
gun-launched, instrumented seabed pen- fluid vents from the Oregon underthrust
etrometer system. Oceans 81 Conference margin: implications for active subduction
Record, Vol. 2. Institute of Electrical and of the Juan de Fuca plate. Abstract, EOS
Electronics Engineers, and Marine Tech- Trans Am. Geophys. Union 65(45), 1090.
nology Society, pp. 656-661. 76. Kvenvolden, K. A. and Barnard, L. A.
64. Jefferies, M. G. and Funegard, E. 1983. 1982. Hydrates of natural gas in continental
Cone penetration testing in the Beaufort margins. Studies in Continental Margin
Sea. Geotechnical Practice in Offshore Geology (Eds J. S. Watkins and C. L.
Engineering (Ed. S. G. Wright). American Drake), AAPG Memoir No. 34, American
Society of Civil Engineers, New York, pp. Association of Petroleum Geologists, Tulsa,
220-243. pp. 631-640.
65. Johns, M. W., Taylor, E. and Bryant, W. 77. Lambert, D. N. 1982. Submersible
R. 1982. Geotechnical sampling and testing mounted in situ geotechnical instrumenta-
of gas-charged marine sediments at in situ tion. Geo. Mar. Lett 2, 209-214.
pressure. Geo. Mar. Lett. 2, 231-236. 78. Ledoux, J. L., Menard, J. and Soulard, P.
66. Justice, J. H., Hinds, R. and Stirbys, A. F. 1982. The penetro-gammadensimeter.
1984. The use of vertical seismic profiling in Penetration Testing, Vol. 2 (Eds A. Ver-
geotechnical site investigation. Offshore ruijt, F. L. Bel·ngen and E. H. de Leeuw).
Technology Conference Proceedings, Vol. A.A. Balkema, Rotterdam, pp. 679-681.
2, Paper 4756, pp. 391-396. 79. Lee, H. J. In press. State of art: laboratory
67. Kallstenius, T. 1961. Development of two determination of the strength of marine
modern continuous sounding methods. In soils. Strength Testing of Marine Sedi-
Fifth International Conference on Soil ments: Laboratory and In-Situ Measure-
Mechanics and Foundation Engineering ments (Eds R. C. Chaney and K. R.
Proceedings, Vol. 1, pp. 475-480. Demars). Special Technical Publication
68. Keller, G. H. 1982. Organic matter and the 883, American Society for Testing and
geotechnical properties of submarine sedi- Materials, Philadelphia.
ments., Geo. Mar. Lett. 2, 191-198. 80. Le Tirant, P., Fay, J. B., Brucy, F. and
69. Keller, G. H. 1965. Nuclear density probe Jezequel, J.-F. 1981. A self-boring pres-
for in place measurement in deep-sea sedi- suremeter for deep sea soils investigations.
ments. Transactions of the Conference and Offshore Technology Conference Proceed-
Exhibit, Marine Technology Society and ings, Vol. 2, Paper 4019, pp. 115-126.
American Society of Limnology and 81. Lewis, L., Nacci, V. and Gallagar, J. 1970.
Oceanography, Vol. 1, pp. 363-372. In situ investigations of ocean sediments.
70. Kermabon, A. and Cortis, V. 1969. A new Civil Engineering in the Oceans II Pro-
Sphincter corer with a recoilless piston. ceedings. American Society of Civil
Mar. Geol. 7, 147-159. Engineers, New York, pp. 641-654.
71. Kirkpatrick, W. M. and Khan, A. J. 1984. 82. Lunne, T., Lacasse, S., Aas, G. and Mad-
The reaction of clays to sampling stress shus, C. In press. Design parameters for
relief. Geotechnique 34(1), 29-42. offshore sands; use of in situ tests. Paper
72. Kraft, L. M., Jr., Ahmad, N. and Focht, J. 17, this volume.
A. 1976. Application of remote vane results 83. The Marine Geotechnical Consortium. In
to offshore geotechnical problems. Offshore press. Geotechnical properties of north-
IN SITU TESTING AND SAMPLING OFFSHORE 161
west Pacific pelagic clays: Deep Sea Dril- Conference Proceedings, Vol. 2, Paper
ling Project Leg 86, Hole 576A. Leg 86 Ini- 3461, pp. 901-908.
tial Reports of the Deep Sea Drillfng Pro- 95. Prindle, R. W. and Lopez, A. A. 1983. Pore
ject, Vol. 86 (Ed. G. R. Heath). US Print- pressures in marine sediments - 1981 test
ing Office, Washington. of the geotechnically instrumented seafloor
84. Marsland, A. and Windle, D. 1982. probe (GISP). Offshore Technology Con-
Developments in offshore site investiga- ference Proceedings, Vol. 1, Paper 4463,
tion. Oceanology International Exhibition pp. 173-180.
and Conference, Vol. 1, Paper 2.7. 96. Prior, D. B. and Doyle, E. H. 1984. Geolog-
4a. McCoy, F. W. and Selwyn, S. 1984. The ical hazard surveying for exploratory dril-
hydrostatic corer. Mar. Geol. 54, 33-41. ling in water depths of 2000 meters. Off-
85. Moeyes, G. and Hackley, M. 1983. Soil shore Technology Conference Proceedings,
investigations in the Troll area. Offshore Vol. 2, Paper 4747, pp. 311-318.
Northern Seas Advanced Project Confer- 97. Prothero, W. A., Jr. 1984. Ocean bottom
ence, Stavangar, Paper T6. seismometer technology. EOS Transac-
86. Montarges, R., Le Tirant, P., Wannesson, tions, Am. Geophys. Union 65(13),
J., Valery, P. and Bethon, J. L. 1983. 113-116.
Large-size stationary-piston corer. In Deep 98. Reece, E. W., Ryerson, D. E., Kestly, J.
Offshore Technology 2nd International D. and McNeill, R. L. 1979. The develop-
Conference and Exhibition, pp. 63-74. ment of in situ marine seismic and
87. Moore, D. G. and Richards, A. F. 1962. geotechnical instrumentation systems.
Conversion of 'relative shear strength' POAC 79. In Fifth International Confer-
measurements by Arrhenius on east Pacific ence on Port and Ocean Engineering
cores to conventional units of shear Under Artic Conditions Proceedings, Vol.
strength. Geotechnique 12, 55-59. 1. The Norwegian Institute of Technology,
88. NEA, 1984. Seabed Disposal of High-Level Trondheim, pp. 331-344.
Radioactive Waste. Nuclear Energy 99. Reid, W. M., St. John, H. D., Fyffe, S. and
Organisation for Economic Co-operation Rigden, W. J. 1982. The push-in pres-
and Development, Paris. suremeter. In Symposium on the Pres-
89. Nieuwenhuis, J. K. and Smits, F. P. 1982. suremeter and Its Marine Applications,
The development of a nuclear density probe 37. Editions Technip, Paris, pp. 247-261.
in a cone penetrometer. Penetration Test- 100. Richards, A. F. 1972. Instrumentation of
ing, Vol. 2 (Eds A. Verruijt, F. L. two submersibles for in situ geotechnical
Beringen and E. H. de Leeuw). A.A. measurements in cohesive sea floor soils.
Balkema, Rotterdam, pp. 745-749. 2nd International Ocean Development
90. Ohya, S., Ogura, K. and Imai, T. 1984. The Conference Preprints, Vol. 2, pp.
suspension PS velocity logging system. 1329-1346.
Offshore Technology Conference Proceed- 101. Richards, A. F. 1982. Review of marine in
ings, Vol. 1, Paper 4680, pp. 291-298. situ geotechnical testing equipment.
91. Okumura, T. and Matsumoto, K. 1981. Oceanology International Exhibition and
Marine auto sampler and sample quality. Conference Papers, Vol. 1, Paper 2.1, 10
Soil Mechanics and Foundation Engineer- pp.
ing Tenth International Conference Pro- 102. Richards, A. F. 1984. Modelling and the
ceedings, Stockholm, Vol. 2, pp. 537-540. Consolidation of Marine Soils. In Seabed
92. Percival, C. M., McVey, D. F., Olson, L. Mechanics (Ed. B. Denness). Graham &
O. and Silva, A. J. 1984. In situ heat trans- Trotman, London, pp. 3-8.
fer experiment (ISHTE). Mar. Geotech. 103. Richards, A. F. and Chaney, R. C. 1981.
5(3-4), 361-377. Present and future geotechnical research
93. Pheasant, J. 1984. A microprocessor con- needs in deep ocean mining. Mar. Min.
trolled seabed rockdrill/vibrocorer. 2(4), 315-337.
Underwater Technol. 10(1), 10-14. 104. Richards, A.F. and Hartevelt, J. J. A.
B4. Preslan, W. L. and Babb, L. 1979. Pie- 1981. Marine engineering geology; scope
zometer measurement for deep penetration and new developments. In 27th Interna-
marine applications. Offshore Technology tional Geological Congress Proceedings,
162 ADVANCES IN UNDERWATER TECHNOLOGY
Vol. 17. NVU Press, Utrecht, The Nether- 114. Rose, V. C. and Roney, J. R. 1971. A nuc-
lands, pp. 261-275. lear guage for in-place measurement of sed-
105. Richards, A. F. and Parks, J. M. 1977. iment density. Offshore Technology Con-
Geotechnical predictor equations for east ference Proceedings, Vol. 1, Paper 1329,
central North Pacific nodule mining area pp.43-52.
sediments. Offshore Technology Confer- 115. Ruiter, J. de 1982. The static cone penetra-
ence Proceedings, Vol. 1, Paper 2773, pp. tion test state-of-the-art-report. Penetra-
377-386. tion Testing (Eds A. Verriujt, F. L.
106. Richards, A. F. and Zuidberg, H. M. In Beringen and E. H. de Leeuw). Proceed-
press, a. State of art: in situ determination ings Second European Conference on
of the strength of marine soils. In Strength Penetration Testing, Vol. 2. A.A.
Testing of Marine Sediments: Laboratory Balkema, Rotterdam, pp. 389-405.
and In Situ Measurements (Eds R. 116. Schaap, L. H. J. and Hoogendoorn, H. G.
C.Chaney and K. R. Demars). Special 1984. A versatile measuring system for
Technical Publication 883. American Soci- electric cone penetration testing. Field
ety for Testing and Materials, Philadelphia. Measurements in Geomechanics, Vol. 1
107. Richards, A. F. and Zuidberg, H. M. In (Ed. K. Kovari). A.A. Balkema, Rotter-
press, b. Sampling and in situ geotechnical dam, pp. 313-324.
investigations offshore. First Shanghai 117. Schaap, L. H. J. and Liefting, J. H. 1984.
Symposium on Marine Geotechnology and Geotechnical applications of solid state
Nearshore/Offshore Structures (Eds R. C. recorders. Field Measurements in
Chaney and H.-Y. Fang). Special Technical Geomechanics, Vol. 1 (Ed. K. Kovari).
Publication, American Society for Testing A.A. Balkema, Rotterdam, pp. 325-333.
and Materials, Philadelphia. 118. Schlosser, F. 1978. Recent advance in in
108. Richards, A. F., McDonald, V. J., Olson, situ testing. Bulletin de Liaison des
R. E. and Keller, G. H. 1972. In-place Laboratories des Ponts et Chaussees, Spe-
measurement of deep-sea soil shear cial Issue VI E, 15-43.
strength. Symposium on Underwater Soil 119. Schultheiss, P. J. 1982. Geotechnical prop-
Sampling, Testing and Construction Con- erties of deep sea sediments: a critical
trol. Special Technical Publication No. 501, review of measurement techniques. Report
American Society for Testing and Materi- No. 134, Institute of Oceanographic Sci-
als, Philadelphia, pp. 55-68. ences, Wormley.
109. Richards, A. F., 0ien, K., Keller, G. H. and· 120. Schultheiss, P. J., McPhail, S. D., Pack-
Lai, J. 1975. Differential piezometer probe wood, A. R. and Hart, B. In press. An
for an in situ measurement of sea-floor instrument to measure differential pore
pore-pressure. Geotechnique 25, 229-238. pressures in deep ocean sediments: Pop-
110. Rietsema, R. A. and Viergever, M. A. Up-Pore-Pressure Instrument (PUPPI).
1979. In situ measurement of permeability. Report No. DoE/RW, Institute of
In Design Parameters in Geotechnical Oceanographic Sciences, Wormley.
Engineering, Vol. 2. Seventh European 121. Schuttenhelm, R. T. E., Kuijpers A and
Conference on Soil Mechanics and Founda- Duin, E. J. Th. 1985. The geology of some
tions Engineering. British Geotechnical Atlantic abyssal plains and the engineering
Society, London, pp. 261-264. implications. Paper 3, this volume.
111. Robertson, P. K., Campanella, R. G. and 122. Shephard, L. E. and Bryant, W. R. 1983.
Gillespie, D. 1985. Seismic CPT to measure Geotechnical properties of lower trench
in-situ shear wave velocity. Paper pre- inner-slope sediments. Tectonophysics 99,
sented at the American Society of Civil 279-312.
Engineers, April 1985. Denver. 123. Silva, A. 1985. The Comparison ofin situ and
112. Rona, P. A. 1984. Hydrothermal mineral- ship-board vane measurements on a deep-
ization at seafloor spreading centers. sea clay. Paper 14, this volume.
Earth-Sci. Rev. 20, 1-104. 124. Silva, A. J., Babb. J. D., Lipkin, J., Piet-
113. Rona, P. A., Bostrom, K., Lucien, L. and ryka, P. and Butler, D. In press. In situ
Smith, K. L. (Eds) 1984. Hydrothermal vane system for seafloor strength investig-
Processes at Seafloor Spreading Centers. ations. IEEE J. Oceanic Engng.
Plenum Press, New York. 125. Singh, J. and Chomg, M. K. 1984.
IN SITU TESTING AND SAMPLING OFFSHORE 163
165
166 ADVANCES IN UNDERWATER TECHNOLOGY
:2) emphasis on extensive, advanced soil investigations, which do not easily provide
testing on board the survey ship; reference criteria for control. The main con-
(3) the comprehensive use of in situ testing tractor's quality assurance procedures were
with several types of tests for reference used as a standard throughout the survey.
and extrapolation purposes. Quality control was established at all levels
of the investigation and was performed as a
Quality assurance and control in geotech- separate function with separate personnel at
nical investigations has not been typically all important stages.
- rmalized, mainly due to the nature of the
168 ADVANCES IN UNDERWATER TECHNOLOGY
TECHNICAL DESCRIPTION
,
--~
Technical Specifications
iAREA 5
Technical specifications for the geotechnical
field work were included in the tender and
~ ,
AREA 1
contract documents. In particular, detailed
specifications covered the following equip-
ment used for the Troll East Field soil
AREA 4 investigation:
• drillship;
• positioning systems;
• seabed reaction frame (including hyd-
raulic units, TV camera and current
----r-----.-------- -- ---
meter);
LEGEND' • sampling equipment (hammer, standard
• BOREHOLE WITH SAMPLING. DOWNHOLE CPT .VANE, push and piston sampling as well as grav-
~
PRESSURE METER TESTING
SURFACE PIEZOCONE TESTING ity coring);
+ SURFACE VANE TESTING
• cone penetration tests including
Fig.3 Soil investigation programme measurement of pore water pressure
(piezocone) both downhole and surface
Laboratory Testing modes;
• vane tests, both downhole and surface
Specialist laboratory equipment was taken modes;
offshore, including an X-ray used to select • push-in pressuremeter tests.
high-quality portions of samples for testing. The purpose of these specifications was to
Results from the offshore testing pro- ensure that sampling and testing were car-
gramme were to be used as a basis to quantify ried out to accepted standards and the
possible effects of storage time and trans- relevant data and descriptions reported.
port on sample quality, to provide high- Detailed specifications were provided for
quality data input for designing the exten- equipment and testing procedures, instru-
sive onshore laboratory programme, and for ment calibration prior, during and after field
selecting preliminary soil design parameters work, and data acquisition and reporting.
at an early design stage.
Drillship
In Situ Testing
MV Bucentaur, as shown in Fig. 4, was con-
Detailed technical specifications were estab- tracted by N orsk Hydro for its maiden voy-
SOIL INVESTIGATION AT THE TROLL EAST FIELD 169
Positioning Systems
tern is capable of operating in water depths readout signals from the meter, were
up to 900 m. transmitted through the existing sea-floor
The sea-floor jack has sensing units to jack umbilical.
monitor both the roll and the tilt of the unit,
the pulldown pressure of the vertical rams Downhole Sampling and Cone Testing
and the hydrostatic pressure at the sea
floor. Also associated with the jack is a bit Downhole sampling and cone testing was
guide, seen on top of the jack in Fig. 5, performed using an updated version of the
which runs along the two-part heavy lift line Stingray system as described by Semple and
to guide the drill string into the jacking unit. Johnston (1979). Cone penetrometer and
vane testing were undertaken to a max-
Seafloor Jacking Unit Stability imum of 6 m beyond the base of the
borehole. Vane and pressuremeter tools
During the tender stage for the Troll East were also pushed into the soil to the
Field project settlement of the Stingray required penetration and held motionless
sea-floor jacking unit was assessed, since during the test. Figure 6 shows the various
very soft soils were known to exist at sea stages in operating the system for downhole
floor. Consequently, a skirt perimeter was cone testing.
added around the base of the jack extending In soft soils, the normal drill-string heave
to a depth of about 1.2 m. This skirt compensation systems will not function, as
increased the soil-bearing capacity by a fac- weight cannot be set off on the bit. In order
tor of three, essentially eliminating settle- to 'drive' the drill-string compensator it was
ment of the sea-floor unit for the expected
duration of soil borings.
Current Meter
-I
Remote Vane Testing
TECHNICAL SPECFICATOONS
ponder system. This enabled control of mul- seven-conductor electric wireline to service
tiple tests at each penetration. After a test four transducers in the cone.
the system was recovered and the vane set Five channels of data, including CPM
at a different distance under the frame so thrust, were logged during continuous oper-
that sea-floor and shallow soil strengths ation. Data is stored in the active computer
could be measured. memory until testing can be interrupted to
allow transfer to disc storage. A multichan-
nel strip chart recorder was used to display
Data Acquisition continuously analogue test data for the
operator in adjustment and test-control
A schematic diagram of the data acquisition operations.
system used for surface and downhole cone
penetration and vane testing is shown in Push-in Pressuremeter
Fig. 11. Signal conditioning at the cone
using downhole amplifiers allows the The push-in pressurementer (PIP) was ini-
+20
LJ
8VDC
Reg~Yated StriP chart
~
Regulator COM for
-20 supply operator control
C s PPPTHCP
H Amplifier
L
r
c ot--
'-5 s
8VDC
U
~
Regulator r---- p p
Attenuator
r-- pp
circuit
pp
r-- T TH
r- CP
H
TH
Amplifier
h
D 8VDC
L I~" c s P PP TH CP
~
Cone
I
Regulator
Penetra tlon
HP 3497
Data logger
---
~~.
H Amplifier
L Teleh~~~th~"sTut
water depth
HP87XM
Computer
@
Temperature
I Temperature
circuit
.+
t-'"
h
Water depth
striP chart
HP 82901
DIsk
transducer
Armour Screen
1*1
Fig.11 Cone penetrometer data aquisition system
174 ADVANCES IN UNDERWATER TECHNOLOGY
TABLE 1
I
I
I \
Procedure used on board Bucentaur during
I I \ 1984 soil investigation at troll
pr.fil~/ : \ A
ople I I \ i
inder ~-----:--=w 1. Sample comes to X-ray container
'ay film A J 2. Lead numbers from 0 to 100 (in cm) are attached to
sampling tube
TV mOnitor
3. Sample is exposed to radiation. The sample quality
Picture
Intensifier is assessed on TV monitor
4. Message is given to drilling deck or soil laboratory
about sample quality and 'usable length'
5. Radiographs of whole sample length are taken (one
radiograph gives a picture of about 40 cm of
TV cham
sample)
6. Sample is sent back to soil laboratory
A B 7. Radiographs are developed and soil is later
~. Alprofile
described in detail
8. Description from radiographs is given to soil
~Tr.Ck.n laboratory for use during visual sample description
steel balls and allocation of specimens.
SECTION A-A
den stress and saturation of the porous fil- boring to evaluate the undrained shear
ters. All further steps were regulated by strength of cohesive samples. Anisotropic
and logged by the microcomputer. At the consolidated undrained tests were per-
end of the survey, data were transferred to formed with the initial horizontal and verti-
NGI's main computer for final data plotting. cal stresses equal to the estimated in situ
stresses. A number of samples were further
Direct Simple Shear Test Apparatus and consolidated to represent the state of stress
Procedures under the weight of a gravity platform.
Efficiency, Operational Statistics and ous cone penetration test performed at loca-
Cost tion CPT 2001 to a depth of 43 m. CPT 2001
is a typical profile for the normally consoli-
The efficiency of a soil investigation may be dated clay of the Troll East Field with point
reflected in a statistical evaluation of per- resistance qe, generally increasing linearly
formance for the various types of opera- with depth from 0 MPa at seabed to 2 MPa
tions. For comparison purposes the perfor- at 43 m penetration. The dynamic pore-
mance units used are hours per meter of bor- pressure ratio (defined as the excess pore
ing or meter of cone testing (Amundsen and pressure generated during testing divided
Lauritzsen, 1982). To enable such an evalua- by the measured or total cone resistance) is
tion, however, a detailed breakdown of typically in the range of 75 to 100%. Figure
activities must be made. The activity is 15 presents the combined point resistances
given as a percentage of the time spent in for CPTs performed in Area 2. Point resis-
the field, exclusive of weather downtime tance from three profIles have been
periods. superimposed on one figure and show con-
Some 75% of the field time was spent sistency and uniformity in the soil condi-
operating. Actual drilling and sampling and tions.
in situ testing consumed 51%, pipe and
equipment handling 22% and positioning Dissipation Test Results
2%. The remaining 25% was standby time
due to the maintenance and testing of Several dissipation tests were performed
state-of-the-art geotechnical equipment and during the surface cone tests. Penetration
some malfunctioning and repair (19%), drill- was halted so that pore pressure and cone-
ship (2%), positioning problems (1%) and point resistance could be recorded for a
other factors (3%). Applying the above data period of up to 30 minutes. Results of two of
to the production figures, 1.05 hours were these tests performed in CPT 2003 are pre-
required per meter of boring and 0.36 hours sented in Fig. 16. When restarting penetra-
were required per meter of surface CPT. tion after each dissipation test, an increase
These figures compare well with similar in sleeve friction was recorded indicating
surveys. The 1983 soil investigation at Troll soil set up around the cone sleeve. This
West Field, on behalf of Norsk Shell, increase in sleeve friction for a number of
resulted in 1.16 hours per metre and 0.38 dissipation tests is shown in Fig. 16.
hours per metre, respectively.
In terms of cost the statistics also show Temperature Test Results
that, on average, in 1984 each CPT cone
lasted 164 m penetration before being lost or Two continuous cone penetration tests were
replaced. Similar statistics show the 'work- undertaken using a temperature cone.
ing' life of drill pipe, cone rods etc. The total Temperature is measured by a thermistor,
cost of the soil investigation was 18 million mounted at the base of the cone shaft, which
NOK with approximately three-quarters has an accuracy of about 1°C. The results of
spent on the offshore operation; the remain- two temperature profIles are tabulated in
ing quarter related to onshore laboratory, Table 2. The measured temperature
data interpretation, administration and increases by about 2°C over a penetration of
reporting. 40m.
NOTES
1 q. C IS adlus ted tor hydrostatic
pressure
2 q . IS correcled toInclude th e
effec t Olex.cesspore
pres su re generated durtn g
te sting
" .2
Poinl resistance
,. ••
•• ••
•e .e
>0 20
n
i
22 (
>4 24 ",
~
POin t resistance
2. \
~.
>e 28
i •.
30
I
30
J> J
\
34 \
\
36
\ J'.'" 36
\ ~
3e \ 3e
\
'0 1 .., 40
I
\ 'J
42 ( '2
and adjacent to Boring 2.6. The results for time was not available to record the residual
both types of test show the mobilized shear shear strength.
stress in kPa with rotation of the vane Results of all downhole tests undertaken
blades. Peak shear stress occurs at about 7 in Boring 3.2 and all surface tests adjacent
to 8 degrees of rotation. Due to the required to Boring 2.6 are shown in Fig. 17.
slow speed of rotation of the vane motor,
SOIL INVESTIGATION AT THE TROLL EAST FIELD 179
2 .................
4
r-r--.
6
8
1()1 1()' .0'
LOG T.ME - SEC
10
CPT2003-Oeptll30m ~ 0
::7 r-
E 12 :It>
o ~
~ ! 1"---1'--
-....
i..
14
40
~
rn
3: 16 60
9w ~
1
III 18
z
o 80
~ 20
a: 100100 10 ' 10' I(P
>- LOG T'ME- SEC
~ 22
~ Sleeve friCt.... eaI . . . .ing dissipation tests
..
24
CPT Pw>. ~~o ~~':I.~ttest ~fn:t_
MPo ~'of_
26 100' 9 0005 0007
21 DOl. 0015
TABLE 2
Temperature test results
8 10
2
...<
...J
••
W
(/)
~ 2or-----+---~+_----4_----~ 4
9
w
III
Z
~ 30 r-----+-----+------...--+----~ 6
I!:w
zw
11.
40L-----L-----~----~----J 8
40r-----.-----.-----,-----~ 40
I I
BOREHOLE VANE SHEAR TEST HALIBUT VANE SHEAR TEST
BORING 3 .2 ADJACENT TO BORINF 2.6
30
I
Tesl penelration
Aillesis 4m
/"'0.,
10
v~ ~
TeSI penelralion
~
- - 1 7.9m
-.-.- 18.8m
---- 19.7m
00~----,~0-----~~--~~3O~---J40 00 10 20 30 40
ROTATION - DEGREES ROTATION - DEGREES
TABLE 3
Soil properties
Typical values
Depth
,oil interval W WL Ip YT Su
ayer m Soil description % % % kN/m 3 kN/m 3 % Clay OCR
0 Clay, very soft to 50-60 55-75 35-50 15.0-17.2 5-60 35-50 1.3-2.2
18-26 firm, high plasticity, increasing increasing
slightly overcon- with depth linearly
solidated with depth
18-26 Clay, slightly sandy 17-24 32-37 17-23 20.0-21.0 50-300 23-31 1.3-1.8
44-95 and silty, firm to increasing increasing
hard, medium plasticity, with with depth with depth
Slightly overcon- depth
solidated
44-95 Clay , sandy, silty, 20-28 35-50 25-30 19.0-21.0 400-1000 34-43 2-6
hard to very hard,
overconsolidated,
glacially reworked
material
.,
IIA T£R CONTENT. % lJI()RAIr-I'O StEAR ST RENGTH. ,Nil"
10 40 60 80 40 60 80 100
.......---ac· ..
,
.',.'
' ~.
:=t 11.0
11.0
LABORA TO RY UNORAINED
SHEAR STRENGTH
'j ,J.;,~.
1 Ie
I1.S Of TERHINA TIONS
11.9
169
11.l Q ( AU, TEST
1'-
16.0
- OSS TEST
20.0 • LABORATORY VANE
19.9
".tl 19.8
10.S
10.5
...........
20.2
20.6
:; ':
20.1
20.9
r
10,4 IN SITU UNORAINED
20.S
SHEAR STRENGTH
20.2
OETE RHINA liONS
~
...............
a
eSURFACE VANE
20.0
20.1 o $OOWN- THE-HOLE
............... VANE
o PUSH- IN -PRESSUREHETER
,
19.9
40 .~ 20.2 , " RA
NGE INTERPRETEO
\:' FROM PCPT's
4S :'
.\ + 20.1
10.4
9
o
~
20.2
••
20.6
so ~
• • 2o.a
o
SOIL PROFILE
UPPER SOm,
AREA 2
Fig. 19 Typical borehole profile
182 ADVANCES IN UNDERWATER TECHNOLOGY
e
a
I'
000 5 I'
o 1\
10
,
\ I
~~
o oe e I
D I I
,
10 I I
\
o
E \ \
....
",,'
I
I.
\
I~ OCR = 1.8
l!; 15
c
e
,
'v< I
OCR = 1.4
\ \
\ , ,,
••
D e 20
I
£7
o
..". ,
'. \
e
25 P'. \ \ ',
\
, ,
DO
\
.
30
D \
\
__ \
\
30L-________ ~ __ ~ ~ ______ ~
e
D Po' = EFFECTIVE OVERBURDEN STRESS
10
r---
P. 100
p', I
/ fo-
1000 0
-- I
-l.
P.
~
100
IP
200
14
12
/1 \ ~
'/ 1\
~
4 - - 1-;-- 10
[I
A \ 1
--1- I\ I
\ I 1\ I
\ \ \ \ 1
\ \ \ -1-1-- - I~
\ \ f--+--
I" - - \
10
\V H \ t-- -
- -
- 1-----:---
\
I~ I
---
12
\ r- ----
f----- ~ I
14
i
010 000100 001000 010000
k. m/year
r-- \
I -
- -
--
y/ (,
1-
.j 2
1\
-
~
11 --~
collection, interpretation and feedback, and land-based system. Figure 22 illustrates this
programme review. On board, laboratory process. During the operational process the
data were collected and reported daily. number of CPTs was reduced due to gen-
Similarly all in situ test data were edited, eral uniformity of stratification throughout
plotted and reported within 24 hours of each the Troll East Field in the upper 50 m. The
test. The operation involved frequent com- amount of surface vane testing (Halibut)
munication with the home offices where was increased due to some regional varia-
geotechnical engineers and geologists at tions in strength of the upper 5 m. A number
Norsk Hydro and NGI participated in the of boreholes were added due to detection of
continuous process of comparing all new a moraine layer in Area 3, stretching south
data from the field laboratory and in situ east from this area. An example of the field
testing with the data used as a basis for the programme performed in Area 2 is shown in
original field programme. Advice for modifi- Fig. 23.
cation of the field programme was relayed to Geotechnical data reporting was part of
the ship in writing. Communication was the administrative reporting system. Here
made possible by extensive use of telefax- the team leader of NGI reported every 24
imile through the automatic Nordic Mobile hours at midnight to the client's representa-
Telephone system. This same telephone sys- tive. This report consisted of a detailed log
tem was used for telephone conversations of activity, a summary of off-hire periods
and is as easy to operate as the ordinary with descriptions and explanations of inci-
184 ADVANCES IN UNDERWATER TECHNOLOGY
CONTINUOUS
REVISION OF
DRILL AND TEST
PROGRA ..
SA .. PLES: I
SA .. PLES
* PISTON SA .. PLES
"* PUSH SA ..PLES
* HA .... ER SA .. PLES IN SITU TEST DATA:
* GRAVITY CORING
"* CONE TESTING
* VANE TESTING
* PRESSURE METER
TESTING
TO ONSHORE
TESTING:
* GEOTECHNICAL
"* GEOLOGICAL
* GEOCHEMICAL
c-
ACKNOWLEDGEMENTS
REFERENCES
1. Amundsen, T. and Lauritzsen, R. 1982. Time
eo
and cost planning for offshore soil investiga-
tions. In Proceedings, Third International
al
_r.-----, Te 101..
Conference on the Behaviour of Qf"f"shore
'JJ'
Structures, Boston, Vol. 2, pp. 417-424.
2. Ehlers, C. J. and Babb, L. V. 1980. In situ
Fig. 23 Soil investigation carried out in Area 2
soil testing: remote vane. McClelland
in 1985
Engineers' Soundings, Spring 1980.
- 3. Lacasse, S., Iversen, K., Sandbaekken, G.
and Morstad, P. 1984. Radiography offshore
positioned on location in the dynamic posi- to assess sample quality. Nordiska Geotek-
tioning mode. nikermotel, Linkoping, Sweden, June 1984,
Technical specifications were established Vol. 2, pp. 575-584.
for sampling and in situ testing equipment 4. Lunne, T., Christopherson, H. P. and Tjelta,
and procedures. Implementation of these T. 1. 1985. Engineering use of piezocone
results in North Sea clays. To be presented at
specifications enhanced the quality of 11th International Conference on Soil
results obtained from the investigation. Mechanics and Foundation Engineering,
State-of-the-art sampling and testing San Francisco.
equipment were used, and a comprehensive 5. 0stmo, S. R. and Kleiven, A. 1984. Shallow
soil-testing programme was carried out off- seismic investigation Troll Block 31/3, 5 and
shore for the first time. The total cost of this 6. Report No. ST8393.01, Norsk Hydro,
project was 18 million NOK with almost 75% Statoil and Saga Internal Report.
of this amount allocated to fieldwork; the 6. Reid, W. M., St John, H. D., Fyffe, S. and
remainder was allocated to laboratory test- Rigden, W. J. 1982. The push-in pressureme-
ing, interpretation and reporting. The ter. Symposium on the Pressuremeter and its
report describing the soil design parameters Marine Applications. Paris, April, pp.
247-261.
186 ADVANCES IN UNDERWATER TECHNOLOGY
7. Sandbaekken, G., Berre, T. and Lacasse, S. 8. Semple, R. M. and Johnston, J. W. 1979. Per-
1985. Oedometer testing at the Norwegian formance of Stingray in soil sampling and in
Geotechnical Institute. Presented at ASTM situ testing. In Proceedings, Conference on
Symposium on Consolidation of Soils Offshore Site Investigation, Society of
Laboratory Testing, Ford Lauderdale, Underwater Technology, London, March, pp.
Florida, January, 1985. 167-182.
12
Use of the P AM Self-boring
Pressuremeter and the STACOR
Large-size Fixed-piston Corer
for Deep Seabed Surveying
J-B. Fay, R. Montarges, P. le Tirant and F. Brucy, Institut Franc;ais du
Petrole, France
To meet the needs of geotechnical soil sur- the seabed of more than 50 m. The system is
veying at great water depths, two systems currently operational and can be used at
have been specially developed and investi- water depths of up to about 1000 m.
gated by the Institut Franc;ais du Petrole in The STACOR corer was designed in col-
recent years: these are the PAM Self-boring laboration with Elf-Aquitaine and Total for
Pressuremeter and the STACOR Large-size taking core samples at water depths of sev-
Fixed piston Corer. eral thousand meters. It is made up of a core
The PAM pressure meter consists mainly barrel 20 to 30 m long, a weight-stand of up
of a framework set on the seabed, a control to 10 tons, and a piston held in a fixed posi-
cabin on the surface and an electric umbilical tion while the core barrel penetrates into
cable connecting the cabin and the the ground. The fact that the piston is kept
framework. The self-boring probe of the fixed enables relatively undisturbed core
PAM is derived from the P AF 76 used by samples to be taken. These samples are 11 cm
the Laboratoires des Ponts et Chaussees for in diameter and have a length equivalent
onshore applications. The PAM provides to the depth to which the core barrel penet-
two types of information: self-boring logs rates the ground. Comparative results have
recorded in real time, and pressuremeter been obtained for core samples taken by the
measurements performed everyone or two STACOR operating normally (with a fixed
meters. The PAM has been operated suc- piston) and by the same device operating
cessfully on various offshore sites in the Gulf like a standard oceanographic corer. The
of Lions and offshore Spain on the Casa- results show the considerable advantages of
blanca and Montanazo fields, at depths rang- the fixed-piston STACOR for taking undis-
ing from 50 to 625 m, with penetrations into turbed samples.
187
188 ADVANCES IN UNDERWATER TECHNOLOGY
Fig. 1 The PAM self-boring pressuremeter Fig. 2 The PAM: frame and mains equipments
THE PAM AND STACOR 189
PRESSUREMETER
THE LPC/IFP PROBE
o ---
C :m
.Z--
DP vuHl (65m)
elKtu(: ..,.nbihul
1 J
.-::?-~~.
... .
;;
Fig. 7 Self-boring logs Q and Pi
..., / ,
,,
i
I
r- - - i-
!
t • tangent Go and secant Gp2 moduli (for 2%
deformation) are also used directly for
calculating foundation design by the pres-
I
I .,
j .
I
··
suremeter method.
"e ~"/c.'~~~>(Ioot;
• pressures P 2, P 5 and P 20 characterize the :'!t 6~/~
;' /'b'O / ~e
" \
\, -
~.
\ '.
\
siltet sable
\ silt and slInd
\ 1\
'"
\ 1\ '"
,, "-
\
. "JJ,i ~
.I
~
cailloutis
r'..
pebbles
"'-
\
rr
~
~
Free fall
Golfedu lion re leasl~
SOmd'eau
"g System
SOm w.t.rclttpth · 6m i"
0 .8 o ~F~~~t
1.--- ---,
5.4 0
\
Pull-out
coble
:-
tt Corer pipe
4.6 0
\/
~
t o.60
l/ ,
V
-~
I
/~ r-- I
Bose plate
Piston
3.80 I I
I
I
0 .4 o
Counter
wei~nt
I
/ Descent Free fall Penetration
3.00 0 .2 o
~/ Fig. 12 Principle of the STACOR
/
.,'
10/ 6vV , %,
..,/
5 10 15
meaningful results. The steady development
20 25
of petroleum production on the continental
Fig. 11 Pressuremeter curve and shear curve slope, at water depths of several hundred
meters, raises two types of geotechnical
problems linked both to the sizing of new
suremeter curve. In soft or medium stiff structural foundations (tension-leg plat-
clays, the peak of the shear-strength curve forms, guyed towers, articulated gravity
gives the maximum shear strength 1:[ gener- towers) and to the evaluation of the instabil-
ally attained for a volume deformation of 1 ity risks of seabed slopes and their consequ-
to 2% (Fig. 11). For these soils, the C u value ences for seabed installations. The offshore
obtained from Menard's correlation self-boring pressuremeter meets the needs
corresponding to the surveying of such soils.
C = P20 - Po
u 5.5
Foundation Design
is in good agreement with that deduced from
the C u triaxial test. In a way similar to the pressuremeter rules
drawn up by Menard for the standard pres-
Application of PAM Measurements suremeter test, the Laboratoires des Ponts
et Chaussees have developed semi-empirical
The PAM is particularly well suited for sur- rules using self-boring pressure meter
veying fine sediments (soft or very soft soils results for calculating the limit load of
with a cohesion C u < 50 kPa, loose carbo- onshore or offshore foundations (Baguelin et
nate sediments which are extremely sensi- al., 1978). These cover:
tive to disturbance, etc.). It thus covers a • lateral friction of piles;
large portion of seabed soils - especially • piles subjected to horizontal loads;
those beyond the continental shelf - in • limit punching pressure at the base of
which conventional techniques of core sam- foundations;
pling or in situ measurements cannot give • settlement of shallow foundations.
194 ADVANCES IN UNDERWATER TECHNOLOGY
Slope Instability Evaluation (a) the probe driving force can be corre-
lated with the penetrometer tip
An evaluation of the instability risks of soft resistance (CPT);
soils on the continental slope is indispens- (b) coefficient P deduced from the pres-
able prior to the installation of petroleum suremeter curve provides a very
structures on the seabed. This evaluation is reliable identification of the soil,
generally made by using the limit equilib- equivalent to that obtained from
rium approach requiring solely the deter- core samples;
mining of the intrinsic parameters (C ' , 'P') of (c) pressure P20 (close to limit pressure)
the soil or the undrained cohesion CU. Soil is directly involved in calculating the
surveying on the continental slope with the bearing capacity by the pressure-
PAM thus determines the C u so as to inves- meter method;
tigate the instability risks of subsea soils (d) pressure Po can be used to deduce an
under undrained conditions. approximate value of the coefficient
Ko of earth pressure at rest;
Application of the PAM for Planning (e) maximum shear strength r:f of the
Deepwater Production Facilities soil is derived from the pressure-
meter curves;
Production from hydrocarbon fields on the (0 in situ soil deformation moduli can be
continental slope, at water depths of several obtained solely by the pressure-
hundred meters, raises the problem of meter.
investigating soils which are often soft or
very soft, prior to the installation of risers, Furthermore, the quality of the data
anchoring piles and flowlines. At such water acquired by the PAM is independent of the
depths and in this type of soil, the PAM, water depth, despite the fact that the dis-
which is designed for water depths of up to turbing of core samples commonly increases
1000 m, has a great many advantages of considerable with water depth on account of
both a technical and economic nature com- both sampling conditions and decomposition.
pared to other soil investigation methods
currently in use. Economic Advantages of the PAM
core samples. This condition, which was by a cable running over two pulleys situated
previously fulfilled by the corer developed at either end of the core pipe. When the core
by Kermabon and Cortis (1969), was a prime pipe penetrates into the soil the piston
consideration in the STACOR corer. remains at the level of the baseplate.
Principle of the Corer The STACOR corer (Fig. 14) mainly con-
sists of (Montarges et al., 1983):
The STACOR is a free-fall gravity corer
with a fixed piston (Fig. 12). It is lowered at • the corer pipe .with its liner;
the end of a hoisting cable wound around a • the weight-stand;
winch. At a predetermined distance above • the baseplate and the fixed-piston sys-
tem;
the seabed, between zero and several met-
ers, a triggering system causes the free fall • the free-fall releasing mechanism.
of the device so that the corer pipe can pene- The corer pipe is made up of modular
trate into the soil. Then the corer is pulled pieces decreasing in thickness towards the
out of the soil by a cable which had been cutting edge. This pipe protects a liner made
allowed to uncoil freely during the free fall of plastic or reinforced resin. The adjustable
and penetration of the corer. weight enables the weight of the corer to be
The true stationariness of the piston, varied from about 5 to 10 tons with a corer
which is impossible to obtain with conven- pipe length of 30 m.
tional corers of the Kullenberg type,
requires the positioning reference to be
taken in relation to a baseplate falling with
the corer pipe and sitting on the seabed
while the corer penetrates into the soil. The
piston and baseplate are connected (Fig. 13) j FrH loll r~1Nn1fH}
System
"pullout cable
!!yeight·stand I W~/Qh,_s,ond
\
corer pipe
CO", pIpe
W i th lmer
- PIS/on co~
I 8QS~ p!ole
I
Cori ng Operations
Experiments of the Corer and Results A filling ratio of 95% is the criterion for a
Obtained good-quality core sample, as checked out,
moreover, by a comparison of the core sam-
Two STACOR test campaigns have been ples taken with a fixed-piston corer with
carried out from the Nadir: those taken with a Kullenberg-type corer
(or acting like one) .
• the first, in April 1982, was to test the A comparison of shear strengths meas-
reliability of the implementation proce- ured on core sample 4, which was correctly
dures and the operation of the corer; taken (filling ratio of 97%), and core 5, which
TABLE 1
STACOR test campaign results in the normally consolidated silty
clays of the Gulf of Lions
1 50 15 11.1 10.2 92
2 510 15 14.8 13.9 94
3 510 20 18.3
4 510 20 18.6 18.0 97
5 510 20 18.8 13.4 71
6 1225 20 17.0 15.3 90
198 ADVANCES IN UNDERWATER TECHNOLOGY
201
202 ADVANCES IN UNDERWATER TECHNOLOGY
Soil description
Density
I Mg / m')
• wc
PL
,
LL % Clay
Cone resistance
I MPa)
Friction ratio
%
carbonate
%
2.0 2.2 2.4 10 20 30 40 50 20 30 40 50 2 4 6 8 10 12
0
, ,' 6
•
I L
Weathered brown 17.SVR 3 / 4 ) 0
~ cP8 " "
6
stiff stony clay till . Some 8.5
2
fissures 0 •• • II 6 " 9 .1
3 00'
8 \1 l' a .8
4 ® o~ •
•• """
9.9
10.2
- - - - - - - - - - - -' • 10.9
o~
• I---------l
5
Unweathered dark grey brown • " "
15.8
6 (10 VR 3 / 2) stiff stony clay
i•1 "" 17.9
.
till. Stones mainly chalk
« 17.5
0
" 18.8
.,
, •
8
9 ®
'""" 19.3
19.5
~ • ""
]
10 ------------
S~ty
© sand
.c 11
Q.
------------ '6 "
14.3
~ 12 ® 0
As "
13
14 0
'S
0
15 0
16
12.0
"
..,.
18 Sand and grave) @ max 27.4
19
------- ----- 8 •
20 %
Dark grey stiff stony
21 '0
..
clay till; but fewer
,tOMS than@ • ~ 19.2
22 0
21.6
23 c9
•~ <- 26.4
24 ® 19.0
2S
that the tills were deposited directly from, (as determined from 865 mm diameter plate
Dr close to, the ice, mainly subglacially, but tests) of between 90 and 140 kPa, which is
also that some flow and melt-out compo- comparable with the mass strengths meas-
nents may be present. The predominance of ured in heavily overconsolidated London
clay tills and their consistency over the site, Clay. The few tests carried out below a
as well as the changing depths and thick- depth of 20 m gave much higher strengths of
nesses of sand and gravel, were confirmed between 210 and 280 kPa, and this lower till
by static cone tests. was probably deposited during an earlier
The relatively uniform nature of the clay advance of the ice.
tills down to the levels investigated can be Oedometer tests indicate that the
seen from the typical profIle in Fig. 1. There unweathered clay till down to a depth of 10 m
is a distinct change of colour from brown to has been subjected to overconsolidation
very dark greyish brown at a depth of 4-5 m, pressures approximately twice the present
and this is considered to be due to a zonal overburden pressures, but estimates are
weathering of the till. The tills are inor- diffIcult in these well-graded soils. Pore-
ganic, well graded (poorly sorted), fairly water pressure measurements made during
dense and of low activity (0.4-0.6). For undrained triaxial tests on samples at their
depths between 3 and 17 m (till B), the bulk in situ moisture content show moderate to
densities obtained from good-quality 98 mm strong dilation. Typical curves showing the
diameter pushed samples are almost con- stress-strain behaviour of the tills are
stant with values between 2.2 and 2.23 shown in Fig. 2. A fuller description of the
mg/m 3. Over this depth range, the percen-
tage of clay-sized particles is usually within 5OOr-----~----~----~----~-.
site geology, soil properties and engineering pressure in the measuring cell was increased
parameters can be found in Marsland and in equal increments and maintained at each
Powell (1985). pressure for a given period of time. In 'stan-
dard' tests the pressure was increased in
50 kPa increments which were maintained
EQUIPMENT AND TEST for 1 min. Volume change readings were
PROCEDURES recorded at intervals of 15 s, 30 sand 1 min.
Pressure corrections for the effects of the
In view of the stony nature of the clay tills at resistance of the membranes were obtained
the site, it was considered advisable at the by expanding the probe in the unconfined
time of the main investigations to make the state. Detailed calibration tests in which
pressure meter tests in pre-drilled pockets pressure increments of 20 kPa were main-
(the SBP being unable at the time to bore in tained for various intervals of time showed
these soils, and the PIP not having been that holding each increment for periods
developed). These first tests were therefore greater than 5 min had little effect on the
carried out using an MPM pressuremeter. corrections. All the tests described in this
paper have been corrected using membrane
Menard Pressuremeter (MPM) calibration curves obtained by applying 20
kPa increments for 5 min. This gave a total
The commercially available type GC Menard calibration time of 20-30 min, which was
pressuremeter was used. Since detailed approximately the same as the time
descriptions of this equipment are readily required to carry out the 'standard' tests. In
available (Baguelin et al., 1978), only the addition to the detailed calibrations, quick
most essential features will be described calibrations (50 kPa pressure increments
here. The probe consists of a rubber- held for 2 min) were made prior to and fol-
sheathed central measuring cell, covered by lowing each test in order to check that no
a longer sheath which forms guard cells at significant changes had occurred in the
either end of the central cell. In its deflated behaviour of the probe. Checks were made
condition the probe has an external diame- for possible volume changes due to the
ter of 58 mm and a central measuring por- expansion of the connecting tubes and for
tion 216 mm long. compression of small air bubbles trapped in
The central cell (capable of being the probe. Errors in the final pressures in
expanded to twice the original volume) is the stiff soils are considered to be less than
expanded by water under gas pressure 2%.
measured on a bourdon gauge, the volume
changes being measured by a burette. The
Self-boring Pressure meter (SBP)
guard cells are expanded by gas pressure
which is slightly lower than the central cell
pressure. The self-boring pressuremeter (SBP) used
A test pocket approximately 1 m deep and its method of operation were similar to
with a diameter slightly larger than the that described by Windle and Wroth (1977).
probe was freshly drilled for each test using The SBP is an instrument that should be
either a hand bucket type auger, or a pow- capable of being inserted into the ground
ered rotary flight auger attached to a dril- with minimum disturbance to the surround-
ling rig. The pressuremeter probe was ing soil. It consists of a rubber-sheathed
inserted as quickly as possible into the test pressuremeter section placed behind a dril-
pocket. The time interval between comple- ling head. The device has a constant diame-
tion of drilling and reapplication of the initial ter of 83.8 mm and is 1.05 m long, the pres-
in situ stresses to the walls of the boreholes suremeter section being 0.5 m long with its
was about 10-15 min. During a test, the central expanding section 0.4 m back from
MENARD, SELF-BORING AND PUSH-IN PRESSUREMETER TESTS 205
the leading edge of the instrument. Gas derson et al. (1979) and only a brief descrip-
pressure is used to expand the membrane tion of the pressuremeter is given here. The
and is monitored by a total pressure trans- Stress probe consists of three basic units; the
ducer mounted within the instrument, while pressuremeter, the pressure developer and
the movement of the membrane is moni- the control and data acquisition systems.
tored by three displacement transducers. The pressuremeter comprises a hollow
The method of insertion is similar to that 78 mm diameter stainless steel cylinder
of a tunnelling machine in that, as the 695 mm long onto which is mounted the
instrument is advanced, the soil within the 330 mm long inflatable membrane. At the
SBP is chipped by a rotating cutter and the lower end, the cylinder was fitted with a
chippings flushed to the surface. The SBP is cutting shoe which incorporated a stress-
pushed into the ground using hydraulic rams reducing step so that soil could pass unre-
which react against kentledge, casing, or stricted up the inside of the cylinder. When
ground anchors. The power to drive the used in very stiff soils or soils that had a
rams and the rotating cutter is controlled tendency to plug before they were fully
through a series of valves using a hydraulic sampled, a 50 mm diameter pre-core of soil
power pack. The cuttings are flushed to the was first removed from the base of the hole.
surface by pumping water down the inner Testing reported in this paper has been per-
rotating rods to the cutter and then back up formed both with and without pre-coring.
the 50 mm diameter outer rods. After inser- The inflatable membrane was protected
tion the SBP was left for an average of by long stainless steel strips attached at the
30 min before the membrane was expanded lower end to the cutting shoe. At the upper
at a constant rate of strain of about 1%per end, the strips were attached to a split ring,
minute. This rate was controlled by means which was free to move axially in order to
of a feedback control unit, which also limited accommodate the expansion of the mem-
the maximum expansion to about 20% of the brane. The membrane was inflated at a con-
original volume. An unload-reload cycle stant rate of 2% volume increase per minute
was included when about 2-3% cavity strain by oil pumped by the pressure developer to
had been reached. The data from the test a maximum expansion of 20% of the original
were logged by a microcomputer and stored volume. The pressure and volume of oil
on floppy disc. Total testing time generally delivered were monitored electrically. The
amounted to 20-30 minutes. volume was measured by recording the
Frequent calibrations for membrane stiff- displacement of a ram within the pres-
ness and equipment compliance were made sure developer, and the pressure by two
and applied to the data (maximum mem- piezoelectric pressure transducers, one to
brane correction was approximately 50 kPa). measure total pressure, the other to meas-
Small stones tended to collect at the ure differential pressure.
base of the hole and after several advances Membrane calibrations were carried out
had to be removed by withdrawing the SBP periodically throughout the test programme
and boring past the stones using flight (maximum correction at full expansion being
augers on a Minute Man drilling machine. about 80-90 kPa). During testing, the mem-
brane was expanded under strain-controlled
Push-in Pressuremeter (PIP) conditions (stress-controlled tests are poss-
ible). The test procedure was to expand the
membrane to a maximum volume increase
The push-in pressuremeter (PIP), now of 20% of the original volume and to include
known as the 'Stressprobe', may be inserted at least one unload-reload loop over a range
lnto undersized pre-cored holes or into the of up to 3% volumetric strain. The reload
bottom of the borehole without pre-coring. loop was usually performed at the end of the
":'he equipment is fully explained by Hen- test before final deflation.
206 ADVANCES IN UNDERWATER TECHNOLOGY
600 600
400
200
Pho
Pho Menard (a) (b) (e)
.... creep cu rve
0
o~--~----~--~~--~~
600 ~ ~ ~ ~~O~~~M~ ~ ~ ~ o
Total volume V (em 3) Cavity strain • log. ('Z,v)
Fig. 3 Typical pressure vs volume curve and deduced curves
MENARD, SELF-BORING AND PUSH-IN PRESSUREMETER TESTS 207
_ PL - Pho
200 or Cu - N (2c)
p
100
and Np = 1 + loge(G/c u )
where N p = pressuremeter constant (taken
as 6.18 corresponding to a bearing capacity
factor N c = 9.25). Equation (2c) yields the
Cavity strain <%
'limit-pressure' value CU.
Fig. 4 Position of reload cycles for different
pressuremeters for the determination of shear Shear Modulus (G)
moduli
Values of secant shear moduli have been
by Palmer (1972) which could yield full obtained using the expressions
stress-strain curves from the pressure-
meter test using the equation or
dp
T = 'd;-lo-g-e(i7L\"""V""'/V"") (1) for initial loading curves (when applicable)
where T = shear stress; p = current pres- for unload-reload loops (see Figs 3c and 4)
sure; V = current volume at pressure P; Vo
= volume at reference pressure Pho; L\V = V where E = current cavity strain
- yo; and Cu = T (max). change in radius of borehole due to pressure (P - Pho)
For the strain-hardening soils reported in radius of borehole at reference state (p = PhJ
this paper, the p versus loge(L\ V IV) curve
becomes essentially linear in the later stages L\E = change in E due to pressure increase
and the slope of this part of the curve has L\p.
been taken for Cu (see Fig. 3c). In later sec-
tions of this paper the shear strength from
.~quation (1) will be referred to as the TEST RESULTS
'Palmer' value.
In situ stresses measured or calculated for
Limit pressure the site are shown in Fig. 5 a and b. The
'""ibson and Anderson (1961) analysed the estimates of in situ horizontal stress in Fig.
lse of a pressuremeter test in an elas- 5b were made using results from oedometer
208 ADVANCES IN UNDERWATER TECHNOLOGY
't.
0 _, xX.
7 ~ \ \
~
8 0
8
8
9 I
\.
\
\
9 9 o .~\\ .
~: • \\
" ~''', \
10 , \ o
Pvo 10
11~
u \
I/"" . 0 . \\
:~:
11
12 12
\ . •
Pvo
~13 ~
(, \) \ 13
\
13
'4 \ {
\ \15 ,
14 14
,
\
6
15 :\ 15 ,/1>6
16
! \.\ \ 16 16
17
18
19
~
:
"
/
"
.
\.
\
17
lIo
(b)
[]
•
Oedometer tests
17
\ i \
21 lit Pko Self boring pressureme1er
\
22
23 \ \
\\
24 (81 \
25 ~--------------------- Fig. 5 In situ stresses
tests (as in Schmidt, 1966), spade cells (as in not taken, this problem can also allow the
Tedd and Charles, 1983), and suction meas- engineer to weight the interpretation
urements on pushed samples. The results towards a desired answer, if this is known.
show reasonable agreement and their mean The values of P ho assessed from the SBP
has been considered as a 'best estimate' of using the 'lift-off method, and shown in Fig.
P ho for comparison with the pressuremeter 5c, are seen to scatter quite wildly. In gen-
results. eral they considerably overestimate the best
Estimates of P ho from the MPM and SBP estimated values. This may indicate the dif-
tests can be seen in Fig. 5c. The 'original' ficulties of the self-boring technique in these
Menard approach gave values significantly soils and the possibility of significant over-
lower than the 'best estimate', whereas the stressing during installation (in the extreme
Marsland and Randolph method gave values tending towards the PIP test). It should be
closer to the 'best estimate' but in general noted that in some cases it was not easy to
still lower. The application of the Marsland decide when significant movement of the
and Randolph method to tests in strain- membrane first occurred on the basis of the
hardening soils, such as those in this paper, average curve of three displacement arms.
is difficult. The difficulty is due to the fact However, this average curve in general
that the stress-strain relationship in these showed movement when the first arm
soils shows greater curvature at relatively moved, and the assessments therefore had a
low strains than in London Clay, for which tendency to a lower bound estimate.
the method was developed. If great care is Typical corrected curves of pressure (p)
MENARD, SELF-BORING AND PUSH-IN PRESSUREMETER TESTS 209
Ca,,-,ilV stra in e %
against cavity strain (E) are given for MPM, tests show the greatest variation. It should
SBP and PIP tests, over the depth range be noted that the SBP curves show shapes
6-8 m, in Fig. 6 (unloaded-reload loops ranging from the flatter MPM results
have been omitted for clarity). Scatter is through to the steeper, over-stressed, PIP
evident in all the plots, although the SBP tests (see Fig.6a).
900 r---------------------------~
17_61
800 800 18.51 16.451
(6.81
700
600
100 100
.... ig. 6b Pressure vs cavity strain for self- Fig. 6c Pressure vs cavity strain for Push-in
boring pressuremeter pressuremeter
210 ADVANCES IN UNDERWATER TECHNOLOGY
. - - - -- -- -- - -- - - - - , 1200
1100
Aange of 1000
IImh pr••sura,
Rang. of {
Iimil P'e; urH 900
800
1/
II 700 700
600 ..
600
...~c
500
400
: ' 500
400
~
300 300
(6.6) 17.01
18.51
17.712 200 200
16.7)2
100 100
-~7-~_8~-_~5-~~~-_~
3 -~
_2~-_~'~00
log .. ,~ )
Fig. 7b Pressure VB loge volumetric strain Fig. 7c Pressure VB loge volumetric strain
for self-boring pressure meter for push-in pressuremeter
MENARD, SELF-BORING AND PUSH-IN PRESSUREMETER TESTS 211
2
0 00 0
4
00 0 0
6 o 00
o
o
8
o 00
] 10 o 0
d, o
12
" o
i
.ci
" o
14 "
Cl 16
18
"\ :\
20
22
(al (bl (el o
24
o
26 "
Plate tests Cu Pressuremeter tests Cu = eqn 2 Pressuremeter tests Cu = eqn 1
Fig. 8 Comparison of undrained shear strength from pressuremeter and plate tests
Marsland and Powell, 1979, for detailed Figures 9 and 10 show plots correspond-
interpretation). A mean line drawn through ing to those in Fig. 8 for the SBP and PIP
this data has been used to define a typical results. The limit-pressure approach again
shear strength profile and has been used in gives the more consistent results. However,
subsequent figures as a marker. the scatter of shear strengths from limit
It is not suggested that the plate test val- pressures is now greater, partly as a result
ues of shear strength are 'the values', as it of the difficulty of extrapolating from the
must be remembered that shear strength smaller final volume expansions. There is
has no unique value and is test dependent. also some reduced scatter in the Palmer val-
Possible reasons for differences between the ues, especially for the PIP tests. This is felt
plate and pressuremeter values of shear to be due either to the reduced or more 'con-
strength are discussed elsewhere (Powell et sistent' disturbance that possibly occurs
al., 1983). Undrained shear strength Cu (kPa)
6
ments on Fig. 7 are again relevant (Figs 8,9
8
and 10). Figs 8 band c show that for MPM
tests much more consistent results are I 12 .
10
\
cussed elsewhere (Bagelin et ai., 1978; Pow-
\
20
ell et ai., 1983), and it is sufficient here to 22
Bay that the method is particularly sensitive 24
(a) (b)
c" Undrained shear strength C" (kPa) c" 12. A certain degree of agreement is evident
00 100 200 300 0r----=-:;:"----"200r-------'3"T00"--..:::400r----'5"T00~ between the different devices. The SBP
results, however, show the greatest scatter.
4 The MPM tests had only a few points on
6 their reload loops and so any hysteresis
8 effect was less clearly defined, giving little
10 difference between the 1 and 2% values.
I 12 However, Fig. 12 seems to indicate that
.
.;,
.ci
.t=
14
some of the differences shown in Fig. 11 can
15.
~
0
16
be explained when shear moduli are com-
18
pared over similar ranges. It would be
:\
20
interesting to see if even better agreement
22
24
(a) (b) could be obtained if all reload loops were
26
done towards the end of the loading stage of
the test, especially in the MPM and SBP
Pressuremeter tests C . . = eqn 2 Pressure meter tests Cu = eqn 1
tests when initial disturbance might have
Fig. 10 Comparison of undrained shear less effect.
strength from push-in pressure meter and plate In Fig. 13 reload shear moduli from the
tests pressuremeter tests (the PIP range from
Fig. 12) are compared with values measured
with the SBP and PIP tests as compared to from unload-reload cycles on 865 mm
the MPM tests. diameter plates (Marsland and Powell,
Shear moduli from all the pressuremeter 1985). The values from the plate tests are in
tests are presented in Fig. 11. The MPM general higher than those from the pres-
results for initial and reload moduli were suremeter. However, if the average values
calculated over a stress range of P ho to P ho + for shear strains under a plate test (Simpson
cu. The SBP and PIP moduli were calculated et al., 1979) are considered, then the plate
over their full unload-reload loop, irrespec- test reload loops are over about 0.5-0.75%
tive of stress or strain range. A similar basic shear strain. In Fig. 13 additional values of
pattern is evident with depth, the MPM ini- shear moduli for PIP and SBP tests have
tial values forming a lower bound with MPM been plotted for 0.5% volumetric strain
reloads as an upper bound and PIP and SBP (which is approximately 0.5% shear strain at
scattering in between. It has been shown small strains). Considering the limitations of
(Reid, 1982) that in general the values of the above approach, the agreement with
reload shear moduli are unaffected by the plate test values is somewhat surprising and
test stage at which the reload loop is per- interesting; it must be stressed that this is a
formed, provided the loops are over similar very tentative comparison. Recent work
stress or strain ranges. However, the mod- (J ardine et al., 1985) has shown remarkable
uli values are affected by variations in the agreement in shear moduli values from
strain range considered (e.g. Fig. 4). For laboratory tests using internal strain meas-
these reasons it would seem acceptable to urements on triaxial samples, and from
compare reload moduli values from the plate tests when comparisons are made over
three types of pressuremeters, irrespective similar strain ranges.
of location within a test (Fig. 4), provided Data from the SBP tests is replotted in
similar strain ranges are used. It should be Fig. 14 as the ratio of Palmer C u to 'limit-
noted also that in all tests the reload loops pressure' C u against the 'overstress' ('lift-off'
were kept within a pressure range of 2c u , as pressure minus 'best estimate' of P hJ. A
suggested by Wroth (1982). marked trend is evident with the results
Secant reload shear moduli for 1, 2 or 3% tending to those deduced from PIP tests at
volumetric strain ranges are given in Fig. higher 'overstresses'.
M enard pressuremet er Self boring pressuremeter Push ·j n pressure meter Menard pressu re meter G(MPal Self boring pressuremeter G(MPal Push in pressuremete r G (MPal
G (MPe) G (M Ps) o 10 20 30 oI 10 20 30 0 10 20 30
i i
0 0I .-I -- 10 20 o 10 20 30
I
I
2 MPM SB P PIP
c SBP
2 MPM PIP
3 \).
3 .•
0 " \ '\.. ~ ~
4 .. "" 6
0 c '" / o.
4
•• 0 co '"o . /
" 5 I~ , j /
• 0 ""6
5
· 0 '" 0
6
0'1".
~
c
6 •• 0 " 'r ~ •
\\
7
\\.
' / 1/
00 c. o , W
7" D
• • •0
" 8 \ \
• • 0 0 c • .\. \..
8 c " .......... ~ ;--
f~ -'______
9 . ..... --- .
c c " /
9 • 0 co . 10 /
/
./
• o. Vi> •
10
0
11
11
• I I I
c 12
12 • 0 " E
13 I I I
V6 .
I
I C •
-ci>
13 0
c . n 14 \ \ I
14 J: \
• 0 I .\ \'1> 6 _ _ 6
c.
. IS
0 ~-
15 • i> \
16 "-.."- \ '- "-
° v I>
16 •
0 17 I
17
i>
18
---
I / / I I
I
,
I
18 I
19 I / ~'A6 6
19
. 20
I / ,
20
21 ~
21 • 0
/
"'. '\ "II
v i> 6
i> / ./
2T i>
22 • 0
.. 23 0 •
.""
23 0
\\
• 24 ,
V6 6
24 \
\, 6 '..
•
25 • 0 " 6 2t
26 \ \ 'v
6
26
27
0 ____ _ .1 ..
27
28 L- ~ - - -----
• G @ 1% VOlumetric strain • G @ 1% Volumet ric strain 6 G @ 1% Volumetric stra in
• V alues from initial slope
o Values from reload cycle o G @ 2% Volumetric str ain D G @ 2% Vo lumetri c strain .. G @ 2% Volumet ric str ain
v G @ 3% Vo lumet ric st rain
""
Fig. 11 Shear modulus vs depth Fig. 12 Shear modulus (G) vs depth. Shear
modulus calculated at different values of
volumetric strain (~V/V)
214 ADVANCES IN UNDERWATER TECHNOLOGY
Shea, Modulus G IMPa l -3e,'
o 10 20 30 40 50 2 _ '~-----------------------'
•
Ith~
Plate reload • 2.2
l!Y
2~
PIP 0.
S8P v 0
2.0
4 11 0 •
~;,
6 \01/
\.
~
1,6
.~ c
---
1/ .
; \
8
I.'
;,.. ~
10 1.2 - -- ..... _ A ___ ____ ___ __ _
]
•
• 1.0
____ ~:~_~~_~__L
I
12
I
..,.;, 0,8
"
-"
a~
Cl
14
\ \ 0,7
100 300 400
I/ " "
At low values of overstress, it becomes
""~
20
SSP MPM
"", .
..... ..
300 l> 0 eqn
..
~ .. • eqn
~ 250 . ~
•• •
..
-5i 150 o -- _
o 00
00 0
• •
0 _ _o o
-.-. -
ic: 0'"
o 0 '" '" o
'§ l> ~%~~§~
-g 100 Range of plate tests .A' 8~@~
:::>
50
Increasing disturbance
OL--~--~-~--~-~--~-~--~-~--~-~--~-~
o 2 3 4 5 6 7 8 g 10 11 12 13
Shear modu lus G (M al
and SBP tests, which both show the suscep- pressure values meant that shear strengths
tibility of the Palmer approach (equation 1) deduced from them, using a pressuremeter
to the value of initial shear modulus which factor N P' were also consistent. In using this
itself is an indication of the degree of distur- limit-pressure approach for shear strength,
bance. it will be necessary to consider the assess-
A similarity between Figs 14 and 15 is ment of Np and what G/c u ratio is relevant
evident. PIP tests with their marked over- for a particular deposit (also at what strain
stress or 'pseudo' high initial shear modulus level the shear modulus G should be asses-
would fit into both patterns plotted to the sed).
right. They further highlight the susceptibil- All pressuremeter types led to values of
ity of the Palmer type of analysis to test shear strength from the Palmer approach
quality. which were sensitive to disturbance and
showed considerably more scatter than the
limit-pressure approach.
CONCLUSIONS Shear strengths were in general higher
than those obtained from plate tests, which
Pressuremeter tests have been performed were similar to those assessed from laborat-
llsing three different types of pressuremeter ory tests (see Marsland and Powell, 1985,
which were installed in the ground in mar- for discussion of laboratory testing) even
kedly different ways. The test-bed site used though K ° is close to 1. This behaviour has
was a glacial clay till, which over the upper been found elsewhere (Ghionna et al., 1983;
15 m was a relatively insensitive, well Wroth, 1984). Reload shear moduli from the
graded, strain-hardening deposit with K ° three devices were in reasonable agreement
values close to 1. when compared over similar strain ranges.
Rates of testing were similar in terms of In general, only the MPM tests could be used
total test time for all three devices, so to give initial shear moduli and these were
minimizing the problems of rate effect which significantly lower than the reload values. A
otherwise could have confused the compari- tentative comparison with plate test shear
sons (Anderson, 1979; Powell et al., 1983). moduli showed the possibility of better
Estimating the in situ horizontal stress agreements.
(P hJ from MPM and SBP test data was far The difficulty of self-boring a pressureme-
from easy. While the Marsland and Ran- ter into a stony deposit of this type is almost
dolph approach for MPM tests gave values certainly the reason for the scatter of results
tending to the 'best estimated', the proce- for P ho and shear modulus with the SBP test.
dure could become operator sensitive owing Behaviour ranging from that similar to the
to the significant curvature of the MPM test to that similar to the PIP test
stress-strain curves. Estimates of P ho from appears to have been possible with the SBP
the SBP tests using the lift-off method test. The PIP test appears to produce
tended to overstimate in situ P ho, although repeatable results which may be the result
the lower bound values came close to the in of more repeatable installation; however,
situ values. estimation of P ho was impossible. MPM tests
The test results have shown that, irrespec- were susceptible to the variable disturbance
tive of device, method of installation or test- possible during installation.
ing procedure, all three pressuremeters can Finally, it should be stated that these
yield results which tend towards the same findings will be very much soil-type depen-
ultimate or 'limit' pressure. The smaller dent. Similar comparisons are needed in
maximum volume expansions of the SBP other deposits and work is currently in hand
and PIP devices as compared to the MPM at both BRE and the Norwegian Geotechni-
led to slightly greater scatter in values of cal Institute (Aas, 1984) to extend the range
limit pressure. The consistency of limit of soil types.
216 ADVANCES IN UNDERWATER TECHNOLOGY
simple interpretation for the pressuremeter 26. Tedd, P. and Charles, J. A. 1983. Evaluation
test. Geotechnique 22(3), 451-457. of push-in pressure cell results in stiff clay.
22. Powell, J. J. M., Marsland, A. and Al Int. Symp. on Soil and Rock Investigations
Khafaji, A. N. 1983. Pressuremeter testing by In Situ Testing, Paris, Vol. 2, pp.
of glacial clay tills. In Proc. Int. Symposium 579-584.
on In Situ Testing, Paris, Vol. 2, pp. 27. Windle, D. and Wroth, C. P. 1977. The use
373-378. of a self-boring pressuremeter to determine
23. Reid, W. M., Fyffe, S., St John, H. D. and the undrained properties of clays. Ground
Rigden, W. J. 1982. The 'Push-in pres- Engng 10(6), 37-46.
suremeter'. In Proc. Symp. on the Pres- 28. Wroth, C. P. 1982. British experience with
suremeter and its Marine Applications, the self boring pressuremeter. In Sym-
Paris, pp. 247-262. posium on the Pressure meter and its
24. Schmidt, B. 1966. Discussion of 'Earth pres- Marine Applications, Paris, pp. 143-164.
sure at rest related to stress history'. Can. 29. Wroth, C. P. 1984. The twenty-fourth Rank-
Geo. J. 3(4), 239-242. ine Lecture: the interpretation of in situ soil
25. Simpson, B., Calabresi, G., Sommer, H. and tests. Geotechnique 34(4), 474-492.
Wallays, M. 1979. Design parameters for
stiff clays. In Proc. 7th European Conf. on
SMFE, Brighton, Vol. 5, pp. 91-126.
14
Comparison of In Situ and
Ship-Board Vane Measurements
on a Deep-Sea Clay
Armand J. Silva, University of Rhode Island, Kingston, Rhode Island
well-defined surface, and, with reasonable driving. Hriaud (1980) gives a detailed
assumptions regarding stress distribution, review of available methods and instru-
it is a simple matter to obtain a relationship ments.
between the torque required to produce In this paper, a short review of vane shear
failure and the shear strength. If torque testing systems that are presently available
versus rotation angle is monitored through will be presented. It is anticipated that
the full test range, it is possible to estimate references will be made to the work of the
the shear modulus and to determine residual following authors: Aas (1965), Briad (1980),
strength. The remoulded strength and sen- Doyle et al. (1971), Flaate, (1966), Lee
sitivity can also be determined if another (1979), Monney (1973), Perlow and Richards
test is conducted at the same site, after full (1972), Richards et al. (1972), ASeE (1975).
rotation to assure complete remoulding.
The main purpose of this paper is to pres-
ent and compare some data on in situ and DEVELOPMENTAL HISTORY OF
laboratory vane measurements made on a URIjISV SYSTEM
homogeneous illitic deep-sea clay in the
north central Pacific. The in situ system General Requirements
used in this study was designed specifically
for use in deep water (6000 m) or on a A more detailed description of the develop-
bottom-supported platform. In the present ment history can be found in recent papers
configuration it can take a series of approx- (Babb, 1982; Babb and Silva, 1983; Silva
imately 25 separate measurements within a et al., in press). The motivation for thein situ
depth of about 1.5 m to produce a profile of vane (ISV) system described here came
strength versus depth. For the first from the US Subseabed Disposal Program
deployment, in 5800 m water depth, the test (SDP) which is a study to determine the
sequence was reduced to 10 measurements. feasilibility of burying solidified high-level
Two large-diameter piston cores were pro- nuclear wastes within certain geologically
cessed on board the ship within 10 hours stable deep-sea sediments (Hollister et al.,
from recovery. In addition to physical prop- 1981). A similar but smaller system had
erty data and subsampling, shear-strength been designed for another project (Silva and
measurements were made with a laboratory Pekin, 1981) and the experience gained was
miniature vane device. This ocean test pro- applied in the development of the ISV sys-
vided an opportunity to compare the in situ tem. The first long-term in situ experiment
measurements, conducted under carefully of the SDP will be the In Situ Heat Transfer
controlled conditions, with those on very Experiment (ISHTE) to determine the
good quality core samples. thermal, geochemical, and geotechnical
response to a heat source in the sediment.
As currently planned, this experiment will be
BACKGROUND ON VANE SHEAR carried out over a one-year period in 5800 m
METHOD water depth in the north central Pacific
ocean. Various components (thermal sen-
There are several methods and instruments sors, piezometers, pore-water sampler, etc.)
to measure in situ strength and deformabil- will be mounted on a tubular structure
ity of marine sediments, including the vane, approximately 4.3 m long and 2.3 m high and
cone, and pressuremeter. Depending on the a 400 W isotopic heat source will be
water depth and sediment type, in situ tests implanted into the sediment to a depth of 1 m
can be performed from a sea-floor vehicle, below the sea floor. The heater and vane
from a sea-bottom platform, through the will be situated in such a way as to be out-
drill string, from a submarine, by a diver, by side the zones of significant stress influence
the impact penetration method or by vibro- from the support pads. The vane shear
IV SITU AND SHIP-BOARD MEASUREMENTS ON A DEEP-SEA CLAY 221
measurements will be made at the end of the Silva, 1983). A tank of seawater was
one-year experiment with the vane passing mounted above the test bed and the entire
within 20 mm of the heater. Comparisons of apparatus was pressurized to 550 bars and
in situ measured responses with the predic- maintained at a temperature of 4°C for a
tions of numerical models for thermal, period of 30 days.
mechanical, and chemical b~haviour will be The ISV configuration was modified for
used to evaluate the applicability of the use in this simulation experiment. Because
techniques being developed in the SDP of space limitations, this prototype (Model
(Percival, 1983). A) was shortened to a 0.6 m penetration but
The present configuration of the ISV sys- was designed for operation in the physical
tem has in part been dictated by the needs of and geotechnical environment of the actual
the ISHTE, but the same basic design can seabed experiment. Therefore the full-scale
be used for other geotechnical applications. ISV system (Model B) is very similar in con-
Based on preliminary studies, the basic cept. The Model A ISV system consists of
design requirements for the ISHTE version three major integrated packages:
or-the ISV system were defined as follows:
• vane shear probe mechanical system
Water depth (pressure) 6000 m (600 bar) (with internal volume/pressure compen-
Sediment strength 0.5-70 kPa sator);
Profile depth 1. 5 m • electronic control and data acquisition
Time on bottom 1 year system;
Sediment temperature 1.5-300°C • power supply.
It was decided to make the ISV system aut- The mechanical system is contained in a
onomous, with its own power, controller and pressure-compensated housing with pres-
data acquisition system, so that it could be sure equalization accomplished by two com-
used in other sea-floor studies. For the pensator tubes equipped with 'floating' pis-
ISHTE application, the system will be acti- tons. One end of each tube remains exposed
vated by the ISHTE platform master con- to the seawater, allowing pressure changes
troller and a subset of the vane shear data to displace a teflon piston and compress the
will be sent to this controller for eventual mineral oil compensating fluid in order to
telemetry on a surface ship through an equalize internal and external pressures.
acoustic link (Backes et al., 1981). In the For the simulation experiment, power
present configuration, the ISV system can was supplied from a source external to the
be mounted on a bottom-supported platform pressure vessel. The Model A system was
or used from a submersible. used successfully during the 30-day simula-
tion experiment while at a pressure of 550
Large-scale laboratory Experiment and bar, and three strength profiles were
ISV-Model A obtained:
(1) unheated profile prior to sediment heat-
Because of the complexity of ISHTE and the
ing and pressurization;
fact that most of the instrumentation is (2) heated profile near the heater (within
entirely new, a short-term (30 days) scale 20 mm) after 30 days' heating under 550
model (0.287 scale) laboratory experiment bar pressure;
was conducted in the 3 m diameter pressure
(3) unheated profile (within 70 mm of hea-
vessel at the Naval Ship Research and
ter) after sediment cool-down and
Development Center, Annapolis, MD. The
depressurization.
test bed for this ISHTE Simulation Experi-
ment consisted of a 1 m diameter by 1 m Comparison of data from the pretest unhea-
deep tank of saturated, reconstituted, ted and heated profiles indicates that a ten-
reconsolidated north Pacific illite (Babb and to twelve-fold increase in shear strength
222 ADVANCES IN UNDERWATER TECHNOLOGY
occurred at the midplane depth of the heater pheric pressure and are thus contained in
in the temperature zone of approximately high-pressure housings on the platform. As
200°C. This strength increase is greater in Model A, cabling between these packages
than the six- to ten-fold increase predicted is oil filled to provide volume compensation
from earlier laboratory experiments (Had- and special marine connectors are used. The
ley et al., 1980). controller is programmed with fixed opera-
tion sequences for the vane with the appro-
ISV, Model B, Used in Deep Ocean Test priate sequence selected by commands
transmitted from the ISHTE master con-
The in situ vane shear device (ISV, Model troller. In this section, some details are pre-
B) that will be used on the ISHTE platform sented on the mechanical system, followed
consists of four integrated packages: by a discussion of the electronic controller
and information on the power supply.
• mechanical system; Experience gained from using the in situ
• electronic controller and data acquisition vane device developed for the simulation
system;
experiment (Model A) was employed in
• volume compensator; designing the device needed for the ISHTE.
• power supply. For example, it was determined that pur-
Both the mechanical system and volume chased internal components used in Model A
compensator are oil filled and designed to were capable of operating at ambient deep-
operate at deep-ocean pressure. The power ocean temperature and pressure after a
supply and controller operate at atmos- dormant period of one month. It was con-
To platform master
controller Shalt/Sleeve
Two Ball
Lead Screws
Bronze Bushing
Ball Seal Hous ing
Oil- Filled
Oil fi l led - - ! Vane Protector
vane [)
protector • ~ Vane Blades
Mechanical Elast cop Elastorneric Membrane
system and Temporary Cap
Fig. 1 Schematic diagram of in situ vane sys- Fig. 2 Assembly drawing of ISV Model B,
tem penetration depth of 1.5 m
IN SITU AND SHIP-BOARD MEASUREMENTS ON A DEEP-SEA CLAY 223
eluded that these same components could be compensating vane protector housing with
used for the ISHTE vane, but some signific- an elastomeric cap to isolate the vane and
ant design modifications and refinements seals from seawater until just before penet-
were incorporated in Model B. The new ration.
arrangement is shown schematically in Fig. 1. The heart of the electronic system is an
Figure 2 shows some details of the Intel 8751 microcontroller which has 4k of
mecnanical unit and Fig. 3 is a photo of the Eprom program memory, 128 bytes at
entire system mounted on a platform that RAM, a full duplex serial port, two 16-bit
was used during deep-water tests in the counter/timers, and two external inter-
autumn of 1984. Non-metallic components rupts. System configuration characteristics,
are incorporated in less critical areas to such as the number of sample sites, their
reduce the weight, and the volume- locations, gear ratios, output data format,
compensating bladder is mounted externally. etc., are stored in. a standard 2716 Eprom
Another new feature of the design is a self- which allows the user to easily modify the
system configuration to meet changing
requirements. Interlaced to the 8751 micro-
controller chip are a Sensotec model 41, a
force transducer to measure the penetration
force, a Lebow Model 2120, a torque trans-
ducer to measure the shear torque, a Datel
LPS-16 cassette tape deck, and four Airpax
stepper motor-drive cards. Sensors of dif-
ferent ranges can be substituted to match
the system's sensitivity to the sediment
conditions.
Power for the entire system is provided
from batteries housed in a 0.43 m diameter
Benthos glass sphere. There are three sepa-
rate battery packs within the sphere, with
10 Ah capacity for the vane controller elec-
tronics, and a 12 V pack with 40 Ah capacity
for the two motors.
The system electronics are normally in a
powered-down state, except for the compu-
ter interface board which remains powered
up to wait for commands from the master
computer system. After some initial read-
ings, the microprocessor computes and
issues a number of individual commands to
the stepper motor-drive card that controls
the penetration motor. Each step of the
penetration force is averaged to yield one
force sample per 3 mm of penetration. The
averaged data is stored on magnetic tape
and placed in RAM for transmission to the
master computer on the ISHTE platform.
After reaching the depth of the next sta-
tion, the microprocessor selects a rotation
Fig. 3 ISV Model B mounted on component direction and rotates the vane at a rate of
test cruise platform l°/s. The direction of the vane is alternated
224 ADVANCES IN UNDERWATER TECHNOLOGY
at each successive station, keeping the vane retraction, shuts the power off to all boards
blade rotation within the same quadrant. except the computer interface board.
The output of the torque cell is read and the
raw measurement stored on magnetic tape
while the RAM receives averaged data. The RESULTS OF THE AUTUMN 1984
data is averaged so that 100 data points are DEEP-WATER TEST
placed in memory for each series of torque
measurements. The first 80 data points rep- Cruise Plan and Site Characteristics
resent the first 45° of rotation, and the last
20 points represent the final 45° of rotation. The main objective of the 1984 cruise was to
This averaging technique was chosen to test all the components - except for the
highlight the area where the sediment is heater that will be incorporated in the
expected to fail, usually within the first 10 to ISHTE - near the deep-water site selected
20° of rotation. At the end of the rotation, for the one-year experiment. Almost all the
just before the penetration to the next sta- instrumentation has been developed espe-
tion, a final torque measurement is made to cially for the ISHTE. The main instruments
determine if a residual torque is being are the following:
exerted upon the vane shaft. If a residual
• thermal sensors, on heater and in sedi-
torque is sensed, the vane rotation direction ment;
is reversed to relieve the torque before
penetration to the next station. However, a • thermal conductivity probes (line
sources);
maximum reversal of 15° is imposed as a
limit. The amount of reversal is subtracted • piezometers for pore water pressure
monitoring;
from the next 90° rotation sequence so as to
keep the test rotation within the same quad- • pore-water sampler for geochemical
analysis;
rant. The microprocessor uses the same
8-bit A/D converter to process the torque • vane shear s.rstem for geotechnical
analysis;
data. The torque measurement resolution is
equal to the rated torque of the torque cell • ion migration experiment with overcorer;
divided by 128. For a 1.412 N m (200 inf oz.) • hydrostatically activated corers;
torque cell this yields a sensitivity of 0.011 • cameras to monitor conditions at site and
around heater.
N m (1.5625 inf oz.) per bit. Finer sen-
sitivities than this can be achieved by chang- In addition, a special acoustic tracking sys-
ing torque cells. tem with a transponder network and a mas-
Data from a pair of rotation limit switches ter controller and data acquisition system
mounted upon the motor shafts are collected have been developed (Olson et al., in press).
and stored on the tape and in RAM. These For the 1984 cruise, the instrumentation
limit switches allow verification of proper was mounted on two separate platforms and
vane extension and rotation during post- a heat source was not provided.
mission data analysis. The vane shear measurements reported
Once the vane has completed its full here were taken in a water depth of 5845 m
extension and rotation sequence, it signals in an abysall hill region at 30 20.827'N,
0
the master computer. The master computer 157°50.921'W within a SDP study site
system then requests that the data collected designated as MPG-I (MPG-I lies between
in the RAM be telemetered over a 2400 band latitudes 30° and 31°30'N and longitudes
serial link. After verification of the data, the 157° and 159° and 159°W). The sediments in
master computer issues a command to this area have been studied quite exten-
retract the vane fully. The microprocessor sively and the upper few meters are gener-
then retracts the vane, and, upon full ally characterized as being fine-grained
TN SITU AND SHIP-BOARD MEASUREMENTS ON A DEEP-SEA CLAY 225
TABLE 1
Summary of geotechnical properties in MPG-J (Upper 4 m, 26 cores)
Average
w(%) Silt (%) Clay (%) OCR
110 3.05 90 50 33 66 0.85 10.6
98-137 2.56-3.33 81-101 42-71 0.46-1.50 1.2-48.00
illite-rich clays of medium sensitivity, low vertically and hydraulically extruded incre-
strength, low permeability and high com- mentally by using the corer piston and ram.
pressibility (Table 1). A motorized laboratory miniature vane
(Wykeham-Farrance) was used with a tor-
Coring, Sediment Sampling and que transducer, 12.5 x 12.5 mm vane and
Geotechnical Characteristics 60 o /min rotation rate. The vane apparatus
was rigidly attached to the core barrel to
The hydrostatically actuated corers (HLC) minimize relative movement between the
used on the September 1984 cruise two. Samples were taken for water content
(ATLAS-1-84) were designed especially for determinations and several types of sub-
the ISHTE and use the ambient water pres- samples were obtained for detailed laborat-
sure to drive the core tube into the sediment ory analysis. At the time of writing this
while preventing movement of the piston paper, only the water content and vane
(Percivaletal., 1984). The core has an inside shear data were available.
diameter of 102 mm, a smooth outside bar- Approximtely thirty large-diameter (102
rel, and tapered nose cone. Because of the mm) gravity cores were taken in MPG-I
very controlled way in which the corer is with the closest (DS1208-81) being within
taken, it should be expected to recover an 1 km of the ATLAS-1-84 site. The upper few
excellent quality core with minimal distur- meters of sediment in MPG-I exhibit very
bance. However, at the site, the manganese uniform physical property characteristics
nodule cover is estimated to be 30-40% and and it seems reasonable to make compari-
it is possible that nodules could be dragged sons between these two sites. Therefore
down into the sediment. data from two gravity cores are also pre-
There were two lowerings of the main sented here.
platform. On the first lowering, one full Some typical physical property data for
suite of ISV measurements was made and the upper four meters of the illite-rich clay
then a core (HLC-1) was taken at a horizon- in MPG-I are shown in Table 1. Throughout
tal distance of 370 mm from the vane. On the the region the upper 3-4 m shows high
second lowering only a few measurements 'apparent' overconsolidation with OCR val-
were taken with the ISV and two cores were ues of more than 3 down to 1 m depth (Silva
taken (HLC-2, 3). HLC-2 was taken adja- and Jordan, 1984). The water content profile
cent to the 82 mm heater implant rod and is for HLC-1 (Fig. 4) shows that there are var-
therefore disturbed, but HLC-3 was at a iations downcore. Comparison with other
distance of 742 mm from this rod. The hori- cores in MPG-I indicate that this variability
zontal distance between the two lowerings is fairly typical with a rapid decrease from
was 66 m. The core samples were inverted over 130% at the surface to less than 110%
226 ADVANCES IN UNDERWATER TECHNOLOGY
20 1. Vane size
40
j Since the vane used in the ISV tests (30 x 45
'\ mm) was considerably larger than the one
'-------- used on the cores, a study was made using
/'
~
laboratory remolded illite from MPG-I to
[:::c. determine how this might affect the results.
r The results from several measurements
\
100
slightly lower values (within 2%).
120 2. System calibration
/
140 Several sets of calibration tests, using a pul-
Fig.4 Water content versus depth, ATLAS ley and weight apparatus, were conducted
1-84, HLC-1 with the assembled mechanical system in
both the horizontal and vertical orienta-
tions. All the tests, including one set done at
within 0.1-0.2 m depth. The water content 9°C, were quite consistent. The factors used
increases somewhat within the 0.2-0.4 m in converting controller output data to tor-
depth (in HLC-3 the increase was from 110 ques were as follows:
to 117%) and then decreases again to less
than 105% within the 0.4 to 0.7 m depth. clockwise rotation 34.18 ozf in(V
There is a zone of higher water content anticlockwise rotation 34.07 ozf in/V
(over 120% in both HLC-1 and HLC-3
down to about 1.4 m and then another sig- 3. Pressure and temperature effects on
nificant decrease to about 110% at 1. 7 m). torque sensor
In summary, the water content profile below
the surface for the upper 2 m consists of A special variable load system was
three zones of low water content (including developed to apply known torques to the
the zone below 1. 7 m) separated by two transducer in an oil-filled, cooled pressure
zones of higher water content. chamber. Tests were conducted in both
rotation directions at room temperature
(22°C) and atmospheric pressure, and at 2°C
ISV Preparations, Calibrations and Data and 58.6 MPa (8500 psi) pressure. Compared
Conversion to the factory calibration the calibration fac-
tor at high pressure and low temperature
Because of delays in fabrication and the was only 1.9% greater, and the room-tem-
need to meet the ship schedule, only pre- perature atmospheric pressure calibration
liminary calibrations and environmental was less than 1% greater. The final adjust-
effects studies were made prior to the 1984 ment made to account for the combined
cruise. Since that time a great deal of effort pressure and temperature effect was to
has been put into the determination of all increase the calibration factors of the sys-
possible effects on the ISV measurements. tem by 1.8% in the clockwise direction and
Following is a summary of the major items 2.1 % in the anti-clockwise direction. The
that were checked out and an indication of end result was that the calibration factor
IN SITU AND SHIP-BOARD MEASUREMENTS ON A DEEP-SEA CLAY 227
was the same in each direction, i. e. 34.79 ozf R = 17.3 mm, where R is the radius of a
infV· circular arc at the vane bottom (see Fig. 2),
and assumes a triangular stress distribution
4. Mechanical friction effects on the top and bottom surfaces.
1\
~o r-
-.'"
\ "
\~
'
The length and diameter of the exposed
shaft above the vane are constant at 96 mm HLC-~)
' ;J
~ 60
and 6.35 mm respectively. A correction for ~
the torque on this portion was made by
assuming a remoulded strength of 50% of
)//
the actual strength. This correction (3.7%) /
was subtracted from the measured torque to
yield the corrected torque used to calculate 100 ~<: ~
shear strength around the vane blades. ). 1
/ ,/ '\
120
<." ,
\
7. Conversion to shear strength "-
..
during extrusion. A few points are worth turbance imparted to these two cores was
noting. There is extreme variability in the probably much less.
upper 0.2 m with some points higher and one Results from two large-diameter gravity
lower than the ISV results. This may have cores from the DS 1208-81 cruise (less than
been caused by compression during the 1 km away) are shown in Fig. 6, along with
extrusion process but is more likely to have the ISV results. Except for the lower por-
been due to some difficulties caused by ship tion, GC-12 shows strengths very similar to
vibrations. Below 0.2 m the HLC-1 and ISV the ISV results. The results for GC-13 are
trends are remarkably similar, although considerably less than the ISV strengths.
below 0.9 m the decrease shown in the core
results is much greater.
The HLC-3 core was taken on a different
SUMMARY
lowering (66 m away). Visual observations
indicated some disturbance - possibly by The results of a deep-water test with a new
intrusion of nodules - but overall the core in situ vane shear system have been
seemed to be in good condition. Except for described. A few summary comments fol-
one point at 0.1 m, the shear strength of this low.
core was considerably lower than both the
in situ results and the HLC-1 results. It (a) All subsystems of the ISV system oper-
should be noted that this corer was hauled ated successfully, there were no major
aboard manually and therefore may have malfunctions or failures and the system
been subjected to more disturbance than design appears to be sound.
HLC-l. However, relative to normal piston (b) The strength profiles for HLC-1,
coring or gravity coring, the mechanical dis- HLC-3, and the ISV measurements fol-
low very similar trends for most of the
1.4 m depth. However, the ISV pro-
SHEAR STRENGTH (Kpa) file generally shows slightly higher
-'-;\
o 3 6 9 12 strengths than HLC-1 and considerably
0
\
higher strengths than HLC-3. Some of
the extreme variability in the lower
20 '. parts of the two cores does not show up
-t~INS
40 by disturbance rather than actual
OS1 08-81
TU VANE sedimentary conditions.
G -12 (c) The strength profile for one nearby «1
Y
"~ 60
~ km) gravity core is similar to the ISV
profile, but the other is considerably
i::
~ 80
\ lower.
OSl208~~
/)
1:)
(d) In order to quantify the differences
GC-13 /
between core and ISV measurements,
100 i
a numerical integration was made
/
/
/
I
\
120
" '''''.. respective curves. Ratios were then
calculated to characterize the differ-
14 0 ences; the results of these analyses are
Fig. 6 Shear strength versus depth in situ
summarized in Table 2.
vane versus DS 1208-81 GC-12, GC-13 MPG-I The most direct and reliable compari-
area son is for core HLC-1, since this was
TN SITU AND SHIP-BOARD MEASUREMENTS ON A DEEP-SEA CLAY 229
TABLE 2
Shear strength comparisons (average from 0.20 to 1.15 m depth)
% Reduction Ratio
Core no. ISV to core ISV/core Comments
Minimum 12 1.13
Maximum 31 1.44
Average 20 1.26
taken on the same lowering as the ISV. ured values of in situ shear strength of
As shown in Fig. 5, the agreement be- clays. In Proc. 6th Int. Conf. of Soil Mech.
tween ISV and core measurements be- and Found. Engng, Montreal, Vol. 1, Div.
tween 0.5 and 0.9 m is excellent, but the 3, pp. 141-145.
disparity above and below this zone is 2. ASCE, 1975. In Situ Measurement of Soil
Properties, Vols I and II. Proc. Geot. Eng.
such that there is an average of 13% Division Specialty Conference, 1975.
reduction. Based on this one core, the 3. ASTM, 1972. Field vane shear test in cohe-
average core strengths would need to be sive soil. Designation D2573-72.
increased by 15% to obtain in situ 4. Babb, J. D. 1982. Development of an in situ
strengths. The corresponding values for vane for strength measurement of deep sea
HLC-3 are much higher (25 and 33%). sediments. MS thesis, University of Rhode
There is a greater difference between Island.
the two gravity cores (12 and 31% 5. Babb, J. D. and Silva, A. J. 1983. An in situ
reduction). vane system for measuring deep sea sedi-
The average percentage reduction for ment shear strength. IEEE/MTS Proceed-
all cores is 20%. Therefore the core ings, Oceans '83, Vol. 1, pp. 598-602.
6. Backes, J. L., Bell, B. M. and Olson, L. O.
results would need to be increased by 1981. Long-baseline deep ocean acoustic
26% to obtain the in situ strengths. tracking and telemetry system. IEEE /MTS
(e) The cores used in the comparisons are Proceedings, Oceans '81, Vol. 1, pp. 1-8.
considered to be of very good quality, 7. Briaud, J. L. 1980. In-situ tests to measure
since two of them (HLC cores) are soil strength and deformability for offshore
large-diameter piston cores taken in a engineering. Research Report, Texas A&M
controlled manner from a bottom- Research Foundation, College Station, TX.
supported platform, and two (GC cores) 8. Doyle, E. H., McClelland, B. and Ferguson,
are large-diameter, thin-walled gravity G. H. 1971. Wire-line vane probe for deep
cores. Smaller diameter cores, espe- penetration measurements of ocean sedi-
ment strength. Offshore Technology Confer-
cially standard piston cores, used in ence, Paper No. OTC 1327.
deep water probably produce greater 9. Flaate, K. 1966. Factors influencing the
sample disturbance and the corrections results of vane tests. Can. Geotech. J. 3(1).
suggested above may not be applicable. 10. Hadley, G. R., McVey, F. F. and Morin, R.
1980. Thermophysical properties of deep
ocean sediments. In Marine Technology '80.
HEFERENCES Marine Technology Society.
11. Hollister, C. D., Anderson, D. R. and
1.• Aas, G. A. 1965. A study of the effect of Heath, G. R. 1981. Subseabed disposal of
vane shape and rate of strain on the meas- nuclear waste? Science 213, 1321-1326.
230 ADVANCES IN UNDERWATER TECHNOLOGY
12. Lee, H. J. 1979. Offshore soil sampling and E. and Keller, G. H. 1972. In-place meas-
geotechnical parameter determination. Off- urement of deep sea soil shear strength.
shore Technology Conference, Paper No. ASTM, STP, pp. 55-68.
OTC 3524. 19. Seabed Programs Division, 1983. The sub-
13. Olson, L. 0., Backes, J. L. and Miller, J. B. seabed disposal program: 1983 Status
In press. Communication, control and data Report. SAND83-1367, Sandia National
acquisition systems on the ISHTE Lander. Laboratories, Albuquerque, N. Mex.
IEEE. 20. Silva, A. J. and Jordan, S. A. 1984. Consoli-
14. Monney, 1973. Analysis of sediment shear dation properties and stress history of some
strength at various rates of shear. In The deep sea sediments. In Seabed Mechanics.
Physical and Engineering Properties of Edited Proceedings of IUTAM Symposium,
Deep Sea Sediments (Ed. A. L. Inderbit- University of Newcastle Upon Tyne, Sep-
zen), Virginia. tember 1983, pp. 25-40.
15. Percival, C. M. 1983. The subseabed dis- 21. Silva, A. J. and Pekin, O. 1981. An in situ
posal program In Situ Heat Transfer Ex- geotechnical measurement system for deep
periment (ISHTE). SAND80-0202, Sandia sea surficial sediments. Trans. AGU 62(45).
National Laboratories, Albuquerque, N. (Abstract No. 01-2-C-5.)
Mex. 22. Silva, A. J., Babb, J. D., Lipkin, J., Piet-
16. Percival, C. M., McVey, D. F., Olson, L. O. ryka, P. and Butler, D. In press. In situ
and Silva, A. J. 1984. In Situ Heat Transfer vane system for seafloor strength investiga-
Experiment (ISHTE). Mar. Geotech. 5(3/4), tions. IEEE.
361-378. 23. Silva, A. J., Criscenzo, S. J., Jordan, S. A.
17. Perlow, M. and Richards, A. F. 1972. In- and Babb, J. D. 1983. URI Geotechnical
place geotechnical measurements from sub- Program of the In Situ Heat Transfer
mersible Alvin in Gulf of Maine soils. In Off- Experiment. ISHTE Annual Report No.2,
shore Technology Conference, 1972. University of Rhode Island.
18. Richards, A. F., McDonald, V. J., Olson, R.
15
The I nfluence of Geological
Processes and Test Procedures
on Measured and Evaluated
Parameters
A. Marsland, Building Research Establishment, UK
of foundations. The traditional method of Sea was free from ice (Boulton et al., 1977;
attempting to overcome this problem uses Behre et al., 1979). Whether the Scandina-
empirical correlations between data vian and British ice coalesced during part of
obtained from specific standard tests and the Devensian is still a matter for specula-
the behaviour of particular types of founda- tion, but there seems to be general agree-
tions or elements of foundations. However, ment that both reached the Dogger Bank in
many of the available correlations have the central North Sea. Superimposed on the
either been derived for a particular local local glacial conditions were the changes in
situation or show such a large scatter that sea-level produced by the increases and
their use is limited to the design of found- decreases of the global ice masses. Various
ations where the overall factor of safety is sources suggest that the sea-levels, when
high. The purpose of this paper is to high- the Devensian ice caps surrounding the
light some of the outstanding problems in North Sea reached their maximum (about
assessing relevant parameters for offshore 18 000 years BP), were probably 130 m lower
design and to suggest ways in which they than at present (Jardin, 1979; Jelgersma,
may be improved. 1979). Even neglecting the continuing set-
tlement due to techtonic movements, there
is little doubt that the southern and central
GENESIS OF NORTH SEA SOILS North Sea were above sea-level for appreci-
able periods after the commencement of the
Most of the soils in the North Sea and the Anglian glaciation about 300 000 year ago.
adjacent parts of the continental shelf which The situation in the northern North Sea is
are of interest to the geotechnical engineer more debatable, but even there the sea
were deposited during the Quaternary. Dur- would have been shallow for substantial
ing this period the area was exposed to large periods with much of the area covered with
variations in climate, which resulted in grounded ice and the possibility of some
glaciations and intervening warmer periods. ground exposure. From the above discus-
The continual changing positions of ice sion it is evident that a full range of glacial,
fronts, sea levels and climatic conditions led glaciomarine, fresh-water depositional
to sequences of errosion, deposition, con- environments occurred within the North
solidation, and post-depositional modifica- Sea Basin at some time. The location of the
tions. Studies of land borings and expos- ice fronts, shorelines and glacial lakes at any
ures, combined with more recent data from one time are not known with any degree of
offshore borings and geophysical surveys, precision and are the subject of much ongo-
show that ice-sheets which developed in the ing discussion (Aarseth and Sejup, 1984).
highlands of Scotland and Norway extended There is, however, general agreement that
well into the North Sea on several occasions a situation similar to that portrayed in Fig. 1
(L~ken, 1976; Boulton et al., 1977; Oele and probably occurred on more than one occas-
Schuttenhelm, 1979). At its maximum sion during the various glacial periods.
(200 000-300 000 years BP), the ice covered Much of the material deposited during
the whole of the North Sea area, including earlier glacial advances was removed or
the UK, Scandinavia and the northern parts reworked by later advances. In the southern
of Germany and the Netherlands. During parts of the North Sea deposition was
this period the sediments of the North Sea mainly from sheets of land ice, but in the
were subjected to extensive erosion, rede- northern parts of the North Sea there were
position and high total overburden pres- more periods of marine deposition with the
sures. The extent of the ice during the last soil being supplied by subglacial streams
major glacial advances during the Deven- and ice calving. Evidence from studies on
sian (10 000-75 000 years BP) were less present-day land areas adjacent to the
extensive and much of the southern North North Sea indicate that many of the ice-free
GEOLOGICAL PROCESSES AND TEST PROCEDURES 233
o·
_
~
E
observations around present-day ice-sheets
(Boulton, 1972), suggest that subpolar and
thick polar glaciers, while frozen to their
beds in their outer zones, are underlain by
thin films of liquid water elsewhere. The
transition from one subglacial regime to the
other provides the mechanism for the incor-
poration of debris by basal freezing with the
concentration of material becoming progres-
sively greater towards the ice margins as a
consequence of the downwasting of the ice.
Deposition of this debris is affected by a var-
iety of processes, including melt-out, flow
and lodgement. The complex sedimentary
sequences produced by these processes
around modern polar and subpolar glaciers
bear a striking resemblance to the sequ-
ences commonly seen in British glacial
deposits. This led Boulton (1972) to propose
that at their maximum extents the British
ice-sheets were of polar or subpolar type.
The wide range and complexity of the
deposits formed in land-ice and glaciomarine
environments are discussed by Boulton and
Deynoux (1981). Further aspects of glacio-
marine sedimentation are summarized in a
... recent paper by Powell (1984) .
Ijjl/; LAND §:=i .LAKES ~ SEA f ICE Maximum deposition occurs near the ice
fronts present at the particular time. The
Fig. 1 Generalized map of possible conditions amount and mode of deposition depends on
in the North Sea during part of Devensian Glaci- the activity and fluctuations of the ice front
ation; based on data published by Boulton et al.,
1977; Jardine, 1979; Jelgersma, 1979 and L0ken, at the particular locality. For example, dur-
1976 ing the late Devensian advances an exten-
sive surging lobe developed in the western
part of the southern North Sea adjacent to
land areas within the North Sea Basin were the east coast of England (Boulton et al.,
sUbjected to very cold conditions which 1977). This deposited deep layers of mainly
would have caused deep freezing to occur in subglacial till that are found on land in Hol-
some areas. This ties in with the concept derness (north of Hull) in eastern England
that the main streams of warm, moist air and extend well out into the North Sea
were further south during the glacial within the area which was covered by this
periods than they are today (Liljequist, lobe.
1974) and as a consequence cold, dry conti- In addition to the wide range of primary
nental conditions existed, particularly in the depositional processes, the soils have often
mid and northern parts of the North Sea. been subjected to post-depositional changes
The dynamic conditions of the ice masses, other than simple loading and unloading.
Ilarticularly as influenced by the thermal Grounded ice causes appreciable remoulding
regime, affect the patterns and processes of and shearing of subglacial deposits. Simi-
glacial erosion and deposition. Models of larly, advancing ice produces substantial
cift entrainment and deposition, based on disturbance of the local bedrock and mater-
234 ADVANCES IN UNDERWATER TECHNOLOGY
ial deposited during previous glacial re- these studies are given by Marsland et al.
treats. Shearing also occurs as a result of (1982), Derbyshire and Love (1985), and
differential ice loading, slipping or 'flowing' Love and Derbyshire (1985). The fabric fea-
of soft wet deposits, and melting of buried tures found in a particular deposit are a
ice during deposition. The degree of consoli- reflection of the composition of the parent
dation that occurred during ice loading rocks, the degree of comminution, deposi-
would have depended both on the thickness tional processes and environments, as well
of the ice and the ease or otherwise with as post-depositional changes due to environ-
which drainage could occur. Drainage could mental and stress conditions.
have been prevented by ground freezing and While the presence of soil fabric features
the long drainage paths from ground below and their importance for the engineering
thick ice areas would have restricted the properties of land-based soils was becoming
degree of pore water pressure dissipation. increasingly recognized in the mid-1960s and
Freezing and drying caused further disrup- early 1970s (Ward et al., 1965; Marsland and
tion which substantially modified the effects Butler, 1967; Bishop and Little, 1967; Rowe,
of deposition and post-depositional consoli- 1972; Marsland, 1971b, c, 1975), surpris-
dation. ingly little detailed attention has been paid
to fabric features in offshore deposits. The
potential importance of fabric studies in the
SOIL FABRIC FEATURES IN NORTH evaluation of the geotechnical properties of
SEA AND ADJACENT LAND DEPOSITS offshore soils was stressed by the author
(Marsland, 1977), and some typical exam-
'Soil fabric' as used in this paper refers to ples of macrofabric in samples obtained from
the nature, spacing, and directional proper- the North Sea Frigg field (near location D,
ties of all types of fabric features at all Fig. 1) were given. Schjetne and Brylawski
scales. In order to obtain a reasonably com- (1979) gave an example of a finely laminated
plete picture it is necessary visually to clay from the east Shetland basin and
examine 'broken' surfaces of as many sam- described the use of X-ray inspection to dis-
ples as possible. The use of suitable illumina- cover fabric features and possible distur-
tion and a large low-powered magnifier are bance. Schjetne et al. (1979) mentioned fis-
essential. Careful progressive separation of suring of clays and the possible importance
the samples along visible discontinuities of freezing and chemical cementation. An
provides data on their nature and spacing illustration of a clay containing polished,
and often reveals the presence of further striated discontinuities (obtained from the
smaller-scale discontinuities. Selected areas Aberdeen ground beds just north of the
of a typical sample should be examined at Forties field by the Institute of Geological
increasing magnifications, such as 25, 50, Sciences) was published by Redding (1976).
100, 250, 500 and 1000 times, with occasional As part of the overall studies of the
higher magnifications being used to inspect geotechnical properties of both land and
individual particles. Such multiscale inspec- North Sea deposits, the BRE contracted the
tions are necessary because features that geography department of Keele University
can be easily observed at one particular to make detailed studies of the sedimen-
scale can be easily missed at other scales. In tological and fabric properties of soils from
addition there is often a heirarchy of typical locations. In the early stages of the
important features. Such studies can be offshore investigations one of the main dif-
readily made using a scanning electron micro- ficulties was to obtain sufficient good-
scope, and the examples given in this quality samples in order to give a reasonable
paper were obtained as part of a BRE coverage of the variable fabrics at specific
research contract undertaken at Keele U ni- sites and to relate these to geotechnical
versity. Further details of some aspects of properties. In spite of the difficulties of
GEOLOGICAL PROCESSES AND TEST PROCEDURES 235
Fig.2 Photomicrograph of till from a depth of 14.5 m at the B.R.E. test site at Cowden (location A
Fig. 1); photograph width represents approximately 0.1 mm
TABLE 1
Some references on macrofabric features in glacical clays observed in land
areas around the North Sea
subglacial shearing. However, detailed vis- pressures and the measured in situ horizon-
ual inspection of the tills in 100 mm diameter tal stresses were surprisingly low. High
samples and in exposed sides of large diam- densities can, however, be achieved by
eter boreholes taken to a depth of 22 m thorough mixing of well-graded soils and
showed that observable macro discon- this, together with the general absence of
tinuities were restricted to the weathered observable macro discontinuities, suggests
zone in the upper 4 to 5 m. This is contrary that these tills were probably deposited in a
to the general findings in subglacial tills relatively wet state.
found in many other parts of Britain, as can Further north (location C, Fig. 1) the
be seen from Table 1. Even though the deposits in the upper 100 m become more
Cowden tills contained few macro discon- variable due to periods of deposition in gla-
tinuities, both the onshore and offshore tills ciallakes and under glaciomarine conditions,
in this area (locations A and B, Fig. 1) con- which gave rise to more laminated deposits
tained numerous microshears which are typ- such as that shown in Fig. 4. Dense subgla-
ically 0.05 to 0.2 mm long, as can be clearly cial tills are, however, still present as illus-
seen in Fig. 2. The highly mixed nature of trated by the micrograph in Fig. 5. Other
the tills both at Cowden and at West Sole tills which have a more open structure (see
(locations A and B, Fig. 1) is evident from Fig. 6) are also present. There is also evi-
the micrographs in Figs 2 and 3 where par- dence in this area of substantial post-
ticular note should be made of the clay depositional shearing, as shown by the mul-
wrapped round the silt and sand particles. tiple shears in Fig. 7 and the highly polished
These tills contain about 30% of clay-sized microshears in the photograph published by
particles and have densities between 2.2 and Redding (1976).
2.24 mg/m2. In view of these relatively high Further north still (around locations D and
densities the evaluated over consolidation E) the deposits show more evidence of
GEOLOGICAL PROCESSES AND TEST PROCEDURES 237
Fig. 3 Photomicrograph of till from a depth of 11.4 m at West Sole near B in Fig. 1; photograph
width represents approximately 0.2 mm
Fig. 5 Photomicrograph of dense well-mixed fabric from depth of 22.0 m near C in Fig. 1; photo-
graph width represents approximately 1.6 mm
tions and magnitude of the stresses applied tinuities. This is reflected in the values of
in the tests faithfully reproduce those on the cone factor N k obtained when values of Cu
typical elements in the ground surrounding back analysed from large plate tests are
the foundation. The influence of the relative used (Marsland, 1979; Marsland and Quar-
scale of a static cone and a soil containing terman, 1982). In stiff clays values of 30 or
reasonably randomly orientated discon- more are obtained when the cones are small
tinuities at different spacings is illustrated compared to the discontinuity spacing as in
in Fig. 15. For the condition shown in A, the C, and values of around 15 or less are
cone will provide a reasonable measure of obtained when the condition in A applies. In
the large-scale strength for failures similar typical strength stiff clays free of macro dis-
to that around a cone. However, in case C, continuities, the values of N k are close to 15
the cone will provide a measure of the (Marsland and Powell, 1985). The lower val-
strength of the intact clay between the dis- ues sometimes obtained in clays containing
(>l)ntinuities which in stiff clays is often sev- closely spaced discontinuities, as in A, could
"al times that mobilized along discon- be due to loose packing of the clay caused by
240 ADVANCES IN UNDERWATER TECHNOLOGY
Fig. 6 Photomicrograph of fairly loosely packed till from a depth 14.0 m near C in Fig. 1; photo-
graph width represents approximately 0.35 mm
processes such as freeze-thaw. In practice to Fig. 16. It is fairly obvious that the smal-
the situation is never as simple as that illus- ler tests will measure the strength of the
trated in Fig. 15, since the discontinuities clays between the macro discontinuities,
are seldom randomly orientated and there is even when they are fairly closely spaced.
often a hierarchy of discontinuities at dif- This will generally differ significantly from
ferrent scales, at different orientations, and the larger-scale strengths which take into
when there are significantly different sur- account the discontinuities. Possible excep-
face properties. For example, highly tions are when the clay between the macro
polished slickensided surfaces are often discontinuities is also highly fractured at the
larger than the discontinuities found in sam- macroscale as in Fig. 14. Even here some
ples which are often less polished. An indi- differences are likely to be due to the prob-
cation of the different influences of scale and able different nature of the smaller discon-
direction of shearing in the various tests tinuities and directional effects. The reduc-
used offshore can be obtained by reference tion in strength due to the presence of dis-
GEOLOGICAL PROCESSES AND TEST PROCEDURES 241
Fig. 7 Photomicrograph showing microfolding and shears in a sample from a depth of 4 m near C in
Fig. 1; photograph width represents approximately 4 mm
continuities depends On the degree of over- stresses at the surface of the probe. Thus
consolidation and the properties of the soil strengths evaluated from pressuremeter
within or along the discontinuity (McGown, tests in stiff clays may be significantly
1979; Lupini et al., 1981). higher than the in situ strengths if the spac-
The scale of the fabric features is even ing between vertical discontinuities is com-
important in tests which stress a relatively parable or greater than the radius of the
large volume of soil when the stress levels probe. The pressuremeter will also give
applied by the tests vary within the stress strengths approaching the intact strength of
field. This occurs in the pressuremeter test a stiff clay if the discontinuities are predo-
where the highest stresses occur adjacent to minantly horizontal.
the expanding probe and decrease with the The degree of disturbance during sampl-
radial distance from the centre of the probe. ing and specimen preparation, or during
The area within the annulus between the insertion of in situ test equipment, can also
probe surface (broken circle) and the outer depend On the nature and scale of the soil
dashed line shown in Fig. 16 is subjected to fabric and the stiffness of the soil between
1 average stress which is about 70% of the the discontinuities.
242 ADVANCES IN UNDERWATER TECHNOLOGY
Fig. 9 Well-developed platy-lenticular struc- Fig. 11 Finer, more distorted lenticular fabric
ture in sample from a depth of 66 m near C in in sample from depth of 50 m near E in Fig. 1;
Fig. 1; sample width 72 mm sample width 72 mm
GEOLOGICAL PROCESSES AND TEST PROCEDURES 243
Fig. 12 Macroped fabric developed by a single artificial freeze-thaw cycle in a sample of previously
intact uniform glacial lake clay from Winnipeg; photograph width represents 70 mm
B c
Fig. 13 Fabric features caused by freezing and
thawing (after Van Vliet and Laughor, 1981): (a)
Foliated or platy (0.2-3.0 mm thick and 0.5-10.0
mm long), (b) Platy or lenticular subangular Fig. 14 Photomicrograph of microped fabric
blocks (3.0-15 mm thick and 8-50 mm long), (c) within macroped features from a depth of about
• ngular blocky to prismatic (dimensions 10-50 18.4 m at a location near E in Fig. 1; photograph
mm) width represents approximately 1 mm
244 ADVANCES IN UNDERWATER TECHNOLOGY
Even when the pore water is free from dis- vital to use the best available sampling
solved gases, pore suctions within samples techniques, but even these result in some
gradually break down when the reduction in sample disturbance. Direct assessment of
effective stress exceeds a threshold value, the degree of disturbance of offshore sam-
which differs for each soil (Marsland and ples taken by the best available techniques
Windle, 1982; Kirkpatrick and Khan, 1984). by comparing the results obtained with
The presence of macro- and microdiscon- those from 'truly' undisturbed samples, such
tinuities can reduce this threshold and speed as carefully taken block samples or reliable
up the reduction in the all-round effective large-scale tests, is not possible.
stresses locked up in the sample. Shear The first requirement in assessing the
stresses induced in stiff soils during sampl- likely degree of disturbance is to have avail-
lng and extrusion can cause dilation and par- able a precise account, including the times, of
tial cavitation of parts of the samples. As a all the significant stages of drilling and sam-
~onsequence the effective stresses main- pling or insertion, together with details of
tained within a sample after stress release operational conditions and specific problems
lecrease with increasing sample distur- which may have arisen. For example, the
Dance. Since there is increasing evidence possible reasons for poor or nil recovery and
that undrained shear strengths are depen- for obtaining heavily disturbed samples may
lent on the initial mean effective stress be very relevant since many geotechnical
within the sample (Marsland and Powell, problems arise due to the missing or poor
1985), any decrease in the mean effective recovery of strata which are critical in a par-
stress can lead to the measurement of low ticular foundation problem.
shear strengths. Whether or not this is the Indications of the degree of sampling dis-
~ase depends not only on the degree of turbance can be obtained by examining
Dreakdown of pore suctions during sampling radiographs of undisturbed samples, by vis-
:wd storage but also on the increases in ual inspection of split samples and by
~ffective stresses which occur during observing thin slices as they dry. Careful
lpplicaion of confining pressures on the now measurements of the moisture content gra-
::>artially saturated soil. When stiff soils dients, variations in the microfabric and
which have suffered small but significant small-scale strength (using small penetro-
shear strains during sampling are subjected meters or fall cones) in different parts of a
;0 soil pressures significantly greater than sample can provide more positive evidence.
;he effective overburden, the mean effective Several investigators (for example,
;tresses which develop in the sample may be Broms, 1980) have suggested that the com-
mbstantially higher than the mean effective parison of the mean all-round effective
n situ stresses. As a consequence the use of stress (0" m) on a sample with that which
~onfining pressures equal to the total in situ would act on an 'ideal' sample from the same
werburden (including water load) can lead location may provide a more quantitative
;0 serious overestimation of strengths in measure of disturbance.
lome stiff clays. Not only can cavitation However, in practice problems arise in
Nithin the pore water of the clays result in the determination of both the 'ideal' and the
;he measurement of incorrect strengths, it measured values of 0" m' As a first
~an also lead to incorrect estimates of in situ approximation it is adequate to assume that
stresses. stiff clays behave as elastic solids when
The assessment of the likely effects of , 2'
O'v+ O'k
sample disturbance on the parameters
IT easured in laboratory tests is one of the 0"
In (ideal) = - - -
3
ITlost difficult and neglected aspects in the
overall evaluation of laboratory data. For all The measured values depend on the extent
,otechnical and fabric measurements it is to which cavitation has occurred within the
246 ADVANCES IN UNDERWATER TECHNOLOGY
Obser-
Relatively undisturbed primary
I.<:
vations
Description of Pnnclpal F'g Fabric sedimentary features (Laminations
C. "
c Fabric Features Nos Type Dustlngs, Partings. inC. crossed Flaser
o"
o
N Bedding, Clastic Dykes etc) Sulphide A
Laminae
I • • ClaVT!!1 micro cracks
No large discontinuities
12(aMb)
-- --
~----------------- --- ----
Disturbed primary bedding with primary
•• •• --------------------
Lammatlons of slit and clay
-.-•
13(a) BC. EF features still remaining dominant. B
-.-
Trans-sample discontinUities
1-- ~t!'~cl
14{a) A-C
Near hOrizontal primary macro
~
•• •
features stilt eVident EF
111(,) Numerous small scale macro 15 C. EF Disrupted primary features, more
fissures and micro cracks 14(b) transgressive dislocations across and
• Trans-sample discontinUities utilising primary boundaries. C
~- • ------------------
~
• • Primary deposition features completely
dIsrupted numerous small scale macro
16(b) D. EF Dislocations not wholly transgressive of
primary boundaries but effectively
••
111(11)
peds Some transample discontInUities 16(a)
Highly developed miCro ped fabnc obliterating these structures. D
-------------------
••
-- -~
111(111)
• Primary depositional features completeh
disrupted Highly developed small
scale macro ped fabric but larger and
17 CD. E
(few F)
more mterlocked than In III (II) High angle 60° - 90° discontinuity
~ Some trans-sample discontinUities planes. E
1--
• -------------------- - - - - ~-- -
~-- -. -- j--------------------
Small scale macro ped fabric well CD
Fig. 18 Typical macrofabric features observed
IVIII) developed Some major discontinUities E-F at location near E in Fig. 1. Marsland et al. (in
literature)
Fig.17 Fabric profile based on observations at
Index strength s
location near E in Fig. 1. Note: figure numbers Field cone strengths (N k - 17)
given in column 6 refer to figures in Marsland et
O.B 1.0 1.2 1.4 1 .6
al. (in literature)
~
solidation, radiocarbon dating, stratigraphi- 0
0;
..c 20 III(ii )
cal variations from geophysical measure-
ments, profiles obtained by static cone tests 0
..
.l:
0.
25
•
and piezocone tests, and borehole logging
techniques. A good overall model of the 'fos- III (i ii)
sil' geological processes, combined with 30
observations of present-day depositional
processes in similar environments, provides 35 L.---~-
an indispensable background against which Fig. 19 Example of plot of ratios of two differ-
to interpret the soil data and its relevance to ent scales of test down profile at location near E
the large-scale situation. in Fig. 1: Marsland et al. (in literature)
GEOLOGICAL PROCESSES AND TEST PROCEDURES 249
Plots of the ratio of results from different mens, could help in the evaluation of the
types of tests when correlated with fabric possible effects of disturbance during sampl-
types, such as in the example in Fig. 19, can ing and insertion of in situ test equipment.
provide additional clues to the value of data In order to explain the varying ratios bet-
from particular tests and could even indicate ween results from different types and scales
the presence of fabric features not already of tests and between tests and large-scale
observed. behaviour, it is necessary to consider the
changes in fabric features down the soil pro-
file as well as differences between the stress
CONCLUSIONS paths relevant in each case. More correla-
tions between large-scale behaviour and
Detailed knowledge of 'fossil' geological pro- high-quality test data in soils with well-
cesses is essential for the adequate under- documented fabrics are urgently required.
standing and evaluation of geotechnical The interdependence of data from all the
data. Detailed studies of soil fabric provide relevant sources and the importance of con-
an important link between the geological tinual interaction throughout an investiga-
processes and geotechnical behaviour. tion of specialists in the various fields cannot
Fabric features at all scales have an impor- be overestimated.
tant effect on both measured and large-scale
geotechnical parameters. The effects of soil
fabric on different types and scales of tests ACKNOWLEDGEMENTS
vary. Reliance should not be placed on the l
res~lts. of one or two different types of test. The work described in this paper forms part
It I~ Important to compare the results of a research programme of the Building
obtamed from ~ rang~ of tests a~d t~ relate Research Establishment and is published by
the~ to the soIl fabrIc, stres~ hIstorIes and permission of the Director. The author
apphed .stress paths. In. thIS respe~t t~e wishes to thank Dr E. Derbyshire and Mr
NorwegIan Geotechmcal In~tItute s M. A. Love of Keele University for their
approach (Andresen et al., 1979), WhICh uses close co-operation in many aspects of the
a combination of triaxial compression, triax- work.
ial extension and simple shear, has much to
commend it. However, a more critical
examination needs to be made of estimations REFERENCES
of OCR and in situ stresses by taking into
account all past geological processes. More 1. Aarseth, I. and Sejup, H. P. (Eds) 1984.
consideration also needs to be given to the Quaternary stratigraphy of the North Sea.
soil fabric and the representative nature of Abstract Volume, Symp. Univ. Bergen,
the test specimens. The inclusion of more Dec. 1984, Bergen, Norway.
comprehensive profiles of tests on uncon- 2. Andersen, A., Berre, T., Kleven, A. and
solidated specimens could provide additional Lunne, T. 1979. Procedures to obtain soil
checks and control. Measurements of pore- parameters for foundation engineering.
water pressure responses prior to and dur- Mar. Technol. 3(3), 201-266.
ing shearing should be made in as many of 3. Behre, K. E., Menke, B. and Streif, M.
the unconsolidated specimens as possible 1979. The Quaternary geological develop-
ment of the German part of the N. Sea. In
(Marsland, 1977). Additional control could
The Quaternary History of the North Sea
1: e obtained by simple measurements of the (Eds E. Oele, R. T. E. Schuttenhelm and A.
nean effective stresses within specimens J. Wiggers). Acta Univ. Uppsala, Symp.
soon after sampling. Such measurements, U niv. U ppsala Annum Quingentesimum
together with appropriate observation of Celebrantis, Vol. 2. Uppsala, Sweden, pp.
IiI fabric in different parts of the speci- 85-113.
250 ADVANCES IN UNDERWATER TECHNOLOGY
4. Bishop, A. W. and Little, A. L. 1967. The Conference on Soil Mechanics and Founda-
influence of size and orientation of sample on tion Engineering, Brighton, September
the apparent strength of the London Clay at 1979, Vol. 4, pp. 57-65.
Maldon, Essex. In Proc. Geotechnical Con- 16. Jardine, W. G. 1979. The Western (United
ference on the Shear Strength Properties of Kingdom) Shore of the North Sea in Late
Natural Soils and Rocks, Oslo, 1967, Vol. 1, Pleistocene and Holocene times. In The
pp.89-96. Quaternary History of the North Sea (Eds
5. Boulton, G. S. 1970. On the deposition of E. Oele, R. T. E. Schuttenhelm and A. J.
subglacial and melt-out tills at the margin of Wiggers). Acta Univ. Uppsala, Symp. Univ.
certain Svalband glaciers. J. Glacoil. 9, Uppsala Annum Quingentesimum Celebran-
231-245. tis, Vol. 2. Uppsala, Sweden, pp. 159-174.
6. Boulton, G. S. 1972. Modern Arctic glaciers 17. Jelgermsa, S. 1979. Sea level changes in the
as depositional models for former ice sheets. N. Sea Basin. In The Quaternary History of
J. Geol. Soc., London 128, 361-393. the North Sea (Eds E. Oele, R. T. E. Schut-
7. Boulton, G. S. and Deynoux, M. 1981. tenhelm and A. J. Wiggers). Acta Univ.
Sedimentation in glacial environments and Uppsala, Symposium Univ. Uppsala Annum
the identification of tills and tillites in Quingentesimum Celebrantis, Vol. 2.
ancient sedimentary sequences. Precam. Uppsala, Sweden, pp. 233-248.
Res. 15, 397-422. 18. Johnston, G. H. (Ed.) 1981. Permafrost
8. Boulton, G. S. and Paul, M. A. 1976. The Engineering Design and Construction. John
influence of genetic processes on some Wiley, pp. 52-53.
geotechnical properties of glacial tills. 19. Kirkpatrick, W. M. and Khan, A. J. 1984.
Q. J. Eng. Geol. 9, 159-194. The reaction of clays to sampling stress
9. Boulton, G. S., Jones, A. S., Clayton, K. M. relief. Geotechnique 34(1), 29-42.
and Kenning, M. J. 1977. A British ice-sheet 20. Liljequist, G. H. 1974. Notes on meteorolog-
model and pattern of glacial erosion and ical conditions in connection with continental
deposition in Britain. In British Quaternary land-ices. In Pleistocene Gelogiska Fore-
Studies: Recent Advances (Ed. F. W. Shot- ningens i Stockholm Forhandlingar, Vol.
ton). Clarendon Press, Oxford (551. 79). 96, pp. 293-298.
10. Broms, B. B. 1980. Soil sampling in Europe. 21. L0ken, T. 1976. Geology of superficial sedi-
J. Geotech. Engng Div. Proc. Am. Civ. Eng. ments in the Northern North Sea. In Proc.
106(GTl), 65-98. 1st Int. Conf. on Behaviour of Offshore
11. Chamberlain, E. W. and Gow, A. J. 1978. Structures, Trondheim, August 1976, pp.
Effects of freezing and thawing on the per- 501-515.
meability and structure of soils. In Proc. 22. Love, M. A. and Derbyshire, E. 1985.
Int. Symposium on Ground Freezing, Ruhr Micro fabric of glacial soils and its quanti-
University, Bochum, pp. 31-34. tation measurements. In Proc. Int. Conf.
12. Derbyshire, E. and Love, M. A. 1985. Mic- Construction of Glacial Tills and Boulder
roshears in diamicts from the North Sea Clays, Edinburgh, March 1985, pp.
Basin. In Scanning Electron Microscope in 129-133.
Geology (Ed. W. B. Whalley). Geobooks, 23. Lupini, J. F., Skinner, A. E. and Vaughan,
Norwich. P. R. 1981. The drained residual strength of
13. Derbyshire, E., Love, M. A. and Edge, M. cohesive soils. Geotechnique 31(2), 181-213.
J. 1985. Fabrics of probable segregated 24. Mackay, R. J. 1974. Reticulate ice veins in
ground ice origin in some sediment cores permafrost Northern Canada. Can. Geotech.
from the North Sea Basin. In Soils and J. II, 230-237.
Quaternary Landscape Evolution (Ed. J. 25. Marsland, A. 1971a. The shear strength of
Boardman). John Wiley, Chap. 11. stiff fissured clays. In Proc. Roscoe Memor-
14. Eden, W. J. 1970. Sampler trials in overcon- ial Symp, Engineering Department, U ni-
solidated sensitive clay. Proc. Symp. on versity of Cambridge, 29-32 March 1971. T.
Sampling of Soil and Rock, Toronto, J. Foulis, Henley-on-Thames, pp. 59-68.
Canada. STP 483, American Society for 26. Marsland, A. 1971b. Large in-situ tests to
Testing and Materials, pp. 132-142. measure the properties of stiff fissured
15. Gens and Hight. 1979. Proc. 7th European clays. In Proc. First Australian-New Zea-
}EOLOGICAL PROCESSES AND TEST PROCEDURES 251
47. Schjetne, K., Anderson, K. H., Lauritssen, undrained strength tests. Geotechnique 15,
R. and Hansteen, O. E. 1979. Foundation 321-344.
Engineering for Offshore Gravity Struc- 50. Wroth, C. P. 1984. The interpretation of
tures. Mar. Geotech. 3(4). in-situ soil tests. Geotechnique 34(4),
48. Van Vliet, N. and Langohr, R. 1981. Corre- 449-489.
lation between fragipans and permafrost - 51. Young, A. G., Quiros, G. W. and Ehlers, C.
with special reference to Weichsel silty J. 1983. The effects of offshore sampling and
deposits in Belgium and Northern France. testing on undrained soil shear strength.
Cantena 8, 137-154. Proc. 15th Annual Offshore Technology
49. Ward, W. H., Marsland, A. and Samuels, S. Conference, Houston, Texas. Paper OTC
G. 1965. Properties of the London clay at the 4465.
Ashford Common Shaft: in-situ and
16
Evaluation of Geotechnical
Parameters from Triaxial Tests
on Offshore Clay
D. W. Hi~ht, q-eotechnical Consulting Group, UK,
A. Gens, Umversidad de Catalufla, Barcelona Spain
and R. J. Jardine, Imperial College of Science and Technology, Lond~n, UK
253
254 ADVANCES IN UNDERWATER TECHNOLOGY
~-----------------------------T--------------------------------'----------
8 S Sieve sizes
f-Im r---------f'-m----~I----------------
m-I1l----·----
r--------------------- --------+-·------~----~-------------------------1
Me .~ ,
.;.~
-ri
I
-
I
1---+-+-+H-++tI---+--b41!' I
I
I
I
I
I
I
I
I-----+--H-++t-ttt----I[-:rl~f'l 11---+-++. - --+I I
I
' +1--1-+-+--++-IH-~
--- -t++-++-~
I
~~H+~~1--~~frl . ~:=+-+H~-~~:r-~4+~~+-~++4
I
;~~~+i:::::::~~ I
I
002 02 2 20
j---:F..,::IJ:,:::l•.....,JJ...,:M:.:,:.:;,U;;;:,IJ:..:.;JlJ...L.::::Co::;:.::..:rs::..e+-.:..;F,~
ne'--...L::
IM",e.:::;
d ,::.,:um:,;.llL..::::
Co::;:a::..:rs::.. I
m:.J_Coarse Cobble,
e _j----:IF....:JO"'e'-.-LIM"'e::.:uc.:'u'7
I Silt I Sand I Gravel '--'
-
Symbol Clay I I (0'.) P l (0/0) P I (0'.)
--
LCT lower Cromer Till 25 12 13
R North Sea Clay 32 15 17
l london Clay 75 28 47
North Sea Clay Groups
(:;(.;: <;:.~;i' A 26 - 30 1 2 - 14 13 - 15
~""'-~'" B 35 - 40 14 - 1 7 18 - 23
. 00 ~ 00 0 C 42 - 48
50 - 55
17 - 20 25 - 28
0 18 - 22 31 - 34
Fig. 1 Gradings and index properties of the clays
E
---;..~~ ...
-~'Y //
B - - - - /....
~ -/
,.....
,
" ,
"
100
C
...a.
250 6 8 10 12 14
(av' · a H') '2(kPa)
· 50
2 ,
'ig. 3 Effective stress paths and stress-strain curves for undrained shear in Lower Cromer Till
• Peak
18
o Ultimate
17
16
15
t
'E 14
~
~
v
W
0 13
~
20 50 100 o
CUp.CUull, Pp , Pt' (kPa) Cup = Peak undrained strength
CUuH ='Ultimate' undrained st~ngth
Pp = Mean cf_ct~ st . . . at CUp
p\ = Mean cffe~ sIres at CUult
Fig.4 Undrained strength and water content relationship for Lower Cromer Till
100
a.... 0 . 20
300 350
("~ . "H)l2 (kPa)
-to
---., ~ '= 22
---... 5°
Fig.5 Pattern of undrained behaviour for reconstituted London Clay (axial strains indicated in %)
VALUATION OF GEOTECHNICAL PARAMETERS 257
stress paths form part of a bounding sur- stress-strain properties have been explored
face to the undrained behaviour of the in more detail, particularly over the strain
K o-consolidated soil. range 0.001 to 1%, using internally mounted
displacement-measuring devices based on
learly, the overall behaviour is governed electrolevels (Burland and Symes, 1982).
Clearly, the overall behaviour is governed The internally measured axial strains are
by the pre-shearing position on the swelling shown plotted along each stress path in Figs
line, EF in Fig. 2, and by direction of shear- 5 and 6, and contours of equal axial strain
ing. The changes in behaviour with OCR can have been sketched.
be considered to result from changes in The internal measurements of strain have
stress state and water content. N ormaliza- revealed an initially very stiff and non-linear
tion to eliminate the effects of water content response inside the 0.1% major principal
changes reveals that it is the effect of the strain contour. This response appears to be
stress state that dominates. This has an typical of a wide range of soils (Jardine et
important bearing on sampling since, as will al., 1984) and has an extremely important
be seen, one of its effects is to change the bearing on their in situ behaviour and on the
effective stress state. interpretation of in situ tests carried out in
This same pattern of behaviour has been them (Jardine et al., 1985). The 0.1 % maj or
found to apply to other young sedimented principal strain contour will be taken as the
clays of higher plasticity. Figures 5 and 6 boundary to the small-strain region or zone.
show the effective stress paths from a simi- These features of the stress-strain prop-
lar series of triaxial compression and exten- erties can best be visualized by plotting
sion tests on reconstituted specimens of stress-strain data on a semi-logarithmic
London Clay and the North Sea clay. In the scale; data for the North Sea clay are shown
experimental work on these two clays, the
+1=300
/
100
50
-50
, ~ .oo
s]) "--...
'~ .'=30·
6 Pattern of undrained behaviour for reconstituted North Sea clay (axial strains indicated in %)
258 ADVANCES IN UNDERWATER TECHNOLOGY
140
R8
>
b -
~20
01---e--_---~
- ~OhO~I~----~0~OO~5~0~Dttl--------~0~,0~5--~0.~I--------'0~,5~~I~--------~5~~10
Axia l strain. £L ("10)
TABLE 1
Stiffness indices
(E U )O.Ol% L = E"'O.l)% p~
p~ E ",0.001)% OCR (kPa)
'In situ'
R1 830 0.185 1 267
R1.4 2180 0.270 1.4 206
R2 2270 0.353 2.05 158
R4 2130 0.386 3.73 106
R8 1740 0.407 7.4 65
*Before sampling
tEstimated
,VALUATION OF GEOTECHNICAL PARAMETERS 259
I.
o
RI •
agc~
1. R1.4~. \
R2'~\\
1.4 \\
u,
1.2
· \~4
¥,
''\
R8D-_ i-
to
~ww
: i
, ,
O.B
" 0.6
w
-'
0.4
02
sedimented clays are: imperfect sampling into block and tube sam-
piing. Although it is recognized that block
(a) the size of the small-strain region is sampling is not feasible offshore, its treat-
dependent on OCR and p' 0; ment allows the effects of sampling to be
(b) for each OCR, the initial stiffness varies introduced in stages. No account is taken
with strain, although the degree of non- here of the effects of advancing the bore-
linearity reduces with increasing OCR; hole.
(c) yielding, evident from a sharp deviation
in the effective stress path, occurs at PERFECT SAMPLING
axial strains less than 0.2%;
(d) major dilation/contraction and, there- The advantage of considering perfect samp-
fore, major distortion to the inital soil ling is that it is a well-defined path which
fabric begins at strains of 1 or 2%. illustrates certain features of behaviour
This general pattern of behaviour for that can then be anticipated in imperfect
sedimented clays, especially the location of samples. It is well known that, even for
the bounding surface and the Ei 1 = 0.1 % con- the hypothetical case of perfect sampling
tour, is modified by both rate of shear and there can be a change in strength and
ageing. These considerations are beyond the stress-strain behaviour from that for the in
scope of this paper. situ soil (e.g. Skempton and Sowa, 1963).
This is illustrated for the case of the North
SAMPLING EFFECTS Sea clay in Fig. 10; perfect sampling is
shown from two in situ conditions:
Traditionally the effects of sampling have
been considered by dealing separately with: (1) normally consolidated, perfect sampling
path AB and reloading path BC (test
(a) perfect sampling , in which the in situ PSI); and
shear stress is removed and undrained (2) overconsolidated soil (OCR = 7.4), per-
shear to failure starts from an isotropic fect sampling path DE and reloading
stress, and path EF (test PS8).
(b) imperfect sampling, in which some arbi-
trary stress path is assumed to be The path for reloading in test PSI has
applied before undrained shearing to moved inside the bounding surface provided
failure. by the stress paths for triaxial compression
and extension on the normally consolidated
We will retain this distinction and subdivide soil. The path intercepts the bounding sur-
G-~
2.0 30
Ax ial strain,Eo (.,.)
- 50 Normally consolidate d soil - 50
AG - ' insitu' path
AB- ptrlect sampling path
Be - triaxial compression afltr perfect sampling
Overconsol idated soil (OCR 7. 4)
DEF - 'insitu' path (R8)
DE - p~ftct sampling path
EF - triaxial compression alwr perfrlc t sampling (PS8)
face at a lower point than the path for the in of the zones and also alter their position in
situ soil, and, as a result, its peak undrained relation to the state of stress prior to shear.
strength and undrained brittleness are The following points should be noted:
reduced by perfect sampling. Since the
(a) When the direction of the stress path
water content has not been changed, the
reverses (as for PSI) the full small-strain
ultimate strength is unaffected and this zone is traversed; this can result in an
parameter provides a lower bound to
increase in (E JO.Ol% even if the overall
strength after sampling.
size of the zone has diminished.
The changes in the stress-strain proper-
(b) When the direction of the stress path in
ties may be viewed in terms of the two indi-
triaxial compression is a continuation of
ces (Ea)O.Ol%/P'O and L. The former, based on that followed during sampling (PS8)
P' 0 prior to shear, increases from 830 to 1371
some or all of the small-strain zone is
(cf. Rl and PSI in Fig. 9 and Table 1); the
missed.
latter increases from 0.185 to 0.404 (cf. Rl
(c) When the small-strain zone approaches
and PSI in Fig. 8 and Table 1).
the bounding surface it becomes dis-
For the case of the overconsolidated soil,
torted and the stress-strain characteris-
in which K 0 > 1, the perfect sampling path is
tics are modified accordingly.
identical to the initial section of the triaxial
compression path from in situ conditions. In reality, sampling involves a release of
There is, therefore, no change in strength. total stresses and a delay between unloading
However, the stress-strain properties of and subsequent reimposition of total stres-
the perfect sample have as their origin point ses and shearing to failure. Both delays and
E, as opposed to point D for the in situ soil. the reimposition of stress are likely to mod-
Part of the small-strain region has already ify the location of the small-strain zone.
been traversed along DE. Consequently, Evidence suggests that the distance to the
both indices of stiffness and linearity are nearest point on the zone's boundary is
modified; in this case (EU)O.Ol%/P'OO has been likely to increase with time.
reduced and L increased (see Figs 8 and 9 On the basis of the simple stress paths
and Table 1, and compare PS8 and R8). involved in perfect sampling, we can antici-
The contrasting changes in stress-strain pate that, as a result of imperfect sampling,
characteristics after perfect sampling of there will be changes in strength and
specimens with OCRs of 1 and 7.4 reveal an stress-strain response in subsequent
additional important aspect of sampling, undrained loading. Furthermore, the size
namely the direction of its final stress path and form of these changes will depend on the
relative to that followed in the subsequent initial in situ stress state and stress history,
triaxial tests. This can be illustrated in and on the directions of straining involved in
terms of the small-strain zone which, based the sampling process.
on the data for PSI, appears to be carried
around stress space. In Fig. 11 this small-
strain zone is sketched for specimens which
are:
(a) normally consolidated (OCR = 1);
(b) perfectly sampled from a normally con-
solidated state (PSI); and
(c) heavily overconsolidated (R8/PS8).
These zones have been identified on the
basis of the data shown in Figs 6 and 10 and DEF
EF
- R8
- PS8
r ""l
i... ...:
SU99qsted limits
to small strain zonq
on additional tests. Processes such as per- - 50
fect sampling can modify the size and shape Fig. 11 Small-strain zones
262 ADVANCES IN UNDERWATER TECHNOLOGY
. o
E
.
_J:
b
~ 033
-J:
b
-~
J:
b
0~T-----~~~L-~-LL-~0.~67'---------1~0'
o 4
80
(cr~ • 0:,) I (cr~ • cr~) max Eo ('1.)
50
25
(;
.,.
n.
~
xb
.
>
~
00 50 100
(cr~.cr~) 12 (kPo)
OCR before sampling (OCR = 2). It is evi- The levels of distortion which occur as soil
dent that both the initial stiffness and the enters a sampler have been most elegantly
degree of non-linearity have been reduced. elucidated by application of the Strain Path
Method (Baligh, 1984). Figure 14 has been
taken from Baligh's work and illustrates the
TUBE SAMPLING sequence of strains experienced by an ele-
ment of soil on the centreline of samplers of
In contrast to perfect sampling, the stress different geometry. The strains involve
or strain paths involved in tube sampling triaxial compression followed by triaxial
and subsequent extrusion are complex. We extension. Their effect can be assessed from
will simplify the problem and consider the the pattern of behaviour displayed in Fig. 3
conditions at the periphery of a tube sample and from the results of an extensive study of
and along its centreline. these soils under cyclic triaxial compression
and extension loading. In Figs 15 and 16 the
B
estimated stress paths corresponding to the
2.0 strain paths in Fig. 14 are shown for nor-
O.5
I1II mally and heavily overconsolidated soil. In
the former, there is a major reduction in
/ 1// effective stress, in the latter an increase in
I
ID
III O.5 /
BIt=~
2a/,1I 1H---lI-.-
effective stress.
N
t-- df'-... l~ c
Z
Z For certain geometries of sampler, the
.§ 0 -...., r- .JJ-- 1
""'r--"---'- strain levels predicted for an element on the
o
.2 -0 5
") b centreline in Fig. 14 exceed those quoted
c 'I Va v
V slt=IO i above as leading to gross distortion of the
.. -I0
E
OJ
1// \/ o initial fabric. There must, in these cases, be
uncertainty as to whether the true
IlL!
"0 -I 5
~ (Extension)
~ - 2'0_4 -3 -2
(Compr~Jsion) undrained characteristics of the in situ soil
-I o 3 4
can ever be measured on unconsolidated
Vertical strai n. E zz ·1.
retrieved samples.
Fig. 14 Strain paths for an element on the The distort1cn which occurs around the
centreline of a tube sampler (after Baligh, 1984) periphery of a tube sample is often apparent
Tubf: Sampli ng
Tube: Sampling
obc .. compre:s$lon cycll" ,
edt: - tJltension cyclq t See Fig, 14
c:fgh " extrusion
[j)-"
Fig. 16 Stress paths for tube sampling of heavily overconsolidated soil
when such a sample is split to expose its on the sample. Its path is indicated arbitrar-
fabric. Obviously the strain paths followed ily by efgh in Figs 15 and 16. Reductions in
in this outer zone are inadequately modelled effective stress which would result from the
in triaxial tests but we can, on the basis of pore-pressure parameter B being less than
Fig. 3, reasonably anticipate that: unity are neglected.
From the findings on perfect sampling,
(a) soil in an initially normally consolidated
or lightly overconsolidated state will the direction of the extrusion path relative
develop positive pore pressure incre- to that followed in subsequent triaxial test-
ments (Fig. 15); ing is of major importance in determining
(b) soil in a heavily overconsolidated state the initial stress-strain properties that are
will develop negative pore pressure observed. If the final phase of the process is
increments (Fig. 16). to unload from compression, then the soil's
stiffest response would be shown in tests
Since shear is actually taking place against a that recompress the sample.
steel interface, residual fabric will be The response that could be anticipated in
developed and the thickness of the distorted normally consolidated and overconsolidated
zone will be affected by surface roughness, soil after tube sampling and extrusion is
soil plasticity and rate of displacement. shown by path IS in Fig. 17. This response
There will be equilibration of the pore- is compared to that for the in situ soil, and
pressure differences between the centreline for the in situ soil after perfect sampling
and periphery of the sample. At equilibrium (PS) and block sampling (BS). For com-
there will have been an overall reduction in pleteness, the response of soil subject to
pore pressure in the heavily overconsoli- gross distortion (G) is also shown. Even
dated soil and an overall increase in the after extrusion, further changes in effective
normally consolidated soil. Correspond- stress state can occur as a result of: changes
ingly, there will have been redistribution of in temperatures; gas exsolution; drying;
water content. sample trimming and installation. Provided,
Extrusion involves additional distortion, however, that these changes are not exces-
somewhat analagous to triaxial compres- sive, the basic effects of sampling will not be
sion, and final removal of the total stresses altered.
266 ADVANCES IN UNDERWATER TECHNOLOGY
3
C
a.
~ 200
N
;::.
6'I
>
b
100
Fig. 18 Stress paths for triaxial compression tests on tube samples from the Northern
North Sea (axial strains indicated in %)
TABLE 2
Comparison of 'in situ' behaviour with that measured on tube samples of North Sea clay
Compression to Higher stiffness than in Behaviour similar to Stiffness lower than expected
ea = 0.1% situ, more pore pressure in situ in situ, less pore pressure
generated generated
Compression to In situ sample fails, strain 12 shows yield and slow In situ sample yields and
ea = 1.0% softens and produces large strain hardening, with some behaviour is similar to 13
pore pressures. 11 stress pore-pressure generation
path climbs almost
vertically
Compression to 11 shows a slow strain 12 shows strain softening 13 shows strain hardening in
ea = 10.0% hardening with sample but with less pore-pressure a similar way to in situ
dilating. In situ behaviour generation than in situ behaviour
shows continuous strain
softening and pore pressure
generation
Peak strength Less than in situ Greater than in situ Approximately equal to
in situ
Axial strain at 11 8%, in situ 0.1 % 122.5%, in situ 0.1% 13 25%, in situ 10-15%
failure
EVALUATION OF GEOTECHNICAL PARAMETERS 265
__ - ,n .itu
Normaly Consolidated I __
~~-==::==-~-
Allie! 5train
PS - pcrf~ct sampling
85 -block sampling
rs -lmpcrfttt sampling (_ .' . small strain bmits
G - 'ijross dislurbanc~ \ ._.1 ( Diagrammatic)
AKiol strain
generally slightly lower than those expected Nevertheless, UU tests can be used to
in situ, but the large-strain behaviour determine ultimate strength, although
becomes progressively more realistic with efforts must be made to minimize the oppor-
increasing OCR. tunities for uncontrolled water content
change.
The use of the measured strengths in the
SUMMARY OF SAMPLING EFFECTS ratio C J (1' v allows an estimate to be made of
the in situ degree of overconsolidaiton. Such
Sampling modifies the effective stress state apparent OCRs give some guidance in
in an element of soil so that the behaviour evaluating the in situ horizontal stresses,
measured in unconsolidated undrained triax- but there is a clear requirement for the
ial tests does not match that of an element in independent measurement of this stress in
situ, subject to the same stress path. Impor- situ.
tant differences in stress-strain charac- For samples estimated as having a high
teristics, peak strength and undrained brit- OCR, isotropic swelling to the in situ mean
tleness can occur. In these young sedi- effective stress has been found to provide a
mented clays, peak strengths for triaxial reasonable model of in situ behaviour. This
compression may be reduced by as much approach was proposed by Gens and Hight
as 30%; the ultimate strength is not (1979) on the basic of work with the Cowden
affected, provided that water content Till. Reimposition of anisotropic stresses, if
changes are avoided. Stress-strain proper- these are known, will improve the fit for
ties are modified by the changes in size and stress-strain properties.
location of the small-strain zone which For soils estimated as being normally or
accompany changes in effective stress state. lightly overconsolidated engineers face a
The potential effects of different forms of dilemma. Reconsolidation to estimated in
sampling on normally consolidated and heav- situ stresses does not completely recover in
ily overconsolidated soil are summarized in situ behaviour (refer to test 12 and Table 2).
Fig. 17. Reconsolidation beyond the in situ stresses
following the SHANSEP approach (Ladd
and Foott, 1974) recovers the features
RECONSTRUCTING IN SITU observed in young clays, but eliminates the
BEHAVIOUR FROM TESTS ON TUBE potential effects of ageing and cementing.
SAMPLES We therefore advocate an overlapping
approach which relies on:
Provided that the sampling process does not
involve excessive strains, the behaviour (a) UU tests to identify special features of
measured in unconsolidated undrained triax- behaviour, such as bonding;
ial tests fits into the overall pattern of (b) tests run on samples anisotropically con-
behaviour for the soil. However, the impor- solidated to estimated in situ stresses
tant effects of stress history are largely which, together with the results of UU
obliterated by the changes in effective tests, enable the small and intermediate
stress state which accompany sampling strain response to be bracketed; and
(Fig. 12). Unconsolidated undrained triaxial (c) either the interpretation of the data and
tests are not, therefore, a good indicator of the evaluation of design parameters
stress history or of detailed in situ within the framework described in this
behaviour. Furthermore, measurements of paper, or the performance of a suite of
initial effective stress on relatively small SHANSEP tests on identical samples.
tube samples retrieved from the seabed are
unlikely to provide a reliable means of With regard to interpretation, recognition
:;timating the in situ stress state. should be given to the existence of:
268 ADVANCES IN UNDERWATER TECHNOLOGY
ABSTRACT INTRODUCTION
Traditionally, North Sea sands have been The use of in situ tests is particularly impor-
tested by laboratory tests on reconstituted tant in sands where it is often impossible to
samples and by in situ cone penetration take representative undisturbed samples.
tests. A number of other in situ test devices Traditionally, in situ testing in connection
are already available for offshore use or will with offshore soil investigations has predo-
most likely be available in the near future. minantly consisted of cone penetration tests
These include the piezocone, the pressure- (CPTs), including frequently for the last
meter, the dilatometer, nuclear density three years measurement of pore-water
and electrical resistivity probes, the seismic pressure (piezocone). A few other in situ
cone and the screw plate. tests applicable in sandy soils have been
The present status of interpretation of the used in selected cases, and other in situ
above tests in terms of the following soil tests that have proved particularly useful on
design parameters is discussed: soil density, land will undoubtedly be used offshore in the
drained shear strength, deformation charac- future. These include the use of pressure-
teristics, initial shear modulus, and soil meter, dilatometer, special sand-density
layering and soil identification. In addition, methods, and tests measuring shear-wave
the need for obtaining samples and perform- velocity and screw plate.
ing laboratory tests is stressed. The results Some of the interpretation methods have
{rom in situ tests give the basis for recon- a purely theoretical basis, but the most
Rtituting specimens for laboratory testing commonly used interpretation methods are
soil density and consolidation stresses). semi- or purely empirical. In the latter case,
269
270 ADVANCES IN UNDERWATER TECHNOLOGY
the methods are mostly based on research • Test must have been shown to given
data from large calibration chamber tests meaningful results in terms of soil design
and to some extent on actual field tests per- parameters in sandy soils.
formed at uniform, well investigated test
Cone Penetration Test (CPT)
sites.
The cone penetration test measures cone-tip
resistance, q c' and sleeve friction, is. The
newer devices (piezocones) also measure
IN SITU TEST TYPES CONSIDERED pore-water pressure (u). The standard cone
has a 60° apex angle and cross-sectional area
The test types selected for the present of 10 cm 2 (Fig. la). In the North Sea, 15 cm 2
review had to obey three criteria: cones are frequently used. For more details,
see, for instance, de Ruiter (1982).
• Test must be applicable in sand.
• Test must be operational for offshore use Special Density Probes
or likely to become operational offshore in Special density probes directly or indirectly
the near future. measure in situ density or porosity. They
.. .
..
O,ci llolc:ope
Data transmission
through cable
.....
and photomul tiplierl
lead shield
/ShO,. We"
Radio-active source
(3mCi isotope 137 caesiUII)
All dimensions in mm
44""
After Tjelta et. 0.1. (198S) S. i,m ic
Cone Pltnetr ometu
cl
d) After Campanella nnd Robertson (1981,)
Fig. 1 Sketches of in situ devices used in sand: (a) cone penetration test with pore-pressure
measurement; (b) electrical resistivity probe; (c) nuclear density probe; (d) seismic cone;
DESIGN PARAMETERS FOR OFFSHORE SANDS 271
have been developed by the Delft Soil determined by comparing the energy level
Mechanics Laboratory. of the emitted photons from the source and
The electrical resistivity probe measures the photons entering the detector after
cone resistance and sleeve friction in addi- backscattering from the soil mass. The
tion to the electrical resistivity of the soil equipment is calibrated by lowering the
volume (Fig. Ib). A separate probe meas- probe into fluids with known densities be-
ures the resistivity of the pore water. Read- tween 10 and 22 kN/m3. The nuclear density
ings are generally taken at 0.2 m intervals. probe was recently used offshore (Tjelta et
The field measurements are correlated to al., 1985).
laboratory calibrations on specimens recon-
stituted to different porosities (see Kroezen, Seismic Cone
1981). In the seismic cone a piezocone is combined
The nuclear density probe has a radioac- with a set of miniature seismometers built
tive source and detector built into a 15 cm 2 into the cone (Fig. Id). The device was
cone (Fig. lc). The bulk density of the soil is developed at the University of British Col-
....
.-
....
L.
L.a.
:::I 0
.. > Flush
L.
1>.0
..
Water/Mud
Return
So ill Mud
Pres sure Drill
pipe
Spacer
Expanded E
Trans- Membrane :
330111111
ducer (2)
Helllbrane
wi Chinese
Honitored ~
8y 3 arlu I
lantern
DIMENSIONS IN mill : Soil cutter
Cutting
ONSH. OFFSH.
shoe ~ = 78mm Cutting shol
w 94 77
t 14 16 .. = 80mm
~ 60 53
PUSH-IN SELf-BORING
- offshore (CAMBRIDGE)
- onshor e
e) DllA TOMETER f) PRESSUREMETERS
t
e--load cell
Displ. Transducer
Rods
L:::::::\ ~~~j1- Cross Bealll
H .. = lS0mm
g) SCREW PlATE TES T.
(onshorel
umbia (Campanella and Robertson, 1984). fb', dilation angle and drained shear modulus
The cone resistance, sleeve friction and are derived from the field curve.
pore-pressure measurements are used to The self-boring pressuremeter (SBP) is
determine the soil layering at a site during also used offshore (Fay and Le Tirant,
penetration, and the downhole shear-wave 1982). The device tunnels itself into the
velocity is measured at intervals (normally ground and should induce the least distur-
1 m). A profile of the in situ dynamic shear- bance of all the in situ devices. The experi-
wave velocity, and initial modulus (i. e. at very ence offshore with the SBP is so far limited.
small strain level) is therefore obtained. When successful, self-boring tests offer the
During on-land tests a sledgehammer is best alternative for measuring the in situ
used to generate the shear wave. The wave horizontal stress in a sand deposit.
generation system is presently being mod-
The Screw Plate
ified for use offshore in the Beaufort Sea
The screw plate (Fig. Ig) can perform load
(Campanella, 1985).
tests at depth more economically than plate
On land the crosshole technique is also
load tests. The plate consists of a flat pitch
used to measure shear-wave velocity. How-
auger screwed into the ground and loaded
ever, it is not practical to perform this test vertically (Janbu and Senneset, 1973; Ber-
offshore except perhaps underneath already
zins and Campanella, 1981). The settlement
installed platforms.
versus time is measured at selected depth
Dilatometer Test (DMT) intervals, but penetration is limited to about
The dilatometer test measures soil contact 15 or 20 m in a medium dense sand. The
pressure and the 1 mm expansion pressure screw-plate test has been used only onshore
(Fig. Ie). The device, 94 mm wide and 14 up to now.
mm thick for onshore use and 77 mm wide
and 16 mm thick for offshore use, provides STRATIFICATION AND SOIL
nearly continuous data with depth. The test IDENTIFICATION
is performed very rapidly, as the expansion
test every 20 cm takes about 15 seconds. General
Empirical correlations enable one to predict
in situ horizontal stress, stress history, con- It is always important to establish a clear
strained modulus and effective friction picture of the soil layering or stratigraphy at
angle. The device is promising in sand
a site. In situ tests are extremely useful in
where few reliable methods for obtaining K 0 this context, and they are a necessary sup-
are available, with the exception of the more plement to laboratory tests on obtained
sophisticated pressuremeter. Robertson samples.
and Campanella (1984) also use dilatometer Very frequently CPTs are performed
results to predict the liquefaction potential prior to sampling and a careful interpreta-
of sands. The device has been used offshore tion of CPT results can yield optimum plan-
in limited water depths (Burgess et al., ning of the sampling programme. Further,
1983), but a device for deep-water use is since performing in situ tests is usually
under development. much cheaper than obtaining and testing
Pressuremeter Tests samples, in situ test results are used to
Pressuremeter tests provide in situ extrapolate the laboratory results and to
stress-strain curves (Fig. If). The push-in interpolate between boreholes.
pressuremeter (PIP) was developed mainly
for offshore use (Reid et al., 1982). The tests Cone Penetration Test
are usually carried out alternately with
other tests in the same borehole. Tests in It is generally accepted that the cone penet-
sands are fully drained. The friction angle ration test with measurement of pore pres-
DESIGN PARAMETERS FOR OFFSHORE SANDS 273
CONE RESISTANce q, IMP"
o
..
0.05 0.10 0.150.20 0.25 0.300.35 1.6 1.8 2.0 2.2 0 0 .4 0.8 1.2 1.6 2 .0
SOIL SLEEVE FRICT ION f IMP I ,
DENSITY (II " PORE PRESSURE u IMP"
oP~
~ r r> ~
1 I
I
'(,
)
"'\
<r--.,
l.
r. ..-- ---
10
c='
~
~~ . F5! iI:::::..
.'
~E:.
~ "'iii.. I~~
s
~ -= ~
.:,
20
t" -. < 1"""""'1\.
-
-:, .. l!=-
- ._-
--
,::.
.. .
-'1l"
_0.
-!~~
..:!
I:t:-
~
$,-
_ .i'':;'-
1-
::z:
=-
~
~
~; '-!!!!
,~ iii;;.,. I~
'" I
",r;>
~
$
J~ .::; ~~
.~~ Cone resis tinu I'-~
.<:; "
;--
Fig. 2 Piezocone and nuclear density test results, Gullfaks 'C' (after Tjelta et al., 1985)
sure is the best and most efficient tool for piezocone in determining soil stratigraphy is
logging stratigraphy at a site. A thin pore- evident from this figure.
pressure filter (either at the tip or immedi- With reference to soil classification from
ately behind the cone tip) can sense layers CPT results, several charts relate the fric-
only a few centimetres thick. However, tion ratio to soil type. The most comprehen-
Robertson and Campanella (1983) point out sive work on this has been done by Douglas
that, for interbedded deposits, the thinnest and Olsen (1981). Their chart (Fig. 3) corre-
stiff layer for which the cone resistance lates CPT data with other soil-type indices,
measures full response is about 10-20 cone such as those provided by the Unified Soil
diameters. For the standard 10 cm 2 cone, Classification System_ The measured sleeve
the minimum stiff-layer thickness to ensure friction is usually far less accurate than the
full or plateau response of cone resistance is cone reading, and more liable to variability
therefore about 35-70 cm. from one cone to another. Among other fac-
Figure 2 shows as an example the results tors, the pore pressure induced during cone
of a piezocone profile performed at Gullfaks penetration affects the sleeve friction meas-
'C' in the Norwegian sector of the North Sea urements (e.g. Campanella et al., 1982).
(Tjelta et al., 1985). The usefulness of the For classification purposes, one should
274 ADVANCES IN UNDERWATER TECHNOLOGY
.00
10
60
14
<0
10
10
• IB = u, - u•
q q, p
0
~
~;
z
«
::l'"0: 4
~
1
z 0.1
S
0.6
0.,
1.2 Bq
0.1
r-J 11 e
~i· ~~ .::::
~; ~ ~ 500
~
1ft
.a CLAY
~ ~' ~ vi
=>
10 r--' - -- ~ p~ 5--' 200
~. ~ ~D_~
0-
C
I ....
o
! a""
... .
I ---J
20
:
j ~. I
25 ..- · f - 1-
<tt{ . i
)0 . 1 MUD OR PEAT
I I
ol 5 ~~~~_~~~-L_~ __~
0.1 0.2 0.)3 0.6 0.8 1.2 1.8
o Medium sand ( 2 soundings)
b) MATERIAL INDEX, 10
• Plastic sensitive clay ( 2 soundingsl
~ ,,26 - 40% (Lacasse and Lunne, ((hart also gives unit weight - not
1982) s~own here).
For years the relative density (DR) has mine the engineering behaviour of a sand.
been used by engineers as an important Other very important factors include stress
parameter to describe sand deposits. For history, age, sand structure, grain size and
example, the in situ relative density is shape, mineralogy and cementation.
often evaluated from CPTs or standard It is essential to have a good estimate of
penetration tests and used with empirical the in situ soil density or relative density
relationships to evaluate the angle of inter- since running laboratory tests on sand
nal friction, the liquefaction potential, etc. requires recompacting the material to an in
Recent trends try to bypass the inter- situ density. For these reasons, a reliable in
mediate determination of DR and rather situ method for measuring sand density is
correlate the in situ data directly to the very important.
relevant engineering parameter. The most
important reasons for this are that the max- Electrical Resistivity Tests
imum and minimum laboratory densities
vary with the testing method used and that The in situ porosity of a sand can be deter-
the relative density alone does not deter- mined from electrical resistivity measure-
276 ADVANCES IN UNDERWATER TECHNOLOGY
101- ----4-1--1-----i
•
0
_.. :>-
....:
or---' I':.>
I) '
I'
x
iL
'"o '01-----'1>------ t - - - I
-f--- . ~
.(
0
.. I---+-~~t---I • ~
~~
.0 - .c;.....j" --1----1 0 ~
or Sornl)IQ
Fig. 6 Example result from electrical resistivity test (after Kroezen, 1981)
ments and laboratory calibrations. For this the penetrated sand. Figures 2 and 7 show
test, sand samples have to be taken the results of a test performed in 220 m of
throughout the sand profile. It is therefore water at the Gullfaks Field in the N or-
routine also to determine maximum and wegian Sector of the North Sea (Tjelta et
minimum porosity, and to calculate the rela- al., 1985). The detector is placed 210 mm
tive density, DR. Figure 6 shows an example from the radioactive source inside the 15 cm 2
of measurements (Kroezen, 1981). The elec- cone, and the result of this test is perhaps
trodes of the soil probe are 0.5 m apart, and more influenced by the disturbance around
it is expected that the bulk of the soil in the penetrating cone than in the case of the
which the average soil resistivity is meas- electrical resistivity test. However, the test
ured is within 0.1 to 0.5 m from the probe. It is less cumbersome to perform than the elec-
is therefore thought that the disturbance of trical resistivity test where two probes need
the soil caused by the penetrating 10 cm 2 to be used and soil samples are required for
cone in most cases does not influence the calibration purpose. The nuclear density
resistivity readings significantly, i.e. that test can also be used in cohesive soils. As for
the measured soil resistivity is representa- the electrical resistivity tests, the accuracy
tive for undisturbed soil. However, little of the nuclear density test needs to be
data have been provided which can be used documented.
to check the accuracy of the electrical resis-
tivity probe. Cone Penetration Test
Nuclear Density Tests The most commonly used method for deriv-
ing relative density from CPT results is that
This test gives directly the soil density of given by Schmertmann (1978), who per-
DESIGN PARAMETERS FOR OFFSHORE SANDS 277
IlEI/SITY ItI,.l, IlEI/SITY It 1 ..', CONE RESISTANCE, q , (MPal
,
2. 2
00~_1~0_~20~~3rO__-+___5~0_-,__-,70
-20 1~/===;~::;;:;~
.....
o = -1-ln~100 %
:loo,......::y ----j r 2.91 61o~
q, and c', in kPa
IE IE
:-
....
~ -1 0 1--_........,.... .,. . -1 r
fu
C>
-30 1--- """"';jHr-j -'
<t
~
~ 250
u.J
>
:::""
.....
w
-35 I--- - -- j ~ 0, =20% 40% 60% 80% ' 90% 100%
~ 350 1----1--+----1---\--·-+----"[-- 1
Rela tions hip ,ugge,lod by S,h .." t m.nn 1191al dashed
400 L __.l.-__....L.__--1...__--'-__---L_ _ _ _' - -- - - '
-40 L -_ _ _- - '
Fig. 8 Recommended relationship among (J~, qc
and DR for NC fine-medium quartz sand (from
+ DfIlSITY ~ASUREO FROM SAMPLES
Lunne and Christoffersen, 1983)
o IlEI/SITY COMPUTED FROM WATER CONTENTS
- IDISITY FROM IN-SITU DENSITY TEST
~ .....-
ent overconsolidated cone resistance, qC,co ,
may be found by the following formula sug-
r
;,v~ a gested by Schmertmann (1978):
5 a
qc,oc = 1 + /Ko,oc - 1) (4)
a
?:> a qc,ne Ko,ne
.,.~ a where
10
16
, K
,~ a
'.Q a --.2e: = OCR P (ratio of Ko-value for
~
'Measured
on obtained
Ko,ne normally consolidated and
15 '.~ sam~les
overconsolidated sand)
e ~~
~
I-
...,
a.
, Q
'\ .
-9
, Q
,{ and f3 are experimental constants, and nc
Cl and oc denote normally and overconsoli-
~"
a
20 a dated respectively (Schmertmann sug-
gested ,( = 0.75 and f3 = 0.42).
'! Jamioli<owski .t . • 1.
Baldi et al. (1985) tentatively suggested
l unn. and '~V 11985)
that P increases with DR' i.e. P = 0.27 +
(hrist offorson (1981) /, ~ a
25 • a 0.26D R , and that 1[ decreases with OCR,
Schm.. tmann (1978)/
a being about 0.50 for OCR = 2 and 0.25 for
OCR = 15. The authors recommend the use
a
30L-____________~a____________~
of Baldi et al.'s findings when evaluating
qc,ne, and hence DR, for overconsolidated
Fig. 9 Relative densities computed from CPT sands. Unfortunately, a reliable assessment
results in medium-to-coarse sand of OCR and/or Ko of natural sand deposits is
a very difficult task. However, as discussed
As an example, the cone resistance (mean of in the next section, the dilatometer shows
42 tests) for a normally consolidated medium promise in this respect.
sand has been interpreted using the
methods listed above (Fig. 9). Generally, Dilatometer Test
the absolute value DR predicted by
Schmertmann's correlation is 10% higher Robertson and Campanella (1984) proposed
than that predicted by Lunne and Christ a relationship between dilatometer para-
offersen's chart, while the DR-values based meters and relative density. Empirical data
on Jamiolkowski et al. (1985) lie slightly are available for DR between 15 and 70%.
above Schmertmann's values.
Figure 9 also includes relative densities as
derived from measured densities on 'undis- IN SITU HORIZONTAL STRESS
turbed' samples. The CPT predicted lower
densities than the ones based on the actual General
sand samples. Because the samples are
likely to become densified during sampling The determination of the horizontal stress
and handling, the authors believe that the and the coefficient of earth pressure at rest
CPT values may be more representative of in sand is probably one of the most difficult
the in situ conditions. tasks of in situ testing. The action of the
The q c' (1' va' DR relationships referred to measurement itself defeats the purpose of
above are only valid for normally consoli- obtaining an at-rest condition. The more
DESIGN PARAMETERS FOR OFFSHORE SANDS 279
promising devices for such measurement are the disturbance caused by the insertion
the self-boring pressuremeter test and the method.
dilatometer tests, although both methods
present drawbacks. Dilatometer Test
l
o
:
,
O!
0
I B' ,I ~,
I ! : ! I
~ o~~ " t '
~-j----1-
0
I " , I
~O, OJ
' ~:,I I ", ' I i
0
I
'I
0 0 0\
0 ,0 "'1
o '8 II
Ii
0 0," '" '0 .....
0,
0
0 1 ~ I. ' r-~ 0.1
't> I,
0
0
o ~,
'I" ' ~ 0 0 ' 4~
r
~!o I .", ... o
,
0
0
-..~--l-
0
__ ~
0.\ I oo.~ ~"'"
o o~oo ~
&. 'co ,
o
;p .1>. 0 ","
10 ....
0
o ~o , '& ~ '" ..
,
80 0 , , " 0
°
~~~.. o :
~~ '" 00
° 0
, "
'<>=l to 0 °
to L.
..
0,
,0
0° ,1
8 ,
, ,' ," ,
°0
00
0
,0
~: " , ,, 0 , ,
- c - -0 ,,
° , 0
o 0 .......
0
, 0
, ,
Vc 0 ...... 0
~~t.o "'';.
0
08 0 " ,0
o _9,0 ° ,
I~~ o~ u 00 t 0 ,,
~ :0
0
~ ...
'0
: 0
,8 ,
....
0
, °
,
~5 ,, 10°'/,£ '
o (I)..~~.
10 W"~ to
o 8~~ 0 1 ....
14'ig. 10 Dilatometer parameters I Dand KD and derived Ko and OCR in medium-to-coarse sane (four
profilings)
280 ADVANCES IN UNDERWATER TECHNOLOGY
0..
Au- a.
"to'" .0
Sand
10 MP ,
600
500
40 0
be t ... een Ian 'll...k and
N. fo r nc. sands
6u-0
\ ~ 300
:5
o
\
\
a: a:
100 \ 0 200
=>
\\---~.:-:'.,
""a:
0-
W
UJ
> :1.
o
w >- 100
> <-
80
E
w 100 6u> 0
U
~
0. 60
u..
tb kP,
~
u 50
40
'-'l Meyerhof (19511
;;; 30
a: General bea r ing
la) Principle of interpretation
~
UJ
cap acit y fac t ors
1000
"" 20 N q (modifoed
Prand t l solu hon )
c:r for strip foundation
z:
0:::
600
~-
__ ~'. .' ..~.
~~~~<t;':::'::.!._ n=-15· 0.8 0.9 ~O 11
0
300 '" FRICTIO N TAN 'll~.'k
1-
LJ
200
Definit ion of j1
n=O· Fig. 12 Bearing capacity factor N q vs tan ¢'
'" '"
<:(
n=·lS·
LL /
>- 100
// /' peak from tests on NC sand (after Lunne and
t:::: / I.~/ ./ '" Christoffersen, 1983)
60
///1
LJ
<:(
a.. / '"
<:(
LJ 30 //
'" /"f
L:J
z:
20 '" I." / tion or pile-driving resistance, a high ¢' may
,,~" "
Range recommended
0:::
<:( 10 ~~y Janbu and Senneset be considered conservative. If the Janbu
I..LJ
co
/" 119751
Note : Nq = ~
CIc.a and Senneset method yields a value for
5 0.4 0.6 0.8 1.0
attraction (cohesion), this must be taken
into account when selecting the design
FRICTION. TAN 'P'
value. For sands with low relative density,
Fig. 11 Janbu and Senneset's interpretation the structure of the sand may significantly
method of drained strength in sand influence the triaxial (and in situ) ¢' value.
In forming sand for tests in the laboratory a
Find DR from the revised q c> (1' v, DR
chart in Fig. 8 and carry out triaxial
tests on sand specimens reconstituted to
different DR values. Find 4>' by interpo-
lation. If no triaxial tests all available, 4>'
may be found by the chart given in Fig.
13, as recommended by Schmertmann.
For overconsolidated sands, OCR and Ko
should be estimated by using the relation-
ships recommended above under the head-
ing 'Pressuremeter Test'.
For both NC and OC sands, the use of the
more conservative ¢'-value computed by the 10 20 ]0 &.0 So '0 10 .0 90 we
three methods for the problem at hand is RElA TlVE DENSITY o,lPER (ENTI
recommended. For computing bearing
capacity, a low ¢' is considered conserva- Fig. 13 Relationship between ¢' and DR sug-
ive, but for the calculation of skirt penetra- gested by Schmertmann (1978)
282 ADVANCES IN UNDERWATER TECHNOLOGY
As mentioned earlier, two methods exist for Fig. 14 Effective friction angle predicted from
obtaining an effective friction angle in sands dilatometer test
from dilatometer tests. Figure 14 gives
examples of friction angles obtained in a
medium sand before and after compaction.
The two methods provide slightly different
answers. However, it is not clear whether al. (1977), is the one most commonly used
the dilatometer friction angle should be cor- today. The field expansion curve is plotted
related with the friction from a drained or on log-log scales, and the linear portion of
undrained triaxial test. The dilatometer the curve gives a slope related to the friction
penetrates very quickly, and one should angle (Fig. 15b). The inital strain can also be
consider whether the bearing capacity fail- adjusted if necessary. The friction and dila-
ure takes place under undrained conditions. tion angles are found by assuming a fIJ' cv
Future developments with measurements of value in Fig. 15c. Figure 16 compares the
pore pressure in connection with the expansion curves of push-in and self-boring
dilatometer test will reduce or eliminate this pressuremeter tests in sand from the same
uncertainty. depth. Neither of the tests predicted the in
situ horizontal stress well. The two tests
Pressuremeter Test result in different effective friction and dila-
tion angles. It is interesting to note that the
The pressuremeter test results may be two expansion curves are very similar, and
interpreted, as illustrated in Fig. 15, to that the PIP test appears to have had a
determine the effective friction angle fIJ'. translation of about 100 kPa along the stress
The method shown, proposed by Hughes et axis.
DESIGN PARAMETERS FOR OFFSHORE SANDS 283
200 600
c; . ' : 29'
v : 0'_ _ -
~ 500
a --1 It'''"
.,...- ~
0.. Q.
111
~
,.; ~ 400 J
/" ~
~100r---~~---r----+---~~~-4 to
.... ~
--
II>
/
]00
-siP 7-
/-
g:
b
...
.....
-- I
~
Q. 200 .' : 34'
~
o ;;i . 'n : )0·
~
O~O----~--~----~----~~~
4 8 12 16 20
e 100 /
",'" v : 10'
]0
behaviour (see 'Initial Shear Modulus'
below),
~ Gj'cy: friction
20 angle at constant
.....
I ....
i 60
I:
, , - -----r----,----.--------r------,.,---- -----. tion, which is related to crush ability of sand
-i 1. 0 - -- ____ _ grains.
vi
:::>
5 110
However, despite these limitations,
o
o
I:
engineers need correlations like those given
100
above between M and q c because there is a
'<I: 10 lack of technically and economically valid
'"o~ 60
alternatives. For future applications, the
c relationship dilatometer may become an attractive alter-
8 1.0 ~
native.
z
«
~z
8 Dilatometer Test
10 20 30 .0 50 60
(ONE RESISTANCE. q . HP.
Marchetti (1980) provides empirical correla-
LEG EN O,
-
NGI
- - BALDI ET A[(19611
i .. ........
_ .. _ .. -
WESS 119751 } Field
NGI tosts
tions for the calculation of the vertical con-
strained modulus from the horizontal stress
- ' - (HAPMAN 119691
- - - - - - THOMAS 119661
(HAMBER NOTES ESTIM ATED APPROX LINEAR
TESTS CU RVE FIT CORRESPONDING TO M.
index K D, the material index i D, and the
_ . _ . _ •• VE IS MANIS 11975) IN THE FIELD dilatometer modulus ED' The correlations
- lost.' 965 are known to work well in Norwegian clays
Fig. 17 Summary of relationships between (Lacasse and Lunne, 1982) but have been
constrained modulus M 0 and cone resistance qc little used in sand. Figure 18 gives examples
for NC sands (based on Lunne and Christof- of the constrained modulus obtained from
fersen, 1983 dilatometer tests on a medium coarse sand
before and after compaction and compares
them with modulus values obtained via
sands should be estimated as follows: Janbu's (1970) relationship for normally con-
solidated sand:
Mo = 4 qc for qc < 10 MPa
M 0 = 2 qc + 20 MPa for 10 MPa < qc M = m(po'Pa)1/2
< 50 MPa where M = constrained modulus; m = mod-
Mo = 120 MPa for qc > 50 MPa ulus number = 150-250 for medium sand; p 0'
For overconsolidated sands with OCR > 2 = effective overburden stress; P a = refer-
they proposed: ence pressure = 100 kPa.
The comparison of the modulus based on
Mo = 5 qc1 for qc < 50 MPa Janbu's relationship and the modulus based
Mo = 250 qc for qc > 50 MPa on dilatometer tests on the normally con-
The constrained modulus applicable for the solidated sand before compaction is rela-
stress range (1' vo to (1' vo to 11(1' v can be esti- tively good.
mated as
Screw-plate Test
M = M0 = ((1' vo + ,11(1' v/ 2 )! In general, the first load increment of the
(1 vo screw plate is p' 0' and the load is then
increased in steps to 500 or 1000 kPa. The
The relationships given above are meant to deformation pattern is shown in Fig. 19a.
be used as a guide for a first estimate of The sand volume compresses vertically, and
settlements. As pointed out by Jamiol- a small shear deformation also takes place
kowski et al. (1985), both M jq c and E jq c horizontally. This shear deformation is
depend on several factors including stress believed to have little significance at low
and strain history, stress level, relative stresses, but becomes significant at high
density of sand and mineralogical compos i- stresses. One should calculate the modulus
DESIGN PARAMETERS FOR OFFSHORE SANDS 285
'.ONSTRAINEO M
ODUlUS. M=lIMV IMPA I
o
o
0
0
15
• . 30
01-
45 60
(
15
..
90
.. • •
o 0 0
LEGEND<
0 • Oedomeh r tests on
0
•
• •
0 " undisturbed" somples
00
\,
00
00
.. 0\ ~ 'Screw plate tests
0 0 o
.. .
0
6~:~ ....
~
. .
0 '\ Oilot ometer tests
.. . • • . ..
0
0"
0
00
0
0
.. '2 t--~t---i~------i
)~ .. . o 0
.
~ , ILEGEND'
0 A. 0 Be-fore (Qmpoctlon
.. \~ ....
'0;.~ Aft., compac tion
101 Medium dense sand Ibl Medium 10 tine sand (att .. Aas. 1980)
Fig. 18 Constrained modulus from dilatometer test: (a) medium coarse sand; (b) medium-to-fine
sand
~ _ 02_
1, P.
T (.. ll~'",,+-
DE DOMETER E :.:~ I ,,' •• 10tH/a
- Coarse Sand} "1osjoen
6 = Volume change ...
oS . • 110kNl. 2
........ Fine sand
- Coars e silt.
south
0--
:z Verdalsora
w
:E:
w
-'
0-
0-
"., .,
w • • I'OIlNl.'
VI
' \
~ -',, ' !2.S. I
E .\ "" 'r:.rH~ltr
, I ,l E
oS
w-
1-1-
<C:z
E I
1
~ . . " f~: I-
:z
w
Mos jiien south
-'w
a..:E:
, , 1)111: I \~ i'.. j'". 0
a:i
-'
- Fine sand (10/:21%)
....... Medium sand (101:21%
:J: ~ l-
wI-
<rl-
ww
VI Vl S _\ 'HO ~S \O t;
VI
•• _. Coarse sand (101:22%
• ~
,.
cl
!<'ig. 19 Results of screw-plate tests in different sand/silt types (Aas, 1977): (a) deformation
n'lttern in oedometer and screw-plate tests; (b) time-settlement curves in 3 soil types; (c) compres-
sion curve from oedometer and screw-plate tests in three sand types
286 ADVANCES IN UNDERWATER TECHNOLOGY
based on screw-plate tests for the stress One dilatometer test was also carried out.
range p' to 2p' o. The screw-plate test pro-
0 Figure 18 compares the constrained mod-
vides a time-settlement curve which can ulus based on oedometer, screw-plate and
be used to calculate the coefficient of con- dilatometer tests. The agreement of the
solidation, as shown in Fig. 19b. Figure 19c screw-plate and dilatometer tests is very
compares screw-plate and oedometer com- good.
pressibility curves on three sand types. The Aas (1981) compared the modulus num-
screw plate tends to give a higher modulus bers backfigured from settlement measure-
numberm(m =dM/da'v) than the oedometer ments with the modulus numbers obtained
tests (Aas, 1977). from screw-plate and oedometer tests at six
Screw-plate tests were run in a loose sand sites where silt and fine, medium and coarse
in Drammen and compared to the results of sands were found (Fig. 20). The modulus
oedometer tests on 'undisturbed' samples. numbers are presented as a function of
1000
.
800
.
Sere" plot.
Oedom~ti!'r
- -----....
.
, E
",,-
600
I
t •
.
.,•
~
z
VI
400
3::> . r- ~ ~
I ..
~
200
l oose groyel ~
~
\" ~
• 0
o
\. • . .~~
~
o 10 20 30 40 50 60
MEASURED POROSITY n %
400 1 - - -- 1 -- \
E
ffi3001----~~
~ Hosjoen SQuth
~ D'Qmmon (11 SITE SA.OO
~ 200 I----I-----i~ Hosjoen Fine to coorsi!'
i 100 I----f----+-~
south
Hosjoen
north
Rni!' sand to
grov.1
Vordalsoro Fine sand and
silt
L.... vik Fine to COO fse
i)-am"'tn Medium sand
30 40 SO
MEASURED POROSITY, %
(bl Modulus numbor in sand from 6 tost sitos
(bQcktigured from settlements measurementsl
Fig.20 Modulus numbers from six test sites, based on screw plate and oedometer tests (Aas, 1981):
(a) comparison of modulus numbers from oedometer and screw plate; (b) modulus number in sand
from 6 test sites
DESIGN PARAMETERS FOR OFFSHORE SANDS 287
90
porosity. The porosity was calculated from 60 Boldi . t. 01. !lUll
NorrtQUy (onsolidated t icino sand
the water content, the unit weight and the Hcdiull' dtnse. 0, : 46-/.
Ocnu 0, ~ 70-/.
specific gravity of solids measured on undis- Very dense. Ot :: 90-1.
Young's Modulus, E
For cases which are not one-dimensional, Fig. 21 Secant Young's modulus values for
Young's modulus, E, is more appropriate uncemented n.c. quartz sands (after Robertson
than the constrained modulus, M. Young's and Campanella, 1983)
modulus depends on several factors, includ-
ing stress level, stress and strain history,
relative density and sand type. caution, since the increase depends on the
Robertson and Campanella (1983) degree of overconsolidation and density
reviewed available calibration-chamber test (Baldi et al. , 1982).
results in which cone resistance had been When using the relationship given in Fig.
compared to drained secant modulus from 21 it is important to be aware of the limita-
parallel triaxial tests prepared in the same tions of this approach, which are the same as
manner as the large calibration samples and for the M/q c relationship presented earlier.
to the same densities. Robertson and
Campanella (1983) argued that E 25 (secant Pressuremeter Test
modulus at 25% failure stress) is most
appropriate since the overall safety factor Figure 16 compared the expansion curves of
against bearing capacity failure is usually push-in and self-boring pressuremeter tests.
around 4 for foundations on sand. For com- The unload-reload loops shown are used to
puting bearing capacity of piles, E 50 may be compute shear modulus. As shown by this
more adequate. Calibration-chamber test particular test, the modulus values are very
results give values of E zJq c varying between similar. In practice, however, this drained
1. 5 and 3.0 which are in good agreement with modulus is little used in geotechnical calcu-
the recommended value of 2 given by lations. Campanella et al. (1985) describe a
Schmertmann (1978) for the computation of new cone pressuremeter which may be used
settlements of shallow foundations on sand. to determine modulus and soil strength. The
Schmertmann (1978) has since amended his equipment is still being developed. It is also
earlier suggestion by recommending two intended for offshore use.
values, 2.5 and 3.0, to allow for the variation
of shape factors for square and strip foot-
ings. Robertson and Campanella (1983) INITIAL SHEAR MODULUS
recommend the relationship given in Fig. 21
for the computation E 25 or E 50. Cone Penetration Test
Results from calibration-chamber tests
suggest higher values of EzJqc for overcon- Shear modulus at any strain level can as a
solidated sands (Le. 6 ~ a ~ 18). However, first approximation be estimated from the
the application of these larger factors to initial tangent modulus (G max) using the
overconsolidated sands should be used with reduction curves given by Seed and Idriss
288 ADVANCES IN UNDERWATER TECHNOLOGY
1000
Girt ~ 800kPo OYHAMI( SHEAR MODULUS, Ci.. ••• 1"'I.P:t
or· ___'~·__~2~0_ _~l~
O _ _~'~
O _ _~S.~~.~0_ _~'~
O ~I.
2S00
t.OOkPa
0
.::
"-
:I:
---
./
./
'"
2000
./
",/ -- IInOI ' Tonouchi 1191!1 21
Ci ..... I USN· ···1
vi
=> /.// 200kP. Iq,I N • 4.SI
--' ,SOO
=>
c
0
:I:
ex
~ 1000
:l:
VI
u
...z
'i:
>- '0
Cl
•
100 200 lOO .0 0 SOO
..~
!
'2
[ONE RESISTANCE, q'" HP.
are available. 20 •
qobfr t iOO and Cor.pa.nll!'llQ' s chort
Cron · l'Klle
L1. Durgunoglu, H. T. and Mitchell, J. K. 1975. developments in field and laboratory testing
Static penetration resistance of soils, I-II. of soils. Paper to 11th International Confer-
In Proceedings of Conference on In Situ ence on Soil Mechanics and Foundation
Measurement of Soil Properties, Raleigh, Engineering, San Francisco, August 1985.
North Carolina, 1975. American Society of 22. Janbu, N. C. 1970. Grunnlag i Geoteknikk.
Civil Engineers, Vol. 1, pp. 151-189. Tapir, 426 p.
L2. Eidsmoen, T., Gillespie, D., Lunne, T. and 23. Janbu, N. and Senneset, K. 1973. Field
Campanella, R. G. 1985. Tests with UBC compressometer - principles and applica-
seismic cone at three Norwegian research tions. In Proceedings of 8th International
sites. NGI and UBC joint report, 59040-1, Conference on Soil Mechanics and Founda-
20 December 1984. tion Engineering, Moscow, Vol. 1.1, pp.
13. Fahey, M. and Ranolph, M. F. 1984. Effect 191-198.
of disturbance on parameters derived from 24. Janbu, N. and Senneset, K. 1975. Effective
self-boring pressuremeter tests in sand. stress interpretation of in situ static pen-
Geotechnique 34(1), 81-97. etration tests. In Proceedings of European
14. Fay, J. B. and Le Tirant, P. 1982. Offshore Symposium on Penetration Testing, Stock-
self-boring pressuremeter for deep water. holm, 1974, Vol. 2.2, pp. 181-193.
Symposium on the Pressuremeter and its 25. Kroezen, M. 1981. Measurement of in situ
Marine Application, Paris. Editions Tech- density in sandy/silty soils. Can. Geotech.
nip. Paris, pp. 305-324. Soc. Newslett. 18(4), 13-15.
15. Campanella, R. G., Gillespie, D. and 26. Lacasse, S. and Lunne, T. 1982. Penetration
Robertson, P. K. 1982. Pore pressures dur- tests in two Norwegian clays. In Proceed-
ing cone penetration testing. In Proceedings ings of 2nd European Symposium on Penet-
of the 2nd European Symposium on Penet- ration Testing, Amsterdam, 1982. Balkema,
ration Testing, Amsterdam, Vol. 2, pp. Rotterdam, Vol. 2, pp. 661-669.
507-512. 27. Last, N. 1985. Proceedings from Seminar
16. Campanella, R. G. and Robertson, P. K. on Cone Penetration Testing in the Laborat-
1983. Flat plate dilatometer testing: ory, Southampton University, November
Research and development. Soil Mech. 1984.
Series No. 68, Department of Civil 28. Lunne, T. and Christoffersen, H. P. 1983.
Engineering, University of British Colum- Interpretation of cone penetrometer data for
bia, January 1983. offshore sands. In Proceedings of 15th Off-
17. Campanella, R. G. and Robertson, P. K. shore Tech. Conf., Houston, Texas, 1983,
1984. A seismic cone penetrometer to meas- Vol. 1, pp. 181-192.
ure engineering properties of soil. Paper 29. Marchetti, S. 1980. In situ tests by flat
presented at the Society of Exploration dilatometer. ASCE, JGED, V. 106, No.
Geophysicists' 54th Annual Meeting, GT3, pp. 299-321.
Atlanta. 30. Marchetti, S. and Crapps, D. K. 1981. Flat
18. Campanella, R. G. 1985. Personal communi- Dilatometer Manual. Gainesville, Florida.
cation. 31. Meyerhof, G. G. 1961. The ultimate bearing
19. Campanella, R. G., Robertson, P. K., Gil- capacity of wedge-shaped foundations. In
lespie, D. G. and Greig, J. 1985. Recent Proceedings of 5th International Conference
developments in in situ testing of soils. on Soil Mechanics and Foundation
Paper to 11 th International Conference on Engineering, Paris, Vol. 2, pp. 105-109.
Soil Mechanics and Foundation Engineer- 32. Mitchell, J. K. 1976. Fundamentals of Soil
ing, San Francisco, August 1985. Behaviour. John Wiley, New York.
20. Chapman, G. A. 1979. The interpretation of 33. Mitchell, J. K. and Lunne, T. A. 1978. Cone
friction cone penetrometer tests in sand. resistance as measure of sand strength.
PhD thesis, Department of Civil Engineer- Proc. Am. Soc. Civ. Engrs 104 (GT7) ,
ing, Monash University, Australia. 995-1012.
~:1. Hughes, J. M. 0., Wroth, C. P. and Windle, 34. Mitchell, J. K. 1985. Personal communication.
D. 1977. Pressuremeter tests in sands. 35. Parkin, A., Holden, J., Aamot, K., Last, N.
Geotechnique 27(4), 455-477. and Lunne, T. 1980. Laboratory investiga-
1. Jamiolkowski, M., Ladd, C. C., Germaine, tions of CPT's in sand. Report 52108-9,
J. T. and Lancelotta, R. 1985. New Norwegian Geotechnical Institute.
292 ADVANCES IN UNDERWATER TECHNOLOGY
36. Parkin, A. K. and Lunne, T. 1982. Boundary cone penetration test: performance and
effects in the laboratory calibration of a cone design. Report TS-78-209, Department of
penetrometer for sand. In Proceedings of Transportation, Federal Highway Adminis-
2nd European Symposium on Penetration tration Offices of Research and Develop-
Testing, Amsterdam, 1982. Balkema, Rot- ment, Washington, DC.
terdam, Vol. 2, pp. 761-768. 42. Schmertmann, J. S. 1982. A method for
37. Reid, W. M., St John, H. D., Fyffe, S. and determining the friction angle in sands from
Rigden, W. J. 1982. The push-in pres- the Marchetti dilatometer test. In Proceed-
suremeter. 37th Symposium on the Pres- ings of 2nd European Symposium on Penet-
suremeter and its Marine Applications. ration Testing, Amsterdam, Vol. 2, pp.
Editions Technip, Paris, pp. 247-261. 853-861.
38. Robertson, P. K. and Campanella, R. G. 43. Senneset, K. and Janbu, N. 1984. Shear
1983. Interpretation of cone penetration strength parameters obtained from static
tests. Part I: Sand. Can. Geotech. J. 20(4), cone penetration tests. ASTM Symposium,
718-733. San Diego, 1984.
39. Robertson, P. K. and Campanella, R. G. 44. Tjelta, T. I., Tieges, A. W. W., Smits, F.
1984. Liquefaction potential of sands using P., Geise, J. M. and Lunne, T. 1985. In situ
the DMT. Technical note, submitted to density measurements by nuclear backscat-
ASCE, JGED, April 1984. ter for an offshore soil investigation. Off-
40. Ruiter, J. de 1982. The static cone penetra- shore Technology Conference Proceedings,
tion test state-of-the-art report. In Proceed- May 1985, Paper 4917. (In press.)
ings of 2nd European Symposium on Penet- 45. Vesic, A. S. 1973. Analysis of ultimate loads
ration Testing, Amsterdam, Vol. 2, pp. of shallow foundations. Proc. Am. Soc. Civ.
389-405. Engrs 99 (SM1), 45-73.
41. Schmertmann, J. H. 1978. Guidelines for
18
The Applications of Centrifugal
Modelling to the Design of
Jack-up Rig Foundations
w. H. Craig, University of Manchester, UK,
and M. D. Higham, McClelland Ltd, UK
In recent years the use of physical modelling Since their introduction in the early 1950s,
aboard large centrifuges has been consider- the number of jack-up mobile drilling plat-
ably extended in many areas of geotechnical forms available for petroleum exploration
research. In offshore and coastal engineer- and production purposes has increased to its
ing, where loadings on structures are often present level of more than 400. Most of
complex and non-static, such models have these jack-up rigs are supported by three or
been widely used and are believed to have four independently jacked legs with indi-
given considerable qualitative insights into vidual tank-type footings (spud-cans),
the modes of behaviour of a number of dif- although earlier models sometimes had
ferent structural forms on both uniform and more legs. The alternative jack-up concept
stratified soil deposits, as well as yielding is that of the mat-supported rig. Two excel-
quantitative data for performance predic- lent review papers, one by McClelland et al.
tions which add to the confidence of desig- (1983) and the other by Young et al. (1984),
ners, operators, insurers and regulatory cover most aspects of the operation of all
agencies. Perhaps surprisingly, the mobile such rigs. While there are wide variations in
jack-up structure has only recently been the design specifications, individual spud-can
subject of model studies of this type. Such foundations are commonly of the order of
studies as have been carried out are very 100-150 m 2 carrying loads up to 30 MN and
r-reliminary and the present paper aims to working at bearing stresses of 200-300 kPa.
add to these and consider possible areas of Such footings may be operating at depths up
seful future extension. to 30 or 40 m below the seabed, and in such
293
294 ADVANCES IN UNDERWATER TECHNOLOGY
instances huge volumes of soil are subjected (3) Footing instability caused by scour
to gross shearing at the time of installation. undermining footings.
In contrast, the mat-supported rigs may (4) Mass movement of the sea floor.
have bearing areas an order of magnitUde (5) Jack-down and extraction of footings,
larger and working stresses an order lower, where pull-out loads exceed available
being particularly suitable for work on very forces.
soft cohesive seabeds where working depths
are typically 2-3 m. Category 5 is generally associated with
For both types of rig the normal practice deeply embedded footings, while category 3
is to pre-load the foundations to the max- is more likely at shallow depths in cohesion-
imum design load on first arrival on site. less soils.
The design load includes the fairly reliable For mat-supported rigs bearing failure at
known all-up weight of the rig itself, the pre-load stage is less likely to be catas-
together with an assessment of environmen- trophic, but this form of structure is, by vir-
tal loads which may be known to a consider- tue of its low bearing pressures and shallow
ably lower degree of precision. Thus, unlike penetration, particularly vulnerable to fail-
permanent structures which may never be ure under environmental loading with slid-
subjected to the design loads at any time in ing displacements being difficult to predict.
their working life, each mobile structure is
proof loaded to the design load at each loca-
tion. A high proportion of jack-up rig acci-
dents occur during this proof testing stage PUNCH-THROUGH FAILURES
of spud-can foundations, when punch-
through failures occur beneath individual Most cases of punch-through occur in
footings as they penetrate through a stiff stratified soil profiles when there is a rela-
stratum into a softer one. While most tively thin layer of sand or stiff clay overly-
authors in the literature indicate that ing a weaker layer. When installing a rig on
analytical methods are adequate to assess this type of soil profile instability may occur
the likely performance of a single footing after the footing has been temporarily sup-
under static vertical loading, regardless of ported by the upper layer. A further
the soil statification, provided the soil pro- increase in footing pressure which exceeds
file and properties are known, the practice the combined bearing capacity of the two-
of installing rigs on sites without adequate layered system will result in the footing
pre-investigation continues and this may punching through the upper soil layer and
largely account for the high incidence of fail- penetrating into the underlying layer until
ure at the pre-load stage. In short, site sufficient bearing resistance is encountered
investigation is as crucial for a mobile struc- at some lower level, or the applied load is
ture as for a fixed structure. distributed elsewhere. The resulting dam-
Young et al. (1984) list footing-supported age to the rig can range from minor struc-
rig failures which have resulted in major tural damage to the leg and its jacking
economic losses in recent years. In addition mechanism to the complete loss of the rig.
to these there are no doubt a greater If it is required to install a rig at a site
number of failures which were not included with a potential for punch-through type fail-
because the consequences were less severe ure it is important that a reliable estimate of
- the table may in reality only offer up the the ultimate bearing capacity of the
tip of the iceberg. Five problem areas are stratified soil profile be made. Two basic
identified: methods have been proposed for predicting
the capacity of foundations supported by
(1) Punch-through during pre-load such a soil profile, these being the Meyerhof
(2) Excessive penetration during storm and Hanna method and the Projected Area
loading. method.
CENTRIFUGAL MODELLING FOR JACK-UP RIG FOUNDATION DESIGN 295
o
H
ailurf !tanf
where q b = ultimate bearing capacity of
lower layer as before
Clay qt = ultimate bearing capacity of
upper layer as before
(alThin Sand Lay..- (b)ThlCk Sand LaYfr
P = slope of failure plane in upper
layer
ig. 1 Failure of soil below footing on dense
and overlying soft clay (after Meyerhof, 1974) All other symbols as before.
296 ADVANCES IN UNDERWATER TECHNOLOGY
Experimental Programme
to
A total of eleven tests were completed in a
preliminary study. Except for the fIrst test,
to
in which a uniform sand profIle was used, all
the tests were carried out using a soil profIle
with a bed of compacted air-dry Mersey Fig. 3 Model footing configurations (75 mm
River sand of between 175 and 230 mm over- diameter) (All dimensions in millimetres)
lying a block of remoulded saturated Cow-
den clay which had been consolidated into a
container with internal dimensions 560 mm preparation and the measured stress levels;
square x 460 mm deep. A total of three beds and these were the values used in analysing
of clay were used in the test programme the test data for punching shear failure. The
with values of undrained shear strength of tests were carried out using four model foot-
36, 60 and 240-260 kPa. The fIrst two were ings manufactured from mild steel, two hav-
specially prepared for the study and, being ing a diameter of 50 mm and the other two a
the full size of the container, were large diameter of 75 mm. Two different vertical
enough to allow for four tests in each. A profIles were machined for each footing
smaller bed of stiff clay was made available diameter (Fig. 3) in order to study the
and only two tests were performed in this. effects of the profIle on the measured bear-
For each clay bed an indication of its ing capacity, with the two 50 mm diameter
undrained shear strength was obtained profIles being 2/3 scale models of the 75 mm
prior to the preparation of the sand bed by footings.
use of a small penetrometer. A more accu- For each test the soil bed was rotated in
rate assessment was later obtained by a the Simon Engineering Laboratories cen-
series of static unconsolidated undrained trifuge at 90 rpm, corresponding to a radial
triaxial tests on samples taken after comple- acceleration of 25.2 g at the radius of the
tion of the test programme. The friction clay/sand interface. With the centrifuge
angle of the sand layer was determined by rotating at the required speed, the model
fItting a straight line to the initial part of the footing was driven into the soil model by the
bearing capacity profIles for each test and use of a servo-controlled hydraulic jack (Fig.
then backfIguring a value from the bearing 4) to a depth of 300 to 350 mm measured
~apacity factors given by Vesic (1975). The relative to the point of maximum area of the
calculated values for the angle of friction of footing. During the test the load applied to
37-38 were within the range expected for
0
the footing was measured by a load cell on
he type of sand used, the method of bed the hydraulic jack and the movement of the
298 ADVANCES IN UNDERWATER TECHNOLOGY
Natural gravity ( 9 )
Centrifuge
acceleration Gas filled
-up to 120.g 0. 0.1 02m
L....L....J
Hydroulic actuator by
R.D. P. Howden Ltd.
(±So.kN,500mm stroke)
transducer
Position of lateral
loading actuator
Fig. 4
footing was measured by use of a L VDT the point of the maximum bearing capacity.
with a stroke length of 500 mm. The load The value of the shape factor, S., was calcu-
and displacement measurements were lated using the conservative values for the
recorded automatically by use of an on-line punching shear coefficient, K., given by
computer system. A general description of Meyerhof and Hanna (1978). Figure 5 shows
the centrifuge and a discussion of the vari- a typical bearing capacity versus depth rela-
ous design considerations are given by Craig tionship for one test using the clay bed of
and Rowe (1981). Full details of the particu- intermediate strength and a theoretical
lar experiments are given by Higham 'best fit' curve based on Meyerhof and
(1984). Hanna analysis. Figure 6 shows similar fit-
ting for the same test using Jacobsen
Analysis of Results analysis and also includes data from a paral-
lel test with a deeply profiled footing. The
Table 1 summarizes details of each test per- values of K .S. and S. calculated from a series
formed. For all the tests in which the footing of such constructions are plotted against the
underwent punching shear failure , the test bearing capacity ratio qJql in Figs 7 and 8
data were analysed to determine the value respectively, with the values of a plotted
of the bearing capacity factor, K 8S" and a with against the bearing capacity ratios qt/qband
regard to Meyerhof and Hanna's and Jacob- qJq2 in Figs 9 and 10 respectively.
sen's methods respectively, such that the The data from the three tests performed
theoretical bearing capacity profile gave the by Herdy and Townsend (1983), in which
best possible fit to the plotted test data at punch-through failures were observed, were
CENTRIFUGAL MODELLING FOR JACK-UP RIG FOUNDATION DESIGN 299
TABLE 1
Summary of centrifuge tests performed
Maximum bearing
stress (kPa)
Footing Footing Sand depth Clay strength
Test no. diameter (mm) shape (mm) (kPa) (Continuous increase)
0 75 flat 460
5 50 flat 230 36 3750
10 50 profiled 230 36 3600
15 75 flat 230 36 2900
20 75 profiled 230 36 2350
25 50 flat 2lO 60 3700
30 50 profiled 210 60 3200
35 75 flat 210 60 2200
40 75 profiled 210 72 2400
45 50 flat 175 240 4400
50 75 flat 175 260 3600*
analysed and plotted in the same manner, (Figs 7 and 8), the calculated values of the
and these calculated values are incorporated . bearing capacity factor K.S. showed no dis-
into the following discussion. cernible trend with regard to the bearing
In the analysis of the test data with capacity ratio qJqJo whereas the calculated
regard to Meyerhof and Hanna's method values of the shape factor, S., decreased for
HI
4'0 4'0
- - -'" =D'495
a·o 8-0
1 0 · 0 1 L - - - - - - - - - - - ---' 10·0L--- - - - - - - -- - - - - l
Tj'ig. 5 Bearing capacity vs. depth profile for Fig. 6 Bearing capacity VS. depth profile for
calculated values of punching shear factors calculated values of projected area coefficients
300 ADVANCES IN UNDERWATER TECHNOLOGY
15.-------------------------------------------~
b!gend
Hogham 119841
• Flat footings
• Profiled footings
Herdyand Townsend 119831
vi" 10
,.r' + Flat footings ks for _;37-5·
>- +
!
+
5
g'
i
ID
Fig. 7 Variation of bearing capacity factor KsSs with the bearing capacity ratio q2/ql
Fig. 8 Variation of bearing capacity factor Ss with the bearing capacity ratio qdql
I~r-------------------------------------------__,
1·2
+
+
~OL---------~5-----------I~O----------~5~---------=W
Fig. 9 Variation of bearing capacity factor a with bearing capacity ratio qT/qn
CENTRIFUGAL MODELLING FOR JACK-UP RIG FOUNDATION DESIGN 301
·6...-------------------,
O~~---~S~---~IO~---~IS~---~20
BearirYil CapaCIty Ratto,Q./qt
Fig. 10 Variation of bearing capacity factor ex with bearing capacity ratio qt!q2
an increase in the value of the bearing capac- where failure is by a combined punching and
ity ratio q2/q), with the value calculated for squeezing mechanism, the ultimate value of
tests carried out using essentially flat foot- bearing capacity is less than the value given
ings being less than unity for values of bear- by considering the failure of the soil to be
ing capacity ratio greater than 0.5. For val- solely due to punching shear.
ues of bearing capacity ratio ofless than 0.5,
the use of a shape factor value of unity, as Theoretical Considerations
suggested by Meyerhof and Hanna (1978),
should prove to be conservative. For the Meyerhof and Hanna method, the
In the analysis of test data with regard to ultimate bearing capacity of a foundation
Jacobsen's method, Figs 9 and 10 show that under vertical load is given by the equation
an equation of the form ex = 0.35 + 0.083qJq2
would give a better prediction of bearing qu = qb + 2'yH(H + 2D) KsSs tan ¢lB
capacity for the majority of tests than that + y'V/A
proposed by Jacobsen et al. (1977). Consid-
By use of the substitution D = D' - H) where
ering the form of this equation, it would
D' is the depth to the soil interface, the
seem necessary to limit the value of ex for
above equation can be rearranged to give a
high values of the bearing capacity ratio
quadratic of the form
q)/q2. From the range of values calculated in
analysing the test data a limiting value of qu -a mH2 + bmH + Cm
=
ex = 1.5 should be conservative.
where
For the final two tests in the test prog-
ramme the calculated values of the bearing am = 2y'KsSs tan ¢/B
capacity ratio qJq) were greater than unity.
b m = 4D'y'Ks S s tan ¢/B
For such cases it has been suggested
(Meyerhof, 1974) that the failure of the soil and
beneath the footing is a combination of
punching shear failure and lateral squeez- Cm = qb + y'V/A
ings of the overlying stiff layer. Because the Similarly, for the projected area method,
failure mode cannot be considered to be the equation for the value of ultimate bear-
:;olely punching failure, for these two tests, ing capacity of a foundation under vertical
·;he calculated values of KsSs and ex cannot load
strictly be compared with those values cal-
culated for the preceding tests. However, qu = qb(1.0 + 2{JH/B)2 + y'V/A
hese two tests indicate that, in those cases can be rearranged to give a quadratic of the
302 ADVANCES IN UNDERWATER TECHNOLOGY
/
'!...../
...,.
I
I
I
I
/
I method can only be used to predict
I
/
the peak values of bearing capacity.
I
/
When taking a shape factor value of
..
c /
z;" , /
/
unity, the values of bearing capacity
ic ;-
/
culated in this study, an upper limit in its capacity and the on-board loading
for the value of IX of 1.5 is suggested. equipment was similarly limited. Thus pro-
(b) For a footing penetrating a cohesionless totype sizes and loads were low relative to
soil layer, the particular vertical profile the offshore situation, but stresses were
used resulted in a reduction of the relatively high.
measured bearing capacity of only some James and Tanaka (1984) used a bigger
15% relative to a 'flat' footing - this is machine capable of modelling footings much
in contrast to the results of James and larger than those actually simulated. They
Tanaka (1984). could possibly cope with a 100 m 2 field pro-
(c) For a footing penetrating a cohesive soil totype at 1:100 scale in the future but are
layer, the shape of the vertical profile of currently limited by the capacity of loading
the footing has no appreciable effect on equipment. The stresses they have used are
the measured bearing capacity. again higher than those typical offshore.
In our study the centrifuge was capable
of modelling the full field dimensions using
ASSESSMENT OF CENTRIFUGE 100 mm diameter footings at accelerations up
TESTS TO DATE AND FUTURE to 110 g. The loading equipment available had
POTENTIAL a very high capacity (80 kN) and was exces-
sively powerful for the job in hand. In view
Table 2 summarizes the three sets of cen- of the unstable nature of the punch-through
trifuge experiments referred to above. It is mechanism, which causes the massive
clear that to date none of the experimental servo-hydraulic loading actuator to plunge
programme has been carried out with com- into the model under its own increased
plete similarity to the typical jack-up loads self-weight in the centrifuge, it seemed
and dimensions considered earlier. The sensible to limit the accelerations used and
reasons for this are largely practical and to utilize relatively high strength soils. This
stem from three independent studies using led once more to model stress levels being
existing or readily assembled equipment - greater than those in the field. Beds of sof-
each set with its own inherent limitations. ter soils are, in a number of ways, harder to
Each set of experiments has been acknow- manufacture and handle in the laboratory
ledged as preliminary in nature. but, with the experience now gained in the
In the case of Herdy and Townsend way in which all the equipment behaved,
(1983), the centrifuge available was limited there should be no difficulty in simulating
TABLE 2
Comparison of model studies with typical field data from Young et al. (1984)
Model footing
diameter (mm) 30 50-100 50-75
Acceleration
factor (g) 20-50 1 and 10-60 25
Maximum bearing
stress (kPa) 450-1150 1100 2200-4400
Prototype
diameter (m) 0.6-1.5 0.1-3.0 1.3-1.9 12
Bearing
stress (kPa) 450-1150 1100 2200-4400 300
304 ADVANCES IN UNDERWATER TECHNOLOGY
realistically both field dimensions and stres- ible with the use of prototype particle sizes
ses. Moreover, with the knowledge that all in model studies which may be desirable
these studies - despite these deficiencies - from a number of other points of view.
have confirmed recognized analytical proce- The performance of spud-can foundations
dures for the static vertical bearing case, in the face of massive sea-floor movement
there is every confidence that useful exten- may not be a realistic modelling ambition
sions to such modelling can be carried out except in very limited site-specific contexts.
into areas less adequately understood at Another site-specific possibility is the
present. assessment of the likely performance of
In looking at the potential areas of future foundations placed adjacent to previous
activity it is pertinent to consider again the spud-can locations or adjacent to other
recognized field problems. For spud-can structures, e.g. piles or gravity platform
foundations the static punch-through can be bases where complex interactions can be
adequately modelled and a range of studies expected.
can be contemplated, assessing, for exam- The prediction of jack-down and spud-can
ple, the effects of relative layer strengths extraction can be readily assessed for the
and layer thicknesses and of various footing idealized conditions of a model. To a limited
profiles. Addition of lateral loading capabil- extent the models can also incorporate some
ity, as already initiated by James and of the techniques used by rig operators to
Tanaka, causes little difficulty for surface or free jammed footings, e.g. jetting.
near-surface footings, or for footings instal- The mat-supported jack-up does not
led in model beds at depth before centrifug- appear to have yet been subjected to any
ing. Where gross changes of geometry are specific modelling. Of particular interest is
contemplated, there are mechanical prob- the performance of structures on soft cohe-
lems with lateral loading for models involv- sive soils where strength increases linearly
ing deep penetration or punch-through. In with depth and where possible thin, near-
cohesive soils the need for performing the surface sand or silt horizons could lead to
large penetration at high g may not be pres- localized liquefaction and loss of sliding
sing, but when structural performance resistance under cyclic loading.
depends upon granular soils the evidence Limited studies have been started on
from tests on other deep foundations indi- small surface footings on beds of homogene-
cates that 'in flight' installation is paramount ous clay with strength increasing with
(Craig, 1984). depth. Leung et al. (1984) and Nakase et al.
The effects of cyclic vertical and horizon- (1984) used strip footings 80 and 50 mm wide
tal loads are readily incorporated into the respectively under centrifuge accelerations
model tests, and herein lies a particular of 40-80 g and 50 g with applied static, ver-
attraction since reliable field data on the tical bearing stresses rising to 100 and
effects of cyclic loading are virtually absent 25 kPa. Similar conclusions can be drawn
from the published record. Although proof- from these studies as from the jack-up tests
testing by preload may be sufficient for reported above: modelled dimensions and
quasi-static vertical loading, provided ade- stresses may not be in good similarity with
quate capacity is available, preload cannot the field cases, but the potential for
proof-test foundations against lateral load- development is clearly demonstrated.
ing or the degrading effects of shear-stress A substantial amount of work has previ-
reversals. At shallow depths in cohesionless ously been reported at the other end of the
soils such loading may lead to loss of support foundation spectrum - that of the very
due to pumping action underwater, but the large surface and near-surface founded
removal of material by current-activated gravity structures, e.g. Rowe and Craig
scour is not realistic - the modelling of sed- (1980). Models have been subjected to com-
iment transport is difficult and not compat- binations of static and cyclic, vertical and
CENTRIFUGAL MODELLING FOR JACK-UP RIG FOUNDATION DESIGN 305
I am grateful to the Organising Committee that were covered on the first day. First, we
for the invitation to attempt to sum up the had Dr Riemersma talking about positioning
proceedings. Summing up is not really the requirements, and it seemed to me to be an
appropriate phrase - it is a difficult job to unhappy reflection on human frailty that he
do justice in a summary to the amount of was concentrating so much on the errors in
material that has been presented over the the system and on the human factors that
two days of the conference. Clearly, each led to trouble, emphasizing that the techni-
paper merits further individual attention in ques are vastly superior to the ability of the
order to reflect on its content. What I am human beings who used them. Then, Dr
going to say must necessarily be an unbal- Palmer talked about a fascinating case his-
anced critique, because we are considering a tory of the Ocean Thermal Power Project;
whole range of knowledge and experience in this was of particular interest because most
a wide diversity of topics, and my comments of the other stories we heard were not so
are bound to be biased by my own interests. specific and not about such a novel project.
My first general comment is that I found Dr Schiittenhelm and Mr Stoker, in their
the second day of the conference livelier and presentations, reported significant additions
of more interest than the first - I think to the general store of knowledge about the
there are two reasons for this. First of all, geological properties of the North Sea sec-
the second day's theme was concentrated on tor around the north of the UK and the
geotechnical aspects, so it was one part of Abyssal Plains. This must be very useful
the tripartite structure of geophysics, geol- background information for site investiga-
ogy and geotechnics. Secondly, the papers tions in these areas, and I think it is laudable
were focused much more on specific details to attempt to add geotechnical information
and specific jobs. to the geological data that have been
Let me remind the audience of the topics obtained. However, I wonder whether this
307
308 ADVANCES IN UNDERWATER TECHNOLOGY
is a little too dangerous and whether the soil is not the homogeneous, continuous
maps should be issued with a warning to the material that is assumed in so many
effect that they are an aid, and not a substi- theories. Messrs Hight and Jardine focused
tute, for proper site investigation. our attention on the problem of attempting
We then had three papers that can be high quality laboratory testing, and indi-
grouped together, presented separately by rectly underlined the value of in situ test-
Loevik, Hutchins and Games, which all ing. At the same time, I respond to Mr
dealt with important improvements in vari- Lunne's plea that in situ testing is not the
ous aspects of seabed surveys. Mr Green panacea of all site investigation, and that its
then proselytised about his integrated interpretation must be backed up with high-
approach to site investigation, which quality laboratory testing which must
launched us into a debate about the philoso- include appropriate reconsolidation of
phy of education and some lively discus- specimens. Mr Lunne gave us a most useful
sions. Finally we had Mr Sarginson, essen- review of in situ testing in sand and recent
tially making a plea that the contractors developments in that area. These develop-
providing services should be treated in a ments have been backed up by Professor
more honourable manner. Jamiolkowski, whose very careful work with
The second day's programme started with his Italian colleagues, testing these various
Dr Richards giving a very good review of instruments in controlled circumstances in
in situ testing and sampling in deep water, laboratory test chambers, provides invalu-
which he defined as' deeper than 300 m. I able verification for their acceptance in
found this a fascinating glimpse of develop- practice.
ments that are taking place, and the limits of Finally, Mr Craig told us about a very real
time undoubtedly prevented him from doing problem, that of jack-up rig foundations,
justice to a lot of valuable information. This which he has studied in centrifugal model
was followed by Christophersen telling us tests at Manchester. I warmed to his plea
about the site investigation of the Troll that industry should be much more sophisti-
Field and the very detailed information cated in its approach to this important prob-
obtained about this one site. lem and should take account of current work
Subsequently, our attention was drawn to and present understanding.
various in situ testing devices. Two papers If I could attempt to draw some more
were about pressuremeter tests - M. Fay general conclusions about the conference, I
talking about the French version of the should like to repeat what we heard from Mr
self-boring pressuremeter developed for off- Zuidberg about the very steady and marked
shore use, and Mr Powell with a valuable development that has taken place since the
comparison of data obtained from three dif- last conference in 1980. He commented on
ferent pressuremeters at one site in this the increasing reliability of all the equip-
country. These were followed by Professor ment, the maturity of the industry and the
Silva discussing the development and growing confidence that people now have in
deployment of an in situ vane, and I must the results that are obtained from any of the
congratulate him on some very commend- surveys or site investigations. I am sure
able real engineering - to have got that that this will continue as exploration moves
instrument to work successfully first time in steadily into deeper waters. Economists'
such a depth of water was a great achieve- jargon refers to whether or not develop-
ment. ments are due to 'demand pull' or 'technol-
We then diverted to some comments ogy push'. It seems to me that in the off-
about laboratory testing, and Mr Marsland shore industry at the moment it is a case of
provided timely reminders of the impor- 'demand pull': it is the move into deeper
tance of soil fabric as well as the fact that waters which is causing the relevant tech-
SUMMARY 309
corrected end bearing (qc - (1'vo) that should the Organising Committee, chaired by
be correlated with other soil properties. Dennis Ardus, supported by Sid Green,
I shall close by making one or two more Eugene Toolan and Tim Freeman. I imagine
general comments. About 120 people have that it was their idea to hold the conference,
registered for OSSI '85, a good number for a their initiative and drive that made it poss-
conference. We have a good mixture of dis- ible. Of course, the conference would not
ciplines within civil engineering, geotechn- have happened without the organisational
ics, geology and geophysics, and a cross- ability of the Society for Underwater Tech-
section of representatives from industry, nology, of its Secretary, David Wardle, and
academia and government research estab- of the Conference Organiser, Jean Pritch-
lishments. It is a disappointment that there ard, and her team. We also must thank the
is no one here from the Headquarters of the session chairmen, who guided us ably
Departments of Energy and Environment, through the discussions; the authors, who
who have so much influence on and involve- prepared the papers and presented them so
ment in the offshore site investigation well; and those who entered the discussion
industry. Dr Richards made the point about from the floor. Finally, I should like to hear-
the different attitudes to development from tily congratulate all those who have contri-
industry on the one hand, and from research buted. to a most successful conference.
establishments and universities on the
other. It is valuable to exchange ideas,
experience and attitudes in a conference REFERENCE
such as this.
My final duty - and a most pleasant one Wroth, C. T. 1984. The interpretation of in situ
- is to express thanks on behalf of all the soil tests. 24th Rankine Lecture Geotech-
delegates to a number of people. First, to nique 34 (4), 449-489.