Marine and Petroleum Geology 04 "0887# 438462

The use of dipmeter data to constrain the structural geology of the

Gullfaks Field\ northern North Sea
J[ Hesthammera\\ H[ Fossenb
a

Statoil\ GF:PETEK!GEO\ N!4919 Ber`en\ Norway

Department of Geolo`y\ University of Ber`en\ Alle`t[ 30\ N!4996 Ber`en\ Norway
Received 11 September 0886 ^ revised 19 April 0887 ^ accepted 14 April 0887

Abstract
Analyses of 12 km of dipmeter data from 37 wells on the Gullfaks Field\ northern North Sea\ have helped to constrain and
understand the structural geology of the area[ The analyses have veri_ed the general structure obtained by seismic interpretation^
i[e[ a main\ western domino system of rotated fault blocks and non!planar bedding and an eastern horst complex\ separated by a
collapsed anticline structure[ In addition\ subseismic structures have been revealed by the integrated use of dipmeter data\ well log
correlation\ seismic data and core data[
Within the domino system\ dip of bedding decreases towards the main faults to the west "large!scale drag#[ The e}ect of such
large!scale drag decreases with depth\ probably due to a vertical increase in consolidation and mechanical strength at the time of
deformation[ Within the accommodation zone\ a modi_ed fold structure is seen\ whereas strata within the horst complex are generally
subhorizontal[
More than half of all faults on the Gullfaks Field have developed a zone of local drag[ Generally\ only the northerly!trending
faults are associated with drag[ The interval a}ected by drag is typically some tens of meters wide\ and is consistently wider in the
hanging wall than the footwall[ Due to the e}ect of drag\ total o}set may be up to one order of magnitude larger than the amount
of missing section identi_ed from well log correlation[ There is no apparent relationship between interval a}ected by drag and
lithology[ However\ drag is less abundant in more consolidated rocks at deeper stratigraphic levels[
Minor faults are less abundant than expected from a power!law down!scaling of seismically detectable faults\ and are not
particularly common within the large!scale drag zones[ It is suggested that a signi_cant amount of subseismic deformation occurs
on a scale below resolution of dipmeter data\ probably by {ductile| ~ow rather than discrete faulting[ North!trending minor faults
are most common and will restrict ~uid ~ow in an EW direction due to cataclasis and phyllosilicate smear associated with abundant
deformation bands "micro!faults#[ 0887 Elsevier Science Ltd[ All rights reserved[
Keywords] Structural geology ^ Dipmeter data ^ Gullfaks Field

0[ Introduction
As many new oil and gas _elds and remaining well
targets in producing _elds are economically marginal\ the
need for a sound understanding of the structural geology
in such areas increases[ Integrated use of di}erent types
of data is necessary to achieve this goal[ In _eld devel!
opment\ this involves both seismic\ production and vari!
ous types of well data[ In structurally complex areas\
seismic data are typically of poor quality\ and production
data may be di.cult to interpret[ In such cases\ it is
important to utilise available well information such as
dipmeter data\ which is the theme of the present article[
 Corresponding author[ Tel[ ] 36 44881029 ^ fax ] 36 44881986 ^
e!mail ] jonhestatoil[no

Dipmeter data re~ect dip and azimuth of bedding

"Schlumberger\ 0875 ^ Serra\ 0878 ^ Schlumberger\ 0889 ^
0880#[ Variations of these parameters along the bore
holes help improve our understanding of the struc!
tural geology in faulted reservoirs "Bengtson\ 0870 ^
Schlumberger\ 0875 ^ Etchecopar and Bonnetain\ 0878 ^
Bigelow\ 0882 ^ Goetz\ 0883#[
The present study focuses on how the analyses of 12 km
of dipmeter data from the structurally complex Gullfaks
Field in the northern North Sea have helped to outline
the structural characteristics of the _eld[ We believe that
the large amount of data available makes the Gullfaks
Field particularly well suited for this type of study\ and
may serve as a guide for similar analyses on other _elds[
In the following\ description and discussion of the results
are provided[

S91537061:87:,*see front matter 0887 Elsevier Science Ltd[ All rights reserved
PII ] S 9 1 5 3 7 0 6 1 " 8 7 # 9 9 9 1 7 1

449

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

1[ Structural setting
The Gullfaks Field is situated in block 23:09 in the
western ~ank of the Viking Graben in the northern North
Sea\ south of the Snorre Field and southeast of the
Statfjord Field "Fig[ 0#[ The _eld covers an area of ca 44
km1 and is developed from three platforms under the
Norwegian license group consisting of Statoil "operator#\
Norsk Hydro and Saga Petroleum[ Total recoverable
reserves amount to ca[ 209 mill[ Sm2 of oil and some 29
bill[ Sm2 of gas\ and are produced from the Middle Jur!
assic Brent Group and the Lower Jurassic Cook and
Statfjord Formations "Fig[ 1#[
The Gullfaks Field occupies the eastern half of a major
"0914 km wide# NNESSW trending fault block "the
Gullfaks fault block ^ Fossen et al[ in press# which is
bounded by faults with km!scale displacement to the
east and west[ The major fault to the west separates the
Gullfaks block from the giant fault block containing the
Statfjord and Brent oil _elds\ whereas the bounding fault
to the east borders the deeper parts of the Viking Graben[
At least two major rift phases have a}ected the area
"e[g[ Ziegler\ 0871 ^ Beach et al[\ 0876 ^ Giltner\ 0876 ^
Badley et al[\ 0877 ^ Thorne and Watts\ 0878 ^ Gabrielsen
et al[\ 0889 ^ Roberts et al[\ 0889 ^ F%rseth et al[\ 0884#[
The _rst rift phase is de_ned as a Permo!Triassic phase
that a}ected the total width of the northern North Sea
"Roberts et al[\ 0884 ^ F%rseth\ 0885#[ The second rift
event is termed the late Jurassic phase "e[g[ Badley et al[\
0877#\ and is more localised to the central portions of the
northern North Sea "Viking and Sogn Grabens#[ Whereas
the Permo!Triassic phase is at least as signi_cant as the
late Jurassic phase\ late Jurassic deformation is more
obvious on commercial seismic lines[ Furthermore\
because the majority of wells in the North Sea are con!
_ned to the Jurassic and younger section\ the late Jurassic
phase is best known from well data[ The present study is
mainly concerned with data from the Jurassic layers in
the Gullfaks Field\ and thus with the late Jurassic rift
phase[

2[ Geology of the Gullfaks Field

2[0[ Stratigraphy
The deepest well on the Gullfaks Field "23:09!02# was
drilled to ca 2249 m depth and penetrated 0239 m of
Triassic sands and shales of the Hegre Group[ The Scy!
tian!Rhaetian Hegre Group "Fig[ 1# comprises inter!
bedded sandstones\ claystones and shales deposited in
a continental environment[ Overlying this sequence are
alluvial sandstones of the Rhaetian!Sinemurian Statfjord
Formation[ The Sinemurian!Toarcian Dunlin Group
comprises marine clay! and siltstones of the Amundsen
and Burton Formations\ regressive\ marine\ silty clay!

stones\ muddy sandstones\ and sands of the Cook For!

mation\ and marine shales and siltstones of the Drake
Formation[ The mainly Bajocian!Early Bathonian Brent
Group forms the upper and main part of the reservoir[
The group is subdivided into the Broom\ Rannoch\ Etive\
Ness\ and Tarbert Formations deposited in a deltaic
environment[ A broad lithological subdivision can be
made between the shaly Ness Formation and the sandy
intervals below and above[
A time gap of up to 099 My is represented by the
base Cretaceous "late Cimmerian# unconformity on the
Gullfaks Field[ This discontinuity separates Triassic and
Jurassic sediments from Upper Cretaceous sediments and
post!dates the major part of the faulting history of the
Gullfaks Field[ Up to 099 m of shales of the Upper
Jurassic Heather Formation are locally preserved in the
hanging walls to the main NS trending faults\ par!
ticularly in the western part of the _eld[

2[1[ Structural outline

A general description and discussion of the structural
geology of the Gullfaks Field is given in Fossen and
Hesthammer "0887#\ and only the main points will be
repeated here[ Based on seismic interpretation and sup!
ported by dipmeter data "as described in this work#\ the
Gullfaks Field can be divided into three contrasting com!
partments "Fig[ 0# ^ a western domino system with dom!
ino!style fault block geometry\ a deeply eroded eastern
horst complex of elevated subhorizontal layers and steep
faults\ and a transitional accommodation zone or col!
lapsed fold structure[
The domino system\ which constitutes the main and
western part of the Gullfaks Field\ is characterised by N
S trending faults "hereafter referred to as main faults#
with displacements in the order of 49499 m[ The faults
dip 1429> to the east\ whereas the strata within the fault
blocks dip shallowly "typically 04># to the west[ These
main faults typically show increasing complexity "bifur!
cation# towards higher reservoir levels[
The domino fault blocks are compartmentalised by
several minor faults with throws less than 49 m[ These
faults have much more variable trends than the main
faults[ NS striking\ east!dipping minor faults generally
have slightly steeper dips than the main faults when situ!
ated in the hanging wall\ and slightly lower dips if they
are located in a footwall position "footwall collapse struc!
tures#[ Minor faults striking in an EW direction typically
have steeper dips "3489># and are believed related to
internal block deformation during di}erential slip along
the main faults[ NWSE and NESW striking minor
faults have variable throws and intermediate dips[ Minor
faults oriented antithetically to the main faults are less
common and have variable\ but mostly NS strike[ Core
analyses typically indicate a relatively narrow {damage

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

440

Fig[ 0[ "a# Structure map of the Statfjord Formation on the Gullfaks Field[ Well locations indicated by a star are projected down from shallower
reservoir levels[ "b# EW pro_le through the _eld showing the three structural domains[ "c# The triangular drag zone and high dip bu}er zone within
the domino system[ Inset map shows the location of the Gullfaks Field[

)
441

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

zone| "usually a few meters or less# associated with both

minor and main faults[
The average dip of bedding in the domino area varies
from 02> for the Brent Group to 05[5> for the deeper
Statfjord Formation "calculated from depth!converted
seismic interpretation#[ In addition\ there is a general
decrease in dip from the east "footwall position# to the
west "hanging wall position# within the domino fault
blocks[ This change in dip gives rise to a gentle hanging!
wall syncline\ and is hereafter referred to as large!scale
drag[ The e}ect of this large!scale drag is greatest at
shallower reservoir levels[ Main faults in the horst com!
plex are generally much steeper and more planar than
those in the domino area\ with dips of 5969> being
common[ Both east! and west!dipping faults occur with
a constant NS strike[
The horst complex is characterised by poorer seismic
data quality\ which complicates mapping of minor faults[
Seismically mappable minor faults are relatively steep
"3469># and run subparallel to the main faults[ Bedding
within the horst complex is sub!horizontal or shallowly
west!dipping[ The accommodation zone is a graben struc!
ture which accommodates the di}erent structural styles
between the domino area and the horst complex[ The
zone is typically de_ned by a collapsed anticline with a
west!dipping "ca 04># western limb and a subhorizontal
to gently east!dipping eastern limb[ The fold axis plunges
very gently to the NNW\ and the opening angle is
approximately 059> with a steeply dipping axial plane[
The accommodation zone is bounded by steep "54>#
faults to the east and lower!angle "1429># faults to the
west[ Minor faults strike mostly NS or EW and have
variable dip[ Physical modelling "Fossen and Gabrielsen\
0885# and well data indicate that the accommodation
zone is a complex area with many minor faults of which
we have been able to identify only a small fraction[
3[ Data
Several 2D re~ection seismic surveys have been col!
lected in the Gullfaks area[ The current seismic interpret!
ation is based on a 0874 acquisition which consists of
some 09[999 km of seismic lines covering an area of ca
08 km in EW direction and ca 03 km in NS direction[
The oil!bearing part of the structure covers an area of ca
44 km1[ More than 069 wells have been drilled on the
Gullfaks Field since block 23:09 was awarded to the
license group in 0867\ yielding over 001 km of well section
within the reservoir "Table 0#[
Dipmeter data from a total of 12 km well section col!
lected from 37 wells on the Gullfaks Field have been
analysed with respect to structural dip of bedding[ The
instruments used on the Gullfaks Field for providing the
information are Schlumberger|s HDT\ SHDT\ FMS\
and OBDT "for detailed description of the tools\ see
Schlumberger\ 0875 ^ Serra\ 0878 ^ Schlumberger\ 0880#[

Fig[ 1[ Stratigraphic column for the Jurassic and Triassic reservoir units
in the Gullfaks Field "modi_ed from Tollefsen et al[\ 0883#[

CMYK Page 441

442

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Table 0
Amount of well data and dipmeter data available for interpretation[ Simple trigonometric expressions were used to calculate the amount of available
data projected into EW and NS direction[ The nearly equal amount of data in these two directions suggests that there are no preferred orientations
of drilling on the Gullfaks Field[ It is important to be aware that\ when dividing the number of faults observed from well log correlation and dipmeter
data by the projected amount of well data in any direction\ a too high number will result "simply because the projected amount of well data will
always be less than the total amount#[
Formation

) of
drilled
resevoir

Meters
penetrated

Meters
penetrated
EW

Meters
penetrated
NS

m analysed
dipmeter data

m analysed
dipmeter data
EW

m analysed
dipmeter data
NS

Heather
Tarbert
Ness
Etive
Rannoch
Broom
Drake
Cook
Burton
Amundsen
Statfjord
Hegre Group
All formations

4[3
02[0
03[2
2[3
01[2
0[2
09[5
04[4
9[4
7[7
6[0
6[6
099

5\011
03\670
05\001
2[758
02\806
0\244
00\843
06\371
353
8\770
6\756
7\608
001\412

1\644
6\613
7\682
1\097
5\481
472
4\699
6\349
071
3\136
2\251
3\207
42\703

3\039
7\590
7\027
0\657
6\145
474
3\885
6\610
86
2\976
1\117
1\871
40\488

104
1\015
2\198
790
1\584
063
1\052
2\524
23
1\440
1\242
1\697
11\553

51
609
0\138
163
0\976
66
663
815
01
454
506
487
5\840

82
860
0\172
177
884
60
631
814
03
447
251
390
5\692

In order to take full advantage of the information

provided by the dipmeter tool\ the resistivity curves must
be processed to yield information on either structural or
sedimentary bedding patterns[ The processing methods
used in this study are developed by Schlumberger "Sch!
lumberger\ 0875 ^ 0889# and include the CLUSTER\
MSD\ OBDIP and FMS IMAGE processing methods[
The quality of the dipmeter data varies with the hole
conditions\ temperature and depth\ type of mud used\
rock type and tools[ Generally\ FMS data are of better
quality than SHDT data due to the increased button
density and borehole coverage[ Similarly\ SHDT gives
better results than HDT due to improvements of the tool
and the algorithms used in processing[ Data from the
OBDT are generally of poorer quality than any of the
resistivity tools "related to the use of oil based mud#\ and
results must accordingly be treated with care[
4[ Large!scale structures
Forward modelling of the Gullfaks Field "Fossen and
Hesthammer\ 0887# suggests that the accommodation
zone underwent considerably more internal deformation
than both the horst complex and the domino system[ The
geometries within the horst complex require no internal
"sub!seismic# deformation\ whereas the domino system
can be explained by a combination of ca 29> tilting and
an average shear strain "g# of 9[3 with the shear plane
oriented synthetic to the main faults[ This model implies
an extension in the order of 79)\ i[e[ considerably more
than the ca 3949) indicated by a rigid block model
"Rouby et al[\ 0885#[

In general\ there is good agreement between dipmeter

data and seismic data\ and the characteristic orientation
of bedding in the domino system "westerly tilted layers#\
the horst complex "subhorizontal layering#\ and the
accommodation zone "collapsed anticline structure# are
re~ected both in seismic "Fig[ 2# and dipmeter data "Figs
34#[ Figure 3 displays stereonet plots from the three
wells shown in Fig[ 2[ Both sets of data show west!dipping
strata in the domino system as seen in well 23:09!2 which
penetrates one of the rotated fault blocks[ The anticline
structure in the accommodation zone is veri_ed by
dipmeter data from well 23:09!C!0[ The eastern fold limb
shows very shallow dips to the east\ whereas the western
limb shows somewhat steeper dips to the west "C!0 in
Fig[ 3#[ Dip of strata in the hinge zone is subhorizontal[
Dipmeter data from well 23:09!C!6\ which is located in
the horst complex\ indicate subhorizontal strata in this
area[
Figure 4 shows changes in dip and dip direction with
respect to depth from the three wells[ Dipmeter data from
well 23:09!2 "Fig[ 4a\ domino system# show that the strata
dip in a westerly direction with the amount of dip decreas!
ing with depth[ Although the same type of plot for well
23:09!C!0 "Fig[ 4b\ accommodation zone# shows much
scatter "due to use of oil based mud ^ Stuart!Bruges\ 0873 ^
Dumont et al[\ 0876 ^ Goetz\ 0877#\ it can be seen that
the dip of strata changes from an easterly direction above
1299 mMD "metres measured depth# to westerly dip
below 1399 mMD[ This example demonstrates that
although dipmeter measurements from tools used in oil
based mud are more scattered than those from water
based mud\ reliable structural dip and useful information

)
443

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 2[ Seismic pro_les through "a# well 23:09!2 within the domino system\ "b# well 23:09!C!0 from the accommodation zone\ and "c# well 23:09!C!6
located within the horst complex[ See Fig[ 0 for location[

CMYK Page 443

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

444

Fig[ 3[ Stereonet plots of dipmeter data "dip direction:dip# from the wells displayed in Fig[ 2[ See text for discussion[ Depth interval indicated in
lower right!hand corner "measured depth in well#[ Strata penetrated by well 23:09!C!0 ] eastern limb ] Cook and Amundsen Formations ^ hin`e zone ]
lower part of the Amundsen Formation and upper part of the Statfjord Formation ^ western limb ] lower part of the Statfjord Formation and upper
part of the Lunde Formation[

can be achieved "see also Dumont et al[\ 0876#[ The zone

from 12991399 mMD represents the hinge zone seen in
Fig[ 3[ Strata in well 23:09!C!6 display shallow dips at
all depths "Fig[ 4c\ horst complex#\ and as a result\ the
dip direction vs[ depth plot shows much scatter[
4[0[ Domino system
Figure 4"a# showed that amount of dip in well 23:09!2
decreases with depth[ This change is consistent with the
gradual decrease in dip towards the hanging wall side of
the domino blocks which can be seen from the seismic
data "Figs 2"a# and 5#[ The change in dip is most pro!
nounced at shallower depths[ A pro_le through the ver!

tical exploration well 23:09!7 "Fig[ 6#\ illustrates that the

change in dip is most pronounced within a triangular
area in the hanging wall part of the fault blocks\ and that
the decrease in dip occurs both to the west and with
depth[ Dipmeter data from the well verify this change in
dip "Fig[ 7#[ The dip vs depth plot "Fig[ 8# shows that\
just below the base Cretaceous unconformity\ the dip is
relatively high "around 08>#\ but decreases to sub!
horizontal at 1099 mMD[ Strata in this interval dip to
the west[ Below 1099 mMD\ dip increases with depth
but dip direction changes to southeast[ This pattern is
characteristic of a fold structure with a subhorizontal
fold axis "i[e[ a non!plunging or shallowly plunging fold#[
The dip of strata increases to approximately 14> at 1079

445

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 4[ Dip vs depth and azimuth "dip direction# vs depth plots from
"a# well 23:09!2\ "b# well 23:09!C!0\ and "c# well 23:09!C!6[ By plotting
the changes in dip and azimuth in two di}erent diagrams\ changes in
azimuth are more clearly displayed than by using the standard tadpole
plots\ thus allowing for analyses of the curvature of bedding "Bengtson\
0870#[ See text for discussion and Fig[ 0 for location[

)
J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

446

Fig[ 5[ Seismic line through well 23:09!7[ Curved bedding traces appear in the hanging wall to the main fault and may in part be described as large!
scale drag con_ned to a triangular zone[ See Fig[ 0 for location[

Fig[ 6[ Pro_le through well 23:09!7 based on seismic data and dipmeter data[ Strata are dragged towards the main fault in a hanging wall position[
Note how displacement at shallower levels are much less than at deeper levels[ This large!scale drag "km!scale# de_nes a triangular zone and must be
distinguished from local drag which a}ects the strata only some tens of meters away from the fault "see inset _gure ^ the thin white line illustrates the
changes in dip caused by local drag#[

mMD where a fault with 03 m missing section intersects

the well[ Below this fault\ data are not reliable enough
for any con_dent interpretation[ Since the well
approaches the hanging wall side of the rotated domino
fault block with depth\ the decrease in dip with depth
therefore re~ects both lateral and vertical changes[ The
lateral decrease in dip towards the hanging wall part of
the domino blocks is also seen from Fig[ 09\ where the

CMYK Page 446

average dip from the Rannoch\ Amundsen and Statfjord

Formations is plotted against distance to the next main
fault to the east[
The large!scale drag within the triangle zone in a hang!
ing wall position to the main faults is most easily explai!
ned by internal shear synthetic to\ and steeper than the
main faults "Fig[ 00#[ In this model\ shear strain increases
towards the hanging wall side of the rotated fault blocks[

447

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 8[ The dip vs depth and dip direction vs depth plots from well
23:09!7 show an example of ~attening drag "Bengtson\ 0870# towards
a main fault[ Bedding is dragged from westerly dip at shallower reservoir
levels to easterly dip at the base[ The well does not penetrate the main
fault\ although several minor faults are encountered[ See text for further
discussion and Fig[ 0 for location[

The model implies that the hanging walls to the rotated

fault blocks are more deformed than the footwalls[ Other
examples where hanging walls are more deformed
than footwalls are common both o}shore "e[g[ well
29:2!6S near the Veslefrikk oil _eld and well 23:09!21
in the Gullfaks So
r oil and gas _eld# and onshore "Fig[
01#[
Since the amount of shear strain in the hanging walls
to the main faults decreases with depth\ the o}set across
the faults increases at deeper stratigraphic levels "see Fig[
6#[ In some cases\ fault displacement at the top of the
Brent Group may be only 09 m\ whereas the displacement
at the top of the Statfjord Formation may be as much as

300000000000000000000000000000
0
Fig[ 7[ Stereonet plots of dip direction:dip from the di}erent formations
penetrated by well 23:09!7[ Dip of bedding in the uppermost formations
is to the west but decreases rapidly with depth[ Within the Cook Forma!
tion\ dip of bedding is to the east[ This change in dip describes large!
scale drag in a triangular zone on the hanging wall side to the main
faults within the rotated fault blocks of the domino system[ See Fig[ 0
for location[

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

448

Fig[ 09[ Dip estimated from dipmeter data and from depth!converted seismic interpretation plotted against the distance to the nearest main fault to
the east for the Rannoch Formation "a# and for the Statford:Amundsen Formation level "b# "domino area#[ Each data point represents a single well
around which the dip is estimated from the seismic interpretation\ and where the highest concentration of dip is estimated from dipmeter data[ Both
data sets show a clear decrease in dip to the west\ consistent with the geometry shown in Fig[ 0 "bottom#\ Fig[ 2 and Fig[ 6[ A binomial test "with
con_dence interval of 84)# of the data indicates that dipmeter estimates are slightly higher than seismically determined dips in "a#\ while no systematic
signi_cant di}erence is detected for data shown in "b#[

149 m[ The reason that the deeper stratigraphic levels are

less a}ected by large!scale drag is likely that these rocks
were more consolidated at the time of deformation[ In

fact\ since deformation of the Gullfaks area started

immediately after deposition of the Brent Group\ the
sands of the Tarbert Formation were little or not con!

459

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 00[ The large!scale drag within the triangular zone may be easily explained by applying increasing shear strain towards the hanging wall side to
the main fault[ The shear angle is synthetic and somewhat steeper than the main fault[ Thus\ the amount of shear will decrease with depth as well as
to the east[

Fig[ 01[ Field example from the Bartlett fault near Moab\ Utah[ A large normal fault separates the massive Entrada sandstone "footwall# from
interlayered sandstones and shales of the Cedar Mountain Formation "hanging wall#[ Strata in the hanging wall have developed a several hundred
meter wide drag zone[ The displacement along the fault is ca 299 m[

solidated at all and would therefore be less resistant

against soft and ductile deformation
Wells located outside the triangular zone tend to show
a downward increase in dip of bedding[ Well 23:09!A!7
is located mainly in a footwall position to one of the main
faults[ The dip vs depth and dip direction vs depth plots
"Fig[ 02# show a change from shallow dip "ca 04># to the
southwest in the Tarbert Formation to steeper dip "ca
12># to the west in the stratigraphically deeper Drake
Formation[ This small increase in dip with depth is

thought to be related to di}erential compaction[ After

rotation of the layering within the domino fault blocks\
the Jurassic and Triassic strata were overlain by several
hundred meters of Cretaceous and Tertiary sediments[
Loosely consolidated rocks of the Tarbert Formation
was then more compacted and dips of layers thereby
decreased more than those of the Statfjord Formation
"Fossen and Hesthammer\ 0887#[
In some cases\ the western boundary of the triangular
zone is not de_ned by one of the main faults\ but rather

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

450

zone# does not exist\ it is di.cult to separate large!scale

from minor faults based on dipmeter data alone[
Figure 04"a# shows dip vs depth and dip direction vs
depth plots for well 23:09!C!5 which is located within the
horst complex[ Due to the high amount of scatter\ it is
hard to distinguish any break at 1719 mMD\ which is
the location of a fault with 71 m missing section[ The
cumulative dip plot "Fig[ 04"b##\ however\ clearly shows
a change in the gradient of the curve at about 1719 mMD[
Above this depth\ the curve has a relatively shallow gradi!
ent\ suggesting a fairly rapid increase in the cumulative
dip[ Below the fault\ the gradient is steeper\ indicating
shallower dip of bedding[ This is consistent with obser!
vations from seismic data "Fig[ 04"c## where the hanging
wall dips to the east\ and the footwall is subhorizontal[
Fig[ 02[ The dip vs depth plot for well 23:09!A!7\ which is located
outside the triangular zone of large!scale drag\ shows how amount of
dip increases slightly with depth[ See Fig[ 0 for location[

by a minor fault "Fig[ 03"b## or simply by an abrupt

change in dip "Fig[ 03"a##\ in which case the o}set across
the structure is entirely accommodated by continuous or
ductile deformation[ In these cases\ the main fault is
located farther to the west[ The dip within several such
{high dip bu}er zones| have been veri_ed by dipmeter
data "from wells 23:09!A!04\ 23:09!A!14A\ 23:09!A!16A\
23:09!B!7#\ showing that strata between the minor fault
and the main fault to the west are not a}ected by large!
scale drag\ and are tilted in the direction of opening of
the zone\ typically to the north[ Because strata are steep
within the high dip bu}er zone\ the amount of shear
strain is likely low "assuming that the rocks underwent a
component of rigid fault block rotation so that a decrease
in the dip of bedding represents an increase in shear
strain ^ see Fossen and Hesthammer\ 0887#[ This implies
that the rocks on the hanging wall side of the main fault is
no more deformed than the footwall at this stratigraphic
level[ Shear strain will however increase drastically across
the minor fault or across the change in dip to the east[
The {high dip bu}er zones| may possibly represent hard!
linked relay!ramp structures "e[g[ Trudgill and Cart!
wright\ 0883#[
4[1[ Faults
Larger faults are usually identi_ed both on seismic
data and from stratigraphic well log correlation[ From
dipmeter data\ the main faults within the domino system
can usually be identi_ed by the change in dip across
the faults\ as described above[ The drastic change from
shallow dips in the hanging wall to much steeper dips
in the footwall is readily distinguished "Ro
nningsland\
0889#[ In areas where such a change in dip "triangular

5[ Small!scale structures
Dipmeter data are especially useful for identi_cation
of faults "Werner et al[\ 0876 ^ Koepsell et al[\ 0878a ^
0878b ^ Devilliers and Werner\ 0889 ^ Zhang\ 0882#\ par!
ticularly those below seismic resolution\ since it is com!
monly di.cult to obtain information of the fault|s
geometry by other means "Adams et al[\ 0881#[ In the
Gullfaks Field\ faults are commonly associated with a
gradual or abrupt change in amount of dip and:or dip
direction of bedding[ If no such changes exist\ it is nor!
mally not possible to detect faults from dipmeter data
alone[ Similarly\ dipmeter data alone can not separate
between a fault and an unconformity in those cases where
abrupt changes occur[ The cumulative dip and dip direc!
tion plots "Fig[ 04"b# ^ Hurley\ 0883# are especially suit!
able for recognising faults where dip of strata changes
abruptly across the fault plane\ whereas the statistical
curvature analysis technique "Bengtson\ 0870 ^ see also
Kaya and Norman\ 0882# provides the best mean for
recognising and describing drag features[ Stereonet plots
are generally not suitable for fault detection\ since they
do not yield information on changes in dip and azimuth
with depth[
Figure 05 shows an example of local drag "a}ecting
some tens of meters on each side of the fault plane# related
to a fault with 8 m missing section within the Amundsen
Formation in well 23:09!C!2[ Whereas well log cor!
relation has placed the fault at 1610 mMD\ dipmeter data
suggest that the zone of highest shear strain is located at
1602 mMD[ It is thus likely that the location of the fault
is slightly misplaced by stratigraphic well log correlation[
The interval a}ected by drag in this case covers an area
of approximately 099 m and is not easily observed on
seismic data[ Analyses of dip direction of drag!folded
bedding give an indication of the fault orientation "the
bedding within the drag zone is usually rotated towards
parallelism with the fault#[ Since there is no apparent
change in dip direction of the bedding in Fig[ 05\ it is

)
451

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 03[ Several places within the domino system\ the large!scale drag terminates against a minor fault rather than against the main fault which is
located farther to the west[ Between the minor fault and the main fault is a zone of high dip\ termed the {high dip bu}er zone|[ Figures "a# and "b#
show two seismic sections through two such zones[ "c# Schematic illustration of the high dip bu}er zone[

likely that the fault dips to the west[ The amount of dip
increases to 39> or more at the location of the fault[ This
increase in dip towards the location of the fault is termed
{steepening drag| "Bengtson\ 0870#\ and occur when both
bedding and the fault dip in the same direction[ Since
bedding is bent towards parallelism with the fault plane\
this gives a minimum estimate of the amount of dip of
the fault plane[ It is thus possible to achieve an under!
standing of the geometry of bedding around a fault as
well as the orientation of the fault plane itself[
It is important to try to quantify the e}ect of drag as
it may a}ect the prediction of stratigraphic juxtaposition
and fault!sealing potential[ A geological pro_le through
well 23:09!C!2 was constructed based on the dipmeter
information from the well "Fig[ 06#[ Dip isogons are

CMYK Page 451

drawn parallel to the fault\ and the orientation of the

fault is estimated from the dip and azimuth vs depth
plots[ Although the missing section is only 8 m\ the result
shows that the total o}set of layering outside the area
a}ected by drag is more than 099 m[ The example clearly
illustrates the importance of such geometries[
A similar example is seen in well 23:09!A!4H "Fig[ 07#[
A fault with 06 m missing section has been identi_ed from
well log correlation[ The area a}ected by drag covers an
interval of approximately 49 m along the wellbore hole[
Strata above the fault dip shallowly to the west[ At
approximately 0829 mMD\ the apparent dip decreases to
zero before it increases again\ this time to the east[ Below
0844 mMD\ the trend becomes a mirror image of that
above[ This pattern of ~attening drag "Bengtson\ 0870#

)
J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

452

Fig[ 04[ "a# The dip vs depth and azimuth vs depth plots for well 23:09!C!5 show much scatter[ It is di.cult from these data to observe the fault with
71 m missing section identi_ed from well log correlation[ "b# The cumulative dip plot "Hurley\ 0883# shows a clear change in gradient at the location
of the fault[ This type of plot can thus help in the identi_cation of faults where data quality is poor[ "c# A seismic pro_le through well 23:09!C!5[ The
observations from dipmeter data "a!b# are consistent with observations from seismic data[ See Fig[ 0 for location[

is caused by dragging of west!dipping strata towards

parallelism with an east!dipping fault[ Before strata can
dip to the east\ it must _rst be subhorizontal[
A second fault with 20 m missing section has been
identi_ed in well 23:09!A!4H at 0868 mMD[ This fault
has no related drag structures\ and can only be identi_ed
from dipmeter data by a drastic change in amount of dip
"especially apparent on the longitudinal dip component
plot ^ Fig[ 07#[ Without the help of well log correlation
and detailed knowledge of the sedimentology in the res!

CMYK Page 452

ervoir\ such a break in amount of dip could be mistaken

for an unconformity or a sequence boundary[
When the location of a minor fault is known\ stereonet
plots can be made for the interval above and below the
fault[ In an example from well 23:09!A!00 "Fig[ 08#\ dip
of strata above a fault "hanging wall# with 09 m missing
section is somewhat steeper than dip of bedding below
the fault "footwall#[ This di}erence is related to drag in
the hanging wall\ whereas the footwall is hardly a}ected
by drag[ The dip of strata in the footwall is therefore

453

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 05[ The dip vs depth and dip direction vs depth plots for well 23:09!
C!2 clearly identify an area of steepening drag "Bengtson\ 0870# related
to a fault with 8 m missing section[ The dip direction is to the west at
all depths[ See Fig[ 0 for location[

Fig[ 06[ An exercise was carried out on dipmeter data from well 23:09!
C!2 in order to quantify the amount of missing section relative to total
o}set[ The results show that\ where drag is present\ the total o}set may
be more than one order of magnitude larger "099 m# than the missing
section recorded by well log correlation "8 m#[

more consistent with the general dip of bedding in the

area[
In order to check for more systematic di}erences in
deformation between the hanging wall and footwall of
minor faults\ the average intervals a}ected by local drag
on both sides of 35 faults where examined[ The result
"Fig[ 19# shows that the interval a}ected by drag is more

Fig[ 07[ Transverse and longitudinal dip component plots for parts of
well 23:09!A!4H[ The plots show apparent dip of the formation for a
chosen pro_le direction "094> for the transverse dip component plot
and 904> for the longitudinal dip component plot#[ This display method
was _rst documented by Bengtson "0870# and allows for analyses of the
curvature of structures such as folds and faults[ A fault with 06 m
missing section\ located at 0844 mMD\ is associated with a zone of
~attening drag which is easily identi_ed on the transverse dip com!
ponent plot[ The fault located at 0868 mMD is clearly not associated
with such a zone of drag[ Instead\ the fault is identi_ed from dipmeter
data by a change in amount of dip[ This is seen best on the longitudinal
dip component plot where the bedding above the fault dips in a north!
erly direction whereas bedding below the fault dips in a southerly
direction[ See Fig[ 0 for location[

than twice as wide in the hanging wall "average of 59 m

along the wellbore# than in the footwall "average of 18
m# for all formations[
Figure 19 does not indicate any signi_cant di}erences
in the interval a}ected by drag between the di}erent
formations[ These formations range from thick sand!
stone units "Tarbert\ Etive\ Rannoch\ Cook and Statfjord
Formations#\ via interlayered sandstones and shales
"Ness Formation#\ to thick shale units "Drake and
Amundsen Formations#[ There is thus no clear cor!
relation between lithology and width of the drag zone[
This is likely due to the fact that the sandstones were
poorly consolidated at the time of deformation\ and
therefore had shear strengths close to that of shale[
When deformation started\ the bedding was more or
less horizontal[ The vertical axis in Fig[ 19 therefore also
re~ects increasing depth at the onset of deformation "the
Statfjord Formation was located deeper than the Brent
Group prior to the deformation#[ With the possible
exception of the Statfjord Formation "where only three
faults with associated drag were identi_ed#\ there is no
signi_cant correlation between the average interval
a}ected by drag and the depth at the time of deformation[
A comparison of formation vs the normalised number
of faults with related drag is shown in Fig[ 10"a#[ From

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

454

Fig[ 08[ Stereonet plots of dip direction:dip of bedding from the hanging wall and footwall of a minor west!dipping fault penetrated by well 23:09!
A!00[ Dip of bedding in the hanging wall is clearly steeper than dip of bedding in the footwall[ This is consistent with steepening drag related to a
west!dipping fault[ See Fig[ 0 for location[

Fig[ 19[ Average interval "measured along the wellbore hole# a}ected
by drag for various formations clearly demonstrates that the defor!
mation in the hanging walls to the faults a}ects a much wider interval
than in the footwalls[ No clear relationship is seen between lithology
and width of the drag zone[ Since the width of the drag zone is measured
along the trace of the wells\ which is not necessarily perpendicular to
the fault\ the intervals represent maximum estimates[

this plot\ it appears to be fewer faults at deeper strati!

graphic levels that have developed associated drag[ This
_nding suggests that the lithologies at deeper strati!
graphic levels were somewhat more consolidated than
shallower strata at the time of deformation\ and is con!
sistent with the observation that large!scale drag "tri!
angular zone# is more widespread in the upper parts of
the reservoir[ The ratio between faults with identi_ed
drag from dipmeter data and faults without drag "ident!
i_ed by other means# is plotted in Fig[ 10"b#[ The _gure
indicates that faults within the shales of the Drake and
Amundsen Formations are less likely to exhibit drag[
Drag folding associated with faulting is most commonly
observed within the Ness\ Rannoch and Statfjord For!
mations[ A total of 41 faults with related drag were ident!
i_ed from dipmeter data[ In comparison\ 72 faults have
been identi_ed from other well log correlations within
the same intervals[ This gives a weighted ratio of 9[52\
and suggests that drag is associated with more than half
of all faults in the Gullfaks Field[
In most cases where the faults are associated with drag
folding\ it is possible from dipmeter data to indicate the
dip direction of the faults[ The results from analyses of
28 faults on the Gullfaks Field is illustrated in Fig[ 11"b#[
Almost the entire population "21# of faults with associ!
ated drag dips either in an easterly "most common# or
westerly direction[ Very few north! or south!dipping
faults exhibit drag zones[ This is partly due to the general
strike of bedding which is also to the north "it is easier to
change only the amount of dip rather than both dip and
azimuth#[ Also\ easterly trending faults commonly show
a strike!slip component "Rouby et al[\ 0885#\ which will
prevent the development of drag folds[ Although drag
folds with north!trending fold axes are most common\
dipmeter data are capable of recording drag features
which involves changes in both dip and azimuth "Fig[
12#[
The average EW density for ca NS trending faults
with related drag is 3[5 per km "21 faults within 5[8 km

455

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 10[ "a# Plot of faults with related drag per kilometre with respect to lithology[ Fewer faults appear to have developed associated drag at deeper
stratigraphic levels[ "b# A plot of the relationship between drag!related faults identi_ed from dipmeter data and faults identi_ed from well log
correlation "same interval# with respect to the di}erent formations does not support the common assumption that drag is more common in shales
than in sandstones[ Instead\ drag is less commonly observed in the shaly Drake and Amundsen Formations[

Fig[ 11[ "a# XY scatter plot showing interval of strata a}ected by drag
near faults vs missing section[ There is no clear relationship between
interval a}ected by drag and the size of the fault as identi_ed from
missing section[ "b# A plot of the dip direction of faults as identi_ed
from dipmeter data "based on associated drag of bedding# shows that
most faults that developed associated drag dip to the east "most com!
mon# or the west[

of analysed dipmeter data in EW direction#[ The density

of all faults in EW direction as identi_ed from well log
correlation is 3[3 per km "126 faults within 42[7 km of
drilled reservoir in EW direction#[ It is thus likely that\
due to the e}ect of drag and little missing section\ not all
faults are identi_ed from routine well log correlation[
A plot of missing section vs interval a}ected by drag
"Fig[ 11"a## reveals no signi_cant correlation[ From this

plot\ it is clear that large!scale faults do not necessarily

develop wider local drag zones than minor faults[ In fact\
with the exception of the domino system where the main
faults are associated with a triangular zone of large!scale
drag\ there are surprisingly few di}erences between large!
scale and small!scale faults in terms of change in dip and
azimuth of bedding and width of the local drag zone[ This
important observation suggests that local drag started to
develop prior to the initiation of the fault plane\ or as a
process zone ahead of the growing fault surface "Hobbs
et al[\ 0865 ^ Scholz et al[\ 0882#[ In this model "Fig[ 13#\
a fault plane develops when the amount of shear strain
"related to displacement across the drag zone and its
width# exceeds a critical value[ This is so because ductile
deformation "at some scale# is only feasible until the
critical strength of the rock is exceeded and a fracture
forms[ After this point\ deformation is more easily
accommodated by slip along the fracture "brittle faulting#
than by ductile deformation in the vicinity of the fracture[
Deviations from this idealised model may be explained
by changes in the geometry and kinematics of the fault
zone during the slip history[

6[ Integrated use of well data and seismic attribute maps

Recognition of faults and related structures at the limit
of\ or below seismic resolution\ is perhaps the largest
challenge for structural geologists involved in oil _elds
under production[ Down!scaling of assumed self!similar
fault attributes have been applied in some cases "e[g[

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

456

Fig[ 12[ Stereonet plots of bedding orientation "dip direction:dip# for the hanging wall and footwall of a minor fault penetrated by well 23:09!A!4H
"see Fig[ 0 for location#[ The _gure shows clear di}erences in dip of strata in the hanging wall "shallow dips to the northwest# and footwall "steeper
dips to the west#[ In addition\ there are clear evidence of drag in a northerly direction[ Drag is to the northeast\ suggesting a northeast!dipping fault[

Gauthier and Lake\ 0882#[ A statistical study of fault

populations in the Gullfaks Field "Fossen and Ro
rnes\
0885# indicates that this approach would not be appro!
priate in the Gullfaks area[ However\ integrated use of
the large number of dipmeter data\ other well data and
seismic data enables us to approach this problem without
the use of statistical fault scaling models "Hesthammer\
0887a^ 0887b#[
Seismic attribute analyses have helped to detect faults
with displacements\ in some cases\ down to 4 m\ although
conventional 2D seismic interpretation typically allows
only identi_cation of faults with displacements more than
0414 m[ Seismic timedip maps of the strong intra!Ness
Formation re~ection show abundant "curvi!#linear fea!
tures "Fig[ 14#[ These approximately NS trending fea!
tures are especially numerous in the eastern "footwall#
parts of the domino fault blocks\ and relief maps indicate
that most "ca 74)# are directly associated with down!to!
the!west o}sets of seismic re~ections "Hesthammer and
Fossen\ 0886a#[ If the linear features represent minor
faults\ it would indicate that the footwalls of the rotated
domino fault blocks are more deformed than the hanging
walls and that the deformation occurred along discrete
west!dipping faults and associated fault block rotation[
However\ there are several reasons why such an interpret!
ation may not be valid[ For example\ structural modelling
for the Gullfaks domino system "Fig[ 00# suggests that
most deformation is located in the hanging wall part of
the domino fault blocks[ In addition\ the average density
in EW direction of the west!dipping seismic features is
approximately 8 per km "estimated from timedip maps
of the intra!Ness and top Statfjord Formations ^ Hes!
thammer and Fossen\ 0886b#[ Most of the "curvi!#linear
features are related to apparent o}set of re~ections in the
range of 0929 m[ Generally\ all faults with o}sets in this
range can be identi_ed by well log correlation "Fossen
and Ro
rnes\ 0885#[ The signi_cantly lower fault density
calculated from well log data\ 1[0 or 3[3 per km if the

number of faults is divided by the amount of drilled

reservoir projected into an EW direction\ indicates that
many do not represent faults "note that the method of
dividing the total number of faults identi_ed from well
log correlation by an amount of well data projected into
any direction will result in a too high number of faults
per km#[ In addition\ recent studies of 041 faults with
displacement less than 29 m that have been penetrated
by wells on the Gullfaks Field demonstrate that less than
14) of these faults can be observed in the seismic data[
Structural analyses of 5 km of core data "work in pro!
gress# are consistent with the above observations and
suggest that {ductile| and widely distributed grain reor!
ganisation rather than discrete micro!faulting is respon!
sible for most of the deformation within the domino
system on the Gullfaks Field[
Detailed analyses of dipmeter data give further support
to this conclusion[ Analyses show that approximately
59) of all faults detected by stratigraphic well log cor!
relation have developed local drag\ and that drag is pref!
erentially related to NS trending faults "i[e[ considerably
more than 59) of NS trending faults have developed
local drag#[ Most of the NS trending faults are therefore
detectable from dipmeter data[ Twelve west!dipping "dip
direction is 169:11[4># faults have been identi_ed
from dipmeter data on the basis of drag\ i[e[ an average
density per km of 9[4\ or 0[7 if projected onto an EW
section "a total of 5[8 km of analysed dipmeter data exists
in an EW direction ^ see Table 0 for detailed expla!
nation#[ Assuming that 79) of the NS trending faults
developed drag\ about 49 faults should have been ident!
i_ed from dipmeter data if all the "curvi!#linear features
with associated down!to!the!west o}set "density  8 per
km# observed on the timedip maps were faults[
Another approach is to compare apparent dip at the
seismic and sub!seismic scales[ If subseismic faults exist\
the dipping layer "re~ection# will be imaged with an aver!
aged dip on a seismic section\ while the true dip of bed!

457

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 14[ Timedip map of an intra!Ness Formation re~ection[ Most of

the N!S trending "curvi!#linear features observed on the dip map result
from interference of dipping coherent noise with the intra!Ness For!
mation re~ection[ See text for discussion[

Fig[ 13[ Schematic illustration of how a layered sequence "a# may be

deformed by localised shearing "b# to form a zone of drag[ Eventually
the deformation results in discrete slip "a fault# along the line of
maximum shear deformation "c#[ Given that tensional stresses give rise
to horizontal extension across the zone\ the drag is expected to decrease
after the establishment of the fault due to rebound of the elastic com!
ponent of strain associated with the drag structure[

ding may be lower or larger\ depending on the geometry

of the sub!seismic faults "Fig[ 15#[ This prediction can be
demonstrated by comparing dip from dipmeter data and
seismic interpretation[ If the true dip is detectable from
dipmeter data\ then this dip must generally be di}erent
from that detected from seismic interpretation[ Fur!
thermore\ if the "curvi!#linear features re~ect sub!seismic
faulting\ the amount of dip from dipmeter data should
be lower than the seismically interpreted dip "Fig[ 15#[
Figure 09 showed a comparison of dip from dipmeter
data and seismic dip for the Rannoch and Statfjord For!
mations[ Statistical analysis shows that dip from dipmeter
data is slightly hi`her than seismic dip for the Rannoch
Formation "Fig[ 09"a##\ whereas no signi_cant di}erence

exists for the Amundsen and Statfjord Formations "Fig[

09"b##[ This _nding indicates that\ while there may be
some support for minor east!dipping faults\ minor west!
dipping faults are relatively uncommon\ and not com!
parable to the high distribution of "curvi!#linear features
on the attribute maps[ Hence\ we conclude that much of
the intra!block deformation is not re~ected by dipmeter
data\ and thus occurs on a very small scale[ Comp!
lementary core analysis indicates that much deformation
is by {ductile| grain!reorganisation processes\ particularly
in the Brent Group which was very poorly consolidated
during deformation[
We suggest that the "curvi!#linear features observed on
the seismic timedip maps are related to the interference of
dipping coherent noise with real re~ections "Hesthammer
and Fossen\ 0886a ^ 0886b ^ Hesthammer and Lo
kkebo
\
0886^ Hesthammer\ 0887b#[ The e}ect of this interference
is clearly demonstrated in Fig[ 16\ which shows a seismic
pro_le through exploration well 23:09!4[ Until recently\
the 23:09!4 area was believed to be a}ected by footwall
collapse with the presence of two east!dipping minor
faults "Fig[ 16"a##[ Bedding within these two collapsed
fault blocks was\ based entirely on seismic data\ inter!
preted to be to the west[ However\ dipmeter data clearly

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 15[ Cartoon showing that if the features observed on a seismic

timedip map in the footwalls to some of the rotated fault blocks in the
domino system represented faults\ dip from dipmeter data should be
di}erent from seismic dip[ Thus\ by comparing dip from dipmeter data
with seismic dip\ it is possible to obtain information on shear angle
within the fault blocks provided that the strain was accommodated by
discrete faulting rather than by a more widely distributed homogeneous
grain reorganisation[

show that bedding within the two fault blocks is sub!

horizontal "Fig[ 16"b##[ The area was thus reinterpreted
based on the dipmeter data[ In the section above the base
Cretaceous unconformity where layers are known to be
subhorizontal "Fig[ 16"c##\ it is clear that the interference
of west!dipping coherent noise "the seismic data set con!
tains abundant west! and east!dipping coherent noise#
with a subhorizontal re~ection causes apparent rotation
of the re~ection\ and the appearance of west!dipping
strata[ This example demonstrates how important it is to
use all available data to distinguish between real seismic
re~ections and seismic noise\ and how dipmeter data
combined with seismic data can be used to approach this
common type of problem in reservoir characterisation[
7[ Implications for oil production
The observation that the hanging wall is generally more
deformed than the footwall has implications for well
planning in the Gullfaks Field and in structures that
evolved under similar circumstances[ While oil producers
may be placed very close to the main faults in a footwall
position\ the wider zone of shear deformation in the hang!
ing wall may make it necessary to place injectors farther
away from the main fault[ This is particularly important

458

where ~uid ~ow might be restricted by deformation struc!

tures "faults with several meters displacement and associ!
ated deformation bands# in an EW direction\ e[g[ where
minor faults within the triangular zone are parallel to the
east!dipping shear planes[ Pressure and production data
indicate that this may be the case in the Gullfaks Field[
In addition\ collapsed pore space in the large!scale drag
zone may reduce permeability in the hanging wall[ The
onshore analogue example from Utah "Fig[ 01# shows a
299399 m wide drag zone in the hanging wall to a normal
fault with 299 m displacement[ Silica!rich ~uids have
cemented the rocks in the drag zone\ while the almost
undeformed sandstones in the footwall have retained
their original porosity a few meters away from the fault[
A similar situation may exist in the hanging walls to two
main faults on the Gullfaks Field\ where wells 23:09!C!
20 and 23:09!B!25 have shown highly calcite!cemented
reservoir rocks[
We have described a downward decrease in the amount
of large!scale drag in the hanging walls to the main faults
in the Gullfaks Field\ resulting in increasing throw with
depth[ For example\ for a fault with only a few meters
missing section at the upper Brent Group level\ the throw
at the Statfjord Formation level may be as much as 149
m[ This will a}ect communication across the fault and
must be considered when planning additional wells in
deeper strata[
Figure 03 demonstrated the presence of a zone of high
dip possibly related to a relay structure[ The amount of
shear strain within this zone is likely much lower than in
the low!dip zone immediately to the east[ When placing
an injector within the {high dip bu}er zone|\ ~uid ~ow
may be preferentially in the opening direction of the zone
"generally to the north#\ limited by the highly sheared
zone and:or minor fault to the east\ and the main fault
to the west[
The presence of local drag\ which should not be con!
fused with large!scale drag observed within the triangular
zone "Fig[ 6#\ is also of importance in oil production[
Missing section estimated from standard well log cor!
relation methods does not account for the e}ect of drag[
Also\ local drag is usually not identi_ed from seismic
data due to resolution limitations[ Consequently\ seismic
interpretation of faults intersected by wells will com!
monly be based on amount of missing section reported by
the sedimentologists[ The result may be that the measured
missing section may be much less than the total o}set
measured outside the zone a}ected by drag "Fig[ 06#[ Due
to the e}ect of drag\ a well located in the hanging wall
may intersect the interpreted horizon signi_cantly deeper
than planned[ Similarly\ a well located in the footwall
may penetrate the strata at shallower levels than
expected[ In the example from well 23:09!C!2 "Fig[ 06#\
the missing section was estimated to only 8 m\ whereas
the total o}set of layering outside the area a}ected by
drag is more than 099 m[ When more than half of all

)
469

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Fig[ 16[ "a# Seismic interpretation of footwall collapse around well 23:09!4 before the incorporation of available dipmeter data[ "b# Dipmeter data
show that dip of bedding within the two fault blocks is subhorizontal\ and not west!dipping as indicated on seismic data[ "c# Seismic data show the
presence of dipping coherent noise above the reservoir[ Within the reservoir\ this noise interferes with subhorizontal re~ections and causes the
appearance of west!dipping strata[ See Fig[ 0 for location[

faults the Gullfaks Field have associated drag\ this e}ect

must obviously be considered during well planning and
interpretation of seismic data[
Drag is also important when calculating sealing
capacities of faults[ Estimating sealing potential involves
the use of structure maps from seismic interpretation and
construction of fault juxtaposition diagrams "e[g[ Knipe\
0886 ^ Needham et al[\ 0885#[ Such estimates will generally
not take into consideration the e}ect of local drag\ since
this is not observed in the seismic data[ Whereas the total
o}set across a fault structure may predict a sealing sand!
to!shale contact\ the presence of drag may provide a
communication path for the ~uids "Fig[ 17#[ Alter!
natively\ where a sandsand contact is expected\ drag of
a shale layer between the two sandstone units may destroy

CMYK Page 469

communication between these two units[ Calculations of

a faults sealing potential ideally should therefore not be
based purely on seismic interpretation\ where the e}ect
of local drag is not easily observed[ Drag also represents
a potential source of error in statistical analysis of fault
populations "e[g[ Walsh and Watterson\ 0876# which may
be used in reservoir models[ Detailed analysis of dipmeter
data may o}er a means of compensating for such possible
errors[

8[ Conclusions
Integrated use of dipmeter data helps to constrain the
structural interpretation\ particularly where the seismic

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

"2#

"3#

"4#
Fig[ 17[ The presence of local drag may give rise to communication
across a fault which from seismic interpretation appears to be sealing[
Also\ drag of a shale layer may in some cases restrict communication
between two sand units[ The lower _gure shows the results if no drag
is associated with the fault[

data quality is poor and generally where smaller struc!

tures cannot be resolved seismically[ The analyses of 12
km of dipmeter data from 37 wells on the Gullfaks Field\
northern North Sea\ have proven invaluable for a tho!
rough understanding of the structural complexity and
evolution of the _eld[ Even where dipmeter data show
much scatter\ statistical analyses may give important
information on variations in bed geometry[ In the
Gullfaks example\ the general structure as interpreted
from 2D seismic data is con_rmed[ However\ dipmeter
data have resulted in local reinterpretations and helped
describe bedding and fault geometries in more detail[
In particular\ subseismic geometries are revealed and:or
understood by the integrated use of dipmeter data\ well
log correlation data\ core data\ and seismic data[ The
following main points are emphasised ]
"0# Dipmeter analyses from the Gullfaks Field verify the
existence of three structural domains ^ the domino
system\ the horst complex and the accommodation
zone[
"1# Dip of bedding within the rotated fault blocks of the
domino system decreases towards the main faults to
the west as a result of large!scale drag[ The drag is
caused mainly by some mechanism other than brittle
faulting\ probably by a widely distributed "ductile#
~ow at the grain scale[ The e}ect of large!scale drag
decreases with depth\ due probably to a downward
increase in the degree of consolidation at the time of
deformation[ In some areas of the Gullfaks Field\ a
{high dip bu}er zone| of low internal shear strain

"5#

460

separates the triangular large!scale drag zone and the

main fault to the west[
Dipmeter data help the interpretation of seismic attri!
bute maps\ and indicate that most curvilinear features
on timedip maps from the Gullfaks Field represent
seismic noise rather than small faults[ This conclusion
is consistent with independent _eld data\ and is con!
sistent with the conclusion that much of the defor!
mation was by grain reorganisation rather than
discrete faulting[ North!trending minor "and major#
faults are most common and will restrict ~uid ~ow in
an EW direction[
Dipmeter data show that more than 59) of all faults
on the Gullfaks Field are associated with local drag[
Drag is not common for minor EW trending faults\
probably because they are internal accommodation
structures in~uenced by lateral slip[
The total o}set across the complete fault:drag zone
may be as much as one order of magnitude larger
than missing section identi_ed by well log correlation[
The interval a}ected by drag is wider in the hanging
wall than the footwall\ but there is no apparent
relationship between interval a}ected by drag and
lithology[ However\ drag is less abundant at deeper
stratigraphic levels where the sediments are more con!
solidated[
To fully take advantage of dipmeter data\ several
di}erent display methods should be used[ The use of
stereonet plots and cumulative dip and dip direction
plots are especially valuable where much scatter of
dip and azimuth measurements exists[ If the data
quality is good\ the statistical curvature analysis tech!
nique of Bengtson "0870# is more appropriate[ If very
detailed analyses are needed\ the normal {tadpole|
plots may be used after a thorough quality check of
the raw data[

Acknowledgements
The authors want to thank Norsk Hydro\ Saga Pet!
roleum and Statoil for permission to publish these results[
The article has bene_ted from extensive reviews by Roy
Gabrielsen\ Mike Badley and one anonymous referee[
Assistance by Astri Ro
rnes\ Margrethe M[ Faaberg\ Lars
Aamodt\ and Sidsel Haugland during the work is appreci!
ated[

References
Adams\ J[ T[\ Ayodele\ J[ K[\ Bedford\ J[\ Kaars!Sijpesteijn\ C[ H[\ +
Watts\ N[ L[ "0881#[ Application of dipmeter data in structural
interpretation\ Niger Delta[ In A[ Hurst\ C[ M[ Gri.ths and P[ F[
Worthington\ "Eds[#\ Geolo`ical Applications of Wireline Lo`s II\
"pp[ 136153# Geolo`ical Society of London\ 54\ Special Publication[
Badley\ M[ E[\ Price\ J[ D[\ Rambech Dahl\ C[\ + Adestein\ T[ "0877#[
The structural evolution of the northern Viking Graben and its

461

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

bearing upon extensional modes of graben formation[ Geolo`ical

Society of London\ 034\ 344361[
Beach\ A[\ Bird\ T[\ + Gibbs\ A[ "0876#[ Extensional tectonics and
crustal structure ] deep seismic re~ection data from the northern
North Sea Viking Graben[ In M[ P[ Coward\ J[ F[ Dewey\ and P[
L[ Hancock\ "Eds[#\ Continental extensional tectonics\ "pp[ 356365#
Geolo`ical Society\ Special Publication\ 17[
Bengtson\ C[ A[ "0870#[ Statistical curvature analysis techniques for
structural interpretation of dipmeter data[ American Association of
Petroleum Geolo`ists Bulletin\ 54\ 201221[
Bigelow\ E[ L[ "0882#[ Multiwell dipmeter interpretation[ 3th Annual
SPE et al[ Archie Conference Proceedings\ 6279[
Devilliers\ M[ C[\ + Werner\ P[ "0889#[ Example of fault identi_cation
using dipmeter data[ In A[ Hurst\ M[ A[ Lovell\ and A[ C[ Morton\
"Eds[#\ Geolo`ical Applications of Wireline Lo`s\ "pp[ 176184#[
Geolo`ical Society\ Special Publication\ 37[
Dumont\ A[\ Kubacsi\ M[\ + Chardac\ J[ L[ "0876#[ The oil!based mud
dipmeter tool[ 17th Annual SPWLA Logging Symposium\ 1\ 04 pp[
Etchecopar\ A[\ + Bonnetain\ J[ L[ "0878#[ Cross sections from dipmeter
data[ American Association of Petroleum Geolo`ists Bulletin\ 4\ 510
526[
Fossen\ H[\ + Gabrielsen\ R[ H[ "0885#[ Experimental modeling of
extensional fault systems[ Journal of Structural Geolo`y\ 07\ 562
576[
Fossen\ H[\ + Hesthammer\ J[ "0887#[ Structural geology of the Gullfaks
Field\ northern North Sea[ In M[ P[ Coward\ H[ Johnson\ and T[
S[ Daltaban\ "Eds[#\ Structural `eolo`y in reservoir characterization\
Geolo`ical Society of London\ Special Publications\ 016 "pp[ 120
150#[
Fossen\ H[\ Odinsen\ T[\ F%rseth\ R[ B[\ + Gabrielsen\ R[ H[ "in press#[
Detachments and low!angle faults in the northern North Sea rift
system[ In A[ No
ttvedt "Ed[#\ Inte`rated Basin Studies\ Geolo`ical
Society of London\ Special Publications[
Fossen\ H[\ + Ro
rnes\ A[ "0885#[ Properties of fault populations in the
Gullfaks Field\ northern North Sea[ Journal of Structural Geolo`y\
07\ 068089[
F%rseth\ R[ B[ "0885#[ Interaction of Permo!Triassic and Jurassic exten!
sional fault!blocks during the development of the northern North
Sea[ Geolo`ical Society of London\ 042\ 820833
F%rseth\ R[ B[\ Sjo
blom\ T[ S[\ Steel\ R[ J[\ Liljedahl\ T[\ Sauar\ B[ E[\
+ Tjelland\ T[ "0884#[ Tectonic controls on Bathonian!Volgian syn!
rift successions on the Visund fault block\ northern North Sea[ In
R[ J[ Steel et al[\ "Eds[#\ Sequence strati`raphy on the Northwest
European mar`in\ "pp[ 214235#[ NPF Special Publication 4[
Gabrielsen\ R[ H[\ F%rseth\ R[ B[\ Steel\ R[ J[\ Idil\ S[\ + Klo
vjan\ O[
S[ "0889#[ Architectural styles of basin _ll in the northern Viking
Graben[ In D[ J[ Blundell\ and A[ D[ Gibbs\ "Eds[#\ Tectonic evol!
ution of the North Sea rifts\ "pp[ 047068#[ Clarendon Press\ Oxford[
Gauthier\ B[ D[\ + Lake\ S[ D[ "0882#[ Probabilistic modeling of faults
below the limit of seismic resolution in Pelican Field\ North Sea\
o}shore United Kingdom[ Bulletin American Association of Pet!
roleum Geolo`ists\ 66\ 650666[
Giltner\ J[ P[ "0876#[ Application of extensional models of the northern
Viking Graben[ Norsk Geolo`isk Tidsskrift\ 56\ 228241[
Goetz\ J[ F[ "0877#[ Reliable dip data in oil based mud ] Hardware and
software considerations[ 00th SPWLA Norwegian Chapter et al[
Europe Formation Evaluation Symposium\ Y\ 10 pp[
Goetz\ J[ F[ "0883#[ A common sense approach to dipmeter interpret!
ation[ Lo` Analyst\ 24\ 1029[
Hesthammer\ J[ "0887a#[ Evaluation of the timedip\ correlation and
coherence map for structural interpretation of seismic data[ First
Break\ 05\ 040056[
Hesthammer\ J[ "in press\ "b## Integrated use of well data for structural
control of seismic interpretation[ Petroleum Geoscience\ 3\ 86098[
Hesthammer\ J[\ + Fossen\ H[ "0886a#[ Seismic attribute analysis in
structural interpretation of the Gullfaks Field\ northern North Sea[
Petroleum Geoscience\ 2\ 0215[

Hesthammer\ J[\ + Fossen\ H[\ "0886b#[ The in~uence of noise in struc!

tural interpretation of seismic data[ First Break\ 04\ 198108[
Hesthammer\ J[\ + Lo
kkebo
\ S[ "0886#[ Combining seismic data sets to
enhance data quality[ First Break\ 04\ 092004
Hobbs\ B[ E[\ Means\ W[ D[\ + Williams P[ F[ "0865#[ An outline of
structural geology[ Wiley International\ New York\ 460 pp[
Hurley\ N[ F[ "0883#[ Recognition of faults\ unconformities\ and
sequence boundaries using cumulative dip plots[ American Associ!
ation of Petroleum Geolo`ists Bulletin\ 67\ 00620074[
Kaya\ M[ A[\ + Norman\ T[ N[ "0882#[ Verifying a geological structure
by applying SCAT "statistical curvature analysis techniques#
method on dipmeter data at the Umurca Field in the Turkish
Thrace[ 3th Annual SPE et al[ Archie Conference Proceedings\ 56
61[
Knipe\ R[ J[ "0886#[ Juxtaposition and seal diagrams to help analyze
fault seals in hydrocarbon reservoirs[ American Association of Pet!
roleum Geolo`ists\ 70\ 076084[
Koepsell\ R[ J[\ Jenson\ F[ E[\ + Langley\ R[ L[ "0878a#[ Formation
imaging*0 ] Formation imaging yields precise fault orientation\
minimizes dry o}sets[ Oil + Gas Journal\ 76\ 4447[
Koepsell\ R[ J[\ Jenson\ F[ E[\ + Langley\ R[ L[ "0878b#[ Formation
imaging*1 ] Imaging aids visualization of faults[ Oil + Gas Journal\
76\ 7475[
Needham\ D[ T[\ Yielding\ G[\ + Freeman\ B[ "0885#[ Analysis of fault
geometry and displacement patterns[ In D[ A[ Nieuwland\ and P[ G[
Buchanan\ "Eds[#\ Modern Development in Structural Interpretation\
Validation and Modellin`\ "pp[ 078088#[ Geolo`ical Society of
London\ Special Publications\ 88[
Roberts\ A[ M[\ Yielding\ G[\ + Badley\ M[ E[ "0889#[ A kinematic
model for the orthogonal opening of the Late Jurassic North Sea
rift system\ Denmark!Mid Norway[ In D[ J[ Blundell and A[ D[
Gibbs\ "Eds[#\ Tectonic evolution of the North Sea Rifts\ "pp[ 079
088#[ Clarendon Press\ Oxford[
Roberts\ A[ M[\ Yielding\ G[\ Kusznir\ N[ J[\ Walker\ I[ M[\ + Dorn!
Lopez\ D[ "0884#[ Quantitative analysis of Triassic extension in the
northern Viking Graben[ Journal of Geolo`ical Society of London\
041\ 0415[
Rouby\ D\ Fossen\ H[\ + Cobbold\ P[ R[ "0885#[ Extension\ dis!
placement and block rotation in the larger Gullfaks area\ northern
North Sea ] determined from map view restoration[ Bulletin Amer!
ican Association of Petroleum Geolo`ists\ 79\ 764789[
Ro
nningsland\ T[ M[ "0889#[ Structural interpretation of dipmeter
results in the Gullfaks Field[ In A[ Hurst\ M[ A[ Lovell and A[ C[
Morton\ "Eds[#\ Geolo`ical Applications of Wireline Lo`s\ "pp[ 162
175#[ Geolo`ical Society of London Special Publication\ 37[
Schlumberger "0875#[ Dipmeter interpretation[ Schlumberger Limited\
New York\ 65 pp[
Schlumberger "0889#[ Data services catalog[ Schlumberger Educational
Services\ Texas\ 097 pp[
Schlumberger "0880#[ Wireline services catalog[ Schlumberger Edu!
cational Services\ Texas\ 000 pp[
Scholz\ C[ H[\ Dawers\ N[ H[\ Yu\ J[!J[\ Anders\ M[ H[\ + Cowie\ P[
A[ "0882#[ Fault growth and fault scaling laws ] preliminary results[
Journal of Geophysical Research\ 87\ 1084010851[
Serra\ O[ "0878#[ Formation MicroScanner Image interpretation[ Sch!
lumberger Educational Services\ Texas\ 006 pp[
Stuart!Bruges\ W[ P[ "0873#[ A dipmeter for use in oil based muds[
8th International SAID "division of SPWLA# Formation Evaluation
Symposium\ 04 pp[
Thorne\ J[ A[\ + Watts\ A[ B[ "0878#[ Quantitative analysis of North
Sea subsidence[ American Association of Petroleum Geolo`ists Bull!
etin\ 62\ 77005[
Tollefsen\ S[\ Graue\ E[\ + Svinddal\ S[ "0883#[ Gullfaks development
provides challenges ] Pt[ 0[ World Oil\ 104\ 3443[
Trudgill\ B[\ + Cartwright\ J[ "0883#[ Relay!ramp forms and normal!
fault linkages\ Canyonlands National Park\ Utah[ American Associ!
ation of Petroleum Geolo`ists Bulletin\ 095\ 00320046[

J[ Hesthammer\ H[ Fossen:Marine and Petroleum Geolo`y 04 "0887# 438462

Walsh\ J[ J[\ + Watterson\ J[ "0876#[ Distributions of cumulative dis!
placement and seismic slip on a single normal fault surface[ Journal
of Structural Geolo`y\ 8\ 09280935[
Werner\ P[\ Piazza\ J[ L[\ + Raiga!Clemenceau\ J[ "0876#[ Using
dipmeter data for enhanced structural interpretation from the seis!
mic[ 17th Annual SPWLA Logging Symposium\ 1\ 04 pp[

462

Ziegler\ P[ A[ "0871#[ Geological atlas of western and central Europe[

Shell International Petrol[\ Maatshappij B[ V[
Zhang\ Y[ "0882#[ Fault identi_cation in Qiu Ling oil_eld with diplog
data[ Well Lo``in` Technolo`y\ 06\ 174177[