You are on page 1of 17

SPE 25054

I Society of Petroleum Engineers I


The Gullfaks Field Development: Challenges and Perspectives
Svein Tollefsen,' Eirik Graue, and Stein Svinddal, Statoil AIS
'SPE Member
Copyright 1992, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the European Petroleum Conference held in Cannes, France, 18-18 November 1992.
This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submined by the author@). Contents of the paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
of Petroleum Engineers. Permission to copy is restricted to an ebstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 750833836, U.S.A. Telex, 163245 SPEUT.
ABSTRACT
The Gullfaks Field is located in the Norwegian Sector of
the ~or t h' Sea, block 34/10, and currently has the
capability of producing more than 70,000 Sd/ d of oil
(440,000 stb/dl from three CBS platforms. The
reservoir san& comprise shallow marine to fluvial
sediments of the Cook Formation, Statfiord Formation
and Brent Group, ranging in age from Early to Middle
Jurassic. Water injection is the major drive mechanism
for maintaining reservoir pressure above bubble point.
Development wells have confirmed a complicated
structural picture with numerous faults beyond seismic
resolution, causing major impacts on predicting field
reserves and flow patterns. Reverse faulting in an area
of predominantly normal faulting further emphasises the
structural complexity.
Complex geology along with field performance after
water breakthrough resulted in several changes in the
initial development strategy. Production from the highly
productive Tarbert and Staffjord sands was accelerated
in order to compensate for the loss of production from
the Lower Brent sands caused by sandproduction after
waterbreakthrough. Development of complex Ness and
low productivity Cook sands have recently commenced.
Gravel packing, implemented in the Upper Brent field
development, provided sand control and increased
production rates. Various types of chemical sand
control are currently being evaluated and tested in the
field. Lowering pressure in gravel packed wellbore
region below bubble point may increase the production
rates even further.
Following advances in drilling technology, highly
deviateoP~orimntal wells improve recovery and
accelerate field development by combining production
from several reservoirs in one single well. A test
programme for surfactant and WAG flooding has been
implemented on the field. Other EORmethods, such as
gel and polymer flooding, are currently being
investigated for potential use.
Referencon and illustrations at the end of paper.
377
INTRODUCTION
The Gullfaks Field, situated in the Norwegian Sector of
the North Sea, block 3411 0 (Fig. 1 ), is the first licence
ever run by a fully Norwegian joint venture corporation.
The licence group consists of Statoil (operator), Norsk
Hydro and Saga Petroleum. The field currently has the
capability of producing more than 70,000 Sm3/d of oil
(440,000 stbld) on a stream day basis from three main
reservoirs of Jurassic age. These are the Cook Fm.,
Statfjord Fm. and Brent Group, with total recoverable
reserves estimated to 230 mill. Sm3 of oil (1.5 bill. stb)
and 25 bill. Sm3 of associated gas (890 bill. scf). The
Brent Group has been subdivided into the Upper Brent,
Tarbert and Ness fms. and Lower Brent, Etive and
Mnnoch fms.
The field produces from three separate CBS platforms,
the Gullfaks A, B and C. Gullfaks A and C are fully
independent processing platforms, with three
separation stages. The crude oil is stored in the
concrete base of the platforms and loaded into tankers
via two separate mooring buoys. Associated gas mainly
passes into the Statpipe network. Some gas is,
however, recompressed and reinjected into the
reservoir. The Gullfaks B platform provides processing
facilities for single stage separation only, and transports
partly stabilised crude to both the A and C platforms
for further processing. The particular option of further
processing on two different platforms has ensured a
high degree of flexibility for oil delivery from the
Gullfaks B platform. A separate pipeline transports the
associated gas to Gullfaks A. The platform infra-
structure and capacities are shown in Fig. 2.
This paper presents the history, current status and
Mure perspectives of the Gullfaks Field Development,
the technology available, as well as new technology
required for optimising such complex field develop-
ments. By comparing early development plans to the
present situation, emphasising in particular all the
adjustments required to get there, this paper provides
a broad base of experience which may be of great
value t o both further development of this particular
field, as well as to other field developments of similar
complexity in the years ahead.
2 THE GULLFAKS FIELD DEVELOPMENT - CHALLENGES AND PERSPECTIVES SPE 25054
HISTORICAL REVIEW
Exploration history
Block 3411 0 was awarded t o the licence group in 1978
during the 4th concession round. The first exploration
well, 3411 0-1 (Fig. 31, encountered a 160 m oil column
in the Brent Group and penetrated water bearing Cook
and Statfjord fms. Exploration well 3411 0-3 established
the oil-water contact (OWC) in the Brent Group. The
next three wells confirmed the OWC and appraised the
western part of the field.
Early use of 3D seismic introduced a new era of
exploration and appraisal drilling in the eastern part of
the field. Well 34110-7 proved a deeper hydrocarbon
system in the Cook Fm., whereas well 34110-1 1 in the
north-eastern corner of the block succeeded in proving
a deep OWC and a new oil system in the Statfjord Fm.
The exploration and appraisal phase was completed by
the end of 1983 with a total of 14 wells.
Development, Phase I
Based on well information and seismics a field
development plan in two phases was proposed.
Following the Commerciality Report in late 1980, the
authorities approved a Field Development Plan (FDP) for
the western part, Phase I in October 198 1. This plan
included the production of the Brent Group reserves
west of an assumed sealing northsouth trending fault
(Fig. 1). Two CBS platforms were required for
producing these reserves.
The overall production philosophy was t o produce the
reserves successively shallower within the Brent Group
Water injection would be the major drive mechanism,
maintaining reservoir pressure above bubble point. In
order t o avoid potential sand problems during produc-
tion, selective perforating would be used. At this stage,
the total Gullfaks recoverable reserves were estimated
t o approximately 230 mill. Sm3 of oil (Table 1).
Development, Phase I1
Based on an updated FDP, which included the eastern
part of the Gullfaks Field, the authorities approved the
Phase II development in 1985. Another CBS platform
was required for producing the Brent, Cook and
Statfjord reserves in the complex Phase II area. A 60"
deviation was considered a limit in drilling platform
wells. The overall production strategy in Phase II was
unchanged compared t o the FDP in Phase I.
Production Start
The Gullfaks A platform came on stream on December
22nd. 1986. In order t o accelerate production start, a
total of five subsea wells were predrilled and put on
stream initially. A 6th subsea well drilled on a separate
segment was tied back some months later (Fig. 2).
Gullfaks B started production in February 1988, whilst
Gullfaks C came on stream in December 1989.
During the first production period, it became clear that
the process and water injection capacity had t o be
upgraded in order t o optimise production. This is now
implemented. Fig. 4 shows production profiles as
planned in 1988 compared t o current estimates.
GEOLOGY
Geophysics And Structural Geology
Structural mapping of the Gullfaks Field is complicated
due t o poor seismic data and complex structural
geology. The first 3D survey was shot in 1979,
providing a great improvement compared t o previous
available 2D data. A second 30 survey collected in
1985 further improved the data. Recent processing
techniques made it possible t o improve seismic quality
even more. As such, a reprocessing of the survey was
completed during the spring of 1992. This resulted in
higher resolution data, particularly towards the eastern
parts of the field.
The Gullfaks Field was highly deformed during the late
Jurassic extensional period. This resulted in rotated
fault blocks with low angle normal faults, dipping
approximately 60' t o the east and typically spaced 1 -
1.5 km apart. Strata within the fault blocks normally
dip 15"-25" t o the west (Fig. 5). In addition, many
small-scale, east-west trending normal faults occur on
the field. While most faults were non-sealing during
hydrocarbon migration and filling of the reservoir,
several faults act as barriers during production, limiting
horizontal flow and pressure communication across the
field.
In the eastern part of the field a horst block dominates
the structural picture, resulting in a graben west of the
horst. Coexistence of westerly and easterly dipping
faults may have caused spatial problems accompanied
by local reverse faulting. Strata within the graben and
horst blocks are mostly horizontal or even slightly
dipping t o the east. Large scale normal faults as well as
a major drop in the Base Cretaceous Unconformity
define the northern, eastern and southern limits of the
Gullfaks structure (Fig. 5).
The stratigraphy of the Gullfaks Field show progressive
erosion towards the east. Whereas the Middle Jurassic
Brent Group is not eroded in the western part of the
field, some 600-800 m of Middle and Early Jurassic
sediments are eroded on the horst block t o the east
(Fig. 5). Structural depth maps of top Brent Group, top
Cook Fm. and top Statfjord Fm. are shown in Fig. 3.
SPE 25054 SVElN TOLLEFSEN, ElRlK GRAUE AND STEIN SVINDDAL 3
Stratigraphy And Reservoir Description
The lithostratigraphic subdivision used on the Gullfaks
Field is described by Vollset and Dor8 (Fig. 6) . This
subdivision is used on group and formation level, and at
members level in the Statfjord Fm. A further internal
field zonation is shown as well. Note that the petro-
physical interpretation presented in the first column
shows only sand, shale, calcite and coal as rock
forming constituents in the Cook and Statfjord fms. In
the Brent Group feldspar and mica are extracted as
separate additional constituents. The name "Sverdrup
Member" is proposed for the shallow marine reservoir
unit in the lower part of the Amundsen Fm.
Hydrocarbon/P/essure Systems
The Gullfaks Field is the shallowest Mesozoic oil
accumulation in the Tampen Spur area, northern North
Sea. The majority of hydrocarbon volumes have
migrated into the structure from the east and north-
east, with minor contribution from the west. A recently
discovered hydrocarbon accumulation in the south-
western part of the field suggests filling from the
southlsouth-east as well. The various hydrocarbon
contacts and volume distributions for the different
systems are listed in Table 2.
Upgrading And Redistribution Of The Reserves
During the field development planning and early
production period the Lower Brent, Rannoch and Etive
fms. were regarded as the main reservoir in the
Gullfaks Field. Early production and injection wells,
however, revealed a promising thickness expansion and
quality improvement of the parameters applied t o
reserves estimates for the Tarbert Fm. As drilling
proceeded it became clear that the Tarbert Fm.,
particularly in northern parts, had been greatly
underestimated. This was mostly due t o a re-
interpretation of well 34110-6, which revealed an
undetected fault with a 60 m missing section in the
Tarbert Fm. In addition, several development wells
drilled through major north-south trending faults
demonstrated segmentation of the faults, contributing
towards an upgrade of the reserves (Fig. 7). Current
interpretation of central parts of the field rarely shows
any Heather sequence (Upper Jurassic shale) between
base Cretaceous and the OWC. The space is replaced
by the Tarbert Fm. Furthermore, better communication
during production is obtained from segmented faults.
As such, the Tarbert Fm. is now considered the main
reservoir of the Gullfaks Field.
It has been beneficial t o optimise locations of develop-
ment wells also with respect t o the evaluation and
testing of structural closureslprospects in structurally
deeper layers. In fact, an additional 11 % of the
STOOIP is proved by development wells in six such
structures (Table 2). Even though Gullfaks is a
structurally complex field, with many different oil
systems, the main Brent oil system nevertheless
contributes towards some 70 % of the STOOIP.
Additional potentials, still not incorporated in the
STOOIP figures, are represented by the Triassic
sequence in the eastern part of the field. Some of the
wells drilled in this area have penetrated deeper oil
systems than seen elsewhere on the field.
During six years of production from the Gullfaks Field,
the reservoir development and management plans have
been modified somewhat. This is due t o improved and
updated geological, reservoir and well performance
information, as well as new technology.
Development strategy
A primary set of producers and injectors are dedicated
t o each reservoir unit in order t o obtain sufficient
recovery. The number of wells needed is a function of
reserves and the communication complexity for the
reservoir. The production wells are situated high on the
structures in the eastern parts of the fault blocks,
whereas the injectors are placed downdip near the
OWC (Fig. 5). Generally, producers are supported by
injectors within the same "major fault block".
The primary drive mechanism for oil production is water
flooding, maintaining reservoir pressure above bubble-
point. Water injection is and will continue t o be the
major pressure supporting mechanism together with
natural basin influx.
Communication
The communication within the reservoirs and between
fault Mocks on the Gullfaks Field is a subject of great
attention. The best possible picture of the com-
munication relationships throughout the field is formed
through the interactive use of structural and strati-
graphic information, openhole pressure data, open and
cased hole logs, production data, permanentlmemory
pressure gauges3, transient well testing and radioactive
tracers.
The Tarbert and Lower Brent reservoirs show excellent
lateral and vertical communication both internally and
between fault blocks. Regional lateral calcite layers
may locally restrict vertical flow, but good lateral
communication still exists.
Extensive faulting causes flow patterns which are
difficult t o predict in heterogenous reservoirs like the
Ness and Statfjord fms. Generally, pressure data and
well performance indicate good communication within
each fault block. Faults with a throw greater than 30-
1 00 m (reservoir dependent), however, distortleliminate
this communication. Major faults are mostly detected
THE GULLFAKS FIELD DEVELOPMENT - CHALLENGES AND PERSPECTIVES SPE 25054
by seismic interpretation. Nevertheless, in as much as
70 % of the Gullfaks wells additional small-scale faults
not previously seen on seismic are detected. This
further complicates communication patterns.
Well Placement
A total of approximately 100 wells are needed t o
develop the Gullfaks Field. Generally, for every t wo
producers one injector is needed. In good homogenous
reservoirs like the Tarbert Fm. a set of high capacity
injectors ("super water injectors") situated far t o the
west in fault block row 5 is able t o sweep and maintain
the reservoir pressure in the formation from west t o
east. Due t o good lateral communication it is possible
t o place these injectors at a considerable distance from
the OWC. This provides a uniform rise of the water
level and hence further postpones water breakthrough
in the oil producers. On the contrary, a closer
configuration between producers and injectors is
needed in heterogenous reservoirs.
Due t o improved drilling and completion technology,
high anglelhorizontal wells may be drilled parallel t o
faults or erosion planes, maximising distance to the
OWC, and thereby postponing water breakthrough and
improving drainage of attic oil (Fig. 5). Consequently,
oil recovery is acceleratedhmproved and the required
number of primary wells is reduced. Furthermore,
platform wells may now reach areas which were
previously planned for subsea development.
Drainage Phi/osophy
According t o previous FDPs, development was to start
with the Brent reservoir in both Phase I and Phase II
areas. The Statfjord and Cook reservoirs were t o be put
on stream when a production plateau was reached in
order t o compensate for production loss from the Brent
reservoir due t o water breakthrough. Development of
the Brent reservoir was t o start with the structurally
lower Etive and Rannoch fms. (Lower Brent), continuing
with the Ness Fm., and finally producing from the
Tarbert Fm.
During 1990-1 991 this strategy was changed. The
Gullfaks Field faced severe water breakthrough in the
Lower Brent producers. In addition t o a substantial
decline in oil production, the total liquid rates dropped
due t o sand production. Consequently, Lower Brent oil
production level stabilised prior t o reaching the planned
plateau rate for the field based upon the positive pro-
duction level experienced during 1989. In order t o com-
pensate for this, development of Tarbert and Statfjord
sands were accelerated (Fig. 4). Simultaneously, sand
control actions were performed in Lower Brent wells,
successfully restoring liquid rates t o original level, and
i n some cases increasing the oil production.
For this reason, the development of the less productive
Ness and Cook fms. were postponed. Simulation
studies show that both reservoirs need time in order t o
reach a sufficient recovery. Further delay will therefore
have a negative influence on recovery. Furthermore,
simulation studies predict water breakthrough in some
Tarbert oil producers by 1992, hence making oil
production from Ness and Cook wells important in
compensating for Tarbert production loss. As such,
development of the Ness and Cook reservoirs currently
have high priority.
Originally, the eastern area (Phase Ill should produce
from the Brent reservoir first, followed by the Statfjord
and Cook fms. Evaluation work carried out prior t o
production start indicated that alternating drilling of
development wells into the Brent and Statfjord
reservoirs would be most beneficial, providing sufficient
time t o evaluate and incorporate critical data prior to
planning of the next well in the same reservoir.
The first well drilled from Gullfaks C was originally
planned as a high productivity Lower Brent producer.
Due t o a geological surprise the well was completed as
a low productivity Cook producer. Remapping and
revised drainage strategy have resulted in the Statfjord
reservoir being more important in increasing the
production on Gullfaks C than previously anticipated.
Recent drilling in southern parts of fault blocks 1 and 2
indicates an increase in Statfjord reserves, further
confirming its major role in the Phase II development.
Reducing the number of primary wells and accelerating
the oil production make commingled production and
injection a subject of consideration. A major challenge
is the ability t o control and monitor production and
injection from the various reservoirs. Productivity and
pressure contrasts may initiate cross flow and
unbalanced flow profiles. This could result in
operational problems due t o sand production and
reduced recovery from low productivity zones.
Lower Brent
Lower Brent, being the first reservoir on production,
now faces severe water breakthrough i n most of the
producers in the Phase I area4. Phase II wells have
produced for only 2 years and water breakthrough is
not expected until 199211 993.
The characteristic coarsening upward sequence in
Lower Brent causes override of injected water, thus
causing poorly flushed Rannoch-I and Rannoch-2
sections. Once water breakthrough occurs, the water
cut increases rapidly before stabilising at 50-70 %.
Perforating low in the Rannoch Fm. does not alone
prevent water from entering the Etive Fm.
Water production causes sand production at lower
liquid rates, hence significantly reducing MSFR.
However, the application of sand control methods in
water producing wells has increased the total liquid
rates, and hence the oil production. In order t o improve
the recovery in the low productivity intervals, horizontal
wells and propped hydraulic fracturing in the low
SPE' 25054 SVElN TOLLEFSEN, ElRlK GRAUE AND STEIN SVINDDAL 5
I
381
permeable Rannoch-112 sections have been
successfully implemented.
Tarbert-2 and Tarbert-3
The Tarbert-213 units are the most productive reservoir
units on the Gullfaks Field today, generally with
permeabilities in the order of 1-10 Darcies (Fig. 6).
Good productivity and communication throughout the
field give this reservoir highest priority. The reservoir
pressure is stabilised at a depletion from initial pressure
of 2,000-3,000 kPa throughout the field, although the
Tarbert Fm. in the Phase II area seems t o be somewhat
isolated from the rest of the Tarbert Fm. Pressure data,
however, indicate communication with the Lower Brent
reservoir in the Phase I area.
Only two producers on the western flank currently
produce water (A-21 and B-l3), however, water break-
through is expected in wells in the central area (fault
blocks 4 and 5) in 1992. The oldest Tarbert well, 9-1,
has produced approximately 5.2 mill. Sm3 of oil since
1988. The primary drilling programme will be com-
pleted in 1993. Additional offtake points in the Tarbert
Fm. will be obtained by sidetracks, recompletions and
finally commingled production in high anglelhorizontal
wells together with NesslTarbert-1 units.
Although the NesslTarbert-1 reservoirs have several
highly productive sands, small sand bodies with limited
lateral and vertical communication within the units
make reservoir management challenging. Pressure and
log evaluation confirm the complexity of fluid flow.
Local pressure differences between Ness sands may be
as high as 3,000 kPa. Tarbert-213 and Lower Brent
production influence fluid movement in NesslTarbert-1.
Pressure depletion was observed prior t o development
of the NessKarbert-1 reservoirs. Tarbert-213 influences
production in upper parts of the Tarbert-1 and Ness
reservoirs, whereas Lower Brent affects production in
the Ness-1 and Ness-2 units.
The NessKarbert-1 units will be developed with higher
well density than the other reservoirs. This will partly
be obtained by commingling and recompleting both
Tarbert-213 and Lower Brent wells which penetrate the
Ness Fm. Unconsolidated formations, pressure varia-
tions and expected rapid water breakthrough demand
sand control. Reservoir simulation models and field
observations suggest water breakthrough within the
first year of production. The ability t o process produced
water as well as handling associated sand production
are key factors in the NessKarbert-1 development.
Statfjod Formation
Production from the Statfjord Fm. started in 1990. In
contrast t o the characteristic westerly structural dip in
most of the Gullfaks Field, strata within fault block 1
dip gently t o the south. Heterogeneity and faulting of
the reservoir have complicated the prediction of
communication patterns. Besides, there is virtually no
water zone present, enforcing water injection into the
oil column. Placing injectors is therefore a subject of
great attention, aiming at postponing water
breakthrough and improving reservoir sweep.
Sand control is important in order t o achieve good
recovery. Selective perforating, gravel packing and
chemical techniques have been used and proven
successful in various wells.
Cook
Production from the Cook Fm. in Phase I started in
1986 when the formation was unexpectedly found oil
bearing in the forth subsea development well, A4H.
The well was completed as an alternating gas injector
and oil producer. The Cook-3 unit is the most important
reservoir in the Phase I area, whereas Cook-2 has small
volumes above the OWC and low productivity (average
permeabilities of 1-4 and 0.01-0.1 Darcies
respectively). Final development of the Cook Fm. i n
Phase I starts during 1992, upon which a total of 5
wells will be drilled.
Two thirds of the Cook reserves in the Phase II area are
located i n the low productivity Cook-2 sands, thus
representing a challenge in increasing recovery beyond
the expected 15 % resulting from conventional wells
and water drive. In addition t o the planned low
deviation wells producing from Cook-3 sands via
induced fractures and perforations in Cook-2, horizontal
wells will be considered drilled through the entire Cook-
2 section. As such, all barriers may be penetrated with
the longest sections possible in each sand body.
Vertical producers will be fractured and stimulated
using proppants in order t o maximise productivity from
Cook-2. Final development of the Cook Fm. in the
Phase II area is t o commence in 1993.
WELL COMPLETION
The Gullfaks wells have traditionally been completed
with a standard 7" liner and 5 'k" tubing configuration
(Fig. 8). But since a wide variety of well operations are
required in the various wells, the standard completion
configuration, along with increasing well deviations and
complex well paths, have frequently complicated such
operations. To improve access, the completion
configuration was therefore simplified by introducing
the MONOBORE concept (Fig. 8). In this configuration
the 7" liner is directly connected t o a 7" tubing
configuration t o the surface, resulting in a smooth
pipeline from bottomhole t o surface. This approach has
significantly reduced the number of workovers which
would otherwise be required, e.g. having t o pull the
tubing prior t o gravel packing "old" wells. Thus, the
6 THE GULLFAKS FIELD DEVELOPMENT - CHALLENGES AND PERSPECTIVES SPE 25054
risks of higher operational costs and delayed drilling
programmes are reduced as well.
THE SAND PRODUCTION PROBLEM
The highly productive reservoir sands of the Gullfaks
Field are poorly cemented and consolidated. For this
reason the high production potential in most oil wells in
the field could not be fully utilised due t o sand
production.
In order t o overcome this problem, the strategy
originally focused on selective perforation in consoli-
dated sand intervals, primarily in Lower Brent wells.
Thus, oil in the surrounding highly permeable sands of
poor consolidation was produced through these
perforations. The consolidated intervals were identified
from core data, openhole log data and past experience.
The strategy was successful for some time, however,
upon experiencing water breakthrough in the wells,
MSFR was significantly reduced. As such, sand control
would be required in order t o maintain high production
rates at the field. Gravel packing has so far been the
major method of sand control, but other methods are
also currently being testedlinvestigated.
SAND CONTROL - A REVIEW OF
PAST EXPERIENCE AND RECENT
DEVELOPMENTS
Gravel Packing - An Established Method of
Sand Control
The first gravel packed well at the Gullfaks Field, well
A-23, was packed across a 41 m perforated interval in
May 1989. The method has been greatly improved ever
since.
Productivity loss due t o the potential collapse of poorly
gravel filled perforation tunnels, and subsequent fill-up
of poorly sorted reservoir sand, is prevented by
increased underbalance during perforation and improved
prepacking techniques.
Due t o the danger of bridging during gravel packing,
particularly in highly deviated wells, the maximum
length which could be gravel packed in one single stage
has been restricted. Previously, therefore, sand control
actions across long intervals implemented use of multi-
stage gravel packing techniques, e.g. well A-33, which
was gravel packed in 4 stages across a total interval of
150 m. However, this is a time consuming and
expensive procedure. By introducing the shunt tubular
system these problems were significantly reduced. The
shunts provide a passage for the gravel pack slurry in
tubes with restricted leak off and reduce the danger of
bridging. The tubes are mounted along the outside of
the screens. Consequently, it is possible t o gravel pack
much longer intervals in one single stage, thereby
reducing operational duration and costs due t o delayed
production. The longest single stage gravel pack t o
date was installed in the 68O deviated well 8-20 in
February, 1992, with a total length of 1 10 m.
Another costly aspect of gravel packing wells with 5 H "
tubing configurations was that the operations had t o be
carried out using the drilling rig, deferring other drilling
operations and eventually reducing the field production
revenue. Recent estimates indicate that total costs of
delaying drilling operations 1 day amounts to approxi-
mately NOK 5 mill. (USD 800,000). Although the intro-
duction of the 7" MONOBORE system (eliminating the
necessity of pulling production tubings) and enhanced
gravel packing technology reduced the duration of
gravel packing operations, the costs involved are still
significant. For this reason, a separate snubbing rig
now carries out such operations. While the snubbing rig
is occupied gravel packing one well, the drilling rig may
drill the next well at another location on the platform.
Additional costs of operating the snubbing unit are
minor compared with the savings achieved by pre-
venting deferments on the primary drilling programme.
To date, 17 oil wells (35 % of all the production wells)
have been gravel packed. Production experience from
these wells is good, and close t o half the total field
production is currently produced from these wells.
Stimulation of Gravel Packed Wells
Although the extensive gravel packing scheme
implemented in Gullfaks oil wells has increased MSFR
and thus production rates t o a significant degree,
decline in productivity from some of the wells is
observed over time. The tendency is most pronounced
in the older wells which were perforated in low under-
balance and gravel packed without first being
prepacked5. Even though the mechanism by which the
productivity starts declining is not yet fully understood,
significant efforts have been made in order t o develop
a satisfactory stimulation procedure for restoring
productivity in gravel packed wells.
The first stimulated well was A-23 in December 1990.
Since then, several gravel packed wells have been
stimulated, the experience from each job contributing
towards a continuous improvement of the procedures.
The present "state of the art", aiming primarily at
dissolving and stabilising clay minerals, involves the
successive injection of mud acid followed by clay acid6.
In order t o gain experience from each stimulation job,
wells are, whenever possible, production tested both
prior to the stimulation and immediately after. Multirate
well testing generally confirms a significant increase in
productivity immediately after stimulation. Similar tests
run at a later time, however, show that productivity
starts t o decline again after some time. Such obser-
vations enforce a somewhat frequent stimulation sche-
SPE 25054 SVElN TOLLEFSEN, ElRlK GRAUE AND STEIN SVINDDAL 7
dule. Several wells, among those well A-23, have been
restimulated. Because of the fairly high level of costs
involved, particularly due t o a relatively long bean-up
period after stimulation, the timing of such jobs must
be carefully considered in an overall cashflow analysis.
Producing Gravel Packed Wells With Pressure
In Gravel Packed Region Below Bubble Point
There is a somewhat significant pressure drop across
the gravel pack and the transition between crushed
zone and prepacked perforation tunnels (Fig. 9). As
such, it is possible t o increase the production rate in
gravel packed wells by lowering the flowing pressure
within the gravel packed region below bubble point,
whilst at the same time maintaining the near-wellbore
reservoir pressure above bubble point.
Due t o turbulence effects, this pressure drop is clearly
rate dependant, and may typically vary between 500-
3,000 kPa. The pressure drop has been quantified by
the application of multirate well testings. Hence, by
plotting sandface pressure and pressure drop across the
gravel pack as functions of production rate, it is
possible t o estimate an additional production potential
(Fig. 10). The principle is applied in several gravel
packed wells at the field and has increased production
rates in these wells by approximately 20 %. Naturally,
the gain would be at its highest in wells having
significant pressure drop across the gravel pack. Such
wells are, however, normally restricted by minimum
well head pressure, and would probably therefore at
some point be subjected to acid stimulation in order t o
reduce the pressure drop as much as possible. By
applying this principle it is possible, nevertheless, t o
defer such stimulation jobs t o a significant extent.
Chemical Sand Control Methods
Generally, there are two alternatives of chemical sand
control methods available today:
Squeezehnjection of phenolic resin coated
proppants (resin slurries) into cavities behind
casing and in the perforation tunnels
Direct injection of consolidating chemicals into
the formation
The injection of resin slurries has successfully been
implemented in several wells on the Gullfaks Field. The
first well, B-2, was treated in August 1991, increasing
the maximum sand free liquid rate by approximately 50
%. Carbolite or just normal, carefully sized sand have
been used as proppants. The proppants are squeezed
into the perforation tunnels and other cavities behind
the casing caused by previous sand production and act
essentially as porous plugs preventing sand grains from
entering the wellbore.
The obvious advantage of this method compared t o
standard gravel packing is the non-restricted wellbore,
providing improved access t o operations in the well.
Should the method fail, it is still possible t o install a
gravel pack across the treated zone. On the other hand,
it is necessary t o drill and circulate out consolidated
excess materials afterwards. This is carried out using a
snubbing rig. The application of alternate pumping of
resin slurry and chemical diverters have made it
possible t o treat long intervals in one single stage. Well
C-8 was recently treated by this method over a 1 13 m
zone containing 8 separate perforation intervals.
A 3 m perforation interval in well B-20 was subjected
t o direct injection of consolidating chemicals in April
1992. The injected chemicals, an epoxy resinlafterflush
system, were placed using snubbing equipment.
Significant sand production afterwards indicates failure,
most probably due t o incomplete placement of the
chemicals. Unfortunately, the well operations time
schedule did not permit any logs t o be run afterwards
in order t o confirm the cause of failure. The well is
currently being gravel packed across this interval.
Propped Hydraulic Fracturing
By this method a fracture is created in a low permeable
reservoir zone (with high degree of sand consolidation)
and filled with proppants in order t o keep it "open".
The fracture is designed such that it extends into the
overlying highly productive reservoir sands with low
consolidation7. The technique was recently
implemented in several oil wells in the Lower Brent
reservoir. A major problem was initially related t o
proppant flowback during production, primarily due t o
extremely low net confining pressure along with
required high flow rates. To rectify this problem, resin
coated proppants have later successfully been applied.
LOOKING AHEAD
Continuous Reservoir Development
Pressure support in the heterogenous formations like
Ness and Statfjord will represent a major challenge in
the years t o come. Economical and technical
considerations may not justify the drilling of injectors in
small fault blocks with restricted communication t o
adjacent fault blocks. Thus, concepts such as
producing wells with reservoir pressure below bubble-
point must be carefully considered in order to maintain
production rates and achieve satisfactory drainage.
The horizontal well concept as an approach for
improving oil recovery was recently applied t o the
Gullfaks Field upon the drilling of Lower Brent well A-
34A. The plans ahead are t o implement this concept in
the Upper Brent reservoirs by drilling horizontally
through both the stratified Ness Fm. as well as the
8 THE GULLFAKS FIELD DEVELOPMENT - CHALLENGES AND PERSPECTIVES SPE 25054
Tarbert Fm. Instable hole conditions during drilling and
completing the well with satisfactory sand control are
aspects which require careful planning and sophisti-
cated technical solutions. Horizontal cased hole gravel
packing across a 400 m perforated interval is currently
being planned for sand control purposes in well 8-23
(which will be the first horizontal Upper Brent well on
Gullfaks).
The efforts of reducing injection water override effects
in the Lower Brent and Ness reservoirs will continue.
Established methods, such as zone isolations,
perforating wells structurally deep, varying perforation
density and drilling horizontal wells in low permeable
sands will be applied and further improved in the
future. The approaches of WAG injection and various
high productivity reservoir blocking techniques, will be
also be considered for application on a field scale basis.
Production With Reservoir Pressure In Near-
Wellbore Region Below Bubble Point
The Gullfaks Field is classified as an undersaturated oil
reservoir. Upon proposing the original field development
plans, attention was therefore paid t o potential prob-
lems of reduced recovery and productivity due t o the
creation and production of secondary gas. As such, the
field was planned exploited by a complete reservoir
pressure maintenance scheme, maintaining the
reservoir pressure above bubble point.
Detailed reservoir simulation studies have, however,
identified a potential of accelerated oil production in
some wells by reducing near-wellbore pressure below
bubble point. This increases pressure drawdown and
hence production rate. By producing the wells at a rate
corresponding t o a particular pressure drawdown
profile, aided by detailed reservoir simulation, it is
possible t o control secondary gas movement such that
all liberated solution gas may be produced directly and
evenly distributed into the well. A pilot programme was
initiated in April 1992 in the horizontal well A-34A,
including the careful monitoring of productivity develop-
ment and GOR/wellhead pressure in this well and the
stratigraphically higher surrounding wells B 4 and 8-7.
The programme resulted in an immediate increase in
production rate by approximately 40 % in well A-34A.
No decline in productivity due t o gas blocking or re-
duced relative permeability t o oil is observed yet. Other
wells are currently being evaluated for potential produc-
tion with near-wellbore pressure below bubble point.
The creation of secondary gas caps for the purpose of
draining additional oil reserves structurally updip of the
wells (attic oil), may be a possible approach t o
enhanced recovery provided it is possible t o control gas
movement in the reservoir. Producing oil wells downdip
below bubble point may as such represent an
alternative t o secondary gas injection.
Enhanced Oil Recovery - Advanced Methods
Motivation
Reservoir simulations based on the current field
development strategy predict that more than 300 mill.
Sm3 of oil will remain in the Gullfaks reservoirs at the
end of field life. A substantial part of this oil is
contained in rocks with excellent reservoir properties.
All the major platforms are installed, and uncertainties
concerning reservoir architecture, fluid distributions and
flow performance decrease as more field history
becomes available. It is an obvious challenge t o recover
as much of this oil as possible, which is one of the
main goals of the Current reservoir management efforts.
An extensive programme has been initiated t o verify
techniques or methods capable of improving recovery
on the Gullfaks Field. The stages i n a process of
selecting and verifying EOR methods are as follow:
1. Identify the EOR potential.
2. Identify applicable EOR methods.
3. Conduct laboratory testing and simulation
studies.
4. Perform well tests and field pilots.
5. Implement on field scale.
Potentials
Remaining oil in the Gullfaks reservoirs are mainly
associated with:
high residual oil saturation
poor vertical sweep efficiency
attic oil residuals
low permeability regions
Method selection
Three systems are identified t o have an EOR potential
specifically for the Gullfaks Field (Fig. 11 1. These are:
Water-alternating-gas injection I WA GI,
improving vertical sweep efficiency and
recovering attic oil.
Thin polymer gels, - blocking high permeable
reservoirs t o improve vertical sweep efficiency
in low permeable reservoirs.
Surfactant flooding, reducing residual oil
saturation by improving microscopic sweep
efficiency.
SPE 25054 SVElN TOLLEFSEN, ElRlK GRAUE AND STEIN SVINDDAL 9
Proiect status
The WAG project started in 1990. Simulation studies
were performed in order t o identify the improved oil
recovery potential. A pilot test was designed and the
necessary modifications on the water injector ended in
January 1991. The pilot test commenced in March
1991, and up t o June 1992 three gas injection periods
were performed in the central area of the field (fault
block 3F). Gas breakthroughs were observed in the
main target wells. Maximum water cut decreased from
54 % t o 42 % in the main target well. Post test
simulation studies show improved recovery of
approximately 100,000 Sm3 of oil. A more extensive
WAG implementation, including more water injectors,
is currently evaluated.
The thin polymer gels project was initiated in 1989 for
the development of a chemical system capable of
selectively blocking override zones some 50 t o 100
meters into the reservoir. Three different systems are
identified with interesting properties. A field test is
planned in 1993.
wells drilled from the B-platform. Natural water influx
is expected t o provide sufficient pressure support.
Drilling is due in 1993.
The Tordis Field, located in block 3417 (Fig. 1) and
operated by Saga Petroleum A/S, was discovered in
1987. Pressure data showed depletion of oil in the
Brent Group caused by the Gullfaks Field production.
The FDP involves a subsea development connected t o
the C-platform with pressure maintenance by water
injection. Production start is planned for 1994.
Production well C-7 proved oil in the Lunde Fm. in the
eastern part of the Gullfaks Field (fault block 1). Oil
was earlier observed in exploration well 3411 0-1 3. The
Lunde reservoir was production tested in C-7, but
showed an immediate decline in reservoir pressure,
indicating a somewhat limited communication in parts
of the formation. Evaluation of development strategy
will start in near future.
The surfactant flooding project started in 1989 by
screening studies and optimising of a surfactant system
for the Gullfaks Brent oil. The system reduced residual
oil saturation in core flooding experiments from 3540
% t o 5 %. Two field tests have been performed in two
different wells t o verify laboratory results.
Measurements of residual oil saturation were carried
out using the Partitioning Tracer Test Technique, along
with an extensive logging programme. Data are
currently being evaluated. Depending on the results, a
further pilot field test will be considered.
Discoveriiw and prospects
A number of recent discoveries and promising
prospects are located in the vicinity of the Gullfaks
Field. Three discoveries with proven significant reserves
are the "Gullfaks Snr", the "Gullfaks Vest" and the
"Tordis" Fields. In addition, the Lunde Fm. in fault block
1 was found oil bearing in well C-7 (Fig. 1).
Exploration well 34110-2, drilled in 1978, led t o the
discovery of the Gullfaks Ssr Field (Fig. 1). Further
appraisal drilling proved oil and gas in both the Brent
Group and the Statfjord Fm. The field is structurally
complex with many different OWC's and pressure
regimes. An FDP has not yet been approved by the
licence, but preliminary plans indicate a separate
platform with flowlines t o the main field for processing
and transport. One or t wo long range wells may be
drilled from the,P-platform, accelerating production
start from the oil ;one.
Exploration well 3411 0-34, drilled in 1991, led t o the
discovery of the Gullfaks Vest Field (Fig. I), proving oil
in the Tarbert and Ness fms. According t o current
The Gullfaks Field area is structurally complex. 70
% of the wells prove minor or large faults not
previously seen on seismics.
Continuous production drilling has resulted in a re-
distribution of the estimated reserves. The Tarbert
Fm. is now the main reservoir on the field.
Accelerating the development of the highly
productive Tarbert and Statfjord fms. increased
the plateau rate of the field.
Advances in drilling and well completion,
commingled productiontinjection and high capacity
water injectors accelerate oil production and
reduce the total number of wells required.
Gravel packing is currently the main technique of
sand control on the Gullfaks Field. Recent
improvements have reduced costs of implemen-
tation and improved production behaviour.
Squeezehnjection of resin coated proppants has
proved a successful and cost effective method of
chemical sand control in "old" production wells.
plans, the field will be woduced from t wo horizontal
385
10 THE GULLFAKS FIELD DEVELOPMENT - CHALLENGES AND PERSPECTIVES SPE 25054
ACKNOWLEDGMENT
3.
The authors wish to express their sincere thanks to all
colleagues at Statoil Bergen Operations for their advice
and encouragement in producing this paper.
The authors would also like to thank the management
4.
of Den Norske Stats Oljeselskap a.s IStatoilA/SI, Norsk
Hydro A/S and Saga Petroleum A/S for their permission
to publish this paper.
REFERENCES
6.
1. Vollset, J. and Dor6. A.G.
"A Revised Triassic and Jurassic tithostratigraphic
Nomenclature for the Norwegian North Sea'
Norwegian Petroleum Directorate Bull. No.3, 1-53, 1984
2. Pettorson, 0.. Storfi, A., Ljosland, E. and Massie, I. 7 .
"The Gultfaks fie1d:Geolog y and Reservoir Development"
North Sea Oil and Gas Reservoirs-ll, The Norwegian
Institute of Technology, Graham & Trotman, 1990
TABLE 7: Estimated Recoverable Reserves
Unneland, T. and Haugland. T.
"Permanent Downhole Gauges Used in Reservoir
Management of Complex North Sea Oil fields'
Paper presented at the 3rd Latin American Petroleum
Congress, CONEXPO ARPEL, Rio de Janeiro, Brazil,
October 18-23, 1992
Anes, H.M., Haga, O., Instefjord, R. and Jakoben, K.G.
"The Gullfaks Lower Brent Waterflood Performance'
Paper presented at the 6th European Symposium on
Improved Oil Recovery, Stavanger, Norway, May 21-23,
1991
Svendsen, 0.. Kleven, R., Abnes, N. and Hartley, I.
'Stimulation of High Rate Gravel Packed Oil Wells
Damaged By Clay and fines Migration: A Case Study,
Gultfaks field, North Sea"
Paper SPE 24991, presented at the European Petroleum
Conference, Cannes, France, November 16-1 8, 1992
Unneland, 1. and Waage, R.I.
"E'xperience and Evaluation of PIoduction Through High
Rate Gravel Packed Oil Wells, Gultfaks field, North Sea'
Paper SPE 22796, presented at the 1991 SPE Annual
Technical Conference and Exhibition, Dallas, Oct. 6-9
Bale. A.. Owen, K. and Smith, M.B.
"Propped Fracturing as a Tool for Sand Control and
Reservoir Management "
Paper SPE 24992, presented at the European Petroleum
Conference, Cannes, France, November 16-1 8, 1992
at Gullfaks
TABLE 2: Hydrocarbon Contacts and Volume Distributions at Gullfaks
Brent
Statfjord
Cook
Y w
1981
1984
1988
1992
Statfjord TOTAL
Reservoir
Brent
Cook
Statfjord
Cook
Brent
Lunde
Ateatfault block OWC ST001P
I 1 I I
Phase 1/11
lexcl. 2A,6A,6B, 6AI
Phase II
Phase II
lo" sm=
230.0
21 0.0
210.0
230.0
lo= sm3
19
33
26
Lowga Brsnt Upper Brent
"I Reserves proved by development ws/s, partly incorporated in recoverable reserves
%
9
16
11
Cook
' ' ' ' ' '
lob sm3
135
109
81
7 5
-
IcPSmJ
5 5
70
83
108
.
lob Sm=
40
12
13
21
' ' ' " "' '
%
59
52
39
33
' "" "
%
24
33
39
4 7
%
17
6
6
9
SPE 2 5 0 5 4
The Gullfaks Field
LOCATION MAP
Fig.1 Location map of the Gullfaks Field.
SPE 25054
- OIL
- 6' i C
.......... GAS
.,.-. WATER
. FL.OWI.INE
SUBSEA WELl
, @ SUBSEA COMPLETED INJECTION
WELL
il Q s rn. 2, : k r r ~
- - -
Fig.2 Field facilities, design capacities and infrastructure.
producer Intersection Base Cretaceous
ir, Injector unconformity and the fault plane:
...... : OWC - T. Brent eroded
m B. Brent eroded
6 BEo: number o 0.5 1.5
--- --
tructural depth map,
Structural depth map,
Top Statfjord Fm.
km
Fig.3 Structural depth maps
SPE 25054
Production rate
(Smsd)
- -
1
EBIRannochIEtive Tarbert 133 Ness Statfjord I Co o k '
I
3c FDP -88
- A
Fig.4 Current history and forecast 1992 update compared to Field Development Plan 1988
SWI TP M TP~/~'~T/N!' NIRI CI LBP LBP CP SP SI
I
TP . TARBERT PROWCER SWI : SUPER WATER NECTOR
NP : NSS PROWCER. N : NSS NECTOR ' Hwizmtal scab apfm&mt&y 3 x vertical j - PERMEABLE NOVERTICAL BARRERS
LBP: LOWER BRENT PROWCER CI : COOK NECTOR
CP . CO( X PROWCER SI : STATFJORD NECTOR
SP . STATFmD PROWCER
a IWi PERMEABLE-. VEfllCAL BARRERS
0 LOW PERMEABLESANS. VERTICAL BARRLRS
Fig.5 An east-west cross section showing structural styles, setting and we1 placement of the GuUfaks Field
YULE PROOELNi
DEEP MARINE SHALES
RESERVUR
cam
FI WECRUW NEARWORE
DEEP MARINE %ALES
Fig.6 A composite log display. Composed of wells 8-9 (Brent),
C-3 (Cook) grid C-2 (Statfjord).
BEFORE DRI LLI NG
AFTER DRI LLI NG
0-3
Fig.7 Segmentation of a fault leaves more of the
reservoir above the OWC, compared to a "one big
fault" situation. ( Modifiedafter Petterson et.al. )
m 5 1/2" TUBING COMPLETION r MONOBORE C o MP ma
llNlMUM ID: 3.7" MINIMUM ID: 6"
Fig.8 5 112 Tubing and 7" Monobore completion
configurations
Fig.9 Shematic perforation tunnel in gravel packed
wells showing various pressure regimes.
SPE
ig. 10 Production potential in gravelpacked wells.
SURFACTAN T
OI L
GEL WATER
0 GAS
SURFACTANTS+
MOBlLlSED OI L
g.11 Improved vertical (GELIWAG) and microscopic
(surfactant) sweep efficiency.

You might also like