You are on page 1of 15

SPE 134953

Post-Plateau Reservoir Management in the Magnus Field


Tim P. Moulds, SPE, Phillip Trussell, SPE, Demet Erbas, SPE, David J. Cox, Ewan D. Laws, and Chris Davies, SPE,
BP Exploration Operating Company Ltd., and Neil Strachan, SPE, SPEC Ltd.

Copyright 2010, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Florence, Italy, 19–22 September 2010.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed
by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or
members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is
restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract

Magnus is a high productivity oil field in the northern North Sea. First oil was produced in 1983 and the plateau of 150 MSTB/D
ended in 1995. In the post-plateau period a variety of reservoir management techniques have been employed to arrest decline rate.
Two major projects have been executed post-plateau: in 2002 a gas injection EOR scheme was initiated and more recently
additional drilling slots were added to the platform to increase reservoir access. These major projects have rejuvenated the field
development options and will enable significant oil production beyond the next decade.
The EOR scheme exemplifies the synergies that are important to continuing to extract value in a mature oil province. The
gas that is injected is sourced from several other fields and provides an export route for associated gas that would otherwise be
stranded. After seven years of operation the impact of the EOR scheme is readily quantified, as are the challenges inherent in
operating such a scheme. This paper describes the reservoir performance of EOR, discusses observations that may impact gas
relative permeability under WAG operation, describes the tools that have supported reservoir management decisions and considers
the operational issues that continually challenge delivery and the surveillance approach taken to mitigate these.
The extended slots project used slot-splitter technology to increase the number of wells. Wells from the new slots are
now the longest wells that have been drilled from the platform enabling recovery from targets at the periphery of the field.
The combination of EOR and additional slots has generated a position where the field is once again opportunity-rich.
Together with technology developments in 4D seismic and revised geological description a new tranche of opportunities are being
identified that combine initially dry oil attic targets along with the underpinning volumes from EOR delivery.

Introduction

The Magnus field is located in blocks 211/7 and 211/12, it is the most northerly field on production in the UKCS, discovered in
1974 with first production in 1983. There are over 100 well penetrations in the field comprising exploration, appraisal and
development wells and sidetracks. The current owners of the Magnus field are BP Exploration Operating Company Ltd (Operator
– 85%), Nippon Oil Exploration and Production (NOEPUK) (7.5%), Eni (UK) Ltd (5%), Marubeni North Sea Ltd (2.5%).
The reservoir is formed by stacked turbidite sandstones of Late Jurassic age with the Magnus Sandstone Member (MSM)
overlying the Lower Kimmeridge Clay Formation (LKCF) (MacGregor et al, 2003). A type log is shown in Figure 1(a). The
MSM is a high NTG (0.8 – 0.9) series of stacked high-density turbidites. Individual lobes are typically 2-7 m thick fine-coarse
sandstones with occasional interbedded shales. Units are named A-G from base to top and many units are fining-upwards cycles,
or show fine tops. Two shales are ubiquitous and act as regional seals, the B and F shales. Other shales are locally present and act
as vertical flow barriers. The underlying LKCF is a mud dominated low-density turbidite system with NTG nearer to 0.3; the
thickest sand unit is c.5m thick.
The total reservoir thickness is up to 200mtvt and the depth is 2800mss at the crest. The Original Oil Water Contact is at
3160mSS. The trap is a large tilted fault block dipping to the East, as illustrated in Figure 1(b). Faulting and stratigraphic
pinchouts have been demonstrated to have an impact on the performance of some flow units. STOIIP is in the order of 1.5 bstb of
39° API light sweet crude with bubble point pressure of 2600 psi and solution GOR 700 scf/stb. The MSM has the majority of the
2 SPE 134953

STOIIP (1.2 bstb) and has a higher expected recovery factor than the LKCF. Initial reservoir pressure was 6653 psi at datum depth
of 3050mSS.

Figure 1. (a) Well log type section through the Magnus reservoir, (b) Magnus Field Illustrative Cross Section

The field has a single steel jacket platform which originally had 20 well slots, 14 of which are used for production wells and 6
for injection wells. Recently 4 additional slots have been added to the jacket with the capability to drill 2 wells from each new
slot. Subsea wells have also been used over field life to provide additional well capacity; there are currently four active subsea
injection wells and one subsea production well from the South Magnus satellite field.
The original development plan was to waterflood the field, maintaining average reservoir pressure above bubble point to
maximise recovery. First oil was produced in 1983 through pre-drilled subsea production wells, the first platform producer came
on line in April 1984 and water injection commenced in July 1984.

Magnus Field Production (and Gas Injection) History Magnus Field Oil Production History and Major Events
200 200
Water Rate Precipitous decline with onset of
Oil Rate major Barium Sulphate scaling
Production and Injection Rates mboed

Gas Injection
Decline stabilised by activity:
150 150
Oil Production Rate mstbd

Well interventions / clean-out


New wells from platform slots
Side-track existing wells
Introduce gas lift completions
Additional subsea water injectors
100 100 Single zone completion policy

Commence gas
injection for EOR
50 50 Additional
well slots

0 0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010

1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010

Figure 2. (a) Magnus Production and Injection History, (b) Production History annotated with major events

Production History

Production climbed rapidly to plateau in 1985 and the field was constrained by facilities limits to approximately 140 mstbd until
the mid 1990s, Figure 2(a). At this time extensive sea water breakthrough occurred. The decline caused by increasing watercuts
was exacerbated by the formation of Barium Sulphate scale as a consequence of the mixture of sea water and formation.
Subsequently a period of extensive activity ensued to arrest the decline rate. Actions included well interventions, further new
wells and sidetracks, implementation of gas lift, introduction of additional water injection through subsea wells and a change from
multi-zone to single zone completions; these have been reported previously (Day et al, 1998, & Ravenscroft et al, 1996). These
SPE 134953 3

activities successfully slowed the decline rate, Figure 2(b). The focus of this paper is to consider the more recent post-plateau
period from 2002 onwards following the implementation of a gas injection EOR scheme.

Gas Supply for the EOR scheme

In 2002 Magnus commenced gas injection under a water-alternating-gas (WAG) process for EOR (Moulds et al, 2005). The
source of the injection gas was primarily from oil fields producing from the West of Shetlands (Schiehallion, Foinaven, Loyal and
now Clair). A major construction project was completed with new pipelines to transport the gas from the West of Shetland fields
to Sullom Voe in the Shetland Islands and then onwards to Magnus (Figure 3). The scheme had many attributes including
obtaining value for gas that would otherwise be stranded, reduced flaring at the West of Shetland fields, a supply of lean gas for
power generation at Sullom Voe and incremental oil recovery at Magnus.

M agnus Platform
20 inch 200 km -New pipeline riser
pipeline -Gas injection compressor
-LPG recovery
-6 w ell conversions
well

Sullom Voe Terminal


-Fuel gas to pow er
20 inch 200 km -Rich gas treatment & compression
pipeline -Fiscal metering

Shetland Islands
West of Shetland
-Subsea tie-
tie-ins
-Pipeline end manifold (PLEM )

Figure 3. Magnus EOR Scheme System Overview

The EOR process

For any tertiary EOR scheme to be effective the first step is to understand the habitat of the oil target that remains after secondary
recovery. For Magnus the secondary waterflood process leaves target oil for EOR in three areas: firstly waterflood residual oil,
secondly oil that has been by-passed during the waterflood and thirdly attic oil updip of production wells.
The waterflood of the Magnus MSM sands has been very effective in both areal sweep and at pore scale where the residual oil
saturation at typical reservoir pore volume throughput is about 25%. The residual oil saturation is not exceptionally high and it
presents a modest target for a miscible flood, although with 1.5 billion stbo originally in place, the prize from further reduction of
oil saturation in the waterflooded regions becomes substantial if carried out at field scale. The opportunity to develop a miscible
displacement is due to the favorable reservoir fluid properties. The light reservoir oil in Magnus combines with typical North Sea
export gas compositions to form a vaporising gas drive which gives rise to a multi-contact miscible displacement at pressures
above the minimum miscibility pressure (MMP) of 5000 psi. The vaporising drive mechanism results in the miscibility pressure
being independent of the injected gas composition; at least for the range of hydrocarbon gas compositions that are available to the
scheme. The early development work for the scheme had envisaged the reservoir being operated at lower pressure and the
construction project included provision to enrich the lean gas from the West of Shetland fields with additional propane and butane.
Today the target reservoir pressure is close to MMP. Greater dispersion occurs during field scale displacement compared to
laboratory slimtube experiments and gas enrichment helps to mitigate this which leads to some increase in oil recovery at field
scale if the gas was enriched prior to injection. However, the incremental oil benefit due to gas enrichment does not merit the
purchase cost of the enriching components, and indeed results in a net loss in boe terms, and so the imported gas is not enriched.
The aim of the EOR scheme is to achieve miscible displacement but it is useful to understand how the process may change in
the event that operational constraints cause fluctuation in the target reservoir pressure. In fact, the reservoir fluid properties are
such that even if the pressure falls below MMP there is still a significant benefit from injecting gas. The saturation pressure for
4 SPE 134953

Magnus oil is about 2600 psi and the target reservoir pressure in the EOR areas of the field is about 5000 psi. In the event that the
pressure falls below MMP the injected gas will continue to dissolve in the under-saturated oil with the result that it both swells the
oil and reduces its viscosity. Both oil swelling and the viscosity reduction become more effective as pressure increases, and with
multi-contacts between oil and gas phases. Nonetheless, for a substantial range of pressure below the target operating pressure
these effects alone can re-mobilise waterflood residual oil and enhance its capture.
In addition to the reduction in residual oil by miscible displacement there are also other recovery processes that occur as a
result of gas injection. The MSM sands in Magnus comprise repeated fining upward sequences. Waterflooding these sands leads
to a degree of water slumping which leaves unswept oil under the bed tops; this has been confirmed by many open-hole logs from
wells drilled through waterflood zones. The injected gas migrates upwards and underruns the bed tops and so mobilises the
bypassed oil.
If gas was injected continuously then it would quickly finger through the mobilised oil and would result in little incremental
recovery. In order to increase capture of the mobilised oil the scheme operates a water-alternating-gas injection scheme. The role
of the injected gas is to mobilise the residual and by-passed oil. The role of the water cycle is to create a more favorable mobility
ratio so that the oil mobilised during the gas cycle can be driven more effectively towards the production wells. Of course, during
the water cycle the mobilised oil will tend to resaturate areas previously swept out by gas and it may again become immobile. By
implementing water-alternating-gas injection this ensures that the mobilised oil is carried forward to the production wells through
successive WAG cycles.
The bypassed oil target and the favorably phase behaviour below MMP help to mitigate reservoir performance in the event
that, for operational reasons, the reservoir pressure cannot be maintained at or above MMP. In the event that the reservoir
pressure does fall below MMP there is a gradual falling away in recovery efficiency, but it is not precipitous.

Reservoir Performance under EOR

First gas injection for EOR was in October 2002. For reservoir management purposes the field in sub-divided into panels with
voidage replacement calculation applied to each panel. Four WAG injection wells were available at, or shortly after, the start of
gas injection, the well locations are shown in Figure 4(a). Recent 4D seismic response in Figure 4(b) illustrates the movement of
the injected gas away from the injection wells and towards the producers in the central and A3:B3 panels.

EOR Panels M 4 7 :A 3

A3B3 M SM

M 4 2 Y :B3
M 3 9 Z:B6

Central M SM

M 2 4 :B4
North
M SM
M 4 3 :B7

M 4 8 :C4

South M SM

Central LKCF M 3 4 :C3

Figure 4 (a) EOR patterns (b) 2001 to 2007 SNA map illustrating 4D response to injected gas around WAG injection wells

Two of the WAG injection wells are in the central panel, well M48:C4 injecting to the upper MSM sands and well M34:C3
injecting to the MSM-A sands. One WAG injection well, M42Y:B3 is in the A3:B3 fault block and well M45:C5 is in the
SPE 134953 5

southern panel. In 2006 a further WAG injection well, M53:C2, was drilled to the northern panel. This well encountered a high
pressure unit at the top of the reservoir which, due to poor cement isolation, continues to raise the bottom hole pressure above the
maximum that can be applied to initiate gas injection and the well has remained on water injection. In 2009 well M38:C6 was
converted from water injection to WAG injection service and was intended to be the first dedicated WAG injector in the LKCF
reservoir. Unfortunately the change from water to gas injection exposed a poor cement bond and the injected gas immediately
travelled behind the liner and entered the upper, better quality, MSM reservoir and was produced at the nearby MSM production
well M24:B4. Well M38:C6 is due to be sidetracked in 2011 to complete its objective of initiating WAG injection to the LKCF
reservoir. The inability to inject gas to either M53:C2 or M38:C6 meant that until early 2009, when an additional EOR production
well, M56ZE8, was drilled in the southern panel, the original four WAG patterns have remained constant. The WAG injection
schedule over the life of the EOR scheme is shown in Figure 5.

2002 2003 2004 2005 2006 2007 2008 2009

B3 M42y

C3 M34

C4 M48

C5 M45

C6 M38

GAS INJECTION WATER INJECTION

Figure 5 Gas and Water Injection Cycles by Well

The upper MSM central panel contains one WAG injection well and two EOR production wells. This is the largest EOR panel
and has the thickest MSM section in the field, this panel has received the largest volume of injected gas 57 bcf. The nearest
production well to the WAG injection location is M24:B4 which is 1.5 km from the injector; this is also the highest gross
production rate well on the field. Gas breakthrough at well M24:B4 took 12 months. The second production well in the panel,
M39Z:B6, is further updip and to some extent in the shadow of well M24B4. Gas breakthrough took 16 months from first gas
injection. Both of the production wells were on stream and had seen water breakthrough prior to the start of WAG injection. This
presents the opportunity to fit a waterflood decline curve to the wells based on the extensive knowledge of waterflood decline
characteristics for the field acquired over the previous fifteen years and allows good estimation of the oil recovered through the
wells that is in excessive of the expected oil production from waterflood alone, see Figure 6. These patterns are now mature with 5
gas injection cycles at well M48:C4. These two wells have produced incremental oil production due to EOR, i.e. above the
watercut decline curve, of 7.1 mmstb.

M24:B4 EOR Oil Production M39Z:B6 EOR Oil Production


8 8

6 6
Oil Rate, stbopd
Oil Rate, stbopd

4 4

2 2

0 0
2003 2004 2005 2006 2007 2008 2009 2010 2003 2004 2005 2006 2007 2008 2009 2010

Figure 6. Oil production from EOR wells M24B4 and M39Z:B6 shown with waterflood decline curve
6 SPE 134953

The MSM-A in the central panel is developed with one WAG injection well, M34:C3, and one production well, M43:B7. The
volume of gas injected is 30 bcf although the injection well is completed in both the upper MSM and the MSM-A sands so there is
uncertainty on how much of the gas is entering each zone. PLT logging indicates that about 30% of the injected gas is entering the
upper-MSM and 70% to the MSM-A sands. The production well also has a small perforated interval in the upper MSM sands
although it is thought that most of the EOR recovery, about 1.9 mmstbo, is due to sweep of the MSM-A sands.
The A3:B3 fault has a single injector producer pair. It is the most mature of the WAG patterns and has the highest ratio of
injected gas to pore volume. Communication between the A3:B3 fault block and the main field is baffled which helps to reduce
uncertainty when characterizing the EOR performance. The distance from the injection well to the producer is 1.2 km and gas
breakthrough occurred after 6 months. This pattern is also advantaged through having the injector slightly updip of the producer.
The recovery from this area has been very good with 2.6 mmstbo of EOR oil recovered at a gross gas efficiency of 9.6 mscf/stb
and a net usage of 2.5 mscf/stb. This indicates a very efficient reservoir process with about 1.1 reservoir barrels of gas left in the
reservoir for every reservoir barrel of oil recovered.
These three areas have each seen at least 4 WAG cycles and are relatively mature so it is appropriate to consider the overall
efficiency that has been achieved for the EOR process. This is done by combining the results from the three areas as this helps to
reduce the uncertainty associated with metering and allocation errors and to remove dependence on zonal splits from PLT
interpretation. In total these three areas have injected 112 bcf, have produced 11.5 mmstb of EOR oil (i.e. oil that lies above the
waterflood decline curve). The gross gas usage to deliver the EOR oil is 9.8 mscf/stb and the net usage is 3.5 mscf/stb; in term of
reservoir efficiency about 1.5 rb of gas is left in the reservoir for 1 rb of incremental oil produced.
The EOR development of the southern panel is considerably less mature than the other areas. Gas injection commenced at well
M45:C5 at the start of the EOR scheme in 2002 at the time when there were two associated production wells. However, the first of
these production wells was sidetracked within a few months to an LKCF target and the second well M29:A7 developed an Annular
Safety Valve problem that prevented gas lift operation and so the well was limited to low offtake rates. Well M45:C5 continued
to follow a WAG injection in anticipation of further production wells becoming available. This took longer than anticipated with
the first new production well in the panel, well M56Z:E8, completed in 2009 and a further production well M60:A6 expected to be
completed in 2010. The results from these wells have been very encouraging (see section on New Well Results). However, prior
to the new wells the flow regime in the southern panel has been largely quiescent and very different from the other EOR areas.
This produces different challenges to prediction of EOR performance and these are considered in the following section that
discusses the full field dynamic simulation model.

60 EOR contribution to Magnus Main Field Oil Production


80%
non-EOR Oil Production
50 EOR Oil Production EOR contribution

60%
40

30
40%

20

20%
10

0 0%
2003 2004 2005 2006 2007 2008 2009 2010 2003 2004 2005 2006 2007 2008 2009 2010

Figure 7 (a) EOR oil contribution to Magnus Main Field Oil Production 2002 to 2010; (b) EOR oil percentage contribution

In the seven years since EOR started it has become an increasingly important contributor to the field production. Figure 7(a)
shows the field oil rate associated with EOR and with non-EOR (secondary waterflood) recovery processes. In these figures the
data up to early 2009 are derived solely from the difference between established waterflood decline and actual performance in the
EOR production wells. From 2009 onwards the EOR rate also includes all production from the new well M56Z:E8 which
produced at an elevated GOR from start-up. This well does not have any waterflood only history to anchor a waterflood decline
curve and so is not directly comparable with the oil production rate based solely on EOR mechanisms which were being tracked
for original EOR production wells. Figure 7(b) shows that with the expanded definition of EOR oil to include all oil from any
new wells drilled as part of the EOR scheme the percentage contribution of EOR has increased throughout the period until it now
accounts for about 40% of the field production rate.
SPE 134953 7

Full Field Model

The primary tool for setting future development strategy is the full field model (FFM). The construction of the FFM was described
in an earlier paper (Moulds et al, 2005). The model is a finite difference simulator, the grid cells are 100m*100m with a nominal
thickness of 4 m, fluids are described by a 6 component Peng-Robinson equation of state and the model contains about 102k active
cells. Relative permeability curves were developed through an upscaling study. Both the pseudo relative permeability generation
and original history match were assisted by utilizing genetic algorithm techniques in BP’s TDRM™ technology (Williams et al,
2004). Five years after its construction and seven years after the start of EOR operations it is a good time to ask whether the
choice of model was appropriate.
The FFM has been updated periodically for production history and where new well information has indicated some local
revisions to the grid structure and properties have been applied. The FFM has proved to be a robust tool. It has been simple to
maintain the history match and has provided good predictions for both current wells and when new wells have been added. The
current history match at field level for all fluids is shown in Figures 8(a) and 8(b). The errors at field scale are 0.8 mmstb (0.1%)
for oil, 1.1 bcf (0.2%) for gas and 11 mmbw (2%) for water. At individual well level there is more variation but errors remain
acceptable.

Figure 8. FFM history match to July 2009, (a) cumulative oil, gas and water production; (b) watercut and GOR

The FFM match appears to be good. However, for the model to be appropriate for EOR prediction requires a higher level of
accuracy than may be required for waterflood. The match to EOR incremental oil and returned injection gas are key indicators of
whether the FFM captures the EOR process sufficiently well that it is an appropriate tool for EOR prediction. Figure 9 shows the
FFM match to EOR performance for the current version of the model (2009).

14 100
FFM : summed EOR production w ells FFM : summed EOR production w ells
Incremental Oil (mmstb)

12 Surveillance Surveillance
80
Returned Gas (bcf)

10

8 60

6
40
4
2 20

0
0
0 20 40 60 80 100 120 140
0 20 40 60 80 100 120 140
Gas Injected (bcf)
Gas Injected (bcf)

Figure 9. FFM history match to July 2009 for Incremental EOR oil and Returned Injection Gas
8 SPE 134953

Relative Permeability in the FFM


During the model updates from 2005 to 2009 the relative permeability curves have required little adjustment to maintain the
history match. Figure 10 shows the relative permeability curves that were used in the original construction of the model and those
that are in the current model. The Krow, Krwo, Krog and Krg curves are almost unchanged, only the Krwg curve has been revised
so as to provide a small improvement to the returned gas match (the 2005 curve produces slightly better match to incremental oil).

1.0 1.0 1.0

KRG 2005
KRWG 2005
KRG 2009
0.8 0.8 0.8 KRWG 2009
KROG 2005

RELATIVE PERMEABILITY

RELATIVE PERMEABILITY
RELATIVE PERMEABILITY

KROG 2009

0.6 0.6 0.6


KRW 2005
KRW 2009
KROW 2005
0.4 0.4 0.4
KROW 2009

0.2 0.2 0.2

0.0 0.0 0.0


0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
WATER SATURATION GAS SATURATION GAS SATURATION

Figure 10. Relative Permeability curves in 2005 and 2009: (a) Krow and Krw, (b) Krog and Krg and (c) Krwg

These curves have been applied globally in the FFM with the exception of two areas. In the southern panel the injection well
M45:C5 continued to follow a WAG injection cycle for several years during a period when there was no significant production
from the panel. Cross-flow into the central panel did occur but flux rates were low and the viscous-gravity ratio quite different
from the other panels. The pseudo relative permeability curves that successfully match EOR performance in the active panels
were not generated to match the quiescent flow conditions in the southern panel. Two pieces of evidence demonstrate that flow
conditions in the southern panel were outside the bounds for which the pseudo relative permeability curves are applicable. The
first evidence was that the production well M56Z:E8 drilled in 2009 produced oil at an elevated GOR of 2000 scf/stb from start-up
(solution gas is 700 scf/stb) but the FFM predicted that gas had not reached the well location. The second evidence was from a 4D
seismic survey that indicated that gas in the southern panel had migrated further than indicated by the FFM. The match to well
M56Z:E8 was re-established by adjusting the gas relative permeability curve to be linear between its end-points during the period
prior to 2009. The linearization of the Krg curve in the southern panel raises the gas mobility at low gas saturation which enables
more rapid vertical segregation of gas and mimics the movement of a thin zone of gas under the zone tops. It is also consistent
with the gas saturations being less disturbed by the water injection cycles given that the gas is already at the top of each zone.

1.0
reduced gas mobility
usual Krg

0.8
RELATIVE PERMEABILITY

0.6

0.4

0.2

0.0
0.0 0.2 0.4 0.6 0.8 1.0
GAS SATURATION

Figure 11. Impact of Reduced Gas Mobility on History Match to Cumulative Oil Production and GOR for successive WAG cycles

A second observation concerning relative permeability is that the EOR performance of the A3:B3 fault block appears to be
consistent with a reduction in gas mobility during successive WAG cycles, such as has been proposed by (Larson & Skauge,
SPE 134953 9

1998). Although the match to the first gas cycle in the A3:B3 fault block is well matched it is apparent that subsequent cycles are
less well matched. The FFM does not support cycle dependent relative permeabilities but a sensitivity was run where the gas
mobility at higher gas saturations was much reduced. The results in Figure 11 show that the reduced gas mobility improves the
match to both returned gas and recovered oil during the later gas cycles. There may be other explanations for this behaviour,
perhaps related to flux between the A3:B3 fault block and the main field, but the results appear to be directionally supportive of
reducing gas mobility during successive WAG cycles.

Operational Challenges

Operating a tertiary EOR WAG injection scheme is inherently more complex than operating a primary or secondary recovery
process. This can be illustrated by considering the number of systems that need to be aligned and the rapidity with which they
interact with each other. For example the gas volume available for injection needs to be aligned with the number of wells that are
lined up to inject gas rather than water, these wells in turn need to be operated on a staggered schedule so that the gas travelling
through the reservoir does not all arrive back at the platform at the same time which would lead to cutting back production due to
produced gas handling constraints. In the case that excessive produced gas does lead to a production cut-back then in order to
remain within the target tolerance for reservoir pressure this forces a reduction in the injected gas rate which in turn forces a
reduction in the amount of import gas that can be accepted by the platform for injection.
The high level of inter-dependence between sub-systems means that EOR operations can quickly become sub-optimal if the
early warning signs of poor performance are not recognized. The situation is exacerbated if intervention is delayed. Magnus is an
aging offshore production platform located in a harsh environment that requires a large commitment to maintaining platform
integrity, and which has an active drilling programme. Under these operating conditions then both early recognition of problem
behaviour and the ability to intervene present significant challenges.

WAG changeovers
The EOR scheme has operated with 4 WAG injection wells since 2002. The target operation has been to operate 2 wells on
gas injection with the other 2 wells on water injection. The scheme has operated at nominal WAG ratio of one injecting equal
reservoir volumes of water and gas each cycle. The operation to changeover from water to gas injection, and vice versa, is a 3 day
operation that involves physical removal of the injection line for one phase to be replaced by the hook-up of the injection line for
the other phase followed by tubing displacement with solvent to eliminate hydrate problems. These changeovers need to be
scheduled in advance to build into the platform activity plans and, unfortunately, because they are perceived as small activities it is
very easy for those planning the platform activity to consider that there is little cost to delaying them. In the early years of the
EOR scheme when the patterns were less mature the target of 2 wells injecting gas and 2 wells injecting water was maintained
reasonably well, see Figure 5. As the patterns have matured the injected gas travels through the reservoir more quickly and the
ability of the system to absorb unexpected events is reduced.

Oil Loss due to Plant Gas Constraint

10
9
8 Oil losses
7 predominantly
Oil Loss, mbd

6 from A3:B3 gas


5 storage panel
4
3
2
1
0
Jan 09 M ar 09 M ay 09 Jul 09 Sep 09 Nov 09 Jan 10

Figure 12. Oil losses associated with Produced Gas constraint

The challenge of maintaining appropriate WAG ratios is illustrated by events in late 2008 and early 2009 when one of the
WAG injection wells, M45:C5, ceased any injection in order to manage the reservoir pressure in the southern panel prior to
drilling well M56Z:E8. In order to maintain field gas injection rates the other three WAG injection wells were taking more gas
and, consequently, less water. This change to the effective WAG ratio was already causing some challenge to the gas handling
plant when gas injection to the LKCF well M38:C6 was initiated and found to be cycling very quickly to an MSM production well.
10 SPE 134953

This exceeded the ability of the platform to manage the volume of returned gas and production from all EOR production wells was
cut-back. Figure 12 shows the oil losses associated with the production cut-backs. The imbalance between gas and water injection
volumes was remedied by ceasing all gas injection for three months apart from retaining the ability to inject to the A3:B3 block
which was required for operational safety reasons. The remedy was successful as shown by recovery from the oil losses that were
occurring in early 2009. The experience showed that the EOR scheme can accommodate only a limited degree of delay to routine
operations; it has also elevated the priority given to WAG changeovers when planning platform activity.

Metering
Accurate metering and allocation of gas volumes into and out of the reservoir is important to maintain smooth operations and
efficient WAG scheduling and to determine the efficiency of the EOR process. The EOR project had included a requirement that
all wells that inject gas should have individual meters and also that the test separator should be capable of accurate gas
measurements. In practice, retro-fitting meters to the gas injection wells was challenging as it competed for offshore time
alongside the Magnus Extension Project (MEP), which was a major project to construct addition drilling slots, and later against
drilling activity. The installation of meters also faced technical issues because the corrosion protection on the internal surface of
the flowlines created difficulty with sonic meter readings. A solution was found by locating the meters at a non-optimum location
on the flowline, away from a long straight flow section, which although having decreased meter accuracy was adequate to manage
reservoir pressure safely and to estimate the EOR efficiency for the different patterns.

Reservoir Pressure Management


Reservoir pressure within the EOR panels is managed with a view to maximising recovery through miscible displacement by
maintaining average pressure at the MMP of 5000 psi. However, the pressure rating for the EOR production wellheads is 5000 psi
which places an upper limit on the reservoir pressure. After allowing for safety margins and worst case conditions in which the
production tubing would be full of the lightest gas in the system, i.e. injection gas, the combination of MMP target and wellhead
pressure rating create a narrow target range. As a consequence of metering and allocation uncertainty there is a requirement to
routinely monitor the reservoir pressure in the EOR panels. The newer wells that are currently being drilled are equipped with
continuous read-out downhole gauges, but the older wells which date back to secondary waterflood operations do not have
downhole gauges. In the smaller pore volume panels which are most susceptible to pressure fluctuation due to metering error the
bottom hole pressure is recorded by gauge every three months and in the larger panels every six months. This policy is strictly
enforced and in the event that pressures are not obtained on schedule then the voidage replacement to those panels is severely
restricted until pressures are again validated by downhole gauge.

Reservoir Management Tools


In response to the operational challenges new tools have been developed to assist the management of the EOR scheme.

1000
Returned Injection Gas Prediction
monthly gas volumes mrb

800 Tank Model

600

400

200

0
Jan 02 Jan 03 Jan 04 Jan 05 Jan 06 Jan 07 Jan 08 Jan 09 Jan 10

Figure 13. Gas Tank Model for A3:B3 panel showing match to historic data and prediction usage

The first of these is a tool built to assist the short term prediction of returned gas volumes that will arrive back at the platform.
This is the returned gas Tank Model which combines a rudimentary material balance constraint with an empirical match to historic
performance. This tool is based on tank model methodology (Wingard and Redman, 1994) which has found much use in
estimating EOR performance in very large fields with many tens of WAG patterns such as Prudhoe Bay in Alaska. Although
Magnus has a well matched fully compositional full field model its’ primary role is to build robust field development plans to be
applied over many years into the future; it is not an ideal tool for short term forecasting over a period of a few months. The
SPE 134953 11

returned gas tank model is tuned to each pattern using just three parameters: the system volume (related to pattern pore volume),
the number of tanks in the models and the recovery factor for the gas. The model is then history matched against previous gas
cycles in an EOR pattern and provides a short term forecast of returned gas volumes based on near-term gas injection volumes and
gross fluid off-take at the associated production wells. Figure 13 shows both the history match and prediction phases for the
A3:B3 panel.
The second tool is an EOR dashboard that is continuously updated with injection and production volumes for each EOR panel.
A series of key plots in the dashboard are the volume of gas injected to each WAG injection well and a colour coded test of
whether it has exceeded the injected water volume immediately preceding it, which would indicate that the pattern has exceeded
the amount of gas appropriate to its target WAG ratio of one. Figure 14 shows this plot for the M48:C4 WAG injection well. This
visual descriptor is very helpful both to the production operations team who are scheduling WAG changeovers and also as an aid
to those who are involved with EOR operations but may not have an in-depth of knowledge of what is required to maintain EOR
delivery.

Figure 14. WAG cycle tracking (red indicates excessive gas injection)

Gas supply management


The EOR scheme on Magnus depends on imported gas from producing oil fields. Operational issues at these other fields lead
to a variable gas supply at Magnus. If the gas supply was more uniform and more predictable it would be possible to achieve more
efficient sweep through the EOR panels. Now that some of the EOR panels are very mature there is an opportunity to use one of
these panels to store gas and provide a more secure supply that could supplement the imported gas and smooth out the gas
available for injection. The A3:B3 panel is the most mature EOR pattern and is in baffled communication to the main field so is
unlikely to interfere with main field performance. This panel has been used for the last year to provide Magnus with its own gas
storage facility. Also, as it has been maintained at relatively low pressure it can accept injection gas when this is required for
operational purposes, such as at WAG changeovers in other panels. The operation of the A3:B3 fault block for gas storage
initially created some stability issues for the plant, as shown in Figure 12, but these are now resolved and the gas is used to
supplement the imported gas volumes without causing plant instability or oil losses.

New Wells Drilling

The major themes for reservoir development in Magnus at this mature stage of its field life are optimisation of the water-
alternating-gas Enhanced Oil Recovery (EOR) scheme, exploitation of satellites and peripheral areas of the field and continuing
development of the low net-to-gross LKCF reservoir. The MEP major project was sanctioned to support these field development
12 SPE 134953

objectives with the installation of 4 additional slots on the east side of the Magnus jacket. Each of these slots uses conductor
sharing wellhead technology which allows two wells from each new slot (Hicks et al, 2009). A combined subsurface, drilling and
completions team successfully installed the world’s first dual unguided 18-5/8” slot splitter in 2008. This was the first time that
two 18-5/8” casing strings had been installed inside a conductor without structural guides. The first well from the new slots, an
EOR producer, was delivered in early 2009. Further wells from the new slots have been drilled including two to North West
Magnus, a satellite accumulation, over 6.9km from the platform, the injection well being 7.6km which is longest well drilled from
the Magnus platform. To April 2010 five new wells have been drilled. The results from these wells have addressed all aspects of
maintaining production from the Magnus Field and all five wells delivered to, or exceeded, the pre-drill expectations.

New Well Targets and Results

The results from the first well to be drilled from the new slots, well M58Z:E8, are of particular importance to opening up further
targets for the EOR scheme. This well targeted a crestal area of southern panel in the main field; its principal objective was to
provide an effective off-take point for the WAG injection at well M45:C5, and so re-invigorate the EOR development, although
with an upside that it may also exploit attic oil along the crest of the structure. It was 10 years since any wells had been active in
the neighbourhood of the target location and those wells had been abandoned at high watercuts in excess of 95%. Also the first
gas injection at well M45:C5 had occurred as early as 2002, so there was uncertainty as to the effect that the injected gas may have
had and whether it may have caused any re-distribution of oil after the waterflood.
The timing of well M56Z:E8 coincided with interpretation of 4D seismic surveys. The crestal edge of the MSM sands, which
is the target location for well M56Z:E8, has been a continuing challenge to seismic interpretation throughout Magnus development
because the thinning sands are below seismic resolution. There are three seismic surveys on Magnus which are available to
evaluate 4D response; these were acquired in 1992, 2002 and 2007. The first two 4D surveys were acquired either side of a major
re-pressurising of the field. The pressure impact is clearly visible on the sum of negative amplitudes response and this very
effectively illuminates the edge of the MSM sands in the Southern Panel, Figure 15(a).

M56z(E8)
M45:C5
Pressure illuminating WAG
edge of fairway Injection

Abandoned
wells:

water
injection

production

appraisal

possible M29:A7
unswept oil production
updip of
historical
producers
01d92 m110-m150 SNA significance

Figure 15. (a) Seismic 4D interpretation over the Southern Panel; (b) toe-thrust play-type

The location for well M56Z:E8 was selected following a review of the possible target locations for bypassed oil. Improved
techniques for seismic visualization are being combined with an updated geological model to produce a better understanding of
remaining oil locations. A number of 2D seismic lines stretching from the Magnus structure towards the East have been
interpreted to find indications for the extent of the Magnus low angle slump faults (slides) that occur within the MSM sands. The
top of the slides can be mapped on the Eastern slope of Magnus structure where a slide scarp is exposed but also overprinted by
erosion of the base Cretaceous. The slide toes can be mapped down the flank of the structure where they show expressions of
thrusting. Additionally some internal heterogeneity can be seen in the seismic data.
The potential for bypassed oil arises from the possibility of creating isolated reservoir compartments. These compartments are
the result of post depositional alteration of the turbidites in the Magnus field. By definition the turbidites are deposited in deep
water environments and contain a significant volume of water. During lithogenesis the interbedded shales between the reservoir
SPE 134953 13

sandstones compact and de-water, the water migrates into the reservoir sandstones and the pore pressure of the sandstones
increases. The Magnus field is inclined at eight degrees and immediately post the deposition of the Magnus Sandstone Member
(MSM) in the early Cretaceous there was significant tectonic activity associated with the opening of the North Atlantic to the west.
The tectonic activity in conjunction with the slope of eight degrees and over pressured sandstones interbedded with impermeable
shale intervals created the ideal conditions for slides and slumps. Across the Magnus field the expression of slides and slumps are
mappable on 2D and 3D seismic surveys on the Base Cretaceous horizon. The slides and slumps are characterised by extensional
scarps at the back where reservoir material has slumped and migrated downslope, and compressional mounds at the foot of the
slope where the material has coalesced, see Figure 15(b). It is the coalesced mounds that have provided the targets for the recent
Magnus drilling activity M56Z:E8 and M60:A6 and have provided the additional benefit of dry attic oil in addition to the predicted
EOR volumes. Although this play can be lucrative it can also be problematic. The danger with this type of structural play in
Magnus is that where there is a local water injection support well in close proximity to such a slump target there is the possibility
that the reservoir compartment can be isolated and over pressured by the injector, which can lead to drilling problems, and if the
well is a WAG injector, an inability to inject gas as in M53:C2.
Both well E56Z:E8 and the later production well to the southern panel, well M60:A6, target this “toe-thrust” play-type. Well
M56Z:E8 found 35m of net sand and mainly dry oil and confirmed the understanding of the target type. When first brought on-
line the well was found to be producing at an elevated GOR of 2000 mscf/stb compared to the solution gas GOR of 700 mscf/stb.
The elevated GOR is due to the gas injected at well M45:C5 which has migrated updip. The well sustained an average rate of
about 6 mstb/d over the first year with good pressure support being maintained from the WAG injector. Importantly, the well
demonstrated that this play type offers the early benefit of dry oil supported in the longer term by larger volumes from the EOR
flood.
The overall delivery from the first five wells after drilling resumed has been very successful. The five wells comprise two
wells to the North West Magnus satellite accumulation, an LKCF production well and two production wells in the southern panel
that combine bypassed oil with EOR delivery from toe-thrust targets. The impact of these wells on the platform oil production
profile which includes production from satellites South Magnus and North-West Magnus is shown in Figure 16.

M56Z:E8 M57Z:E7 M58Z:E3 M59Z:E4


30
Magnus Platform Oil Rate (mstb/d)

20

10

0
Sep-08 Jan-09 May-09 Sep-09 Jan-10

th
Figure 16. Impact of first four wells from new slots on Magnus Platform Oil rate (5 well M60:A6 scheduled on-line April 2010)

Future Development

The success of the first five wells completed since drilling resumed in late 2008, together with the demonstrated delivery from
the EOR recovery mechanism, has given confidence that there is substantial opportunity remaining in the field. Eight additional
targets which continue the reservoir development themes of EOR optimization, drilling to peripheral areas of the reservoir and
further development of the LKCF reservoir are to be drilled over the next few years. Each of the proposed targets builds on the
learnings from the first five wells with improved understanding of play-type supported by improved target visualisation and by
ERD experience in the case of the longer step-out wells. The location of the first five wells and the future targets are shown on the
field map in Figure 17.
14 SPE 134953

1 M 56Z:E8 Southern panel EOR producer


4 7
2 M 57Z:E7 North-West M agnus producer
3 M 58Z:E3 LKCF reservoir producer
2
4 M 59Z:E4 North-West M agnus injector
9 5 M 60:A6 Southern panel EOR producer
future target
12
12

3
5

1 6
6

10
10
8
8a
11
11
13
8b

Figure 17: Location of first five wells (post resumption of drilling in 2009) and future targets

Further opportunities are also being studied and additional data needs to support these are being defined. The recognition of a
play-type that combines initially dry oil with longer term EOR volumes has focused attention on target imaging. Further seismic
acquisition is being considered, both to capture 4D images for gas sweep and also to improve resolution of the static image through
ocean-bottom-cable (OBC) seismic.
For further development drilling to be economic the lessons learnt to date will be invaluable, as will the continuing updating of
technology. New “slot-splitter” drilling technology has been game changing for field development; drilling performance will need
to continue to be innovative as smaller pools are targeted. In Operations, the understanding of the key role that surveillance and
accurate metering play in maintaining stable and reliable EOR operations needs to remain embedded in the way the field operates.
This will be assisted as new wells will be equipped with permanent downhole gauges.
The prize from the EOR scheme is reinforced with each of the recent wells (M57Z:E7, M58Z:E3 and M60:A6) that have
drilled through the MSM sands. In each case these wells have confirmed, see Figure 18, the presence of either an updip attic target
or unswept oil underneath several shales which both offer attractive oil targets for WAG injection. Opportunities for acquiring
core data from gas swept zone are being assessed, as was done on the Ula field (Thomas et al, 2008), and this may lead to further
opportunity for the EOR scheme if gas sweep conformance away from the WAG injection well locations can be more fully
characterized.

M 56Z:E8 M 60:A6 M 58Z:E3


Southern panel Southern panel
producer producer M SM sands
show n here
although w ell
drilled and
completed as
an LKCF
producer

Figure 18. Recent wells drilled through MSM sands illustrating EOR target oil remaining under shales after waterflood
SPE 134953 15

People

Magnus operations are ambitious. The platform operates a complex tertiary EOR scheme whilst simultaneously carrying out
extended reach drilling and developing multiple reservoirs. The Magnus platform itself requires significant maintenance and a
rejuvenation project is underway, after almost thirty years of producing oil in the harsh North Sea environment. In order to
succeed in these plans whilst working within the tight personal-on-board (PoB) constraint requires very close co-operation
between many teams. BP has recently moved to a functionally based organization, the organization outline having been developed
and road tested in Aberdeen with Magnus operations being a significant component. The success that has been achieved on
Magnus has made great use of the in-depth technical knowledge that can be harnessed through a functionally based organization
whilst also creating a shared vision, between different functional teams, as to what success looks like.

Conclusions

• The tertiary WAG EOR scheme on Magnus is now well characterized and has delivered sufficient volumes of oil that it can be
seen to have changed both field life expectations and created the opportunity for further reservoir development.
• Operating a tertiary WAG EOR scheme is very much more complex than primary or secondary recovery; success requires a
greater in-depth knowledge within each function including reservoir, production operations, drilling and safety.
• The operational complexity for a tertiary WAG EOR scheme is not always easily communicated outside of the immediate
team and additional visual tools to assist discussion are presented in this paper.
• The compositional full field model has been shown to provide a good tool for predicting performance several years forward;
for shorter term forecasting a simpler tank model, which is easier to tune to recent well performance, is more appropriate.
• The combination of a tertiary EOR gas injection scheme together with ERD drilling and multiple reservoir developments
requires close co-operation between the different functional teams. Delivery in Magnus has been enhanced by the transition
to a functionally based organization that also recognizes the need to work effectively across the functions.

Nomenclature
Krow Relative permeability of oil to water
Krwo Relative permeability of water to oil
Krog Relative permeability of oil to gas
Krg Relative permeability of gas
Krwg Relative permeability of water to gas

Acknowledgement
The authors would like to thank the management of BP plc and the Magnus partners: Nippon Oil Exploration and Production
(NOEPUK), Eni (UK) Ltd and Marubeni North Sea Ltd for permission to publish this paper. The observations and opinions
expressed in this paper are those of the authors and are not necessarily shared by BP plc, the Magnus Field operator, or its partners.

References
1 Day, S., Griffin, T. and Martins, P. 1998. Redevelopment and Management of the Magnus Field for Post-Plateau Production. Paper
SPE 49130 presented at SPE Annual Technical Conference and Exhibition, 27-30 September, 1998, New Orleans, Louisiana
2 Hicks, S.E., Moore, A., Honey M., Farmer I.R., Smart, B., Ekseth, R. and Brown, D. 2009. Magnus: Utilisation of Conductor Sharing
Wellhead Technology to Access Additional Hydrocarbons via a Slot-Constrained Platform. Paper SPE 124233 presented at 2009 SPE
Offchore Europe Ol & Gas Conference and Exhibition, Aberdeen, 8-11 September 2009.
3 Larsen, J.A. and Skauge, A., 1998. Methodology for Numerical Simulation With Cycle-Dependent Relative Permeabilities. SPE
Journal, June 1998.
4 MacGregor, A.G., Trussell, P., Lauver, S., Bedrock, M., Bryce, J., and Moulds, T.P. 2003. The Magnus Field – Extending Field Life
through Good Reservoir Management and Enhanced Oil Recovery. Presented at North-West Europe and Global Perspectives - 6th
Petroleum Geology Conference, London, October 6-9, 2003.
5 Moulds T.P., Trussell, P., Haseldonckx, S.A. and Carruthers, R.A. 2005. Magnus Field: Reservoir Management in a Mature Field
combining Waterflood, EOR and New Area Developments. Paper SPE 96292 presented at Offshore Europe 2005, Aberdeen, Scotland,
U.K., 6–9 September 2005.
6 Ravenscroft, P.D., Cowie and L.G., Smith, P.S. 1996. Magnus Scale Inhibitor Squeeze Treatments - A Case History. Paper SPE 36612
presented at SPE Annual Technical Conference and Exhibition, 6-9 October, 1996, Denver, Colorado.
7 Thomas, S., Duncan, J., Haajizadeh, M. and Williams, J. 2008. Ula Field-Life after the Flood: Core & Log Experience from behind a
maturing WAG front. Presented at SPWLA 49th Annual Logging Symposium held in Edinburgh, Scotland, May 25-28, 2008.
8 Wingard J.S. and Redman R.S. 1994. A Full-Field Forecasting Tool for the Combined Water/Miscible Gas Flood at Prudhoe Bay.
Paper SPE 28632 presented at SPE 69th Annual Technical Conference and Exhibition, New Orleans, U.S.A., 25-28 September, 1994.
9 Williams, G.J.J., Mansfield, M., MacDonald D.G. and Bush, M.D.: 2004, Top-Down Reservoir Modelling. Paper SPE 89974 presented
at SPE Annual Technical Conference and Exhibition held in Houston, Texas, U.S.A., 26-29 September 2004

You might also like