Signe Berg Verlo and Mari Hetland Development of a field case with real production and 4D data from the Norne Field as a benchmark case for future reservoir simulation model testing Trondheim, 06.06.2008 Master thesis NTNU Norwegian University of Science and Technology Faculty of Engineering Science and Technology Department of Petroleum Engineering and Applied Geophysics
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Signe Berg Verlo and Mari Hetland
Development of a field
case with real production
and 4D data from the
Norne Field as a
benchmark case for future
reservoir simulation model
testing
Trondheim, 06.06.2008
Master thesis
NTNU
Norwegian University of Science and Technology
Faculty of Engineering Science and Technology
Department of Petroleum Engineering
and Applied Geophysics
Preface
The work presented in this Master thesis was conducted in the 10th semester of the Master
of Science studies at NTNU. It was written at the Department of Petroleum Technology and
Applied Geophysics, spring 2008. The work was prepared by the authors with Professor Jon
Kleppe as academic adviser.
We would like to express our gratitude to Professor Jon Kleppe for guidance and advice
throughout the thesis work. We are also very grateful for the help we have received from
StatoilHydro by Anna Fawke, Kristin Seim and Trine Alsos. Finally we would like to thank
Stein Krogstad from Sintef and Alexey Stovas, Jan Ivar Jensen, Knut Backe, Bjarne Foss and
Egil Tjåland at NTNU for their help and support.
Trondheim, 6. June 2008
Mari Hetland Signe Berg Verlo
I
Abstract
Reservoir simulation models are essential tools for the development of oil and gas elds. These
realistic models are used for calculating reservoir volumes, for well planning, and to predict
future behaviour of elds. Building and maintenance of robust, reliable reservoir models are
time-consuming and expensive. Research based on simulation models could improve existing
methods and tools utilized for this work. One of the objectives of the research program 2 in the
Center for Integrated Operations in the Petroleum Industry (IO Center) at NTNU is to develop
methods for rapid updates of the reservoir model or geological model for petroleum elds, based
on 4D seismics, production data and other available data. As a pilot case, the Norne Field is
selected.
It is of great importance to have a real model which is open for several research institutions,
to compare various methods used on the same set of data. Currently, there exists no model with
real data. Therefore, NTNU in collaboration with StatoilHydro wish to establish a model based
on the Norne Field.
The Norne eld is located in the Norwegian Sea. It was discovered in December 1991 and
started producing November 1997. The Norne reservoir rocks were deposited from Late Triassic
to Middle Jurassic. Petrophysical results from the Norne Field are mainly based on the results
from exploration wells. A total of 49 wells are drilled, 3 exploration wells and 46 production and
injection wells. Seismic surveys were acquired in 2001, 2003, 2004 and 2006. These surveys have
good quality and 4D seismics has been extracted. The base case simulation model is history
matched until December 2006, and predicts reservoir development until January 2022. The
model is run in Eclipse 100, which is a standard black oil simulator, and input can easily be
converted for use in other reservoir simulators.
The objective of this master thesis has been to shape a reservoir model using real data from
the Norne Field. The assignment emphasizes the design of a benchmark case for research, and
focus on the utility value of a model with real data, open for several research communities.
A number of possible cases can be designed when constructing a benchmark case. The Norne
benchmark case should exploit the good quality data which is available, and promote comparative
studies of alternative methods for history matching. Through this study it has been found that
the potential for the Norne benchmark case is great, and release of the data set could benet a
number of institutions. The challenges will be to provide synchronized data and ensure thorough
contact between StatoilHydro and the IO Center. In addition, there is a need for qualied
personnel to maintain consistency in published data and to provide support for the users.
Reservoir simulation models are powerful and essential tools for the development of oil and gas
elds. These realistic models are used for calculating reservoir volumes, for well planning, and to
predict the future behaviour of the eld. Building and maintenance of robust, reliable reservoir
models are time-consuming and expensive. The objective of this master thesis is to shape a
reservoir model with real data from the Norne Field in the Norwegian Sea.
The purpose of designing this model is to provide a benchmark dataset with real data, which
can be used by dierent institutions to compare the performance of dierent simulators and
simulation methods. The Norne Field is suitable because it is a rather young eld, with high
quality 4D seismic data and production data. It has been in production since November 1997
and has still many years left of production. StatoilHydro, which is the operator of the eld, is
positive to collaborate with NTNU and the Center for Integrated Operations in the Petroleum
Industry (IO Center). The IO Center will use the Norne model in their research program 2;
Real-time reservoir management. The release of the model is to be decided by StatoilHydro
and partners. If released, several users of the dataset could discuss their results of simulations
from a common basis. The IO Center is planning to design and provide a benchmark case for
use in reservoir simulation model testing. This benchmark case could be the rst of its kind
with real data. The aim is to establish a collaboration between several research institutions,
universities and companies and communicate results of the testing.
The master thesis is divided into two main parts. The rst consists of a description of the
Norne Field independent of simulation models. The second part comprises a description of the
base case simulation, with basis in the Eclipse input data. The goal is to make a good description
of a test case for simulation and history matching based on real data. This master thesis might
be used as a foundation for the test case planned by the IO Center.
The thesis starts with an introduction to the Norne Field in Chapter 2. Chapter 3 presents a
detailed description of the eld. This includes geology, petrophysics and all the wells on the eld.
Both exploration, and production/injection wells are described. The chapter also contains an
introduction to 4D seismic data as well as available seismic data from the Norne Field. Finally,
the production data is described and rates are included in table format in appendix C.1 and C.2.
Chapter 4 provides a description of the reservoir simulation model. It comprises the base case
1
simulation, the Eclipse simulator and the Eclipse input le with explanations of all keywords
used in the data le. A proposal for how the Norne benchmark case can be designed is presented
in Chapter 5. It also gives suggestions about what types of data that should be provided and at
what time. A discussion of the utility and potential for this kind of reservoir simulation models
is stated in Chapter 6. The chapter also deals with challenges connected to the shaping of a
benchmark case. Chapter 7 nally presents a conclusion.
2
Chapter 2
Introduction to the Norne Field
The Norne Field was discovered in December 1991. Development drilling began in August 1996
and oil production started November 6th 1997. The eld is located in the blocks 6608/10 and
6508/10 in the southern part of the Nordland II area in the Norwegian Sea, as seen in gure 2.1.
Sea depth in the area is about 380 m.
Figure 2.1: The location of the Norne Field [Statoil, 2001c]
Norne consists of two separate oil compartments; Norne Main Structure (Norne C-, D- and
E-segment), which contains 97% of the oil in place, and the North-East Segment (Norne G-
segment).
A 135 m hydrocarbon-bearing column was discovered from the exploration well, 6608/10-2
3
consisting of a 110 m thick oil leg with an overlying gas cap. The hydrocarbons were found in
the rocks of Lower and Middle Jurassic age. [Statoil, 2001c]
The eld is being developed with a oating production and storage vessel tied to six subsea
templates. An illustration of the eld is shown in gure 2.2. The well stream is carried by
exible risers to the vessel, which rotates around a cylindrical turret anchored to the sea oor.
The vessel has storage tanks for stabilised oil and a processing plant is located on the deck of
the ship. Figure 2.3 shows the vessel.
Figure 2.2: Development of the Norne Field [Statoil, 2001c]
Figure 2.3: The Vessel [NPD, 2008]
Approximately 0.403 mill Sm3 of oil was produced from 11 well slots in December 2007.
Water is injected in 8 wells. The Norne Field has produced 77 mill Sm3 of oil in total per
December 2007 [NPD, 2008]. That is approximately 86% of recoverable reserves. The Norwegian
Petroleum Directorate estimated the 31st of December 2006 the recoverable reserves to be 90
mill Sm3 of oil and 10.70 bill Sm3 of gas. Remaining reserves are estimated to be 17.3 mill Sm3
of oil and 5.7 bill Sm3 of gas. Gas production started in 2001. The eld is producing about
0.052 bill Sm3 of gas/month. [NPD, 2008]
4
Figures 2.4, 2.5, 2.6 and 2.7 illustrates the gross production of oil, gas, oil equivalents and
water per month from April 2006 until March 2008. The graphs shows that the production of
oil and gas gradually decrease while the water production increases.
StatoilHydro is the operator of the Norne eld with Petoro AS and Eni Norge AS as partners,
with respectively 39.1, 54.0 and 6.9 per cent interest.
Several studies have already been performed on the Norne Field. The results are described
by [Huseby et al., 2005], [Kowalewski et al., 2006], [Selle et al., 2008], [Steensen and Karstad,
1995], [Al-Kasim et al., 2002], [Boutte, 2007] and [Ouair et al., 2005].
Figure 2.4: Gross Production of Oil, April 2006 - March 2008 [NPD, 2008]
5
Figure 2.5: Gross Production of Gas, April 2006 - March 2008 [NPD, 2008]
Figure 2.6: Gross Production of Sm3 o.e., April 2006 - March 2008 [NPD, 2008]
6
Figure 2.7: Gross Production of Water, April 2006 - March 2008 [NPD, 2008]
7
Chapter 3
Detailed description of the Norne Field
3.1 Geology
The Norne eld is located in the blocks 6608/10 and 6508/10 on a horst block in the southern
part of the Nordland II area in the Norwegian Sea. The horst block is approximately 9 kmx 3 km [Ouair et al., 2005]. Figure 2.1 shows the location of Norne, while gure 3.1 shows the
structural setting of the eld. The eld is situated at the transition between the Nordland Ridge
and the Dønna Terrace, in an area called the Revfallet Fault Complex as seen in the gure. The
Nordland Area, which includes the Norne Field, has been exposed for two periods of rifting; in
Perm and Late Jurassic - Early Cretaceous. During the rst rifting, faulting aected a wide part
of the area. Especially normal faults, with NNE-SSW trends, are common from this period. The
second rifting period can be subdivided into four phases ranged in age from Late Bathonian to
Early Albian. The trend during this rifting was footwall uplift along the Nordland Ridge, and
erosion of high structures. Between the two rifting periods the tectonic activity was limited,
although some faulting occurred in the Mid and Late Triassic. This period was dominated by
subsidence and transgression. Some unconformities are discovered, possibly related to tectonic
activity. These unconformities are found between the Tofte and Tilje Formations, and within
the Ile Formation. After the last rifting no major structural development aecting the Norne
reservoir has taken place. The reservoir has gradually been buried deeper, allowing the oil and
gas to form and to accumulate within the reservoir. The rocks within the Norne reservoir are of
Late Triassic to Middle Jurassic age.
The reservoir sandstones in the formations Tilje, Tofte, Ile and Garn, are dominated by
ne-grained and well to very well sorted sub-arkosic arenites. The sandstones are buried at a
deep of 2500-2700 m and are aected by diagenetic processes. Mechanical compaction is the
most important process which reduces reservoir quality. Still, most of the sandstones are good
reservoir rocks. The porosity is in the range of 25-30 % while permeability varies from 20
to 2500 mD. [Statoil, 2001c]
The source rocks for the oil and gas in the Norne Field are believed to be the Spekk Formation
from Late Jurassic and coal bedded Åre Formation from Early Jurassic [NPD, 2005]. A source
rock is a rock of high organic content, which under the right circumstances, temperature and
8
Figure 3.1: The structural setting of the Norne Field [Statoil, 1994a]
pressure, will form oil and gas.
The cap rock which seals the reservoir and keeps the oil and gas in place is the Melke
Formation. The Not Formation behaves as a cap rock, preventing communication between
the Garn and Ile Formations. Also keeping the hydrocarbons in place is the rotated fault
blocks, in this relation called traps. Oil and gas is lighter than water and will migrate upward
until it is trapped. Both a cap rock and a trap is needed to preserve the hydrocarbons in the
reservoir. [Statoil, 1994a]
3.1.1 Zonation
The present geological model consists of 17 reservoir zones. Today's reservoir-zonation is slightly
altered from earlier subdivisions. The main dierence is that the Ile and Tofte zones have been
further subdivided, and the Tilje zones have been simplied. An illustration of the zonation
from 2001 can be seen in gure 3.2. The zonation is made to correspond as good as possible to
9
Figure 3.2: Stratigraphical sub-division of the Norne reservoir [Statoil, 2001c]
the actual change of lithology in the layers of the reservoir. Hence, boundaries between zones
are chosen at sequence boundaries and maximum ooding surfaces. Litological boundaries and
distinct breaks in porosity or permeability that correlates across the eld can also be basis for
the zonation. [Statoil, 2001c] Oil is mainly found in the Ile and Tofte Formations, and gas in the
Garn Formation [NPD, 2008].
The geological zonations from 2002 and 2006 are illustrated in gure 3.3. As seen from the
gure; Not is called Not 1, Garn has changed name to Not 2 and Lower Melke Formation has
changed name to Not 3. The Tilje Formation is still divided into four zones with no further
subdivision. The old names will still be used in the continuance of this thesis as the geomodel
considered is from 2004.
10
Figure 3.3: Old and new zonation [Fawke, 2008]
11
3.1.2 Stratigraphy and sedimentology
The entire reservoir thickness, from Top Åre to Top Garn Formations, varies over the Norne
Field from 260 m in the southern parts to 120 m in the northern parts [Statoil, 1994a], see
gure 3.4. The reason for this dierence is the increased erosion to the north, causing especially
the Ile and Tilje Formations to decrease in height [Statoil, 1995]. This has been found from
seismic mapping [Statoil, 1994a].
Figure 3.4: Cross-section Through Reservoir Zone Isochores [Statoil, 1994a]
The Åre Formation is the lowest formation within the Norne Field and has a heterolitic
composition. It is mainly comprised of channel sandstones which are 2-10 m thick and interbed-
ded with mudstones, shales and coals. The Åre Formation was deposited during Hettangian to
Early Pliensbachian, see gure 3.5. The total thickness of the formation varies a lot; from 200 min the southern Haltenbanken Area, to a more than 800 m thick column discovered in well
6608/10-2. An increased sand/shale ratio eastwards is discovered. The depositional environment
was probably alluvial to delta plain setting, transported from a source area to the east. [Statoil,
1994a]
12
Figure3.5:
Stratigraphicchart[InternationalCom
mission
onStratigraphy,2004]
13
The Tilje Formation was deposited in a marginal marine, tidally aected environment.
Sediments deposited are mostly sand with some clay and conglomerates. The source of the
sediments was located west of the Norne Area. The formation is thinning to the north due to
decreased subsidence rate during the deposition, along with increased erosion to the north/north-
east at the base of the overlying Tofte Formation. An unconformity is discovered at the top of
the Tilje Formation. This hiatus was most likely created due to uplift, followed by subaerial
exposure and erosion. It was probably the result of an important tectonic event. The hiatus
marks the transition from heterolitic sediments of the Åre- and Tilje Formations into thicker
marine sandstones of the overlying formations. The Tilje Formation is divided into four reservoir
zones based on biostratigraphic events and similarities in log pattern. Tilje 1 is not cored in
either of the wells 6608/10-2 nor 6608/10-3, but it is believed to consist of two sequences of
sand that is coarsening upward and a more massive sand at the top. Tilje 2 has a heterolitic
A total picture of the porosity of the Norne Field is obtained by relating the core porosity
to the density log. As a consequence, the water saturation has to be calculated using Archie's
formula. The net to gross ratio and permeability were also estimated in this study. For the
G-segment, separate values for net to gross ratio, porosity, water saturation and permeability
were calculated. [Statoil, 2001c]
Since the rst study, other wells have been cored on the Norne Field. This includes wells
6608/10-D-1 H, 6608/10-C-4 H and 6608/10-F-1 H. Based on these new cores, revision has been
worked out on porosity/permeability relations and the water saturation. [Statoil, 2001c]
The petrophysical parameters have been modelled in the geological model using co-located
co-kriging to acoustic impedance [Fawke, 2008].
3.2.1 Data
Well Information
Well 6608/10-2 was spudded October 28th 1991. The well was located at
66°,00',49.35"N08°,04',26.48"E
Total depth (TD) of the well was at 3678 m below Rotary Kelly Bushing (RKB), and this
depth was reached December 16th the same year. In January 1992, there were carried out four
drill stem tests on this well, which tested gas in the Garn Formation, oil in the Tofte Formation
and water in the Tofte/Tilje Formation.
The well discovered a hydrocarbon column of 135 m in the rocks of Lower and Middle
Jurassic. 110 m was oil, and the rest was an overlying gas cap.
Well 6608/10-3 was located at
66°,02',06.66"N08°,04',57.97"E
This well was spudded January 1993 and Total Depth (TD) was reached at 2991 m February
19th 1993. The month after, one drill stem test was performed, which tested oil in the Ile
Formation.
The well conrmed the test results from well 6608/10-2, and proved the extension of the eld
to north.
20
66°,02',25.26"N08°,09',41.74"E
Well 6608/10-4 was spudded in the end of 1993 and was located at
This well was drilled in the northeast segment, which is located approximately 3 km east of
the main structure. An oil column of 30.5 m was discovered in the same structures as the main
eld.
Figure 3.7 illustrates the location of the exploration wells. Alternating red and green indicates
that there exist both oil and gas. Green represents oil, while red represents gas.
Figure 3.7: Location of exploration wells [NPD, 2008]
21
Log data
The wells 6608/10-2, 6608/10-3 and 6608/10-4 have been logged with generally good quality.
Logs give important data for geophysical interpretation of the area. The dierent logs used for
acquiring data in the eld are mentioned below along with the logging interval given in m.
Well 6608/10-2:
mwd - 465-3335lwd-cdr cdn - 2100-2573difl acl gr - 867-3661zdl gr - 867-1525zdl cnl cal gr - 1520-2141zdl cnl cal gr - 2559-3644dll mll sl - 2559-2758diplog gr - 1520-2140diplog gr - 2559-3332diplog gr - 3329-3661fmt hp gr - 2579-2800fmt hp gr - 2650-2650cbl vdl gr - 394-1520acbl gr - 1563-2559acbl gr - 2505-3319velocity - 930-3640
Well 6608/10-3:
mwd - 472-2920difl acl gr - 863-1587cdl cnl gr - 1575-2914difl dac gr - 1574-2555dipl mac sl - 2430-2915dll mll gr - 2539-2800fmt hp gr - 2498-2862fmt hp gr - 2650-2650cbl vdl gr - 646-2871diplog gr - 1900-2555hrdip gr - 2563-2905swc - 894-2901vsp - 1240-2900
22
Well 6608/10-4:
mwd - 477-2558difl mac sl - 2175-2795zdl cnl gr - 2465-2794dll mll gr - 2465-2650hrdip gr - 1396-2555fmt gr - 2485-2662cbl vdl gr - 800-2746swc gr - 1430-2774vsp - 500-2750
[NPD, 2008]
The layers Ile 2, Ile 1, Tilje 4, Tilje 3 and Tilje 2 are eroded in well 6608/10-4. This can be
seen for instance from logs as demonstrated in gure 3.8, which illustrates correlation of wells
in the Norne Area.
Figure 3.8: Correlation of Wells in the Norne Area [Statoil, 1995]
Logs from the wells B-1 H, D-1 H and E-1 H are included on a CD accompanied with this
thesis. These logs are attached in relation with the 4D seismic data in section 3.4.3. Few of the
wells on the Norne Field have been logged with sonic logs, i.e. dt or dts. Only D-1 H has sonic
data of the three wells B-1 H, D-1 H and E-1 H. The log for D-1 H is edited and corrected for
mud ltrate invasion, and are suitable for modelling. The logs used for this well is gr, phie,
phit, rhob_v, vp_v and vs_v. For the two other wells there exist data for dt_synt, gr,
phif, and rhob. dt_synt is a synthetic dt log made with linear relation and are not logged
in the bore hole.
23
Core data
Core data has also been used as a basis for determination of the petrophysical properties of
the Norne Field. From well 6608/10-2 there has been cut six cores, eleven cores are cut from
well 6608/10-3 and 7 from well 6608/10-4. All this data has been depth shifted to match the
zdl-cn-gr. Photos of cores from the dierent formation are included in gures 3.9-3.14.
Use of core measurements is introducing some uncertainties which should be mentioned.
When drilling the cores, the transportation of the cores and the treatment of the core material
are vital. When performing measurements on the cores, there can be systematic errors connected
to equipment and methods. The plug may not be of general reservoir quality and will because
of that give incorrect results.
Figure 3.9: Cores from well 6608/10-2, inter-val 2600-2605 in the Garn Formation [NPD,2008]. Sandstones deposited near shore withsome tidal inuence
Figure 3.10: Cores from well 6608/10-2, interval 2611-2616 in the Not Forma-tion [NPD, 2008]. Grey to black claystonewith siltstone lamina, deposited in quiet ma-rine environment
24
Figure 3.11: Cores from well 6608/10-2, in-terval 2627-2632 in the Ile Formation [NPD,2008]. Sandstones deposited in shoreface en-vironment
Figure 3.12: Cores from well 6608/10-2, interval 2661-2665 in the Ror Forma-tion [NPD, 2008]. Very ne grained/shalysand, deposited in lower shoreface environ-ment with low sediment supply
25
Figure 3.13: Cores from well 6608/10-2, interval 2724-2729 in the Tilje Forma-tion [NPD, 2008]. Sand, with some clay andconglomerates, deposited in a marginal ma-rine, tidally aected environment
Figure 3.14: Cores from well 6608/10-2, interval 2674-2679 in the Tofte Forma-tion [NPD, 2008]. Channel sandstones
Test data
Well 6608/10-2: Test data from four drillstem tests (DST) has been reported for this well.
One of the tests showed evidence of Joule-Thomson eect as the temperature decreased when
the gas owed from the reservoir to the wellbore [Schlumberger Oileld Glossary]. As this test
was performed close to the gas-oil contact it is likely that the eect is a result of coning. All the
other DST's produced uids in accordance with the petrophysical evaluation made here [Statoil,
1994a].
DST 1 tested the interval 2715-2720 m in the lower Tofte Formation. Max bottom hole
temperature here was 100 C. 310 Sm3 water/day was produced through a 2" choke.
DST 2 tested the interval 2673-2695 m in the upper Tofte Formation. The production rate
measured was 1165 Sm3/d oil and 108667 Sm3/d gas through a 1.5" choke. Gas-Oil Ratio
was 93 Sm3/Sm3, oil density was 0.856 g/cm3, the gas gravity was 0.65 and the gas contained
1.8% CO2 and 4 ppm H2S. Max bottom hole temperature was 98.4 C.DST 3 tested the interval 2605-2610 m in the lower Garn Formation. The test produced 33 Sm3
26
condensate and 582600 Sm3 gas/day through a 19.05 mm choke. Measured GOR was 17654 Sm3/Sm3,
and max bottom hole temperature was 91.4 C.DST 3B tested the interval 2590 2603 m in the Garn Formation. Measured rates recorded
were 100 Sm3/d condensate and 9645000 Sm3 gas/day through a 38.1 mm choke. GOR were
recored to 9450 Sm3/Sm3. The condensate density was 0.783 g/cm3, the gas gravity was 0.645
and the gas contained 1.1% CO2 and 0.5 ppm H2S. Maximum bottom hole temperature measured
was 95.5 C. [NPD, 2008]
Well 6608/10-3: One drill stem test was carried out in this well. The test was performed
in the Ile Formation, in the perforated interval 2617-2648 m. The production was measured
to 1250 Sm3/d oil with density of 860 kg/m3 at standard conditions. 102500 Sm3/d gas was
produced with relative density of 0.65. The choke was of the size 60/64". [NPD, 2008]
Well 6608/10-4: In this well, three drill stem tests were performed.
DST 1 tested the Tofte Formation in the interval 2635-2640 m. No formation uid was
produced to the surface. Minifrac tests were performed at the end of this test, and the fracture
closing pressure was evaluated to 405 bar bar.DST 2 tested the Garn Formation in the interval 2566.2-2582.2 m. This test produced a
maximum of 900 Sm3/d oil with a density of 858 kg/m3 at standard conditions. 75000 Sm3/dgas with a relative density of 0.648 was measured. The choke was of size 80/64" (31.75 mm).
Minifrac tests were performed at the end of this test, and evaluated the fracture closing pressure
to be 410 bar.DST 3A and DST 3B tested the Melke Formation. DST 3A in the intervals 2484.5-2599 m
and 2505-2514 m, and DST 3B in 2524-2531 m. No formation uid was produced to the surface.
This test proved that the Melke Formation was tight with oil in place. [NPD, 2008]
FMT-data
The nal data type used for the petrophysical evaluation was the Formation Multi Tester (FMT)
log. This tool enables conrmation of a water bearing reservoir using pore pressure gradient.
It also allows sampling of the formation water. [NPD, 1994] Evaluation of the FMT-data gives
a base case oil-water contact at about 2688.5 m TVD/MSL for both well 6608/10-2 and well
6608/10-3. Well 6608/10-4 had a oil-water contact at 2574.5 m. Dierent gas-oil contacts were
observed in wells 6608/10-2 and 6608/10-3, while well 6608/10-4 did not contain any gas [Statoil,
1995]. Well 6608/10-2 had a gas-oil contact at 2580 m TVD/MSL and in well 6608/10-3 the gas-
oil contact was at 2575 m TVD/MSL. The FMT data also suggests that there is a small pressure
barrier in the northern segment (Segment E), caused by the presence of the Not Formation.
Figure 3.15 illustrates this feature.
However, it is shown by uid analysis that it is the same composition of oil above and below
this barrier. The calculated gradients are given in table 3.1. Reference depth used in the oil
zone was 2639 m and the formation pressure was 273.2 bar. [Statoil, 1994a]
27
Figure 3.15: Fluid model, from [Statoil, 1994a]
Table 3.1: Calculated gradients, with some uncertainty [Statoil, 1994a]Fluid Gradient
[ g/cm3]
Gas 0.19Oil 0.72
Water 1.02
3.2.2 Interpretation parameters
a, m and n The lithology factor, a, the cementation factor, m, and the saturation exponent,
n, have been estimated based on core analysis from wells 6608/10-2 and 6608/10-3. For the rst
two parameters the values were found from plug data with overburden measurements. Estimated
values are; a = 1.0 and m = 1.84. The saturation exponents are found for three dierent zone
groups, from Resistivity Index (RI) measurements. The groups and the n values are given in
table 3.2. 6 plugs from group 1, 9 plugs from group 2 and 5 plugs from group 3 are used as a
basis for the RI-measurements. [Statoil, 1994a]
Grain density
The average grain density for the entire reservoir, based on all core data from both wells are
ρma = 2.67 g/cm3. Zones of dierent grain densities are Tofte 3 and 2, 2.65 g/cm3 and Tofte
1, 2.71 g/cm3. [Statoil, 1994a]
Overburden corrections
The overburden pressure was calculated to correct results accordingly. To calculate the overbur-
den pressure, the density logs in wells 6608/10-2 and 6608/10-3 were integrated. A minimum
horizontal stress at depth 2673 m of 389 bar was indicated in a minifrac test [Statoil, 1992]. At
that depth, the pore pressure was 273 bar, hence the minimum horizontal stress is 116 bar and
28
Table 3.2: n-values for the zone groups [Statoil, 1994a]Group n- Formationnumber value names
Garn 2+11 1.84 Not
Ile 32 2.02 Ror
TofteGarn 3
3 2.20 Ile 2+1Tilje
the dierence between the horizontal and the vertical stress is 123.5 bar. Due to rock mechanicsthe conning pressure will be 123.5/3 + 116 bar. In [Statoil, 1994a] the equations for porosity
and permeability are given as:
Φres = 0.967Φatmos
Kres = 0.856Katmos1.004
Water resistivity
The resistivity of the formation water is found from the water sample from DST 1 in well
6608/10-2. It is temperature corrected using Arps formula. The resistivity is:
Rw = 0.054 Ω at 98.3 C
[Statoil, 1992]
Formation temperature
Both the formation temperature and the temperature gradient were determined from the DST.
They are:
T = 9.83 C at depth 2639 m TVD/MSL
∆T = 3.5 C/100 m
These values were in good agreement with the estimation carried out in [Statoil, 1992], where
the temperature was estimated to be 121.8 C at 3322 m MD/RKB. [Statoil, 1994a]
29
3.2.3 Evaluation
Porosity
Generation of total porosity is executed by use of the equation
φ = a+ b ∗ ρb
ρb is the bulk density, while a and b are constants. Crossplots of overburden corrected core
porosity vs. density log are used to nd these constants. The constants are found for the dierent
zones, which are grouped together for improving correlations. Some uncertainties are related to
the determination of the constants a and b from crossplots. [Statoil, 1994a]
Fluid contacts
As mentioned in section 3.2.1 there was a common oil-water contact at 2688.5 m TVD/MSL
for wells 6608/10-2 and 6608/10-3, while well 6608/10-4 had a oil-water contact at 2574.5 mand did not contain any gas. There were two dierent gas-oil contacts for wells 6608/10-2 and
6608/10-3; 2580 m and 2575 m respectively. The gas systems seem to be common over the entire
eld. That is also the case for the oil systems, except the oil above the Not Formation in well
6608/10-3. These contacts were also determined by FMT and DST data.
Formation resistivity
Calculations of the true formation resistivity in both the hydrocarbon zones and the water
zones were performed. The logs used for the calculations were environmentally corrected. In
the hydrocarbon zones the dll-mll log was used along with [Western Atlas Logging Services,
1985], while the deep induction logs were used for the water zones.
Water saturations
Two dierent models; Archie and Capillary pressure, were used to determine the water satura-
tion. These models are described in the following.
Archie The Archie equation is given below, and was used to evaluate Sw assuming clean sand.
The parameters needed for the equation are given in section 3.2.2.
Sw =(Rwa
Rtφm
)1/n
The average values of the water saturation, are given in tables 3.4, 3.5, 3.6 and 3.7.
It was assumed that Archie's equation could be used to estimate water saturation in the two
wells, and the constant a was treated without uncertainty. [Statoil, 1994a]
30
Capillary pressure The capillary pressure model is based on core data and was compared to
the log model, Archie. Estimations of water saturation were only made for the oil zones.
To normalize the capillary pressure data a J-function was used. The only assumption needed
for this method was: from a set of capillary pressure measurements for a reservoir, a single curve
of J vs. Sw can be drawn and used to determine the water saturation for a eld. The Leverett-
Figure 3.17: CPI-plot Well 6608/10-2 [Statoil, 1994a]
38
Figure 3.18: Log from NPD Well 6608/10-3 [NPD, 2008]
39
Figure 3.19: Log from NPD Well 6608/10-4 [NPD, 2008]
40
3.2.5 Uncertainties
Uncertainties in the study of petrophysics are associated to the used methods' assumptions
and simplications, input data and core measurements. Assumptions connected to the use of
Archie's equation, are mentioned in the section 3.2.3. Input data may have both random and
systematically errors. Measurements from cores have uncertainties connected to the handling
of cores from drilling and transport to measurement performance. The uncertainties introduced
by use of core measurements are discussed in section 3.2.1
3.2.6 Conclusions
The petrophysical evaluation of the eld has established parameters for; porosity, net to gross,
water saturations and permeability. These are used in the geological model and in the reservoir
simulation.
A continuous gas system for the eld is found in the upper part of the Garn Formation. Oil
is present below and down to the lower part of the Tofte Formation and the upper part of the
Tilje Formation. The oil system is divided in two parts; one in the Garn Formation in segment
G which is isolated by the Not Formation, and a continuous oil system for the rest of the main
eld. The initial GOC and OWC in the dierent formations and segments are listed in Table 3.9
and illustrated in gure 3.20.
Table 3.9: Initial GOC and OWC on the Norne Field [Statoil, 1994a]Formation C-segment D-segment E-segment G-segment
OWC GOC OWC GOC OWC GOC OWC GOCGarn 2692 2582 2692 2582 2618 2582 2585 No gas capIle 2693 2585 2693 2585 2693 2685 Water lled Water lled
Tofte 2693 2585 2693 2585 2693 2585 Water lled Water lledTilje 2693 2585 2693 2585 2693 2585 Water lled Water lled
Both well 6608/10-2 and well 6608/10-3 give good petrophysical properties. The average
porosity is in the range of 20-30%, permeability 20-2500mD, net to gross values in the range of
0.7-1 and water saturations 12-43% for hydrocarbon zones. Small variations in reservoir quality
between the three wells occur. Best reservoir quality is found in the upper part of Garn, Ile, Ror
and the upper part of Tofte. These are the most homogeneous parts of the reservoir. The Not
Formation consists of organic shale and net to gross here is zero. Sand intervals of good quality
are also found in Tilje and lower parts of Garn. However, these parts are more laminated and
cemented.
High permeability values are found in sand with good porosity. More shaly and cemented
sand has lower permeability. It is from this study generated a recommendation of what kind
of permeability that should be used for reservoir simulation. In accordance with permeabilities
from the DST tests, it is recommended to use the arithmetic means of permeabilities.
41
Figure 3.20: NE-SW running structural cross section through the Norne Field with initial andindications of present uid contacts, and current drainage strategy [Statoil, 2006a]
42
3.3 Wells
The Norne Field is being developed with a oating production and storage vessel. The vessel
is connected to six subsea wellhead templates named B, C, D, E and K, as seen in gure 2.2.
Template K was placed on the sea bottom in 2005, south of B, C and D templates. The K
template has 4 slots available; 3 for production and 1 for injction or production. The Norne
Field was discovered with well 6608/10-2 in 1991. Well 6608/10-3 conrmed the result of hydro-
carbons in the discovery well, while well 6608/10-4 encountered oil in the North-East segment.
Development drilling started with well 6608/10-D-1 H in August 1996. [Statoil, 2006a]
3.3.1 Exploration wellbores
To test the hydrocarbon potential in the sandstones and appraise oil accumulation in dierent
formations, 4 exploration wells were drilled. Several intervals were perforated and tested.
An overview of the exploration wells is presented in table 3.10 below.
43
Table3.10:Exploration
wellbores
[NPD,2008]
Nam
eUTM
En try
Date
Com
pletion
Purp ose
Status
Contents
TrueVertical
Hydrocarbon
coordinates
Date
Depth
[m]
form
ation(s)
6608/10-2
7321933.62N,457994.68E
28.10.1991
29.01.1992
WILDCAT
Plugged
and
OIL/G
AS
3677
FangstandBåt
Abandoned
6608/10-3
7324321.37N,458426.47E
07.01.1993
11.03.1993
APPRAISAL
Susp.
OIL/G
AS
2920
F angstandBåt
reenteredlater
6608/10-3R
7324321.37N,458426.47E
08.08.1995
17.08.1995
APPRAISAL
Plugged
and
OIL/G
AS
2920
FangstandBåt
Abandoned
6608/10-4
7324847.23N,462006.74E
15.12.1993
06.03.1994
WILDCAT
Plugged
and
OIL/G
AS
2800
Melke
andGarn
Abandoned
44
3.3.2 Description of exploration wells
Well 6608/10-2
Well 6608/10-2 was the well that rst discovered oil and gas on the Norne Field. The drilling
started the 28th of October 1991. The objectives were to test the hydrocarbon potential of
the Fangst Group of Middle Jurassic age, and to see if it was a sandy equivalent to the Rogne
Formation in the Viking Group. The plan was to drill a near to vertical well into rocks of Triassic
age at a total depth of 3225 m.
There were some problems with tight hole during drilling, which lead to extension in required
time. The extension was accepted because hydrocarbons had been proved and the goal of drilling
into the rocks of Triassic age was important. When drilling, it turned out that the Triassic
Formation were located deeper than expected, and the total depth of the hole ended at 3678 m.
Oil and gas were encountered in both the Båt and Fangst Groups from Lower and Middle
Jurassic. The uid contacts were found from logs and test, and revealed a gas-oil contact
at 2605 m and a oil-water contact at 2713.5 m. Cores recovered from well 6608/10-2 have a
total length of 141.5 m from the Båt and Fangst Groups. Two FMT samples have also been
collected; one gas sample and one oil sample.
The well was permanently plugged and abandoned on the 29th of January 1992. [NPD, 2008]
Well 6608/10-3
The second exploration well drilled on the Norne eld was 6608/10-3. This well was spudded
7th of January 1993. The purpose of was to evaluate the accumulation of oil in the Båt and
Fangst Groups in the Northern fault block on the Norne Field.
The well was drilled to a total depth of 2921 m, into Lower Jurassic formation. Oil and gas
were encountered in both Båt and Fangst Groups. A total of 11 cores were cut from Lower
Melke to Tilje Formations. In addition, four FMT samples were taken, containing mudltrate,
oil and gas.
The well was suspended as an oil and gas appraisal well the 11th of March 1993. A re-entry
of the well was performed the 8th of August 1995, and the well was permanently plugged and
abandoned as an appraisal well the 17th of August 1995. [NPD, 2008]
Well 6608/10-4
Well 6608/10-4 was the rst exploration well to be drilled on the North-East area. Its purpose
was to prove the presence of oil in the Middle Jurassic sandstones in the G-segment.
Drilling started on the 15th of December 1993, and the well was drilled to a total depth
of 2800 m. It reached rocks of the Lower Jurassic Åre Formation. As anticipated, oil was
encountered in the Melke and Garn Formations of Middle Jurassic age. A total of 8 cores were
cut from the Cretaceous Nise Formation to the Åre Formation. FMT samples were extracted
from the Melke, Garn and Ile Formations with content varying from only mudltrate to also
containing traces of oil and gas.
45
The 7th of March, well 6608/10-4 was plugged and abandoned as an oil and gas discov-
ery. [NPD, 2008]
3.3.3 Development wellbores
The Norne Field has 4 templates for production and 2 templates for injection. Each template
has 4 slots available. Oil was produced from all 12 slots in January 2006, and all 8 injection wells
were used for water injection. This was before the wells on the K-template were completed. The
eld is developed only with horizontal producers today. Some of the producers were rst drilled
vertical to some deviated, to accelerate the build-up of well potential until plateau production was
reached. These wells have been sidetracked to horizontal production wells. [Statoil, 2001c] New
well technology has been implemented on Norne to increase recovery, for instance multilateral
wells.
Both gas and water have been injected into the reservoir, but the gas injection was stopped
in 2005 [Statoil, 2001c]. However, injection of gas from the C-wells started again in 2006 for an
extended period to avoid pressure depletion in the gas cap [NPD, 2008].
The decision of wellbore locations is based on these principles:
Water injectors located at the anks of the reservoir
Gas injectors located at the structural heights of the reservoir
Oil producers located between gas and water injectors for delaying gas and water break-
through
Oil producers located at some distance from major faults to avoid gas inow
The principles presented above are used for all well locations as an initial location. The locations
are thereafter optimized with regard to gas and water breakthrough times by use of reservoir
simulation studies.
Total number of active wells in December 2006 was 17, with 11 oil producers, 3 water injectors
and 3 gas injectors. The wells are completed in dierent formations depending on the drainage
strategy. A summary of each well is presented in section 3.3.4. [Statoil, 1994b]
Drilling history
10 wells were predrilled to obtain plateau production from the production start-up. 7 of the 10
wells were oil producers with good productivity and late breakthrough of gas and water. 3 wells
were predrilled for injection; 1 for injection of produced gas, and two for pressure maintenance
water injection. The water injectors were perforated below oil-water contact, and gas injectors
in the top Garn Formation. [Statoil, 1994b]
The development wells are presented in table 3.11 below. A more detailed description of all
the development wells are given in section 3.3.4. Appendix B.1 contain plots of oil production
rate, water cut and gas-oil ratio for all production wells and injection rates for the injection
wells.
46
Table3.11:Developmentwellbores
[NPD,2008]
Nam
eUTM
En try
Date
Com
pletion
Purp ose
Status
Contents
Total
coordinates
Date
Depth
[m]
6608/10-B-1
H7322128.37N,457125.62E
26.01.1999
05.04.1999
PRODUCTIO
NPLUGGED
OIL
4300
6608/10-B-1
AH
7322128.37N,457125.62E
06.11.2005
03.12.2005
OBSE
RVATIO
NPLUGGED
NA
3478
6608/10-B-1
BH
7322128.37N,457125.62E
04.12.2005
09.01.2006
PRODUCTIO
NPRODUCING
OIL
2976
6608/10-B-2
H7322122.85N,457121.88E
13.12.1996
09.12.1997
PRODUCTIO
NPRODUCING
OIL
3862
6608/10-B-3
H7322135.86N,457122.07E
21.05.1999
05.07.1999
PRODUCTIO
NPRODUCING
OIL
4150
6608/10-B-4
H7322136.28N,457114.14E
12.01.1998
06.02.1998
PRODUCTIO
NPLUGGED
NA
2555
6608/10-B-4
AH
7322136.28N,457114.14E
13.06.2001
12.07.2001
OBSE
RVATIO
NPLUGGED
NA
3900
6608/10-B-4
BH
7322136.28N,457114.14E
13.07.2001
07.08.2001
PRODUCTIO
NPLUGGED
OIL
4346
6608/10-B-4
CH
7322136.28N,457114.14E
03.06.2004
19.06.2004
OBSE
RVATIO
NPLUGGED
NA
3630
6608/10-B-4
DH
7322136.28N,457114.14E
20.06.2004
10.07.2004
PRODUCTIO
NPRODUCING
OIL
2870
6608/10-C-1
H7322024.28N,457190.08E
12.02.1998
20.07.1998
INJE
CTIO
NINJE
CTING
WATER
3255
6608/10-C-2
H7322026.87N,457182.56E
01.10.1998
27.11.1998
INJE
CTIO
NINJE
CTING
WATER
4421
6608/10-C-3
H7322029.45N,457175.54E
06.04.1999
20.05.1999
INJE
CTIO
NINJE
CTING
WATER
3800
6608/10-C-4
H7322034.03N,457180.02E
18.11.1996
18.08.1997
INJE
CTIO
NPLUGGED
GAS
2900
6608/10-C-4
AH
7322034.03N,457180.02E
15.11.2003
13.01.2004
INJE
CTIO
NINJE
CTING
WATER
3638
6608/10-D-1
H7321942.88N,457269.01E
28.09.1996
18.11.1996
PRODUCTIO
NPLUGGED
NA
3500
6608/10-D-1
AH
7321942.88N,457269.01E
28.05.2002
25.06.2002
OBSE
RVATIO
NPLUGGED
NA
2897
6608/10-D-1
BH
7321942.88N,457269.01E
26.06.2002
05.09.2002
PRODUCTIO
NPLUGGED
NA
4852
6608/10-D-1
CH
7321942.88N,457269.01E
30.09.2003
07.11.2003
PRODUCTIO
NPRODUCING
OIL
4575
6608/10-D-2
H7321938.31N,457264.41E
09.01.1997
05.01.1998
PRODUCTIO
NPRODUCING
OIL
4174
6608/10-D-3
H7321948.37N,457254.48E
05.07.2000
04.08.2000
PRODUCTIO
NPLUGGED
NA
4198
6608/10-D-3
AH
7321948.37N,457254.48E
05.08.2000
30.08.2000
PRODUCTIO
NPLUGGED
OIL
5100
Continued
onNextPage...
47
Table3.11
Continued
Nam
eUTM
Entry
Date
Com
pletion
Purpose
Status
Contents
Total
coordinates
Date
Depth
[m]
6608/10-D-3BY2H
7321948.37N,457254.48E
12.08.2000
25.09.2005
PRODUCTIO
NSU
SP.ATTD
OIL
5400
6608/10-D-3BY1H
7321948.37N,457254.48E
06.07.2005
07.10.2005
PRODUCTIO
NSU
SP.ATTD
OIL
5400
6608/10-D-4
H7321952.94N,457259.08E
07.01.1998
18.06.1998
PRODUCTIO
NPLUGGED
NA
3137
6608/10-D-4
AH
7321952.94N,457259.08E
11.01.2003
09.06.2003
PRODUCTIO
NPRODUCING
OIL
5829
6608/10-E-1
H7325447.22N,459204.35E
28.05.1999
19.06.1999
PRODUCTIO
NPRODUCING
OIL
4350
6608/10-E-2
H7325441.40N,459199.73E
16.10.1999
21.11.1999
PRODUCTIO
NPLUGGED
OIL
4075
6608/10-E-2
AH
7325441.40N,459199.73E
28.07.2005
15.08.2005
PRODUCTIO
NPLUGGED
OIL
3775
6608/10-E-2
BH
7325441.40N,459199.73E
23.11.2007
OBSE
RVATIO
NPLUGGED
4204
6608/10-E-2
CH
PRODUCTIO
N
6608/10-E-3
H732545.14N
,459189.80E
29.07.1998
23.09.1998
PRODUCTIO
NPLUGGED
OIL
3110
6608/10-E-3
AH
732545.14N
,459189.80E
02.10.2000
12.12.2000
PRODUCTIO
NPLUGGED
OIL
4849
6608/10-E-3
BH
732545.14N
,459189.80E
09.03.2005
03.04.2005
OBSE
RVATIO
NPLUGGED
NA
3259
6608/10-E-3
CH
732545.14N
,459189.80E
04.04.2005
07.05.2005
PRODUCTIO
NPRODUCING
OIL
4018
6608/10-E-4
H7325455.72N,459194.27E
05.02.2000
12.03.2000
PRODUCTIO
NPLUGGED
NA
4508
6608/10-E-4
AH
7325455.72N,459194.27E
12.03.2000
01.06.2000
PRODUCTIO
NPRODUCING
OIL
6069
6608/10-F-1
H7325354.98N,459309.39E
29.04.1999
27.05.2005
INJE
CTIO
NINJE
CTING
WATER
3170
6608/10-F-2
H7325350.39N,459304.92E
18.09.1999
15.10.1999
INJE
CTIO
NINJE
CTING
WATER
3048
6608/10-F-3
H7325357.56N,459301.87E
02.12.1999
05.02.2000
INJE
CTIO
NINJE
CTING
WATER
3370
6608/10-F-4
H7325364.72N,459299.20E
10.06.2001
06.07.2001
INJE
CTIO
NINJE
CTING
WATER
4280
6608/10-F-4
AH
7325364.72N,459299.20E
01.10.2007
08.11.2007
INJE
CTIO
NSU
SP.ATTD
WATER
4080
6608/10-J-2H
7325822.28N,462456.23E
02.11.2005
22.12.2005
PRODUCTIO
NPRODUCING
OIL
3290
6608/10-K-1
H7321915.19N,457092.53E
18.10.2006
20.12.2006
PRODUCTIO
NSU
SP.ATTD
3795
6608/10-K-3
H7321926.4N
,457087.91E
04.09.2006
17.10.2006
PRODUCTIO
NSU
SP.ATTD
3849
Continued
onNextPage...
48
Table3.11
Continued
Nam
eUTM
Entry
Date
Com
pletion
Purpose
Status
Contents
Total
coordinates
Date
Depth
[m]
6608/10-K-4
H7321918.24N,457095.73E
29.03.2007
29.10.2007
PRODUCTIO
NSU
SP.ATTD
OIL
4104
49
3.3.4 Description of development wells
Well 6608/10-B-1 H
The tenth development well to be drilled on the Norne Field was well 6608/10-B-1 H. This was
a horizontal well, producing from Ile 2 and Tofte 3 Formations. The purpose was to drain oil
from the C-segment, mainly the north eastern parts. Reasons for drilling this well was the desire
to achieve low GOR and at the same time rapidly build up to plateau production. Production
start was 1st of April 1999.Two horizontal segments were the producing parts of the B-1 H well. The rst, in the
heal of the well, was 400 m long and located in the top of the Ile 2 Formation. The second
horizontal segment, located in the Tofte 3 Formation, was 600 m long and was the toe of the
well. Further completions were possible and also side-tracking toward the northern parts as a
horizontal producer in the Ile Formation. The well was plugged in October 2005. [Statoil, 1999a]
A pilot well, 6608/10-B-1 AH, was drilled as a sidetrack to 6608/10-B-1 H to conrm the
location of the OWC as interpreted from the 2004 4D seismic data. This was done to optimise
the placement of the production well 6608/10-B-1 BH, which started production January 2006.
Pressure data from the dierent reservoir sections were acquired in the pilot. [Statoil, 2005a]
Well 6608/10-B-2 H
As the third development well to be drilled, well 6608/10-B-2 H started to produce the 9th of
December 1997. It produces from the eastern part of the C-segment with a horizontal section
from northwest to southeast in the top of the Ile Formation.
The horizontal section of the well is 850 m long and is completed in the Ile Formation for
production. At a later stage the whole reservoir can be completed for production, from top Garn
to total depth. This will allow for both oil and gas production. [Statoil, 1997a]
Well 6608/10-B-3 H
At 1st of July 1999, well 6608/10-B-3 H started to produce from the western part of the D-
segment and the southern part of the E-segment. The well is completed in Ile 2 and Tofte 3
Formations.
The well was drilled in two horizontal sections to enable production from both Ile and Tofte.
As there are two major faults in this area, the completed intervals are much shorter than in
B-2 H for instance. The depths of both sections were set based on the goal of not having early
water break through or high gas-oil ratio. The well was initially completed in the Ile and Tofte
Formations, and can be further completed along the entire reservoir interval in the future if that
is needed. [Statoil, 1999b]
50
Well 6608/10-B-4 H
Well 6608/10-B-4 H was the fth development well drilled on the eld. It was a vertical pro-
ducer, drilled through Garn, Ile, Ror, Tofte and Tilje Formations. The well was planned to
drain the western part of the C segment and started producing the 27th of April 1998. It was
perforated only in the Tofte 3 Formation, while the whole interval is available for perforation,
and modications have been performed. The well was shut May 31st 2001. [Statoil, 1998a]The well 6608/10-B-4 AH was drilled as a pilot for well 6608/10-B-4 BH to locate the present
oil-water contact and the formation tops in the D-segment. With this information, the placing
of well B-4 BH in the exact right spot was easier. A better understanding of the pressure balance
was also achieved. The pilot well B-4 AH was a success. [Statoil, 2002a]
The objective of drilling well 6608/10-B-4 BH was to make a 600 m long horizontal well
within the Ile 2 Formation. The oil-water contact was actually deeper than rst anticipated,
and the result was a 483 m long horizontal section in the upper part of the Ile 2 Formation.
Perforations were made in this section as well as in the Garn 3 Formation where a gas lift valve
was installed. The production from B-4 BH started the 1st of August 2001 and lasted until the
1st of September 2003 when it closed due to high water production. [Statoil, 2002a]
Well 6608/10-B-4 CH was planned and drilled as a pilot for well 6608/10-B-4 DH to verify
the uid contacts in the location where B-4 DH was planned. The pilot were drilled because of
uncertainties about location of the gas-oil and oil-water contacts in the C-segment. In addition to
this, a calibration of the contacts to the 2003 4D seismics was important in order to place B-4 DH
in the optimal position. Two dierent gas-oil systems with dierent levels of the uid contacts
were discovered in the pilot drilling. Some were higher than expected and some lower. Residual
gas was found below the Not Formation in addition, this came from the C-3 H injector. [Statoil,
2005b]
The objective of drilling well 6608/10-B-4 DH was to drain oil from the upper Ile Formation
in the south western area of the C-segment. This was done with a horizontal production well.
As a result of the pilot drilling, the planned well path was changed to the alternative location
to avoid the large amounts of injected gas around well C-3 H. The rst attempt to drill this well
was stopped due to failure in the PowerDrive BHA, and the hole was plugged and abandoned.
The next attempt was called B-4 DHT2. This was side-tracked in the Melke Formation and
drilled much further to the east, away from C-3 H. The well was drilled through the Garn
Formation and into the upper Ile Formation where it has a 357 m long perforated, horizontal
section. Production from this well started the 4th of July 2004. [Statoil, 2005b]
Well 6608/10-D-1 H
This well was the rst development well to be drilled on the Norne Field. The plan was to drill
it as a producer in the Ile, Ror and Tofte Formations in the southern part of the eld. Average
inclination of the well from top Ile to total depth was 44. As this was the rst well to be drilledin this area, results from the well was important for the further development of the eld and
numerous tests were performed. The production start in this well marks the start of the life of
51
the Norne Field; the production start date was the 7th of November 1997. The well was shut
the 1st of September 2002. [Statoil, 1997b]When well D-1 H was shut, a side-track was planned. A pilot, 6608/10-D-1 AH, was to be
drilled rst to log the formation, nd uid properties and the oil-water contact in the southern
part of the C-segment. When the drilling of the pilot started, the drillstring got stuck and the
well needed to be redrilled to run the logs. It was decided not to do this due to a relatively
high cost and risk compared to gain. The well was plugged back and drilling of the producer,
6608/10-D-1 BH, commenced. The plan for well 6608/10-D-1 BH was to have a highly deviated
section placed in the Ile Formation using geosteering. Logging While Drilling (LWD) through
the Garn, Not, Ile 3 and Ile 2 Formations was performed to achieve a sucient production
interval in Ile 3 and Ile 2 Formations. A 350 m long interval was achieved in Ile 3, while in Ile
2 a 800 m long interval was achieved. The gas lled Garn Formation, is open for perforation at
a later stage. This well started to produce the 2nd of November 2003. [Statoil, 2003]
Well 6608/10-D-2 H
The plan for well 6608/10-D-2 H was to drill a horizontal producer through the Ile Formation
in the C-segment. Because of lost cones in the hole, the rst track was plugged back soon after
entering the Ile reservoir.
The second track, 6608/10-D-2 T2H, was more successful and reached its target in Ile 2 and
Ile 3 with a near horizontal section of almost 1.1 km. This well was abandoned for a short while
with the plan of perforating it in Ile 2 for production. The well started producing the 24th of
December 1997. [Statoil, 1997c]
Well 6608/10-D-3 H
The well 6608/10-D-3 H was drilled as a pilot to conrm the location of the oil-water contact in
the C-segment. The producer in the area was planned to be well 6608/10-D-3 AH. The plan was
to make a horizontal producer through Ile 2 and Tofte 3 reservoirs. The result was according to
plans with a 53 m long section of Ile 2 Formation and a 998 m long section of Tofte 3 Formation
penetrated. At rst, only the Tofte 3 reservoir was perforated for production, which started the
28th of August 2000. The well was closed the 2nd of June 2005. [Statoil, 2001a]When well D-3 AH was shut, a multilateral side-track was planned. This was to consist of
one lateral to drain oil from the Ile 2 Formation in the C-segment, 6608/10-D-3 BY1H, and one
lateral to drain oil from the Ile 2.2 Formation in the western part of the D-segment, 6608/10-D-3
BY2H. [Statoil, 2006b]
The rst lateral, D-3 BY1H, was side-tracked from D-3 AH in the Spekk Formation and the
goal was to drill through the Spekk, Melke, Garn and Not Formations and then land horizontally
in the Ile 2 Formation. The rst attempt on this failed because the required buildup angle was
not achieved. The hole was cemented back and a new attempt was made. This attempt, D-3
BY1HT2, was again side-tracked from the Spekk Formation, and this time it was a success. The
target was reached and logging indicated hydrocarbons along the entire reservoir section.
52
The second lateral, D-3 BY2H, was side-tracked from D-3 AH in the Lyr Formation. The
well was successfully drilled down to the Not Formation and into the Ile 1.3 and 1.2 Formations
where it went almost horizontal until it reached a main fault. Then it went through a short
section of the Not Formation before it was drilled through the Ile 2.2 and 2.1 Formations. Due
to complications when the production liner was to be run, this lateral was abandoned and the
D-3 BY1HT2 was completed as a single bore producer. [Statoil, 2006b]
The production from D-3 BY1HT2 started the 26th of February 2006.
Well 6608/10-D-4 H
This sixth development well to be drilled on the Norne Field, 6608/10-D-4 H, was a deviated
production well with an inclination of 40 through the Garn, Not, Ile, Tofte, Tilje and Åre
reservoir intervals. The purpose was to drain the eastern part of the C-segment and to contribute
to a rapid build-up to plateau production. At rst, the well was perforated in the Ile and Tofte
reservoirs and started production the 17th of June 1998. At a later stage, the entire reservoir
section can be completed for production or the well can be side-tracked to a more north-eastern
prospect. [Statoil, 1998c] Production from well 6608/10-D-4 H was shut the 16th of November
2002 because of water breakthrough.
When well D-4 H was shut, a plan for the side-track was made; well 6608/10-D-4 AH. The
plan was to drain oil from the Garn Formation in the north-eastern part of the D-segment
with a highly deviated well. To reach the goal, the well was to be drilled with Gyro. When
rigging up for this, the drill string got stuck and it was shot o. Well D-4 AH was plugged and
abandoned. [Statoil, 1998b]
A second attempt of drilling the producer was made, called 6608/10-D-4 AHT2. The plan
was to perforate large intervals of the Garn 3 and Garn 2 reservoirs. A section of 300 m of the
Garn 3 reservoir was planned to be perforated initially. Thereafter 100 m of the Garn 2 reservoir
was to be perforated. When the well was drilled, an unexpected fault was penetrated in the
Garn 3 reservoir. The result was 155 m long perforation in the Garn 3 Formation and 122 mlong perforation in the Garn 2 Formation, with a possibility of expanded perforation interval
in the Garn 2 Formation. The total length of the perforation was according to plan, but the
perforations in the Garn 3 Formation were reduced due to the fault. [Statoil, 1998b] Production
from D-4 AHT2 started the 4th of June 2003.
Well 6608/10-E-1 H
Well 6608/10-E-1 H was the ninth oil production well and the fourteenth development well
drilled on the eld. It was designed as the fth horizontal production well, to drain oil from the
southern part of segment E. Low GOR oil was planned to be produced from the well, and it
should facilitate the rapid build up to plateau production on the eld.
Ile 2 and Ile 3 Formations were completed, but the entire reservoir interval can be completed
if necessary in the future. [Statoil, 2000a] Production from well 6608/10-E-1 H started September
1999.
53
Well 6608/10-E-2 H
The tenth oil producer and sixteenth development well to be drilled on Norne was well 6608/10-
E-2 H. The well was horizontal, and should drain oil from the southern part of the E-segment.
This well was planned for producing a low GOR oil, and facilitating the rapid build up to plateau
production.
The reservoir was at 2604 m TVD MSL, and the well was drilled horizontal at that depth. Ile
3 and Ile 2 Formations were perforated. The location of GOC and OWC were studied when the
well was to be placed to prevent early water break through and high GOR oil. [Statoil, 2000b]
Well 6608/10-E-2 H started producing oil November 1999 and was producing until July 2005.
The objective for well 6608/10-E-2 AH was to drain the remaining oil in segment E. The
well trajectory was planned as a horizontal section below the Top Ile Formation, over the OWC
at approximately 2606 m TVD MSL. It was drilled deeper than planned and penetrated higher
than the anticipated OWC, before it was steered back through Ile 2.1 Formation. [Statoil, 2006c]
The well started to produce oil in August 2005.
Well 6608/10-E-3 H
Well 6608/10-E-3 H was the eighth development well and rst production well planned in the
northern part of segment E. An inclination of 16 in the well path through Garn, Not, Ile, Tofte
and Åre Formations was used. The central part of segment E was the target for draining. The
well was designed to contribute to a low GOR oil production, and provide a reference point in
the northern part of the eld to conrm reservoir communication.
Ile and upper Tofte Formations were completed, but the entire reservoir interval was planned
to be available for completion at later stages. In addition, the well can be sidetracked as a
horizontal well toward the western part of segment E. [Statoil, 1999g] Well 6608/10-E-3 H
started production December 2000 and was plugged May 2000.
Well 6608/10-E-3 AH was designed as a horizontal well to drain oil from the Garn Formation
in the northern area of Segment E. It was assumed that the OWC was at 2688.5 m TVD/MSL
and the path was planned thereafter. During drilling it was found that the OWC was at a
shallower depth than rst expected. The consequence was a drilling stop and plugging back
before sidetracking as well 6608/10-E-3 AHT2.
Well 6608/10-E-3 AHT2 penetrated the Garn Formation horizontally. The well was located
in sands containing oil the whole section except for an interval in the water zone. The OWC
could in that way be dened in the Garn Formation to be 2617 m TVD MSL in the central
part of Segment E. [Statoil, 2002b] It started up production December 2000 and produced until
January 2005.
Well 6608/10-E-4 H
Well 6608/10-E-4 H was a pilot well drilled to test the depth of the Garn Formation in Segment
G. Bad weather suspended drilling of the well, and the BHA was pulled into the casing. When
54
the hole was reentered, the BHA hit an obstruction which could not be bypassed. The solution
was to sidetrack the well to 6608/10-E-4 HT2.
Well 6608/10-E-4 AH was the eleventh oil producer drilled, located in the G-segment. It
was horizontal, placed 5-10 m TVD below the top of the Garn Formation, sidetracked from pilot
well 6608/10-E-4 H. During completion, problems occurred and the well had to be sidetracked
to 6608/10-E-4 AHT2.
Well 6608/10-E-4 AHT2 was completed in the Garn Formation, with a 600 m interval per-
forated. [Statoil, 2002c] Well 6608/10-E-4 AHT2 started production in June 2000. Then there
was a stop in production from June 2001 until August 2002 and from July 2005.
Well 6608/10-K-1 H
The actual trajectory of well 6608/10-K-1 H did not follow the planned wellbore because the
main fault between C and D segments was greater than prognosed. The well was cemented to
the Not 1 Shale, and sidetracked with the K-1 HT2 through Ile 2.2 and Ile 2.1 Formations. The
well entered the Ile 1.3 before it crossed the main fault and entered the Ile 2.2 where it was
perforated.
The well was planned to drain remaining oil from the Ile Formation in the north-western
part of segment C and south-western part of segment D. K-1 H was designed as a producer only
and the whole interval was planned to be available for perforation. In the future, it is possible
that K-1 H may be sidetracked from Not Formation or Melke Formation. [Statoil, 2007a]
Well 6608/10-K-3 H
Well 6608/10-K-3 H started to produce oil 15th of October 2006. It was the rst production welldrilled from the K-template. This well was also used to drill the exploration well 6608/10-11 S
Trost, before proceeding down, deviated to horizontal, to the base of the Melke Formation. The
well was completed in the Ile 2.2 Formation.
The primary objective of the well was to drain the remaining oil in the Ile Formation in
segment C. The well can be sidetracked at later stages from the Not Formation or Melke For-
mation. [Statoil, 2007b]
Well 6608/10-K-4 H
Well 6608/10-K-4 H was designed as a horizontal producer through the Ile 2.2 Formation. When
drilling through Not 1 Shale, it collapsed, and the wellbore was abandoned. Thereafter the
sidetracked K-4 HT2 was steered according to the plan.
Primary objective of the well was to drain oil from the Ile Formation in the north-western
part of Segment C. The well was designed as a producer only, where the whole reservoir interval
was planned to be available for completion in the future. [Statoil, 2008]
55
Well 6608/10-C-1 H
Well 6608/10-C-1 H was the seventh development well and the rst water injection well drilled
on the Norne Field. It injects water into the water leg. This well could also inject gas at a later
stage if needed. All injectors on the C template can convert between water and gas injection.
An inclination of 12 was used and the well was drilled through Garn, Ile, Ror, Tofte, Tilje and
Åre Formations. Completion of the well was performed with a perforated cemented liner within
the base Tofte and upper Tilje Formations, and the injection started the 21st of July 1998. Thewell can be perforated through the whole interval later if needed. Side-tracking of the well in
north-east direction is also possible if water support is required in the Norne G-segment. [Statoil,
1999c]
Well 6608/10-C-2 H
The second water injector to be drilled on the Norne eld was the 6608/10-C-2 H injector. The
plan was that this injector should support the already existing injection into the southern part of
the eld provided by C-1 H. This well can also easily be converted to a gas injector if needed. It
was drilled through the Garn, Ile, Tofte, Tilje and Åre Formations with an inclination of 50-45.The well was perforated within the Tilje 3 and 4 Formations. The entire reservoir interval is
available for perforation at a later stage and there is a possibility of side-tracking toward the
southern parts of the C-segment. [Statoil, 1999d] The injection started the 21st of January 1999.
Well 6608/10-C-3 H
Well 6608/10-C-3 H was the third water injection well to be drilled on the Norne eld. The
plan for this well was to support the existing injection from C-1 H and C-2 H in the southern
part of the eld, by injecting water into the water leg. As for the other injection wells at the
C-template, C-3 H can easily be converted from water injection to gas injection. The well was
drilled through the Garn, Ile, Tofte, Tilje and Åre reservoir intervals with an inclination of
15-10. The perforation started about 10 m TVD above the oil-water contact, in the Tofte 3
Formation, and continued within the Tofte 2, Tofte 1 and Tilje 4 Formations. Injection start
was on the 21st of May 1999.
The well is located in the south-western part of the C-segment with the bounding faults of
the main eld to the north and southwest. When the well was pressure tested it was discovered
that there were poorer communication between Ile, Tofte and Tilje than expected. This was the
reason why the injection from C-1 H and C-2 H increased the pressure only in the Tilje Formation
and not in the Tofte Formation. To enhance the pressure support in the Tofte Formation the
perforations were made higher up than originally planned. [Statoil, 1999e]
Well 6608/10-C-4 H
Well 6608/10-C-4 H was drilled in the north-western part of the C-segment as the second devel-
opment well. The well penetrates the Garn Formation and is a vertical gas injector. Perforations
56
are made with a cemented liner in Garn 3. The injection started the 22nd of November 1997
and lasted until the well was shut the 18th of November 2003. Well C-4 H was then plugged and
side-tracked to well C-4 AH. [Statoil, 1999f]
The reason for shutting well C-4 H was that it contributed to a high gas-oil ratio and water
cut in the neighbouring production wells. Well 6608/10-C-4 AH was drilled as the rst injector
in the Ile Formation on the C-segment. It was placed there to provide pressure support, enhance
the oil sweep from the Ile Formation and to verify the oil-water contact in the Garn Formation.
The well was drilled to total depth in the Åre Formation with an inclination of less than 20.The initial perforations were a 38 m long section in the Ile Formation, with the possibility of
extending to cover the entire reservoir section, from Garn to Tilje, at a later stage. As for the
other wells at the C-template, it can easily switch between water and gas injection. [Statoil,
2004]
Based on new seismic data, the original target was moved about 100 m to the southwest to
be able to verify the oil-water contact in the Garn Formation. The contact was not proved in
the well and suggested that the oil-water contact still corresponds to the initial of 2692 m TVD
MSL. The injection from C-4 AH started the 20th of January 2004. [Statoil, 2004]
Well 6608/10-F-1 H
Well 6608/10-F-1 H was the fourth water injector to be drilled, located to the north of the Norne
E-Segment. The well was designed to inject water in the water leg in northern part of the eld.
All wells on the F-template can easily be converted from water to gas injection. Well F-1 H was
drilled vertically through Garn, Ile, Tofte, Tilje and Åre Formations. The well was perforated
approximately 23 m TVD below the oil-water contact in the Ile and Tofte Formations. Injection
from this well started September 1999.
The entire reservoir interval can be perforated in the future. Pressure testing from the well
has proved good communication between Ile and Tofte. [Statoil, 1999h]
Well 6608/10-F-2 H
The fth water injector drilled on Norne was well 6608/10-F-2 H, located to the north of the
Norne D-Segment. An angle of 13 was used on the trajectory through Garn, Ile, Tofte, Tilje
and Åre Formations. The well was perforated within the interval of Ile and Tofte, approxi-
mately 31.5 m below the oil-water contact. The well started injection of water in October 1999.
As for well F-1 H, the entire reservoir interval is available for further completion, and pressure
testing from the well has proved good communication between Ile and Tofte Formations. [Statoil,
2000c] The well can easily be converted from water to gas injection. [Statoil, 1999h]
Well 6608/10-F-3 H
This was the sixth water injector drilled on the eld, located in the south-western part of the
E-segment. The well was drilled with an angle of up to 50 in the top hole section and less
than 20 in the reservoir. It was perforated in the Tofte and Tilje Formations [Statoil, 2001b].
57
Injection start was in September 2000. As for the other wells in the F-template it is easy to
convert from water to gas injection. [Statoil, 1999h]
Well 6608/10-F-4 H
Well 6608/10-F-4 H was the seventh water injector drilled with purpose to inject water into the
water leg south of well E-4 AHT2 in the G-segment. This well had no pressure support and had
to be shut in for a period in 2001 and 2002 due to low reservoir pressure. [Statoil, 2002d] The
injector started injecting water in September 2001 and it can easily be converted to inject gas.
58
3.4 4D seismic data
3.4.1 Introduction to 4D seismic data
4D seismic data is 3D seismic data acquired over the same area at dierent times. Time-lapse
seismics is another word for this technology, which purpose is to detect changes in the subsurface
during production of hydrocarbons. The observed changes are changes in uid location and
saturation, as well as in pressure and temperature. This kind of seismic data can be acquired
either on the surface or in a borehole. [Schlumberger Oileld Glossary] It is important to have
the various surveys surveying the exact same locations to achieve reliable results. The best
results are obtained if receivers are permanently placed at the seabed so that the signals are
recorded from the exact same places during each survey.
Statoil has used 4D seismic data in the reservoir management for approximately 70% of
their operating elds. The data has produced important information used to locate remaining
hydrocarbons in the reservoirs. [Ouair et al., 2005] On the Norne Field, a total of 5 seismic
surveys have been carried out, starting with the rst conventional base survey in 1992. The next
four surveys have been rendered with a Q-marine vessel in 2001, 2003, 2004 and 2006. [Statoil,
2006a] The survey area is shown in gure 3.21. Repeatability is good and the survey data is of
high quality. The only place where the data is poorer, is in the area around and beneath the
Norne production vessel. Undershoot was performed in the monitor surveys in order to generate
coverage beneath the vessel. This gives a fairly acceptable repeatability in this area. [Ouair
et al., 2005] The next Q-marine survey is to be performed during June 2008 [Cheng and Osdal,
2008].
The survey performed in 2001 was a 40 km2 single source survey. It was named ST0113 and
was intended as a time-lapse survey. ST0113 was compared to the survey from 1992. Earlier
in 2001, a survey on the Norne Area was performed with reservoir characterisation as purpose.
The survey was called ST0103 and data from this was included in the processing of ST0113
to ensure the necessary migration aperture. The Q-marine survey acquired in June 2003 was
named ST0305. It covered 85 km2 and was carried out as identically as possible to ST0113. The
3rd Q-marine survey, ST0409 covered a larger area, approximately 146 km2. It was acquired in
July 2004, as identically as possible to the 1st and 2nd survey. The 4th Q-marine survey, ST0603,
was shot in July/August 2006, as identically as possible to the 3rd survey. Time-lapse changes
in the reservoir between the years 2001, 2003, 2004 and 2006 could now be identied. Several
undershoot lines were acquired to monitor beneath the Norne production vessel. Two dierent
undershoot vessels and source were used. The rst was used in the 2001 and 2003, while the
second was used in the 2004 and 2006. [WesternGeco, 2007]
Seismic data available in this work:
3D seismic survey from 2006 with near, far, mid and full osets
4D cubes from the years 2006-2001, 2006-2003, 2003-2001 and 2004-2001
interpreted top reservoir horizon
59
interpreted faults
well paths for all wells
interpreted oil-water contacts from 2001, 2003, 2004 and 2006
interpreted cubes of pressure and water and gas saturations from the years 2006-2001,
2006-2003 and 2003-2001
2 velocity cube for conversions, both time and depth
All the data can be acquired by requesting [Department of Petroleum Engineering and Ap-
plied Geophysics].
Figure 3.21: Map of the seismic survey area, with wells
The usage of the 4D seismic data at the Norne Field has been to observe the dierence
in amplitude and acoustic impedance. Results have then been used to adjust the simulation
model [Cheng and Osdal, 2008]. The 4D results have indicated changes in the saturations,
which the simulation model did not predicted. In the Garn Formation, water was predicted
from the model as migrating to the northwest and south of one of the wells. However, 4D
inversion clearly indicated migration of water to the east. This demonstrates the importance
time-lapse seismics has for a eld with complex geology as Norne. Two additional cases have
60
involved issues where the model could not predict future behaviour with condence, and 4D
data provided the required data. [Boutte, 2007] The 4D seismic data is also an important tool
in the process of targeting the remaining oil. [Statoil, 2006a]
3.4.2 Seismic processing
The processing of each survey was performed in two phases; a generation of a fast-track cube and
a full processing. The pre-stack and post-stack portions of the full seismic processing ow of the
2006 survey are illustrated in gures 3.22 and 3.23. A detailed description of all the processing
steps can be found in [WesternGeco, 2007], attached digitally.
Figure 3.22: The pre-stack portion of the full seismic processing ow [WesternGeco, 2007]
61
Figure 3.23: The post-stack portion of the full seismic processing ow [WesternGeco, 2007]
62
4D Quality Control
The quality control (QC) of 4D seismics included generation of 3D stacks of the full volume. It
also involved analysis of amplitude, phase and time dierences between data sets from dierent
years, normalised rms dierence amplitudes and visual inspection of inline and crossline dier-
ence data sets. These attributes were computed in a 2000-3000 ms window, after use of a 5-40
Hz bandpass lter. A full 4D QC was performed at the following stages:
- Missing shot interpolation- SRME (Surface Related Multiple Elimination)- Taup mute and radon demultiple- Swath dependent time shifts- Dip-moveout- Inverse dip-moveout- Final stack- Final post processing
No problems of importance were observed in the 4D QC. The 4D QC steps performed
throughout the processing ow, see gures 3.22 and 3.23, indicate that the processing ow was
performed as intended. [WesternGeco, 2007]
63
3.4.3 Seismics on Norne
The SeisWorks® 3D software is used for viewing of seismics in this thesis. Seisworks provides
innovative 3D viewing and interpretation capabilities and is an industry standard software.
3D seismics
To get an impression of how the Norne Field looks like in the subsurface, 3D seismics can be
studied. Geophysicists can interpret faults, horizons and other trade terms, based on available
geological information and seismic images. Knowledge of the locations of top reservoir and
oil-water contacts, is important information required for calculation of the reservoir volume.
Observed changes in the horizons over time are also of interest. When the oil-water contact
moves up, it denotes that the amount of hydrocarbons left in the reservoir decreases. A change
in the position of the top reservoir horizon can suggest that there has been a change in the
pressure, accordingly a compaction.
For this master thesis, line number 1100 and trace number 1600 are selected to represent the
eld. However, all lines and traces are available for the work. Three wells located in the vicinity
of line 1100 are selected to be represented by logs. These are wells are B-1 H, D-1 H and E-1 H.
The gures B.51- B.54 in appendix B.2.1 demonstrates the oil-water contact at dierent years.
The top reservoir is represented as the horizon called Top Not 2. Several faults are marked, and
the three wells B-1 H, D-1 H and E-1 H are also included in the gures. The same properties
are shown for trace 1600 in gures B.55-B.58.
Logs from the three chosen wells are attached to the thesis digitally on a CD. A few of the
wells on Norne have been logged with sonic logs, i.e. dt or dts. Only D-1 H has sonic data
of the three wells B-1 H, D-1 H and E-1 H. The log for D-1 H is edited and corrected for mud
ltrate invasion, and are suitable for modelling. The logs used for this well is gr, phie, phit,
rhob_v, vp_v and vs_v. For the two other wells, there exist data for dt_synt, gr, phif
and rhob. dt_synt is a synthetic dt log made with linear relation and is not logged in the
bore hole.
Changes of the oil-water contact from the rst survey, 2001, to the last, 2006, are shown in
gures 3.24 and 3.25 for the line and the trace, respectively. The gures are made in time, i.e.
the y-axis. As can be seen, the wells are not located far from the oil-water contact in 2006, but
both the production wells, B-1 H and D-1 H, was sidetracked to higher formations before this
survey.
64
Figure 3.24: 3D seismic, line number 1100 showing oil-water contact in 2001 and 2006
65
Figure 3.25: 3D seismic, trace number 1600 showing oil-water contact in 2001 and 2006
66
4D seismics
The Q-marine surveys shot in 2001, 2003, 2004 and 2006 are used for 4D seismics. Time-lapse
changes in the reservoir between the dierent years are identied by use of these data. In
this work, data cubes with dierence between acoustic impedance between the following years:
2001-2003, 2001-2006 and 2003-2006 are studied. These dierences are extracted by subtraction.
4D data with dierence between 2001 and 2006 are shown in gures 3.26 and 3.27. Result for
the same line and trace for the years 2001-2004 and 2001-2003 are given in gures B.60, B.62, B.59
and B.61 in appendix B.2.2, respectively.
Figure 3.26: 4D seismic, line number 1100, 2001-2006
Changes in acoustic impedance are due to pressure or saturations changes which lead to a
dierent velocity. The interpretations are made by geophysicists in StatoilHydro working with
the eld on a daily basis. To be able to show the interpreted changes in relation to the seismic
area it belongs to, it is necessary to display both pictures. This can be done with the overlay
function in seisworks. The pressure or saturation change is shown as variable density, while the
4D cube is put on top as wiggle. In order to do this, it is important that both the 4D data and
the interpretations are made in either depth or time. It is possible to convert the cubes between
67
Figure 3.27: 4D seismic, trace number 1600, 2001-2006
depth and time by use of a velocity cube.
The gure 3.28 shows 4D seismics overlaid interpretation of pressure changes from 2001 until
2006.
4D seismics is an important tool in connection with well planning. By studying the water
saturation changes in the reservoir, water ooded areas can be located and avoided as possible
well locations. To avoid high gas-oil ratio, the gas saturation changes should be studied. The
4D seismics can also be utilized in the work of history matching by comparing real seismics with
synthetic seismics created from the simulation output. Agreement and disagreement between
the simulation model and the historical data can be discovered from such a comparison.
A history matching of the production on the Norne Field was performed in 2005. The period
covered is from production start in 1997 to a revision stop August 22th 2004.To assess the reliability and prediction capability of a simulation model, it is possible to
history match the model up to a certain date, and not include all available data. When the
history match is accomplished, a prediction can be made. The prediction is run until the date
of the last available eld data. Then the results from the simulator can be compared to the real
data and the reliability of the history matched model can be assessed. The period from August
2004 to June 2005 was used to compare the real data to the prediction.
The history matching is performed by use of pressures from FMT logs, GOR, water cut and
oil-water contact rise interpreted from 4D data. The transmissibilities of faults and vertical
barriers are adjusted, and some relative permeability curves are changed to provide a better
t. [Statoil, 2005c] To assess and minimize the mismatch between observed and simulated data
in computer-assisted history matching, objective functions are used. Because uncertainties are
weighted and put into the objective function, it is important that they are properly assessed. It
is especially hard to assess the uncertainties in Time-lapse seismics because of the complexity
of data acquisition, survey repeatability, seismic processing and seismic inversion. [Ouair et al.,
2005] The history match has been updated since 2005. The present model is matched until
December 1st 2006.
79
4.2 Description of the base case
The simulation model is divided into two main parts; the history period and the prediction
period. The rst part covers 9 years in great detail, and the second part covers the next 15
years. The CPU times for these models are approximately 4 and 2 hours for the history and
prediction, respectively [Statoil, 2005c]. The next sections will describe these two parts which
compose the base case.
4.2.1 History Period
The simulation base case starts the 6th of November 1997, when production starts in well D-1 H.The model is history matched until the 1st of December 2006. At that time there are 12 activeproducers; B-1 BH, B-2 H, B-3 H, B-4 DH, D-1 CH, D-2 H, D-3 BH, D-4 AH, E-1 H, E-2 AH,
E-3 CH and K-3 H, along with 8 active injectors; C-1 H, C-2 H, C-3 H, C-4 AH, F-1 H, F-2 H,
F-3 H and F-4 H. The last well that started producing was K-3 H which started 15th of October2006. Wells that have been shut down or sidetracked are; B-1 H, B-1 AH, B-4 H, B-4 AH, B-4
Injection uids have been both gas and water. The wells on template F have only injected
water while the wells on the C-template have injected both water and gas in an irregular pattern.
How the injection and production strategies have changed since the start up is shown in gure 4.3.
Red illustrates gas, green is oil and blue is water.
Figure 4.3: The drainage strategy for the Norne Field from pre-start and until 2005 [Statoil,2006a]
The base case model is based on the geological model from 2004, which has been updated
both in 2005 and 2006. This model consists of 44 431 active cells, and the eld is divided into 22
layers. Figure 4.4 and 4.5 shows the change in oil saturation in the eld from start of production
until 1st of December 2006.
80
Figure 4.4: Oil saturation applied to the reservoir simulation model seen from above at simulationstart
81
Figure 4.5: Oil saturation applied to the reservoir simulation model seen from above at the endof the history period
82
Plots of base case simulation results and historical values for the history period are shown
in gures 4.6-4.10; where the base case results are shown in blue, while the actual results are
shown in pink.
Figure 4.6: Field Oil Production Rate, History Period
The oil production rate has some discrepancies over the entire simulation time, varying
between higher and lower values than the actual case, see gure 4.6. The match is quite good,
but improvements can be made - especially for the last year.
When total oil production is considered, the errors in the total oil production match is not
of signicance. This is illustrated in gure 4.7. The only discrepancy that should be improved
is the one in 2006, which is the one that inuences the major dierence in the total amount of
oil produced.
There is no record of the actual eld pressure in the simulation model. It is therefore dicult
to make comments on the reliability of the calculated eld pressures. However, as seen from
gure 4.8, the pressure is rst declining, but in mid 1999 the pressure starts to build up again, and
eventually exceeds the initial pressure. This denotes a greater injection volume than production
volume.
83
Figure 4.7: Field Oil Production Total, History Period
Figure 4.8: Field Pressure, History Period
84
Figure 4.9: Field Gas-Oil Ratio, History Period
The gas-oil ratio for the eld lies between 160 and 300 during the entire base case simulation,
see gure 4.9. As for the oil production rate, the match is quite good, but for the gas-oil ratio
the match improves in 2006 when the oil production rate match was rather poor. Due to the
higher oil production rate and the match in as-oil ratio, the gas production rate in the base case
is higher than the actual. The ratio also decreases at this time and is only about 130 Sm3/Sm3.
The eld water cut is steadily increasing over the simulation time of the history period, see
gure 4.10. In the last part, 2006, the actual water cut is higher than the calculated. This is
connected to the discrepancy in the oil production rate. The simulated base case is producing a
higher amount of oil and gas than the historical data, hence decreasing the amount of water it
should be producing and the result is discrepancy in the water cut.
85
Figure 4.10: Field Water Cut, History Period
Total amounts of produced and injected uids per 1st of December 2006 are:
Field Oil Production Total = 73711304 Sm3
Field Gas Production Total = 15.293217 109Sm3
Field Water Production Total = 15642383 Sm3
Field Water Cut = 0.47463846 fracField Gas-Oil Ratio = 205.56699 Sm3/Sm3
Field Water Injection Total = 103.96547 106Sm3
Field Gas Injection Total = 8.6845399 109Sm3
A total oil production of 73711304 Sm3 equals 45.9% of the original oil in place in the
simulation model [Statoil, 2005c]. Another comparison is recoverable reserves calculated by the
Norwegian Petroleum Directorate [NPD, 2008], where the base case production equals 81.9% of
the recoverable oil reserves.
4.2.2 Prediction Period
A prediction has also been made until 1st of January 2022. The drainage strategy for this periodis shown in gure 4.11. The uids produced in the prediction are both oil and gas. As the
pressure in a eld decreases, more of the dissolved gas becomes free gas. Most of the remaining
hydrocarbons on the Norne Field lie in the upper parts of the reservoir, in the Garn and upper
Ile Formations. The injection uid is water and it is injected into the lower Tofte Formation.
During the prediction only water is injected in all wells. All injection wells except F-4 AH
86
Figure 4.11: The drainage strategy for the Norne Field from 2005 and until 2014 [Statoil, 2006a]
inject at constant rates given in table 4.3. The rate in well F-4 AH is under group control,
and it is injecting its share of a group target with a maximum rate of 2500 Sm3/d. Rates for
this well is steadily increasing during the prediction, from 1400 Sm3/d, January 2007, until it
reaches the injection limit in September 2019. The rate is then kept constant at 2500 Sm3/duntil simulation end in January 2022.
Table 4.3: Injection rates during the predictionWell WIRname Sm3/day
Figure 4.12: Oil saturation applied to the reservoir simulation model seen from above at the endof the prediction period
88
No historical data is available during the prediction period. On basis of that, no discussion
of the model's reliability with comparisons of simulated results and actual values, are included.
However the performance of the eld can be assessed by studying some plots, see gures 4.13-
4.17.
Figure 4.13: Field Oil Production Rate, Prediction Period
The oil production rate is steadily declining, see gure 4.13, as a consequence of end of
plateau production. A small peak in the rate can be observed in the beginning of 2014. That
might be the result of the new well P-20 that opens the 11th of January 2014.
The total amount of oil produced is also increasing, but not as fast as before, see gure 4.14.
This is due to the decreasing production rates.
From gure 4.15 it can be seen that the reservoir pressure continues to increase during the
rst years of the prediction period. The turning point is in 2014 when the pressure starts to
decline quite fast. It is due to the start-up of the gas producer P-20, which starts to produce
gas from the Garn Formation. The nal reservoir pressure is 234 bar.Field gas-oil ratio is rather constant until the gas producer P-20 starts to produce in 2014,
see gure 4.16. The gas-oil ratio increases rapidly the rst months and continues to increase
until the end of 2018, when it starts to decrease.
The eld water cut is steadily increasing over the entire prediction period. As can be seen
from gure 4.17, there are only a few minor exceptions to this, which is when new wells start to
produce in 2007 and 2014.
89
Figure 4.14: Field Oil Production Total, Prediction Period
Figure 4.15: Field Pressure, Prediction Period
90
Figure 4.16: Field Gas-Oil Ratio, Prediction Period
Figure 4.17: Field Water Cut, Prediction Period
91
Total amounts of produced and injected uids per 1st of January 2022 are:
Field Oil Production Total = 100.43956 106Sm3
Field Gas Production Total = 25.503697 109Sm3
Field Water Production Total = 162.8764 106Sm3
Field Water Cut = 0.94593531 fracField Gas-Oil Ratio = 701.75629 Sm3/Sm3
Field Water Injection Total = 294.80131 106Sm3
Field Gas Injection Total = 8.6845399 109Sm3
The total oil production of 100.43956 106Sm3 equals 62.5% of the original oil in place in
the simulation model [Statoil, 2005c]. It exceeds the recoverable reserves calculated by the
Norwegian Petroleum Directorate [NPD, 2008] by more than 10 106Sm3.
92
4.3 Eclipse reservoir simulator
In order to run numerical simulations on a reservoir model, a simulator is needed. Reservoir
simulations divides the reservoir into a number of small blocks and applies the fundamental
equations of uid ow through porous media, phase behaviour and conservation to each block.
The result is display of variations in reservoir rock and uid parameters in space and time. The
numerical work involved in actual simulation problems is very large and requires the use of high
speed computers. For the Norne Field base case, the Eclipse reservoir simulator has been used.
This simulator consists of two separate simulators; Eclipse 100 and Eclipse 300 specializing in
black oil modelling and compositional modelling, respectively.
For the Norne Field, Eclipse 100 is used. It is a fully implicit, general purpose black oil simu-
lator that can handle up to three phases in three dimensions. It also has a gas condensate option.
Eclipse is written in FORTRAN and will run on any computer that has an ANSI-standard FOR-
TRAN90 compiler and sucient memory, or it can be run in a parallel mode [Sch, 2007b]. Other
important options available in Eclipse are corner-point versus block-center geometry and radial
versus cartesian coordinate systems [NTNU, 2007].
An input le is needed to be able to run a simulation. This le must contain all data
concerning the reservoir and how it is exploited. A special name format has to be used for
the Eclipse input le, namely FILENAME.DATA. The input le is constructed using certain
keywords used in the right order. There are eight main sections that can be included in the input
le. These are given by the section-header keywords runspec, grid, edit, props, regions,
solution, summary and schedule, where all are mandatory except for grid, regions and
summary. Each of these sections is followed by multiple keywords, where some are required
while others are optional. They will be thoroughly described in the next section, by use of the
Eclipse Reference Manual, reference [Sch, 2007a].
The Norne input les are included on a CD, because of the enormous amount of data.
However, the .DATA le is attached in appendix D as a sample.
93
4.4 Section Keywords
4.4.1 runspec
The runspec section is required as the rst section of an Eclipse data input le. It includes
title, start date, problem dimensions, switches, phases/components present etc.
Several keywords are introduced in the runspec section. These turn on various modelling
options or contain data. The set of runspec keywords included in the Norne le will be
presented below.
Keywords in the RUNSPEC section
The dimens keyword denes the number of blocks in X, Y and Z directions. The numbers for
the Norne eld are 46, 112 and 22.
The gridopts keyword requests additional options for processing the grid data. It is followed
by two items; the rst allows the alternative transmissibility multipliersmultx-, multy-,multz-
etc. to be used in the grid, edit or schedule sections. The keyword is also used if the
alternative diusivity multipliers diffmx-, diffmy-, diffmz- etc. are used. When YES is
typed as item 1, keywords as multx-, diffmx- etc. may be used. Item 2, nrmult, is the
maximum number of multnum regions entered in the grid section. This apply either to inter-
region transmissibility multipliers, using the multregt keyword, or pore volume multipliers
using the multregp keyword. This item is set to zero in the Norne le, which means that any
multiplier is applied between ux regions entered using fluxnum, see section 4.4.2.
The active phases present in the runs are dened by typing their names. The Norne Field
has oil, water, gas, disgas and vapoil included. These words represents oil, water, gas,
dissolved gas and vaporized oil in wet gas.
Unit convention in the le is dened as eld, metric or lab units. The Eclipse le of Norne
uses metric units.
The hysteresis option is enabled by use of the hyst keyword. If the hyst keyword is selected,
imbnum values must be entered in the regions section.
The start date of simulation is entered after the start keyword. The start of the Norne run is
November 6th 1997. eqldims consists of three items and species the dimensions of equilibration
tables. The rst item, ntequl denes the number of equilibration regions entered by use of
eqlnum in the regions section, see section 4.4.5. 5 regions are established in the Norne
case. The second item gives the number of depth nodes in any table of pressure versus depth
constructed internally by the equilibration algorithm. The number entered here is 100. Finally,
the maximum number of depth nodes, which is 20, in any rsvd, rvvd, rswvd, rtempvd,
pbvd or pdvd table entered in the solution section to dene the initial Rs, Rv, Tr, Pb or Pdversus depth is typed.
eqlopts denes several options for equilibration. The keyword is followed by one item in the
Norne Eclipse le. This is thpres which enables the threshold pressure option. When thpres
is entered, ow will be prevented from occurring between dierent equilibration regions until
94
the potential dierence exceeds a threshold value. The threshold value is to be specied with
the keyword thpres in the solution section. Also, if named faults have threshold pressures,
the option is required.
Dimension data of regions are entered under the regdims keyword. The data consists of
4 items in the Norne case, and these describe the maximum number of regions associated with
miscellaneous keywords in other sections. An illustration of how it is done is show below. Item
1, which is 22, is the maximum number of uid-in-place regions (NTFIP) dened with keyword
fipnum in the regions section, see section 4.4.5. Item 2 (NMFIPR) is the number of sets of
uid-in-place regions. 3 sets are present. Item 3 (NRFREG) denes the maximum number of
independent reservoir regions. This option is set to 0 for Norne. The nal item (NTFREG)
gives maximum number of ux regions for the Flux option, or the maximum number of regions
used by the fluxnum keyword in the grid section, see section 4.4.2. 20 ux regions is the
maximum number for Norne.
Tracer dimensions and options are introduced in the runspec section. The tracers are
described and options for the tracer tracking algorithm are included here. The tracers keyword
is followed by up to six items, but only one is included for the Norne case. This is the maximum
number of passive water tracers entered using tracer in the props section, section 4.4.4. 10
water tracers are the maximum amount in the Norne case.
Well dimensions are given under the keyword welldims. The data can consist of up to 10
items, but for the wells in the Norne eld only 4 items are used to describe the dimensions of
the well data to be used in the run. The entered numbers in the Norne le are as follows; the
maximum number of wells in the model is 130, the maximum number of connections per well
is 36, the maximum number of groups in the model is 15 and the maximum number of wells in
any group is 84.
Dimensions of tables are dened by use of the tabdims keyword as shown in the gure
below. The data describes the sizes of saturation and PVT tables, and the number of uid-in-
place regions used in the run. The Norne le uses 6 items to describe table dimensions. Item
1 gives the number of saturation tables in the props section, which is 107. Item 2 denes the
number of PVT tables in the props section. There are 2 such tables in the BC0407.DATA le.
Maximum number of saturation nodes in any saturation table is given as item 3, and the number
is 33. Item 4 is the maximum number of pressure nodes in any PVT table or rock compaction
table. 60 pressure nodes are the maximum in the Norne le. Item 5 gives the maximum number
of FIP regions given in the regions section under the fipnum keyword, see section 4.4.5. 16
such regions are the maximum here. The last item denes the maximum number of Rs nodes in
a live oil PVT table or Rv nodes in a wet gas PVT table, which is 60 for the Norne eld.
The vfpidims keyword denes injection well VFP table dimensions. The data consists of
95
three items and describes the dimensions of the injection well Vertical Flow Performance tables
entered in the schedule section using the vfpinj keyword. Item 1 is the maximum number
of ow values per table. Item 2 is the maximum number of tubing head pressure values per
table, while item 3 gives the maximum number of injection well VFP tables. The numbers are
respectively 30, 20 and 20 for Norne.
As for the injection wells, the VFP table dimensions must be described for the production
wells. This is done by use of the vfppdims keyword. The data consists of six items. These
are as follows; 1: The maximum number of ow values per table, 2: The maximum number of
tubing head pressure values per table, 3: The maximum number of water fraction values per
table, 4: The maximum number of gas fraction values per table, 5: The maximum number of
articial lift quantities per table and 6: The maximum number of production well VFP tables.
Numbers used for Norne are shown below.
Dimensions for fault data are specied by use of the faultdim keyword. One single item of
data denes the maximum number of segments of fault data entered in the grid section with
the faults keyword, see section 4.4.2. Maximum number of fault segments is 10000 here.
The pimtdims keyword is used to describe the number of tables of PI scaling factor versus
maximum water cut entered in the pitmultab keyword, and the maximum number of entries
in any table. The two integers for the Norne case are 1 and 51.
The nstack data represents the size of the stack of previous search directions held by the
ORTHOMIN linear solver. By increasing the value of nstack, the memory required for a run
is increased as well. The stack size is 30 in the Norne case.
The keywords unifin and unifout indicate that input les and output les, which can be
multiple or unied, are to be unied.
The option keyword activates special program options. The options are principally of a
temporary or experimental nature. They can also act to restore back-compatibility with earlier
versions of the code. The option keyword is followed by a number of integers. Each of these
activates a special option. A value equal to zero switches o the special option, while a value
other than zero activates a special option. In the Norne Eclipse le, there are 77 integers which
are set equal to 1. These 77 options are described in detail in the Eclipse Reference Manual[Sch,
2007a].
96
4.4.2 grid
The purpose of the grid section in the BC0407.DATA le is to specify the geometry of the
computational grid, and to set rock properties for the grid blocks in the grid. Based on this
information, Eclipse calculates grid block mid-point depths, pore volumes and inter block trans-
missibilities.
The system in the Norne le is of cartesian geometry and the keywords used in this section
depends on the geometry option.
All keywords used in this section are described in the following.
Keywords in the grid section
When the keyword newtran is set it means that the transmissibilities are calculated from the
cell corner points. It also enables automatic calculation of fault transmissibilities.
The gridfile keyword is used to control the output of the geometry. It is followed by one
or two integers. The rst integer denes whether a .GRID le is to be written and the extension
of this, while the second integer denes the .EGRID le which is to be written and what format
it should have. In this case both extended .GRID and .EGRID les are written.
The keyword mapaxes is used to enable storage of the origin of the maps used to generate
the grid. For post-processing purposes, the origin is available through the .GRID le. The
number of items following the keyword is six, consisting of three pairs of coordinates. The rst
pair gives the coordinates of one point of the grid y-axis relative to the map, the second gives
the coordinates of the grid origin relative to the map origin, and nally the coordinates of one
point of the grid x-axis relative to the map. For this case the values are 0 100 0 0 100 0.
To specify the grid data units, the gridunit keyword is used. The keyword is followed by
two items where the rst states the unit of length of the grid data, while the second indicates the
relation of the measured grid data. The second item is set to MAP if the grid data is measured
relative to the map, or is left blank if it is relative to the origin given in the mapaxes keyword.
In the Norne case the grid data units are given in meters and are relative to the origin given by
the mapaxes keyword.
The init keyword requests that an .INIT le should be created and outputted. Such a le
contains a summary of all the data entered in the grid, props and regions sections. An .INIT
le can be either formatted or unformatted. The later is the case here.
If there is a desire to reset print and/or stop limits for messages of any severity type, the
messages keyword is used. There are 6 levels of severity in Eclipse from the informative
MESSAGE, to the suspected programming error printed as a BUG. The rst 6 items following
the messages keyword resets the print limit for each of the severity levels, while the last 6
items resets the stop limits for each of the severity levels. For this case all of the print limits,
and the stop limits for the two least severe messages is set to 10000, while the stop limits for a
WARNING, PROBLEM, ERROR and BUG are 20000, 10000, 1000 and 1, respectively.
To activate the minimum pore volume a cell has, the minpv keyword is used. It is followed
by a single, positive number which is the minimum pore volume of an active cell. For the Norne
97
Field the minimum pore volume for an active cell is 500.
A pinch-out is when a layer of rock is terminated by thinning or tapering out against another
type of rock [Schlumberger Oileld Glossary]. To generate connections across such pinched-out
layers the pinch keyword is used. It can be followed by up to ve items. The rst item states the
pinchout threshold thickness, while the second item controls the generation of pinchouts when a
minimum pore volume has been set by the minpv keyword. As a third item, the maximum empty
gap allowed between cells in adjacent grid layers where non-zero transmissibility is wanted, is
set. The fourth item states in which way the pinchout transmissibilities should be calculated.
It can be done either by using harmonic average of the z-direction transmissibilities of all cells
nearby (ALL), or only by half-cell z-direction transmissibilities of active cells on each side of
the pinchout (TOPBOT). The nal item is used to account for multz through a pinched-out
column, but is only used if the fourth item is set to TOPBOT. The transmissibility multiplier
that will be used can be the multz (TOP) or the minimum of this value for the active cells
at the top of the pinchout (ALL). For the Norne simulation model the treshold thickness is set
to 0.001 m. The generation of pinchouts is set to GAP which indicates that non-neighbouring
connections are allowed across cells that are inactive even if the thickness exceeds the treshold.
As maximum empty gap in item 3 the value is set to innity, 100* 1018. The last two items are
set to TOPBOT and TOP, respectively.
To reduce the amount of print-outs from a run or to avoid the out-put of large included les
the keyword noecho can be used. Here it is used to avoid the print-out of all the included les
into the .PRT le.
Dening the grid In the Norne case Corner Point geometry is used. It requires that all the
corner point are given, but there is no requirement for the corner angles to be right.
This model consists of 46x112x22 grid blocks in the x-, y- and z-direction, respectively.
The coordinate system that denes the grid is given in UTM, Universal Transverse Mercator,
coordinate system for the x and y coordinates and depth in meters for the z coordinates.
The grid is dened in the IRAP_1005.GRDECL le. The rst keyword in this le is the
specgrid keyword. This keyword repeats the specication of dimensions, number of reservoirs
and type of coordinates dened in the runspec section, it is an optional keyword used only
to control the settings. The rst item is the number of grid blocks in the x-direction, second
the number of grid blocks in the y-direction and thirdly comes the number of grid blocks in
the z-direction. Item number four and ve are number of reservoirs and type of coordinates,
respectively. The coordinate type is either cylindrical (T) or cartesian (F). For the Norne Field
the values are as follows; 46 112 22 1 F.
The next step is to dene coordinate lines between two points, which is done under the
coord keyword, see gure below. These lines dene possible positions for the grid block corner
points. The depth of each corner point is given in the same le under the zcorn keyword,
see gure below. With this information the x- and y-coordinates for the corner points can be
calculated, hence specifying all the grid blocks in the model.
98
Active cells
The entire model consists of 113344 cells, where 44431 are active cells. To dene active cells,
integers 0 or 1 are used for each cell in the ACTNUM_0704.prop le under the actnum keyword.
Active cells are assigned the value 1, while the inactive cells are assigned the value 0. Grid blocks
are ordered with the index for the X-axis cycling fastest, and then followed by the Y- and Z-
axis, respectively. Starting with block (1,1,1) moving to block (2,1,1) then, for a 2x2x2 system,
moving on to (1,2,1) and (2,2,1) before moving to (1,1,2), (2,1,2), (1,2,2) and nally (2,2,2).
Faults
The faults are dened in the FAULT_JUN_05.INC le under the faults keyword. First the
fault name is given. Then the position of the fault, which grid blocks it is connected to, by giving
the lower and upper I-, J- and K-values of the grid blocks. Finally the face of the fault, which
states what side of the grid block the fault is connected to, is dened. It is done by entering the
name of the face X, Y or Z or the corresponding negative face.
To set the transmissibility of the fault the keyword multflt is used. This keyword is
presented in the FAULTMULT_AUG-2006.INC input le. It is followed by the fault name and
the corresponding transmissibility multiplier. The multiplier in this eld ranges from 0.00075 to
20, where a low multiplier seals the fault.
Porosity The porosity is imported from the geological model and is calculated for each grid
block in the model.
The porosity is included in the Eclipse le with keyword poro in the PORO_0704.prop le.
A part of the included porosity le can be seen below.
99
Net-to-gross Net-to-Gross Ratios are dened in the grid section with the keyword ntg.
Net-to-gross values are calculated for each reservoir zone.
The ntg keyword is followed by a non-negative real number for every grid block, as a
fraction. Gross thickness is the thickness of the rock between top and bottom. The amount of
gross thickness that is of reservoir quality is called Net thickness. To convert from gross to net
thicknesses the values specied in the le NTG_0704.prop are used. These converted values
act as multipliers of grid block pore volume and transmissibilities in the X and Y directions.
In addition, the values are used on DZ for the calculation of well connection transmissibility
factors. The ntg keyword with input data from the Norne le can be seen below.
The grid blocks are ordered with the X-axis index cycling fastest, followed by the Y- and
Z-axis indices.
Permeability Permeability is dened under the keywords permx, permy and permz in the
PERM_0704.prop le. The values are calculated for each reservoir zone. The permeability is
an arithmetic average of the permeability in the net sand interval. [Statoil, 2001c]
permx species the permeability values in the X-direction as shown in the gure below. All
values for Y- and Z-direction are copied from the permx array under the copy keyword.
In addition, equals and multiply keywords are used to specify the permeability for the
various segments, wells and layers. By using equals, the array is set to a constant in the
current box. The rst item after the keyword denes the name of the array to be modied, the
second item states the constant to be assigned to the array specied by item 1. Items 3-8 are
used to redene the input box for this and subsequent operation within the current keyword.
Item 3 points out the number of the rst block that is modied on the X axis, while item 4
declare the last modied block on the X-axis. Item 5 and 6 denes the same on the Y axis, and
subsequent item 7 and 8 are doing the same for the Z-axis. multiply have identical method of
use as equals, but now the array is being multiplied by item 2. The next gures show how this
is done.
100
Transmissibilities between layers To dene the transmissibility in the z-direction between
the grid blocks, a le with multipliers is created. This le, MULTZ_HM_1.INC, consists of
the multz keyword followed by one number for each grid block. In this case the numbers are
mainly ones and zeroes. For some areas the transmissibility in the z-direction has been altered
as part of history match studies. The altered le is called MULTZ_JUN_05_MOD.INC. In
this le the equals keyword is used to alter some of the multz or transmissibilities between
certain layers. There the rst value is the array to be modied, in this case the multz, next is
the new value and nally the X-, Y- and Z-ranges of the grid blocks.
Flux regions and transmissibilities The FLUXNUM_0704.prop le is used to dene re-
gions in the model. Each cell is given an integer from 0-20 under the fluxnum keyword. 1 is
default, see gure below. These 20 regions can acquire dierent properties independent of the
other regions. A region can also be run separately from the entire model using ux boundaries.
In the Norne Field there has been dened four regions for each geological layer; Garn, Ile, Tofte,
Tilje-top and Tilje-bottom. The regions are C, D, E and G, where the rst three belong to
the main structure, while the G region belongs to the smaller structure north-east of the main
structure.
Transmissibilities between neighbouring regions can also be dened. In Eclipse this is carried
out by using the multregt keyword. The le MULTREGT_D_27.prop contains the speci-
cation of this, by rst setting the region number to start from, then the region number of the
last region. Finally the transmissibility multiplier that is to be used between these two regions
is set. This is given in table 4.5.
101
Table4.5:
Tansm
issibilitiesbetweenregions,from
includeleMULT
REGT_D_27.prop
Garn
Ile
Tofte
Tilje4&
3Tilje2&
1F orm
ation
CD
EG
CD
EG
CD
EG
CD
EG
CD
EG
Segm
ent
12
34
56
78
910
1112
1314
1516
1718
1920
Fluxnum
FLUX
11
10.005
00
00
00
00
00
00
00
00
1C
Garn
FLUX
11
10
00
00
00
00
00
00
00
02
DFLUX
11
00
00
00
00
00
00
00
00
3E
FLUX
10
00
00
00
00
00
00
00
04
GFLUX
11
10.01
11
11
0.1
0.1
0.1
0.01
00
00
5C
Ile
FLUX
10.05
11
11
10.1
10.1
0.1
00
00
6D
FLUX
11
11
11
0.1
0.1
0.1
0.1
00
00
7E
FLUX
11
11
10.1
0.1
0.1
0.1
00
00
8G
FLUX
11
10.01
11
11
0.001
00
09
CTofte
FLUX
11
11
11
10
10
010
DFLUX
11
11
11
00
0.001
011
EFLUX
11
11
10
00
112
GFLUX
11
10.01
0.0008
00
013
CTilje4&
3FLUX
11
10
0.1
1*10−
60
14D
FLUX
11
00
0.05
015
EFLUX
10
00
0.001
16G
FLUX
11
10.1
17C
Tilje2&
1FLUX
11
118
DFLUX
11
19E
FLUX
120
G
102
4.4.3 edit
The edit section includes instructions for modications to calculated pore volumes, grid block
centre depths, transmissibilities, diusivities and non-neighbour connections computed by the
program from data in the grid section.
Keywords in the edit section
In the Norne case, modications in the edit section are connected to transmissibilities for dierent
wells and faults. The keywords to overwrite transmissibility array values used are tranx and
trany. These keywords are used through the operational keywords multiply and equals.
tranx and trany are the transmissibility for the current input box in respectively X and Y-
directions. An example is shown in gure below.
4.4.4 props
The props section contains input of uid properties and relative permeability of the reservoir.
Multi-tabular keywords are used, and only one entry of any keyword is accepted. The runspec
section of the le has specied which tables that are needed and the maximum size of these.
The correct length and number of tables must be provided.
Keywords in the PROPS section
The noecho keyword is disabling the echo of the data input, see explanation in section 4.4.2.
PVT and rock properties PVT properties are given by use of the PVT keywords, and are
included in the le called PVT-WET-GAS.DATA. Two PVT regions are present in the model.
Region 1 includes the C-, D- and E-segments, while region 2 consists of the G-segment.
pvtg denes tables with PVT properties of wet gas. Item 1 gives the gas phase pressure
Pg given in bar, item 2 is the vaporized oil-gas ratio for saturated gas at pressure Pg. The gas
formation volume factor for saturated gas at Pg is item 3 and the last item gives the gas viscosity
for saturated gas at Pg in centipoise(cP). The gure below shows how this is done.
103
The pvto keyword is used for PVT properties of live oil. The data is given as 4 numbers.
The rst number is the dissolved gas-oil ratio, Rs. The second is the bubble point pressure, Pbubfor oil with dissolved gas-oil ratio given by Rs. The oil formation volume factor for saturated
oil at Pbub is entered in place 3, while the oil viscosity for saturated oil at Pbub is given as item
4. A part of the input le can be seen below.
Water PVT functions are given by use of the pvtw keyword. Item 1 gives reference pressure
Pref for items 2 and 4, thereafter the water formation volume factor, Bw at reference pressure
(Pref ) is dened. Water compressibility is the third item, and water viscosity at reference
pressure the fourth. The last item includes the water "viscosibility" which is zero in the Norne
case. A part of the input le can be seen below.
The rock keyword denes the compressibility of the rock for each pressure table region.
Each record can consist of 6 items of data, but in the Norne le there are only two items. The
rst is the reference pressure (Pref ) and the second the rock compressibility.
Surface densities of the reservoir uids for the two PVT regions are given under the density
keyword. Three numbers are used, respectively values for oil, water and gas densities.
Set up of tracers are done by use of the tracer keyword as shown below. Each tracer is
associated with a particular uid used in the run. The keyword is followed by one line for each
tracer, which includes the name of the tracer and the name of the uid connected to the tracer.
7 tracers are introduced in the Norne le; all of these have water as uid.
Relative Permeability and Capillary Pressure The swof_mod4Gseg_aug-2006.inc le
contains data for oil-water imbibition curves. Drainage curves are equal to imbibition curves.
The swof keyword is used in runs where both oil and water is present as active phases. It is
applied by including tables containing the following 4 columns; water saturation, water relative
permeability, oil-in-water relative permeability and water-oil capillary pressure. The rst value
104
in column 1 is interpreted as the connate water saturation, while the last value is interpreted as
Sw=1-Sor. Two dierent relative permeability curves are included for the oil-water system; one
for the Tofte Formation, and one for the remaining formations. A part of the le is show below.
The gas-oil drainage curves are dened in the sgof_sgc10_mod4Gseg_aug-2006.inc le where
the keyword is sgof and includes gas-oil saturation functions versus gas saturation. Each table
consists of 4 columns where column 1 is the gas saturation, column 2 the corresponding gas
relative permeability, column 3 the corresponding oil relative permeability and the last column
the corresponding oil-gas capillary pressure.
WAG hysteresis model Wag hysteresis parameters model is activated by using thewaghystr
keyword in the waghystr_mod4Gseg_aug-2006.inc le. This enables a better modelling of the
WAG injectors. The required data for the model is presented here. The keyword is followed
by 8 items of data. Item 1 is Land's parameter, C, and governs the trapped gas saturation on
imbibition and the shape of the imbibition curve. The following equation is used:
Sgtrap = Sgcr +(Sgm − Sgcr)
(1 + C ∗ (Sgm − Sgcr))
where
Sgtrap is the trapped gas saturation, Sgm is the maximum gas saturation attained and Sgcr is
the critical gas saturation.
Item 2 is the secondary drainage reduction factor, α. The third item is the gas model ag,
where YES indicates that WAG Hysteresis Model for the gas phase relative permeability is used,
while NO means that the WAG Model is turned o and drainage curves are used instead. In the
Norne case, WAG Model is used. The fourth item is the residual oil ag. If modication of the
residual oil in the STONE 1 3-phase oil relative permeability model is needed, YES indicates
that trapped gas saturation will be used for this. NO will not modify the oil relative permeability
this is the case for the Norne simulation. Item 5, called water model ag has YES and NO as
options. If YES is typed the WAG Hysteresis Wetting Model is applied to the water phase,
while NO indicates that the WAG hysteresis model is not applied. The WAG hysteresis model
is not applied for the Norne case. Item 6 has the number 0.1 which is the imbibition curve
linear fraction. This is the fraction of the curve between Sgm and Sgtrap that uses a linear
transformation. 3-phase model threshold saturation is given in item 7 as 0.1. The nal item
gives the residual oil modication fraction as 0.0. A part of the input le can be seen in the
gure below.
105
4.4.5 regions
Reservoirs can have dened dierent regions with certain, common properties; for instance
uid in place, saturation table number, imbibition saturation function number, PVT data or
equilibration. The regions section divides the computational grid into such regions.
Keywords in the regions section
Fluid-in-place regions Fluid-in-place regions are dened by use of fipnum keyword. Each
grid block is given a region number, see gure below. All grid blocks in the same region share the
same initial uid in place volumes/saturations. For the Norne Field there are 16 dierent regions
dened in FIPNUM_0704.prop. The uids in place for these regions are given in table 4.6.
Additional Fluid-in-place regions As an addition to the Fluid-in-place regions dened un-
der the fipnum keyword, Fluid-in-place regions based on the geological- and numerical layers
are dened. The keywords used are fipgl and fipnl for geological and numerical layers respec-
tively. How the formations are divided into the fipgl and fipnl regions are shown in tables 4.7
and 4.8.
106
Table4.6:
Fluid-in-placeforeach
region
from
includeleBC0407.PRT
Region
OOIP
[Sm
3]
OOIP
[Sm
3]
OOIP
[Sm
3]
WOIP
[Sm
3]
GOIP
[Sm
3]
GOIP
[Sm
3]
GOIP
[Sm
3]
Pressure
Porevolume
number:
Liquid
Vapor
Total
Total
Free
Dissolved
Total
[barsa]
[Rm
3]
15357744
311330
5669074
11495451
5426677329
591643401
6018320730
268.94
44732385
23229039
88235
3317273
3087927
1534993705
362558460
1897552164
269.03
14752345
32280208
60909
2341116
13980845
1049854062
256698501
1306552563
269.64
22490709
45279750
1062
5280812
7625996
19474906
497484252
516959158
267.68
14729288
541754724
89346
41844070
22392417
1549276984
4587679020
6136956005
270.87
85477884
611111392
6308
11117700
4241475
109292523
1215300430
1324592952
271.26
19512663
710874378
3068
10877446
13886905
53139636
1182176434
1235316070
272.23
28917966
80
00
9038812
00
0274.66
9384376
947567775
047567775
45677106
05109706183
5109706183
273.78
109569600
1014007759
014007759
9657870
01503314643
1503314643
273.89
28325189
1111412025
011412025
33667931
01223933869
1223933869
274.12
49835664
120
00
30552206
00
0275.23
31718517
134818368
04818368
112636220
0514599558
514599558
275.37
123139730
141427140
01427140
25011391
0152394634
152394634
275.62
27813651
151136787
01136787
49097756
0121386548
121386548
275.52
52405951
160
00
9966663
00
0280.08
10342360
107
Table 4.7: Numerical layers from include le EXTRA_REG.incNumerical
Region number layer number Formation name
1 1 Garn 32 2 Garn 23 3 Garn 14 4 Not5 5 Ile 2.26 6 Ile 2.1.37 7 Ile 2.1.28 8 Ile 2.1.19 9 Ile 1.310 10 Ile 1.211 11 Ile 1.112 12 Tofte 2.213 13 Tofte 2.1.314 14 Tofte 2.1.215 15 Tofte 2.1.116 16 Tofte 1.2.217 17 Tofte 1.2.118 18 Tofte 1.119 19 Tilje 420 20 Tilje 321 21 Tilje 222 22 Tilje 1
Table 4.8: Geological layers from include le EXTRA_REG.incFrom numerical To numerical
Region number layer number layer number Formation name
−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Input o f f l u i d p r op e r t i e s and r e l a t i v e pe rmeab i l i ty
−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− equ i l i b r ium data : do not inc lude t h i s f i l e in case o f RESTART
−−−−INCLUDE
' . /INCLUDE/PETRO/E3 . prop ' /
−− r e s t a r t date : only used in case o f a RESTART, remember to use SKIPREST
−−RESTART−− 'BASE_30−NOV−2005 ' 360 / AT TIME 3282.0 DAYS ( 1−NOV−2006)
THPRES
1 2 0.588031 /
2 1 0.588031 /
1 3 0.787619 /
3 1 0.787619 /
1 4 7.00083 /
4 1 7.00083 /
/
−− i n i t i a l i s e i n j e c t e d t r a c e r s to zero
−− 01 . 01 . 07 new VFP curves f o r producing we l l s , matched with the l a t e s t we l l t e s t s in Prosper . lmarr
−−INCLUDE
' . /INCLUDE/VFP/B1BH. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/B2H. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/B3H. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/B4DH. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/D1CH. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/D2H. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/D3BH. Ecl ' /
−−
82
INCLUDE
' . /INCLUDE/VFP/E1H. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/E3CH. Ecl ' /
−−INCLUDE
' . /INCLUDE/VFP/K3H. Ecl ' /
−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−=======Production F lowl ine s========−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− 16 . 5 . 02 new VFP curves f o r southgoing PD1,PD2,PB1 ,PB2 f l ow l i n e s −> pd2 .VFP
−−INCLUDE
' . /INCLUDE/VFP/pd2 .VFP' /
−−−− 16 . 5 . 02 new VFP curves f o r northgoing PE1 ,PE2 f l ow l i n e s −> pe2 .VFP
−−INCLUDE
' . /INCLUDE/VFP/pe2 .VFP' /
−− 24 . 11 . 06 new matched VLP curves f o r PB1 va l i d from 01 .07 . 06
−−INCLUDE
' . /INCLUDE/VFP/PB1 .PIPE . Ecl ' /
−−24.11.06 new matched VLP curves f o r PB2 va l i d from 01 .07 . 06
−−INCLUDE
' . /INCLUDE/VFP/PB2 .PIPE . Ecl ' /
−−24.11.06 new matched VLP curves f o r PD1 va l i d from 01 .07 . 06
−−INCLUDE
' . /INCLUDE/VFP/PD1.PIPE . Ecl ' /
−−24.11.06 new matched VLP curves f o r PD2 va l i d from 01 .07 . 06
−−INCLUDE
' . /INCLUDE/VFP/PD2.PIPE . Ecl ' /
−−24.11.06 new matched VLP curves f o r PE1 va l i d from 01 .07 . 06
−−INCLUDE
' . /INCLUDE/VFP/PE1 .PIPE . Ecl ' /
−−24.11.06 new matched VLP curves f o r PE2 va l i d from 01 .07 . 06
−−INCLUDE
' . /INCLUDE/VFP/PE2 .PIPE . Ecl ' /
−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−=======INJECTION FLOWLINES 08 .09 .2005 ========−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− VFPINJ nr . 10 Water i n j e c t i o n f l ow l i n e WIC
−−INCLUDE
' . /INCLUDE/VFP/WIC.PIPE . Ecl ' /
−− VFPINJ nr . 11 Water i n j e c t i o n f l ow l i n e WIF
−−INCLUDE
' . /INCLUDE/VFP/WIF.PIPE . Ecl ' /
−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−======= INJECTION Wells 08 .09 .2005 ========−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− VFPINJ nr . 12 Water i n j e c t i o n we l lbore Norne C−1H−−INCLUDE
' . /INCLUDE/VFP/C1H. Ecl ' /
83
−− VFPINJ nr . 13 Water i n j e c t i o n we l lbore Norne C−2H−−INCLUDE
' . /INCLUDE/VFP/C2H. Ecl ' /
−− VFPINJ nr . 14 Water i n j e c t i o n we l lbore Norne C−3H−−INCLUDE
' . /INCLUDE/VFP/C3H. Ecl ' /
−− VFPINJ nr . 15 Water i n j e c t i o n we l lbore Norne C−4H−−INCLUDE
' . /INCLUDE/VFP/C4H. Ecl ' /
−− VFPINJ nr . 16 Water i n j e c t i o n we l lbore Norne C−4AH−−INCLUDE
' . /INCLUDE/VFP/C4AH. Ecl ' /
−− VFPINJ nr . 17 Water i n j e c t i o n we l lbore Norne F−1H−−INCLUDE
' . /INCLUDE/VFP/F1H. Ecl ' /
−− VFPINJ nr . 18 Water i n j e c t i o n we l lbore Norne F−2H−−INCLUDE
' . /INCLUDE/VFP/F2H. Ecl ' /
−− VFPINJ nr . 19 Water i n j e c t i o n we l lbore Norne F−3 H
−−INCLUDE
' . /INCLUDE/VFP/F3H. Ecl ' /
−− VFPINJ nr . 20 Water i n j e c t i o n we l lbore Norne F−4H−−INCLUDE
' . /INCLUDE/VFP/F4H. Ecl ' /
TUNING
1 10 0 .1 0 .15 3 0 .3 0 .3 1 .20 /
5* 0 .1 0 .0001 0 .02 0 .02 /
−−2* 40 1* 15 /
/
−− only p o s s i b l e f o r ECL 2006.2+ ve r s i on