Re S
Yuefe
eserv Snøh
feng Gao
voir hvit
Ba
o
Cha Field aren
racte d, No nts S
eriza orwe
ea
ation egian
n of
n
R Sn
Rese øhvi
Fa
ervoi it Fie
Ma
D aculty of
ir Ch eld, N
Y
aster The Discip Departme
f Mathem
Unive
01
harac Norw
Sea
Yuefeng
esis in G pline: Ge ent of Ge matics and
ersity of
1.06.201
cteriz wegia
Gao
eoscienc eology
oscience d Natura
f Oslo
3
zatio an B
es
s
l Science
on of Baren
es
f
nts
© Yuefeng Gao, 2013
This work is published digitally through DUO – Digitale Utgivelser ved UiO http://www.duo.uio.no
It is also catalogued in BIBSYS (http://www.bibsys.no/english)
All rights reserved. No part of this publication may be reproduced or transmitted, in any form or by any means, without permission.
I
Acknowledgements
With great help of my two supervisors, Nazmul Haque Mondol and Manzar Fawad, this thesis is finished in its best version. Their kind and reliable teaching and conscientious supervision are the foundation stone of this thesis. The most important thing they taught me is not specific geological concepts but the way to think and solve problems independently as well as cooperatively.
I also want to thank Michael Heeremans and IT staff of Department of Geosciences for their data management works and software support. They created a fast and efficient working environment.
My group fellows, Fahad Ashraf, Shahzeb Haider and Omer Saeed: we have spent time together during the study tenure. I would like to express my appreciation for every discussion and every help we provided to each other.
My parents are my strong backups. I give thanks for their feeding and teaching when I was young and for their unconditional support at the beginning of my study abroad career. Their selfless love and encouragement accompany me every day.
I will leave the last acknowledgement for my beloved soul mate Yujing Yu. Her tender consolations and inspirations helped me to get through the toughest times.
Yuefeng Gao
II
Abstract
Snøhvit field is the only field in production in Norwegian Barents Sea. Most of the hydrocarbons are gas. Complicated tectonic history of this field resulted in low matured source rocks, highly faulted structures, low porosities and trap breach.
Three reservoirs were studied in this project; Stø, Nordmela and Tubåen Formations.
Stø is the main producing formation in Snøhvit field. It is gas saturated and formed in shallow marine environment. Most sections of Stø Formation are composed of very clean sands. Properties in Nordmela are much worse for its high volume of shale and clay contents formed in complex estuary and tidal environments. Only minor gas and oil were found in it. Clay studies show evidence of deep burial and uplifted history in this region.
Although oil and gas did not exist in fluvial Tubåen Formation, it is the main object of study because it was the target formation of the previous CO2 injection project operated by Statoil. Four zones were divided in this formation but the uppermost one represents a transgression event rather than fluvial environments. Models of this formation explained the storage capacity problems.
Different methods were used to estimate porosities and permeabilities for all three reservoirs. Calculations show previous overestimation of permeabilities and connectivity for channels. Local good CO2 warehouses were isolated and the surrounding worse reservoirs become barriers under high injection pressure.
Statoil started a new injection target which is Stø Formation. A preliminary and qualitative analysis shows thick and widespread sandstones. These sandstones may support the long lasting injection project.
III
Table of Contents
Acknowledgements ... I
Abstract ... II
Table of Contents ... III
Chapter 1 Introduction ... 1
1.1 Background ... 1
1.2 Motivation and Objectives ... 3
1.3 Study Area ... 3
1.4 Database ... 4
1.5 Limitations ... 6
1.6 Chapter Descriptions ... 6
Chapter 2 Geology of the Study Area ... 8
2.1 Tectonic History and Geological Evolution ... 8
2.2 Stratigraphic Descriptions ... 11
2.3 Petroleum Systems ... 16
Chapter 3. Methodology and Theoretical Background ... 24
3.1 Borehole Structure ... 25
3.2 Log Editing and Quality Control ... 26
3.3 Shale Volume Calculation ... 28
3.4 Porosity Estimation ... 30
3.5 Permeability ... 31
3.6 Crossplot ... 34
3.7 Polarity and Phase in Seismic Record ... 34
3.8 Variogram ... 35
Chapter 4 Well Log and Seismic Interpretation ... 38
4.1 Lithology Characterization ... 38
4.2 Facies Analysis from Well Log ... 51
IV
4.3 Horizons and Faults Interpretation ... 54
Chapter 5 Reservoir Properties Study ... 60
5.1 Basic Parameter Determinations ... 60
5.2 Archie’s Law and Porosity ... 61
5.3 Neutron Porosity and Density Porosity ... 62
5.4 Permeability Estimation and Calculation ... 69
Chapter 6 Reservoir Models ... 75
6.1 Fluvial Settings ... 75
6.2 Facies Models ... 84
6.3 Property models ... 86
Chapter 7 Discussion and Conclusion ... 90
References ... 93
1.1 Bar The Nor wes cent and trea Sva refe kno
F
In 2 sout equ
Backgro ents Sea lie e total area rway and R stern and ea
ter (Smelro
‘western B ats it as an albard is alw ers to the ar own as the ‘T
Figure 1.1 Loc
2013, accord thwest Bare ivalent), wi
C
und es north of N
of the Bar Russia. In astern part b or et al., 20 Barents Sea’
important ways consid
reas betwee Treaty area
cation map fo
ding the ca ents Sea has
ith an unce
Chapter
Norway and ents Sea is
geological by a huge m 009). But tr
’ are almost potential ar dered separa en the Norw
’ according
r the southwe Harland a
alculation o s expected r ertainty rang
r 1 Intr
d Russia in about 1.3 concepts, monoclinal s
raditionally, t synonyms area for oil
ately, the w wegian coas to the Svalb
est Barents Sea and Dowdesw
of Norwegia resources to ge of 55 -
roduction
the southern million km
Barents Se structure (ce , the words
and often r and gas ex ord ‘southw st and 74°N
bard Treaty
a and the impo well, 1988)
an Petroleu otal approx
565 million
n
n part of the m2 and explo ea has bee entral Baren
‘Norwegia replace each xploration s west Barents N. The north
y (Figure 1.1
ortant islands
um Director . 300 millio n Sm3 o.e.
he Arctic Oc ored mainly en divided
nts High) in an Barents h other. Nor since 1979 s Sea’ norm hern part is 1).
(modified aft
rate (NPD) on Sm3 o.e.
The petrol cean.
y by into n the Sea’
rway . As mally
also
ter
, the (oil leum
2
volu Bar 27.0 Sea the
Figu platf
Up in th basi was 1.2) imp perm geo In 2 was simu Lac and
Res
umes presen ents Sea are 02.2013). C has a more Barents Sea
ure 1.2 Princip forms, basins
until now, S he SW Bare in. In 2000, s also locate ). Because portant and meability a logical even 2008, Statoi s chosen to b
ulate and pr ckner, 2009;
Tappel, 20
ervoir charac
nt could be e expected t Compare wi e complex g
a is compar
pal structural e and the main
Snøhvit fiel ents Sea (N another ma ed in the Ha of the lon difficult tas nd hydroca nts such as il started the
be the targe redict the st
; Pham and 004) estimat
cterization of
considerabl o be gas. Ab th North Se geological e
atively limi
elements and fault systems
d, which wa NPD facts, 2
ajor oil disc ammerfest b ng-lasting co
sk is to pre arbon satura
subsidence, e CO2 injec et formation torage pote Aagaard, 2 ted that in th
f Snøhvit Field
ly greater. M bout 15 perc ea and Nor evolution hi
ited.
discoveries in (Ostanin et al
as discovere 2013). This
overy in No basin, 50 ki
omplex tec edict reservo
ation and t , uplift and
tion program n. In the foll ential of CO 2011; Pham he followin
d, Norwegian
Most of the r cent are exp rwegian Sea story but th
n the SW Bare l., 2012)
ed in 1984, field is loca orwegian Ba ilometers so ctonic histo oir properti try to find erosion.
m in Snøhv lowing year O2 in Tubåen
et al., 2011 ng 30 years,
n Barents Sea
resources in pected to be a, the south he geologica
ents Sea. Outli
is the only ated inside t arents Sea k outheast of S ories of Snø es which in out its rela vit field. Tub rs, theses we n Formation 1). An early
Tubåen Fo
a
n this part o oil (NPD N heastern Bar al knowledg
ined are the h
field develo the Hamme known as G Snøhvit (Fi øhvit field, nclude poro ationship to båen Forma ere publishe n (Estubiler y model (Ma
ormation sh f the News,
rents ge of
ighs,
oped rfest Goliat gure , the osity, o the
ation ed to r and aldal hould
be able to store approximate 23 million tons of CO2. However, according to the numerical model by Pham et al., 2011, the limited lateral permeability will make the bottom hole pressure increase much higher than the fracture pressures, thus they conclude the previous aim of 23 million tons is ‘unrealistic’.
Compared with other CO2 injection projects on Norwegian coast, operation in Snøhvit has more challenges, not only because of the extreme weather conditions, but also due to the less understanding of the target horizon. In 2012, Statoil announced that they have already plugged off Tubåen Formation and moved the target of injection to Stø Formation (Gilding et al., 2012), because Tubåen Formation did not ‘show the extra CO2 capacity needed’ after perforation during the intervention of the only injection well 7121/4-F-2 H.
1.2 Motivation and Objectives
Despite Statoil abandoned Tubåen Formation for now, it is still necessary to study and analysis the properties of it, as the presence consequences strongly suggest the lack of understanding of Tubåen for CO2 storage perspective. Furthermore, although the only CO2 injection well (7121/4-F-2 H) changed its target, well 7120/6-2 S has been suspended since 2007, for a possible future injector. Although Tubåen Formation has good sandstones and relative high net-to-gross ratio, sandstones were commonly separated into several thin layers by shales. The complex depositional environment made difficulties to predict reservoir properties.
Because of the existence of oil and gas, Stø and Nordmela Formation were researched in detail in the past years. Besides property study for all three reservoirs, the main aim of this study is providing geological models for Tubåen Formation in order to understand the distribution of different properties. The final results include the property analysis and facies simulations.
1.3 Study Area
Besides Snøhvit, several other discoveries are also located in Hammerfest basin (Figure 1.2), but this study will not involve them.
Snøhvit field became a hot research topic in geological, geophysical and engineering areas in the past few years (Berger et al., 2003; Maldal and Tappel, 2004; Eiken, 2005;
Eiken et al., 2011; Pham et al., 2011). Geologists and geophysicists are interested in the intricate geological history and features in the southwest Barents Sea, while engineers are interested in developing special operations and production techniques.
Based on the present data, Snøhvit field contains 211.13×106 m3 oil equivalent hydrocarbons, most of which is gas, with a small amount of natural gas liquids (NGL) and condensate. It is common for the fields in western Barents Sea that gas is much
4
mor not Abo Tap (Fig deep Unt (1,0
Figu (Gild
1.4 Thr the Som
‘Snø 712 712 disc follo unle The avai 712
Res
re than oil.
include rec out 5–8 mo pel, 2004), gure 1.3). A per Tubåen til it has b 000,000 Sm
ure 1.3 Snøhv ding et al., 20
Database ee kinds of public infor me definitio
øhvit field’
20/6, 7121/4 21/4 was n
coveries lik owing text, ess it is expr ere are 15 ex ilable in thi 21/4-2 and 7
ervoir charac
Although s overy of the ol% CO2 is
the well stre fter CO2 sep Formation een plugge
3), has been
it field and th 12)
e
f data have rmation.
on should b contains 8 4 and 7121 named by t ke Askeladd
the word ‘S ressly stated xploration w is study. Th 7121/5-1. Al
cterization of
ome oil tra e oil zone.
mixed with eam has bee paration, it h
at a depth o ed off, over n injected an
e LNG (liquefi
been applie e clarified b
discoveries 1/5 are also the well 71 d and Alba
Snøhvit’ is d.
wells were d hey are 7120 ll wells wer
f Snøhvit Field
aces have be h natural ga en transport
had been se of approxim
r 1.1 millio nd stored.
fied natural ga
ed in this st before intro s and comp o called ‘S 121/4-2 an atross Sør a
specific ref drilled in Sn
0/5-1, 7120 re drilled in
d, Norwegian
een found, t as, NGL an ed over 160 ent back to th mately 2600
on tons CO
as) plant on M
tudy: well l oducing the prised 8 blo Snøhvit’. An
d called ‘S also belong fer to the dis nøhvit field.
0/6-1, 7120/
n 80’s excep
n Barents Sea
the develop nd condensa 0 km in pipe
he field to r meters (NP O2, or 2000
Melkøya Island
logs, 3-D se e data. Acco ocks. The di
nother disc Snøhvit No to ‘Snøhv scovery ‘71 . Data from 6-2 S, 7120 t 7120/6-2 S
a
pment plan ate (Maldal eline to Melk
re-inject into PD facts, 20
0 tons per
d near Hamme
eismic data ording to N iscovery bl covery in b ord’. The o vit field’. In 121/4-1 Snø seven well 0/8-4, 7121/
S and 7120/
does and køya o the 013).
day
erfest
a and NPD, ocks block other n the
hvit’
s are /4-1, /8-4.
The gam
W na 712
712
7120 712 712 712 712
The surv The to u crop Apa stud
erefore they mma ray log
Well ame
Com n 20/5-1 06.0
20/6-1 02.0
0/6-2 S 22.0 20/8-4 0.12 21/4-1 27.1 21/4-2 14.0 21/5-1 28.0
e 3-D surve vey. The sur e original da use such det pped seismi art from we dy is mainly
Fi
y have more gs.
Table
mpletio date 06.1985
05.1985
07.2007 2.2007 10.1984 04.1985 Ga 09.1985
y was taken rvey covere ata are store tailed recor ic volume w ell logs and y from the p
gure 1.4 Loca
e complete
e 1.1 Summar
Content
Shows
Oil/Gas
Oil/Gas Dry Oil/Gas s/Condensate
Oil/Gas
n in 1997.
ed approxim ed as 32 bit
rds for horiz with 8 bit da
seismic vol public conte
ations of wells
logging da
ry of 7 explora Total dep (MD) [m RKB
2700.0
2820.0
3242.0 2697.0 2609.0 2800.0 3200.0
Only 5 wel mately 486 k
length with zon interpre ata and 3000 lume, the o ents of NPD
s and the area
ata such as
ation wells by pth
B]
O form 0 Fruh
0
Tubå discu Cha
0 S
0 Fruh 0 Fruh 0 Fruh
0 S
lls are locat km2 and cro h 5000 ms T etation and 0 ms TWT i other useful D.
as covered by 3
S-Sonic lo
y NPD
Oldest mation
holmen åen* (see ussions in apter 4)
1
nadd 2
holmen holmen holmen 7
nadd 1
ted within t ossed 4 bloc TWT, but it i modeling.
is used.
information
3-D seismic d
og and Spe
Core
6 cores, 138.5
12 cores, 184.
2 cores, 114.1 N/A 4 cores, 95.4 7 cores, 134.3 10 cores, 188.
the areas of cks (Figure
is not neces In this stud n related to
data
ctral
es
5m in total
1m in total
1m in total A
m in total 3m in total 6m in total
f the 1.4).
ssary dy, a o this
Reservoir characterization of Snøhvit Field, Norwegian Barents Sea
6
1.5 Limitations
1.5.1 Limitations in Facies Study
This study is mainly based on well log and 3-D seismic data. But in order to study the geological features, researches in different methods and subjects should be considered and combined together. Field trips, well core examinations and paleontology studies are crucial to determine sedimentary facies. Inspections of rock slices under microscopes are the key steps to estimate reservoir properties. These direct observations provide not only the geophysical parameters, but more importantly, they gave answers that cannot be measured or calculated from well logs and seismic data. E.g. paleoclimate, grain sorting, cementation, shapes of pores and throats, distribution of clay inside rocks, and so on.
1.5.2 Limitations in Property Study and Modeling
Log traces can be used to read or calculate reservoir properties such as density and porosity, but none of a method can provide the accurate parameters. For permeability, because of the lack of laboratory report, empirical formulas were used to do the calculation, but the results may not precisely reflect the rock behavior in the study area since the Hammerfest basin has a complex tectonic history. As a matter of fact, every empirical formula can only reflect a specific rock sample under the specific experimental condition.
To modeling facies and properties, sufficient previous researches are needed. For example, previous artificially facies maps are very useful for 3-D facies modeling.
When previous results are not available or not sufficient, well-controlled facies belts give good results, too. However, in this study, even the well numbers are limited, variogram studies gave large ranges. The original ideas of variograms turned to mathematics. Geological features could be deviated.
1.6 Chapter Descriptions
This thesis has been subdivided into 7 separate chapters. The first chapter is a general introduction to the study area and an overview of the numbers of Snøhvit field. Chapter 1 also discussed the limitation of this study.
Tectonic backgrounds and Geological evolution processes with a special emphasis on the Hammerfest basin is the main part of Chapter 2. Moreover, Chapter 2 gave a brief list and description of the stratigraphy and petroleum system in the study area.
Chapter 3 reviewed general theories of well log interpretation and background of property estimation. These theories have been used later to discussion findings of the research.
Chapter 4 gives the results from well log and seismic data include lithology correlation, sedimentary facies interpretation, seismic interpretation and a velocity model.
Theories and applications of porosity and permeability estimation and correlation are the main focus of Chapter 5
Seismic data combined with the previous petrophysical analysis, results of facies and property models are discussed in Chapter 6.
Finally, a summary of the entire research and major conclusions are given in Chapter 7.
Reservoir characterization of Snøhvit Field, Norwegian Barents Sea
8
Chapter 2 Geology of the Study Area
Southwest Barents Sea has over 15 km sediments accumulations and thought to have some deepest basin in the world (Faleide et al., 1993). Along the shelf, a series of Paleozoic and/or Mesozoic basins, structural highs and platform areas were formed during the complex tectonic history. Hammerfest basin did not separate from a large regional basin until Jurassic (Berglund et al., 1986, Gabrielsen et al., 1990). Hanisch (1984) believed Hammerfest basin is the younger one which ‘overprinted’ above an older rift system. Although Caledonian orogeny dominated the large scale pattern of the whole Barents Sea, Hammerfest basin only began to accept sediments alone since late Jurassic (Faleide, 1984; Isaacson and Neff, 1999). Faleide et al. (1993) treated Hammerfest basin as one of the Mesozoic basins in SW Barents Sea and claimed the Cretaceous-Tertiary subsidence did not affect it.
2.1 Tectonic History and Geological Evolution
2.1.1 Early Tectonic Settings of SW Barents Sea (Pre-Cambrian to Devonian) Although the word ‘Barents Sea’ refers to a clear geographical position today, it has not become so until the Cenozoic time.
An important Paleo-ocean called Iapetus opened from late Vendian (600Ma) to Early Cambrian. This ocean eventually closed approximately 400 Ma ago as the result of the severe Caledonian orogeny (Figure 2.1). This process lasted until the Devonian (Gulaugsson et al., 1998) and generated the NE-SW faults of the southwestern Barents Sea (Dore, 1995). In Gabrielsen’s classification (1984) of the fault system of SW Barents Sea, these faults belong to the first class, which has regional importance and involved basement.
As the result of collision, consolidation and metamorphism happened in Laurentia and became the basement substructure of the Barents Sea from late Silurian to Early Devonian time (Berglund et al., 1986, Harland and Dowdeswell, 1988, Dore, 1995).
Devonian is the transitional phase from the previous geosynclinals sequences deformed into a Caledonian basement (Harland and Dowdeswell, 1988). Laurentia and Baltica began to separate at late Devonian. Following the Caledonian orogeny, the hinterland got exhumed and eroded from Devonian to Early Carboniferous.
Figu and D turne east north
2.1.
Exp hist Røn of t near regi plat char enti from show gav the
ure 2.1 openin Dowdeswell, ed to close, Ba
of Greenland hward, strike-
.2 Evolu ploration we ory of the H nnevik et al the faults sy
r Hammerfe ional subsid tform succe
racters in P ire Barents m the Finnm
ws evidence e an explan previous cr
ng to closing o 1988). a: Iape altica moves n . d: Baltica be -slip faults hav
ution of Ham ells in the st Hammerfest l. (1982) poi ystem aroun est basin. In dence. The ession of Permian are Sea was co mark Platfor e of Permia nation that t rust extensio
of Iapetus from etus continues northward wh ecame the part ve been forme
mmerfest B tudy area did t basin is rel
inted out th nd the basin n late Carbo
structural r Late Carb e the config ontiguous. A m (Gabriels an arid evap
his was cau on.
m Mid Cambr s to open until hile Laurentia t of Laurentia ed.
Basin (Carb d not reach lated with th he effect of s n. This cau oniferous, th
relief was g boniferous-P guration of At that time, sen et al., 19 porates (Ørn
used by diff
rian to Late De l late Cambria
stay at the equ after collision
boniferous formations he adjacent strike-slip f used comple
he study are gradually in Permian ag f the Panga Hammerfe 990). The lit n Formation
ferential the
evonian (mod n - Early Ordo uator. c: Baltic n. When it con
to Cenozoi in Early Pa structural e faulting in th
ex transpres ea continued nfilled and ge. The m aea supercon
st basin has thology in S ). Gulaugss ermal subsid
dified after Ha ovician. b: Iap ca collided the ntinued movin
ic)
aleozoic, bu elements.
he developm ssion struct d the proces
blanketed b most disting
ntinent and s been separ SW Barents son et al. (1 dence follow
rland petus
e ng
ut the ment tures ss of by a guish d the
rated s Sea 998) wing
10
Com Ham was 199 unc was Bjør chan A r drop cycl Snø Jura In l regi led mid Late whe Form Jura
Figu towa
Sub norm tran
Res
mpare with mmerfest ba s deposited 94). In seism onformity, w s not so ap
rnøya basin nge from de retrogradatio pping. Whe lic shaly an øhvit field.
assic.
late Jurassic ion much, b
a strong rift d Kimmerian
er, new norm en Hammer
mation, wh assic.
ure 2.2 N-S pr ards northwest
bsidence of mal faults nstensional
ervoir charac
Paleozoic, asin was a
and progra mic section,
which indic pparent but
n. Lithologi eep marine on sequenc en the Panga
nd sandy s No signifi c, the Early but later in M
ting phase ( n and began mal faults ap rfest basin
ich is the m
rofile across H t caused by pr
the basin c within Ha strike-slip
cterization of
the Early depocenter ading seque
a regional r cated a regio
clearly thi es show the into shallow ce at the en
aea began to shallow ma
cant depos y Kimmeria Mid Kimme
(Faleide, 19 n to accept s ppeared as a
was passin main source
Hammerfest ba rogradation. (A
ontinued in ammerfest regime acc
f Snøhvit Field
and Middl r during Ea ences exten reflection at onal uplift. F icker North e upward co w marine an nd of Late o break at E arine sequen sition break an tectonic erian, high a 84). Subsid sediments al
a result of th ng Jurassic rock of the
asin, notice th Ahmed, 2012
nto Early Cr basin are companied b
d, Norwegian
le Triassic arly Triassic nded across t the end of
From Mid t hwestward, oarsening se nd continent Triassic in Early - Midd nce formed ks appear fr
phase did n angle norma dence of Ham
lone separat he reactivati and entere e Snøhvit fi
he thickness in , modified fro
retaceous ti explained by an updo
n Barents Sea
subsidence c. Thick ope
s the basin Middle Tria o Late Trias the depoce equence, als
tal environm ndicates the
dle Jurassic d three mai from Mid T
not affect t al faults beg mmerfest ba ted with the ion of basem ed Cretaceo
eld, was de
ncrease of Tria om Gabrielsen
me. The ex to be fo oming alon
a
turned qui en marine s n (Kulke et assic showe ssic, progra enter move so indicating ment.
e sea level c, a collectio in reservoir Triassic to the Barents gan to active
asin occurre e adjacent ar ment weak p ous. Hekkin epodited in
assic formatio n, 1990).
xistence of E ormed unde ng the Sout
ieter.
shale t al., ed an ading ed to
g the was on of rs of
Mid Sea e and ed in reas.
parts ngen Late
ons
E-W er a thern
Loppa High Fault Complex at the end of Jurassic times (Berglund et al., 1986). E-W faults in Hammerfest basin were formed within the isolated blocks determined by the first and second class faults (Figure 2.3, Gabrielsen, 1984; Gabrielsen et al. 1990;
Olivier, 2009; Ahmed, 2012).
Devonian and Early Carboniferous rift related trends became re-activated during Early Cretaceous. The Aptian - Albian faults in the Ringcassøy - Loppa Faults Complex prevented Hammerfest basin from undergoing the same rapid thermal subsidence as the western Tromsø basin (Faleide et al. 1993). In total, the west part of Hammerfest basin subsided relatively faster rate compared to the east. A series of thick deep marine shales with fans were deposited. Until the Late Cretaceous, 70 to 80 million years ago, the Laramide orogeny reactivated the Kimmerian tectonic.
There was a major hiatus in Hammerfest basin from Oligocene to Miocene caused by the creation of new oceanic crust and the whole Barents shelf was uplifted (Faleide et al., 1984). For the Hammerfest basin and Loppa High area, erosions associated with Plio-Pleistocene glaciations are estimated between 1000 and 1500m (Dimakis et al., 1998). After the glacial erosions, approximately 100m soft claystones were deposited (Linjordet and Olsen, 1993).
2.2 Stratigraphic Descriptions
Although the oldest lithology record in SW Barents Sea is Soldogg Formation formed in the braided river in Devonian/Carboniferous period, none of a well in the study area has ever reached there. The most important Group in this study is Kapp Toscana Group, which contain hydrocarbon reservoirs (Stø Fm. and Nordmela Fm.) and the former formation for CO2 storage (Tubåen).
2.2.1 Lower Triassic to Upper Triassic
All seven wells reached upper Triassic formations. Snadd, Fruholmen and the lower part of Tubåen Formation deposited during this period and all belong to Kapp Toscana Group.
Snadd Formation
According to the data of reference well, this formation is thick (approximately 1000m).
Only two wells 7121/5-1 and 7120/6-2 S reached the upper part. This formation is dominated by shale but with different colors from grey at the base to red or brown at the top. Limestones and calcareous interbeds are relatively common in the lower and middle parts of the unit interbedded siltstones and sandstones are also found. Thin coaly lenses are developed locally further up (Dalland et al., 1988). The grey shale at the bottom indicates the distal marine environment and the interbedded silts and sands are thought to be related with storms, while the middle and upper parts were deposited by progradation of deltaic systems.
12 Figu Ham block
Fru The shal mid wer stud
Diff wer and prog Tub This The
Res
ure 2.3 First, s mmerfest basin ks, more local
uholmen Fo e older sect
les and coal ddle parts, b re recognize dy area, only
ferent GR re re changing fluvial sa gradation w båen Forma
s formation e formation
ervoir charac
second and thi n were third cl l and shallowe
ormation tion of this
ls appear gr but the uppe ed, from bas y Akkar and
esponses be from open andstone (R with a depoc
ation is mainly c was divide
cterization of
ird fault system lass faults due er (Gabrielsen
formation radually upw er part is mo
se to top, th d Reke mem
etween the m marine sha Reke Memb centre to the
omposed by ed into three
f Snøhvit Field
ms and their r e to the updom n, 1984).
is mostly g wards. Sand ore shaly. B hese membe mber were r
members in ales (Akkar
ber). These e south (Dal
y sandstone e sections, w
d, Norwegian
representative ming. They we
grey to dark d became th Based on log ers are Akka ecorded in w
dicate the e Member) a e represent lland et al.,
es with little with a lowe
n Barents Sea
structures. Th ere restricted w
k grey shal he main com gging respon ar, Reke and
well 7121/5
nvironment and passing t northward
1988).
amount of s er and uppe
a
he E-W faults with separated
les. Sandsto mposition in nse, 3 mem d Krabbe. In 5-1.
t and litholo g up into co
d fluvio-de
shales and c er sand-rich
s in d
ones, n the mbers n the
ogies astal eltaic
coals.
unit
separated by a more shaly interval. But recent practice has separated Tubåen Formation into four or five zones (Shi et al., 2012). Shale content increases towards the northwest where the Tubåen Formation may intercalate with a lateral shale equivalent (Dalland et al., 1988).
2.2.2 Jurassic Nordmela Formation
Nordmela Formation has a more complex lithology. The components of this formation include siltstones, claystones, shales, sandstones and little coals. The sandstones become more common at the upper part. The depositional environment varied from tidal flat to flood plain environments (Dalland et al., 1988).
Stø Formation
This formation is comprised of moderately to well-sorted and mature sandstones, some thin shales and siltstones within the formation are related with the pause of transgression.
Facies distribution maps for Stø and Nordmela Formation are well studied, as shown in Figure 2.4.
Fulgen Formation
Fuglen formation is the lower unit of the Adventdalen group. It consists of pyritic mudstone with interbedded thin limestones. The shales are dark brown in color. It was deposited in marine environments during a highstand with ongoing tectonic movements (Dalland et al., 1988). This formation has been oxygenated.
Hekkingen Formation
The formation consists of brown to dark grey shale and claystone with occasional thin interbeds of limestone, dolomite, siltstone and sandstone. Dark shales showed a deep water environment with anoxic conditions.
14 Figu Rahm
2.2.
Knu This dolo red dist betw Kol Kolj circ with wel Rin Trom
Res
ure 2.4 Paleog man, 2012)
.3 Creta urr Format s formation omite interb
to yellow b al marine e ween Knurr lje Formati lje is a form culation. Th h little lime ls. In Ham gvassøy – msø basin (
ervoir charac
geography and
ceous tion
n shows as d beds. Sands brown clays environment r and Hekkin
ion
mation which he lithology stones and d mmerfest ba Loppa Faul (Faleide et a
cterization of
d depositional
dark grey to tones are v stone genera
ts with loca ngen Forma
h deposited is consists dolomites. K asin, Kolje lt Complex al., 1993)
f Snøhvit Field
model for Stø
greyish bro vanishing in ally occurs.
al restricted ation is an u
in distal op of dark bro Kolje Form
Formation x (RLFC) a
d, Norwegian
ø and Nordme
own claysto Hammerfe The enviro d bottom co unconformit
pen marine c own to dark mation is thic n thickens as the result
n Barents Sea
ela Formations
one with thin est basin. In onment is th nditions. Th ty.
conditions w k grey shale ck (over 200 gradually w t of therma
a
s (modified fr
n limestone n the upper hought to be The base con
with good w e and clays 0m) in all s westwards al subsidenc
rom
e and part, e the ntact
water stone tudy into ce in
Kolmule Formation
Like Kolje Formation, Kolmule is also deposit in open marine environment with similar lithology, which is dark grey to green claystone and shale, silty in parts with thin siltstone interbeds and limestone and dolomite stringers. Over 700m thicknesses are measured in wells. This formation turns uplift and thinning towards Ringvassøy – Loppa Fault Complex (RLFC) because of the affection of Aptian event (Faleide et al., 1993).
Kviting Formation
Kviting Formation only appears at the eastern parts of the Hammerfest basin. 3 of 7 study wells (7121/5-1, 7120/6-2S, 7120/8-4) have reached this formation. Limestones and calcareous sandstones are the main components for this formation. The upward increasing with sandy claystones at the upper part of this formation indicates the change of deep to shallow shelf environments.
Kveite Formation
This formation was deposited almost at the same time with Kviting Formation. The green to grey shales and claystones with thin limestone and siltstone shows the deep open shelf environment. Kveite became thinner eastwards and passing into the Kviting Formation. Tuff or glauconites appear in some wells (e.g. 7121/4-2). The density of Kviting Formation is much higher than that of Kviting. Therefore the reflectors show high amplitudes in seismic sections towards east (Figure 2.5).
2.2.4 Paleogene and Neogene Torsk Formation
This thick formation (over 600m) consists of grey or greenish-grey claystones without calcareous component. Siltstone and limestone are rare. At the lower part of this formation, tuffaceous horizons are often seen. Fine grains indicate an open to deep marine shelf without significant coarse clastic supply.
Nordland Group
This is the first group once the drill bit crosses the soft sea bed. Although it has an over 1000 thickness in Viking Graben, only less than 100m is shown in Hammerfest basin.
Sandstones and claystones are typical in Barents Sea, the sand content increases upward.
16 Figu 2782
2.3
2.3.
Snø Tria Hek dist thou and The
Res
ure 2.5 Kvitin 2 (above) and
Petroleu
1 Sourc øhvit field h assic shales kkingen For al and deep ught to be r III kerogen e amount o
ervoir charac
ng and Kveite inline 2304 (b
m System
ce Rocks has three po
include Sna rmation. Bo p water mari elated with n and produ
f TOC in S
cterization of
Formation ap below).
ms
ossible sour add and Kob oth Triassic
ine environm the tidal fla ced gas wit Snadd and
f Snøhvit Field
ppear at the s
rce rock can bbe Format
shale and H ments, whil at marshes.
th little amo Kobbe For
d, Norwegian
same depth bu
ndidates. Fr tion, Jurassi Hekkingen f
le shales in All of these ount of oil.
rmations ar
n Barents Sea
ut differ in am
rom old to y c Nordmela formation w Nordmela F e source roc re between
a
mplitudes, cros
young, they a Formation was deposite
Formation w cks have typ 2-8% with
ssline
y are n and ed in were pe II h the
high rock Ham
Figu
Mal low sour not hyd Hek read orga sour diffe (He
hest hydroc ks (Linjorde mmerfest ba
ure 2.6 Cores
ldal and Tap wer ’ Nordm
rce rock an an import drocarbon in kkingen For ding (up to anic conten rce rocks h ferent sourc enriksen et a
carbon index et and Olsen asin at Early
of Triassic sh
ppel(2004) ela 2’ with nd reservoir tant hydroc ndex (130-2
rmation is v 300 API).
nt (TOC 8%
here are mos ce rocks in
al., 2011).
x (200-590 n, 1993). Ea y Cretaceou
hale source roc Format
mentioned extensive sh , but as a s carbon fact 50 mg HC/g very distinc
It is the be
%-20%) so it st immature different tim
mg HC/g T arly Triassi us (Henrikse
cks. Left: Sna tion in well 72
the upper s hale layers.
source rock, tory, for th
g TOC) and ctive in wel est source r t has high p e or early m
me is possi
TOC) amon ic source ro en et al., 20
dd Formation 224/7-1.
sand domin . Thus Nord , shales in N heir low T d thin layer ll sections, rock in the potential to mature. Mix ible when S
ng all three ck began to 11).
in well 7121/
nated ‘Nordm dmela plays
Nordmela F TOC amoun thickness (1 for its extre Snøhvit fi generate oi xing of hyd
Snøhvit fiel
possible so o generate o
/5-1. Right: K
dmela 1’ and s both a pa Formation s
nt (1-4%), 10-15m).
remely high ield for its il. But after drocarbons f ld was form
ource oil in
Kobbe
d the rt of seem low h GR
high r all, from ming
Source Trias shal Nordm Forma Hekkin
Forma
18 Figu Righ
Tabl
e rock Ag ssic
les Trias mela
ation
Earl Juras ngen
ation
Lat Juras
2.3.
Stø 17%
Prog perm Form clea
Res
ure 2.7 Cores ht: Well 7120/
le 2.1 Charact
ge Thickn
ssic 0-6
ly
ssic 10- te
ssic 10-
2 Reser Formation
% and core p grading co meability. In
mation, in G an sandstone
ervoir charac
of Jurassic sh 2-2, Hekkinge
teristics of Snø
ness(m) T (
60 2
-15 1
-50 8
voir Rocks is the best r permeabilit oastal regim
n the study GR logs, th e.
cterization of
hale source ro en Formation.
øhvit source r
TOC
%)
Bou (m 2-8
1-4
8-20
s
reservoir of ies ranging me environ area, most he thick and
f Snøhvit Field
ocks. Left: Th .
rocks (modifie
und hydrocar mg HC/g rock
55-40
2-23
10-62
Snøhvit fie from 150 t nment had of the wells d box shape
d, Norwegian
he lower Nor
ed after Linjor
rbon k)
Hyd (m
ld, with goo to 500 mD d generated s recorded o ed low read
n Barents Sea
dmela shales
rdet and Olsen rocarbon ind g HC/g TOC
200-590
130-250
8-20
od porositie (data from d the high over 100m t dings made
a
in well 7120
n, 1993).
dex C)
Keroge type III/II
III/II
II/III
s reaching u well 7121/4 h porosity thickness of
the baselin
0/6-1.
en Poten hydroca I Gas/min
I Waxy o
I Oil/g
up to 4-1).
and f Stø ne of
ntial arbons nor oil
oil/gas
gas
Nor as m accu Tub repo Form net/
Tub diffe
Fi
rdmela Form more shaly
umulated in båen Format orted in wel mation is /grass ratio båen Forma ferential com
igure 2.8 The
mation has a sandstones n the upper p
tion does n lls 7121/4-1
dominated (75% on av ation increa mpaction an
e upper and low
a more com and lower part of this
ot have oil 1 and 7121/
by transg verage) (Ram
ses towards nd differenti
wer parts of S
mplex litholo porosity a formation.
and gas in /4-2. The up gressive sho
mberg et al s west. Thi ial subsiden
Stø Formation
ogies and fa s well as p
general, bu ppermost pa oreface san , 2008; Hel is phenome nce (Helgsen
, Well 7120/6-
acies. In gen ermeability ut a trace ga art of sandst ndstones an
gesen, 2010 enon may h n, 2010).
-1.
neral, it app y. Hydrocarb
as cap has b tones in Tub nd has a 0). Thickne have caused
pears bons been båen high ss of d by
2 20
F N 2466-246
the form seque sandston
Figure 2.10 Sand a Nordmela Formatio 66.5m, the upper pa mation. Fining upwa nce, medium grains ne with cross beddin
Reservoi
and shale dominated on, Well 7121/5-1
rt of ards
s ngs.
2488-2488.
of the forma grey siltsto wavy c
ir characterization
d section of 5m, the lower part ation. Light to dark one and shale with
cross bedding.
of Snøhvit Field, NNorwegian Barent
Figu Well Light gr sandstone grains. We
2528-2 ts Sea
ure 2.9 Different san 7121/5-1 and Well rey clean
e with fine ll 7121/5-1, 528.5m.
ndstones in Tubåen 7121/5-3
Brown to grey sandstone. W 2015-20
Formation, cross bedding Well 7121/5-3,
015.5m.
To are
Figu Sme
2.3.
The (199 In u clos and Alth ente Hor Tria
sum up, the shown in Fi
ure 2.11 Litho lror et al., 200
3 Traps e E-W majo
95), in Ham upper Albian
sing of the E Grung Olse hough Hekk ered into th rizontal mig assic shale,
e main tecto igure 2.11
ostratigraphy a 09, NORLEX
or faults in mmerfest bas n, horizonta E-W major
en, 1993).
kingen Form he reservoir gration also
however, c
onic events
and major tect , 2011 and Os
Snøhvit fi sin, most of al stress dire
faults befor mation was r by horizo happened in could migra
and litholo
tonic events in stanin et al., 2
ield provide f the discove ection was n re the period deposited a ontally mig
n Nordmela ate verticall
ogies with p
n the area, mo 012.
e structural eries are bou north-south d of hydroc after Stø Fo gration in M
a source roc ly across sa
petroleum sy
dified and com
traps. Acc unded by til . The traps w arbon migra ormation, th Middle Tert cks. Hydroca
ands and sh
system elem
mpiled from
cording to D lted fault bl were forme ation (Linjo he hydrocarb tiary via fa
arbons from hales from
ments
Dore ocks.
ed by ordet bons aults.
m the Late
22
Cret This reas re-a larg Tria clea
Figu up to
The abov not this geo beli from al.,
Res
taceous to E s uplifts de sons (Kulke active faults ge amount assic section ar in seismic
ure 2.12 Scree o the sea bed.
e Oil-water ve the spill full, but the
trap. A r chemical an ieved to rela m the structu
1992, as cit
ervoir charac
Early Tertia estroyed mo e, 1994): the
and source of hydroca n and extend c sections.
enshot of inlin The lower for
contact in S point (Figu e estimated reasonable
nalyses con ate with the u
ure, while th ted in Dore
cterization of
ry.
ost of the h e phase sepa
rocks move arbons along ded up to se
e 3064. The le rmations are s
Snøhvit is lo ure 2.13). T
volume of guess is t nfirmed this
uplift in Lat he gas expan
and Jensen
f Snøhvit Field
hydrocarbon aration, gas e ed out of th g the fault ea bottom. I
eaking gas cau shadowed.
ocated with This phenom f generated h that leakag s theory (L te Cenozoic nded to abo n, 1996).
d, Norwegian
n traps in w expansion, s he oil windo
ts which fr n Snøhvit, t
used a chimne
hin Nordme menon indic hydrocarbon
e existed injordet and c, over 370 × ut twice its
n Barents Sea
western Bar seals were d w. Hammer om the Jur the ‘gas chim
ey from the pro
ela Formatio cates the tra ns should b in Snøhvit d Olsen, 19
× 106 Sm3 oi original vol
a
arents Sea f destroyed by rfest basin l rassic and mneys’ are
oduction form
on (2418m) ap of Snøhv be enough to t field. Sev 993), leakag il were expe lume (Nylan
for 4 y the ost a Late very
mation
) and vit is o fill veral ge is elled nd et
Figu below
ure 2.13 above w: N-S direct
e: E-W cross s ion, crossline
section and pro 3170, clearly
oduction proc y shows the fau
ess of Snøhvit ult block traps
t field (Maldal s.
l and Tappel, 22004),
24
C
Vari Petr rese ratio and give
Befo The deta
Res
Chapter 3
ious types o rophysics (I ervoir prope o, while Pet making ge en below.
fore starting e following
ailed assump
ervoir charac
3. Meth
of computer IP), Micros erties analys trel is used m eologic mod
g the resear of this ch ptions and a
cterization of
hodology
r software ar soft Excel sis such as p
mostly for w dels and dis
Figure 3.1
ch, some b hapter will applications
f Snøhvit Field
y and Th
re used to w and Matlab porosity, sa well correla stribution m
1 Workflow o
backgrounds mention so s will be de
d, Norwegian
heoretica
work for diff b are used aturation, cla ation, synthe maps. The fl
f the study.
s should be ome main c
scribed resp
n Barents Sea
al Backg
ferent purpo to QC the ay volume a etics, seismi
low diagram
e mentioned concepts an pectively in
a
ground
oses. Interac logs and to and net to g ic interpreta m of researc
d and review nd theories n each chapt
ctive o do gross
ation ch is
wed.
, but ter.
3.1 Dur drill than form wall are
Figu not a of th
Bor mud Und logs trac chem
Boreho ring the dril l bit down.
n the pore mation fluid
ls to the for mudcake, fl
ure 3.2 Chart as clear as sho he formation.
rehole struc d composi derstanding s are mostly ces (e.g. gam mical prope
le Structu ling process Because th
pressure o d and leave s
rmation, sev flushed zone
of the boreho own, and shap
ctures are in tion, mud
borehole st y recorded mma ray, so erties of the
ure s, high pres he hydrostat of the form solid residue veral gradua e, transition
le structure an pes of zones l
nfluence by pressure, tructure is v
in the unc onic, density
borehole ra
ssure mud w tic pressure mations, mu es at boreho ally changin n zone and u
nd symbols. N largely depen
y many fac formation very import cased portio
y and neutr ather than th
was poured t of the mud ud filtrate ole walls. T ng concentr un-invaded z
Note boundar d on heteroge
ctors includ property ant for well on of the w ron) may be he influence
to lubricate d column is
will displa herefore fro ic zones are zone (origin
ries between d eneous porosit
ding bit size and mine l log interpr well bore. A e influence
e of formati
and cooling s usually gre ace the orig om the bore e formed. T nal formatio
different zone ty and permea
e, mud den eral sensiti
retation bec A change in of physical ion itself.
g the eater ginal ehole These on).
es are ability
nsity, ivity.
cause n log l and
Reservoir characterization of Snøhvit Field, Norwegian Barents Sea
26
3.2 Log Editing and Quality Control
3.2.1 Basic Quality Control
Once the log data arrive, the first step of workflow is to ensure that the data are available. The following discipline should be performed (Darling, 2005).
1. Check if the total depth of log roughly matches the drilling record of the well.
2. Check if the derrick floor elevation and seabed positions are correct.
3. Check if the log curves are on depth with each other.
4. Check if the caliper is reading correctly inside the casting. It is useful to find out the non-permeable zones which are not washed out.
5. Check the density borehole correction curve.
6. Inspect the resistivity curves. If oil-based mud is being used, the shallow curves will usually read higher than the deep curves (except in highly gas and oil saturated zones), vice versa with water-based mud.
7. Check the sonic log by observing the transit time in the casting, which should read 57 μs/ft
8. Look out for any cycling-type behavior on any of the curves, such as a wave pattern.
This may be due to corkscrewing while drilling or log toll damaged.
9. Check that the presentation scales and units on the log are consistent with other wells or generally accepted industry norms.
In all seven given wells, curves are available after the quality check above.
3.2.2 Well Log Correction
For almost every kind of log, the reading is not only the function of the properties of the nearby formations, but also largely rely on the hole conditions (hole diameter, mud weight, tool models and positions, etc.). To reduce the influence of those unrelated factors and reflect the true conditions underground, many companies published their own charts and/or software. In this study, the Schlumberger Log Interpretation Chart (2009 edition) has been applied to make corrections and to determine the property parameters. The software Interactive Petrophysics (IP) also contains pre-installed charts to automatically perform the calculations and plotting. The following contents will not involve the physical theories of each method, but will discuss the influencing factors for the most common curves used in this study. Porosity logs will be discussed separately.
Gamma Ray
Everything between the formation and the tool absorb gamma rays. The distance between the tool and the formation will certainly affect the final measurements.
Figure 3.3 gives a correlation factor for measured value of formation gamma ray. The input parameter t is calculated as below.
whe Wmu
dh = dsond
As m heav
Figu corre sand
ere
ud = mud we
= diameter o
de = outside might be ex vy muds.
ure 3.4 Repres ection factor f dstone horizon
eight (g/cm of wellbore e diameter o xpected, the
Figur
sentative secti for shale form ns in small hol
.
3) (in) of tool (in)
e correction
re 3.3 Borehol
ons of the resu mation in big an
les.
………
ns are quite
le correlation
ult of borehol nd irregular h
………
e significant
chart of gamm
e correction in oles. Right: li
………(3.1
t in large bo
ma ray.
n well 7120/5- mited correcti
1)
oreholes an
-1. Left: signif ion result for
nd in
ficant pure
Reservoir characterization of Snøhvit Field, Norwegian Barents Sea
28
Sonic logs
The tool of sonic logs is designed to eliminate the affection of the size of hole by using a pair of transmitter - receiver sets. Sonic logs can be used to estimate many properties include density, porosity, elastic properties and identify the overpressure by the shape of the trace. But the strongest requirement of trace editing in this study appears in the steps of synthetics.
To make the synthetic seismogram, the reliable log curve is needed along the whole borehole wall. The damaged logging trace will make a wrong time - depth relation.
Every mistake of the trace will spread out by the time-depth relation. The most common problems of sonic log are cycle skip, surface noise and attenuation.
The main editions of the trace include despiking, smoothing, interpolation etc., but the most useful and accurate way to control the quality is using the checkshots.
Density log
Compensated formation density log records both bulk density (RHOB) and density correlation value (DRHO). Litho-density log records Photo electric effect (Pef or Pe) in addition to the two traces above. As the density logs are still using gamma ray as the tool, the affect factors are similar as the GR log, but for density logs, the composition of mud poses much stronger affect than the hole diameter.
3.3 Shale Volume Calculation
Therotically, the gigantic physical differences between clay and sand would standout for both in almost every kind of log, e.g. GR, resistivity, neutron, RHOB, Pef and sonic reading. But GR log is the most common and fast way to calculate the amount of shale.
Consider the formation compromised of thick pure shale and sand horizons, the GR reading should be close to zero in sand zones (will not be zero because of the radioactive statistical fluctuation) and remain high in shale zones. So the most common and simple equation to calculate the percentage of shale volume is given below:
……… (3.2)
Where, is the gamma ray reading of formation, is the minimum gamma ray (clean sand or carbonate) and is the maximum gamma ray reading (shale).
IGR is the parameter named ‘gamma ray index’.
Equation 3.2 is the first formula that was used to calculate the volume of shale. This formula is based on the ideal condition and rules out any other disturbing factors.
However, in real application, the errors of are larger than expect, partly because the real formations are not pure and the real equation is not linear, the following formulas
are For
Or
For
For
If th upp incr
Figu
given for di normal roc
old (consol
Tertiary (un
he sandston per five equ
reases, resul
ure 3.5 Conve
ifferent app cks,
∗
1.7 lidated) rock
0.33 2 ∗ nconsolidat 0.083 2 ne formatio uations wil lts from diff
ersion of the G
plications.
(Steiber, 19
3.38 ks,
1 (L ted) rocks,
. ∗ 1
ons are thick ll not vary ferent equat
GR index to sh
970) ………
0.7
Larionov, 19
(Lariono, k and pure y much. Ho
tions may h
haliness depen
………
(Clavier, 1
969) ………
1963) ……
e, the calcul owever, wh have the erro
nding on rock
………… (
1971)…… (
………
………
lated shale hen the sh or as high as
types (Ellis an
(3.3)
(3.4)
(3.5)
(3.6) volume by haly compo
s 15%.
nd Singer, 200
y the onent
08)