Reservoir Characterization of Snøhvit Field, Norwegian Barents Sea

(1)

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

(2)

(3)

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

(4)

© 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)

(5)

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

(6)

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.

(7)

III

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

(10)

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

(11)

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

(12)

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.

(13)

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 km²and 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

(14)

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.

(15)

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.

(16)

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.

(17)

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

(18)

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

(19)

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.

(20)

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

(21)

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.

(22)

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

(23)

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.

(24)

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

(25)

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

(26)

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

(27)

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

(28)

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.

(29)

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

(30)

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

(31)

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),

(32)

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.

(33)

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

(34)

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.

(35)

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

(36)

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

(37)

are For

Or

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)

Reservoir Characterization of Snøhvit Field, Norwegian Barents Sea

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

Acknowledgements

Abstract

Table of Contents

C

Chapter

r 1 Intr

roduction

n

Chapter 2 Geology of the Study Area

C

Chapter 3

3. Meth

hodology

y and Th

heoretica

al Backg

ground