A Critical Review of Cerrently Pore Pressure Methods

8/20/2019 A Critical Review of Cerrently Pore Pressure Methods

1/176

•

•

•

http://htt//etheses.dur.ac.uk/policies/http://%20http//etheses.dur.ac.uk/4090/http://etheses.dur.ac.uk/4090/http://www.dur.ac.uk/


2/176

http://etheses.dur.ac.uk/


3/176

A

critical

review of

currently

available pore

pressure

methods

and their input

parameters.

Glaciations and compaction of North Sea

sediments.

A

copyright of this

thesis

rests

with the author. No quotation

from it should be published

without his prior written

consent

and information derived from it

should be acknowledged.

By

Carl Fredrik Gyllenhammar

0

NOV

2003

This

thesis

was submitted as the

fu l f i lmen t

of the requirements for the

degree

o f

Doctor

of Philosophy

Car l

l-redrik

Gyl lenhammar

i


4/176

Abstract

Historically

pore

pressure

evaluation in exploration

areas

was based on

empirical

relationships between d r i l l i n g parameters,

wireline

logs and the mud weight.

Examples include Eaton's Ratio and the Hottman & Johnson Methods,

which

were

based on data

from

the G u l f of

Mexico.

These methods are not readily transported to

other

areas,

such as the North Sea Basin, where the sediments are different in

character and where

burial

and temperature histories are

distinctly different.

Data f r o m

several offshore

North

Sea

wells,

w i t h

high quality wireline

and associated

data have been analysed to determine the most appropriate method to estimate pore

pressure

in mudrocks. The data have led to an understanding of the key parameters

for

successful pore

pressure

estimation. The most

effective

method is shown to be the

Equivalent

Depth

Method,

but

only

where

disequilibrium

compaction is the source of

the overpressure in the mudrocks.

Core samples

f r o m

576

Br i t i s h

Geological Survey sites in the offshore area o f the

B r i t i s h

Islands were compared w i t h

>

10,000 porosities collected

f r o m

the

deep oceans

(DSDP/ODP sites), which show that the porosities in the shallow section in the North

Sea are anomalously low. The shallow section of the

North

Sea includes large

volumes of Pleistocene-Recent sediments deposited as glacial and inter-glacial

deposits. Frequency analysis

(Cyclolog)

of the

wireline

data covering this

interval

in

several

North

Sea wells revealed a pattern in the relative featureless

original

data.

Comparison w i t h

the

global

signature fo r oxygen isotopes for the

same

time period

suggests

that there have been ten cycles of ice

sheet b u i l d

up

(Glacial

period)

followed

by

melting

(Interglacial

period)

during

the last one

m i l l i o n

years.

Glacial

deposits

from

10

individual

glacial cycles have therefore been

identified

in several

exploration wells in the North Sea. Implications of

loading/unloading

o f ice for the

migration

and trapping o f hydrocarbons in the

North

Sea Basin are

assessed.

Carl

Fredrik Gvi l en h a m m a r


5/176

Acknowledgements

First I must thank my supervisor, Richard Swarbrick ( D i c k ) . I met Dick f i rs t time in

December 1995 in London. It was the

f i r s t

GeoPOP meeting I attended representing

Norske Conoco. When a year later I expressed interest in doing a PhD at the

university of Durham, his enthusiasm, despite my age, made what was

laying ahead

possible. My w i f e M a r i t and I l e f t our jobs late 1998, we sold our house in Stavanger

and moved to Durham w i t h our two children,

Elen-Martine

and Fredrik.

A t the university I found Dic k' s continues support and interest i n my subject as w e l l

as the requirement to deliver regular reports to GeoPOP an

assurance

for continued

progress. N e i l

Goulty

was never fa r away to discuss any d i f f i c u l t equation. Both his

and Dick' s enthusiasm for my "ice theory" made this thesis what it is. I had also help

f r o m

Fred Wollard's

long

experience w i t h using principle component analysis. The

last year in Durham I was lucky to share an office

w i t h

M a r t i n Traugott. He shared his

long

experience in pore

pressure

predic tion as w e l l as his own developed software

PresGraf w i t h me.

M y thanks also goes to Norske Conoco, in particularly James Middleton. They

supplied

the w e l l data I used as w e l l as providing financial support fo r the project.

Thanks go to all

those f r o m

GeoPOP who were there to help and exchange ideas,

Toby, Paul, Daniel, Neville, Gareth Yardley, Andy A p l in and Yunlai Yang.

Finally many thanks to my new colleges at BP, for their support during the last year,

Nigel Last, Mark Alberty and M i k e McLean.

Call Freclrik Gyl ienhamrnur

iii


6/176

Table of Contents

Abstract ii

Acknowledgements ii i

Table of Contents iv

List of Tables vi

Lis t of Figures

v

i i

Declaration x i

Chapter 1 Introduction 1

1.1 Background 2

1.2 Data 3

1 3 Introduction 7

1.4

Pressure,

the basic concepts 11

1.5 Aims and layout of thesis 14

Chapter 2 Pore Pressure Evaluation Concepts and definitions

16

2.1 Def in i t i on 17

2.1.1 Mudrock porosity 18

2.1.2 Different porosity evaluation equations 18

2.1.3

Normal compaction curve and trend lines 23

2.1.3.1

Athy-type re lationship

24

2.1.3.2 Soil mechanics relationship 25

2.1.3.3 A t hy

- Soil mechanics: how are they

different

25

2.1.4

Vertical versus

mean

effective stress

28

2.2

Pore Pressure

Calculation Methods 29

2.2.1 Vertical

Methods 30

2.2.1.1 Equivalent depth method 30

2.2.1.2

Harrold

method 32

2.2.1.3 Exp l i c i t

method using the

resistivity

log 33

2.2.2

Horizontal

methods 35

2.2.2.1 Eaton Method 35

2.2.2.2 The pore

pressure

calculation program; PresGraf 37

2.2.2.2.1

PresGraf

normal

compaction trend 37

2.2.3 Seismic 40

2.2.4 ShaleQuant 41

2.2.5 Principle Component Analysis 41

2.3

Dr i l l ing parameters

46

2.3.1 Real time

data

48

2.3.2 D'exponent 48

2.3.3 Torque 54

2.3.4 Hydraulics 55

2.3.5 Bit type and wear 55

2.3.6 Lagged data 56

2.3.6.1 Gas 56

2.3.6.2 Cuttings and Cavings 58

2.3.6.3 Mud temperature in and out 58

2.3.6.4 Mud

resistivity

in and out 58

2.3.7 Mud chemistry and mud-formation chemical reactions 59

Chapter 3 Comparison of

different

pore

pressure

methods using a North Sea wel l 60

3.1 Introduction 61

3.2

Pore pressure

evaluation of

w e l l

1/6-7 in the

North

Sea, Norwegian sector 61

3.2.1 Pore

pressure

evaluation while dr i l l ing

(wellsite)

62

3.2.2 The Post-well analysis • 63

3.2.2.1 Tertiary 67

3.2.2.2 Chalk 67

Carl

Fredrik GvHenhainmar iv


7/176

3.2.2.3

Jurassic

67

3.3 Normal Compaction in the North Sea 69

3.3.1 Palaeocene and Lower Eocene 75

3.3.2 Normal compaction f r o m resistivity data 76

3.4 Wireline log

pore

pressure calculation 78

3.5 Comparing the North Sea

w i th

the

G u l f

of Mexico Basin 86

3.6

Vimto#land#2

88

3.7

Summary

and conclusions 93

Chapter

4 The impact of the Glaciation on the Normal compaction in the North Sea 96

4.1

Introduction 97

4.2 Glacial history 99

4.2.1

The

Neogene

-

Pleistogene sedimentary succession

102

4.3

Ti l l s

103

4.4 Mudrock

porosities 104

4.5 Oxygen

isotope

data

106

4.6 Time series frequency

analysis

(CycloLog) 108

4.7 Ice loading and

pore pressure 112

4.8 Subglacial water f l o w

.....118

4.8.1 Resistivity log response 121

4.9 Hydrocarbon migration 122

4.10

Erosion of the Scandinavia during

Quaternary

123

4.11

Conclusions 124

References: 126

Appendix 1

136

Appendix 2

145

Appendix 3

155

Carl

Fiednk

Gvll enliamma '

v


8/176

L i s t of Tables

Table 2-1 Eigenanalysis of the correlation matr ix 42

Table 2-2 A

l ist

of

measurements

that can be

used

to interpret the

shale

pore

pressure.

47

Table 2-3 Eigenanalysis of the correlation matr ix 52

Table 4-1 BGS sits and exploration wells 105

Table 4-2 The f l o w rates are calculated assuming hydrostatic

pressure

in the aquifer

underlying

the aquitard 120

Curl l-reclrik

Gy l l en h a m m a r v i


9/176

L i s t of Figures

Figure 1.1 The wireline log plot of w e l l 1/6-7

f r o m

seabed to 2000m 4

Figure 1.2 The wireline log plot of

w e l l

1/6-7

f r o m

1900 to 4000m 5

Figure

1 3

The wireline log plot o f

w e l l

1/6-7

f r o m

3000

m

to

4995

m

(TD) 6

Figure

1.4

Schematic of

wireline

logging. The

lithological

column to the right is a

schematic o f a pressure probe (RFT) being

used

to measure the pore pressure in

permeable sandstone

9

Figure 1.5

Pressure

plotted

against

depth in a

fictional w e l l .

The effective

stress

is

equal to the overburden pressure minus the pore pressure and the

overpressure

is

equal to the pore pressure minus the hydrostatic pressure 12

Figure 1.6 The Figure to the right shows how a pressure

versus

depth plot

( l e f t ,

Figure

1.5)

becomes

presented

as

pressure

gradient

versus

depth 14

Figure 2.1

a,b,c,d. la

porosity variation as a function of sonic velocity. B, porosity

versus

bulk density. C, the sensitivity to pore water density. D, the sensitivity to

matrix density 20

Figure 2.2 Log derived porosities in

w e l l Nor-1/6-7

Norway. The low porosity

interval

f r o m

3261m to 4346m depth is the

Cretaceous

Chalk. The

values

are

listed in Appendix 1 23

Figure 2.3 Comparison o f porosity w i t h effective stress for the

A t h y

and the SM

equations. I n i t i a l porosity (sea floor porosity) for

A t h y

is 0.7 (70 %) whi le the

porosity at 100 kPa (approximately 100

meters

below sea

f l o o r )

is 0.66 (66 %)

The compaction factors a re for

A t h y ; 0.00008

and SM; 0.74 26

Figure 2.4 The relation ship between porosity, solidity and

v o i d

ratio is shown. The y-

axis is the compaction as a length reduction. It is

assumed

that a confined

volume is

compressed

beginning

w i t h

a

v o i d

ratio of four 27

Figure 2.5. Porosity

f r o m

a

pseudo w e l l

is plotted

versus

depth. Integrating the density

log and subtracting the hydrostatic pressure calculate the effective

stress.

The

two

normal compaction curves are coming

f r o m

Figure 2.3 30

Figure 2.6 Porosity

f r o m

a

pseudo

w e l l is plotted

versus

mean effective stress. The

two

normal compaction

curves

are coming

f r o m

Figure 2.3. See text for an

explanation for the equivalent effective stress method 32

Figure 2.7 Comparing the porosity derived

f r o m

the sonic, density and neutron log

w i t h the resistivity derived using the equation 1.37 proposed by

A l ix a n t

and

Desbrandes(1991)

34

Figure 2.8 The PresGraf normal compaction trend to the

l e f t

compared w i t h the

A t h y

normal

compaction curve (Figure 2.3) to the right 39

Figure 2.9 Scree plot of the eigenvalue of

each

principal component of the PCA of the

wireline

logs 43

Figure

2.10

a, b, c and d. The plots to the

l e f t

(a and c) are cross

sections

through

data

perpendicular to

PCI,

PC2 and PC3. b and d show the loading values w i t h

respect

to the different principal components 44

Figure

2.11

a and b.

PCI versus

depth in blue and the sonic travel time in green 46

Figure 2.12 The plot

shows

the

d'exponent

and the corrected

d'exponent versus

depth.

The straight lines are trend lines representing one particular pressure gradient

(one mud weight) 49

Figure 2.13 Cross plot of the d'exponent

versus

the sonic travel time, w e l l N 1/6-7. It

suggests

a good relationship between the sonic log and the

d'exponent

in the

Tertiary

section (pink

squares).

The correlation is

less

obvious in the

Jurassic

(yellow

triangles). There is no correlation in the Chalk 50

Carl

Frecii

ik

Gvl l en h a m m a r

vn


10/176

Figure

2.14 Cross

plot of the

d'exponent versus

the resistivity log. The plot show no

correlation 51

Figure

2.15

Scree plot of the eigenvalue of

each

principal component of the PCA of

the

dri l l ing

parameters

53

Figure 2.16 a, b, c, d. The plots to the left (a and c) are cross

sections

through

data

perpendicular to

PCI,

PC2 and PC3. b and d show the loading

values

with

respect

to the

different

principal

components

53

Figure

2.17

a, b and c.

PCI

versus

depth

wi th

the standardized log porosity overlaid in

Figure b and the normalized sonic travel time in Figure c 54

Figure 3.1 The corrected d'exponent plotted

versus

depth

wi th

a normal trend line

overlaid

in green.

Normally

new trend lines

w i l l

be

added

paralleling the green

line

each time a new bit is put on the drill string. The new line w i l l represent the

actual

M W .

An alternative trend line is

suggested

in red. That trend line

w i l l also

result in a

reasonable

calculated pore

pressure

at the target of interest (i.e. 4200

to 4800

m)

63

Figure 3.2 The sonic velocities in the

shale sections

plotted

against

depth on a

semilogaritmic graph. The yellow line is the best visual f i t trend line 65

Figure 3.3 A comparison of the calculated pore

pressure

f r o m the

d'exponent

(blue

dots) and the sonic velocity calculated pore

pressure

66

Figure 3.4

Shale

velocities f r o m the wells in Figure 3.2 compared

wi th

the wells

used

by

Hansen (1996). The Figure to the left has a logarithmic X-axis. At such a plot

the A t h y type normal compaction trend

become

a straight

line.

On the plot to the

right

it is much more obvious that the

wel l used

in this study are different f r o m

the one

used

by Hansen

(1996)

70

Figure 3.5 Porosity

data

f r o m

10000

DSDP/ODP mudrock

samples.

ODP site #336 is

in

the

deep

water Norwegian Sea. The yellow curve is the

suggested

normal

compaction trend drawn through the

data

set 71

Figure 3.6 Compaction

trends

as plotted as vo id ratio

versus

effective stress.

A l l

the

different trends

that

have been tested

in this

chapter

are overlaid. The PresGraf

(in

blue) is an approximation as the real curve is proprietary

data

to BP 72

Figure 3.7.

Different

wireline methods to calculate the volume of clay are compared

in wel l

N 1/6-7. The red dot are V c l f o rm using the GR log, the blue

dots

f r o m

the neural network (ShaleQuant and the green

dots

the neutron and density log.

74

Figure 3.8 Twenty-six wells in 5 offshore

areas

and 4

onshore

fields (MacGregor,

1965). The pink line is a

suggested

normal trend for the

G u l f

of Mexico 77

Figure 3.9

Pore pressure

in

mega-Pascal versus

depth in

meters.

The green solid line

is the overburden and the blue solid line is the hydrostatic pressure. The

equivalent depth method calculated pressure is in blue

dots

while the

orange

is

the University of Durham method. The

dashed

black curve is the

operator's

interpretation while the olive solid line is the mud weight. The red crosses are the

R FT

direct pore pressure

measurements

79

Figure 3.10

Pore pressure

in

mega-Pascal versus

depth in meters. The green solid line

is the overburden and the blue solid line is the hydrostatic

pressure.

The Eaton

equation

wi th

the sonic log as input (red dots) compared

wi th

the Equivalent

depth method

wi th

the porosity as input (blue dots). The red

crosses

are the RFT

direct pore

pressure measurements.

The

values

are listed in Appendix 2 82

Figure

3.11 Pore pressure

in

mega-Pascal versus

depth in

meters.

The green solid line

is the overburden and the blue solid line is the hydrostatic pressure. The Eaton

equation in red

dots

compared

wi th

the Equivalent depth method in solid blue,

Curl Fredrik Gvilenhuff im;ir

V l l l


11/176

both w i t h the sonic log as input. The red

crosses

are the RFT direct pore

pressure

measurements.

The values are listed in Appendix 2 83

Figure

3.12 Pore

pressure

in mega-Pascal versus depth in meters. The green solid line

is the overburden and the blue solid line is the hydrostatic

pressure.

The

Equivalent

depth method tested w i t h two

different

normal trends. The

A t h y

equation used by the operator of

w e l l

Nor-1/6-7

(blue dots)

versus

the DSDP-

ODP

based trend (red dots).The red crosses are the RFT direct pore

pressure

measurements. The values are listed in Appendix 2 83

Figure 3.13 The

shale

porosity (red solid curve to the right) and the

shale

travel time

(green

solid line

to the

right) versus

depth in the

Jurassic

section. The x axis is in

for porosity,

u,sec/ft

for the sonic log. The curves to the l e f t of the overburden

(strait solid green line) is in MPa. Between the overburden and the hydrostatic

pressure ( l e f t

most solid blue) are

f r o m l e f t

the pore

pressure

calculated using the

sonic log as input (blue curve) then w i t h porosity as input (orange curve). The

values are lis ted in Appendix 2 84

Figure 3.14 Figure 3.13, the

pressure

transition

zone

f r o m

4850- 4890 meters. The

values are listed in Appendix 2 84

Figure

3.15 Pore pressure in mega-Pascal versus depth in meters. The green solid line

is the overburden and the blue solid line is the hydrostatic pressure. The Eaton

equation w i t h the resistivity log as input (green dots) compared w i t h the Eaton

sonic (red dots). The red

crosses

are the RFT direct pore

pressure measurements.


Figure 3.16 Depiction of salt features in the area around the basin. The

local

depocenteres

are termed mini-basins (Yardley and Couples, 2000) 87

Figure

3.17

Comparison of the NRG derived

shale

porosity

versus

permeability

curves,

w i t h

some

basin modell ing default curves. A

range

o f

clay-fractions

are

shown,

f r o m

20% to 80% (Yang and A p l i n , 2000) 89

Figure

3.18

Comparison o f the GeoPOP derived

shale

compaction curves w i t h

some

basin modell ing default curves. A

range

o f clay fractions are shown,

f r o m

20%

to 80% (Yang and A p l i n , 1999) 90

Figure 3.19 Pore

pressure

prediction for

V i m t o

#2. The blue curve is the pore

pressure

calculated using the Eaton method and the

shale

sonic

velocity

as input.

The red

line

is pore

pressure

using the equivalent depth method. The

pink

diamonds are the M D T

pressure

points. The red line to the l e f t is the overburden.


Figure 3.20 The reservoir section for

w e l l

Vimto#2.

The MDT

pressures

are generally

50 to

100

psi higher than the calculated

shale pressures 91

Figure 3.21

The red curve is the pore

pressure

calculated f r o m a 2-D model alowing

for lateral transfere (Yardley and Couples, 2000) 92

Figure 4.1 A curve showing the variation in oxygen isotope composition of the sea-

water for the last 6 m i l l i o n years. The oxygen isotope

data

are

based

on

foraminifera f r o m three boreholes near the coast of Ecuador (Shackleton et al .,

1990; Shackleton et a l . , 1995 98

Figure 4.2 Reconstruction o f the Scandinavian and

B r i t i s h

Ice

Sheet

during a glacial

stadial (after Hughes,

1998).

The maximum ice thickness onshore was 2600 m

and the maximum thickness i n the offshore

North

Sea was approximately 1600

m

101

Figure 4.3 Porosity

versus

depth. A compilation of core

measurements

and

wireline

calculated porosities 106

Carl

I-relink Gvl lenhammar

X


12/176

Figure 4.4 To the l e f t is the oxygen isotope data shown in Figure

4 . 1 .

To the

right

is

the GR log followed by the filtered GR log f r o m w e l l

Nor-1/3-2.

The t h i r d curve

is picks representing sudden

changes

i n the c y c l i c i t y of the filtered GR curve.

The t h i r d curve show peaks, positive or negative representing sudden transitions

f r o m

high

to low GR value, shown as a negative peak (to the l e f t ) . The opposite

results in a positive peak. The last curve is the integration of previous curve.

These curve were output f r o m CYCLOLOG*. This curve represents the

cumulative

difference between the predicted log values and the actual log values.

Breaks in the c y c l i c i t y succession may be related to missing sections or abrupt

changes

in sedimentation

rates.

A large positive peak

could

be a condensed

section

109

Figure

4.5 Correlation of

five

wells

d r i l l e d

in the southern part of the Norwegian

sector using

CycloLog

software. The distance

f r o m Nor-1/3-2

to

Nor-2/11-7

is

about 100 km (60

m i l s ) .

The cycles are compared on the

l e f t w i t h

oxygen isotope

signature (See text) 111

Figure

4.6 The

four maps

above present the palaeo-coastline for each

subsequent

crustal motion

model. It is important to note that large parts of the

North

Sea

were dry land after the last deglaciation, in each case for a period of several

1000

years 113

Figure 4.7 The figure to the l e f t show a

typical

pore pressure profile in the

North

Sea

w i t h no seawater just prior to a

glaciation.

The sand at 2000 meters subcrope to

seafloor and has therefore hydrostatic pressure.

During

glaciation of the

North

Sea the overburden pressure and the hydrostatic pressure increase w i t h a pressure

equivalent to the weight of the ice-sheet. I f the sand subcrops under the ice-sheet

the pore

pressure

w i l l

also increase in the sand. But

i f i t

subcrops outside the ice-

sheet,

its pore pressure

w i l l only

vary as much as the sealevel

changes

114

Figure 4.8 The profile

B - B '

shown on Figure 4.7. It show that the maximum

subsidence was i n the centre o f the

Baltic

Sea of more than 400 metres,

while

it

was

potentially

u p l i f t in the

North

Sea ( M i l n e et

a l . ,

1999, Mitrovica et

a l . ,

1994,

Tushingham,

A . M . ,

1991) 116

Figure

4.9

Resistivity

curves f r o m the

North

Sea compared w i t h G u l f of

Mexico.

The

graph to the

l e f t

is raw data

while

the raw

resistivity

curves have been

temperature corrected on the graph to the

right 122

C a r l

Fredr ik Gyllenhammar

x


13/176

Declaration

The content of this

thesis

is the original work of the author (other

people's

work,

where included, is acknowledged by reference). It has not

been

previously submitted

for

a

degree

at this or any other university.

Carl

Fredrik

Gyllenhammar

Durham

September 2003

Copyright

The copyright of this thesis rests w i t h the author. No quotation f r o m it should be

published without his prior written

consent

and information derived

f r o m

it should be

acknowledged.

Cur

Fredrik

Gyl lcnhammar

xi


14/176

Chapter 1 Introduction

Carl

Fredrik Gvilenhummur

1


15/176

h i i - o ; H k ' ; i o n

1.1 Background

Fifteen years ago, most pore pressure studies were undertaken solely for safety

aspects in the design and

d r i l l i n g

o f exploration

wells.

As the need for accurate pore

pressure

evaluation is

growing

due to its general application in exploration studies

such as hydrocarbon

migration

studies, more accurate methods founded on sound

physical

principles, and not just empirical observations, are needed.

Pore

pressure

estimation is a particular challenge i n the North Sea on account of the

complex tectonic and sedimentological history of the region, where the highest

overpressure (pore pressures above the normal , hydrostatic pressure) are found in

Jurassic

and Triassic reservoir

sandstones.

The

presence

o f a thick Chalk section as

w e l l as a variety of

mudrock

types, including a kerogen-rich petroleum source rock,

challenge standard practices fo r pore pressure evaluation which were, in many cases,

developed

in the G u l f o f

Mexico

where the rocks are younger and exclusively

siliciclastic

(sandstone, siltstones and shale mudrocks). The late history (Pleistocene-

Holocene) o f the

North

Sea has

involved

ice loading and the deposition of

glacially-

derived

sediments which add a further component of complexity to the stress and

f l u i d

history of

North

Sea sediments.

The availability of a very high quality set of w e l l data f r o m the Norwegian North Sea

(Central

Graben) provided impetus for this project

which

was designed to test current

methods of pore

pressure prediction,

assess the impact of a late ice-loading and

unloading

history and apply new technology on mudrock compaction (being

concurrently developed in the GeoPOP research group - see below).

There are a number o f complementary data which can be used for pore

pressure

evaluation including basin modelling, seismic velocities, wireline logs and

d r i l l i n g

parameters. Each requires

different

data

input

and interpretation requirements. In this

thesis the emphasis is fo r pore

pressure

evaluation using

wireline

logs. The

response

f r o m

the

d r i l l i n g

parameters was used as an independent

control.

Carl 1-reilrik Gyl lenhammar

2


16/176

Chapter nirouui'i i o n

The thesis was funded by Norske Conoco in Norway and the work was included as

part of GeoPoP. GeoPoP (GEOsciences Project

into

OverPressure) was a j o i n t

research project involving University of

Durham,

Newcastle University, Heriot Watt

University

and

industrial sponsors

such as major

o i l

companies

l i k e

BP, Amoco,

Statoil,

Norsk

Hydro,

Phillips, Conoco, etc. The aim of GeoPOP was to expla in how

pore

pressures

evolve in mudrocks and to evaluate and develop new methods to

predict and calculate the pore

pressure

in

these

sediments.

1.2 Data

Norske Conoco made most of the data available, consisting of

wireline

data f r o m

exploration wells.

The most important

w e l l

was

1/6-7, d r i l l e d

by Norske Conoco in

1989, which is classified as a

high

pressure

( >

10,000 psi) and

high

temperature

(>350°F) (HPHT) w e l l . In addition to w e l l 1/6-7 were a number of offset well s in the

southern part o f the Norwegian shelf. The data set included also

some

wells

f r o m

Haltenbanken and the Barents sea.

W e ll

1/6-7 has

high quality wireline

and mud

logging data particularly w i t h respect to testing pore pressure prediction and

calculation

methods (Figure 1.1, 1.2 and 1.3).

GeoPOP provided the data f r o m the G u l f of

Mexico.

Data f r o m the shallow coring

project

by

B r i t i s h

Geological Survey (BGS) were provided by BGS. Data

f r o m

the

Ocean D r i l l i n g Project (ODP) are freely available on the Internet and were

downloaded free o f any charge.

Frednk

Gvlk 'n luunmar


17/176

Chapter

Intrudm/iiun

W E L L N1 6 7

Q

—

r | i |

" V T C X A T L r j

.CUt L

M L

L E T L

71

H L I B L

I T

L

T

ZL

X

L

S A J T — T M

t

J O

L 1 t

1

i r

1

L_l

L

r.

•

• - n

"W4 I K » * *

-4DD J E :

If

fa •

i •

LDD

L

S ff

•

Lt

r•

[•

l.L [•[•

. 7(,r,

, - r .

t

.

H O [ '

L . M - f

:

L L C C

Li

f t

I

i l l

. 1M

Figure 1.1 The wireline log plot of we ll 1/6-7 fr o m seabed to 2000m.

C a r l Fredrik

Gyllenhammar

4


18/176

Chapter

I t tw

U N

Figure 1.2 The wireline log plot of well 1/6-7 f ro m 1900 to 4000m.

Carl

Fredrik Gyllenhammur

5


19/176

W E L L

N1 6-7

« £ T

CALL

L

I I M I I U I I > I I I

U K liiilli

ill

I I I -1111111 I I I

i i r - i u n i r m i l

l l E ? i M I I I | i

i ll

M I I M i l l l l II

I I I K I I I I I H I M i

mi i i i i i u i i I S M

iiiiiiiiniii

iimiimiii

lllllllllllll

L

illllllllllll E

m i n i u m

?

H I I I S H I I I I I

fiiihiinii i i

HIIBIIIII llli

lllllllllllll I I I

N I I U I I I I I I I

I I I

[ I I I E I I I I I I I I

I I I

H I I E 2 I I I I I I

I I I

I I I I M I I I I I I

I I I E I I I I I I I I

I I I

I W I B I I I I H I I I

[ I I M I I I I I I I IF

' . n u l l u m r:

L U K I I I I I I I I I

i

I ' D l g H U M 111:'

It iSillllll

I I M - . i l l l

I ' E i ^ i i l l l l

I C S U U I I I I :

l l l l l i Ull'

IllrillllSi

lit.

Will

I I I I

nun

mi win

I

' ^ l l l l l

I M I I I I I I I

l .»?ll l l l l .

I H I C t M i

i T M

"m i

IIIIII IIIIII

i i i i J i l l l IIIIII E

siiiiimiiimiiii

211 i n i n i

IIIIII IIIIII mill

mm mm

m i n i

IIIIII IIIIII IIIIII

IIIIII IIIIII

m i n i

IIIIII IIIIII IIIIII i

i n IIIIII

m i n i

IIIIII IIIIII

m i n i

T *

3 §

or

rai

sum

l

immmFh

Figure

1.3

The wireline

log

plot

of

well

1/6-7

from

3000 m to

4995

m

( T D ) .

( ai l Fredrik Gvllenhumniur

6


20/176

1.3 Introduction

I n exploration d r i l l i n g operations pressure f r o m the

circulating

d r i l l i n g f l u i d (mud) is

used to prevent the pore

f l u i d

in the porous rock entering the borehole. The pressure

f r o m the mud at a particular depth is a function o f the average density (MW = Mud

Weight)

and the vertical height of the column

f r o m

that depth to the surface. In low

permeability formations, such as mudrocks, the formation can cave into the wellbore

through

tensile failure i f the pore pressure is higher than the counter pressure f r o m the

mud. The industry has a long history of establishing empirical relationships between

d r i l l i n g

parameters

such as the rate of penetration and the gas measured in the

returning

d r i l l i n g

f l u i d

to the pore

pressure

in the mudrocks. The

uses

o f

d r i l l i n g

parameters are very subjective and prone to large uncertainties. The pressure can also

be calculated

indirectly

f r o m petrophysical

measurements.

Petrophysical

data

can be

acquired while d r i l l i n g or after d r i l l i n g a section. In the former

case

the petrophysical

sensors are placed behind the d r i l l bit in operations known as Logging While D r i l l i n g

( L W D )

or Measurement

While

D r i l l i n g

( M W D ) .

When

data

is acquired

once d r i l l i n g

has been completed, the petrophysical sensors are lowered down the hole

suspended

f r o m

a

wire (wireline

logging) and readings taken by the tools while being reeled back

up. The pore pressures in the reservoir rocks w i t h high permeability are measured

directly

using a

wireline

tool w i t h a pressure

gauge.

A cylindrical probe w i t h a small

aperture is hydraulically forced into the formation (Figure 1.4) and the

tool

remains at

the location u n t i l the pressure stabilizes between the inside of the tool (where the

pressure gauge is located) and the

formation

(where the probe has

been

extended).

The pressure is recorded as pressure vs time. The most common trade acronyms for

these tools are RFT (Repeat Formation), FMT (Formation multi-tester) or MDT

(Modular Dynamics Tester). In mudrocks where permeability is very low, this

tool

cannot be used due to the time it w i l l take for pressure to stabilize. Direct pressure

measurements are also recorded when a hydrocarbon zone is tested, called a D r i l l

Stem Test (DST).

The accompanying petrophysical

measurements

collected at the same time as the

pressure

tests

include sonic, velocity, neutron porosity, density, and resistivity (unless

you intended to l i s t something else). These sensors are all calibrated for the porous

formation

and w i l l tend to give erroneous reading i f any clay minerals are

present.

C'ai

h e d n k

Gvilenlnuriinai'

7


21/176

Inli oduciion

The challenge is therefore to use these

measurements

in mudrocks

w i t h

low

permeability and high clay content. During compaction of compressible sediment,

such as mudrock, water is expelled and the porosity decreases. I f the free water which

needs

to be expelled to maintain equilibrium

w i t h

the imposed

stresses

cannot drain

out of the system, the porosity w i l l not decrease,

w i t h

the result that the pore pressure

increases

above

the hydrostatic pressure. Porosity cannot be

measured

directly in a

borehole. The porosity is calculated indirectly f r o m the sonic velocity, neutron

porosity, density or the resistivity

measurement,

or a combination of

these

measurements.

The effective or inter-granular

stress

is then calculated using a

relationship between the porosity, the normal compaction trend and the total

lithostatic stress

(overburden

stress).

A variety of empirical relationships

have been

developed for calculating mudrock

porosity

f r o m

different log

responses.

Typically, a stress-porosity relationship is not

used

directly, but instead porosity is compared

against

a normal compaction trend,

which would be the porosity

against

depth for the location in question assuming a

'normal'

pressure

profile

equivalent to the hydrostatic

head

of a water column. In this

work it w i l l be shown that the normal compaction trend often yields the biggest

uncertainties in calculating the pore pressure in mudrocks.

Carl Fredrik G y 1

lenhann 1

i;ii

8


22/176

Figure 1.4 Schematic of wireline logging. The lithological column to the right is a schematic of a

pressure probe

( R F T )

being used to measure the pore pressure in permeable sandstone.

Having inferred mudrock porosity f r o m logs and computed or established a normal

compaction trend of expected porosity for normal pressure, the f i n a l step is to f i n d a

relationship

quantifies the pore pressure magnitude associated w i t h a mismatch

between the estimated mudrock porosity f r o m log response and the normal

compaction trend.. This transform or equation might be based on physical principles,

such as the equivalent depth (or effective stress) method, or empirical relationships,

such as the Eaton's method. It

w i l l

be shown that the transform method used for

calculating the pore pressure is

less

important than the choice of normal compaction

trend.

Carl Fredrik Gyllenhammar

9


23/176

Chapter i

The

i n i t i a l

goal for this research was to establish a new method to calculate the pore

pressure

in mudrocks as a function of petrophysical

measurements. During

the

course

o f

this research it became

apparent

that the classical equivalent depth method is a

reliable equation and it

would

be of

l i m i t e d

value to attempt an improvement to it.

Also,

the porosity of the mudrocks can be reliably calculated f r o m a combination of

the available wireline logs. A sensitivity study shows clearly that the biggest

uncertainty is the normal compaction curve. Eaton (1975) summarized it

best:

"The

methods used

to establish normal

trends

vary as much as the number o f people who

do it".

A

normal compaction curve

represents

the

reference

trend describing the compaction

behaviour of

sediments which

are normally

pressured.

The compaction (porosity loss

involving expulsion of fluids) is caused by increases in vertical and /or horizontal

stress. Conventional pore

pressure

prediction uses the normal compaction curve to

estimate

the magnitude of

overpressure.

Data f r o m

which

normal compaction

curves

are derived include shallow buried

sediments

o f the same age and lithology, or

published compaction relationships. For example,

Hansen

(1996) examined

three

wells

in the

North

Sea where he

assumed

that the mudrocks have normal pore

pressure. He established a relationship between the sonic travel time and the mudrock

porosity

used

in this research. Other approaches are based on laboratory

measurements

of compaction such as by Skempton (1970) where he showed a

relationship between compaction and the volume of fine-grained material in the

samples. The shortcoming o f that approach is that the relationship does not take into

account the different compaction behaviours of clay minerals such as montmorillonite

versus

fine-grained quartz, (K.

Bjorlykke (2001)

personal oral cornmun.).

This research

shows

that it is unlikely that any useful normal compaction trend can be

established in the

North

Sea due to

recent glacial

events. The glacial

t i l l s l e f t

by a

earlier glacial event have been overlooked for many years. The nature of these

sediments

is found to be very different from normal marine and non-marine shale

mudrocks. This

suggest

that the previous method of establishing a normal trend by

overlaying

a number of porosity

curves

f o r m offset wells

w i l l

give wrong

results

i f

used

in basins such as the North Sea.

C a r l Frednk Gyllcnliammar

10


24/176

imroiiuaiijn

1.4 Pressure, the basic concepts

Fluids differ

f r o m solids in that they are unable to support

shear stress.

When a body

is

submerged in a

f l u i d

such as water, the

f l u i d

exerts a force perpendicular to the

surface at all locations around the surface of the body. I f the body is small enough so

we can neglect any differences in the vertical water column, the force (F) per

unit area

( A )

is the

same

in all directions. This force per

unit

area

is called the

pressure

P of the

f l u i d :

=

F/A [ E l . l ]

The SI

unit

of

pressure

is Newton per

square

meter ( N / m

2

) ,

which

is called

Pascal

(Pa). The equivalent imperial

unit

is pounds per

square inch

(psi =

l b / i n

).

Liquids found in rocks in the subsurface are relatively incompressible. This means

that the ratio of

mass

to volume, called density is approximately constant. For a

l i q u i d

whose density is constant, the

pressure increases

linearly

w i t h

depth. The

pressure

P at any point in a

l i q u i d

column is:

=P

0

+

pxgxh

[E1.2]

P

is the

pressure

at the surface and h is the vertical

l i q u i d

column. The Greek letter

p

(rho) is the density. Density has the

unit

mass/volume

(kg/m

3

= g/cm

3

).

g

is the

acceleration due gravity at the earth surface and equal to

9.81

m/s

2

.

Figure

1.5 shows a

simplified

diagram of how pore

pressure

may

increase

in a

w e l l .

The hydrostatic

pressure

(often called the normal

pressure)

in sediments underlying

the ocean often

follows

a gradient equal to 0.0101

MPa/m.

That is the

increase

in

hydrostatic

pressure

in water

w i t h

an average density of 1.03 g/cm

3

. The overburden

pressure

is the

pressure

exerted by all

overlying

material, both

solid

and f l u i d . Below

the water bottom, this

line

approximates 0.0226 MPa/m (1

psi/ft)

in a clastic

sedimentary environment. The pore

pressure

is the

pressure

of the

f l u i d

in the pore

space of the rock. It may be equal to or higher than the hydrostatic

pressure,

but not

higher than the overburden

pressure

(Figure 1.5). I f the pore

pressure approaches

the

overburden

pressure

the rock

w i l l

fracture and

release fluids.

However, often

fracturing w i l l occur at a lower pressure, equivalent to the least principal stress,

which

C a r l r iedi ik (Jyllenhammiir

11


25/176

Chapter I

Introduction

i n an extensional basin is

less

than the overburden (the vertical

stress).

I f at a specific

depth of

burial

the

mudrock

permeability

becomes

so low that the excess water

f r o m

normal compaction can no longer

f l o w

out of the system as fast as the

rate

of new

sediments, the pore

pressure

w i l l increase. The maximum

increase

o f pore

pressure

by

this

mechanism called disequilibrium compaction (Swarbrick and Osborne, 1997)-

and is often found to be parallel to the lithostatic gradient (Clayton and Hey, 1994),

indicating,

at depth, transfer of most/all of the load onto the pore f l u i d ,

w i t h

very

little/no increase w i t h

vertical effective stress..

S E A

sun

nidi)

1500

SHALE

s

\

4J

g 2500-

ressure

.

Q

3000 -

-

•

•

3500-

\

\

/J

4000-

SAND

3>\ ^

4500

I I I I I I

5000

70

-lo

l l

id

100

U

Ml

90

Pore

pressure

(MPa)

Figure 1.5 Pressure plotted against depth in a fictional well. The effective stress is equal to the

overburden pressure minus the pore pressure and the overpressure is equal to the pore pressure

minus the hydrostatic pressure.

I n a borehole, the pressure exerted by the

d r i l l i n g

f l u i d to either prevent

i n f l u x

of

pore

fluids

f r o m the formation or prevent hole caving

instability

is equivalent to

C a r l F r e d r i k Gyllenhammar

I 2


26/176

density of the

d r i l l i n g

f l u i d and its column height. Therefore, the formation pore

pressures are often converted into

d r i l l i n g

f l u i d density equivalents so it is clear as

what

d r i l l i n g

f l u i d

density just

balances

the pore

pressures.

Figure

1.6

shows how a

typical

pore

pressure

profile

can be displayed as

pressure

gradient

versus

depth.

I f

one

follows

the

change

in the

pressure

gradient of the pore

pressure

(red curve), every

point

on the curve

represents

a

pressure

gradient and a corresponding

average

f l u i d

density that particular

pressure

at that depth

represents.

The maximum pore

pressure

gradient is reached at the top of the reservoir (3200 meters) equal 0.016

MPa/m.

That

is

equivalent to the

pressure

at the bottom of a 3200-meter vertical f l u i d column

w i t h

an

average f l u i d

density of 1.64 g/cm

3

. In exploration

d r i l l i n g

a

d r i l l i n g

mud is used

where materials such as barite is mixed to f o r m a

l i q u i d

(called

d r i l l i n g

mud)

w i t h

such

high average

density. The terminology used is equivalent mud weight ( E q M W ) .

The

pressure

gradient

plot

illustrates a big challenge

while

d r i l l i n g

these wells.

The

EqMW

has to be

high

enough to

hold

back the f l u i d f r o m the depth where the

formation

has the highest-pressure gradient. However, in

some

formations,

typically

the shallower

ones,

this mud density

would

apply a

pressure

significantly

greater than

the pore

pressures

in

these

formations. This

excess

pressure

may lead to

fracturing

of

the rock and losses o f the

d r i l l i n g

f l u i d .

A

confusing

aspect

in the oil industry

w i t h

regard to

pressure

terminology is the

mixing the terms;

pressure

gradient and density ( E q M W ) . This

becomes

particularly

d i f f i c u l t

and confusing when

working w i t h

a mixture of both imperial and the SI

units.

It has already

been

shown that the

pressure

gradient

equals

density

multiplied

by

the acceleration due to gravity. In the imperial system, the norm is to use

weight

density

rather than density. Weight density is defined as the ratio of the weight of an

object to its volume. The units are pounds per gallon (ppg). As the weight is equal to

the

mass multiplied

w i t h

gravity, both weight density and

pressure

gradient

have

the

same units. The imperial

unit

system has historically been the norm in the oil industry

and the people

involved

has become used to converting directly f r o m weight density

(ppg)

to

pressure

gradient

(psi/ft)

and to

pressure

(psi).

The

word

weight density is

often

shortened to density. This has created a problem when converting to the SI

system. Too often,

while

converting f r o m density (g/cm

3

) to

pressure

gradient

(MPa/m),

density is not

multiplied

by gravity (9.81 m/s

2

). A

typical

example is a

recent paper

t i t l e d

"Pore Pressure terminology" in the Leading Edge

written

to explain

C ' a r i

I redrik Gyllenhammar

13


27/176

Chapter i

Introduction

the problem, but f a i l i n g to explain the difference between weight density and density

(Bruce and Bowers, 2002).

ijrn

Pore p rrnurr

gnoient

Effective stress

I.

O 3000

ft

4500

I—I-

J

I I

I I I I I I I I

L-LJ

000

n

j ;

.014

0.016

0 70 80 90 100 0.01 0.012

i l l

20 30

.i.

50

Pore pressure (MPa) Pore pressure gradient (MPa/m)

Figure 1.6 The Figure to the right shows how a pressure versus depth plot (left, Figure 1.5)

becomes

presented

as pressure gradient versus depth.

1.5 Aims and layout of

thesis

The aims and objective of this thesis are to:

1. Develop a critical review of current methods used to calculate the pore

pressure in mudrocks.

2 .

Establish the uncertainties o f the input variables using in principle

component analysis, applied to the wireline measurements w i t h reference

to the mudrock porosity calculated and the d r i l l i n g parameters w i t h

reference to the calculated

d r i l l i n g

exponents.

3. Identify the variables that have the biggest impact on the estimation of

pore pressure, and how they can be improved.

C a r l F r e d r i k Gyllenhammar 14


28/176

Chapter 1

In i roduct ion

4. Compare the wireline signature of overpressured shales in the North Sea

basin

w i t h

those f r o m the

G u l f

of

Mexico.

5. Examine why the resistivity

measurements

of the mudrocks can be used as

input

parameter

to calculate pore

pressure

in the

G u l f

of

Mexico,

while

this has proved d i f f i c u l t to apply in the estimation of pore pressure in the

North

Sea.

Following the introduction comes Chapter 2 where the pressure concepts w i t h

respects to pore pressure in shallow

sediments

are

discussed.

That is followed by a

discussion of mudrock porosity and normal compaction in mudrocks. Then the

different pressure

calculation methods,

first w i t h

wireline logs as input, then

those

using d r i l l i n g parameters.

Chapter 3 discusses the results f r o m using these different pore

pressure

estimation

methods

on a test

w e l l ,

Nor 1/6-7 in the North Sea. The sensitivity of the input

parameters

are

discussed.

The

results

f r o m the North Sea are then compared

w i t h

the

mudrocks

f r o m

a mini-basin in the

G u l f

of

Mexico,

Chapter 4

examines

the glacial history of the North Sea to explain the

nature

of the

shallow

sediments,

and their physical and petrophysical properties. Use of a novel

application of the software Cyclolog has helped in characterising the glacial

sediments. Finally the relevance o f the glacial history of the North Sea is reviewed in

relation to the petroleum system which has

generated

productive oil and gas fields .


29/176

Ch apter 2 Pore Pressure Evaluation

Concepts

and

definition;-

Chapter 2 Pore Pressure Evaluation

Concepts

and

definitions

C a r l F r e d r i k Gvlienharnmar

16


30/176

C i K i p i c r

2

Poi

v. Pi'fssm'c- )_• \. Ina ion Concep t and

ck'iinilions

2.1 Definition

Underpinning

pore pressure interpretation is the effective stress equation for porous

media (Terzhagi, 1936):


31/176

Chapter 2

Pore Pressure

.Evaluation

Conce pts and d eii niii ons

2.1.1 Mudrock porosity

Blatt

(1970) has defined mudrock

based

on grain size, where

mud

is sediment

composed of clay sized particles. Typically mudrocks contain some

s i l t .

A

mudstone is a sedimentary rock composed of

l i t h i f i e d

mud, and

shale

is a fissile

mudstone. The term porosity has a different meaning in various disciplines as

w e l l

as

being different for

coarse

grain

sandstone

when compared

w i t h

a mudrock. The

porosities

discussed

here w i l l be l i m i t e d to the physical or total porosity, which is the

ratio

of v o i d volume to total volume.

The preferred method of obtaining the porosity in a rock is to carry out laboratory

experiments on

core

extracted

f r o m

the

w e l l

during

d r i l l i n g

operations. The porosity

o f

low permeability rocks such as mudrocks is

measured f r o m

the bulk density, then

drying

the sample,

followed

by

measurement

of the dry density in the laboratory.

This

procedure ideally must be commenced prior to the

samples

drying after reaching

the surface. On

research

vessels

such

used

during the

Ocean D r i l l i n g

Program (ODP),

these measurements

are

done

just after the

samples

are recovered at surface. Mudrock

is generally not cored during exploration

d r i l l i n g .

I f it is cored, the

samples

are waxed

at the wellsite so the water content is

preserved.

Mudrock porosity as

w e l l

as general rock porosity

f r o m

exploration wells is in most

cases

calculated

f r o m

wireline measurements such as the sonic log, the density log or

the neutron log. None of these measurements are a direct

measurement

of porosity.

They are referred to as log-derived porosity to indicate that their

o r i g i n

is

f r o m

wireline

log

responses.

For all

these

instruments, the

tool

response

is affected by the

formation

porosity, f l u i d and matrix. I f the f l u i d and matrix effects are known, the

porosity can be derived

f r o m

the tool response.

I n addition to the above tools, the resistivity response can also

used

to determine

porosity. However the resistivity is greatly influenced by the

f l u i d

saturation.

2.1.2 Different porosity evaluation

equations

Sonic derived porosity

C a r l

I

reclrik Gvllenhammar 18


32/176

W y l l i e et

a l .

(1958) demonstrated that

there

was an approximate linear relationship

between sonic velocity and porosity in

sandstone.

The porosity is calculated f r o m a

linear interpolation between the zero porosity matrix sonic velocity (in principle

slowness

when using |isec/ft unit) and the

100

% porosity

fluid

sonic velocity.

f

t,

-t

log

ma

[E2.5]

where; t

ma

is the matrix velocity (67 (isec/ft in mudrock, 47.5 |isec/ft in chalk, 55.5

iisec/ft in sandstone) and // the fluid velocity equal 189 (isec/ft in fresh water

(Schlumberger, 1989). t\„

is the measured sonic velocity.

Another equation was suggested by

Raiga-Clemenceau

et al. (1988):

0

=

1-

f

At

^

ma

At

[E2.6]

The matrix velocity

t

ma

and x are both constants that are basin specific. Raiga-

Clemenceau called " x " the acoustic formation factor exponent.

Issler (1992) developed this relationship using data f r o m the Beaufort-Mackenzie

Basin, Northern Canada where the shales are quite uniform in their composition. The

matrix transit time is the same as for mudrocks in the W y l l i e equation (67 |isec/ft)

( W y l l i e , 1958) and the x was calculated to 2.19.

, , 6lY

2A9

p =

1

- — i

A t

J [E2.7]

Hansen

11 (1996), using

shale densities

measured on sidewall and cuttings

samples

f r o m the Nor th Sea, modified this equation. He

suggests

using the

f o l l o w i n g

equation

where the shale matrix velocity is 76.5 |isec/ft and x=

1.17:

0 = 1-

^ 7 6 . 5 ^

1 7

At J

[E2.8]

C u r l I icdrik Gvllenh amnia r 19


33/176

Cha pter 2 i'u;e

i ' l v ^ u r c i . wduaia

at

Coincpiv a Vi

dc : retaat--

Figure 2.1a show how the porosity in a mudrock

w i l l

change as a func tion of

sonic

velocity. In the shallow section where the velocities often are

150 L i s e c

/ f t the

Hansen

(1996)

model suggests 44 % porosity versus the W y l l i e

(1958)

equation estimation of

68 %. The

W y l l i e

equation, although

based

on an empirical relationship in

sandstones

is used to calculate mudrock porosity in

several

publications

(e

.g.

Hermanrud

et

al.,

1998).

Iss er

ii 0

Hansen

Wyllie 67

i

:

u

41

Wyllie 76.5

0

fO

Wyllie

(68.8)

0 30 5

P 0.40

u . 0

0

20

0 10

) 00

o c o

2

15

75

5

90 100 110 120 130 140 150 160 170

:

Si

Bulk

Dens i t y

S o n i c u sec / f t )

0.315

o

)?

J

J13

(I 10

0.J11

o ;'5

a-

0 26

)00

4

n ir,7

i: 2.

0 20

tot

2

S

0=

1

04 • 06

1 J?

: OS

01

Matrix

Dens i t y

ater

Dens i t y

Figure 2.1 a,b,c,d. la porosity variation as a function of sonic velocity. B, porosity versus bulk

density. C , the sensitivity to pore water density. D, the sensitivity to matrix density.

Density derived porosity

C a r

F r e d r i k GvUenhammai

20


34/176

Chapier 2

Pore

Pressure Evalua tion Concepts

and definitions

Density log-derived porosity is calculated f r o m the log

bulk

density, the matrix

density and the pore f l u i d density (equation 2.9).

0:

\

Pm a

~

Pb

Pn,a P f

[E2.9]

where p

ma

is the matr ix density, Pf is the pore water density and ph is the

bulk

density

measured by the density

t o o l .

In mudrocks

f r o m

the

North

Sea,

2.715 g/cm

was used

as an

average

matrix density. This value is used by

B r i t i s h

Geological Survey (BGS)

and is

based

on their shallow coring program in the

North

Sea. The pore water density

is

assumed

to

increase

f r o m

1.03

to

1.08

g/cm

3

w i t h

depth. As the

shale

compacts, the

released water has a lower salt concentration than the remaining pore water. Figure

2.1b shows the porosity variation, as a function of

bulk

density. The matr ix density

and the fluid density are kept constant at 2.715

g/cm

3

and 1.03 g/cm

3

while

the

measured

bulk

density increases f r o m 1.75 to 2.75

g/cm

3

.

The porosity varies linearly

w i t h

the measured

bulk

density and an

increase

of 0.1

g/cm

3

changes the porosity by

5.8 %. The second sensitivity

plot

(Figure 2.1c) shows that increasing the f l u i d

density f r o m

1.03

to

1.08 g/cm

3

only

increases the porosity by

1

%. Figure 2.

I d

shows

that the porosity

changes

by 4 % if the matr ix density

changes

by

0.1

g/cm . The

relationship

between porosity and matrix density is not linear, but

near

linear. The

porosity increases slightly

faster at lower matrix densities that at the higher end. The

matrix

density is a

function

o f

mineralogy.

Smectite has a low matrix density

(2.21

to

2.71 g/cm

3

)

while

chlorite matrix density can be as

high

as 2.94

(Fertl

and

Chilingarian,

1989). In the

North

Sea the mudrock compositions vary and therefore

so do the dry densities. Using a constant dry density w i l l therefore result in 10 to 20%

error in calculated the porosity using the density log.

Neutron derived porosity

Neutron-derived

porosity is related to the hydrogen index,

which

is an indication of

the amount of hydrogen in the sediment. As most of the hydrogen in a

formation

is in

the water and hydrocarbon molecules it is for all practical

purposes

a

measure

of the

water and/or hydrocarbon content. In formations

w i t h

phyllosilicates, the bound water

w i l l

be counted as water, and

hence v o i d space,

by the neutron log. When comparing

C a r l Predrik Gvlle nha mma r 21


35/176


36/176

Chapter 2

Fore Pressure

(-.valuation

Concepts and definitions

i or.

Sonic

{Wy t l i e } -

porosity

:i

9 0

S. Hansen

-

porosity

Density

-

porosity

i.«0

Neutron

-

porosity

L O G

-

porosity

n

7f

1

S

0.60

i

i

it

•

0.50

1

.

i

J

i

S

0.30

0 20

0 10

L

C no

500 1000 1500

2000 2500 3000 3500 4000 4500 5000

D e p t h m R K B )

Figure 2.2 Log derived porosities in well Nor-1/6-7 Norway. The low porosity interval from

3261m

to

4346m

depth is the Cretaceous Chalk. The values are listed in Appendix 1.

Based on Figure 2.2 it is important to realize the uncertainties that exit in the porosity

estimates

in

shales

based on wireline logs. This put limitations on the conclusions we

may wish to draw concerning the

mechanisms

underlying the porosity reduction or

compaction. I f one only has available a wireline log-derived porosity profile, one can

clearly

not attribute the porosity change solely either to a mechanical

process

or to a

chemical process.

2.1.3 Normal compaction curve and trend lines

During

normal compaction, a mudrock

undergoes

a monotonic

increase

in effective

stress,

which

causes

an elastoplastic reduction in porosity. Compaction is a result of

grain

reorientation and

breakage.

Mudstone

consists

of clay minerals, fine grained

quartz, feldspar and mica. As the compressibility is dif ferent for different minerals as

w e l l

as for different clay minerals, the mudrock compressibili ty becomes very

d i f f i c u l t

to predict. The resultant relationship between effective

stress

and porosity is

known

as the normal compaction curve (Harrold et

a l . ,

1999). In this

case

equilibrium

is reached such that:

Pf (pore

pressure)

= ph

yc

j (hydrostatic pressure).

Curl

Fredrik Gyllenhammar

23


37/176

Chapter 2

Pore

Pressure b\

alum io n Concept : and definitions

Although

many porosity - depth

data have been

published, details of age, lithology or

effective

stress are generally

absent.

In this study it was chosen to evaluate and

compare two different relationships: (1) A t h y type and (2) Soil-mechanical type.

These compaction curves

assume

mechanical compaction

only,

and are suitable

only

to describe siliciclastic sediments. Below 2-3 km depth ( 7 0 - 1 0 0 ° C ) , mineral

dissolution

and precipitation becomes important (Bjorlykke, 1999). At these

temperatures hydrocarbon generation also comes into play. There have been many

publications on attempts to assess the potential overpressure generated by these

reactions. The results are conflicting in the

sense

that for the same reaction, some

suggest

that no overpressure is generated

while

others

suggest

generation of large

overpressure. The

conflict

lies to a large degree in the assumed permeability. For

many of

these

reactions to

generate

overpressure, the permeability

w i l l

not be low

enough for overpressure to be retained over geological time (Osborne and Swarbrick,

1999).

It is also evident that w i t h all the uncertainties w i t h regard to chemical

compaction or chemical reactions in mudrocks, it would be quite impossible to predict

the normal compaction trend and hence impossible to calculate the pore

pressure.

Chemical compaction w i l l therefore not be taken into account in the

present

study.

2.1.3.1

Athy-type relationship

The exponential curve to describe compaction was introduced by A t h y

(1930).

It was

based on curve f i t t i n g a particular data set and is given as:

Where 0 is porosity at depth of interest, 00 is porosity at sea bed, c the compaction

coefficient and z the depth. Variations of the compaction curve result

f r om

substituting

depth w i t h mean or vertical effective

stress.

The A t h y compaction curve

was later

modified

by Hubbert and Rubey

(1959),

who recognised that porosity is

controlled

by

effective

stress and not by depth:

0 —

0oe

(-«)

( A t h y , 1930)

[E2.10]

0 = K 0

o

e

c

[ E 2 . l l ]

Carl Predrik Gvllenhammar 24

http://e2.ll/http://e2.ll/


38/176

Chapter 2

Pore

Pressure

kvaluation

Concepts and

ueiimhuns

Where


39/176

Chap

Pore Pressure Evaluation Concepts and definitions

reach

0 % porosity, only at

i n f i n i t y .

The sea

floor

porosity is a factor that can be

related to

samples.

The compaction coefficient is a function of rock compressibility,

which

in theory could be

measured

in the laboratory ( A t h y , 1930). But the

compressibility w i l l vary

w i t h

depth as the rock

becomes

more consolidated. In

practice what is being

done

is to use a

w e l l

where the pore

pressure

is

assumed

hydrostatic

based

on the M W

used

and RFT

pressure

points. Then calculate the

mudrock porosity to calibrate the compaction coefficients in the normal compaction

equation.

Figure 2.3

shows

how different the two

equations express

the change in porosity

w i t h

increased

total

stress.

The

soil

mechanical function

suggests

a larger

rate

of porosity

reduction in the shallow section and

w i l l

always end up w i t h zero porosity i f the total

stress gets

large enough. The

A t h y

function suggests a more

moderate

change of

porosity in the shallow section.

W i t h

increasing stress, the porosity

w i l l

move

asymptotically to zero porosity, but never

become

zero.

Normal

compaction

Lo g

derived porosity

0.1

0.4

5000

10000

15000

20000

L0

25000

30000

Athy, 0.7 0.00008

35000

40000

SM , 2 (0.66), 0.74

45000

50000

Figure 2.3 Comparison of porosity with effective stress for the Athy and the SM equations. Initial

porosity (sea floor porosity) for Athy is 0.7 (70 ) while the porosity at 100 kPa (approximately

100 meters below sea floor) is 0.66 (66 ) The compaction factors a re for Athy; 0.00008 and

SM ; 0.74.

Ca r l Fredr ik

Gyllenhammar

26


40/176

Chapter 2

I'orc Pressure valuation Concepts and definitions

100 (porosity in

% )

0

40

60 80

Porosity

0

Solidity

°

1.5

Void Ratio

•

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Void ratio

Figure 2.4 The relation ship between porosity, solidity and void ratio is shown. The y-axis is the

compaction as a

length

reduction. It is assumed that a confined volume is compressed beginning

with a void ratio of four.

Figure 2.4

shows

the

three

different frames of

reference

to

describe

the

loss

of pore

space. This is a theoretical model

bu i l t

in

EXCEL based

on the

definition

of porosity,

solidity

and v o i d ratio. The compaction as length reduction on the y-axis

represents

the proportional thickness reduction and the corresponding porosity,

solidity

and vo id

ration. Solidity

is the volume of

solid

grains as a

percent

of the total volume of

sediment, i.e. (1- ())). The complement to

solidity

is porosity as the volume of pore

space as a

percent

of the total volume of sediment. This is the opposite of what has

been suggested

by

Baldwin

and Butler (1985). The t h i r d

parameter

is v o i d ratio,

which

is the ratio of the volume of pore space and the volume of solids. Figure 2.4

shows

that

there

is a linear relationship between v o i d ratio and compaction

while

the

relationship is non-linear between porosity as

w e l l

as

solidity

to compaction. It

would

therefore be mathematically

easier

to

describe

the compaction as a function of

vo i d

ratio

rather than porosity. The

reason

why the oil industry

uses

porosity is that it has

become

the convention in the reservoir section. By comparison the convention in the

soil mechanics

environment is to use v o i d ratio as compaction is of interest.

Ca r l

Fredr ik Gyllenhammar

27


41/176

2.1.4 Vertical versus mean effective stress

The mean effective

stress

(cr

m

)

is defined as the difference between the mean

�

A Critical Review of Cerrently Pore Pressure Methods

Documents

Transcript of A Critical Review of Cerrently Pore Pressure Methods