Module 8: Relative Permeability

Synopsis
What is water-oil relative permeability and why does it matter?
endpoints and curves, fractional flow, what curve shapes mean

Understand the jargon (and impress reservoir engineers) Wettability

water-wet, oil-wet and intermediate

How do we measure it (in the lab)? How do we quality control and refine data?

Page 2

Applications
To predict movement of fluid in the reservoir
e.g velocity of water and oil fronts

To predict and bound ultimate recovery factor Application depends on reservoir type
gas-oil water-oil gas-water

Page 3

Definitions
Absolute Permeability
permeability at 100% saturation of single fluid
e.g. brine permeability, gas permeability

Effective Permeability
permeability to one phase when 2 or more phases present
e.g. ko(eff) at Swi

Relative Permeability
ratio of effective permeability to a base (often absolute) permeability
e.g. ko/ka or ko/ko at Swi
Page 4

Requirements
Gas-Oil Relative Permeability (kg-ko)
solution gas drive gas cap drive

Water-Oil Relative Permeability(kw-ko)

water injection

Water - Gas Relative Permeability (kw-kg)

aquifer influx into gas reservoir

Gas-Water Relative Permeability (kg-kw)

gas storage (gas re-injection into gas reservoir)
Page 5

Jargon Buster!
Relative permeability curves are known as rel perms Endpoints are the (4) points at the ends of the curves The displacing phase is always first, i.e.:
kw-ko is water(w) displacing oil (o) kg-ko is gas (g) displacing oil (o) kg-kw is gas displacing water

Page 6

Why shape is important

Measure air permeability Saturate core in water (brine) Desaturate to Swir
Centrifuge or porous plate Measure oil permeability ko @ Swir endpoint
Ko = 80 mD

ka = 100 mD Swir = 0.20 (20%

So = 1-Swir Swirr

Oil = Sro Sw = 1-Sro

Waterflood collect water volume

Swr = 1-0.25 = 0.75

Sro = 0.25

Measure water permeability kw @Sro endpoint

Page 7

Kw = 24 mD

Endpoints
1.0 0.9 0.8 0.7 0.6

Endpoint- oil kro = ko/ko @ Swir = 80/80 =1 Swir = 0.20 Sro = 0.25

Relative Permeability (-)

0.5 0.4 0.3 0.2 0.1

Endpoint - water krw = kw/ko @ Swir = 24/80 = 0.30

0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7 0.8 0.9 1.0

Page 8

Endpoints
1.0 0.9 0.8 0.7 0.6

Relative Permeability (-)

Swir = 0.20
0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7

Sro = 0.25

0.8

0.9

1.0

Page 9

Curves - 1
1.0 0.9 0.8 0.7 0.6

Relative Permeability (-)

Swir = 0.20
0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7

Sro = 0.25

0.8

0.9

1.0

Page 10

Curves - 2
1.0 0.9 0.8 0.7 0.6

Relative Permeability (-)

Swir = 0.20
0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7

Sro = 0.25

0.8

0.9

1.0

Page 11

Curves - 3
1.0 0.9 0.8 0.7 0.6

Relative Permeability (-)

Swir = 0.20
0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7

Sro = 0.25

0.8

0.9

1.0

Page 12

Relative Permeability
1

Non-linear function of Swet Competing forces

gravity forces
Relative Permeability (-)

0.9

0.8

0.7

0.6

minimised in lab tests e.g. water injected from bottom to top

0.5

kro krw

0.4

viscous forces
Darcys Law

0.3

0.2

0.1

capillary forces
low flood rates
Page 13

0 0 0.2 0.4 0.6 0.8 1 Water Saturation (-)

Relative Permeability Curves Key Features

Water-Oil Curves
irreducible water saturation (Swir) endpoint
kro = 1.0 krw = 0.0

residual oil saturation (Sro) endpoint

kro = 0.0 krw = maximum

relative permeability curve shape

Unsteady-state Steady-state Corey exponents:
Page 14

Buckley-Leverett, Welge, JBN Darcy No and Nw

Waterflood Interpretation
Welge
fw=1

fw only after BT
Average Saturation behind flood front

Swf , fw | S

wf

fw
Sw at BT

fw =

1 +

k ro . k rw

w o

Page 15

Swc

Sw

1-Sor

Relative Permeability Interpretation

Welge/Buckley-Leverett fraction flow
gives ratio: kro/krw

fw =

1 +

k ro . k rw

k rw o . M= k ro w
M< 1: piston-like M > 1: unstable

w o

Decouple kro and krw from kro/krw

JBN, Jones and Roszelle, etc

Page 16

JBN Method Outline

Johnson, Bossler, Nauman (JBN)
Based on Buckley-Leverett/Welge W = PV water injected Swa = average (plug) Sw fw2 = 1-fo2
fw = 1+ k ro w . k rw o 1

dS wa = fo2 dW
d( 1 ) f WI r = o2 1 k ro 2 d( ) W

pt =0 Ir = pt =i
Page 17

Injectivity Ratio Waterflood rate, q

Buckley Leverett Assumptions

Fluids are immiscible Fluids are incompressible Flow is linear (1 Dimensional) Flow is uni-directional Porous medium is homogeneous Capillary effects are negligible Most are not met in most core floods

Page 18

Capillary End Effect

If viscous force large (high rate)
Pc effects negligible

If viscous force small (low rate)

Pc effects dominate flood behaviour

Leverett
capillary boundary effects on short cores boundary effects negligible in reservoir

Page 19

End Effect
Pressure Trace for Flood zero p (no injection) start of injection water nears exit
p increases abruptly until Sw(exit) = 1-Sro and Pc nears zero suppresses krw Sw(exit) = 1-Sro, Pc ~0 rate of p increase reduces as krw increases

BT

Page 20

After BT

Scaling Coefficient
Breakthrough Recovery (Rappaport & Leas) Affected by Pc end effects At lengths > 25 cm Little effect on BT recovery (LVw > 1) Hence composite samples or high rates

Page 21

Capillary End Effects

Rapaport and Leas Scaling Coefficient
LVw > 1(cm2/min.cp) : minimal end effect

Overcome by:
flooding at high rate
300 ml/hour +

using longer cores

difficult for reservoir core (limited by core geometry) butt several cores together

using capillary mixing sections

end-point saturations only in USS tests (weigh sample)
Page 22

Composite Core Plug

Capillary end effects adsorbed by Cores 1 and 4

Page 23

Corey Exponents Water/Oil Systems

Define relative permeability curve shapes Based on normalised saturations No guarantee that real rock curves obey Corey
kro = SonNo krw = krw(Swn)Nw krw = end-point krw

1 S w Sro Son = = 1 S wn 1 S wi Sro

S wn
Page 24

S w S wi = 1 S wi S ro

Normalisation
Swn = 1
1 0.9 0.8

Water Relative Permeability (-)

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7 0.8 0.9 1 krw at Sro krwn = 1
Sample 1 Sample 2

krwn = 1

Page 25

Corey Exponents
Depend on wettability
Wettability Water-Wet Intermediate Wet Oil-Wet No (kro) 2 to 4 3 to 6 6 to 8 Nw (krw) 5 to 8 3 to 5 2 to 3

Uses:
interpolate & extrapolate data lab data quality control
Page 26

Gas-Oil Relative Permeability

Pore-Scale Saturation Distribution

Test performed at Swir

Gas is non wetting takes easiest flow path kro drops rapidly as Sg increases krg higher than krw Srog > Srow in lab tests
end effects

Srog < Srow in field

Page 27

Sgc ~ 2% - 6%

Typical Gas-Oil Curves: Linear

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Gas Saturation (fractional)

Relative Permeability (-)

1-(Srog+Swi)
kro krg

Sgc

Page 28

Labs plot kr vs liquid saturation (So+Swi)

Typical Gas-Oil Curves: Semi-Log

1

Relative Permeability (-)

0.1

1-(Srog+Swi)
kro krg

0.01

0.001 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Gas Saturation (fractional)

Page 29

Gas-Oil Curves
Most lab data are artefacts
due to capillary end effects
Tests should be carried out on long cores

insufficient flood period

Real gas-oil curves

Sgc ~ 3% Srog is low and approaches zero
Due to thin film and gravity drainage

krg = 1 at Srog = 0
Page 30

well defined Corey exponents

Gas-Oil Curves Corey Method

Oil relative permeability
normalised oil saturation

kro = Son No
Son = 1 Sg Swir Srog 1 Swir Srog

Gas relative permeability

normalised gas saturation
Sgc: critical gas saturation
Corey Exponent No Ng
Page 31

krg = Sgn Ng
Sgn = Sg Sgc 1 Swir Srog Sgc

Values 4 to 7 1.3 to 3.0

Corey Gas-Oil Curves

1

0.1

Swir kro krg' Srog Sgc

Kro No = 4 krg Ng = 1.3 kro No = 7 krg Ng = 3.0

0.15 1.00 1.00 0.0000 0.0300

Relative Permeability (-)

0.01

0.001

Sgc = 0.03

0.0001

0.00001 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Gas Saturation (-)

Page 32

Typical Lab Data - krg

Krg too low
1

Srog too high

0.1

Relative Permeability, krg

0.01

0.001
Composite Gas-Oil Curves Ng : No : Sgc: Srog: krg' : 2.3 4.0 0.03 0.10 1.0

Ng = 2.3; Swir = 0.15 Ng = 2.3; Swir = 0.20 11a-5 # 4 11a-5 # 31 11a-5 # 34 11a-5 #39 11a-7 BEA5 11a-7 BEA7 11a-7 BEB5 11a-7 BEC5

0.0001

0.00001 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Swi+Sg (fraction)

Page 33

Laboratory Methods
Core Selection
all significant reservoir flow units often constrained by preserved core availability core CT scanning to select plugs

Core Size
at least 25 cm long to overcome end effects butt samples (but several end effects?) flood at high rate to overcome end effects?

Page 34

Test States
Fresh or Preserved State
tested as is (no cleaning) probably too oil wet (e.g OBM, long term storage) Native state term also used (defines bland mud) Some labs fresh state is other labs restored state

Cleaned State
Cleaned (soxhlet or miscible flush) water-wet by definition (but could be oil-wet!!!!!!)

Restored State (reservoir-appropriate wettability)

saturate in crude oil (live or dead) age in oil at P & T to restore native wettability
Page 35

Test State
Fresh-State Tests
too oil wet data unreliable data unreliable

Cleaned-State Tests
too water wet (or oil-wet)

Restored-State Tests

Page 36

native wettability restored data reliable (?) if GOR low can use dead crude ageing (cheaper) if GOR high must use live crude ageing (expensive) if wettability restored - use synthetic fluids at ambient ensure cores water-wet prior to restoration

Compare methods - are there differences?

Irreducible Water Saturation (Swir)

Swir essential for reliable waterflood data Dynamic displacement
flood with viscous oil then test oil rapid and can get primary drainage rel perms Swir too high and can be non-uniform

Centrifuge
faster than others Swir can be non-uniform

Porous Plate
slow, grain loss, loss of capillary contact
Page 37

Swir uniform

Lab Variation in Swir (SPE28826)

30

Dynamic Displacement Porous Plate

25

20 Swi (%)
180 psi ???

15

10

200 psi

0 Lab A Lab B Lab C Lab D

Page 38

Centrifuge Tests
Displaced phase relative permeability only
oil-displacing-brine : krw drainage brine-displacing-oil : kro imbibition assume no hysteresis for krw imbibition
oil-wet or neutral wet rocks?
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7 0.8 0.9 1.0

Good for low kro data (near Sro) Computer simulation used Problems
uncontrolled imbibition at Swirr mobilisation of trapped oil sample fracturing
Page 39
Relative Permeability (-)

e.g. for gravity drainage

Dynamic Displacement Tests

Test Methods
Waterflood (End-Points: ko at Swi, kw at Srow) Unsteady-State (relative permeability curves) Steady-State (relative permeability curves)

Test Conditions
fresh state cleaned state restored state ambient or reservoir conditions

Page 40

Unsteady-State Waterflood
Saturate in brine Desaturate to Swirr Oil permeability at Swirr (Darcy analysis) Waterflood (matched viscosity)
o w o = w res lab

Total Oil Recovery kw at Srow (Darcy analysis)

Page 41

Unsteady-State Relative Permeability

Saturate in brine Desaturate to Swirr Oil permeability at Swirr (Darcy analysis) Waterflood (adverse viscosity)
o o >> w lab w res

Incremental oil recovery measured kw at Srow (Darcy analysis) Relative permeability (JBN Analysis)
Page 42

Unsteady-State Procedures
Water Oil Only oil produced Measure oil volume

Just After Breakthrough Measure oil + water volumes

Increasing Water Collected Continue until 99.x% water

Page 43

Unsteady-State
Rel perm calculations require
fractional flow data at core outlet (JBN) pressure data versus water injected

Labs use high oil/water viscosity ratio

promote viscous fingering provide fractional flow data after BT allow calculation of rel perms

Waterflood (matched viscosity ratio)

little or no oil after BT little or no fractional flow (no rel perms) end points only
Page 44

Effect of Adverse Viscosity Ratio

1.0 0.9 0.8 0.7 Fractional Flow, fw 0.6 0.5 0.4 0.3

o/w = 30:1 Unstable flood front Early BT Prolonged 2 phase flow Oil recovery lower o/w = 3:1 Stable flood front BT delayed Suppressed 2 phase flow Oil recovery higher

0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7 0.8 0.9 1.0

Page 45

Unsteady-State Tests
Only post BT data are used for rel perm calculations
Sw range restricted if matched viscosities

Advantages
appropriate Buckley-Leverett shock-front reservoir flow rates possible fast and low throughput (fines)

Disadvantages
inlet and outlet boundary effects at lower rates complex interpretation
Page 46

Steady-State Tests
Intermediate relative permeability curves
Saturate in brine Desaturate to Swir Oil permeability at Swir (Darcy analysis) Inject oil and water simultaneously in steps Determine So and Sw at steady state conditions kw at Srow (Darcy analysis) Relative Permeability (Darcy Analysis)

Page 47

Steady-State Test Equipment

p

Oil in

Water in

Mixing Sections

Coreholder

Oil and water out

Page 48

Steady-State Procedures
Summary
100% Oil:

ko at Swirr ko & kw at Sw(1) ko & kw at Sw(2)

Ratio 1: Ratio 2: . . Ratio n:

ko & kw at Sw(n)

100% Water: kw at Sro

Page 49

Steady-State versus Unsteady-State

Constant rate (SS) vs constant pressure (USS)
fluids usually re-circulated

Generally high flood rates (SS)

end effects minimised, possible fines damage

Easier analysis
Darcy vs JBN

Slower
days versus hours

Endpoints may not be representative Saturation Measurement

gravimetric (volumetric often not reliable) NISM
Page 50

Laboratory Tests
You can choose from:
matched or high oil-water viscosity ratio cleaned state, fresh state, restored-state tests ambient or reservoir condition high rate or low rate USS versus SS

Laboratory variation expected

McPhee and Arthur (SPE 28826) Compared 4 labs using identical test methods
Page 51

Oil Recovery
70 Fixed - 120 ml/hour 60 Oil Recovery (% OIIP) Preferred 360

50

40

120

30

120

20 Bump 10 Lab A Lab B Lab C Lab D

Page 52

Gas-Oil and Gas-Water Relative Permeability

Unsteady-State
adverse mobility ratio (g<<o or w) prolonged two phase flow data after breakthrough drainage tests reliable imbibition tests difficult

Steady-State
kg-ko, kg-kw and kw-kg saturation determination difficult much slower

Gas humidified to prevent mass transfer

Page 53

Drainage Gas-Water Curves (steady-state)

Steady-state test example Log-linear scale (very low krw) Krg > krw Gas saturation increases Krg increases to 1 Krw reduces to close to zero

Page 54

Water-Gas Relative Permeability

Aquifer influx (imbibition) Drainage gas-water curves can be used but
hysteresis expected for non-wetting phase (krg) curve no hysteresis for wetting phase (krw) curve
drainage krw curve same shape as imbibition krw curve

Imbibition tests require

low rate imbibition waterflood kw-kg test
capillary forces dominate

CCI tests for residual gas saturation Hybrid test

Page 55

Imbibition Tests
Waterflood
low rate waterflood from Swi to Sgr obtain krg and krw on imbibition Sgr too low (viscous force dominates)
129.90 g

Counter-Current Imbibition Test

Sgr dominated by capillary forces immerse sample in wetting phase (from Sgi) monitor sample weight during imbibition Determine Sgr from crossplot

Page 56

CCI: Experimental Data

Air-T oluene CCI: Plug 10706: Sgi = 88.8%
70 65 60 Gas Saturation (%) 55 50 45 40 35 30 0 10 20 30 Square R T (se c s) oot ime 40 50 60 Sgr = 33.5%

Page 57

Trapped or Residual Gas Saturation

Sgr vs Sgi North Sea

Low rate waterflood

Page 58

Repeatability of CCI tests

Imbibition Kw-Kg
1

krw@Sgr krg 1-Sgr Swi

Drainage

kr

Imbibition

krw 0
Page 59

Sw

Relative Permeability Controls

Wettability Saturation History Rock Texture (pore size) Viscosity Ratio Flow Rate

Page 60

Wettability

Page 61

Wettability

Page 62

Wettability
Waterflood of Water-Wet Rock

Page 63

front moves at uniform rate oil displaced into larger pores and produced water moves along pore walls oil trapped at centre of large pores - snap-off BT delayed oil production essentially complete at BT water invades smaller pores earlier BT oil remains continuous oil produced at low rate after BT krw higher - fewer water channels blocked by oil

Waterflood of Oil-Wet Rock

Effects of Wettability
Water-Wet
better kro lower krw krw = kro > 50% better flood performance poorer kro higher krw kro = krw < 50% poorer flood performance

Oil-Wet

Page 64

Wettability Effects: Brent Field

Preserved Core Neutral to oil-wet low kro - high krw Extracted Core Water wet high kro - low krw

Page 65

Importance of Wettability - Example

Water Wet
No = 2 Nw = 8 Swir = 0.20 Sro = 0.30, krw = 0.25, ultimate recovery = 0.625 OIIP

Intermediate Wet
No = 4 Nw = 4 Swir = 0.15 Sro = 0.25, krw = 0.5, ultimate recovery = 0.706 OIIP

Oil Wet
No = 8 Nw = 2 Swir = 0.10 Sro = 0.20, krw = 0.75, ultimate recovery = 0.778 OIIP
Page 66

o/w = 3:1

Relative Permeability Curves

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Water Saturation (-)
WW kro WW krw

Page 67

Relative Permeability (-)

Relative Permeability Curves

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Water Saturation (-)
WW kro WW krw IW kro IW krw

Page 68

Relative Permeability (-)

Relative Permeability Curves

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Water Saturation (-)
WW kro WW krw IW kro IW krw OW kro OW krw

Page 69

Relative Permeability (-)

Fractional Flow Curves

1.0 0.9 0.8 0.7 Fractional Flow, fw (-) 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Water Saturation (-)
WW fw

Water Wet SOR = 0.33 Recovery = 0.59

Page 70

Fractional Flow Curves

1.0 0.9 0.8 0.7 Fractional Flow, fw (-) 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Water Saturation (-)

IW SOR = 0.44 Recovery = 0.482

WW fw IW fw

Page 71

Fractional Flow Curves

1.0 0.9 0.8 0.7 Fractional Flow, fw (-) 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Water Saturation (-)
WW fw IW fw OW fw

Oil Wet SOR = 0.63 Recovery = 0.300

Page 72

Costs of Wettability Uncertainty

PV Oil Price Parameter Swi Ultimate Sro Ultimate Recovery Factor SOR Actual Recovery Factor STOIIP (MMbbls) Ultimate Recovery (bbls) Actual Recovery (bbls) "Loss" (MM US$) Water-Wet 0.200 0.300 0.625 0.330 0.588 96 60 56 108 120 MMbbls 30 US$/bbls IW 0.150 0.250 0.706 0.440 0.482 102 72 49 684 Oil wet 0.100 0.200 0.778 0.630 0.300 108 84 32 1548

It is really, really important to get wettability right!!!

Page 73

Rock Texture

Page 74

Viscosity Ratio
krw and kro - no effect ? End-Points - viscosity dependent Hence: use high viscosity ratio for curves use matched for end-points

Not valid for neutral-wet rocks (?)

Page 75

Saturation History
100 %

Primary Drainage

Primary Imbibition

NW

No hysteresis in wetting phase NW

kr
Swi W

kr
Sro

W 0% 0%
Page 76

0%

Sw

100 %

0%

Sw

100 %

Flow Rate
Reservoir Frontal Advance Rate
about 1 ft/day

Typical Laboratory Rates

about 1500 ft/day for 1.5 core samples

Why not use reservoir rates ?

slow and time consuming capillary end effects capillary forces become significant c.f. viscous forces Buckley-Leverett (and JBN) invalidated
Page 77

Flow Parameters
End Effect Capillary Number Flood Capillary Number

Nc end

k o vL
Ncend 2.3 0.07 0.02 0.02 0

Nc =
Rate (ml/h) 4 120 360 400 Reservoir

Nc

Rate (ml/h) 4 120 360 400 Reservoir

1.2 x10-7 103.6 x 10-6 101.1 x 10-5 101.2 x 10-5 10-7

For reservoir-appropriate data Nclab ~ Ncreservoir If Ncend > 0.1 kro and krw decrease as Ncend increases
Page 78

Relative Permeabilities are Rate-Dependent

Bump Flood
1.0 0.9 0.8 0.7 0.6 0.5 High Rate krw ??? 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7 0.8 0.9 1.0 Low Rate krw' Bump Flood krw'

Page 79

Relative Permeability (-)

Flow Rate Considerations

Imbibition (waterflood of water-wet rock)
Sro function of Soi: Sro is rate dependent oil production essentially complete at BT krw suppressed by Pcend and rate dependent bump flood does not produce much oil but removes Pcend and krw increases significantly high rates acceptable but only if rock is homogeneous at pore level

Considerations
ensure Swi is representative low rate floods for Sro: bump for krw steady-state tests
Page 80

Flow Rate Considerations

Drainage (Waterflood of Oil-Wet Rock)
end effects present at low rate Sro, krw dependent on capillary/viscous force ratio high rate: significant production after BT reduced recovery at BT compared with water-wet

Considerations
high rate floods (minimum Dp = 50 psid) to minimise end effects steady-state tests with ISSM low rates with ISSM and simulation

Page 81

Flow Rate Considerations

Neutral/Intermediate
Sro and kro & krw are rate dependent bump flood produces oil from throughout sample, not just from ends ISSM necessary to distinguish between end effects and sweep

Recommendations
data acquired at representative rates (e.g. near wellbore, grid block rates)

Page 82

JBN Validity
High Viscosity Ratio
viscous fingering invalidates 1D flow assumption

Low Rate
end effects invalidate JBN

Most USS tests viewed with caution

if Ncend significant if Nc not representative if JBN method used

Use coreflood simulation

Page 83

Test Recommendations
Wettability Conditioning
flood rate selected on basis of wettability Amott and USBM tests required Wettability pre-study
reservoir wettability? fresh-state, cleaned-state, restored-state wettabilities

beware fresh-state tests (often waste of time) reservoir condition tests most representative
but expensive and difficult
Page 84

Wettability Restoration
Hot soxhlet does not make cores water wet! Restored-state cores too oil wet Lose 10% OIIP potential recovery
USBM
0.0 1.0 STRONGLY WATER-WET

STRONGLY OIL-WET -1.0 -1.0 0.0

Original SCAL plugs Hot Sox Cleaned Flush Cleaned

1.0

Amott

Page 85

Key Steps in Test Design

Establishing Swi
must be representative use capillary desaturation if at all possible
remember many labs cant do this correctly

fresh-state Swirr is fixed

Viscosity Ratio
matched viscosity ratio for end-points investigate viscosity dependency for rel perms normalise then denormalise to matched end-points
Page 86

Key Steps In Test Design

Flood Rate
depends on wettability determine rate-appropriate end-points steady-state or Corey exponents for rel perm curves

Saturation Determination
conventional
grain loss, flow processes unknown

NISM
can reveal heterogeneity, end effects, etc
Page 87

Use of NISM
Examples from North Sea Core Laboratories SMAX System
low rate waterflood followed by bump flood X-ray scanning along length of core end-points some plugs scanned during waterflood

Fresh-State Tests
core drilled with oil-based mud

Page 88

X-Ray Scanner
Coreholder (invisible to Xrays)
X-rays detected X-rays emitted

Scanning Bed

X-ray Emitter (Detector Behind)

Page 89

X-ray adsorption

0 %

Sw(NaI)

100%

NISM Flood Scans

SMAX Example 1
uniform Swirr oil-wet(?) end effect bump flood removes end effect some oil removed from body of plug neutral-slightly oil-wet

Page 90

NISM Flood Scans

SMAX Example 2
short sample end effect extends through entire sample length significant oil produced from body of core on bump flood moderate-strongly oil-wet data wholly unreliable due to pre-dominant end effect. Need coreflood simulation
Page 91

NISM Flood Scans

SMAX Example 3
scanned during flood minimal end effect stable flood front until BT
vertical profile

bump flood produces oil from body of core neutral wet data reliable
Page 92

NISM Flood Scans

SMAX Example 4
Sample 175 (fresh-state) scanned during waterflood unstable flood front
oil wetting effects

oil-wet end effect bump produces incremental oil from body of core but does not remove end effect neutral to oil-wet
Page 93

data unreliable

NISM Flood Scans

SMAX Example 5
Sample 175 re-run after cleaning increase in Swirr compared to fresh-state test no/minimal end effects moderate-strongly waterwet

Page 94

NISM Flood Scans

SMAX Example 6
heterogeneous coarse sand variation in Swirr Sro variation parallels Swirr end effect masked by heterogeneity (?) very low recovery at low rate (thiefzones in plug?) bump flood produces significant oil from body of core neutral-wet

Page 95

Key Steps in Test Design

Relative Permeability Interpretation
key Buckley-Leverett assumptions invalidated by most short corefloods

Interpretation Model must allow for:

capillarity viscous instability wettability

Simulation required
e.g. SENDRA, SCORES
Page 96

Simulation Data Input

Flood data (continuous)
injection rates and volumes production rates differential pressure

Fluid properties
viscosity, IFT, density

Imbibition Pc curve (option) ISSM or NISM Scans (option) Beware several non-unique solutions possible
Page 97

History Matching
Pressure and production
1.66 cc/min
800 Differential Pressure (kPa) 700 600 500 400 300 200 100 0 0,1 1,0 10,0 100,0 Time (min) 1000,0 Measured differential pressure Simulated differential pressure Measured oil production Simulated oil production 1,0 2,0 4,0 Oil Production (cc) 6,0

5,0

3,0

0,0 10000,0

Page 98

History Matching
Saturation profiles
0.8 0.7

Water Saturation

0.6 0.5 0.4 0.3 0.2 0.0 0.2 0.4 0.6 0.8 1.0

Page 99

Normalized Core Length

Simulation Example JBN Curves

Relative Permeabilty Curves Pre-Simulation
1 0.9 0.8

Relative Permeability

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Krw Kro low rate end point high rate end point

Water saturation
Page 100

Simulation Example Simulated Curves

Relative Permeabilty Curves Post Simulation
1 0.9 0.8

Relative Permeability

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Krw Kro low rate end point high rate end point Krw Simulation Kro Simulation

Water saturation
Page 101

Quality Control
Most abused measurement in core analysis Wide and unacceptable laboratory variation Quality Control essential
test design detailed test specifications and milestones contractor supervision modify test programme if required

Benefits
better data more cost effective
Page 102

Water-Oil Relative Permeability Refining

Key Steps
curve shapes Sro determination and refinement refine krw determine Corey exponents refine measured curves normalise and average

Uses Corey approach

rock curves may not obey Corey behaviour
Page 103

Curve Shapes
1 0.9 0.8 0.7 0.6

Semi-log Good data concave down

Kro Krw

Kr

0.5 0.4 0.3 0.2 0.1 0

Water-Oil Rel. Perms.

1
0 0.2 0.4 Sw 0.6 0.8 1

0.1

Kr

Cartesian Good data convex upwards

0.01

Kro Krw

0.001

0.0001 0 0.2 0.4 Sw 0.6 0.8 1

Page 104

Sro Determination
Compute Son high, medium and low Sro
1

low rate, bump, centrifuge Sro

0.1 Sor = 0.40 Sor = 0.20 Sor = 0.35
Kro

Plot Son vs kro (log-log)

0.01

Sro too low

0.001

curves down Sro too high curves up Sro just right straight line
Page 105
1.000 0.100
Son = (1-Sw-Sor)/(1-Swi-Sor)

0.0001 0.010

Refine krw
Refined krw

Use refined Sro Plot krw versus Swn Fit line to last few points
Krw 0.1 1

least affected by end effects Determine refined krw

0.01 0.1 Swn = 1-Son 1

Page 106

Determine Best Fit Coreys

Use refined Sro and krw Determine instantaneous Coreys
No' & Nw'

3.5 3 2.5 2 1.5 1 0.5 0

log(krw' ) log(krw) Nw* = log(1.0) log(S wn )

No Nw

log(kro ) No* = log(Son )

Plot vs Sw Take No and Nw from flat sections Least influenced by end effects

0.2

0.4 Sw

0.6

0.8

Page 107

Refine Measured Data

Endpoints Refined krw and Sro Corey Exponents No and Nw (stable) Corey Curves
Relative Permeability 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 0.1 0.2 0.3 0.4 0.5 Sw 0.6 0.7 0.8 0.9 1 Refined Kro Refined Krw Original Kro Original Krw

kro( refined ) = Son

No

krw( refined ) = krw' Swn

Page 108

Nw

Normalisation Equations
Water-Oil Data

Sw Swi Swn = 1 Swi Srow

k ro n =

k ro k ro end

krw krwn = krwend

Gas - Oil Data

Sgn = Sg Sgc 1Swi SrogSgc

k ro n =

k ro k ro end

krgn =

krg krgend

Page 109

Example - kro Normalisation

1 0.9 0.8

Oil Relative Permeability (-)

0.7 0.6 0.5 Swirr Swn = 0 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7 0.8 0.9 1 Sw = 1-Sro Swn = 1
Sample 1 Sample 2

Page 110

Example - krw Normalisation

1 0.9 0.8

Water Relative Permeability (-)

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 Water Saturation (-) 0.6 0.7 0.8 0.9 1 krw at Sro krwn = 1
Sample 1 Sample 2

Page 111

Normalise and Compare Data - kron

1.0 0.9 Normalised Oil Relative Permeability (-) 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalised Water Saturation (-) Steady State Different Rock Types ? Different Wettabilities?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Page 112

Normalise and Compare Data - krwn

1.0 0.9 Normalised Water Relative Permeability (-) 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalised Water Saturation (-)
1 2 3 4 5 6 7 8 9 11 12 13 14 15

Page 113

Denormalisation
Group data by zone, HU, lithology etc Determine Swir (e.g. logs, saturation-height model) Determine ultimate Sro
e.g. from centrifuge core tests

Determine krw at ultimate Sro

e.g. from centrifuge core tests

Denormalise to these end-points Truncate denormalised curves at ROS

depends on location in reservoir
Page 114

Denormalisation Equations
Water Oil

S w dn = S wn (1 S wi S ro ) + S wi k rodn = k ro end .k ron k rwdn = k rw end .k rwn

Gas-Oil S = S (1 S S S ) + S g dn gn wi rog gc gc

Denormalised Endpoints Water-Oil Swi kro (@Swi) krw (@1-Srow) From correlations & average data

k rodn = koend .k ron k rgdn = k rg end .k rgn

Page 115

Summary Getting the Best Rel Perms

Ensure samples are representative of poro-perm distribution Ensure Swir representative (e.g. porous plate, centrifuge) Ensure representative wettability (restored-state?) Use ISSM (at least for a few tests) Ensure matched viscosity ratio Low rate then bump flood Centrifuge ultimate Sro and maximum krw
Tail ok kro curve if gravity drainage significant

Use coreflood simulation or Coreys for intermediate kr

Page 116