MISCIBLE FLOODING

Fundamental and Principles

EOR-7 by Sarma 1

Types:

i) Hydrocarbon Miscible Floods

- High pressure (or Vaporizing) gas injection

- Enriched (or Condensing) gas injection
- LPG slug injection followed by dry gas

ii) CO2 miscible flood

iii) Miscible liquid flood

EOR-7 by Sarma 2

EOR-7 by Sarma 1
 Field Applications and Criteria:

Depth : >5000ft (i.e. high pressure is favored)

Viscosity : <10cp
Oil Gravity : >25oAPI
Gas Cap : Absent

Over 150 commercial miscible flooding projects worldwide

- 55% hydrocarbon solvent
- 42% CO2
- 3% use nitrogen.

About 62% oil production from HC miscible floods, and

35% from CO2 miscible floods
20% projects contribute 65% of the total incremental oil production
over 80% of HC gas injection projects are located in Canada
most CO2 flood projects are in the USA (West Texas, New Mexico)

EOR-7 by Sarma 3

Golden Spike Alberta Carbonate Imperial 288 1964 HC 8.0 Gravity St

Prudhoe Bay is the largest successful project

A lack of dolomitization in Golden Spike resulted in a really extensive impermeable barriers to the vertical flow of
solvent, causing the vertical sweep of solvent to be poor. Although restarting the solvent flood below the
barriers was considered, it was decided instead to terminate solvent injection and convert to a gas-cycling
project before completing the miscible flood. The reported EOR of 3% OOIP for Golden Spike corresponds to
the oil recovery at the time of conversion. It is worth noting that the gas-cycling project has been successful and
is expected to give a total ultimate recovery of EOR-7
65% OOIP.
by Sarma 4

EOR-7 by Sarma 2
 Assuming miscible flood start at the end of waterflood:

Primary goal of a miscible flood is to recover this

discontinuous oil

Miscible solvent will also displace some oil from regions

not swept by water, particularly where water slumping
has occurred.

Two major factors affecting the performance of a

miscible flood:

- oil displacement efficiency at the pore level

- sweep efficiency on the field scale.

EOR-7 by Sarma 5

Oil Displacement Sweep Efficiency

Efficiency on Pore Scale on Field Scale
Solvent
Oil Swelling Extraction

Oil
Solvent
Water
Rock

Direct Miscible Flow Viscous Channelling Gravity

Displacement Fingering Override

Factors affecting miscible Recovery

EOR-7 by Sarma 6

EOR-7 by Sarma 3
 Sweep efficiency in field scale is Oil displacement at pore scale
affected by: involves:

viscous fingering Direct miscible displacement of oil

channeling
gravity override, particularly when kv is
high (less severe in stratified reservoirs)
 Sweep efficiency can be increased by
- reducing well spacing,
- increasing injection rate,
- reconfiguring well patterns,
- increasing solvent bank sizes,
- modifying WAG ratio
through high permeability paths
Oil swelling due to solvent-solubility
swelling increases oil volume in the pore
space resulting in a favorable oil relative
permeability situation
Extraction that takes place as solvent
contacts by-passed oil
Above mechanisms, together may
account for 20-30% of total miscible
flood recovery.

Also, oil displacement process is

affected by the solvent composition (in
swelling and extraction) and pressure
(direct miscible displacement).
EOR-7 by Sarma 7

What is Miscibility?

When two fluids become miscible in one another,

the IFT and Pc between them vanish.
Theoretically, this should result in 100% recovery of
the oil as it should result infinitely large capillary
number, Nc.

Key Factors Affecting Miscibility:

 composition and characteristics of the fluids

 pressure
 temperature

EOR-7 by Sarma 8

EOR-7 by Sarma 4
 Typical Ternary Diagram Miscibility occurs when compositions of the oil and
gas occupy opposite sections with respect to the
A. light (or volatile) components tangent at the critical point.
(methane, N2, CO2)

B. intermediate components, Oil composition in region C is miscible with gas in

(C2-C6 which play a major role in region A provided the line joining A and C does not
thermodynamic equilibrium, CO2 and H2S) intersect the bubble point or dew point curves of the
C. heavier C7 + components. phase envelope.

Oil in region D is miscible with gas in region B. This

is an example of condensing gas drive.

Oil in region D and gas in region A are not miscible;

whereas, oil and gas in regions B and C are
mutually miscible in all proportions.

2-phase region envelope becomes smaller as P is

increased at a constant T.

When T is increased at a constant P, 2-phase

region expands.

EOR-7 by Sarma 9

High Pressure Gas Injection:

This process is also called a vaporizing gas drive.

The C2-C6 components are vaporized and the oil
and gas achieve miscibility through multiple
contact.

Applied to high pressure formations (i.e. deeper

formations) with the oil rich in intermediate
components.

The desired API gravity of the oil should exceed

35oAPI

Usual injection rate varies between 3000psi to

6000psi.

Gas at G is not readily miscible with the oil at O as the line GO cuts through the two-
phase envelope. As the phase exchanges continue to take place between them, the
gas becomes richer in the heavier components while the oil becomes leaner of them.
In simple words, intermediates C2-C6 are stripped from the oil by the injected gas.
Eventually, the injected gas assumes a composition (shown as gt in figure) which is on
the right of 2-phase envelope. The point Op denotes the limiting oil composition at
which the oil can no longer exchange its components with the gas. The oil at Op
becomes non recoverable.
EOR-7 by Sarma 10

EOR-7 by Sarma 5
 The miscibility is attained within the reservoir away from the injection
wellbore . This distance (few feet to over 100 ft) is a function of the injection
pressure and oil composition. Once the miscibility is achieved, the miscible
front is usually driven by a chase inert gas to cut down the use of the more
expensive natural gas. The inert gas typically used is N2.
Miscibility develops at the leading edge
EOR-7 of transition zone.
by Sarma 11

Enriched Gas Drive:

Also known as the condensing gas drive process, uses

10-20% PV of natural gas enriched with intermediate
components so that the gas can transfer its C2-C6
components to the oil and become miscible with it.

Enriched gas drive process is suitable for reservoirs

with heavier components and under moderately high
pressure and temperature. The injection pressures are
between 2000 to 3000psi.

Initial compositions of gas and oil are at G and O, and are not miscible as GO cuts
through the two-phase envelope. Typically, a slug of methane enriched with ethane,
propane or butane (10-20%PV) is injected followed by slugs of chase fluids. Like in the
vaporizing gas drive process, the miscibility is achieved at Ot due to the progressive
exchange of the intermediate components from the gas to the oil. During this process,
the gas looses its heavier components and becomes drier. At Ot, all the oil becomes
mobile. In this process the miscible zone forms closer to the wellbore by continued
injection of fresh enriched gas. Thus, progressively, the oil around the wellbore becomes
richer in intermediates.
EOR-7 by Sarma 12

EOR-7 by Sarma 6
 Like in vaporizing gas drive, once the miscibility is achieved, the miscible
front is driven by less expensive leaner gas or by water.

The gas injection must be continuously maintained until there is a sufficient

reserve of the rich gas behind the front. This precautionary measure is
required as the miscibility achieved in this process is not as stable as in the
vaporizing gas drive. Miscibility develops at the trailing edge.

Enriched gas is expensive and the requirement of continuous injection

makes it process even more expensive.
EOR-7 by Sarma 13

LPG Slug Injection ( Also known as miscible slug process

About 0.3-0.5PV of slug (e.g., LPG) is injected into the reservoir.

LPG and oil miscibility occurs along LO line.

Miscible bank with reservoir oil is driven by either water or a dry chase gas.
When a chase gas is used, it remains miscible with LPG provided Pres is
higher than the critical pressure of the LPG-gas mixture and LG does not
intersect 2-phase envelope.

Although LPG offers the benefits of lower injection pressure and

applicability in more reservoirs , it is more expensive than other two types of
gas processes.
EOR-7 by Sarma 14

EOR-7 by Sarma 7
 Vaporizing and Condensing Gas Drives
In vaporizing gas drive, miscibility develops at transition-zone
front. In condensing drive, the miscibility develops at the rear.
The relative location of miscible zone is dictated by the
equilibrium ratio, which converges to unity at miscibility.

In vaporizing gas drive, equilibrium ratio converges to 1 at the

leading edge of the transition zone where the critical fluid is
present. In condensing gas drive, at the trailing edge.

In reality, gas-oil displacements exhibit both vaporizing and

condensing gas drive characteristics. In such scenario, the
transition zone front condenses the intermediates while the
trailing edge vaporizes them.

Conventional ternary diagram are too simple to explain

displacement mechanisms. Hence, pseudo ternary diagrams
are often used. However, doing so overlooks the importance
of the distribution of individual group.

Also, the composition path, i.e., the transition zone that links
the injection gas composition and reservoir oil composition
depends on the overall composition which cannot be
EOR-7 by Sarma 15
described by three pseudo components.

Mixing Parameter () for Viscosity

o  s
 Empirical power law m 
equation for “blended”  Ss So 1/ 4 
viscosity: S  S  1/ 4
  s 
S s  So
o
 s o 

 Effective oil and solvent  oe   o1  m

viscosities:
 se   1s   m
 In lab,  = 0.66
In field,  = 0.33
Use above as initial
boundaries for you
guesses simulation
history matches.
EOR-7 by Sarma 16

EOR-7 by Sarma 8
 Impact of Mixing Parameter on Oil Recovery

EOR-7 by Sarma 17

MINIMUM MISCIBILITY PRESSURE (MMP)

MMP is among the most important parameters we need a priori in a gas miscible
process.

What is MMP?
The lowest pressure at which the gas can develop miscibility with the reservoir
crude at the reservoir temperature.

No unanimity of definitions:

(i) MMP is the pressure that causes a minimum 85% Original Oil-in-
Tube (OOIT) recovery at CO2 (gas) breakthrough with 95%-98%
OOIT recovery at 1.2 PV of CO2 injection. (Mungan, 1981)
(ii) MMP is the pressure at which almost total OOIT recovery becomes
possible after 1.2 PV of CO2 injection. (Yellig and Metcalfe, 1980)
(iii) MMP is the pressure that causes 80% OOIT recovery at CO2
breakthrough and 94% recovery at a GOR of 40,000 SCF/STB. (Holm
and Josendal, 1974)
(iv) MMP is the pressure that causes 90% OOIT recovery after 1.2 PV of
CO2 injection. (Williams et al. 1980)
EOR-7 by Sarma 18

EOR-7 by Sarma 9
 Factors affecting CO2 MMP:

 density parity between CO2 gas and the oil

 reservoir temperature (a higher temperature results in a higher

MMP)

 crude composition (the more of C5-C30 results in lower MMP)

 molecular weight distribution of C5-C30 hydrocarbons (low

molecular weight promotes lower MMP)

 type of hydrocarbon (aromatics provide higher MMP)

 Our recent study at U of Adelaide suggests that the MMP is also

impacted by the presence of nC5. It lowers MMP.

EOR-7 by Sarma 19

Effect of Aromatics on MMP

Hagedorn and Orr (1993) suggest that both the molecular size and structure
affect the way the components partition in a CO2-Crude Oil System, and
hence, the MMP.

Multi-ring aromatic compounds appear to partition much less favorably into

the CO2-rich phase than other components with similar molecular size.

The oil containing a significant quantity of multi-ring aromatics have

correspondingly higher MMP. Therefore, presence of significant quantities
of multi-ring aromatics in a crude oil is detrimental forwards the
development of the miscibility.

Hagedorn and Orr studied three different crude oils, including one from
Kubiki oil field. Of the three crude oil samples, the Kubiki oil had the
highest percentages of multi-ring aromatics (9.03% 2-ring and 8.98% 3+ring
aromatics) and low percentage of n-alkane (5.55%).

EOR-7 by Sarma 20

EOR-7 by Sarma 10
 Chemical Type Distributions for Mean, Oil W, and Kubiki
(Hagedorn and Orr, 1993)
Type Means (%) Oil W (%) Kubiki (%)
n-alkane 30.27 5.01 5.55
branched-alkane 23.26 34.19 19.67
cycloalkane 24.70 27.02 32.77
1-ring aromatics 12.37 11.52 3.28
2-ring aromatics 2.71 6.83 9.03
3+-ring aromatics 0.95 0.11 8.98
tetralin 1.33 2.79 4.46
decalin 2.37 11.79 12.11
indene 0.00 0.00 0.56
biphenyl 0.39 0.45 1.26
sulfur 1.33 0.00 0.00
NSO 0.00 0.00 1.33
unknown 0.33 0.29 1.00
Temp (oF) 105 170 125
MMP(psia) 1200
EOR-7 by Sarma
3100 3950 21

Effect of nC5+ in Gas Stream

 MMPCO2 = 2810psia
 MMPGas2 = 2880psia
 Correlations predicted MMPGas2 > 3400psia
 This led to concept that C5+ components may
have greater effect than correlations predict
 Decided to further investigate this effect

EOR-7 by Sarma 22

EOR-7 by Sarma 11
 Reservoir Fluid Compositions
Fly Lake (Oil A) Oil B Oil C Alternate Fly Lake*
Com ponent Mol % Mol % Mol % Mol %
Hydrogen Sulphide H2S 0.00 0.00 0.00 0.00
Carbon Dioxide CO2 15.57 3.98 0.65 16.80
Nitrogen N2 0.73 0.13 0.27 1.18
Methane C1 23.95 13.40 11.88 30.85
Ethane C2 8.71 6.61 1.06 9.34
Propane C3 8.63 8.57 0.58 8.20
Iso-Butane iC4 2.30 2.77 0.19 2.15
N-Butane nC4 4.16 4.83 0.15 4.48
Iso-Pentane iC5 1.46 1.92 0.31 2.20
N-Pentane nC5 1.63 1.92 0.20 2.02
Hexanes C6 3.70 7.35 3.96 2.90
Heptanes C7 + 29.16 48.52 80.75 19.88
TOTAL 100.00 100.00 100.00 100.00

Stream Properties
Molecular Weight 70.03 96.7 144.1 --
API Gravity 50.7 51.8 44.1 --
GOR (scf/stb) 1273 455 105 --
Shrinkage (rb/stb) 2.00 1.21 1.03 --

Heptanes Plus Properties

Mol % 29.16 48.52 80.75 19.88
Molecular Weight 154.1 156.1 173.9 --

*Alternate Fly Lake - composition used in some of the paramteric senstivity studies

EOR-7 by Sarma 23

Gas Compositions: Gas#1-Gas#4

Gas#1 Gas#2 Gas#3 Gas#4

Component Mol % Mol % Mol % Mol %
Hydrogen Sulphide H2S 0.00 0.00 0.00 0.00
Carbon Dioxide CO2 91.75 90.84 89.08 87.38
Nitrogen N2 0.20 0.20 0.19 0.19
Methane C1 8.05 7.97 7.82 7.67
N-Pentane nC5 0.00 0.99 2.91 4.76
TOTAL 100.00 100.00 100.00 100.00

Mol Weight 41.73 42.03 42.61 43.17

Gas Gravity 1.446 1.457 1.478 1.498
Pc (psia) 1037.3 1031.8 1021.3 1011.1
Tc (R) 530.8 533.9 540.0 545.8

EOR-7 by Sarma 24

EOR-7 by Sarma 12
 Results – Gas #2

Gas #2 Temperature (°F)

Correlation 176 212 279
Alston et al. 2608 2859 3433
Sebastian et al. 3067 3298 3795
Yuan et al. 2281 2869 4243
Dong 3132 3435 3952
Kovarik 3344 3549 3899
Emera and Sarma 2693 2951 3544
RBA 2526 2651 2775
Slim Tube -- -- 2880

EOR-7 by Sarma 25

nC5 effect on MMP of CO2-rich gases

MMP Gas with nC5 / MMP Base Gas
1.2

1.1
MMP Gas with nC5 / MMP Base Gas

0.9

0.8

0.7

0.6

0.5
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
nC5 Content in Inj. Gas

MMPnC5
MMPbase

 0.0087  M nC5 
2
 
 0.0876  M nC5  1.0
Where:
MMPnC5 = Minimum Miscibility pressure of the gas stream containing nC5
MMPbase = Minimum Miscibility pressure of the base gas stream (0% nC5)
MnC5 = Mol percent of nC5 in the injection gas

EOR-7 by Sarma 26

EOR-7 by Sarma 13
 MMP Estimation in Laboratory

MMP is to be estimated in the laboratory for the candidate oil.

Only in the absence of laboratory data, it is to be estimated from

theoretical computation or by using an appropriate correlation
available in the literature. Use of such theoretical or correlation is
not preferable as they may give inaccurate MMP

Laboratory Methods:

i) PVT and MW data correlations from literature

ii) Slim tube test
iii) Corefloods
iv) Rising Bubble Apparatus (RBA)
v) Determination of vanishing IFT
vi) P-x diagram based on experimental data (very time
consuming and yet may not be accurate !)

EOR-7 by Sarma 27

Gas (say, CO2)-oil displacement is made to

occur in a slim tube placed in a constant
Slim Tube Test
temperature air bath in a coiled manner.

Tube ID can be as low as 0.25” and 30-130’

long. Small ID minimizes/suppresses viscous
fingering by encouraging mixing by transverse
dispersion. Even if viscous fingering occurs, it
is expected to be fully suppressed within the
long length of the tube. In the selection of the
internal diameter of the tube, one must take
into consideration the critical finger width likely
to be expected for the particular crude-oil-gas
system. The tube diameter so selected
should be smaller than the critical wavelength.
Gardner and Ympa (1982) eqn. to estimate
the critical wavelength. (λc):

where:
 o +  CO K t μo, μCO2 = Oil and CO2 viscosities (mPa.s)
 c = 25/2  2
Kt = transverse dispersion coefficient
 o -  CO v 2 v
(cm2/s)
= average interstitial velocity (cm/s)
EOR-7 by Sarma 28

EOR-7 by Sarma 14
 In gravity-stable displacement between CO2 and oil, where density
difference is small, it is important to have a very low displacement
velocity. Still, if the flow velocity is impracticably low, the length of the
tube is increased and tube ID is further decreased such that the
transverse dispersion is sufficiently large. Pozzi and Blackwell (1963)
criterion for the dimensionless transverse dispersion group:

 Do 
 F + 0.0157v  H d p  L
  > 10
2
vw
where:
Do = Effective molecular diffusion coefficient (cm2/s)
F = Formation factor
φ = porosity
σH = In homogeneity factor the porous medium used.(usually glass-
beads or unconsolidated sand are used in packing of the slim tube.
Correlation of Perkins and Johnson (1965) can be used as an estimate)
dp = particle diameter in the pack (cm)
L = Length of the tube
w = internal diameter of the tube
EOR-7 by Sarma 29

Experimental procedures

Packed slim-tube is saturated with at the reservoir temperature.

CO2 is then injected on a continuous basis into slim tube at a fixed pressure.
As the tube is considered a high permeable conduit, the pressure gradient
within the tube is neglected.

Miscibility is assumed to have been obtained when a certain percentage of

original oil-in-tube (OOIT) has been recovered following a known volume of
CO2 injection, and the visual observations through the capillary sight glass
tube at the outlet end shows no fluid interface.

If not, the experiment is repeated, progressively, at higher pressures. When

the MMP is attained the oil recovery curve tends to flatten out.

Shortcomings of the slim tube test:

- Does not use porous medium.
-As no water is present in the tube, its effect on MMP is not ascertained.

For such, coreflood tests are also recommended to estimate the MMP.
EOR-7 by Sarma 30

EOR-7 by Sarma 15
 EOR-7 by Sarma 31

Rising Bubble Apparatus (RBA) In RBA, the pressure at which a rising gas
bubble vanishes in a column of oil is termed as the MMP. The test is short (1-2 days)
and inexpensive but could be subjective. There has been much debate on the merits
of RBA method particularly because the bubble-rise behaviours are different in
condensing and vaporizing gas-drives. Exact bubble disappearance may be hard to
capture or detect!

EOR-7 by Sarma 32

EOR-7 by Sarma 16
 Method – RBA
 Miscibility criteria:
pressure at which rising
bubble dissolves in oil
 As miscibility is
approached, bubble
shape changes due to
change in oil-gas IFT
 Spherical  Elliptical  Bullet
Shaped  Elliptical top,
skirted bottom  Dissolves
(when miscibility achieved)

EOR-7 by Sarma 33

Our data compared to Holm & Josendal, 1982

EOR-7 by Sarma 34

EOR-7 by Sarma 17
 ARC’s PRIme Apparatus

EOR-7 by Sarma 35

ARC’s IFT MEASUREMENT CELL

TRANSDUCER
METERING
VALVE
THERMOCOUPLE
UPPER BPR
BRINE
RUBY LIVE
HOLDER OIL

LOWER BPR
BRINE
DRAIN BRINE WATER

OIL DROP

WATER
WORKSTATION TO
PROCESS VIDEO DATA

TEMPERATURE CONTROLLED OVEN PUMP

EOR-7 by Sarma 36

EOR-7 by Sarma 18
 ARC’s PRIme obtains MMP by measuring IFT as a function of pressure.
Although the experimental set up is quite elaborate, the technique itself is
quite simple in terms of principles. The pressure at which IFT becomes
zero is the MMP. To obtain MMP using PRIme one needs to generate a set
of IFT vs Pressure data and then extrapolate if to IFT = 0 pressure.

EOR-7 by Sarma 37

P-x diagram method needs a number of careful lab experiments using a

PVT cell. Increasing amount of solvents are mixed with the oil, and BPP,
DPP and gas liberation curves are obtained for each mixed ratio. When the
BPP and DPP curves no longer rises with the increasing amounts of solvent
added, it is assumed that FCM has been achieved. As obvious, these tests
needs accurate monitoring and are time consuning and expensive.

EOR-7 by Sarma 38

EOR-7 by Sarma 19
 Which method do we choose to estimate MMP?

The answer depends on:

I) Objective in terms of needs

(complex, quick & simple)

ii) Resources in-house (experimental and computer time)

iii) Time and budget

For the initial quick, low-cost and simple case, operators use correlations
currently available in the literature. For screening purposes, they
could give a reasonable “first guess” depending on data used.

Lab estimation of MMP is more reliable

and preferred to Correlations

EOR-7 by Sarma 39

Flow Regimes in Miscible Process

Flow regimes in a miscible process are linked to viscous to

gravity ratio (Rv/g):

 o L
R v/g 
k og h
where:
v = superficial velocity
L = distance between wells
h = reservoir thickness
ko = permeability to oil
μo = oil viscosity
g = acceleration due to gravity
ρ = density difference between solvent and oil
EOR-7 by Sarma 40

EOR-7 by Sarma 20
 Regime 1:
Rv/g is very low
A single gravity tongue
Overrides the oil.
Geometry & vertical sweep depends on Rv/g

Regime 2:
Rv/g is higher.
Still a single solvent tongue, but sweep is a
function of Rv/g only after a certain critical
value.

Regime 3:
Rv/g is even higher
Such that critical velocity has been exceeded
and sweep out increases rapidly as Rv/g
increases

Regime 4:
Very high Rv/g
Viscous fingering occurs,
Rv/g no long influence, the sweep.
EOR-7 by Sarma 41

EOR-7 by Sarma 42

EOR-7 by Sarma 21
 Stalkup has coupled Rv/g with the sweep efficiency a breakthrough

2 0 5 0 v ( B / D  ft 2 )  o ( c P ) L ( ft )
R v /g 
k ( m D )   ( g / c m 3 ) h ( ft )

v o L
R v /g  2050
k h

where the following restrictions apply:

 if k H  kV , th e n k  kV k H

 d a t a is f o r li n e a r f l o w
i
 f o r a 5  s p o t , v  1 .2 5
hL
i
 fo r a lin e d r iv e , v 
hL

where i = injection rate (barrels/day per well)

EOR-7 by Sarma 43

Assignment 7.1 on Flow Regime 10 marks

Which one of the following two 5-spot projects is likely to have problems
of viscous fingering? Hint: Use the Stalkup equation

Project A (5-spot) Project B (5-spot)

Spacing 40 acres 40 acres
L 933 ft 933 ft
i-CH4 1500 b/d 800 b/d
Μo 0.5 cp 1.5 cp
kH 70 mD 15 mD
H 30 ft 25 ft
ρ 0.4 g/cm3 0.15 g/cm3
M 25 25
kV = kH = kH

EOR-7 by Sarma 44

EOR-7 by Sarma 22
 Types of Floods (Based on Injection Mode)
Horizontal Pattern Flooding Vertical or Downward (Crestal) Flood
(Injectors/Producers in Well Pattern) (High-Dip or Reef-type Structures)

1. Gravity segregation is a concern 1. Solvent injection in up-dip and oil

because it causes gravity
override/channeling resulting in poor
vertical sweep. A higher density
difference between the injected solvent
and oil may further enhance these
problems. This is particularly more so
in thick formations with high vertical
permeability.
2. VGR and mobility ratio are also
important factors. Viscous fingering
can be severe and may lead to a
failure.
3. Flooding is usually done in WAG
mode.
EOR-7 by Sarma
production from down-dip zone.
2. Gravity stabilized process at low
solvent flood rate -- less viscous
fingering and channeling. Results in a
high volumetric sweep efficiency.
3. In reef-type applications with an
active aquifer, the solvent rate is
carefully monitored and controlled to
avoid coning.
4. High vertical permeability with no
barriers like shale lenses are preferred
so that solvent can move downward
without restriction.
5. Reservoir formation should be
45
thick.

SK-15

CO-1

Horizontal Injection Mode CO-6

CO-2
CO-5

CO-3

 Continuous injection of CO2 driven by chase water: SK-66

CO-4

Residual CO2 is left behind the water displacement front

rather than in oil

Water CO2 Oil

 WACO2 …Later CO2 slugs being progressively smaller

Continuous Water CO2 Oil

 CO2 followed by carbonated water and continuos water

Continuos Carbonated CO2 Oil

Water Water
 WACO2 and then repeated WAG to miscibly displace CO2
behind the water flood front

Flue Gas Water Flue Gas Water

CO2by Sarma
EOR-7 CO2 Oil 46

EOR-7 by Sarma 23
 Vertical/Downward Crestal Injection

CO2 Chase Gas

CO2

Oil Oil

A B

EOR-7 by Sarma 47

Diffusion and Dispersion Concepts in Gas Floods

 Gas conc may be manifest at prod

well even before breakthrough

 Transition from step-change will

Down stream conc

gradually result in a S-shaped

concentration profile with time, due to
Inlet end

mixing, caused by dispersion at

higher rates.
Upstream conc

 At lower rates, molecular diffusion

takes place. Diffusion is a result of
random molecular motions.
Time  Fick’s law, which you have studied in
your chemistry classes, is related to
diffusion and dispersion.

EOR-7 by Sarma 48

EOR-7 by Sarma 24
 Perkins & Johnston Treatment of Diffusion and Dispersion in Vertical
Miscible Flood
[Perkins & Johnston, SPE Journal (1963), Vol. 3, p. 77]

 Miscible process involves molucular

1  x 
diffusion (D) and dispersion (K)
 General equation for solvent C 1  erf  
concentration, C at time, t at a distance 2   2 D t 
measured from original position of  o 
interface. Do is the molecular diffusion
coefficient with dimension of L2/t
 Sign depends on initial conditions. If 
erf   
2  2
step changes occcur:
e d
 @t=0, C=1 for x<0, C=0 for x>0. The
sign is –ve.  0
 @t=0, C=0 for x <0, C=1 for x>0, sign is
+ve.
 Error function, erf() values can be
obtained from tables.
 This is an equation of a straight line
plot on an arithmetric probability paper.
 Generally, C is assumed to lie between
10% and 90% at the transition zone
between two fluids.
EOR-7 by Sarma 49

The width of diffusion zone in an indication of the amount of

mixing between displaced and displacing fluids in a miscible 1  x 
process. Hence, it dictates the injection slug size that needs to 0 .1  1  erf  10 
2  
be designed.   2 Dot 
Equations on RHS can be modified as per the concentration i.e.,
variation; i.e., if C were to change from 1 to 99%, the constant
1  x 
3.625 will change.  0 .8  1  erf  10 
2  
These equations are porosity =100%!   2 Do t 
They also suggest that the width is proportional to t. Noting i.e.,
that x=vt, we can also say that the width of the zone is x10
proportional to the square root of the mean distance travelled.  0.90625
2 Dot
The relative width .(say, = x10-x90) decreases as x is increased;
i.e., mixing zone is small relative to well spacing and hence, we i.e.,
can go for smaller slug size. Note that we are talking about an
x10  1.8125 D o t
ideal situation here and hence only get the minimum zone width.
1 
In reality, it may much larger for various reasons.  x
0 .9  1  erf  90 
For example, let: V =0.0481 ft/hr, Do=0.000346ft2/hr
2  2 D t 
 o 
X = 4ft, 100ft and 400 ft then using eqn. For (x90-x10) we get:
x90  1.8125 D o t

X (ft) X90-x10 (ft) Relative width =(x90-x10)/x*100  x90  x10  3.625 D o t

4.0 0.61 15% i.e.,
100.0 3.07 3.1% 2
1x  x 
Do   90 10 
400.0 6.15 1.5% t  3.625 
EOR-7 by Sarma 50
(source: EOR by Willhite & Green)

EOR-7 by Sarma 25
 Arithmetic Probability Paper

Distance from original interface

Concentration
EOR-7 by Sarma 51

Applications of Perkins & Johnston to porous media

 Recognize:

 Porosity is <100%

 Distance travelled in porous medium longer than, usually assumed as ~2L

 If D is apparent diffusion coefficient,

D/Do = 1/(FΦ) where F is the formation electrical resistivity factor

Factor [1/(FΦ)] usually lies between 0.6-0.7 but for some porous medium it could be
much lower

 Diffusion is a dominant dispersion mechanism when the flow rates are very low.

 Additional mixing takes place when fluids flow through porous medium in
longitudinal and transverse directions.

 Hence … previous equations cannot be use as they are for flow in porous
media.

EOR-7 by Sarma 52

EOR-7 by Sarma 26
 Diffusion and Dispersion as function of flow rates
(or Peclet Number)

Log Kl/Do

Convective
Dispersion
Diffusion dominance
dominance

0.2 2.0
Peclet Number, u dp/Do

EOR-7 by Sarma 53

Obtaining Dispersion Coefficients

 Not so easy from theory as porous media
conditions vary.
 Often K are estimated from experimental studies
or from correlations.
 Kl =f(porous medium structure, particle size, fluid
and front velocities)
 Kt occurs transversely; i.e., perpendicular to flow
direction, with lesser convective effects. It is also
function of similar parameters as Kl
 Molecular diffusion effects exist but its influence
varies depending on the conditions.

EOR-7 by Sarma 54

EOR-7 by Sarma 27
 Longitudinal Dispersion Calculation
 Based equations on RHS and using following
definitions: 1  ( x  ut ) 
L= porous medium length C 1  erf  
T = time to inject or produce 1 PV fluid = L/U 2  2 Kt 
Vp = volume in 1PV
Vi = fluid volume inejected
t = time to inject 1 PV
1  UL 
1  erf  
u = Darcy velocity
U =(Vp-Vi) / Vi C
2  2 KtVt p 
and noting the following boundary and initial
conditions
  
at t=0, x<0, C=0
at t>0, x=0, C=1
(V p  Vi )
at t>0, x , C=0 U
Kl is obtained as: Vi
L
x1  x  ut  x  t
2
1  L(U10  U 90 ) 
T
Kl    If
V pt  3.625  x  L, x1  L  ut
AL
t
uA
EOR-7 by Sarma 55

Taylor’s Formulations
 For capillary tube of radius, a:
Kl = Do + U2 a2/ (48 Do)

 For unconsolidated sand, we need to account for longitudinal convective

dispersion coefficient, E and particle diameter, dp which is assumed
around 0.36 if no data are available. Convective term can be neglected if
(udp/Do) <2.0. Here, u is the interstitial velocity which the Darcy velocity
divided by porosity.

Kl  D  E
or
Kl D E 1 ud p
    1.75
Do Do D0 F Do
EOR-7 by Sarma 56

EOR-7 by Sarma 28
 Perkins-Johnston Empirical Relationships
Kl  D  E
 Account for inhomogeneity or
factor of the pack as a
function of dp. Kl D E 1 Fi ud p
    0.5
Do Do D0 F Do
 Proposed two relationships:
one for Kl and the other for
for :
Kt. Fi ud p
 50
Do
 Notice that at low rates
diffusion dominates while at Kt D E 1 Fi ud p
high rates, convection     0.0157
dominates. In between, Do Do D0 F Do
both contribute to dispersion for :
and unfortunately most of
our applications are in that Fi ud p
“in betweeb” zone!  10 4
Do
EOR-7 by Sarma 57

EOR-7 by Sarma 58

EOR-7 by Sarma 29
 Practice Example from EOR by Willhite and Green on Dispersion & Diffusion

 Given the following data for a Berea sandstone:

Core diameter = 0.165 ft
L = 4.01 ft
Injected water of salinity 30000ppm (30 g NaCl/L) at a flow rate = 2.12x10-4 ft3/hr
Interstitial velocity = 1.155 ft/day
Experimental data obtained from a linear miscible displacement test are as presented as below:

1. Develop and plot a concentration profile as a function of time, t (Hint: erf usage)
at 76.7, 78.4, 81.3, 83.4, 88.0, 92.6 and 95.4 hours.
2. Estimate Kl from experimental data and compare it with Kl obtained using
Perkins-Johnston’s empirical correlation, if D for the brine system =3.87x10-5
ft2/hr and [1/(F)] = 0.6
EOR-7 by Sarma 59

Solution step 1
Estimate the argument value in erf and then read off erf
value from table to estimate C at the exit point; i.e., x =L

EOR-7 by Sarma 60

EOR-7 by Sarma 30
 EOR-7 by Sarma 61

Step 2: plot experimental data on probability paper, draw best straight

line fit and pick 10% and 90% points and use the equation for Kl

EOR-7 by Sarma 62

EOR-7 by Sarma 31
 Perkins and Johnston Correlations for Kl and Kt can be used
within the validity range for these correlations.

Key input data are:

F1 vs dp correlation.
For more homogenous ones,
following data are available. There
are not many data available for
reservoir rocks! For sandstone
formation, F1dp = 0.36 in is often
used as a good approximation.

Molecular diffusion and

formation resistivity index.
Diffusion coefficients could cause
up to 10% error in the calculation.
Overall, the correlations yields
values within  20%.
EOR-7 by Sarma 63

Comparison of Exptl. and Correlation Results for Kl

Parameters/Results Experimental Method Perkins-Johnston Correlation

U10, ft 3/2 0.0105

U90, ft 3/2 -0.0099

n/a
Vp, ft3 0.01766

Time, t, to inject 1PV,

83.4
hrs.
Kl, ft2/hr 3.46 x 10 -4 3.14 x 10 -4
-9.2% (reasonable and acceptable, given the
Error from exptl.
complexities/time in experimental data)

EOR-7 by Sarma 64

EOR-7 by Sarma 32
 Perkins-Johnston Empirical Correlations for Kt

EOR-7 by Sarma 65

EOR-7 by Sarma 33