I-

.._

Tek!n~caPetrolleum Se'Nioo·s ltd: ..

s.u~te.2:5001, S3iQ ~ ,B·thAvenue S.. W. Cal!gl,ary', A~berta.

T2P 358,

June 2001

Enhanced Oil Recovery Course Contents

Chapter I Enhanced recovery methods and screening guildelines 1. Introduction 2. Enhanced Recovery methods 3. Residual Oil Saturation 4. Screening Guildelines

II

1. 2.

Chemical EOR Processes Introduction Mobility Control Processes 2.1 Polymer flooding 2.2 Foam flooding 3. Low-1FT Processes 3.1 Surfactant Flooding 3.2 Alkaline Floods Miscible EOR Processes 1. Introduction 2. Background 3. Phase Behavior -> P-T and P-X Diagrams 4. Terminology 4.1 First Contact Miscibility 4.2 Multiple Contact Misciblity 5. Misciblity Mechanism 5.1 Gas Condensing Mechanism 5.2 Gas Vaporizing mechanism 5.3 Combination of Condensing I Vaproizing Phase Behavior 6. Misciblity Design - Hydrocarbon and C02 6.1 Correlations 6.2 Equation of State 6.3 Laboratory Techniques 6.4 Miscibility Criteria 6.5 Miscibility Equivalence 7. Factors Affecting Displacement Behavior 7.1 Mobility Ratio 7.2 Fluid Dispersion 7.3. Flow Regimes 7.4 Viscous Fingering 7.5 Sweeping Efficiency

III

8. Improvement Sweep Efficiency 8.1 Gravity Stabilized Floods 8.2 Water Injection to Improve Mobility Ratio 8.3 Foam 9. Immiscible Flooding 10. Project Development Considerations 10.1 Simulations 10.2 Economic 10.3 Implementation and Monitoring IV 1. 2. 3. 4. Thermal methods and Application Introduction Hot Water floods Steam floods Combustion

Appendix I Glossary II Details of EOR Projects

Chapter I Enhanced recovery methods and screening guildelines

1. Introduction 1.1 Primary Oil Recovery 1.2 Secondary Oil Recovery 1.3 Efficiency of Conventional Recovery Methods Enhanced Recovery methods 2.1 EaR Principles 2.2 EaR Oil Recovery Factor Residual Oil Saturation EaR Screening Guildelines

2.

3. 4.

1.

Introduction

Oil is found in underground rock formations called reservoirs. The oil resides together with water, and sometimes gas, in very small holes (pore spaces) and fractures. The size, shape, and degree of interconnection of the pores vary considerably from place to place in an individual reservoir. Thus, the anatomy of a reservoir is complex, both micro- and macroscopically. A complete, detailed quantitative description of a reservoir is never possible. The detailed data which can be obtained from wells represents only an infinitesimal fraction of the reservoir volume, even under the best of circumstances. Properties of crude oil and formation water in different parts of an individual reservoir generally vary only slightly, although there are notable exceptions. For different reservoirs, crude properties cover a wide spectrum of differences. Some are thinner than water, while others are thicker than cold molasses. Generally, crude oils are lighter than water. The water in different reservoirs also varies, in salinity and other mineral content. Because of the existence of a wide range of properties of both rock and fluids, reservoirs act differently and must be treated individually.

1.1

Primary Oil Recovery

After discovery, most oil reservoirs undergo a period of production called "primary oil recovery". Primary oil recovery used natural reservoir energy to drive the oil through the complex pore network to producing wells. The driving energy may be derived from one or more of the following: gas that evolved from solution out of the oil; expansion of free gas; influx of natural water; or gravity force. The primary recovery efficiency is generally low, varied from 5 to 20 % OOIP, when gas is the drive agent; much higher recoveries are associated with water drives. Reservoirs with effective gravity drainage can also have higher recoveries. Eventually, the natural drive energy is dissipated. When this occurs, energy must be added to the reservoir to produce any additional oil.

1.2

Secondary Oil Recovery

Secondary oil recovery involved the introduction of energy into a reservoir by injecting gas or water under pressure where the gas is immiscible with oil. Separate wells are usually used for injection and production. The technique was called "pressure

maintenance". The added energy stimulates the movement of oil, providing additional recovery at increased rates. Today, limited use is being made of gas injection because of its low oil displacement effectiveness and the need for gas supplies in the market. When gravity drainage is effective, pressure maintenance by gas injection is effective. Waterflooding is the principal secondary recovery method and accounts for about half the current U.S. daily oil production.

1.3

Efficiency of Conventional Recovery Methods

Ultimately, conventional primary and secondary recovery are expected to produce about 20 to 40 % of the original oil discovered. The recovery from individual reservoirs can range from as low as 5 percent to as high as 80 percent of original oil in place (OOIP). This broad range of recovery efficiency is a result of the variations in the properties of the specific rock and fluids involved from reservoir to reservoir, as well as the kind and level of energy that drives the oil to producing wells, where it is captured.

2.

Enhanced Recovery methods

The term "enhanced oil recovery" refers to any method used to recover more oil from a reservoir than would not be obtained by primary recovery. Since the early 1950's, a significant amount of laboratory research and field testing has been undertaken, and some of the resulting findings have been developed on a commercial scale.

2.1

EOR Principles

The intent of enhanced recovery methods is to • improve sweep efficiency by reducing the mobility ratio between injected and inplace fluids, • eliminate or reduce the capillary and interfacial forces and thus improve displacement efficiency, and • act on both phenomena simultaneously. The basic principles of the most promising EOR methods used are given in the following Table 1. Table 1. Methods of Enhanced Recovery Methods Used Chemical 1. Mobility Control Processes methods Polymer-augmented waterflooding; Foam-COraugmented waterflooding; immiscible CO2 displacement. 2. Low-1FTProcess surfactant flooding; alkaline flooding; Miscible fluid displacement using Miscible CO2 nitrogen, alcohol, LPG or rich methods gas, dry gas. Thermal

Basic Principle Improvement of sweep efficiency; improvement of displacement efficiency.

Improvement of displacement efficiency.

Cyclic steam injection; steam drive; Improvement of both sweep

methods

in situ combustion.

efficiency and displacement efficiency.

Other processes, such as bacterial activity, electrical heating of the reservoir, and so on have been proposed, but their potential for adding to proved oil reserves must be demonstrated. Figures 1 to 5 show schematic different EOR processes being conducted. A briefly introduction for each process is given as follows. 1. Chemical methods of enhanced oil recovery are characterized by the addition of chemicals to water in order to reduce the mobility of displacing agent and/or to lowering the interfacial tension. Polymer flooding, using polyacrylamides or polysaccarides, is conceptually simple and inexpensive, and its commercial use is increasing despite the fact that it raises potential production by only small increments. Surfactant flooding is complex, requiring detailed laboratory testing to support field project design. It is also expensive and is used in few large scale projects. Alkaline flooding has been used only in those reservoirs containing specific types of high-acid-number crude oils. Miscible methods have their greatest potential for enhanced recovery of lowviscosity oils. These processes are mainly in reducing the 1FT to improve displacement efficiency. Among these methods, hydrocarbon gas and CO2 miscible flooding on a large scale is expected to make the greatest contribution to miscible enhanced recovery in the future. Thermal methods are for oil gravity less than 25 degree. These processes provide a driving force and add energy (heat) to the reservoir to reduce oil viscosity and/or vaporize the oil. This makes the oil more mobile, so that it can be more effectively driven to producing wells. Steam injection has been commercially applied in California since the early 1960's and is the most advanced of all EOR methods in terms of field experience. The performance of steam injection can be estimated with less uncertainty than other methods. In situ combustion is normally applied to reservoirs containing low gravity oil, but has been field tested under a wide variety of reservoir conditions. Only a few projects have proven economical enough to advance to a commercial scale. To date, in situ combustion has been most effective for the recovery of viscous oils in moderately thick reservoirs (National Petroleum Council, 1984).

2.

3.

2.2

EOR Oil Recovery Factor

The oil reserves obtained as a result of EOR methods in addition to the primary or conventional reserves may be expressed as the percentage of original oil in place (OOIP). To estimate how much EOR methods can add to oil reserves, the recovery potential of the reservoir has to be known. This is defined by the reservoir's characteristics and prior recovery mechanism.

For instance, the ultimate oil recovery factor of individual reservoirs under primary and/or conventional recovery methods may range from 5 percent of OOIP for the poorest reservoir characteristics for viscous oil, to as high as 55 or 60 percent of OOIP for the best reservoir characteristics or for light oil. To achieve this desideratum the oil reservoirs are classified by several models according to the average of the ultimate oil recovery Rfinalexpressed as a percentage , of OOIP, possibly attained by the respective recovery mechanism, as follows: Rfinal 5-10% Tight oil reservoirs, slightly fractured or heavy oil reservoirs 10-25% Oil reservoirs produced mainly by solution gas drive 25-40% Oil reservoirs producing under partial water drive, gas injection, or gravity drainage 40-55% Oil reservoirs produced by conventional waterflood A possible estimation of the additional oil reserves, percentage of OOIP, that could be recovered as the result of EOR processes is shown in Table 2. These values can be used for a quick evaluation of the improved oil reserves and for the selection of methods that should be applied. For instance, for oil reservoirs where ER 25-40 percent of OOIP obtained by primary or conventional methods, the implementation of the EOR methods would increase these values to 35-55 percent of OOIP. However, the selection of a certain process aimed at enhancing oil recovery must be made only after detailed investigations of the pertinent oil reservoir data (especially residual oil saturation), laboratory tests, and field pilots.

3.

Residual Oil Saturation

Oil and water do not mix. If these two fluids are poured into a vessel and allowed to settle, two distinct liquid phases are apparent, separated by a sharp interface. The oil and water are said to be immiscible. Similarly, oil and natural gas also are immiscible, although natural gas has some solubility in oil. This solubility depends on pressure level, and as soon as the solubility limit is exceeded, two phases form - a gas phase and a liquid phase, separated by an interface. When two immiscible phases flow simultaneously in a porous medium, as for example in the displacement of oil by water, some of the oil being trapped in the pores as shown in Figure 6. This trapped oil is called "oil droplet" which is retained in the reservoir by the interfacial tension and viscous forces. It is this most elusive oil which is the target for "enhanced oil recovery" projects. The flow behavior in a given unit volume of rock is determined by the relative permeability characteristics of the rock. Oil relative permeability decreases with decreasing oil saturation, but oil permeability reduces to zero and oil stops flowing while the oil saturation is finite. This limiting saturation is called the residual oil saturation or, perhaps more precisely, the relative permeability endpoint residual oil saturation to distinguish it from the material balance or average residual oil saturation left in a reservoir after waterflooding. Similarly, water permeability decreases to zero at a finite

value of water saturation called the irreducible water saturation. Oil and water both flow at saturation between the two endpoints, and the shape of the relative permeability curves between the endpoints determines how much fluid must be injected to reduce the oil saturation in the swept rock to a given level. In addition to being affected by rock pore structure, residual oil saturation and relative permeability are affected by rock wettability and by interfacial tension (1FT) at the oil/water interface. Craig defined wettability as the tendency of one fluid to spread on or to adhere to a solid surface in the presence of other immiscible fluids. It can have a pronounced effect on immiscible displacement. For example, relative permeabilities for oil-wet rocks are more unfavorable for waterflooding than are relative permeabilities for water-wet rocks. Other factors being equal, this causes earlier water breakthrough in the oil-wet rock and requires that more water be injected to achieve a given reduction in oil saturation than in water-wet rocks. The effect of oil/water 1FTon residual oil saturation is illustrated in Figure 7. Here the residual oil saturation is plotted vs. capillary number, the product of Darcy velocity and oil viscosity divided by 1FT. Capillary number is an approximate measure of the ratio of viscous to capillary forces. Over ranges of velocity, oil viscosity, and oil/water 1FT found in conventional waterflooding, residual oil saturation is insensitive to capillary number. Figure 7 shows that a drastic reduction in oil/water 1FT,by several orders of magnitude or more, is required to achieve a significant reduction in waterflood residual oil saturation. Although the discussion of this section centered on waterflooding, similar behavior is found when a gas immiscibly displaces oil.

4.

EOR Screening Guildelines

Reference 1 provides the most recent literature review on EOR screening criteria. Table 3 was copied from reference 2 for a quick checking usage. Later in the course note, a more detail guildelines will be discussed for each process. In summary, factors used in the screening are as following. a. reservoir conditions - temperature and pressure b. reservoir fluid properties - oil viscosity and density and formation water salinity c. reservoir geology - rock type and depth and permeability and porosity References 1. Taber, J. J., et ai, "EOR Screening Criteria Revisited," SPE/DOE 35385, presented SPE/DOE Tenth Symposium on Improved Oil Recovery, 1996. 2. Dafter, Ray, "Scraping the Barrel, The Worldwide Potential for EOR" Financial Times management Reports, 1980.

Table 2 Conventional
ER BY PRIMARY AND CONVENTIONAL RECOVERY ADDmONAL OIL, % OF OOIP TO FINAL RECOVERY EOR ME1HODS IN SIT1J COMBUSTION

and EOR Method Effciency

(ER)

5-10% ADDmONAL

10-25% ADDmONAL

25-40% ADDmONAL OIL, % OOIP Ear",., +15% 40%-55% +10% 35%-50% +15% 40%-55% +12% 37%-52% +15% 40%-55%

40-55% ADDmONAL OIL, %OOIP

ERfiuJ

on,

on,
R

%OOIP Ear",., +35% 40%-45% +25% 30%-35%

%OOIP E_

+30% 40%-55% +20% 30%-45% +25% 35%-50%

+10% 50%-65%

STEAM INJECTION

POLYMER INJECTION DRY ANDRICH GAS ALCOHOL, LPG

+8% 48-63%

SOLVENt INJECTION

+25% 35%-50%

SURFACfANT

FLOODING IMMISCIBLE DISPLACEMENT MISCIBLE . DISPLACEMENT

INFILL DRll.LING ALTERNATING WATER-GAS INJECTION

+30% 40%-55% +20% 30%-45%

+15% 40%-55% +10% 35%-50% +15% 40%-55% +2-4%

+10% 50%-65%

CO2 INJECTION

+10% 50%-65%

+7%

+5%

+5%

IMPROVED CONVENTIONAL ME1HODS

WATERFLOODING AFTER GAS INJECTION WATER INJECTION IN SECONDARY GAS CAP PRESSURE PULSING +3-5%

+5% +3% +5%

+5%

GAS INJECTION AND GRAVITATIONAL DRAINAGE (ATIIC OIL) CROSS-FLOODING

'0 as W s: 15 ~ oco~a."O~ ~ I: a.~ Il')'-..,

I

oi:? ::0

0
_Q)

~
~
~

..........

o
~

0.

Nc)"c

-g .~CO0 .Q z EQ) ~ Q) .Q

CI).~{g.s

..

~~

ZO

.... :: 0 :E!
0

-0 CO;;:::

~ .~ .... "0 o0

C)

&l .E

c
Q)

Cti

.~
~

-0 0.._ .- a. "0 0

a. ....
Q)U;

.2
I

8.
I:-C
Q) Q)

en

_ .Q) I:

CO

s:

=0 i

=
o o o
A

===
NOll')

000 000

c;t;

~Lri"N
VI AlAI

N V

~ o
N V

VI

"0

W'
-I

Q;:!

s
A

8 ,...
VI

ld ,...
V

a.

ld ,...
VI

0.

8
~
VI

o ,...
c\1

I
en
Cl

el!?· o C'Il
It) (\J

0 -0 C'Il w I!?
"0
Q) (\J Q)

oct

~C:Q.

0_ -

... c: is 02

Cl"O

Q..

g>

~ °iii zE:;:: I!?

-iU
:::>
c:

:0);

I
C'Il

00

Q)

>-J2 c: o Q)
0-

c: 0;::

Cl ::l •

-§,"E

J:

tr

tlj
~

> a::

>

II~
I
Q)

'0

u;

c: c:

~.
Q)

.9
c:

c:

Q)

c: c:

C'Il C/)

C'Il C/)

_=
:t::Ln _ V

c?1

020

C'Ilo

_=
020
_
C'Ilo
:!:::It)

ZoO

t;co 0-:;

t;co V
0_
::l 00

b
o (\J

~~~~~

OlOlOlOlOl ~~,....,.... ._. ..._. ._. ........ ._.

~~gfG~
.......... (\J(\J(\J

...

b
o (\J
V

g
(\J

VVVVV

- c: &,
.....

... o
A

o~_
c:Q)

-:=2
..... U
A &Cij(.)

(;"0

'0

E ~
A

'0

E o (\J
A

:;::::E 0;:: 0
(.)(\J

co (.)"0
Z

'0 .0
(\J

'0

o (\J
A

(5 AI

AI

8 .....
~

c..

It) (\J

Bv

g
VI

c..

~ c:c:c:
v888
..... (\Jet)

COOO "","COOl c..OlOlOl (,),....,....,....

--VI VI VI

8
(\J

c..

.....

.....

~I

(\J

cis

> ..... et) -J • •

QC'IlC'll

Q..

I
'CI

'0 as ..... ~~ o
Q) '2 "0 .- 0 Q) as 0 zE;;:::: ~ Q) .c
"0

'0 asO

,go:

('t)

O:::CQ. 0'-10 0)"0 CI) C Q)

Q)

Cl

en O
~
J: I-

j ~I
:c

Q)

e :::J
C

~E
00
..t::. ,...

E8

0.

=>

C_

C as Q)..t::.

>a:: w >

Z <C
W W W W

!i~
I
~

11
:::Ju o as
>LL

-Q)

>00

CI) :::J
I

~ Q) oe ~~ s:

~ 0) fii .....

o :i:

Q) CI) ~ ~

.~

.....

-..t::.

....
0

....
0

f}JQ)- C °co··CI)cE:C ..... ; ..... ~et5oJ!!" :::J CI) 0 C :::J._ 0'~ CI) o 010 i5 0.5 o :i: ~ ZC
0.

"O..t::.

:==0 ~ ..... o as Q)..t::.)(··:::Joc Q):=- Q) o .2l- Q)1O :i: ~ CI)_- Q) 1: c !B .2 C • Q) := := o

0)

C.- >. 0 as J!! ~~"

co

10

.....

J:

I-

o
...J

C/)

Q) C ~ ,g ..... C CI) o 0 "0 C as as

.:=.:=
00 00 00
/I. C\lLt)

.:=
0 0 0 10 V

v
.:=.:=
/I.

0 00 00 1010

o
Q.
Q.

:::i
<C
W

i
~I

.co S il) S il)

0

0 0

..... ,_ ,... ,_ ,... ..._....._....._...._.. ........

C\I .....OC\lLt) ,...,...C\IC\IC\I

..... oCOCJ)CJ) c CJ)CJ)CJ)CJ)CJ)

a;
:;::::;

C\I

ooo~o

·c

VI VI VI VI VI

" " "0 z

0 ,...
/I.
0

J: I-

a::

10
C\I

VI

I
:c as
c,

Q)

E Q)

...J"O L1. E 0:0

"0

co

!;;(
W

WC\I

·c
o Z
0

:;::::

C\I

/I.

AI

3:

:c as
0 .,... as
I

I
~I

....

C\I

" 8
V
CO
I

0.

0.

8
C\I

VI

> 10 0 a: 10
C

0.

a;

8 C\I
AI
('t)
I

0.

o ·c c "0 z
:;:::: 10

... Q)

a..

as

0 ,...

::E
W,

<C,...

....

as
I

.... as
C/)

as

'<:t

cb

cb

J
C)

~§
=E
C
-(1).0 ::l

o~

oS 0 Q) c

0C:
::l

c:

o J:

en

u:

gj

>a: w

tu :!:

>

~,
1.0
Q)

,~

c: .9
CIS

en

c:

;I:

o o

;I:

o ~

1.0

"

"

o~ 1.0, No VlO

o ~

V ...

~
V

i
0CIS

(ij 02 0C:
(,')

"5 c.. ~

... " 0
"C

a.

Q)

° ~ ° §
V

I
~ ~

~
-~C")I.O CIS

:c CIS
co

Q)

I-

,,,
CIS CIS

, "
CIS

cb

C» CIS

~,
CIS

REFERENCES FOR TABLE 1

a-1

N.S. Erofeeve, M.L. Surguchev and E.M.Halimov, "Enhanced Oil Recovery is an important way t increase oil reserves'" The Future Supply of Nature-Made Petroleum & Gas; Pergamon, 1977.

a-2

Harold EGilliiand, Continental Oil Company, Ponca City, Oklahoma; Energy Communications 4(1) 83-106 (1978).

a-3

National Petroleum Council, -Enhanced Oil Recovery", Dec. 1976.

Note:

figures refer to "screentnq" criteria for selection of EOR candidates, page 24. a-4 S.M.Farouq Ali and Robert F.Meldau, "Current Steamflood Field Experience", Society of Petroleum Engineers paper SPE 7546, 1978. a-5 Lewin & Assiciates, "Research and Development in Enhanced Oil Recovery, Part I", October 1976; page 11-8. a-6 H.L.Chang, Cities Service Oil Co., ·Polymer Flooding Technology - Yersterday, Today and Tomorrow", Journal of Petroluem Technology, Aug. 1978; pages 1113-1128. a-7 Chieh Chu, Getty Oil, ·A Study on Fireflood Field Projects", Socity of Petroleum Engineers paper SPE 5821, pages 321-337; SPE report on Symposium on Improved Oil Recovery, Oklahoma, March 1976. a-s F.H.Poettmann, ·In situ Combustion; A Current Appraisal", World Oil, part one (April 1964, pages 124-128) and part two (May 1964, pages 95-98). a-s T.M. Geffen, "Oil Production to Expect from Known Technology", Oil & Gas Journal, May 7,1973, pages 66-76. a-10 Office of Technology Assessment, U.S. Congress, "Enhanced Oil Recovery Potential in the United States", January 1978, page 36.

.!.

D U

• "
c
0

!
C

•• ~

..

-• :5
.!
u

uin.5

u•

~.
E

~ U .!.
c

-ucn

N~

'0

., o
o :z: ~
IU

D Ue

:I

~ a: o

IU

>

U
IU

~:I

.. ~'i
E
0

.. • -•
• " ~
~ E"
u
0

'i

•
~

•
u

•• iE
!

V" -• •
_g

•
0 C

.s ...
e
Q)

e;-

:I

U :I

o• U;;

• NG.

• u

c:
IlL

E~ • 0
0

.. -•
~

IU

a: o
>
IlL

o II:
o

• ~

"

..

~~----------~------------~ ii
• ~

...........

"

..

.!

:I

-" :.
c

e .. ....
Q)

.. •

~ ~
'0

~
(U

'0 C
(U

z c
i

U:I

;! o•

e;-

;t

:I
IlL

~ a:

Figure 1 C02 Miscible Flooding Process

C02 AND WATER

OIL AND WATER

..

I
• • Injected ~ ~ and water zone

D Oil and water zone

-crude oil miscible zone

Figure 2 Surfactant Flooding Process

INJECTION FLU.IDS

OIL AND WATER

..

• •
II

Drive water zone Water/polymer zone Surfactant slug zone Oil and water zone

Figure 3 Cyclic Steam Stimulation Process

~ ( Shut - in phase)

Huff ( injection phase)

STEAM

OIL AND WATER

Puff ( production phase)

(for days or weeks) •

(for days) Hot oil, water and steam zone Steam zone Oil and water zone

( for weeks or months)

Figure 4 Steam Drive Process (Steam Flood)

STEAM

OIL AND WATER

..

• •.

Hot water zone Steam zone Oil and water zone

Figure 5 In-situ Combustion Process

OIL AND WATER

AIR AND WATER

COMBUSTION GASES

J.

Injected air and water zone • .~ Air and vaporized water zone Combustion zone Steamzone Simplified cross-section of oil field structure

III

Hot water zone

Oil and water zone

Figure 6 Residual Oil - Oil Droplet

Oil

Residual

Oil

Water

V/'!

W.R. FOSTER LEFEBVRE DU PREY

[:::::1

BARDON

<,

::t 10-4

CD ~

a:

..

z >a:

::::>

10-5

«
...J ...J n,

10-6

10

20

30 -

40
%

50

60

RESIDUAL OIL SATURATION,

Capi Ilary Number versus Residual (Bardon and Longeron)

P.V.

Oi I Saturation

SPE/DOE 27816 ( What Determines Residual Oil Saturation in Three-Phase Flow?

M.J. Blunt, D.H. Fenwick, and Dengen Zhou, Stanford U.
SPE Members
COpyrIght 1994. Society 01 Petroleum Engineers. Inc.

Thts paper was prepared lor presentation at the SPEJDOE Ninth Symposium on Improved Oil Recovery held In Tulsa. OklahOma. U.S.A.. 17-20 April 1994. Thts paper was selected lor presentation by an SPE Program' Committee loIlowing review 01 Inlormation contained In an abstract submitted by the aUlhor(s). Contents 01 the paper •
• ~ted. have not been reviewed by the SocIety any position 01the SocIety 01 Petroleum Engineers. Hs 01 Petroleum Engineers. Permission to copy ISrestricted 01 where and by whOm the paper IS presented. WrHe
I.

01 Petroleum Engineers and are subject to correction by the aUlhor(s). The material. as presented. does not necassarlly reflect olfocers, or members. Papers presented at SPE meellngs are sUbJect to publlcallon review by Editorial Committees 01the SocIety to an abstract of not more than 300 words. illustrationS may not be copied. The abstrect should contain conspicuous acknowledgment Ubrarian, SPE, P.O. Box 833836, Richardson. TX 75083-3836. U.S.A.. Telex 163245 SPEUT.

I'BSTRACT We present a theoretical and experimental treatment of three lphase flow in water wet porous media from the molecular level :10the core scale. We compute the equilibrium thickness of a flat \Ioillayer between water and gas considering both van der Waals lind structural forces. We show that although most oils will spread over water, they will eventually form a film that is only a ,few molecules thick. The film thickness, which is in the range ;0.1-5 nm, is determined by the capillary pressure and by ~\vhetheror not the film was formed from thinning a thicker layer for by encroaching over a hitherto clean interface. Both spreading ~d non-spreading oils can form thin films. Drainage through ; this film would be far too slow to account for the recoveries seen ~ xperimentally. We then present results from drainage .experiments in capillary tubes with a square cross-section. The tube has water wetting the comers and contains oil ganglia '~para,ted by air bubbles along its length. The gravitational Fving force in the experiment allows oil ganglia to drain by 'pwelling the oil film. We predict not only that eventually all the .oil is recovered, but that we can calculate the drainage rate. The ,predicted rate agrees well with the measured values. Finally we :perform a series of gravity drainage experiments in sand ~olumns at high and low oil/water surface tensions. Air is ttmtroduced into the top of the column containing waterflood ~ ..residual oil. At the end of the low surface tension experiment " there is a thick oil layer (several microns across) between water and gas, stabilized by capillary forces, whereas for the high surface tension fluids, only a molecular oil film is present. In both cases oil is displaced by air, since the oil ganglia will drain undergravity by swelling the oil films. However, the low 'surface tension case gives a lower final recovery, since a significant

amount of oil remains in the column in the thick oil layers. In both experiments virtually all the oil can drain from the top of the column. The residual oil saturation in the presence of water and gas can be almost zero under favorable circumstances. INTRODUCTION The displacement of oil by gas in the presence of water is an important recovery process in oil fields and in the clean up of contaminants spilled below ground. Since the work of Dumore and Schols [1974], it has been known that gravity drainage in water wet rock can lead to a high oil recovery, with residual oil saturations of a few percent in the presence of both gas and water, which is much lower than the residual oil saturation in the presence of water alone. Further studies on sandstone cores, bead packs and sand columns [C/uztzis et al., 1988; Kantzas et al., 1988a; Yizika; 1993] confirmed these results. It was suspected that the high oil recovery was due to drainage through thin layers of oil that lie between the water and the gas in the pore space. This film drainage has been observed directly in two dimensional etched glass micromodels by Kantzas et al. [1988b], Oren et al. [1992], Oren and Pinczewski [1992], Kalaydjian [1992] and Soli et al. [1993]. However, it was not clear under what circumstances drainage through an oil layer would occur. It was hypothesized by Oren [1992], Yizika, [Vizika, 1993] and Kalaydjian [1993] that systems with a positive spreading coefficient, which means that the oil spontaneously spreads over a flat water/gas interface, would experience film drainage and high recoveries, whereas non-spreading systems would see lower recoveries. However, Dumore and Schols [1974] saw the same

t,

371

WHAT DETERMINES

RESIDUAL OIL SATURATION

IN THREE PHASE FLOW?

SPE 27816

residual oil saturation (approximately 3%) for both a spreading and a non-spreading system. whereas Oren et at. [1992] • Oren and Pinczewski [1992]. and Viz i k a [1993] found higher recoveries if the spreading coefficient was positive. It was also implied by Kantzas et al. [1988b] that the recovery could be determined by the stability of the film. which is controlled by capillary and intermolecular forces. rather than the spreading coefficient alone. In experiments in micromodels and capillary tubes. no oil films are observed for non-spreading systems. However, in core floods. oil layers stabilized by capillary forces are formed as an oil bank moves through the system. and thick oil layers can exist for both spreading and non-spreading oils. This paper investigates the fundamental mechanisms of oil recovery in three phase flow. starting at the molecular scale. The principal issues are: (1) the thickness and stability of thin oil films controlled by intermolecular forces; (2) the thickness of oil layers during drainage; (3) the flow rate in these layers; (4) the final oil recovery.

containing this oil tilm has a much lowered effective surface tension for which. if there is no pressure difference across the film. the effective spreading coefficient is zero or negative [Adamson. 1990; Gibbs, 1928; Rowlinson and Widom, 1989]. Gas (a)
t

Oil Water

Gas (b)---Water

Figure 2. (a) The free energy. F. of a film of thickness tand area A is F=A{rgo+row)+E,,(t)A+PAt. E" is the interaction energy (principally van der Waals) of the film and P is the pressure difference between gas and water. (b) When the thickness t = O. F = ggwA. This defines E,,(O) = rgw - rgo - row = Cs.

Consider the oil film illustrated in Figure 2(a). The free energy of the film of thickness t and area A is

DOES OIL SPREAD ON WATER?

(a) (b)

F=A(rgo+row)+Ev(t)A+PAt

(2)

Gas

Gas

Solid surface

Solid surface

where Ev is the interaction energy of the film per unit area and P is the imposed capillary pressure between the gas and water. For oil films much thicker than the molecular width ao, the interaction is described by van der Waals forces. When tis of the order of molecular width, Ev may have significant contributions from structural forces. For t < QQ. we imply a loose arrangement of oil molecules at the gas/water interface. By definition, when t 0 (Figure 2(b» the free energy F rgwA, where rgw is the gas/water surface tension. Hence:

Rgure 1. (a) Three fluids in contact. If rgw > rgo + row the oil wiil spontaneously spread between the gas and water. Define a spreading coefficient Cs = rgw - r~ - row· Cs> 0 indicates an oil that spreads. (b) If the oil forms a thin fiim between the gas and water. the system can be in equilibrium with a macroscopic blob of oil if < 0 where = Y'gw- rgo - row and Y'gw is an effective surface tension. representing the free energy per unit area of the oil fiim.

Ev(t = 0) = rgw - rgo - row = Cs

The equilibrium thickness of the film is given by i}F/iJt or:

(3)

= o.
(4)

c:

c:

P = O(t)

= -oEv/ot

H a large drop of oil. such that the size is greater than the range of intermolecular forces. or around 0.1 urn, is placed on water. the initial spreading of the drop depends on the spreading coefficient:

where O(t) is called the disjoining pressure. The equilibrium film thickness, teq, is when the capillary pressure equals the disjoining pressure. We may derine an effective gas/water surface tension in the presence of an oil film, ~ ,by r;w of/oA. r;w is the ~ energy per unit area of the~as/water interface separated by an oil film of thickness t.

Cs

= rgw - rgo -

row

(1)

r;w
where

= rgo + row + Ev(t) + to(t)

t = O. The effective

(5) spreading

where rgw. rgo and row are the gas/water. gas/oil and oil/water surface tensions respectively. If Cs> O. the contact line between the three phases is unstable and the oil spreads. as shown in Figure 1. Most hydrocarbons and crude oil mixtures (see p. 104. Table 5 of [Muskat. 1949]) do have a positive spreading coefficient. On a flat substrate. the oil eventually forms a thin layer of approximately molecular thickness. The water/gas interface

r;w = rgw when

coefficient

c:

is

c: =

Ev(t)+tO(t)

(6)

372

rSPE 27816 ) ;THE INTERACTION

,!.

MARTIN BLUNT, DARRYL FENWICK AND DENGEN ZHOU

ENERGY

'Using quantum field theory. and assuming t » ll(). the following expression for the disjoining pressure between two semiinfinite materials I and 3 separated by medium 2 of thickness t, has been derived [Dzyaloshinskii et al.• 1961; Lifshitz; 1955]: n(t)=~

Table 1. Table of constants used to compute the dielectric response. The data for water and tetradecane comes from Parseglan (1981) and that for pentane and octane from Hough (1980).

Water

Cr
microwave
Wj

to, (1011 radls)

1.05

n=O where:

f~~Er2UI J
1

+/2)p2dp

(7)

infrared

ultraviolet

12 = (

1
~12~32

exp(2 P~n
c

t..fE2) _1)-1
s3-P s3+P
$3

74.8 (1014 radls) 0.319 1.048 1.398 3.038 6.38 126.1 151.9 173.2 197.5 226.3 281.0

fj

gj

1.429 0.735 0.154 0.143 0.079 0.0392 0.057 0.0923 0.156 0.152 0.271

(1014 radls) 0.228 0.577 0.425 0.380 0.851 7.75 13.5 23.4 31.1 45.0 95.1

a-pentane
infrared ultraviolet 5.54 187.7 0.025 0.819

sl =~£I/E2 -1+ p2.

~12 -;: u12

s3 =~E3/£2 _1+p2.
~32=--.

o o o o o

= $1-

p. $1 + P PE1/E2

n-eetane (8)
infrared ultraviolet 5.54 186.3 0.023 0.925

~.

= $1 -

$1 + PE1/E2

-;: u32

= $3 -

PE3/E2 + PE3/E2

tetradecane
infrared ultraviolet 5.54 128 136 147 160 176 196 221 252 288 335 0.019 0.0928 0.1235 0.1466 0.1489 0.1304 0.1178 0.09978 0.08132 0.06177 0.03960 5.47 8.36 11.2 13.8 17.3 21.9 26.9 32.1 38.0

and El. E2. E3 are the dielectric response as functions of the imaginary frequency i~. where ~n =2nkT liz (n = 1.2.3...), k is Boltzmann's constant and Iz is Planck's constant. c is the speed of light in a vacuum and T is the absolute temperature. which we set to 300 K in our calculations. The summation is over integral values of n and the prime on the summation means that the first term in the sum should be multiplied by 1/2. For our case, materials 1 and 3 are water and air respectively and medium 2 is oil. The functions £1 and E2are computed from the adsorption spectra of water and oil respectively. e = 1 for air. The dielectric response functions are calculated using the Lorentz harmonic oscillator equation:

40.7

The first sum represents the Debye microwave relaxation with resonance frequency (J)r. The second sum accounts for the infrared. visible and ultraviolet adsorption spectra. where the constant gj describes the bandwidth of the adsorption frequency Wj. Table 1 lists the data used in the calculations presented here.

Equation 7 is valid in a continuum. where the interaction across the film is determined solely by the dissipation energy of an electromagnetic wave propagated through the system. When the film thickness is close to a molecular size, structural forces become important. In this regime. the disjoining pressure is modeled empirically using an equation proposed by Israelachvili [1991]: n(t)

= K cos(21rt/ao)

e-t/ao

(10)

The constant K can be determined by knowing that the interaction energy at t = 0 is equal to the spreading coefficient. Thus .
K= Cs{I+4n )

ao
373

(11)

WHAT DETERMINES

RESIDUAL OIL SATURATION

IN THREE PHASE FLOW?

SPE 27816

Equations 7 and 10 for the disjoining pressure are not additive [lsraelachvili. 1991]. Equation 7 applies when 1 » ao. while Equation lOis valid for film thicknesses of a few molecular diameters.

RESULTS FOR THIN OIL FILMS

l

r = 10--4 m, P is in the range of 500 Pa, much larger thar maximum values of 0(1) in Figure 3. Therefore, considering van der Waals forces alone, there is no stable film thickness in a porous medium for the hydrocarbons studied here. The only exception would be for oil on a flat plane of water, where an oil film tens of nanometers thick may be stable if P is extremely small (or zero).

Figure 3 shows the disjoining pressure as a function of film thickness for pentane, octane and tetradecane, using the Lifshitz Equations 7, 8 and 9 and the data in Table 1. For small film thicknesses, the disjoining pressure is negative, which means that the film is unstable and tends to spread further. At larger separations, the disjoining pressure becomes positive. If we consider oil spreading in a porous medium, with an imposed capillary pressure difference, P, between water and gas, the equilibrium film thickness is that for which 0(1) = P, which is Equation 4. Characteristic capillary pressures are of the order r / r where r is the surface tension and r is a representative pore radius. If we assume typical values of 0.05 Nm-I and

10000

: : : : :::

8000

6000

r=

a.. as
:::J
Q) "(f) (f)

IIIIIII '
::::::: : : :: ::: : : : : : : :.

i~
g>-O.2 ~ ~.4 iii -0.6

4000

2000

10000

a..
0>
1.0

Q)

1111111111~\~_..~ ..~~'~.:""' . ........

: ::: ::: : ::: ::: : : : : ::: : : : :::: ::: ~

_lh_ic_k_ne_S_s_(n_m_)

lo"-t

8000 c;;E!::.

0.8 0.6 0.4 02 0.0 -0.2 -0.4 -0.6 ·0.8 ·1.0 10 10"

.0 ................. : .....

c: c:

a..
as
~
Q)
(f) (f)

6000

If) If)

:::>

I!!

.!a

-2000

: :: :::: ::: .............. ..............

::::: ::::: ::::

c,

I!!
OJ

-4000

:::J

4000

c '2 '0 'iii'

-6000

.............. ............. ............ ............ ............ , ............. .. ............. .. .... .,. .. .. .. .. .............. .. .......... ., ........... .. .......... ... .......... .. .. .... .......... ............ · . · .

Lifshitz Equat, Structural Force

is

~; ; ;; n ., ............ ..........
__.__._....._.__._.._._,

2000

thickness

(nm)

a..

Q)

._
c:
0

0>

c::

...... ........

· . · . · . · . · · . · .:._. :.....':"':'..:..' . . -10000 L.:...:..... .:...:L..--L-L.....L-J'-'-'...I.-_--' _

1 10

'

102

thickness (nm)
n-octane n-pentane tetradecane
Figure 4. The disjoining pressure for water/n-octane/air computed using van der Waals forces (Lifshitz Equation) and structural forces. Structural effects give an oscillatory force on the film .

-2000

.!a

-4000

-6000

-8000

I I I I

-10000

L--_---'_-IJ-JL......L__.__._._._.L...L..

__

-'--_L--.J.-J......J._._

........

10

thickness (nm)
Figure 3. The disjoining pressure for water/n-alkane/air systems.

Figure 4 shows the disjoining pressure for n-octane between water and air accounting for structural forces. The structural forces were calculated from Equations 10 and II with ao = 0.4 nm and Cs = 0.2 mNm-l. Structural forces become significant at around 4 nm, or approximately 10 molecular diameters. Because of the oscillatory nature of the structural force, the disjoining pressure has both negative and positive values as the fjlthickness approaches zero. The maximum value of O(t) (at 1 : is simply K, which is approximately 2 x 107 Pa, which is mucn higher than typical capillary pressures in a porous medium.

374

SPE27816

MARTIN BLUNT, DARRYL FENWICK AND DENGEN ZHOU

Therefore, structural forces do allow a stable film, of molecular -_ /thickness. Hirasaki [1988] has shown that a stable film has P = n(l) and dfI{t)/dt <0. Thus we see that there will be several stable film thicknesses for a given P, using Equation 10. The film thickness will depend on how the film is formed. If the film is being created on a previously clean interface, the solution with the smallest value of t is chosen, since a film is formed from t = O. If the spreading coefficient is positive and \ P « K this will occur when cos(21ttlao) is small or t == ao/4 and the effective spreading coefficient, from Equation 6, is negative and given by -21tC_se-1I4. If the initial spreading coefficient is negative, the addition of oil to the gas/water interface is energetically disfavored and no film will form. If the film is formed from the thinning of a thicker layer, the solution with the largest value of 1 is chosen. This situation arises when a film is formed as the oil saturation decreases, such as at the trailing edge of an oil bank in gas flooding. In this case, a stable film may form for both positive and negative spreading coefficients. If P « K, we ignore the oscillato~ factor in Equation 10 to find a film of thickness t where Ke - Ia. == P, or t = aoln(KlP), which may be several molecular widths. The effective spreading coefficient may be positive or negative depending on the exact position of equilibrium. Moreover, increasing P may cause the equilibrium thickness to decrease abruptly, as it finds a new position of equilibrium at a thickness approximately one molecular diameter smaller than before. This Jcan be seen in Figure 4, where for P less than approximately 500 Pa, the equilibrium thickness is around 4 nm, or 10 molecular widths. The position of equilibrium then jumps to around 3.6 nm, nine molecular widths, as we increase the pressure beyond 500 Pa. At P = 1,500 Pa, the thickness decreases suddenly again. The same phenomenon has been described for wetting films of water [Hirasaki. 1988]. . Most oils spread on water and form a film, stabilized by structural forces, which may be one or more molecular diameters across. The thickness depends on the way in which the layer was formed and is a function of capillary pressure. The effective spreading coefficient, and hence the contact angle between the gas and water phases separated by the film, are also dependent on the capillary pressure and displacement history. Van der Waals forces on their own do not lead to the formation of thick films (5 nm or larger) in porous media. Systems with a positive initial spreading coefficient always form oil films. Nonspreading systems only form oil films if the film is generated from thinning a thicker layer. This explains why oil films are not observed in micromodel and capillary tube experiments when Cs < O. On the other hand, stable oil layers are possible for nonspreading oils at the trailing edge of an oil bank in sand column and core flood experiments. Since the molecular films are so thin, the flow through them is minimal and they are unlikely on their own to account for the high recoveries seen in gravity drainage experiments. Oren et al. [1992] suggested that the : equilibrium film thicknesses of the order of 100 nm observed in air/soltroUwater systems were due to van der Waals forces. On the basis of these calculations, this is unlikely to be the case. 375

A molecular film provides pressure continuity for the oil phase. This means that during gas injection, previously isolated oil ganglia become connected. Drainage can then proceed by swelling the oil layers. The flow rate through these swollen layers is much greater than through a molecular film and can account for the good recovery and drainage rates observed experimentally.
DRAINAGE RATES IN A SINGLE THROAT

Solid

Figure 5. An oil layer in a square crevice of the porous medium. teq is the equilibrium thickness of the molecular oil film, where the gas/oil interfacial curvature is zero.

Figure 5 shows the distribution of three phases in a single pore throat of square cross-section. Water is considered to be the wetting phase, oil is intermediate wet and the gas is non-wetting. The pressure differences between the phases are given by the augmented Young-Laplace Equation:
Pcgo

= Pg -Po

rgo =-+n(l) rgo

(12)

and
Pcow=Po-Pw= row+nW(tw) row (13)

where llrow and l/rgo are the total curvatures of the oiUwater and gas/oil interfaces respectively. nW(tw) represents the disjoining pressure of the water film adhered to the solid surface. A layer of oil will exist in the pore space if rgo > row. If the thickness of this layer is outside the range of intermolecular forces, it is stabilized by capillary forces, and the disjoining pressure in Equations 12 and 13 is negligible. We will ignore the disjoining pressure in the analysis below. Consider the arrangement of fluid illustrated in Figure 6. Two blobs of oil are separated by a height h. If there is a film of oil that connects the two blobs, there will be pressure continuity in the oil phase which allows the upper ganglion to drain into the

WHAT DETERMINES

RESIDUAL OIL SATURATION

IN THREE PHASE FLOW?

SPE 27816

lower one. During drainage, the oil layer will swell under the influence of the raised pressure in the oil phase until it is of the order of microns thick. and the radius of curvature, row is:

(16)

(17)

air
where:

110

= 2yow

rgpw
t

(18)

If we estimate the oil layer thickness by then:

= (.,fi - 1)( rgo -

row) •

Figure 6. Fluid configuration in a square capillary tube at the beginning of drainage. Two oil ganglia are separated by a gas bubble of height h. The upper ganglion can drain if there Is pressure continuity In the oil phase, which means that there must, at least, be a molecular oil film present Initially. .

The gas/oil radius of curvature is approximately constant, while the water curvature becomes progressively smaller with height The oil layer thickness is zero when z = h2 0 and increases with z, When the oil flows under gravity. the pressure in the oil phase will be larger than that implied by Equation 16, and this will force the film thickness to be greater than zero for all values of h2 and z, To account for flow in the oil phase and to compute the flow rate due to film drainage, we must derive a more accurate expression for the oil pressure. From Equation 14 we can write:

We assume that the water and gas are stationary and in capillary equilibrium and take the water pressure at Z 0 to be zero. Since there is very little pressure difference in the air bubble, the pressure in the upper and lower ganglia must be approximately the same and the oil can drain downwards under gravity. The capillary tube is square with cross-section 2r and inscribed radius r. Thus the total curvature of the gas/oil interface IIr go is equal to 2fr. If the curvature does not vary with height, then the curvature of the oil/water interface at hI must also be 2fr. Hence Po - Pg 2 Ygo/r. The water pressure as a function of height is:

Po -Pw

= Yow = Po +(hJ
row

+h2 +z)Pwg

(20)

Equation 20 is used to find row and then the thickness of the oil layer is: (21) where: (22) and
X=

(14) and the gas pressure is:

(15)

2yow r

(23)

We can write a Darcy like equation for the oil flow: (24)

assuming that the gas density is negligible. difference between the oil and water phases is:

The pressure

376

fSPE 27816
~\ )

MARTIN BLUNT, DARRYL FENWICK AND DENGEN ZHOU

; where Q is the flow rate and the conductance of the oil layer is , Written as ct3, which assumes that we have film drainage with a . free gas/oil interface. c is a geometrical factor that is also affected by the boundary condition at the oil/water interface and is thus influenced by both the oil and water viscosities. Substituting Equation 21 into Equation 24 yields: (25) where: m=

lower oil ganglion (h2 0) this step was omitted. Water was subsequently drained out of the opposite end to allow air to enter the tube. By controlling the amount of water that drained out, we controlled the height of the gas bubble, h. The last step was to allow some oil into the tube, above the gas. In each experiment the volume of oil V above the gas was 0.225 rnm-'. The tube was then placed vertically and we recorded the time for all the oil in the upper ganglion to drain. We repeated the experiment for various values of nz and h. We conducted experiments for high and low oil/water surface tensions. The fluid properties are shown in Table 2.

(./2" -1)

cr3

(26)
Table 2. Table of fluid properties. High surface tension system: water/iso-octane/air

We cannot solve Equation 25 analytically. However, the second term in Equation 25 is always larger than the first, and so to a good approximation, if the density difference between the oil and water phases is small, we may write: (27) which has the solution:

Susface tensions:

rgw = 72.5 mNm-I, row= r80 = 21.8 mNm-1

38.1 mNm-I,

Densities: Pw = 1000 kgm-3, Po = 690 kgm-3 Viscosities: J1.w= 10-3 kgm-1s-I, Jlo= 0.48 X 10-3 kgm-1s-1

)
wherefh andfo are the values offatz

Low surface tension system: water and oil phases formed from a mixture of water, iso-octane and iso-propanol.

(28)

Surface tensions: Densities:

rgw = 36 mNm-I, row= 4.41 mNm-I, Ygo= 20 mNm-1 Pw = 975 kgm-3, Po = 690 kgm-3

= h and z = o respectively.
(29)

Viscosities: J1.w = 2.8 X 10-3 kgm-Is-I, Jlo

= 0.49 X 10-3 kgm-Is-I

fh

= x+(Pw

-Po)gh2

+Pwgh

(30) Figure 7 shows the results for the system with a high interfacial tension (IFT). The triangles are the drainage times for different values of h when there is no initial oil bank (h2 = 0) and the squares are the drainage times for the same fluids for various h and h2 4 cm. Increasing h reduces the drainage time. The solid and dotted curves are our predicted times. The only unknown parameter in the equations is c, the conductance constant for the oil film. This was estimated to match the experimental results - only one parameter was used to match both curves. The agreement between experiment and theory is good. Figure 8 shows the results from the low IFf system in the same capillary tube. Again the agreement between experiment and theory is good. The drainage times are longer because the viscosities of the water and oil phases are higher than for the high IFf fluids.

We can find the flow rate from Equation 26:

(31) where m is given by Equation 28. Q increases with h, reaching an asymptotic value where m = Pwg = Pog when h is very large. The time td taken to drain a ganglion of volume V is: td=V Q (32)

with Q given by Equations 28 to 31.

EXPERIMENTAL

RESULTS IN CAPILLARY TUBES

We tested our prediction for the drainage time, Equation 32, directly. We used a square capillary tube of side 2r 0.15 mm. We first filled the tube with water. Then oil was introduced into the tube at one end to form the lower oil blob. For runs without a

3n

WHAT DETERMINES

RESIDUAL OIL SATURATION

IN THREE PHASE FLOW?

SPE 2781€

10.-~

--, • ...
Predicted for h2=O.Ocm predicted for h2=4.Ocm measured for h2=4.Ocm measured for h2=O.Ocm

lCD

E
E
CD

5
til

8 6 4 2

:;::

~\

m ca

'! Q

... ". i·...

•
15

o ~~~~~~~~~~~~~~~~ 5 10 o
h (cm)

20

Both fluid systems studied had a POSItIve spreading coefficient. Therefore, a film of oil is rapidly established between the gas and water. When the tube is held vertical' . film provides a pathway through which the oil ganglier. _~ drain. It does this by swelling the oil film to form a bulk layer that occupies most of the comer of the tube. !~ The drainage times for a single ganglion are of the order o~ minutes. Our calulations show that the maximum oil laye~ thickness is not a few nanometers, as predicted by th9 intermolecular force calculations, but around 15 urn (froIIJ Equation 19 for large z), which is comparable to the width oftbe tube. This means that in field situations, oil drainage may bel rapid if there is a well connected pathway of water wet crevi,' in the pore space. I We performed similar experiments with the high 1FT flui~ using a capillary tube of circular cross-section. In this case drainage times were greater than a day, or approximately I, ..' times longer than in the square tube. A bulk oil layer could no be established and the oil was forced to drain through a mue thinner film than before. Porous media contain crevices and grooves that would allow a bulk oil layer to form. Experiments in circular tubes artificially supress the natural flow in roughness and corners expected in more realistic media.

'

Figure 7. Comparison of measured and predicted drainage times in a capillary tube of square cross-section for the system with a high oillwater surface tension.

THE EQUILIBRIUM POROUS MEDIUM

ARRANGEMENT

OF FLUID IN A

10 .-

:;::
CD

c
CD

•
8 6 4 2

Predicted for h~.Ocm Measured for ~-O.Ocm

To test what might occur in reservoir rock, an analysis drainage under gravity in sand columns was performed. Figtn ----,shows a schematic of the final fluid configuration in vertica equilibrium. At the oil/water contact we define the oil/ware capillary pressure to be zero and similarly P ego 0 at the gas/oi contact. We will attempt to estimate how much oil is left in tbl transition region above the oil bank.

E
E m ca f!

.5.
Q

Transition zone Gas, water + oil gas/oil contact oil/water contac water

o ~~~~~~~~~~~~~~~~~~ 5 o 10
h (cm)

15

20

Figure 8. Comparison of measured and predicted drainage times in a capillary tube of square cross-section for the system with a low oillwater surface tension.

Figure 9. A schematic graph of the configuration of oil. water and gas in vertical eqUilibrium.

378

SPE 27816

MARTIN BLUNT, DARRYL FENWICK AND DENGEN ZHOU

At the end of drainage, water, oil and gas are all connected _--- ;n the transition region. The oil will remain in a stable layer , ,between the other two phases. The phase pressures are:
;

are only interested in bulk layers of oil, we may ignore the disjoining pressures to find: h - hob Ca-I (41)

(33) (34) (35) \ and we have allowed for a finite gas density. If the capillary pressures can be scaled by a capillary pressure function J, we may write: (36) (37) where So and Sware the oil and water saturations. The J function is a property of the porous medium. From Equations 12 and 13, J is equal to total curvature of the fluid interface. J should monotonically decrease with increases in saturation. Thus at a given depth we expect J(Sw + So) < J(Sw) that indicates:

The larger the value of a, the smaller he, which means that less oil is left in stable layers above the oil bank and the final oil recovery is higher. If a < 1, there is a stable oil layer at all heights above the oil bank, and this layer may retain a significant quantity of oil. If we consider the fluid properties listed in Table 2, we find that a 3.89 for the high 1FT system and 0.49 for the low 1FT case. Regardless of the value of a, oil ganglia in both experiments drained by swelling the oil layer in an attempt to reach capillary equilibrium. However, the final configuration of fluid implies that more oil will be left from gravity drainage of a system with a < 1, since more oil will remain in thick, bulk layers in the pore space above the oil bank. This conclusion is different from that implied by Kantzas et al. [1988b], where a stable oil layer was considered a necessary prerequisite for film drainage. Our capillary tube experiments demonstrated that this is not the case.

Pego-n(t) rgo

< Peow-nw(tw) row

(38)

SAND COLUMN EXPERIMENTS We tested our prediction that a system with a < 1 should give a lower recovery in gravity drainage than one with a > 1 by performing a series of gravity drainage experiments in unconsolidated sand columns. The columns used were cylindrical, being 1.2 m long and 2 em in diameter. The sand had a porosity of 0.278 and a permeability of 48 Darcys. We used the same fluid systems as in the capillary tube experiments. We established a residual oil saturation in water before allowing any air into the system. For the first day of production, the columns were allowed to drain freely under gravity with the oil and water collected from the bottom of the column. Starting from the second day, we pumped air onto the system periodically to remove the liquids accumulated at the bottom of the column. This exercise was continued until there was no additional production of either oil or water. . For the high oil/water 1FT we had an initial oil saturation of 0.19. The final oil saturation was 0.04 and the water saturation was 0.06. For the low 1FT run, the initial oil saturation was 0.17, with final saturations of 0.105 and 0.03 for the oil and water phases respectively. As we expect, the low IFf system gives the lower recovery. A similar set of experiments of gravity drainage in sand columns was performed by Yizlka [1993]. The final oil saturation was 0.11 in a system with a 4.2 (and Cs > 0) and was 0.23 when a 2.8 (and Cs < 0). Again lower values of a give lower recoveries, although both values of the final oil saturation are higher than ours. A more comprehensive series of experiments on unconsolidated columns was performed by Kantzas et al. [1988a], who found high recoveries similar to ours from gravity drainage. They found higher recoveries from controlled

Equation 38 is equivalent to saying that for a connected oil phase 'go> 'ow in Figure 5, and then using Equations 12 and 13 to derive the inequality. This analysis, however, is quite general and makes no explicit assumption about the microscopic fluid ~onfiguration. Substituting Equations 33 to 37 into Equation 38 gives us: z+hob"+ ( )

a<

n z+~---:--

Pw-Po

(39)

(Po -Pg)g
where we define:

row(po -Pg) rgo(pw -Po)

(40)

This is a generalization of an expression first derived by Kantzas et al. [1988b]. If both oil and water form thick layers, then their disjoining pressures will be negligible and can be ignored in the equation above. Equation 39 gives us the condition for the oil phase to remain connected above the oil bank. The parameter a is determined by the fluid properties. If a < 1 the inequality holds for all values of z - the oil is connected at all heights above the oil bank. If a > 1, there is some finite height, he where the oil phase becomes disconnected, apart from a molecular film. If we

379

10

WHAT DETERMINES

RESIDUAL OIL SATURATION

IN THREE PHASE FLOW?

SPE 27816

drainage, where they succeeded in removing the oil bank completely from the system (and some of the transition zone), than free drainage, when all the transition zone was left in the column. These experiments demonstrate that the final amount of oil left in the system depends on the presence of oil in bulk layers in the ppre space. Systems with a large value of a give the best final recoveries.
CONCLUSIONS

Most hydrocarbons spread on water to form a film a few nanometers across, stabilized by structural forces. In a porous medium, the thickness of this film depends on the capillary pressure and the displacement history. If the oil film is formed as a thicker layer of oil recedes, such as at the trailing edge of an oil bank, a stable film will be present for systems that have both negative and positive initial spreading coefficients. The oil film ensures pressure continuity in the oil phase, that allows oil ganglia to drain. The drainage is facilitated by the swelling of the oil layers in crevices in the pore space to thicknesses comparable to the radius of the pore. We predicted the drainage rate in a square capillary tube. The predictions agreed well with experimental measurements. Virtually all the oil can be recovered in gravity drainage from the top of a long water wet column. Oil remains near the bottom in an oil bank or in bulk layers in the pore space. A low oil/water surface tension ensures that these layers are always stable and results in a lower recovery than from high IFf fluids. This prediction was confirmed by experiments on sand columns.
REFERENCES

Hough, D. B., and L. R. White, The calculation of Hamaker constants from Lifshitz theory with applications to wetting· phenomena, Advances in Colloid and Interface Science, 14, 3-41, 1980. Israelachvili, J., 1ntermolecular and Surface Forces, 2nd ed., Academic Press, London, 1991. Kalaydjian, F. J.-M., Performance and analysis of three-phase capillary pressure curves for drainage and imbibition in porous media, SPE 24878, Proceedings of the 67th Ann. Tech. Conf. and Exhib. of the SPE, Society of Petroleum Engineers,. Washington D.C., 1992. Kalaydjian, F. J.-M., J.-C. Moulu, O. Vizika, and P. K.. Munkerud, Three-phase flow in water-wet porous media: determination of gas/oil relative permeabilities under various spreading conditions, SPE 26671, Proceedings of the 68th
Annual Technical Conference and Exhibition of the SPE,

Adamson, A. A., Physical Chemistry of Surfaces, 5th ed., John Wiley & Sons, Inc., New York, 1990. Chatzis, I., A. Kantzas, and F. A. L. Dullien, On the investigation of gravity-assisted inert gas injection using micromodels, long Berea Sandstone cores, and computerassisted tomography, SPE 18284, Proceedings of the 63rd Ann. Tech. Conf. and Exhib. of the SPE, Society of Petroleum Engineers, Houston, Texas, 1988. Dumore, J. M., and R. S. Schols, Drainage capillary-pressure functions and the influence of connate water, Soc. Pet. Eng. J., (November), 437-444, 1974. Dzyaloshinskii, I. E., E. M. Lifshitz, and L. P. Pitaevskii, The general theory of van der Waals forces, Advances in Physics,
10.165-209,1961.

Gibbs, J. W., The Collected Works of J. Willard Gibbs, Vol. 10f Longmans, Green, New York, 1928. Hirasaki, G. J., Wettability: fundamentals and surface forces, Proceedings of the SPEIDOE Enhanced Oil Recovery Symposium, SPE 17367, Tulsa, Oklahoma, 1988.

Society of Petroleum Engineers, Houston, Texas, 1993. Kantzas, A., I. Chatzis, and F. A. L. Dullien, Enhanced oil recovery by inert gas injection, SPE 17379, Proceedings of the SPEIDOE Symposium on Enhanced Oil Recovery, Society' of Petroleum Engineers, Tulsa, Oklahoma, 1988a. Kantzas, A., I. Chatzis, and F. A. L. Dullien, Mechanisms of capillary displacement of residual oil by gravity-assisted inert gas injection, SPE 17506, Proceedings of the SPE Rocky Mountain Regional Meeting, Society of Petroleum Engineers, Casper, Wyoming, 1988b. Lifshitz, E. M., Journal of Experimental and Theoretical Physics, 29, 94, 1955. Muskat, M., Physical Principles of Oil Production, McGraw Hill, Boston, 1949. Oren, P. E., J. Billiotte, and W. V. Pinczewski, Mobilization of waterflood residual oil by gas injection for water-wet conditions, Soc. Pet. Eng. Form. Eval., (March), 70-78, 1992. Oren, P. E., and W. V. Pinczewski, The effect of wettability and spreading coefficients on the recovery of waterflood residual oil by miscible gasflooding, SPE 24881, Proceedings of the 67th Ann. Tech. Conf. and Exhib. of the SPE, Society of Petroleum Engineers, Washington, D.C., 1992. Parsegian, V. A., and G. H. Weiss, Spectroscopic parameters for computation of van der Waals forces, Journal of Colloid and Interface Science, 81, 285-289, 1981. Rowlinson, J. S., and B. Widom, Molecular Theory of Capillarity, Vol. 8 of International Series of Monographs on Chemistry, edited by J. Halpern and J. S. Rowlinson, Clarendon Press, Oxford, 1989. Soli, W. E., M. A. Celia, and J. L. Wilson, Micromodel studies of three-fluid porous media systems: pore-scale processes relating to capillary pressure-saturation relationships, Water Resour. Res., 29(9), 2963-2974,1993. Vizika, 0., Effect of the spreading coefficient on the efficiency of oil recovery with gravity drainage, Proceedings of the
Symposium on Enhanced Oil Recovery, presented before the Division of Petroleum Chemistry, 1nc., American Chemical

Society, Denver, Colorado, 1993.

SPEIDOE 35385

EOR Screening Criteria Revisited

Joseph J. Taber, SPE, F. David Martin, SPE, and R.S. Seright, SPE, New Mexico Petroleum Recovery Research Center Introduction Oil production from enhanced oil recovery (EOR) projects continues to supply an increasing percentage of the world's oil. About three percent of the worldwide production now comes from EOR. Even though EOR production in the U.S. appeared to peak in 1992, Fig. 1 shows that the EOR percentage of the U.S. production is larger than ever; therefore, the importance of choosing the "best" recovery method becomes increasingly important to petroleum engineers. About 100 years ago, oil producers injected gas to restore pressure to their dying oil wells. I Since air was cheaper than gas, air was often injected to -increase production from the older fields. For many years, operators had the choice of air or gas, and sometimes they injected both into the same reservoir.' Naturally, there were safety and other problems with air. However, not until about 1928 did natural gas become the injectant of choice for pressure maintenance.' Water injection was legalized in Pennsylvania in 1921 (it was done secretly before that)." The choice of injectants has widened considerably since those early days, but the petroleum engineer still must choose an injection fluid and an overall process to try to recover the maximum amount of oil from the reservoir while still making a profit. Screening criteria have evolved through the years to help the petroleum engineer make these decisions.':" Some of the early work in this field was done by Geffen':" before there was much field experience with most EOR methods. Many of his criteria have stood the test of time. Perhaps the best known, and most widely used, screening criteria appeared in the 1976 and 1984 NPC reports." We will comment later on some of the predictions based on these criteria. One paper that we are "revisiting" (see title) is SPE 12069 by two of'us." Although we retain the format of some of the tables in SPE 12069, all have been revised. We are basing our criteria in this paper on the results of much more field and laboratory information that has become available. In recent years, computer technology has improved the application of screening criteria through the use of artificial intelligence techniques, but the value of these programs depends on the accuracy of the input data used.":" In this paper, we present screening criteria based on a combination of the reservoir and oil characteristics of successful projects plus our understanding of the optimum conditions needed for good oil displacement by the different EOR fluids. One goal is to provide realistic parameters that can be used in the newer computer-assisted tools for reservoir management.
Copyright 1996. Society of Petroleum Engineers. Inc. This paper was prepared for presentation at the SPEIDOE Tenth Symposium on Improved Oil Recovery held in Tulsa, OK. 21-24 April. 1996 This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been rviewed by the Society of Petroleum Engineers and ere subject to oorrection by the euthor(s). The material, as presented. does not necessarily reflect any position of the Society of Petroleum Engineers. Hs officers. or members. Papers presented at SPE meetings are subject to J)I.bIication review by EdHoriai Committees of the Society of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspiaJOUs acknowledgment of where and by whom the paper is presented. Write Librarian, SPE. P.O. Box 833836. Richardson, TX 75083-3836, U.S.A Telex, 163245 SPEUT.

Abstract Screening criteria have been proposed for all EOR methods. Data from EOR projects around the world have been examined and the optimum reservoir/oil characteristics for successful projects have been noted. The API gravity ranges of the oils of current EOR methods have been compiled and the results are presented 'graphically. The proposed screening criteria are based on both .fIeld results and oil recovery mechanisms. The current state of the art for all methods is presented briefly, and relationships between them are described. Steamflooding is still the dominant EOR method. All chemical flooding has been declining, but polymers and gels are being used successfully for sweep improvement and water shutoff. Only CO2 flooding activity has increased continuously. Screening criteria are useful for cursory examination of many candidate reservoirs before expensive reservoir descriptions and economic evaluations are done. For example, we have used our CO2 screening criteria to estimate the total quantity of CO2 that might be needed for the oil reservoirs of the world. If only depth and API gravity are considered, it appears that about 80% of the world's reservoirs could qualify for some type of CO2 injection. Since the decisions on future EOR projects are based more on economics than on screening criteria, future oil prices are important. Therefore, we examined the impact of oil prices on EOR activities by comparing the actual EOR oil production to that predicted by earlier NPC reports. Although the lower prices since 1986 have reduced the number ofEOR projects, the actual incremental production has been very close to that predicted for $20lbbl in the 1984 NPC report. Incremental oil production from CO2 flooding continues to increase, and now actually exceeds the predictions made for $20 oil in the NPC report, even though oil prices have been below that level for some time.

387

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

SPE 35385

EOR-IOR-ASR-Reservoir Management, etc. In the past few years, the term lOR (improved oil recovery), has been used instead of the traditional EOR, or the more restrictive "tertiary recovery." Most petroleum engineers understand the meaning of all of the words and phrases, but our technical communications are improved if we use the terms with their intended technical meanings. We certainly endorse the wider use ofIOR (and note that our paper is part of this conference's session on "lOR by Reservoir Management"), but we will cling to the technical mfanings of EOR and tertiary recovery. Successful enhanced recovery projects are being conducted as tertiary, secondary, and even enhanced primary operations. The terms should continue to be used with their evolved historic meanings. Tertiary should not be used as a synonym for EOR because some EOR methods to work quite well as either secondary or tertiary projects (e.g. CO2 flooding) while others, such as steam or polymer flooding, are most effective as enhanced secondary operations. To us, EOR simply means that something other than plain water or brine is being injected into the reservoir. We will use the terms "enhanced secondary" or tertiary when necessary for clarity. Others may use the phrase Advanced Secondary Recovery (ASR)'7-21 for EOR in the secondary mode. We are convinced that engineers should consider this improved (enhanced or advanced) secondary option much more often in the future. Classification of EOR Methods. Table I lists more than 20 EOR methods that experienced intensive laboratory and, in most cases, significant field testing. The methods utilize about 15 different substances (or specific mixtures) that must be purchased and injected into the reservoir, always at costs somewhat greater than for the injection of water. The economics ofEOR will be discussed more later, but experience shows that the best profits come only from those methods wherein several barrels of fluid (liquid or gas at reservoir pressure) can be injected per barrel of incremental oil produced.f-" This limits the main methods to either water (including heated, as steam, or as a dilute chemical solution) or one of the inexpensive gases. For some methods (e.g., micellar/polymer) there have been some technical successes but relatively few economic successes. These methods are included in our screening criteria because they are still being studied and applied in the field. If oil prices rise significantly, there is hope that these methods might become more profitable. We provide screening criteria for the nine methods that are either the most important, or still have some promise. These nine methods are shown in bold type in Table I, along with the number of the table for those eight methods that are examined in more detail. These "current" EOR or lOR methods include the three gas (nitrogen, hydrocarbon, CO2), three water (micellar/polymer plus alkaline/surfactant/polymer; polymer flooding; gel treatments) and the three thermal-mechanical (combustion, steam, surface mining) methods. A convenient way to show these methods is to arrange them by API gravity as shown in Fig. 2. This "at a glance" display also provides approximate API gravity ranges for the field projects now underway. The size of the type is intended to show the relative importance of each of the EOR methods in terms of

current incremental oil production. When examining the rationale for some of the screening parameters, it is instructive to consider the oil displacement mechanisms for the EOR methods. Table 2 shows that there are three main mechanisms for displacing additional oil with an injected fluid: (I) solvent extraction to achieve (or approach) miscibility, (2) interfacial tension reduction, and (3) viscosity change of either the oil or water, and/or plus additional pressure added to the injection fluid. There is overlap of the mechanisms. For example, interfacial tension is lowered as miscibility is approached in the "solvent" methods. The reservoir and injection conditions should be chosen to optimize the displacing mechanisms wherever possible (e.g., use a high enough pressure to achieve miscibility in solvent flooding, and look for shallow reservoirs to reduce wellbore heat losses in steamflooding). Note that we have added "enhanced gravity drainage" by gas inj ection to Table 2. Although not shown as a separate method in Table I, it is covered in Table 3 because it is a part of each of the three gas injection methods.

OiVReservoir Characteristics of Successful Projects

The depth, and the corresponding API gravity, of most of the EOR projects in the world are shown in Figs. 3 and 4. We have included projects for which 'data are available from a recent Oil and Gas Journal survey." We are more familiar with the U.S. projects (Fig. 3.) than those in other parts of the world (Fig. 4). In addition to the very broad distribution of the EOR projects, Fig. 3 shows the general trend, ranging from the many steam projects for the heavy oils at shallow depths in California to the / very deep projects for the lightest oils that can be miscibly . displaced by dry gas or nitrogen at high pressures. The waterbased methods utilize oils in the mid-gravity range, while the CO2 projects cover a fairly broad range of oil gravities between 30 and 45° API. Fig. 3 confirms that all CO2-miscible projects are at depths greater than 2,000 ft. Fig. 4 shows that the nonU.S., world distribution of projects is similar, but there are more hydrocarbon and fewer CO2 projects than in the U.S. The incremental oil production from each EOR project is shown in Figs. 5 and 6. The dominance of steamflooding stands out clearly in these figures. Not only are there far more steamfloods, but the oil produced by steamflooding far exceeds that from all the other methods combined. Note that the largest EOR projects (in terms of oil production), are steamfloods, with the "off-scale" (Fig. 6) Duri steamflood in Indonesia producing more than twice as much oil (245,000 bid) as any other project in the world.

Suggested Criteria for EOR Methods

Oil and reservoir characteristics for successful EOR methods are given in Table 3. The table was compiled from field data for the projects shown in Figs. 3-6, and from the known oil-displacement mechanisms for each of the methods. Very brief descriptions of these mechanism are given in the "thumbnail sketches" of the methods in Tables 4-11.

388

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

"3

Note that we have avoided notations such as ~ (equal to or greater than) because we want to emphasize that the suggested '\arameters are never absolute. They are intended to show '~pproximate ranges of the criteria for good projects. In most cases, when we show >x or <y, etc, there is not a specific upper (or lower) boundary to the parameter except for the limits of the oil and reservoir characteristics, as found in nature. For example, we show that nitrogen floods are recommended for oils lighter than 35 API, but this does not mean that the probability of doing miscible nitrogen floods drops to zero at 34 API. This obvious shortcoming of most screening criteria tables has been noted by ~uthors who use artificial intelligence (AI) methods to select EOR processes for specific reservoirs. I I To overcome the problems that arise by using rigid boundaries in their "crisp" expert systems, some AI workers have utilized "fuzzy logic" methods to obtain much more realistic results 12. In Table 3, we attempt to show that, for a given parameter, if >x is feasible, »x may be even better for a given process. By underlining a value, we indicate the average or mean of the parameter for that EOR method. For example, for the oil gravity in miscible nitrogen floods, >35 ..... .....means that the process 48 should work with oils greater than 35 API (if other criteria are met) and that higher gravity oils (..... are better, and that the ) approximate mean or average of current miscible nitrogen projects is 480 API. The ascending arrow is meant to indicate that higher gravity oils may be better yet. In general, the upper and lower values in Table 3 (> or <) have come from process-mechanism understanding (laboratory ,~xperirnents), and they also include parameters of successful .field projects. For example, even though we are unaware of any miscible CO2 projects in reservoirs with oils of less than 29 API, we list 22 API as the lower limit because extensive laboratory work shows that the required pressure (i.e., MMP, see Table 6) can be met in typical West Texas reservoirs with oils of that gravity. Also, we have lowered the oil gravity requirement to > 12 0 API for immiscible CO2 floods to include a successful project in Turkey (see Fig. 6).
0 0 0 0 0

Gas Injection Methods. Gas injection, the oldest EOR method, is a bright spot in EOR technology. Although most EOR production comes from steamflooding, Figs. 5 and 6 show that gas injection methods are next in importance and appear to be growing throughout the world. Oil production from CO2 flooding is the only EOR method that has continued to increase (Fig. 1) in the U.S. in spite of various declines in oil prices through the years, and more projects are planned. Hydrocarbon gas injection is second to steamflooding for the entire world. Thanks to efforts to reduce gas flaring, gas injection should continue to grow in importance as worldwide oil production expands. After years of extensive laboratory and field experience, the gas EOR methods are now well understood, and screening criteria can be recommended with more confidence than before. Although studied most extensively for CO2, the concept of minimum miscibility pressure (MMP) explains the efficient oil displacements by N2, hydrocarbons, and CO2, As long as this MMP can be achieved in the reservoir, good oil recovery (greater than 90% OOIP in the

region swept) should result, although CO2 displacements are usually more efficient than N2 or CH4. Even though the oil gravity-pressure-depth (MMP) requirements are different for the three gases, Table 3 shows that there is overlap of the criteria for the three methods. Thus, any of the methods will work in a high percentage of the deeper reservoirs, and the fmal choice will often depend on the local availability and cost of the gas to be injected. Nitrogen (and Flue Gas) Injection. Other than compressed air, nitrogen and flue gas are the cheapest gases (especially in terms of volumes at reservoir temperatures and pressures) that can be injected. They are considered together because the pressures required (MMP) for good displacement are similar," and it appears that they can be used interchangeably for oil recovery. Indeed, at least three of the current nitrogen projects (Block 31, TX; E. Binger, OK; Hawkins, TX) 24were operated successfully for years as flue gas injection projects.23•26 However, corrosion was a problem (especially for flue gas from internal combustion engines), and all have switched to nitrogen injection with good results. In addition to its low cost and widespread availability, nitrogen is the most inert of all injection gases. Unfortunately, it has the highest MMP, so miscible displacement is possible only in deep reservoirs with light oils. Hydrocarbon Injection. As one of the oldest EOR methods, hydrocarbon injection was practiced for years before the MMP concept was, well understood. When a surplus of a lowmolecular-weight hydrocarbon existed in some fields, they were often injected to improve oil recovery. The three different methods were described by Stalkupf and are summarized very briefly in Table 5, including tirst-contact-miscible (LPG solvent), condensing (or enriched) gas drive and the vaporizing (or highpressure) gas drive. In terms of the pressure required for efficient miscible displacement, we rank the hydrocarbon gases between the very high pressures required for nitrogen and the more modest range of pressures for CO2 (see Table 6 for the reservoir depth requirement for.different gravity oils). This ranking is correct for methane. However, if a shallower reservoir depth requires a lower pressure, it can be achieved by adding more enriching hydrocarbons (usually C2-C4 if the economics permits).":" This fine-tuning method is practiced most in Canada where cheap CO2 is in short supply and hydrocarbon gases are available. CO2 Flooding. There may be more optimism for CO2 flooding in the U.S. than for any of the other EOR methods. As noted before, it is the only method that has had a continuous increase in production since CO2 flooding started more than 35 years ago. The technical and economic reasons for the success of CO2 flooding have been explained before." In the West TexasNew Mexico Permian basin, a large pipeline supply of natural CO2 is available at a low cost compared to methane, and the pipelines are being extended to more fields.":" The screening criteria in Tables 3 and 6 show that a fairly wide range of crude oils and reservoir depths can meet the requirements for miscible CO2 flooding. The solubility of CO2 in oil decreases with temperature, so the MMP required for a given oil must also increase with

389

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

SPE 35385

temperature." Since the reservoir temperature normally increases with depth, the MMP must also increase with depth, as shown in Fig. 7 for a 40 API oil in typical West Texas reservoirs. Fortunately, the pressure required to fracture reservoirs increases faster than temperature with depth. Therefore, there is an MMP ''window of opportunity," as shown in Fig. 7. The correlations in Fig. 7 and Table 6 come from many sources and are reviewed briefly in Refs. 29, 32, and 33. Most of the relationships among temperature, oil composition and pressure come from extensive work by various workers, primarily on oils from fields in the U.S.34-]7The MMP screening criteria in Fig. 7 should work well for oils that have hydrocarbon distributions similar to the average mid-gravity crude oils of the U.S., especially those from the Permian basin of West Texas and Southeast New Mexico. However, if the oil differs significantly from the types of crudes for which the correlation was developed, additional laboratory tests may be required. Hagadorn and Orr" have shown that a high percentage of multi-ring aromatics will raise the MMP significantly because they are extracted so poorly by the CO2 phase.
0

Chemical and Polymer Flooding and Gel Treatment Methods. Figs. 3~ show that there are relatively few chemical flooding projects (shown as polymer or micellar/polymer) in the world and that these projects contribute little to worldwide EOR production when compared to steamflooding and gas injection. For our screening criteria, we want to concentrate most on current technology that can be applied profitably today. Therefore, we have limited our criteria in Table 3 to these broad methods that are often included in the general term "chemical flooding." We are not aware of any pure alkaline floods at present. There are alkaline/surfactant/polymer (ASP) projects that, hopefully, are a low-cost improvement over micellar/polymer or surfactant flooding. Therefore, we have dropped the separate alkaline flooding category and combined it with the two main surfactant (interfacial-tension lowering) methods as shown in Tables 3 and 7: micellar/polymer; alkaline/surfactant polymer (ASP) and alkaline flooding. There is stilI some excellent chemical flooding research and development work underway in laboratories around the world. The polymer injection projects (especially in the U.S., see Fig. 3) far outnumber the other chemical flooding methods. However, there has been some confusion between polymer flooding for enhanced oil recovery and the injection of gelling polymers for water shutoff in either injection or production wells. Therefore, they are considered separately in Table 3 and in Tables 8 and 9. Wettability is another area of importance to waterflooding, and significant progress on understanding the influence of wettability on oil recovery in being made.":" However, it would be premature to try to include wettability in our screening criteria at this time. Micellar/Po/ymer, AlkalinelSurjactanVPo/ymer (ASP) and Alkaline Flooding. The goal of the chemical methods is to reduce the interfacial tension between oil and water, generally to displace discontinuous trapped oil (Sor) that remains after a

waterflood. Since it is approximately ten times more difficult tOI replace trapped oil than continuous oil," the surfactant slugs for. these tertiary processes must be very efficient. The oil displacement mechanics are well understood and many formulations have been devised to give very high recoveries in laboratory experiments with actual reservoir rocks and fluids. There have been some technical successes in the field.42•4] However, there have been fewer economic successes because the cost of the injectant is too high. Therefore, there has been an effort to lower the injectant cost by adding more alkali and less surfactant or cosolvent to the formulations during the past few years.23•44•4S These mixtures are often termed alkaline/surfactant/polymer (ASP) processes, and very large "slugs" can be injected since the cost is so low compared to the classical micellar/polymer formations. The alkali costs much less than the surfactant or cosolvent and it helps to lower the interfacial tension and reduce adsorption of the surfactant on the rock.46•47 In one case, workers were able to reduce surfactant concentration by ten times by adding low-cost alkali, and the formulation still provided very good oil recovery" The ASP process has also been tested in the field." A recent fieldwide project in Wyoming reports costs of$1.60-$3.50Ibbl of incremental oil produced." Polymer Floods and GeiTreatments. In the past, polymer floods and gel treatments were often lumped together as a single technology." However, these processes have very different technical objectives, so here, they will be considered separately. The distinction between a mobility-control process (e.g., a I polymer flood) and a blocking treatment (e.g., involving cross- ., . linked polymers or other gels) is an important concept to understand. For polymer floods and other mobility-control processes, the mobility-control agent should sweep evenly through the reservoir. In other weirds, the polymer should penetrate as far as possible, into the low-permeability zones because that action provides the driving force for displacing and producing unswept oil. , In contrast, for gel treatments, gel penetration should be minimized in less-permeable, oil-productive zones.. Any gel that forms' in the oil-productive zones reduces the oil-displacement efficiency and retards oil produc-

'I

tion."
For existing gels and gelants that are used as blocking agents, the following behavior is observed during flow through porous media. S3-SS First, before gel aggregates grow to approach the size of pore throats, gelants flow through porous media like solutions without crosslinkers. Second, after gelation (or after gel aggregates grow to the size of pore throats), gel movement through porous rock is negligible. Third, in porous rock, the transition from a freely flowing gelant to an immobile gel occurs abruptly. After gel formation, crosslinked polymers, gels, gel aggregates, and the so-called "colloidal-dispersion gels" do not flow through porous rock like viscous polymer solutions. 54 Also, they do not enter and block the most-permeable strata first and then sequentially enter and block progressively less-permeable zones. Gelants and polymer solutions enter all zones simultaneously. S2 (Of course, the distance of polymer or gelant penetration depends directly on the permeability.) Understanding these concepts is particularly important for projects that were designed

390

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

as polymer floods but that used HP AM crosslinked with aluminum citrate (i.e., the "colloidal-dispersion gels")." For these .projects, an important question is, "Would the field response j have been better ifHPAM had been injected without aluminum

citratej?"
Polymer Flooding. Over the past 35 years, a large number of polymer floods have been applied over a remarkably wide range of conditions. 57.58Reservoir temperatures ranged from 46 to 235°F; average reservoir penneabilities ranged from 0.6 to 15,000 md; oil viscosities ranged from 0.01 to 1,494 cp; net pay \ ranged from 4 to 432 ft; and resident brine salinities ranged from 0.3 to 21.3% total dissolved solids (IDS). At project startup, the percent of original oil in place (%OOIP) ranged from 36% to 97.1%, and the producing water/oil ratio ranged from 0 to 100. During the 1980s, polymer floods were applied in sand or sandstone reservoirs about four times more frequently than in carbonate reservoirs. 58 In concept, a polymer flood could improve sweep efficiency during any waterflood. However, a number of technical and economic factors have limited the application of successful polymer floods. The cost-effectiveness of polymers (i.e., the mobility reduction or viscosity provided per unit cost of polymer) is the main economic limitation. For example, if the cost ofacrylamide/acrylate copolymers (HPAM) and xanthan polymers were substantially lower, higher polymer concentrations and larger polymer-bank sizes could be afforded in a given application. This, in turn, would lead to greater oilrecovery efficiencies, higher profits, and a wider range of potential applications. \ Cost-effectiveness also impacts the permeability constraints ) for polymer flooding. For a given polymer, chemical retention increases and the rate of polymer propagation decreases with decreasing rock permeability. Current high-molecular-weight polymers often experience high retention and low propagation rates for rock penneabilities below 100 md." This permeability constraint can be relaxed by using polymers with lower molecular weights. However, the viscosity provided by a polymer decreases with decreasing molecular weight, so more polymer (and a higher cost per viscosity unit) is needed as the rock permeability and the maximum allowable polymer molecular weight decrease. An important issue related to reservoir permeability is that of injectivity (injection rate per pressure drop). In wells that are not fractured, injection of viscous polymer solutions will necessarily decrease injectivity. In order to maintain the waterflood injection rates, the selected polymer-injection wells must allow higher injection pressures. This requirement becomes increasingly difficult to fulfill as the formation penneability decreases, unless the wells are fractured. If injectors are fractured, one must ask, "Will the increased injectivity from fracturing outweigh the increased risk of channeling?" (In a later section, we suggest that horizontal injection wells may alleviate injectivity limitations in some cases.) Cost-effectiveness also affects the temperature constraints for polymer flooding. More than 95% of previous polymer floods were applied in reservoirs with temperatures below .200°F.58 This fact reflects wide-spread doubt that HPAM and xanthan polymers are sufficiently stable at elevated temperatures.

Literature reports 59question whether these polymers are stable for field applications above 175°F. More stable polymers (e.g., scleroglucan and acrylamide/ AMPS/NVP copolymers and terpolymers) are available for high-temperature use, but the cost and cost-effectiveness of these polymers have limited their application to date." Of course, significantly higher oil prices and/or breakthroughs in reducing polymer production costs could change this situation. For many years, water salinity has been an important issue in polymer flooding. 59 In the range from 0 to 1% TDS, the viscosities of HPAM solutions decrease substantially with increased salinity. Thus, high-salinity HPAM solutions are relatively ineffective during polymer flooding. Differences of opinion existed concerning the viability of injecting low-salinity HPAM solutions into reservoirs with high-salinity waters. An important paper that addressed this issue was presented by Maitin." In a well-documented field study, Maitin demonstrated the conditions needed for low-salinity HPAM solutions to be effective in high-salinity reservoirs. In reviewing literature reports of polymer floods, we often noted considerable uncertainty in assessing the benefits after a given project was completed. Most previous polymer floods used relatively small quantities of polymer (both in terms of polymer concentration and bank size)." Consequently, relatively small lOR values (1 to 5% OOIP) were often projected that resulted in small alterations of the oil-production decline curves and the WOR curves. Commonly, these small alterations were difficult to discern when comparing the actual polymer-flood response with the projected waterflood response. In contrast, several polymer floods stand out that showed definitive responses, such as at the Marmul," Oerrel,":" Courtenay.Pand Daqing" fields. Properties of these successful polymer floods are listed in Table 8 along with median values for all polymer floods that were applied during the 1980s. The four. successful floods listed in Table 8 had a number of features in common. These characteristics may be useful as screening criteria for .today's economic environment. First, the floods were applied in high-permeability (>0.87 darcy) sands and low-temperature (86°F to 136°F) reservoirs. High oil saturations (71 to 92 % OOIP) were present at project startup, and the oil/water viscosity ratios (15 to 114) at reservoir temperature were relatively high. The injected polymer solutions contained relatively high HPAM concentrations (900 to 1,500 ppm) in low-salinity waters, and large quantities of polymer (162 to 520 lbs polymer/ac-ft) were injected. Finally, the incremental oil recoveries (11 to 30 % original oil in place or 155 to 499 bbl oil/ac-ft) were high. Gel Treatments. Gel treatments have been applied under conditions as diverse as those listed previously for polymer floods.58.65As mentioned earlier, the technical objective ofa gel treatment should be very different from that of a polymer flood. In most cases, the objective of a gel treatment is to prevent channeling of fluid (usually water) without damaging hydrocarbon productivity. After extensive discussions with experts from the oil and service companies.Y" we developed the criteria listed in Table 9 for selection of gel-treatment candidates for injection and production wells.

391

J.J. TABER, F.D. MARTIN, AND A.S. SERIGHT

SPE 35385

For both injection and production wells, the first criterion indicates that a sufficient target of mobile oil must be present, and realistic calculations should indicate that the oil can be recovered economically. In Table 9A, several qualifications were added to clarify what is meant by "low sweep efficiency." We felt that this clarification was necessary because different operators interpret this phrase in radically different ways. Cases exist where gel treatments were implemented even though more than 50% OOIP had been recovered before the treatment." Point c in the first criterion is particularly important for applications in gray areas. One criticism of this point might be that, often, insufficient manpower or reservoir description is available to adequately predict the benefits of a gel treatment. If that is the case and if the second or third criterion in Table 9A cannot be met, then we feel that a gel treatment in an injection well is unlikely to be successful. Recovery calculations were also felt to be very important in the selection of production-well candidates (Table 9B). An essential element of improving the success rates for gel treatments is adequate identification of the source and nature of the channeling problem. Determination of the importance of fractures as channels is a first priority. For applications in injection wells or waterflood production wells, interwell tracer studies can be the most effective tool in identifying very severe channeling between injector-producer pairs." Very rapid tracer transit times (less than one week) probably indicate that the channel is a fracture or a formation part. In addition to diagnosing the severity of the channeling problem, tracer studies can be very useful in designing the volume of gelant to be injected and assessing the ultimate effectiveness of the treatment. 68 Table 9 indicates that high well injectivities or productivities are important for candidate selection. These requirements recognize that a gel treatment will generally reduce the flow capacity of a well. For a given pressure limit, the injection or production rates will be lower after a treatment than before a treatment. By having excess injection or production capacity before the treatment, the operator will be more likely to tolerate the reductions in flow capacity that result from a gel treatment. Simple injectivity or productivity calculations can aid in establishing the nature of a channeling problem." For unfractured injection wells, Table 9A lists three important conditions that must be present before candidate selection: (1) poor injection profiles must correlate from well to well, (2) effective barriers to crossflow must exist, and (3) gels can be placed in the offending channel without damaging oil zones. To prevent damage to oil productivity, zones usually must be mechanically isolated during gelant placement in unfractured injection wells." This requirement also applies in unfractured production wells. If zones are not isolated during gelant placement in unfractured wells, low-permeability zones can be seriously damaged even in extremely heterogeneous reservoirs (e.g., Dykstra-Parsons coefficient ofO.9).5l Most vendors and operators feel that the candidate well should be in good mechanical condition for applications in injection wells," but this requirement may be less critical for production-well applications. Not all gel applications were directed at in-depth channeling

problems. Often, gels are used as a substitute for cement in fixing casing leaks or flow behind pipe. The main advantage of gels over cements is a superior ability of gelants to penetrate into constricted spaces, such as narrow channels behind pipe, small casing leaks, and fractures. Also, because gelants can penetrate into porous rock, gels can sometimes form a better pipe-formation seal than can be obtained using cement. 67-69 In addition to gels, many different water-shutoff methods are available, including cement, sand plugs, calcium-carbonate plugs, resins, foams, particulates, emulsions, precipitates, microorganisms, mechanical devices (e.g., packers and bridge plugs), well abandonmentlinfill drilling, pattern flow control, and horizontal wells. Each of these methods will work very well in solving some types of problems but not others. The proper choice of water-shutoff material and method depends critically on the cause of the problem. Gel treatments work best for those problems identified in Table 9. Thermal-Mechanical Methods for Heavier Oils and Tar Sands. Thermal methods account for the biggest share of the world's enhanced oil production. The largest EOR operations in many countries (e.g. Canada, Columbia, Germany, Indonesia, Trinidad, U.S. and Venezuela) are either steamfloods or surface mining operations. In the past, the production of bitumen from tar sands has not normally been included in EOR screening. criteria or surveys.perhaps because the mining operations are not considered a part of reservoir engineering. However, the resource is so important that hydrocarbon recovery from tar sands should be included in listings ofEOR or lOR processes. There is a very strong effort to try to recover these extremely viscous oils by in- ' situ methods 70 to avoid the cost of surface mining and to open up vast deeper reserves. One method that shows promise uses horizontal wells in a variation of steamflooding known as Steam Assisted Gravity Drainage or SAGD7I-73 This mechanism is akin to the enhanced gravity drainage by immiscible gas injection mentioned previously and for which screening criteria are given in Table 3. In general, the screening criteria for SAGO would be similar except that the depth, viscosity and API gravity ranges should be extended to include the tar sands. Thermal EOR projects have been successful for more than 30 years, and the methods have been described in many publications and a few books.74-76 We comment here on only a few aspects that relate to screening criteria. In general, .thermal methods have been used for those heavy oil reservoirs that cannot be produced in any other way because the oil is too viscous to flow without the application of heat and pressure. In order to be produced at profitable rates, the sands must have a high permeability and oil saturations must be high at the start of the process. Therefore, the successful projects are almost always enhanced secondary (or even enhanced primary, since primary production was essentially nil in many fields). In-Situ Combustion. In situ combustion seems like an ideal EOR method because: It utilizes the two cheapest and most plentiful of all EOR injectants: air and water.

392

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

- For fuel, it burns about 10% of the least desirable fraction of the oil, and may upgrade the rest. - It works over a wider range of field conditions than steam./ooding, especially in deep reservoirs. This complicated process has been studied extensively" and tried in many different types of reservoirs.":" However, at a recent symposium on in situ combustion, Farouk Ali-claimed that "insitu combustion remains the most tantalizing EOR method. ,,10 At the same symposium, Sarathi and Olsen showed that only one of eight cost-shared projects was an economic success, but it provided valuable information on how to engineer a successful project.11 ' According to Turta, air injection must start in the uppermost part of the reservoir, so the combustion front can propagate down dip, preferably with a line drive well configura.tion." Turta also described benefits of horizontal wells that have shown promising results in two Canadian combustion projects. Efforts are continuing to improve the combustion process and to apply it to different types of fields. For example, oxygenenriched fireflooding continues to look promising for reservoirs that require large volumes of gas at high flow rates where oxygen is cheaper than air. Newer materials and technology should help solve some of the field problems." In another application, horizontal wells are being planned to improve light-oil, in-situ combustion projects (310 -4r API) in North and South Dakota. Air injection has been underway since 1981. The operator hopes that horizontal wells will increase the recovery from the current 20-30% to 50% OOIP.14 Deeper, light-oil reservoirs with significant dip are also targets for a new method of in situ -ombustion that might be considered another variation of )manced gravity drainage by nitrogen or flue gas. IS In this process, air is injected in the formation, and the resulting combustion front moves down-dip to displace the oil either miscibly or immiscibly by the flue gas produced from the combustion." Combustion continues to have great promise for a much wider range of fields than the original heavy oil targets, especially in deeper reservoirs. However, it is a complicated method with safety and corrosion problems that always need attention. These problems and their solutions were described in a recent review." Steam flooding. Steamflooding is the oldest commercial EOR method; the oil displacement mechanisms are well understood. Much of the current emphasis is on improving the economics through better reservoir management. 88 As for screening criteria, the observations in our earlier paper still apply, i.e., good projects require thick, shallow deposits with high oil saturations and good permeabilities," In times of low oil prices, the economics are very tight, especially since the heavy oil has less value than higher gravity crudes. In recent years, the cogeneration of steam and electric power has been very beneficial to both the economics and environmental problems.":" Steamflooding was probably the first EOR method to take advantage of the benefits of horizontal wells." References indicate that their use and other advanced engineering methods should make it possible to extend steamflooding to both lighter and heavier oils." Laboratory tests show that steamflooding is an -fficient mechanism for displacing light oils." Several field tests ... iave also been conducted in light-oil reservoirs and a few have

been successful. 91,92 The Duri project in Indonesia is sometimes referred to as a light-oil project because its 22 API oil is outside Unitar's defmition of heavy oil: 10Y2- 20W API inclusive." As the world's largest EOR project, the Duri steamflood is certainly successful (see Fig. 6). However, its starting oil saturation of 63% is near the average of the successful steamfloods in the world. Most of the other light-oil steamfloods had much lower oil saturations, so economic success was more difficult. In Table 3, we left a question mark for the upper limit to the API gravity for steamflooding a medium gravity oil that could be waterflooded as well as steamflooded. The steamflood should produce much more oil, but an effective waterflood will be cheaper. It will take a careful economic analysis of each potential light-oil steamflood to determine if the additional oil will pay for the additional cost of the steamflood. It does appear that light-oil steamfloods should always be planned as enhanced secondary operations. At the other end of the oil-gravity-steamflooding spectrum are the aforementioned Steam Assisted Gravity Drainage (SAGD) projects in heavy-oil or tar sands. Although different techniques are under development, almost all of these require one or more horizontal wells to inject the steam and withdraw the melted bitumen." Normally" the steam is injected into the upper well of two parallel horizontal wells. With the application of hot steam and pressure, the tar melts and flows by gravity to the lower well, where it is pumped to the surface. Mining and Extraction. Although not normally listed with EOR screening criteria, we include surface mining because the tar sands are such an important hydrocarbon resource, and the production of synthetic crude from recovered bitumen keeps increasing.":" In general, the mining method is used only when the oil is so viscous that it cannot be recovered by any other method because the mining and upgrading of the bitumen are more costly than in-situ recovery methods. For this reason, the tar sands must have a high oil (bitumen) saturation, and the ratio of overburden to tar sand must be low as shown in the screening criteria of Table 3. As mentioned in the previous section, there is an increased effort to producethese viscous hydrocarbons by in-situ methods such as the SAGD process.
0

Utilization of Screening Guides With the reservoir management practices of today, engineers consider the various IOR-EOR options much earlier in the productive life ofa field. For many fields, the decision is not if, but when to inject something. Obviously, economics always play the major role in "go-no go" decisions for expensive injection projects, but a cursory examination with the technical criteria is helpful to rule out the less-likely candidates. The criteria are also useful for surveys of a large number of fields to determine if specific gases or liquids could be utilized for oil recovery if an injectant was available at a low cost. This application of the CO2 screening criteria is described in the next section.

393

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

SPE 35385

Estimation of the Worldwide Quantity of COl that Could Be Utilized for Oil Recovery. The miscible and immiscible screening criteria for CO2 flooding in Table 6 were used to make a rough estimate of the total quantity of CO2 that would be needed to recover oil from qualified oil reservoirs throughout the world. The estimate was made for the lEA Greenhouse Gas R & D Programme, Cheltingham, England, as part of their ongoing search for ways to store or dispose of very large amounts of CO2, in case that becomes necessary to avert global warming. The potential for either miscible or immiscible CO2 flooding for almost 1,000 oil fields was estimated by using the depth and oil gravity data published in a recent Oil and Gas Journal survey." The percent of the fields in each country that met the criteria in Table 6 for either miscible or immiscible CO2 flooding was determined and combined with that country's oil reserves to estimate the incremental oil recovery and CO2 requirements. Assuming that one-half of the potential new miscible projects would be carried out as more-efficient enhanced secondary operations, an average recovery factor of22% OOIP was used, and 10% recovery was assumed for the immiscible projects. A CO2 utilization factor of 6 mcf per incremental barrel was assumed for all estimates. This estimated oil recovery for each country was then totaled by region, and all of the regions were totaled in Table 12 to provide the world totals." The basis for the assumed incremental oil recovery percentage and CO2 utilization factors and other details are given in Ref. 33. Economics were not a part of this initial hypothetical estimate. It was just assumed that pure, supercritical CO2 was available (presumably by pipeline from power plants) for each of the fields and/or regions of the world. Table 12 shows that about 61 billion metric tons (61 gigatonnes-see conversion factors) of CO2 would be required to produce 206 billion barrels of additional oil. The country by country results and other details (including separate sections on the costs of CO2 flooding) are given in Ref. 33. Although not much better than an educated guess with many qualifying numbers, our estimate agrees well with other estimates of the quantity of CO2 that could be stored (or disposed ot) in oil reservoirs." Although this is a very large amount of CO2, when the CO2 demand is spread over the several decades that would be required for the hypothetical CO2 flooding projects, it would reduce the worldwide power plant CO2 emissions into the atmosphere by only a few percent per year. Therefore, more open-ended CO2 disposal methods (such as the more-costly deep ocean disposal) will probably be needed if the complex general circulation models of the atmosphere ever prove conclusively that global warming from excess CO2 is underway.":" However, from the viewpoint of overall net cost, one of the most efficient CO2 disposal/storage systems would be the combined injection of CO2 into oil reservoirs and into any aquifers in the same or nearby fields.33•1°O By including aquifers, this potential for underground CO2 storage would be increased significantly, and the quantity sequestered could have a significant impact on reducing the atmospheric CO2 emissions from the world's power plants.

Impact of Oil Prices on EOR

Major new EOR projects will be started only if they appear, profitable. This depends on the perception of future oil price. Therefore, the relationship between future oil prices and EOR was a major thrust of the two aforementioned NPC reports." These extensive studies used as much laboratory and field information as possible to predict the EOR production in the future for different ranges of oil prices. Now, it is possible to compare the NPC predictions with actual oil production to date. These comparisons were made recently to see how oil prices might affect oil recovery from future CO2 projects." We have extended these graphical comparisons and reproduced them here as Figs. 8-10. In general, the figures confirm that EOR production increases when prices increase and EOR production declines when prices fall, but not to the extent predicted. There is a time lag before the effect is noted. Figs. 8 and 9 show that total EOR production did increase in the early eighties when oil prices were high. This was in response to an increase in the number of projects during this period when prices of even $SOlbbl or more were predicted. Although the rate of increase slowed in 1986 when oil prices dropped precipitously, the EOR production did not decline until 1992; after several years of low oil prices, i.e. less than $20 per barrel. . The 1984 predictions were made while oil prices were high (::: 30Ibbl), but they were not nearly as optimistic as those made $ in 1976, when oil prices were lower. However, the 1984 predictions benefited from experience gained from the field projects conducted in the interim. The only price common to both NPC / reports is for $201bbl. The 1976 $201bbl prediction would be off '. the scale by 1990 if plotted on the 1984 graph of Fig. 9. However, the $201bbl prediction of 1984 is close to the $1 Olbbl value of 1976. Note that the actual oil production does track predictions of $IOlbbI for 1976 and $20 for 1984 in Figs. 8 and 9. Since oil prices were at or below $201bbl for much of the period since 1986, the NPC predictions have merit. The impact of the lower oil prices since 1986 was fmally felt in 1992 when EOR production (except for CO2 flooding) dropped for the first time due to fewer projects. The number' of EOR projects has been declining steadily since 1986, the year that oil prices fell. However, Table 13 shows that the profits from EOR projects did not decline during the recent years of low oil prices. For most EOR methods, Table 13 shows that there was an increase in the percentage of projects that were profitable, presumably because the less-efficient projects were discontinued. ,i The optimism that came from the much higher oil prices in the late seventies and early eighties was probably very fortunate for the CO2 flooding industry in the U.S. The large natural CO2 sources were developed, and pipelines were built during this period. The inexpensive, supercritical CO2 has been flowing into the Permian Basin ever since. The pipelines are being extended, and more projects are being started as CO2 flooding efficiencies continue to increase.P'" Figure 10 shows that (after the long "incubation" period) CO2 flooding has now exceeded the NPC prediction for oil prices of$20lbbl. This is in spite of the fact that oil prices were near or less than $201bbl for much of the time since 1986.

394

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

References ,Future Technical and Economic Improvements ,:Expected Even with the low oil prices, there are many technology advances that should continue to improve the outlook for EOR and lOR. These include: (1) 3D seismic-to learn where the target oil is located, in old as well as new fields, (2) use of horizontal injection as well as production wells," (3) cheaper horizontal injection wells with multilaterals, short radius, and those used in lieu of more costly infill drilling, etc., (4) more efficient reservoir (simulation methods, and (5) foam for mobility control, especially in CO2 flooding. These and other technological advances are expected to improve the process efficiency and cost effectiveness of EOR methods in the future. Conclusions I. Screening criteria and brief descriptions are presented for the major EOR methods. The criteria are based on oil displacement mechanisms and the results ofEOR field projects. The depth, oil gravity, and oil production from hundreds of projects are displayed in graphs to show the wide distribution and relative importance of the methods. Steamflooding continues to be the dominant method, but hydrocarbon injection and CO2 flooding are increasing. 2. If only oil gravity is considered, the results show that there is a wide choice of effective methods that range from miscible recovery of the lightest oil by nitrogen injection to steamflooding and surface mining for heavy oil and tar sands. However, there >isoften a wide overlap in choices. /3. With low oil prices, there is less chemical flooding of the intermediate-gravity oils that are normally waterflooded. Polymer flooding continues to show promise, especially if projects are started at high oil saturations. The criteria for the effective use of gels to control water flow in both injection and production wells are presented. 4. The CO2 screening criteria were used to estimate the capacity of the world's oil reservoirs for the storage/disposal of CO2, If only depth and API gravity are considered, it appears that about 80% of the world's reservoirs could quality for some type of CO2 injection to produce incremental oils. 5. The impact of oil prices on EOR production in the U.S. was considered by comparing the recent EOR production to that predicted by the NPC reports for various oil prices. Although lower oil prices since 1986 have reduced the number of EOR projects, the actual incremental production has been very close to that predicted for $20lbbl in the 1984 NPC report. Incremental oil production from CO2 flooding has increased continuously, and now exceeds the predictions for $20 oil in the NPC report. Acknowledgments The authors express their appreciation to Tommy Morris for assistance in screening many reservoirs for CO2 flooding potential, to Mailin Seldal and Steve Anderson for help on the figures and to Liz Bustamante for valuable assistance in the \preparation of this manuscript.
1. 2. 3. Muskat, M.M.: Physical Principles of Oil Production, McGrawHill, New York (1949) 709-10. Ibid., 715-724. Beecher, C.D.: "Increasing the Recovery of Oil by Repressuring," Discussion in Chapter Viii, "Petroleum Engineering Problems,' Petroleum Development and Technology in 1927, AIME, New York (1928), 77, 379-82. Willhite, G.P.: Waterflooding, Society of Petroleum Engineers, Richardson, TX (1986). Geffen, T.M., "Oil Production to Expect from Known Technology,", Oil & GasJ. (May 1983) 66--76. Geffen, T.M., "Improved Oil Recovery Could Help Ease Energy Shortage," World Oil (Oct. 1983) 84-88. Haynes, H.J. et aI.: Enhanced Oil Recovery, National Petroleum Council, Industry Advisory Council to the U.S. Department of the Interior, (1976). Bailey, RE. et al.: Enhanced Oil Recovery, National Petroleum Council, Industry Advisory Committee to the U.S. Secretary of Energy, Washington, D.C. (1984). . Taber, 1.1. and Martin, F.D.: "Technical Screening Guides for the Enhanced Recovery of Oil," paper SPE 12069 presented at the 1983 Society of Petroleum Engineers Annual Technical Conference and Exhibition, San Francisco, CA, Oct. 5-8. Goodlett, G.O., et al.: "Lab Evaluation Requires Appropriate Techiques," Oil & Gas i., June 23, 1986, 47-54. Parkinson, W.l et al.: "AnExpert System for Screening Enhanced Oil Recovery Methods," presented at AICHE Summer Meeting, Aug. 19..,.~21990, San Diego CA. Parkinson, W.J. et al.: "Screening EOR Methods with Fuzzy Logic," presented at Third International Reservoir Characterization Conference, Tulsa, OK, Nov. 3-5, 1991. Elemo, R.O. and ElmtaIab, L: "A Practical Artificial Intelligence Application in EOR Projects, " paper SPE 26248 presented at the 1993 Society of Petroleum Engineers Computer Conference, New Orleans, LA, July 11-14. Basnieva, I.K. et al.: "Comparative Analysis of Successful Application of EOR in Russia and CIS," paper SPE 28002 presented at the 1994 Society of Petroleum Engineers Centennial Petroleum Engineering Symposium, Tulsa OK, Aug. 29-31. Arnold, M.D.: "Screening Criteria Decide Best Process for Enhanced Oil Recovery," Proc., 41st Annual Southwestern Petroleum Short Course Association Inc. et al., Lubbock, TX (1994) 299-305. Rivas, 0., Embid, S., and Bolivar, F.: "Ranking Reservoirs for Carbon Dioxide Flooding Process," SPE Advanced Technology Series (year?) 2, No. 1,95-103. Ray, R.M.: "An Evaluation of Known Remaining Oil Resources in the United States: Project on Advanced Oil Recovery and the States," DOEIBC/14431-1 (Vol. I), Contract No. DE-FG2289BC14431, U.S. DOE (Oct. 1994). Ray, R.M.: "An Evaluation of Known Remaining Oil Resources in the State of California: Project on Advanced Oil Recovery and the States," DOEIBC/1443 I-I (Vol. 2), Contract No. DE-FG2289BC14431, U.S. DOE (Oct. 1994). Ray, R.M.: "An Evaluation of Known Remaining Oil Resources in the States of Louisiana and Texas: Project on Advanced Oil Recovery and the States," DOEIBC/1443 I-I (Vol. 3), Contract No. DE-FG22-89BCI443 I, U.S. DOE (Oct. 1994). Ray, R.M.: "An Evaluation of Known Remaining Oil Resources in the States of New Mexico and Wyoming: Project on Advanced Oil Recovery and the States," DOElBC/I443 I-I (Vol. 4), Contract

4. 5. 6. 7.

8.

9.

10. II.

12.

13.

14.

15.

16.

17.

18.

19.

20.

395

10

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT No. DE-FG22-89BCI443 I, U.S. DOE (Oct. 1994). Ray, R.M.: "An Evaluation of Known Remaining Oil Resources in the States of Kansas, Illinois, and Oklahoma: Project on Advanced Oil Recovery and the States," DOEIBC/1443 I-I (Vol. 5), Contract No. DE-FG22-89BC 14431, U.S. DOE (Oct. 1994). Stalkup, F.E.: Miscible Displacement, Monograph Series, Society of Petroleum Engineers, Dallas, New York (1983). Taber, 1.1.: "Environmental Improvements and Better Economics in EOR Operations," In Situ (1990) 14 (4), 345-404. Moritis, G.: "EOR Dips in U.S. but Remains a Significant Factor," Oil & Gas J., (Sept. 26, 1994) 51-79. Emmons, F.R. et al.: "Nitrogen Management at the East Binger Unit Using an Integrated Cryogenic Process," paper SPE 15591 presented at the 1986 SPE Annual Technical Conference and Exhibition, New Orleans, LA, Oct. 5-8. Taber, 1.1.: "The Use of Flue Gas for the Enhanced Recovery of Oil," Presented at the Symposium "EOR by Gas Injection," International Energy Agency Collaborative Research Program on EOR, Copenhagen, Denmark, Sept. 14, 1988. Batycky, J.P. and Nagra, S.S.: "The Application of New Technology in the Judy Creek Miscible Project," presented at the Symposium on Field Studies, International Energy Agency Collaborative Research Program on EOR, Sydney, Australia, Sept. 30, 1987. Sibbald, L.R., Novosad, Z., and Costain, T.G.: "Methodology for the Specificiation of Solvent Blends for Miscible Enriched-Gas Drives," SPERE (Aug. 1991) 6, No.3, 373. Martin, F.D. and Taber, 1.1.: "Carbon Dioxide Flooding," JPT (April 1992) 396-400. "C02 Activity Picks up in W. Texas, New Mexico," Oil & Gas J. (July 17, 1995) 26-29. Hsu, C-F., Koinis, R.L., and Fox, C.E.: "Technology, Experience Speed CO2 Flood Design," Oil & Gas J. (Oct. 23, 1995) 51-? Heller, J.P. and Taber, 1.1.: "Influence of Reservoir Depth on Enhanced Oil Recovery by CO2 Flooding," paper SPE 15001 presented at the 1986 Permian Basin Oil and Gas Recovery Conference, Midland, TX, March 13-14. Taber, 1.1.: "A Study of Technical Feasibility for the Utilization of CO2 for Enhanced Oil Recovery," , The Utilization of Carbon Dioxide from Fossel Fuel Fired Power Stations, P. Reimer, lEA Greenhouse Gas R&D Programme, Cheltenham, England (1994) Appendix B, 134-204. Holm, L.W. and Josendal, V.A: "Effect of Oil Composition on Miscible-Type Displacement by Carbon Dioxide," SPEI (Feb. 1982) 87-98. Orr, F.M., lr., and Taber, 1.1.: "Use of Carbon Dioxide in Enhanced Oil Recovery, Science (May II, 1984) 563-69. Yellig, W.F. and Metcalfe, R.S.: "Determination and Prediction of CO2 Minimum Miscibility Pressures," JPT(lan. 1980) 160-68. Orr, F.M., lr., and Silva, M.K.: "Effect of Oil Composition on Minimum Miscibility Pressure-Part 2: Correlation," SPERE (Nov. 1987) 1, 479-91. Hagedorn, K.D. and Orr, F.M., lr.: "Component Partitioning in COiCrude Oil Systems: Effects of Oil Compositions on f:0 Displacement Performance," paper SPE 25169 presented at the SPE International Symposium on Oilfield Chemistry, New Orleans, March 2-5, 1993. Morrow, N.R.: "Wettability and Its Effect on Oil Recovery," JPT (Dec. 1990) 1476-84. Buckley, 1.S., Bousseau, C., and Liu, Y.: "Wetting Alteration by Brine and Crude Oil: from Contact Angles to Cores," paper SPE 30765 presented at the 1995 SPE Annual Technical Conference and Exhibition, Dallas, Oct. 22-25. 41. 42.

SPE 35385

21.

22. 23. 24.

l

43.

44. 45.

25.

46. 47.

26.

27.

48.

49.

28.

29. 30. 31. 32.

50.

51. 52.

33.

53. 54.

34.

55. 56.

35. 36. 37.

57.

38.

58.

39. 40.

59. 60.

Chatzis, I. And Morrow, N.R.: "Correlation of Capillary Number Relationships for Sandstone," SPEI (April 1983) Maerker, J.M. and Gale, W.W.: "Surfactant Flood Process Design for Loudon," paper SPEIDOE 20218 presented at the 7th Joint Symposium for Enhanced Oil Recovery, Tulsa, OK, April 22-25, 1990. Reppert, T.R. et al.: "Second Ripley Surfactant Flood Pilot Test," paper SPEIDOE 20219 presented at the 7th Joint Symposium for Enhanced Oil Recovery, Tulsa, OK, April 22-25, 1990. M. Baviere et al.: "Improved EOR by Use of Chemicals in Combination," SPERE (August 1995) 10, No.3, 187-193. Shutang, G. Huabin, L. and Hongfu, L.: "Laboratory Investigation of Combination of Alkali/Surfactant/Polymer Technology for Daqing EOR," SPERE (August 1995) 10, No.3, 187-193. Lorenz, P.B. and Peru, D.A.: "Guidelines Help Select Reservoirs for NaHC03 EOR," Oil & Gas J. (Sept. II, 1989) 53-57. Eme, V.O.: "Design of AlkalineiSurfactant/Polymer Enhanced Oil Recovery Scheme for a Saudi Arabian Limestone Reservoir," MS Thesis, King Fahd Univ. Petrol. Min. (1994). Yang, C. et al.: "The Alkaline-Surfactant-Polymer Combination Flooding and Application to Oilfield for EOR," Proc., 8th EAPG Impr. Oil Recovery Europe Symp., Vienna (1995) 1, 183-190. Pitts, M.1., Surkalo, H. And Mundorf, W.R.: "Detailed Evaluation of the West Kiehl Alkaline-Surfactant-Polymer Field Project and Its Application to Mature Minnelusa Waterfloods," annual report (January to December 1993), DOElBC/14860-5, U.s. DOE (Nov. 1994). . Wyatt, K. et al.: "Alkaline-Surfactant-Polymer Technology Potential of the Minnelusa Trend, Powder River Basin," paper SPE 29565 presented at the 1995 SPE Rocky Mountain RegionILowPermeability Reservoir Symposium, Denver, March 20-22. Leonard, 1.: "Increased Rate of EOR Brightens Outlook," Oil & GasJ. (April 14, 1986)71-101. Seright, R.S.: "Placement of Gels to Modify Injection Profiles," paper SPEIDOE 17332 presented at the 1988 SPEIDOE Enhanced Oil Recovery Symposium, Tulsa, April 17-20. Hejri, S. et al.: "Permeability Reduction by a XanthanlCr(III) System in Porous Media," SPERE (Nov. 1993) 299-304. Seright, R.S.: "Improved Techniques for Fluid Diversion in Oil Recovery Processes," second annual report (DOElBC/14880-1 0), Contract No. DE-AC22-92BCI4880, U.S. DOE (March, 1995) 5164. . Seright, R.S.: "Gel Placement in Fractured Systems," SPEPF (Nov. 1995), 241-248. Fielding, R.C., Gibbons, D.H., and Legrand, F.P.: "In-Depth Drive Fluid Diversion Using and Evolution of Colloidal Dispersion Gels and New Bulk Gels: An Operational Case History of North Rainbow Ranch Unit," paper SPE .27773 presented at the 1994 SPEIDOE Symposium on Improved Oil Recovery, Tulsa, April 17-20. Manning, R.K. et al.: "A Technical Survey of Polymer Flooding Projects," DOE report DOElBC/I0327-19, U.S. DOE, (Sept. 1983). Seright, R.S.: "Improved Techniques for Fluid Diversion in Oil Recovery," first annual report, DOElBC/14880-5, U.S. DOE (Dec. 1993) 2-72. Sorbie, K.S.: Polymer-Improved Oil Recovery, Blackie, CRC Press, Inc, 1991,61-64,83-114, 148. Maitin, B.K.: "Performance Analysis of Several Polyacrylamide Floods in North German Oil Fields," paper SPE 24118 presented at the 1992 SPEIDOE Symposium on Improved Oil Recovery, Tulsa, April 22-24.

396

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

11

61. Koning, E.J.L., Mentzer, E., and Heemskerk, 1.: "Evaluation ofa Pilot Polymer Flood in the Mannul Field, Oman," paper SPE 18092 presented at the 1988 SPE Annual Technical Conference and Exhibition, Houston, TX, Oct. 2-5. 62. Maitin, B., Daboul, B., and Sohn, W.O.: "Numerical Simulation for Planning and Evaluation of Polymer Flood Process: A Field Performance Analysis," SPE paper 17631 presented at the 1988 SPE International Meeting on Petroleum Engineering, Tianjin, China, Nov. 1-4. 63. Putt, A.G., Bazin, 8., and Pedron, B.M.: "Commercial Polymer Injection in the Courtenay Field, 1994 Update," SPE paper 28601 presented at the 1994 SPE Annual Technical Conference and Exhibition, New Orleans, Sept. 25-28. 64. Wang Demin et al.: "Commercial Test of Polymer Flooding in Daqing Oil Field," SPE paper 29902 presented at the 1995 SPE International Meeting on Petroleum Engineering, Beijing, China, Nov. 14-17. 65. Seright, R.S. and Liang, 1.: "A Survey ofField Applications of Gel Treatments for Water Shutoff," paper SPE 26991 presented at the 1994 SPE III Latin American & Caribbean Petroleum Engineering Conference, Buenos Aires, Argentina, April 27-29. 66. Sydansk, R.D. and Moore, P.E.: "Gel Conformance Treatments Increase Oil Production in Wyoming," Oil & Gas J (Jan. 20, 1992) 40-45. 67. Jurinak, J.J., Summers, L.E., and Bennett, K.E.: "Oilfield Application of Colloidal Silica Gel," SPERE (Nov. 1991) 406-412. 68. Odorisio, V.G. and Curtis, S.C.: "Operational Advances from Field Application of Short-Radius Horizontal Drilling in the Yates Field Unit," paper SPE 24612 presented at the 1992 SPE Annual Technical Conference and Exhibition, Washington, D.C., Oct. 4-7. 69. Sanders, G.S., Chambers, M.J., and Lane, RH.: "Successful Gas Shutoff with Polymer Gel Using Temperature Modeling and Selective Placement in the Prudhoe Bay Field," paper SPE 28502 presented at the 1994 SPE Annual Technical Conference and Exhibition, New Orleans, Sept. 25-28. 70. The Canadian Heavy Oil Reservoir Handbook, The Canadian Heavy Oil Association, Calgary, Alberta, Canada (1991). 71. "Royalty Incentives Spawn Alberta Oilsands Plans," Oil & Gas J. (Dec. 18, 1995) 30. 72. "Thermal System Ups Heavy Oil Flow; Lighter Crudes Eligible?" Improved Recovery Week (Dec. 4, 1995) 4, No. 47, 1,6. 73. "Suncor Set to Try SAGD at Burnt Lake Heavy Oil Property," Improved Recovery Week (Dec. 11, 1995) 4, No. 48, 6. 74. White, D. And Moss. 1.T.: Thermal Recovery Methods, Pennwell, Tulsa, OK (1983). 75. Prats, M.: Thermal Recovery. Monograph Series, SPE, Richardson, TX (1982). 76. Farouq Ali, S.M. and Meldau, RF., "Current Steamflood Technology," J. Pet. Tech. (Oct. 1979) 1332-1342. 77. Farouq Ali, S.M.: "A Current Appraisal of In Situ Combustion Field Tests," J. Pet. Tech. (April 1972) 477-485. 78. Chu, C.: "A Study of Fireflood Field Projects," J. Pet. Tech. (Feb. 1977) 111-119. 79. Chu, C.: "State-of-the-Art Review of Fireflood Field Projects," J. Pet. Tech. (Jan. 1982) 19-36. 80. Farouq Ali, S.M. "Redeeming Features of In Situ Combustion," paper DOEINIPER 609,470 (ISC 1) presented at the 1994 DOFlNIPER Symposium on In Situ Combustion Practices-Past, Present and Future Application, Tulsa, April 21-22. 81. Sarathi, P.S. and Olsen, D.K.: "DOE Cost-Shared In Situ Combustion Projects Revisited," paper DOEINIPER 609,473 (ISC 4) presented at the 1994 DOEINIPER Symposium on In Situ

82.

83.

84. 85. 86.

87. 88. 89.

90.

91. 92. 93.

94.

95. 96. 97.

98. 99. 100.

Combustion Practices-Past, Present and Future Application, Tulsa, April 21-22. Turta, A.: "In Situ Combustion-from Pilot to Commercial Application," paper DOFlNIPER 609,472 (ISC 3) presented at the 1994 DOEINIPER Symposium on In Situ Combustion Practices-Past, Present and Future Application, Tulsa, OK, April 21-22. Shahani, G.H. and Gunardson, H.H. "Oxygen Enriched Fireflooding," paper DOEINIPER 609,471 (ISC 2) presented at the 1994 DOEINIPER Symposium on In Situ Combustion Practices-Past, Present and Future Application, Tulsa, April 21-22. Petzet, G.A.: "Horizontal Projects Buoy Williston Recovery," Oil & GasJ. (Jan. 15, 1995) 21-22. "Light Oil Air Injection to Revolutionize lOR, Say Experts," Improved Recovery Week (Dec. 25, 1995) 4, No. 50, 1. Kumar, V.K., Fassihi, M.R and Yannimaras, D.V.: "Case History and Appraisal of the Medicine Pole Hills Unit Air Injection Project," paper SPEIDOE 27792 presented at the 1994 SPEIDOE Symposium on Improved Oil Recovery, Tulsa, OK, April 17-20. Duncan, G. Khalbad, A. and Stemler, P.: "Enhanced Recovery Engineering-combustion Processes," World Oil (Jan. 1966) 65. Hong, K.C.: Steamflood Reservoir Management, Penwell, Tulsa, OK (1994). Taber, J.J. and Seright, R.S.: "Horizontal Injection and Production Wells for EOR or Waterflooding," paper SPE 23952 presented at the 1992 SPEIDOE Symposium on Enhanced Oil Recovery, Tulsa, April 22-24. Hanzlik, E.J,: "Steamflooding as an Alternative Eor Process for Light Oil Reservoirs," paper SPE 10319 presented at the 1981 SPE Annual T~chnical Conference and Exhibition, San Antonio, TX, Oct. 5-7. Blevins, T.R, Duerksen, J.H., and Ault, J.W.: "Light-Oil Steamflooding: an Emerging Technology," JPT (July 1984) 1115. Gangle, F.J. et a1.:" Light-Oil Steamdrive Pilot Test at NPR-l, Elk Hills, California," SPERE (Aug. 1992) 315-320. Olsen, D.K. et al.: "Light Oil Steamflooding: A Laboratory Study of Oil Recovery from Oil-Wet and Water-Wet Porous Media from a 2-D Physical Model," paper SPE 23692 presented at the 1992 SPE Latin American Petroleum Engineering Conference, II LAPEC, Caracas, Venezuela, March 8-11. Palmgren, C and Renard G.: "Screening Criteria for the Application of Steam Injection and Horizontal Wells," Proc. 8th EAPG Impr. Oil Recovery Europe Symp., Vienna (1995) 2,256-264. "Syncrude Production Keeps Climbing," Oil & Gas J. (Jan. 15, 1996) 23. "Syncrude Operation Claims Production Mark from Oil Sands," Oil & Gas J. (Oct. 9, 1995) 46. "Worldwide production falls as market plays its wild cards," Oil & Gas J. Worldwide Production Report: OGJ Special (Dec. 27, 1993) 37-87. "It's Official: First Glimmer of Greenhouse Warming Seen," Science (Dec. 8, 1995) 270,1565-1567. "1995 the Wannest Year? Yes and No," Science, (Jan. 12, 1996) 271, 137, 138. Ribeiro, 1. and Henry, B.: "Carbon Dioxide Disposal and Storage Technologies: Part I: Current Status; Part II: Prospective Study towards the Year 2020 and Beyond," Institute for Prospective Technological Studies, European Commission Joint Research Centre, Ispra, Italy (Jan. 1995).

397

12

J.J. TABER. F.D. MARTIN. AND R.S. SERIGHT

SPE 35385

Metric Conversion Factors acre acre-ft °API abn bbl cp cu in cu ft of

x x x x x x x x x

4.046873 1.233489 1.013250* 1.589873 1.0* 1.638706 2.831685 CF - 32)/1.8

E+03 E+03 E + 05 E - 01 E-03 E+Ol E-02 E - 01 E+OO E+OO E - 01 E+OO E+OO E-02 E+OO E-02 E - 01

m2 m3 g/cm' Pa . m'

141.5/(131.5 + °API)

Pa-s em'
m3 °C m cm N kg

3.048* ft 2.54* in x Ibf x 4.448222 x 4.535924 Ibm x 1.609344* mile x 6.894757 psi x 2.863640 scf 6.4516* sq in x x 9.290304* sq ft x 9.144* yard *conversion factor is exact.

km
kPa std cm2 m2 m '"

Additional Conversion Factors Useful for Reading COl Flooding Literature

(Some factors are rounded for convenience and quick estimates) 1 barrel (bbl)

= 42

U.S. gallons

= 0.159

m'

1 cubic foot = 0.0283 m' 1000

ftl (mscf or mcf or Mcf or MCF)

= 28.3 m'

Standard conditions in U.S. oil industry = 1 atm and 60°F (1.0l3 bar, 14.7 psia) (may vary in some U.S. states) COl density at standard conditions kg/m' = 0.1166 lb/ff 17,150 lbs)
=

0.001868 glcc or 1.87

ftl of CO2 at 60°F (1 atm)

(weighs) 1 ton U.S. (2,000

1 ton U.S.

= 2,000
=

lb.

= 907

kg (1 kg '" 2.2 lbs)

1 ton U.S. = 0.907 metric ton or tonne 1 tonne COl 18,904 scf@ 15.6° C and 1 atm.

1 Gt (gigatonne) = 1 billion metric tons 1 barrel oil (35° API) = 0.16982 ton U.S. = 0.16895 tonne 1 mscflbbl = 0.31324 tonne CO/tonne oil (35° API)

(Crude oil density typically ranges from 0.8 - 0.95 g/cm' or 800 - 950 kg/m') CO2 utilization factor used for Table 12 tonne of CO/tonne of incremental oil.
=

6 mcflbbl

1.879

398

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

13

Table I. Current and Past Enhanced Recovery Methods

Gas (and Hydrocarbon Solvent) Methods

"inert" gas injection nitrogen injection flue gas injection hydrocarbon gas (and liquid) injection including: high pressure gas drive enriched gas drive miscible solvent (LPG or propane) flooding carbon dioxide flooding

Detailed . Table Number

4

(4) 5

Improved Waterflooding Methods

alcohol miscible solvent flooding micellar/polymer alkaline flooding alkaline/surfactant/polymer polymer flooding gels for water shutoff microbial injection (ASP) flooding (surfactant) flooding

7
7 7

low interfacial waterflooding

8
9

Thermal Methods
in situ combustion standard forward combustion wet combustion 02-enriched combustion reverse combustion steam and hot water injection hot water flooding steam stimulation steamflooding II 10

Surface Mining and Extraction

399

14

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

SPE 35385

Table 2.

Classification of Current Enhanced Recovery Methods (Excluding Gel Treatments) by the Main Mechanism of Oil Displacement

Solvent Extraction and/or "Miscible-Type" Processes

nitrogen and flue gas hydrocarbon miscible methods carbon dioxide flooding "solvent" extraction of mined, oil-bearing ore

Interracial Tension Reduction Processes

micellar/polymer flooding (sometimes included in miscible-type flooding above) alkaline/surfactant/polymer flooding

Viscosity Reduction (Or Oil) or Viscosity Increase (Or Driving Fluid) plus Pressure
steamflooding fireflooding polymer flooding enhanced gravity drainage by gas or steam injection

400

SPE35385

EOR SCREENING

CRITERIA

REVISITED

15

e-

G) f-

S-

8.~

cj

:i

cj

:i

:i
oG) 0-

cj

:i

cj

~I
0 0

~I 1
0 0

k Q.~
.c:: U .!=

0 0

'0

o.

0 0

"

"='"

o.

"

M'G)f "
ani

'0

\}

\}

0 0

~I ,

0 0

<II <II ·c ~ y

fn=

i: G) <II

> <G) Q., <II <II G) .fa y.-ct:: l5

ei]' G)G)

e-

.. cj

<II
cj

"
0 0 00.

0 0 0

~~ 1

0 0 0

M
0 0

an
1

"
V

cj

:i

:i 3
=.9 G)-

cj

:i
<II
~QI)

cj

:i
<II
~QI)

cj

:i

:i
QI) •

~I ,

~z
·i~ 0
= 0

- ...
2i .Cj
.It!
WI

1·1 1-[
~~ ~~

·ll·e j .So 8.
"0 QI) _ ...

"
o

'0

~I ,

"

"l.

0 0

an

0 0

"l.

~-e"O

":":]E!

if
on

~~

"

0 an

8
~

cj

:i

""
I'll

ef-

= .9> :::~Q., o cue.... :f!.

.;;;

:E
a .9 a
WI

,1 , , ._ , , , an .. an I'll I'll I'll

! 3.8 '8 ~a ~~ iu iu

OG) G)tU ==

OG) G)tU G)tU == 30 ==

... OG)

...
~ ~ ~
0

-8

~ii Z

-ss !
II a
u

cj

:i

cj

:i
G)"O =G)

>

Ecj

!
II DII

"
0

t G)
0 N "0

"
0

"

G) G)

j~

e
'CI

:i

~[
~

~c2

8] 3t:

~c2

3t:

] e~ a
<11-

~8 .- 0

~[
~

= .9

8. S 0 U
<II .~ t:

~I
f"'I

I::

"='"

II

eo..

o o ... ... .. o cj ~~ ~~

""
eo..

f"'I

eo..

ii,- ii,N ·-U ·-U ::c ::c

~y ii, ...
!:EU
"1
0

"

II

f"'I

,
"

i
II

.a

II

8.~ .- i ::C<ll

ii,=tJ ii,~

~8 .- 0 19 e~ .!!"o 8.i ii G)<II

<11-

.5
s

!:E~

"'! "
u

1
.a a

:i

r;.iI

II

ii,] G) "'~
;;J

..r.!

.B,~G)

"

e-

~ .a ~ E- an

~"O
~

.5

e §.~ "'= ~l'Ills

~ 1 ~

~·u:Sl

" y~
0·';::

o
u

~~

t
8

:f!.eo.. &,0

"

S- = cj 0!8. I'Il:aS :i

cj

:i
~

e Q.,

8.

.U;~

:>
0

Q.,

~Oy

-e

~ .s:

-g
~
"0

e C

"ti

::E
0
UJ

.-.D

.!! _g

~~z Q

0•

,~I ~I , ,'f- , 3--- " " " " " ..:-a -.:. ~-I ;§~ ~ l·5!S;;.5 l·5 ....a i :i ~c -1 J s g.a S·~]C ]o;;t t z"" i. -I
~ ~ ~ 1
f"'I

~I 1

V
'0

1 1
0 0

an

"
V
0

0 an

0 0

!
V

vi

0 0

= 0 y

:s! ~

\}

CI'I

C'o

e-

'0-

~G)

an
f"'I

.!!

f"'I

~ ~B "I'll

~ an

I'll

G) G)

oJ,

r.!.

_..::

=:a

=-8

.It!

~ ~

yG) :a

.,!..

00.5

Q.,G)

QI)

QI)

...

QI)

[<1 ::EQ.,<10-1""
r--

oQ.,,,"

.D

U
0

::E G)
I'll

"='"

an

'0

00

CI'I

401

16

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

--_._-------

SPE 35385

Table 4. Nitrogen and Flue Gas Flooding Description Nitrogen and flue gas are oil recovery methods which use these inexpensive non-hydrocarbon gases to displace oil in systems that may be either miscible or immiscible depending on the pressure and oil composition (see Table 3 for immiscible criteria). Because of their low cost, large volumes of these gases may be injected. Nitrogen or flue gas are also considered for use as chase gases in hydrocarbon-miscible and CO2 floods. Mechanisms Nitrogen and flue gas flooding recovery oil by: -vaporizing the lighter components of the crude oil and generating miscibility if the pressure is high enough -providing a gas drive where a significant portion of the reservoir volume is filled with low-cost gases --enhancing gravity drainage in dipping reservoirs (miscible or immiscible). TECHNICAL SCREENING GUIDES Crude Oil Gravity Viscosity Composition Reservoir Oil Saturation Type of Formation Net Thickness Average Permeability Depth Temperature Limitations Developed miscibility can only be achieved with light oils and at very high pressures; therefore, deep reservoirs are needed. A steeply dipping reservoir is desired to permit gravity stabilization of the displacement which has an unfavorable mobility ratio. For miscible, or immiscible enhanced gravity drainage, a dipping reservoir may be crucial to the success of the project. Problems Viscous fingering results in poor vertical and horizontal sweep efficiency. The non-hydrocarbon gases must be separated from the saleable produced gas. Injection of flue gas has caused corrosion problems in the past. At present, nitrogen is being injected into large successful projects that formerly used flue gas. >40%PV Sandstone or carbonate with few fractures and high permeability streaks Relatively thin unless formation is dipping Not critical >6000 ft Not critical 10,000-18,500 ft 59-80% VP <0.4 cp High percentage of light hydrocarbons Recommended Range of Current Projects 38-54
0

API (miscible)

0.07-0.3

402

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

17

Table 5. Hydrocarbon Miscible Flooding Description Hydrocarbon miscible flooding consists of injecting light hydrocarbons through the reservoir to form a miscible flood. Three different methods have been used. The first-contact miscible method used about 5% PV slug of liquefied petroleum gas (LPG) such as propane, followed by natural gas or gas and water. A second method, called Enriched (Condensing) Gas Drive, consists of injecting a 10-20% PV slug of natural gas that is enriched with ethane through hexane (C2 to C6), followed by lean gas (dry, mostly methane) and possibly water. The enriching components are transferred from the gas to the oil. The third, and most common method, called High Pressure (Vaporizing) Gas Drive, consists of injecting lean gas at high pressure to vaporize C2--C6 components from the crude oil being displaced. Immiscible criteria are given in Table 3. Mechanisms Hydrocarbon -generating -increasing -decreasing -immiscible miscible flooding recovers crude oil by: miscibility (in the condensing and vaporizing gas drive). the oil volume (swelling). the viscosity of the oil. gas displacement, especially enhanced gravity drainage with the right reservoir conditions.

Crude Oil Gravity Viscosity Composition Reservoir Oil Saturation Type of Formation

TECHNICAL SCREENING GUIDES Range of Current Projects Recommended >23°API <3 cp High percentage of light hydrocarbons >30%PV Sandstone or carbonate with a minimum of fractures and high permeability streaks Relatively thin unless formation is dipping Not critical if uniform >4000 ft Not critical 4,040-15,900 ft 30-98% VP 24-54 °API (miscible) 0.04-2.3 cp

Net Thickness Average Permeability Depth Temperature Limitations

The minimum depth is set by the pressure needed to maintain the generated miscibility. The required pressure ranges from about 1200 psi for the LPG process to 4,000-5,000 psi for the High Pressure Gas Drive, depending on the oil. A steeply dipping formation is very desirable to permit some gravity stabilization of the displacement which normally has an unfavorable mobility ratio. Problems Viscous fingering results in poor vertical and horizontal sweep efficiency. Large quantities of expensive products are required. Solvent may be trapped and not recovered in the LPG method.

403

18

--_----------

.._--_ .._--

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

----- .._----_._-----

- -----------

---_--- ------_

SPE 35385

Table 6. Carbon Dioxide Flooding Description Carbon dioxide flooding is carried out by injecting large quantities of CO2 (30% or more of the hydrocarbon PV) into the reservoir. Although CO2 is not first-contact miscible with the crude oil, the CO2 extracts the light-to-intermediate components from the oil, and, if the pressure is high enough, develops miscibility to displace the crude oil from the reservoir. Immiscible displacements are less effective, but they recover oil better than waterflooding (see below and Table 3 for immiscible criteria).
\

Mechanisms CO2 recovers crude oil by: -swelling the crude oil (C02 is very soluble in high-gravity oils). -lowering the viscosity of the oil (much more effectively than N2 or CH4). -lowering the interfacial tension between the oil and the CO2 -oil phase in the near-miscible regions. -generation of miscibility when pressure is high enough (see below). Depth vs Oil Gravity Screening Criteria I. For COz-Miscible Oil Gravity °API >40° 32-39,9° 28-31.9° 22-27.9° Flooding: Depth must be greater Feet 2,500 2,800 3,300 4,000 than: '. for C02 Flooding

Meters 762 853 1,006 1,219

<2ZO

Fails miscible, screen for immiscible

2. For Immiscible COz Flooding (Lower Oil Recovery) 13-21.9° <13° 1,800 549

All oil reservoirs fail at any depth

<1800 feet (549 meters}-all reservoirs fail screening criteria for either miscible or immiscible flooding with supercritical CO2_

(The arbitrary criteria have been selected to provide a safety margin of approximately 500 feet (152 meters) above typical reservoir fracture depth for the required miscibility (MMP) pressures, and about 20 bar above the CO2 critical pressure for the immiscible floods at the shallow depths. Reservoir temperature is assumed from depth. See text.) -See Table 3 for other screening criteria.

404

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

19

Table 7. Micellar/polymer, Alkaline Surfactant Polymer (ASP) and Alkaline Flooding Description Classical Micellar/Polymer Flooding consists of injecting a slug that contains water, surfactant, electrolyte (salt), usually a cosolvent (alcohol), and possibly a hydrocarbon (oil). The size of the slug is often 5-15% PV for a high surfactant concentration system and 15-50% PV for low concentrations. The surfactant slug is followed by polymer-thickened water. The polymer concentration often ranges from 500-2000 mg/L, and the volume of polymer solution injected may be 50% PV or more. Alkaline/Surfactant/Polymer (ASP) Flooding is similar except that much of the surfactant is replaced by low cost alkali so the slugs can \ be much larger but the overall cost is lower and polymer is usually incorporated in the larger, dilute slug. For Alkaline Flooding much of the injection water was treated with low concentrations of the alkaline agent and the surfactants were generated February 1, 1996 in situ by interaction with oil and rock. Mechanisms All surfactant and alkaline flooding methods recover oil by: -lowering the interfacial tension between oil and water. -solubilization of oil in some micellar systems. -emulsification of oil and water, especially in the alkaline methods. -wettability alteration (in the alkaline methods). -mobility enhancement. TECHNICAL SCREENING GUIDES Crude Oil Gravity Viscosity Composition Reservoir Oil Saturation Type of Formation Net Thickness Average Permeability Depth Temperature Limitations An areal sweep of more than 50% on waterflood is desired. Relatively homogeneous formation is preferred. High amounts of anhydrite, gypsum, or clays are undesirable. Available systems provide optimum behavior over a narrow set of conditions. With commercially available surfactants, formation water chlorides should be <20,000 ppm and divalent ions (Ca ++ and Mg") <500 ppm. Problems Complex and expensive systems. Possibility of chromatographic separation of chemicals in reservoir. High adsorption of surfactant. Interactions between surfactant and polymer. Degradation of chemicals at high temperature. >35%PV Sandstones preferred Not critical >10md <about 9000 ft (see Temperature) <200°F Recommended >20° API <35 cp Light intermediates are desirable for micellar/polymer. Organic acids needed to achieve lower interfacial tensions with alkaline methods.

405

20

J.J. TABER. F.D. MARTIN. AND R.S. SERIGHT

SPE 35385

Table 8. Polymer Flooding Description The objective of polymer flooding is to provide better displacement and volumetric sweep efficiencies during a waterflood. In polymer flooding, certain high-molecular-weight polymers (typically polyacrylamide or xanthan) are dissolved in the injection water to decrease water mobility. Polymer concentrations from 250 to 2000 mgIL are used; properly sized treatments may require 25 to 60% reservoir PV. TECHNICAL SCREENING GUIDES A very broad screening guide is given in Table 3. When identifying polymer-flooding candidates, we recommend that the reservoir characteristics and polymer-flood features be closer to those of the four successful projects listed below. P roperties Property Oil/water viscosity ratio at reservoir temperature Reservoir temperature, OF Permeability, darcys % OOIP present at startup WOR at startup HPAM concentration, ppm Ibs polymer/ac-ft Projected lOR, % OOIP Projected bbl oil/lb-polymer
0fP

oiymer FI00dimg F· Id P ro .ec ts te I Marmul 114 115 15 -92 1 1,000 373 25* 1.2 Oerrel 39 136 2 81.5 4 1,500 162 -13 -1.4 Courtenay 50 86 2 78 8 900 520 30 0.96 Daqing 15 113 0.87 71 10 1,000 271
II

1980s median (171 projects) 9.4 120 0.075 76 3 460 25 4.9 1.1

0.57 155

-230. Projected bbl oil/ac-ft 27 461 499 . incremental over waterflooding . . *IOR over pnmary production for this case only. For the others, lOR IS

406

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

21

Table 9. Gel Treatments Description Gel treatments are applied most effectively when fractures cause severe channeling or excess water or gas production. Gel treatments are also useful as an alternative to cement squeezes when plugging narrow channels behind pipe. They can also be effective in blocking water zones in unfractured wells where impermeable barriers prevent crossflow, but only if hydrocarbon-productive zones are protected during gel placement. Treatments can be effective in both injection and production wells. A. Selection Criteria for Gel Treatments in Injection Wells 1. Reservoir and production data indicates low sweep efficiency during waterflooding. a. Water breakthrough occurs much earlier than expected (i.e., from standard calculations or simulations or from comparison with the performance of other patterns in the field). b. WOR values at offset producers are much higher than expected. c. Recovery calculations indicate that considerable mobile oil remains that could be recovered more cost-effectively if a blocking agent could be realistically placed in the proper location. 2. If barriers to crossflow do not exist, then interwell tracers must show very rapid transit times (probably indicating that fractures or formation parting cause the channeling problem). 3. In un fractured wells, a. Poor injection profiles must be correlatable from well to well. b. Effective barriers to crossflow must exist (very low kjkh' no flow behind pipe, no vertical fractures). c. Gel can be placed in the offending channel without damaging oil zones (e.g., using zone isolation). 4. Reduced injectivity (caused by the gel) can be tolerated. 5. The well to be treated is in good mechanical condition.

B. Selection Criteria for Gel Treatments in Production Wells 1. 2. 3. 4. 5. Recovery calculations indicate that considerable mobile oil remains that could be recovered more cost-effectively if a blocking agent could be realistically placed in the proper location. High WOR values are observed. The source of the excess water production is identified (e.g., using profiles, logs, or tracers). The candidate well exhibits high productivity. The gelant can be placed without damaging oil zones (e.g., using zone isolation).

407

22

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

SPE 35385

Table 10. In-Situ Combustion Description In-situ combustion or fireflooding involves starting a fire in the reservoir and injecting air to sustain the burning of some of the crude oil. The most common techniques is forward combustion in which the reservoir is ignited in an injection well, and air is injected to propagate the combustion front away from the well. One of the variations of this technique is a combination of'forward combustion and waterflooding (COFCA W). A second technique is reverse combustion in which a fire is started in a well that will eventually become a producing well, and air injection is then switched to adjacent wells; however, no successful field trials have been completed for reverse combustion. Mechanisms In-situ combustion recovers crude oil by: -the application of heat which is transferred downstream by conduction and convection, thus lowering the viscosity of the crude -the products of steam distillation and thermal crackling which are carried forward to mix with and upgrade the crude -burning coke that is produced from the heavy ends of the crude oil -the pressure supplied to the reservoir by injected air TECHNICAL SCREENING GUIDES Crude Oil Gravity Viscosity Composition Reservoir Oil Saturation Type of Formation Net Thickness Average Permeability Depth Temperature Limitations If sufficient coke is not deposited from the oil being burned, the combustion process will not be sustained. This prevents the application for high-gravity, paraffinic oils. If excessive coke is deposited, the rate of advance of the combustion zone will be slow, and the quantity of air required to sustain combustion will be high. Oil saturation and porosity must be high to minimize heat loss to rock. Process tends to sweep through upper part of reservoir so that sweep efficiency is poor in thick formation. Problems Adverse mobility ratio. Complex process, requiring large capital investment, is difficult to control. Produced flue gases can present environmental problems. Operational problems such as severe corrosion caused by low pH hot water, serious oil-water emulsions, increased sand production, deposition of carbon or wax, and pipe failures in the producing wells as a result of the very high temperatures. >50%PV Sand or sandstone with high porosity >10 ft >50md >11,500 ft >1000 F 85-4,000md 400--11,300 ft 100--221 of 62-94 Recommended 10--27 API
0

Range of Current Projects 10-400 API (miscible) 6-5,000 cp

<500 cp Some asphaltic components to aid coke deposition

408

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

23

Table 11. Steamflooding Description The steam drive process or steamflooding involves the continuous injection of about 80% quality steam to displace crude oil towards producing wells. Normal practice is to precede and accompany the steam drive by a cyclic steam stimulation of the producing wells (called huff and puff). Mechanisms Steam recovers crude oil by: -heating the crude oil and reducing its viscosity -supplying the pressure to drive oil to the producing well TECHNICAL SCREENING GUIDES Crude Oil Gravity Viscosity Composition Reservoir Oil Saturation Type of Formation Net Thickness Average Permeability Transmissibility Depth Temperature Limitations Oil saturations must be quite high and the pay zone should be more than 20 feet thick to minimize heat losses to adjacent
formations.

Recommended 10-25° API <50,000 cp Not critical but some light ends for steam distillation will help >40% V Sand or sandstone with high porosity and permeability preferred >20 ft >200 md (See Transmissibility) >100 md ftlcp <5,000 ft not critical

Range of Current Projects 8-27°API 10-20,000 cp

35-90% VP

63-10,000 md 150-4,500 ft 60-280 of

Lighter, less viscous crude oils can be steamflooded but normally will not be if the reservoir will respond to an ordinary waterflood. Steamflooding is primarily applicable to viscous oils in massive, high-permeability sandstones or unconsolidated sands. Because of excess heat losses in the wellbore, steamflooded reservoirs should be as shallow as possible as long as pressure for sufficient injection rates can be maintained. Steamflooding is not normally used in carbonate reservoirs. Since about one-third of the additional oil recovered is consumed to generate the required steam, the cost per incremental barrel of oil is high. A low percentage of water-sensitive clays is desired for good injectivity.

409

24

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

SPE 35385

Table 12. Summary of Estimated Worldwide CO2 DemandlUtilization Potential Oil Recovery

and

Oil-Producing gion

Re-

Potential Oil Production by CO2 Injection (from Tables in Ref. 33)

billions of barrels Middle East Western Hemisphere Africa Eastern Europe and CIS Asia-Pacific Western Europe World Totals 141.04 28.78 13.18 10.85 8.59 3.52 205.96

Gtonnes 23.84 4.86 2.23 1.83 1.45 0.59 34.80

Total CO2 Required to Produce Incremental Oil (from Tables in Ref. 33) (G tonnes) 44.80 9.14 4.19 3.45 2.73 1.12 65.43

Urgency, Timing, or Regional Adjustment (%)

Potential CO2 Utilisation (G tonnes)

-12 +10 -5 0 -5 +15 _~-7]·

39.42 10.05 3.98 3.45 2.59 1.29 60.78

• Net reduction in world total, %

-From
r»

Ref. 33

Table 13. Profitability ofEOR Projects in the U.S.

Percent Reported Method Steam Combustion Hot Water CO2 Hydrocarbon Nitrogen Flue Gas Polymer MicellarlPolymer Alkaline or Alkaline/Surfactant as Profitable

1982
86 65

1988
95 78 89 66 100 100 100 92 0

1990
96 88 78 81 100 100 100 86 0

1994
96 80 100 81 100 100

--

21 50 100 100 72 0 40

-0

100 100

100 (1 success)

Table updated from Ref. 23 and 24.

410

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

25

)
12

r----------------,

10

EORas%
of U.S. Production 8

2~~--~~~~~~--~~~~ 1110 , .. 2 1114 1114

1111

,_

'"2

1114

lHe

200.000

1110

, .. 2

1114

1111

1111

,_

'"2

1114

lHe

Fig. I-EOR

production in the U.S. (data from Ref. 24).

10
I

15

Oil Gravity GAPI 20 25 30 35 40 I

-<::]mmiscible Gas -_-_-_-_-_-_-_-_-_-_-_-_-.

<

_--=:::=========::::::: - _ CO2 - Miscf~-f~-_-_-_-_- _

& Micellar

yd rocarbon-----~~~~
Pi~r_- _- -_-:': .: _-: _ _. . ------------- ... _.. --------

< N, & Flue gas .-_-_-_-_-_-_~

45

50
I

55

60
I

c::::::::::: Polymer Floods

C Gel Treatments

--

A1kalinelSurfaclanllPo!ymer

_ ... ----_

Fig. 2-API gravity range of oil that is most effective for EOR methods. Relative production (bid) is shown by size of type. 411

26 0

J.J. TABER, F.D. MARTIN, AND A.S. SERIGHT

SPE 3531

5,000

•• •

• •
IQ)

~
0

~0
0
0

Steamftoodlng

0
0

8§P
cPo
0

Hat Water

+
Of!]
§
0
[J]

00

o 00000
0 0

Combultkln

o
0
[J:>

CO2

•
0

0"",

t
Q
II

10,000

"'"

+
0

HC Miscible & IlI1IT1iacjble

"'"

"'" ~

Nitrogen Miscible & Immiscible

"'"

+ +
Polymer

15,000

"'"
+
20,000 0 10 20 30 40 50 60

0
MlcellarPolymer

X
AIIiaIlnei Surfactant

·API Fig. 3-Depth 24.)

and API gravity of producing EOR projects in the U.S. (data from Ref.

o ~--------------------------------------,~--------~
o

Steamflooding Hot Water Cbmbustion C02

5,000 Io
o

•
•
o

g
Q
II

t 10,000

l-

HC Miscible
& Immiscible

15,000 r-

Polymer

"'"

o
Mlcenar-Polymer
X

20,000 .__ ......... _

_j....._--'_ ..... I.___......... _...I..I--I._-,I,--_.__-,--

1--,_-,

Alkaline!

10

20

30

40

50

60

Surfactant

API Fig. 4-Depth Ref. 24.) and API gravity of producing EOR projects outside the

•
(data from

u.S.

412

~PE 35385 100,000

EOR SCREENING

CRITERIA

REVISITED

27

80,000

~ I: ~ :::II
"CI

60,000

l
~ 40,000 w

20,000

Stum

~ ,*,.,OD Combustion Polymer & MlcellllrIPoIymer

Hydrocarbon

.41

JAil ..

Nitrogen

011Gravity [ • API] Fig. 5-EOR production vs oil gravity in the U.S. (data from Ref. 24).

100,000

.__
80,000

245,000

~ :I
D-

s 60,000

1
Ill: 40,000

fa

20,000

Hydrocarbon

011 Gravity [ • API] Fig. 6--EOR production vs oil gravity outside the U.S. (data from Ref. 24). 413

28

J.J. TABER, F.D. MARTIN, AND R.S. SERIGHT

SPE 3538(

II

.ooo_--r-....,..-.,..--r-,_,..-""T""-"r"--r-~-.,.........,r-...,

10.000

..
a.

a:
en en
Q.

IIJ

:::l IIJ

a:

1000

2··3

9'

10

II

12

DEPTH IN THOUSANDS

OF FEET

Fig. 7-Increase in MMP and fracture pressure with depth for Permian Basin reservoirs (from Ref. 32). " 4,000 ~----------------------------------------------------------~

i;'
~

~
~ 3,000

1976 Constant Dollars

'0
(/)

,/

't:J C

o ..c ~

C'G

2,000

Actual EOR Production

,/ ,/

I::.
0:: 1,000
C)

z-:
/'

,/ ,/ ,/

/'
,/

"/

........

- ..

_ $151 Barrel_ _ _ ---._ _

.. .. ..

..

.. -_._-

'/ 17

'0

~ n,

::l 't:J

~
1980

~ ~

~ ........... -- ~ ~

.....

;",,;;;

.....-..----------:...............
-

$10 I Barrel

--- --.

...........

$ 51 Barrel

o ~----~----~--~----~----~--~----~----~--~----~
1982 1984 1986 1988

1990

1992

1994

1996

1998

2000

Fig. 8-Actual 7 and 24.)

U.S. EOR production vs 1976 NPC predictions. (Extended from Ref. 33, data from Refs.

414

SPE 35385

EOR SCREENING

CRITERIA

REVISITED

29

2,000

C
Q)

...
0.
UJ

m '0
UJ UJ

1,500

1983 Constant Dollars 10% ROR

/

$5Q_/.~a~~\-. __ __ . _.,
"

_,._,.-

/"'

(IJ

,.......,
.,/"'

'U

C
(IJ

1,000

.s
s: .,_ 0:::
(IJ

:J 0

/ /'./ /" ---~ $30 I Barrel Actual EOR Produgti"op/ /" /" /' /_,.-/" .' _,.-_,..'

-_ $401 Barrel -" _--_ -.

.. -

-_
..

.:

,,/7.., .
500

s> .'

.~/

.--/

.'_"-

' .. $~O_lBarrel
- ....

:J 'U

1S

a..

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Fig. 9-Sensitivity of U.S. EOR production to the crude price predicted in 1984 NPC report. (Extended from Ref. 33, data from Refs. 8 and 24.)

C
UJ
(IJ

1,000

>-

m '0
'U
(IJ

(IJ

800

1983 Constant Dollars 10% ROR

/

___ _ _ - $SO/Ba rrel

--$40 I Barrel
/ / / / / /
/ / / / / /

UJ

c
UJ

600

:J 0

Actual CO2 Production

400

.s
(IJ

- .-

.. _--- ..

$30 I Barrel

----.'-

..

..

_._

0:::

/ / /

1S :J

200

e a..
'U

~<~:~:~'_.. .. .. :

--- ---

-----

$ 20/ Barrel
..

_.

-----,

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Fig. 100Sensitivity of U.S. CO2 production to crude oil price predicted by 1984 NPC report. (Extended from Ref. 33, data from Refs. 8 and 24.)

415

CHAPTER II CHEMICAL RECOVERY PROCESSES

Contents 1. 2. Introduction Mobility Control Processes 2.1 2.2 3. Polymer Flooding Foam Flooding

Low-1FT Processes 3.1 3.2 Surfactant Flooding Alkaline Flooding

4.

Conclusions

1.

Introduction

Chemical flooding process is to recover oil by (1) reducing the mobility of displacing agent (mobility control process), and/or (2) lowering the oil/water interfacial tension (lOW-1FT process). This chapter is divided into two sections corresponding to above two processes.

2.

Mobili~yControl Processes

The mobility is defined as the ratio of permeability to the viscosity and the mobility ratio is defined as the mobility of displacing phase (water) to the mobility of the displaced phase (oil).

Mp = k, /

IIp

Mw-o Mw/ Mo = Mobility control processes inject a low-mobility displacing agent to increase volumetric and displacement sweep efficiency. This process includes polymer flooding and foam flooding.

2.1

Polymer Flooding

Polymer flooding is a process that a small amount of polymer is added to thicken brine (water) to reduce the water mobility. Under current technology, most light-oil reservoirs are waterflooded unless a strong natural water drive or an effective gas cap drive is already taking place. The oil recovery that can be expected by water flooding a given reservoir is controlled by two factors - volumetric coverage efficiency and the residual oil saturation. Generally, at least one-half of the original oil remains after water flooding. Under some conditions, the volumetric coverage efficiency can be improved by "thickening" the injected water to improve the mobility ratio. The mobility ratio of a waterflood can be reduced by adding a chemical such as polyacrylamide or XC- biopolymer to the injected water. The effect of mobility ratio on vertical invasion efficiency and on pattern coverage efficiency was covered in the Waterflooding section. decreases. Both of these efficiencies increase as the mobility ratio The main incentive for polymer is in flooding medium viscosity oil As a result, the most attractive oil

reservoirs. The mobility ratio of a waterflood can be effectively reduced by about a factor of 10 by adding polymer to the water. reservoirs for polymer flooding will fall in the 20-200 cp viscosity range. As we will see later, water banks thickened with polymer are used to provide mobility control in pushing miscible orsurfactant banks in tertiary recovery projects.

2.1.1 Types 'of Polymer

Several polymers have been considered for polymer flooding: xanthan gum (XCbiopolymer), hydrolyzed polymer (HPAM), copolymers of acrylic acid and acryamide, copolymers of acrylamide and 2-acrylamide-2-methyl propane sulfant (AM/AMPS), hydroxyethyl cellulose (HEC), carboxymethyhydroxyethylcellulose (CMHEC), polyacrylamide (PAM), polyacrylic acide, glucan, dextran polyethyleneoxide (PEa), and polyvinyl alcohol. Only two basic types of polymer - XC-biopolymer and partially hydrolyzed

polyacrylamides are widely used in field recovery projects. XC-biopolymer is a natural polysaccharide produced by a microbial fermentation process. Polyacrylamides are synthetic chemical products. XC-biopolymer reduces the mobility ratio simply by increasing water viscosity. However, polyacrylamides reduce water mobility mainly by

reducing the formation permeability to water. The effect of each type of polymer is illustrated in Figure 1. Figure 1A shows the viscosity increase as a function of polymer concentration. For a given concentration, XC-biopolymer results in a much greater increase in viscosity. However, as shown in Figure 1B, for a concentration of 600 ppm, the XC-biopolymer reduces water mobility in a core much less than does the polyacrylamide. Obviously, the polyacrylamide reduces the permeability of the core to water. This permeability reduction is relatively permanent and is not removed by a brine flush, as illustrated in Figure 1B. By contrast, when the XC-biopolymer is followed by a brine flush, the original permeability is soon restored.

2.1.2 Polymer Properties

Because of the complexity of the subject, a comprehensive study of polymer properties is not possible. a. b. c. d. a. However, qualitative trends, a few quantitative relations and representative data are presented on these properties: viscosity relations and non-Newtonian effects, retention, permeability reduction, and chemical, biological, or mechanical degradation. Viscosity Relations and Non-NewtonianEffects

A fluid whose viscosity decreases with increasing shear rate (e) is shear thinning. The behavior of the polymer solution is caused by the uncoiling and unsnagging of polymer chains when they are placed in a shear flow. The polymer solution viscosity 1 shear rate relationship may be described by a power law model. lip = K (e) n-1 where K and n are the power law coefficient and exponent, respectively.

Another useful relationship is Meter model. lip where When

= Ilpoo + (Ilpo

-lipOO)

I (1 + (e I e1l2t-1)

e112=

1/2 x (Ilpo + Ilpoo)

applied to permeable

media flow these general trends and equation

continue to apply is usually called "apparent" viscosity and the effective shear rate is based on capillary tube concepts.

where

Uw

= superficial flux of polymer-rich (water) phase

kw

= permeability
= porosity

to polymer-rich phase

o
b.

Polymer Retention

Polymer retention is due to solid surface adoption or mechanical trapping within porous media. Field measurement values of retention range from 20 to 400 Ibm polymer / acreRetention

ft bulk volume with desirable retention level being less than 50 Ibm / acre-ft.

causes the loss of polymer from solution, which can cause mobility control effect to be destroyed. Retention also causes a delay in the rate of polymer propagation.

c.

Permeability Reduction

Polymer retention reduces the apparent permeability of the rock. Permeability reduction depends on the type of polymer, the amount of polymer retained, the pore size

distribution,

and the average size of the polymer relative to the pores in the rock. The to water caused by the polymer is a big advantage since a

reduction in permeability

relatively small polymer bank can be followed by brine without losing mobility control. Mobility control is required for processes using miscible or surfactant banks. Because

these banks remove the residual oil, the permeability to water behind the bank is greatly increased. The bank must be driven by thickened water to compensate for this relative

permeability increase.

This phenomenon

is measured by the following factors: the resistance factor (FR), and

residual resistance factor (FRr).

FR FRr
where

= (kw / f.lw ) / (kp

=
Awb /

/ f.lp )

Awa

AWb

= =

the mobility of water-rich phase before polymer injection. the mobility of water-rich phase after polymer injection

Awa
d.

Chemical, biological and mechanical degradation

The average polymer molecular weight can be decreased to the detrimental of the overall process by chemical, biological, and mechanical degradation. Chemical degradation ca be minimized by restricting polymer usage to low temperature reservoirs and adding oxygen scavengers (e.g., sodium sulfate or sodium sulfite) to the polymer solution. Biological degradation can be eliminated by adding oxygen scavengers and biocides (e.g., formaldehyde or isopropyl alcohol) Mechanical degradation occurs when polymer solution are exposed to high velocity flow, surface equipments (valves, orifices, pumps, or tubing) and downhole conditions (perforations or screens).

2.1.3

Problems with Polymers

Field tests have exposed severe operational problems with both types of polymer. Injectivity problems were encountered with XC-biopolymer water in most field pilots. A combination of water quality and polymer quality caused these injectivity problems. Water quality must be much higher for a polymer test than for ordinary water injection. It was found that the Kelco biopolymer used in these projects contained about 11 weight percent cellular debris which caused much of the plugging. Biopolymer marketed by Pfizer has proven to be much better from the injectivity standpoint. The major operational problem with polyacrylamide is that the material can be mechanically degraded when subjected to high shear stresses. The molecules are stretched and ruptured by shear stresses and their ability to reduce mobility is permanently reduced. High shear stresses can be encountered in surface pumps, valves, and meters as well as at the point where fluids heave the wellbore and enter the formation. problem with the polyacrylamides is related to water salinity. viscosity. Another As water salinity

increases, a greater concentration of polyacrylamides is needed to achieve a target XC-polymer is relatively unaffected by water salinity. The affect a water salinity on viscosity increase for both types of polymer is illustrated in Figure 2. The retention of polymer in the pores of the rock is a very serious problem for either type of polymer. Lost polymer is absorbed on the walls of main flow channels and it

may completely plug off narrow pore channels. polyacrylamide sandstones. pounds per - 300 to 400 pounds By comparison, acre-foot. As

Polymer retention is more serious with per acre-foot in consolidated

are retained

XC-biopolymer pointed out

retention will range from 150 to 300 previously, the greater retention of

polyacrylamide results in a more permanent decrease in reservoir permeability.

2.1.4

Screening Candidatesfor Polymer Flooding

should not be considered for any reservoir unless the following

Polymer flooding guidelines are met: 1. 2. 3. 4.

Oil viscosity should be in the range 20-200 cp, Reservoir temperature should be less than 200°F. Permeability should be greater than 20 md. If the reservoir is being waterflooded, less than 0.2. the producing water-oil ration should be

These guidelines resulted from experience obtained in approximately flood projects, as well as theoretical considerations.

100 field polymer

Very few of the field projects were

successful from the economic standpoint, although in most cases polymer resulted in reduced producing water-oil ratios. The best results were obtained in projects where

polymer was used early in the waterflood life. Currently very few large-scale thickened waterflood projects are in operation.

2.1.5 Field Results

Table 1 shows a polymer floods statics for more than 250 polymer floods from USA by Manning et al.. The table emphasizes oil recovery and screening parameters used for polymer flooding. The average oil recovery (AOR) is about 3.56 % SOIW and 2.69 STB The large variations of

of improved oil recovery (lOR) per pound of polymer injected. AOR and lOR reflects the nature of polymer flooding.

2.2

FoamFlooding

Gas/Liquid foams offer an alternative to polymers for providing mobility control in chemical floods, and have been both field tested as mobility control agents in miscible and steam floods. Foams are dispersions of gas bubble in liquids. Gas/liquid dispersions are normally unstable and usually will break in less than 1 second. If surfactants are added to the liquid, however, stability is improved greatly so that some foams can persist indefinitely. To understand the foam properties requires some discussion of surfactants and their classifications. Most of the discussion applies to MP surfactant as well. A typical surfactant monomer is composed of a nonpolar portion (lipophile - i.e., having strong affinity for oil) and a polar portion (hydrophile - i.e., strong affinity for water); the entire monomer is an amphiphile (has affinity for both oil and water). Table 2 classifies the surfactant into four groups: anionics, cationics, nonionics, and amphoterics. a. Anionics. The anionics surfactant molecule is uncharged with an inorganic metal cation (usually sodium) associated with the monomer. Anionics surfactants are the most common in EOR because they are good surfactants, relatively resistant to retention, stable, and relative inexpensive. b. c. Cationics. In this case the surfactant molecule contains a cationic hydorphile and an inorganiC anion to balance the charge. Nonionics. These surfactants do not form ionic bonds, but when dissolved in aqueous solutions, they exhibit surfactant properties simply by electronegativeity contrasts bewteen their constituents. Nonionics are much less sensitive to high salinities than anionics and cationics. d. Amphoterics. A class of surfactants are those that contain aspects or two or Even small more of the previous classes. Within anyone class there is a huge variety of possible surfactants. variations can change surfactant properties drastically, (e.g., sulfates tend to be less thermally stable than sulfonate). 2.2.1 Foam Physical Properties

Foams are characterized by three measures; quality, texture, and bubble size range.

a.

Quality is defined as the ratio of foam (gas) volume to the total volume. The quality increases with increasing temperature and decreasing pressure both because (1) the gas volume increases, and (2) gas dissolved in liquid phase evolve from solution. A foam with quality greater than 90 % is a dry foam.

b.

Texture is a measure of average bubble size. The texture determines how the foam flow through a porous media. If the average bubble size is larger than the average pore diameter, the foam flows as a progression of films spreading individual gas bubbles. This condition is most nearly realized in permeable flow for high foam quality.

c.

Bubble size range. Foams with a large distribution range are more likely to be unstable because of gas diffusion from large to small bubbles.

2.2.2

Foam Stability The liquid film

The stability of a foam can be illustrated as shown in Figure 4.

separates two gas bubbles in cross section. The hydrophiles of surfactant are oriented into the interior of film and the lipophiles toward the bulk gas phase. The pressure in the gas phase is assumed constant because of its low density if the foam is static, or low viscosity if in motion. The surface tension at the gas/liquid interfaces plays an important role in film stability. The fOllowing three conditions are worth to mention to foam stability. a. b. c. Very low surface tension would not be favorable. Thinning can be caused spontaneously by diffusion of gas from small to large bubbles and by gravity drainage. External effects that may cause the foam to collapse are the presence of a foam breaker (oil or a high electrolyte concentration), local heating, or contact with a hydrophobic surface. 2.2.3 Field Results

Fields tests of foam injection alone have been scarce. The foam has been sucessfully applied to steam and miscible flood for mobility control.

3.

Low-1FT Process

Low-1FTmethods rely on injecting or forming in-situ a surface-active agent (surfactant) which lowers oil/water 1FT,and ultimately, residual oil saturation. High-pH or alkaline processes produce a surfactant in-situ, since these processes rely on reactions with acidic components of crude to generate the surfactant. Processes that inject the surfactant are called "MP" floods because of the tendency for surfactant to form micelles in aqueous solutions and inevitable need to drive the micellar solution with polymer. MPflooding has appeared in the technical literature under many names: detergent, surfactant, low-tension, soluble oil, microemulsion, and chemical flooding. An idealized version of a MP flooding sequence consists of 5 stages: preflush, MP slug, mobility buffer, fresh buffer, and chase water. The basic tool for illustrating how lowering 1FTreduces residual oil saturation (ROS) is the capillary distribution curve (CDC) shown in Figure 5. defined as the ratio of viscosity force to the capillary force as Nc = lIw J.lw / awo The CDC is a nearly horizontal plateau at small Nc until a critical value above which both residual phase saturation decrease. At the second critical Nc the residual phase saturations are zero and complete recovery of original trapped phase. The capillary number is

._ \

3.1

SurfactantFlooding

If anionic surfactant is dissolved in brine, the surfactant disassociates into a cation and a monmer. If the surfactant concentration then is increased, the lipophilic portions of the surfactant begin to associate among themselves to form aggregates or micelles containing several monomers each. Figure 6 shows a curve of the surfactant monomer concentration vs total surfactant concentration. The curve begins at the origin, increases with unit slope, then levels off the critical micelle concentration (CMC). Typically, CMC's are quite small (about 10.3 to 10-4moll L).

3.1.1

Surfactant/Oil/Brine

Phase Behavior

The micellar solution operates misbly with oil and water without phase separation and it differs from microemulsions due to the microscopic size of discontinuous phase, therefore ternary diagrams are always used to represent the micellar solution phase behabior as shown in Figures 7 to 9. MP phase behavior is affected strongly by the salinity of the brine pseudocomponent. The phase behavior was presented originally by Windsor. At low brine salinity, a typical MP surfactant will exhibit good aqueous (water-rich) phase solubility and poor oleic (oil-rich) phase solubility. Thus an overall composition near brine/oil boundary of ternary will split into two phases: an excess oil phase that is eseential pure oil and a (water external) microemulsion (o/w) phase that contains brine surfactant and some solubilized oil. This type of phase environment is called a Windsor Type I system or Type II system, II means that no more than two phases can form. (-) means that the tie lines have negative slope. The right plait point ins such system, PR,usually is located quite close to the oil apex as shown in Figure 7. For high brine salinities, electrostatic forces drastically decrease the surfactant's solubility in the aqueous phase. The overall composition within the two phase region now split into an excess brine phase and an o/w microemulsion. The phase

i \

environment is a Windsor Type 11(+) ystem. The Plait point PL is now close to brine s apex as shown in Figure 8. For an optimal salinity, there is a continuous change between Type II{-) and 11(+) nd a a third surfactant rich phase is formed. An overall composition within three phase region separates into excess oil and excess brine and microemulsion phase whose composition is represented by an invariant point. This environment is called a Windsor type III as shown in Figure 9. Figure 10, a prism diagram, shows the entire progression of phase environment from Type 11(-)o 11(+).The Type III region forms through the splitting of a critical tie line that t lies close to the brine/oil boundary as the salinity increases to Cel (low effective salinity limit for Type III phase environment). A second critical tie line also splits at Ceu (high effective salinity limit for Type III phase environment).

3.1.2

Phase Behavior and 1FT

To investigate the relation between 1FT'sand phase behavior, solubilization parameters between the microemulsion oleic phase, Fmo,and the microemulsion water phase, Fmw will be used. They are defined as Fmo = v.t», Fmw = v.Jv, Figure 13 shows the corresponding behavior of the soulbilization parameters and 1FT's.
(}mo

will be defined from low salinity up to Ceu, and (}mw will be defined from Cel to

high salinities.

3.1.3 Fluid/Rock Interaction

To understand the theory of surfactant flooding, we need to again look at an oil trapped in a pore of the rock and analyze the capillary forces that prevent the droplet from moving out. The trapped oil droplet shown in Figure 3 has been distorted in trying to leave the pore. The residual oil is in a discontinuous phase. . With the aid of the capillary pressure equations, we can write equations for the capillary pressure (oil phase pressure minus water phase pressure) at the constricted downstream end (r1)and at the upstream end of the bubble (r2). respectively: Assuming cos
. 'l% ...
. ~- ~._..

'J

equals 1.0, these capillary pressures are

p
and

cl

200w r1

where: Oow oil-water interfacial tension - dynes/cm = r1 0000 The difference in capillary pressure over the length of the droplet is:

= radius at constricted downstream end-ern

r2 = radius at upstream end-ern

(1)

In order to move the bubble out of the pore, the water phase pressure must drop by an amount equal to ~Pc over the length of the droplet. form of the capillary pressure gradient Equation (1) can be written in the

Me = 28 ow (_!_ _ _!_) L L r, r2
where L = droplet length in direction of flow-cm.

(1A)

The following example will show the magnitude of pressure gradient required to move out a droplet of residual oil in a typical water-wet sandstone reservoir.

Problem Calculate the pressure gradient required .to move a residual oil droplet out of a pore in a typical water-wet consolidated sandstone. r1 =9x10-4cm Use the following data:

r2 Oow

= 4 x 10-3 em
= 30 dynes/cm

droplet length = 4 x 10-2 em

Solution Pressure gradient is calculated by equation (1A).

Me
L
or

2(30) 4(10-2)

(1

9(10-4)

4(10-

1)
3

= 1.29(106) dyne / em'

/ em
',>

Me = 1.29(106)
L

dynes X 1.45(10-6) psi x30.48em (em" x em) dyne / em" ft

~ = 571psi / ft.
The pressure gradient calculated in this example is at least 3 orders of magnitude This is why the

greater than pressure gradients that are normally attained in the field.

oil droplets will remain trapped in a field waterflood regardless of how much water passes through the reservoir. In order to move the droplet under field gradients which are of the order of 0.5 psi/ft, the interfacial tension must be reduced from 30 dynes/cm to a value less than 0.03 dynes/em. This can be done by contacting the residual oil droplets with a surfactant that has been designed for this purpose. Surfactants can be introduced to the reservoir by a simple solution of surfactant in water or by using a microemulsion. The solution generally carries a much lower concentration of surfactant (2-3%) than the microemulsion (10%). Almost the same amount of surfactant is used with either method, however the surfactant bank size is much smaller when the microemulsion is used. A microemulsion is a stable solution of brine, oil, and surfactant plus some other components such as alcohol that tend to stabilize the system. The droplets of the dispersed phase are so small that the emulsion appears transparent and is very stable, having a shelf life of several years. A typical microemulsion might contain 45% oil, 45% water, and 10% surfactant. 3.1.3 Screening Guideline

)
}

Following is a list of reservoir and fluid properties that are considered to be desirable for a surfactant flood. Guideline values are given for certain reservoir and fluid properties. 1. 2. 3. 4. 5. 6. 7. 8. Homogeneous sandstone reservoir. Sand should be at least 10 feet thick. Permeability should be at least 100 md. Reservoir temperature should be less than 160°F. Oil viscosity should be less than about 20 cp, Reservoir water salinity should be less than 50,000 ppm. Divalent ion concentration should be less than 500 ppm. Waterflood residual oil saturation should be greater than about 30 percent.

The numbers quoted above are not absolute limits, but serve as guidelines. A reservoir cannot be ruled out just because a few of the above guidelines are not met. However, a good prospect will meet most of these guidelines. completed. No surfactant project can be started until a detailed reservoir study including an economic analysis has been

3.1.4 Field Results

At least 10 field tests of surfactant flooding processes have been done. effective than lo-concentrations of surfactant. Field experience to date suggests that the high-concentration micellar solution are more

3.2

High-pH or Alkaline Flooding

High-pH or alkaline flooding is an improved waterflood process in which the pH of the injected water is modified for increasing the oil recovery.

High-pH's indicate large concentrations of hydroxide anions (OH} solution is defined as

The pH of an ideal

where the concentration of hydrogen ions, CH+ , in moll L. As the concentration of OH- is increased, the concentration of H+ decreases, since the
{ ,

two concentrations are related through the dissociation of water,

Kw = (COH-) (CH+) I CH20+

and the water concentration is constant. These considerations suggest two means for introducing high-pH's into formation: dissociation of hydroxyl-containing chemicals, such as sodium hydroxide (NaOH), or adding chemicals that preferentially bind hydrogen ions, such as sodium carbonates.

3.2.1 Mechanisms
There are four mechanisms believed to be involved in making alkaline flooding work. These are: a. b. c. d. Interfacial tension alternation, Emulsification and flow, Wettability alternation, and Rigid interfacial films.

a.

Interfacial tension alternation

Figure 13 shows an example of interfacial tension vs pH. Increasing the concentration of NaOH would not necessary decreases the 1FTcontinuous. Rather there are may be an optimum concentration for which the lowest tensions are obtained as shown in Figure 14. By decreasing the 1FT,capillary number (Nc) is increased. To be effective, Nc must be increased to 10 -4. Recall

where Jl is the viscosity,

o is 1FT,and u Darcy velocity.

b.

Emulsion and flow

In the laboratory studies, the produced oil from the alkaline flooding are fine droplets. Therefore it is assumed that the lowering of 1FT due to reaction that takes place between organic acids and the alkaline solution also leads to the formation of emulsion. The emulsion could be OIW or W/O or both depending upon oil. emulsification and entrainment. c. Wettability alternation Two mechanisms in the emulsification process have been proposed to explain the incremental oil recovery:

In waterflooding, one must reduce the capillary force to improve oil recovery. Recall that the capillary pressure defined as Pc = 2
(j

Ir

Changing in the pH activates the surface-active materials to some crudes and brings about gross wettability change. Figure 15 shows an oil droplet trapped under water wet conditions. As the reverse wetting agent arrives at the pore and the concentration of the agent begins to increase the surface of pore becomes progressively more oil-wet. Capillary forces now favor movement of oil droplet into the smaller parts and tight places within the pore. d. Rigid interfacial films.

Crude oils form insoluble films at crude oil-water interfaces upon contact. This forming material was from the family of resins and asphaltenes in the crude oil and that once

formed the film would dissolve but very slowly into oil phase. Therefore these films reduce oil recovery. the addition of NaOH or other alkaline materials to the water neutralizes the acid constituents in the interfacial films thereby solubilizing these films. Hence increases oil recovery.

3.2.2 Field Results

Table 2 summarizes the high-pH flood statics. Note the wide range in reservoir and oil characteristics and in oil saturation at the start of the flood. lOR expressed as a fraction of PV, ranges from 0.0006 to 8, which translates into recoveries, expressed as a fraction of the OIP at the project start that are comparable with but slightly smaller than those reported from polymer floods. There are many features in these data that are common to the response of other chemical flooding process. The oil production rate declines as the total fluid production increases, indicating a declining oil rate.

4.

Conclusions
As the oil viscosity increases, more of the respective chemical is required

Chemical flooding usually is limited to moderate to low oil viscosities because of economics. to obtain good mobility control. The suqqestlons for future chemical flooding development are: a. the development of more cost effective chemicals such as surfactants and polymers that b. c. d. more salinity resistant, have minimal retention, can be manufactured in-site or can be recycle. development of more efficient design procedure such as applying MP technology to high-pH flooding deSign, removing technical limitations on the use water-soluble polymers, as they are temperature sensitivity, and more reliable prediction technique.

"OC
L-O

"O~

ro·c·-

000

ro _0>>

M<OI.OO>C\I <OCO<O,....,....C\I .. ... I.O~<OOOO M ,....

1.0

,....

o
M

co

M ,....

en
E :J E
~
<OM

en
oCt

t= t=

en o o o .J
U.

t-

C 000000

q~~

C\I ......

o
1.0

o
<0

CO

1.0

:E

o E·o Q. :J I . zo...
L-

w >.J
~

L-

O·

.DO>

0>0

f/)

OOCOCOC\lM I.OCOCO,....<OI.O,.... ,....

...... CO

W .J

m
oCt

t-

en
...J

00
I

« x
w

:E

0..

"§ z
o .-

c z

«
en

t-

o
en

« t« LL
LL

::l

a:

o Z o t« o
en en
o

« ...J
N

LL

~.~

00

,1a o
I

.c o E
E
(5 ._ 0 :;, 000~ ._ C. 0. -"'O""'~O Q) ~CN"'O_ Q)._
«S .C :;,

E :;,

E :;,EEoo ~ ._:;,:;, _ E ._ C C c._._ "'"

«S "'0
Q) .._

"'0 C

C1)

:;, s, 'E- .0. '"5

Q.._Q..cn<..)

0-

.L:.

.c
t-

'"

,~
C

<OE"".o~ c_ o - Q) ..oECi>E .0 Q) a. Q) E:O

U·-u

('I')

.... ~.c

('I') 0 0

('I')C\I

o~ .0 o·0

0 ('I') 0 0

0 .....
C\I ('I') 0

0 ....

LO 0 0

..-

..
"0
Q)

en
()

C 0

cr:LL

Q)....

OUa. U~

>-c ....0 Q).-> > .....

('I') 0 0 0

('I') OC\l ·0 CO • 0

0 0

......
0 .....
<D 0 0 0 0

LO

en
W
_J

~ 0

E·-.oE~ -Q)- .c u.o .0

N-

<0::>
U

('I') 0 0

a>

u:
J: e,
I .

LOLO LOLO C\l0

<D 0

..-

CO C\I 0

00::>
C)~ ::Jo

('I') 0 0

m~

..-

......
LOCO

..-

0 C\I

o
0
~ ~ ~
J: LL

J:

0 0

..-

00-

<t

> a:

c -.2 c ~ .... u~o·E !:; +:; ~ cu .... Q)Q)Q);: UcE:

.c o ._-

C\l0

MN

00 LOC\I

C\I 0

C\I 0

..-

C\I

en I M
W
_J

<t
~

.... -oQ):C ._.0 0 uE~

<::JC)

C)

cE

C')

·0 C\I_

~;:

C\I;: ~o o-

<D 0

-~._
a.(J)~()~
>-.c ~

o u ',,.~ Q)EQ)

0 O:C o ()O

C')

~z
N~

00

:c:c
~~

zz

:c
~

·in .c
0 0
I

.s.
t

Q) Q) ..... ..... ~ ~
·in .c
0 0
I

.~
t

fJl

on
~
J.4

~ '"

P4

:>
I

0 til

~ ""' =' ~

...
0

on ~
0

r:I

~ ~
P4

!
....
~ ~
0. 0
0

...

1? ....
CIJ

cJ CIJ

"-4 0

~ ~

oM
orf.

~ ~

:c 0
:E
r::::
'C

'"
0

as

..0

~
CIJ J.4 0 U

a:s
~ In 0
<J

s
In ~

""•
0

uOllnlos ~amAIOa JO ~llllqoR aAllela~

\.
)

s: r::::
~ 0 ~
(1)

o
CIJ "0. .-4

~ ....
CD CD

>.
0 cJ

~ '"

D.._

E >"0
0

.j

).

i
Cd

t' cJ

!:4 0
Po.

0 0 .-4

Q.

a 0.
I. ~

.... :>
~
0

! w
0r-

'0

0 .... u

.... U
0

u
0 0
CIJ cJ 0 U J.4

J.4

Cd ~

u
cJ

J.4

Cd

e
c:s
0 CIJ

-u,

e ::s

C)

0 U
"-4

'"

.-4 0 P4 0

~ ~ ~ ~

.s

.-4

<

d~ - UOllnlos ;[amAIOa }O ~l190~81A

{,

100 80
60
_.

.... -

B·rine (20,.QOO ppm)

Fresh Wa·ter

~;'

#
XC-Biopolymer

40
CJ I 0 .... .u .::s .... . 0 ...
til
,..,f.

'20

10 8
6 4 polya~de 2
~
",-

0 C1.t.

,
(

'6-1

--

....t.. ·co co ....t

0 CJ

.&oJ

:>.

~/

",

/'

:>

1----------------------------~ 1500 1000 o 500

Polymer Concentration - ppm

Figure 2 Effect of Salinity on Water Viscosity Increase

e 5
c
"0 CD
Co Co

Q.

a;

'0
+i
::J

t-

.c
~

is

SURFACE TENSION

./
,/

./

./

."..

__

----

ADSORPTION

/
/

,/

C.M.C.

CONC.

Thinner, Less adsorption, Contraction

Larger surface tension,

Figure 4 Surface Tension and Adsorption of a surfactant vs Concnetration t.

,.'

NORMAL RANGE WATERFLOODS

0_
00::

0:: Z

_, I-

Sorw o 30J---=:""'_--. S,w

-.a:

---

~.a:
~(/)

_,t-

:)

20

<:>.
PHASE

/WETTING

PHASE

(/)W

-0:

O::.a: 10
~~

WI-

NONWETTING

Figure 5 Schematic Capillary Distribution Curves

SURfACTANT MONOMER CONCENTRATION

TOTAL

SURfACTANT

CONClNTRATION

Figure 6 Schematic of the CMC

...

, ,
/

s
WAT~R EXTERNAL MIC ROEMULSION EXCESS OIL

COMPOSITION

OVERALL

Figure 7 Schematic of low-salinity/surfactantlbrine/oil

phase behavior

SWOLLEN MICELlE

.1

No1BRINE
Na+

. No+

5
EXCESS BHINE SINGLE OIL EXTERNAL MICROEMULSION

PHASE

OVERALL COMPOSITION

Figure 8 Schematic of high-salinity/surfactantlbrine/oil

phase behavior

BRINE

(
EXCESS BRINE

EXCESS OIL MIOOLE PHASE MICROEMlJLSION

I NVARIENT ~

OVERALL COMPOSITION

0
phase behavior

Figure 9 Schematic of the optimal-salinity/surfactantlbrine/oil

{
~.

0 .+oJ

c
C

.fn
+oJ +oJ

e
C CD

E ._

c > c
CD CD fn

e
as

.c c.
'I-

0
CD CJ C CD

::s

en c ._

C"' CD fn

'~
7

.c
fn

E
(5

as ._

e en

"C

E fn ._ 'a..
~
0

+ z
t.11:

'"

en .LL

e ::s

'i

10-' ~---------------------------------------,

o ......
&AI

:E

z 10-1 >o

Q
CI)

(J""mo

z
~

;J.
o
0::
&AI

10-·.

:
tZ

o
0.4 0.8
1.2 1.6 SALINITY.
%

NaCI

Figure 111FT and solubilization ratios

J9
c
0 0

CI)

c
:;:J
o

w
(!)

e0

0:

... <0

:r '"

c t)

ca ..., ca

'C

tn

.C.)

..... 0

tn

:s

z
Q~
~:I
(!)

en·
0:

i) 0:

~o ~~
I.L.I.L.

tJ)
0r-

.c C.)

10 E CI)

o
en
o

'"

wt: UQ.

..JZ ..JO

s~ °en
'" 0
..J

e :s .LL

C)

g3
..J

'" en
Q.

'"

f \,

2B.-----------------------------------------------------------------~

24

<,

E E

20

z
z
0

16

en

::z:

I.J <!

t·
\

I-

a: w
::z:

<! u,

12

\
4

pH

Figure 131FT vs pH for a Crude Oil

f

40.0

1\
10.0 ~
~ ~ ~
I-

en
Z
0

"-

5.0~

·ccr. TI.,,-,,,,, - - _. -_ .. _ - --PHloSl

""\;r~·..Js

""o( tt

D"STILLEO "";ECTJI)t,.

~:'T[R ""AHA

l.&J

>-

r.

o ""Oc.uc£o
""toTE Ji

CO~I"'''l[(J

0 Z
(f)

.
0.50-

,_

1.0 ;:- .
~
lI-

l.&J
_.J

....

.... z
0.05

0:: l.&J

< u < lL.

...
~

'\

0.10~ .
~ ~ ~

Q.

<,

I-

-----

_./
J
0.4

I 0.1

I .O.~

I 0.3

f 0.5

CONCENTRATLON OF No OH.· WEIGHT

PERCENT

Figure 141FT vs NaOH Concentration for a crude at 24 oC.

(
(0)
(b)
~.

-.

~.~

WATER WET
~(d)

OIL WET

0·

.
EARLY OESORPTION

~
WATER WET

Figure 15 Mechanism of Trapped Oil Removal by Reverse Wetting

(
\

SPE/DOE 35354

A METHOD FOR SIMPLIFYING FIELD APPLICATION OF ASP FLOODING

Troy R. French, SPE, 80M-Oklahoma,
Copyright This paper 1996,

Inc.

Socie1y of Petroleum Engineers, Inc.

at the SPEIDOE Tenth Symposium on Improved Oil

R8OOII8ry held in Tulsa, OK. 21-24 April, 1996.

was prepared for presentation

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(sl. Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correcbon by the author(s). The material, as presented, doss not necessarily reftect any position of the Society of Petrol8um Engineers, its offioers, or members. Papers presented at SPE meetings are subject to publication review by Ed~orial Committee of the Society of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment of where and by whom the paper is presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A Telex, 163245 SPEUT.

Abstract
An injection strategy that should simplify field application of the ASP (alkali/surfactant/polymer) process was developed. Additionally, this procedure should be useful for increasing oil recovery from on going polymer floods. Process design was accomplished with computer simulations which were )upported with laboratory experiments. Small slugs of concentrated alkali and surfactant, which were diluted by in situ mixing, produced suitable 'conditions for ultra-low 1FT (interfacial tension) at the leading edge of a larger polymer slug. Computer simulation was used to determine diffusion/dispersion parameters so that the degree of in situ reservoir mixing of small chemical slugs could be predicted. A three-dimensional, three-phase, reservoir simulator was used for simulation. The derived diffusion/dispersion coefficients were used for chemical flooding simulations, and the simulation predictions were verified with laboratory core floods. Coreflood effluent compositions matched computer simulation predictions. When followed with polymer, injection of small, concentrated, slugs of alkali and surfactant mobilized as much oil as injection of larger, more dilute, slugs of alkali/surfactant. This procedure should simplify field application of ASP flooding because very small slugs of alkali and surfactant can be easily handled and injected in an oil field. It was also shown that injection of small slugs of alkali and surfactant during an on going polymer flood should improve oil recovery significantly. An additional 23% PV of oil was recovered when small alkali and surfactant slugs were injected during polymer flooding.

Alkaline flooding is recognized as a tk>tentially cost-effective chemical flooding method. However,

Introduction 7ield Experiences.

early field projects conducted with strong alkalis, such as sodium hydroxide or sodium orthosilicate, were usually disappointing. Problems encountered in the field were typically associated with high injection pressures and silicate scale which was precipitated in production wells. 1 In the mid-1980s, rock-alkali reactions were identified as the cause of most problems associated with alkaline flooding projects. It was discovered that the extent of rock-alkali reactions could be reduced by using weaker alkalis such as sodium bicarbonate, sodium carbonate, and buffered mixtures. By 1989, it was also known that the interfacial properties of weak alkalis could be further enhanced by addition of small amounts of synthetic surfactants. Patents on this process are held by Shell Oil Company- and NIPER.3 The field viability of surfactant-enhanced alkaline flooding has been shown by successful field experiments performed in the United States by Shell4 and Surtek.5-7 The process is also reported to have been successfully applied in at least one other country, but there is apparently no official report available for that project at this time. Injection Strategies. Laboratory coreflood tests conducted at this laboratory with relatively large chemical slugs, 0.25 to 0.75 pore volume (PV) in size, yielded the following observations about injection sequence: • Oil recovery was highest when alkaline surfactant was followed by polymer. • Oil recovery was only slightly decreased when the alkali, surfactant, and polymer were simultaneously injected (mixed a in single slug), and oil production occurred faster. • Oil recovery was significantly reduced when surfactant was followed with alkaline polymer. • Oil recovery was lowest when surfactant was followed by polymer, and there was no alkali in any of the injected mixtures. When alkaline preflushes were injected before injection of the main chemical slug, it was observed that: • Oil production increased as the preflush volume was increased from 0 to 0.25 PV. • The alkaline preflush increased oil production in all cases, but the effect was most pronounced when the cores were saturated with brine that contained divalent ions. These observations were derived from corefloods conducted with relatively large injection slugs of alkali and surfactant,

97

A METHOD FOR SIMPLIFYING FIELD APPLICATION OF ASP FLOODING

SPE35354

and another observation was that, generally, oil recovery decreased when the sizes of the dilute chemical slugs were decreased.8 Since large chemical slugs are expensive to handle in the oil field, injection of very small slugs could simplify the method. Since oil recovery was shown to be decreased by injection of very small dilute chemical slugs, researchers here decided to test an alternate strategy-injection of very small concentrated slugs of alkali and surfactant. The hypothesis was that in situ mixing of small concentrated slugs of alkali and surfactant could produce optimal conditions for low 1FT and adequate mobility control when injected ahead of a larger slug of polymer. As in situ mixing occurs, process conditions should become similar to propagation of a small micellar slug ahead of polymer.9-IO This hypothesis was verified with computer simulations and laboratory corefloods. Computer Software
An injection strategy that may simplify field application of the

apparent for surfactant concentration, where the experimental data were affected by surfactant adsorption. The simulation values of 0.02 and 0.01 ft2/day for alkali and surfactant, respectively, were used for the computer simulations of simplified ASP injection strategies. Polymer concentration was derived from a correlation between effluent viscosity and polymer concentration, and the calculated diffusion/dispersion coefficient for polymer was 0.01 ft2/day. Coreflood Procedure Laboratory corefloods were conducted at 46° C in 25.4 em long, high-permeability Berea cores. The cores were first saturated with 0.6 wt% NaCl, then oil-flooded with crude oil to residual water saturation (Soi). Following oil-flooding, cores were flooded to waterflood residual oil saturation (Sow!) with 2.0 PV of 0.6 wt% NaCI. The chemical systems containing surfactant and alkali were then injected, followed by polymer. Frontal advance rate for chemical floods was 1ft/day. The chemical floods were conducted, after waterflooding,by injecting small, separate, slugs of chemical formulations that contained surfactant or alkali. For comparison, additional corefloods were conducted with larger volume slugs of the same chemicals. The total injected mass of surfactant, alkaIi, and polymer was maintained constant for the series of corefloods. Coreflood effluents were analyzed and the results compared to the computer simulations. Chemicals and Crude Oil A crude oil and an ASP chemical system was selected for / coreflood tests of the. Simplified injection strategy. For the process to be successful, dilution of small slugs of alkali and surfactant by in situ mixing must produce suitable conditions so that very low 1FT will occur in the zone near the leading edge of the larger polymer slug. The coreflood experiments were conducted with crude oil from the L. Wendt lease in Government Wells North (TX) field and mixtures that contained 3.75 wt% Na2C03 and 2.5 wt% active Chevron Cbaser™ CF-IOO anionic surfactant dispersed in 0.6 wt% NaCl. Very low 1FT values were measured between this oil and the chemical system up to 64 fold dilution with 0.6 wt% NaCI, and comparative corefloods were conducted with larger chemical slugs that contained 0.75% Na2C03 and 0.5 wt% surtactant.l+ The mobility control buffer contained Pflzer Flopaam™ 3330S polyacrylamide polymer in 0.6 wt% NaCI. During corefloods, polymer was followed with chase brine that contained 0.6 wt% NaCl. Analytical Procedures Coreflood effluent samples were collected in 5-mL increments and centrifuged before the oil and brine phases were separated. Aqueous phase alkaIinity was measured by titration to pH 4.0 with 0.1 N HCl, and surfactant concentration was measured by two-phase titration as described by Rosen and Goldsmith.15 A Brookfield Model LVT-CIP 1.565° cone and plate viscometer was used for measuring effluent viscosity, and interfacial tension between crude oil and aqueous mixtures was measured at 46° C with a Model 300 University of Texas spinning drop , interfacial tensiometer. -,

ASP chemical process is to inject very small slugs of alkali and surfactant ahead of the polymer flood. Computer simulation was used to determine diffusion parameters so that the degree of in situ reservoir mixing of small alkali and surfactant slugs could be predicted. For the process to be successful, dilution of small slugs of chemicals by in situ mixing must produce suitable conditions so that low 1FT will occur in a zone near the leading edge of the larger polymer slug. Diffusion/dispersion coefficients for surfactant, alkali, and polymer were required for input to the simulation program. PC-GEL is a three-dimensional, three-phase (oil, water, and gas), five-component reservoir simulator developed by incorporating a chemical transport model into a black oil computer simulation program, called BOAST. I I The simulator features include component diffusion/dispersion in addition to adsorption, convection, and in situ gelation parameters. The in situ gelation model and simulator were validated against data reported in the literature. The PC-GEL simulator is a useful tool for predicting the amount of in situ mixing that will occur during transport through porous media. Coreflooding diffusion was simulated in one dimension on laboratory scale. The 25.4 em core length was divided into 25 grids in the model with one injection and one production "well," each at the opposite ends. Both the injection and production speeds were 1 ftlday during the simulations. The core porosity and permeability were 0.2075 and 855 md, respectively. Preflush, sample injection, and water flooding procedures were simulated, and changes in component concentrations versus time and in situ distance were simulated Diffusion/Dispersion Coefficients The Diffusion/dispersion coefficients that were used for input to the PC-Gel numerical simulator were derived from data generated during coreflood experiments that were described in a previous report.12 Diffusion/dispersion coefficients were derived from polymer, alkaIinity, and surfactant data generated r1uringthose coreflood experiments. The simulator derivations are symmetrical (normal distribution) curves, which do not .xactly match the experimental data.13 This is particularly

98

SPE35354

TROY R. FRENCH

Simulation of ASP Injection Strategies 'liffusion/dispersion was simulated in the PC-GEL oneJimensional model on laboratory scale, using the diffusion/dispersion coefficients that were described above. A simulation for injection of small chemical slugs of surfactant and alkali followed with a larger slug of polymer is illustrated in Figures 1 through 3. The chemical system simulated in Figures 1 through 3 consisted of injection of 0.10 PV of an anionic surfactant, followed with 0.10 PV of alkali (Na2C03). The injection of surfactant and alkali was followed with injection of 0.60 PV polymer. In each case, the polymer slug was followed with chase brine. Figure 1 shows concentration profiles in the core for each of the three chemical components after 0.21 PV of fluid injection. The figure shows that small slugs of surfactant and alkali began to mix immediately, and that a significant amount of mixing of polymer and surfactant should occur near a wellbore. In the mixing zone, the maximum concentration of surfactant and alkali is about 50% of the original surfactant and alkali concentrations. Therefore, an initially over-optimum salinity condition should be approaching more optimum concentration levels. This effect should reduce adverse surfactant solubility conditions that could cause precipitation of surfactant by contact with concentrated alkali. (The simulator does not take into account alkali-surfactant interaction that could occur.) The well-documented interaction that may occur between polymer and surfactant should be reduced due to the initial separation of surfactant and polymer, ~hich are in separate slugs.16 Figure 2 shows that mixing of )rfactant and alkali continue during transport through the core. Figure 3 shows that alkali and surfactant should be wellmixed in the coreflood effluent. Alkali and surfactant are also shown to be mixed with polymer in the zone near the leading edge of the polymer slug. For comparison, a chemical system is simulated in Figures 4 and 5 that consisted of injection of 0.25 PV of a mixture that contained both anionic surfactant and alkali (Na2C03) in the same chemical slug. Injection of surfactant/alkali was followed with injection of 0.60 PV polymer. Polymer was followed with chase brine. In this case, the alkali and surfactant concentrations are lower than for the previous simulation; i.e., surfactant and alkali concentrations are initially optimum, and as dilution with formation water and polymer occurs, an under optimum salinity may occur at the back edge of the slug. Figure 4 shows that very little separation of surfactant and alkali occurs as the chemical slug is propagated through the core. The shapes and positions of the effluent concentration curves shown in Figure 5 are only slightly different than for injection of small, separate surfactant and alkali slugs that become mixed in situ, which was shown in Figure 3. Laboratory Corefloods Chemical floods were conducted, after waterflooding, by injecting small, separate, slugs of chemical formulations that contained surfactant or alkali. For comparison, additional -orenoods were conducted with larger volume slugs of the he chemicals. (See Table 1.) Coreflood effluent fractions

were analyzed and the results compared with the simulations described above. Coreflood 1 was conducted with small volume chemical slugs: 0.05 PV of a mixture containing 2.5 wt% active Chevron Chaser'" CF-l00 anionic surfactant in 0.6 wt% NaCI was injected, followed with injection of 0.05 PV of 3.75 wt% Na2C03 in 0.6 wt% NaCI. The surfactant and alkali slugs were followed with 0.60 PV of 1,500 ppm Pfizer Flopaam™ 3330S polyacrylamide polymer, in 0.6 wt% NaCI, for mobility control. The polymer solution was followed with chase brine that also contained 0.6 wt% NaCI. Coreflood 2 was conducted with the same injection strategy as core flood 1, except that the alkali solution was injected before the surfactant was injected. Coreflood 3 was conducted with a relatively large-volume chemical slug that contained both alkali and surfactant. A total of 0.25 PV of a mixture containing 0.5 wt% active Chevron Chaser™ CF-l00 anionic surfactant, 0.6 wt% NaCI, and 0.75 wt% Na2C03 was injected. The alkali-surfactant mixture was followed with 0.60 PV of 1,500 ppm Pfizer Flopaam™ 3330S polyacrylamide polymer, in 0.6 wt% NaCI, for mobility control. The total injected mass of surfactant, alkali, and polymer was equal to that of the two previous corefloods, and the polymer solution was followed with chase brine that contained 0.6 wt% NaCI. Oil recoveries from the three corefloods are shown in Figure 6. The figure legend should be interpreted as follows: S~A Surfactant injection preceded alkali injection. A~S Alkali injection preceded surfactant injection. A&S Alkali and surfactant injected simultaneously. Examination of Figure 6 shows that oil recovery was lowest when alkali and surfactant were injected simultaneously, intermediate when a small slug of surfactant preceded alkali, and highest when a small slug of alkali preceded surfactant. Core effluent analyses for the three corefloods are compared in Figures 7 through 9. Comparison of Figure 7 with Figure 6 shows that polymer was detected in the core effluent at the same time that tertiary oil production began. Also, polymer appeared in core effluent at the same time in all three corefloods. Figures 8 and 9 show that both alkali and surfactant persisted in the effluent longer for the coreflood where alkali injection preceded surfactant injection, which could be an artifact of core permeability distribution. Changes in core oil saturation for coreflood 2 described above, A~S, are shown in Figure 10. A nearly identical flood (coreflood 4 in Table 1), also shown in Figure 10, was conducted with higher polymer concentration to confirm that adequate mobility ratio was being used for the series of corefloods. Oil cuts for the two floods are shown in Figure 11. Results from the two corefloods are almost identical, which helps to confirm the potential effectiveness of the process. Additional corefloods (corefloods 5 and 6 in Table 1) were conducted to confum the effectiveness of injecting small alkali and surfactant slugs during an ongoing polymer flood. The chemical systems and injection strategies were exactly the same as for the previously described floods, A~S, except that a large volume of polymer was injected before injection of the mixtures that contained alkali and surfactant. Changes in core oil saturations are shown in Figure 12. Core oil saturations after waterflood were about 44% PV. Polymer flooding, after

99

A METHOD FOR SIMPLIFYING FIELD APPLICATION OF ASP FLOODING

SPE35354

waterflooding, reduced the oil saturations. to about 41% PV, which corresponds to recovery of only 3% PV of oil by polymer. This illustrates the well-known fact that polymer alone does not mobilize much residual oil after a homogeneous core is waterflooded. Examination of Figure 12 shows that after injection of the small slugs of alkali and surfactant (0.05 PV each), core oil saturations were reduced to about 18% PV. This corresponds to recovery of 26% PV of core oil after waterflood. This is 'nearly identical to the amount of residual oil recovered by the similar alkali/surfactant chemical flood (A--+S shown in Figure 6) conducted in a core immediately after waterflood. These corefloods demonstrate conclusively that injecting small alkali and surfactant slugs during an ongoing polymer flood could greatly improve the oil recovery from the polymer flood. To further refine and demonstrate the process, additional corefloods ( corefloods 7 through 11 in Table 1) were conducted with the same injection strategy described for the second coreflood (a small slug of alkali preceding a small slug of surfactant: A--+S), but with decreased surfactant concentrations. For the series of floods shown in Figure 13, all conditions are identical except for changing the amount of surfactant injected. For this oil and chemical system, the amount of incremental oil remained relatively constant until the amount of injected surfactant was reduced below 48 mg. The chemical cost per barrel of incremental oil was at minimum value for this amount of surfactant. This demonstrates that the simplified ASP procedure has the potential to be economic in terms of chemical costs, as well as being simpler to apply in an oil field. Conclusions 1. An injection strategy that should simplify field application of the ASP (alkalilsurfactant/polymer) process was further refined, This procedure utilizes small slugs of alkali and surfactants, diluted by in situ mixing, thereby producing suitable conditions for ultra-low interfacial tension (IFf) at the leading edge of a larger polymer slug. 2. Computer simulations were in good agreement with the analytical results from coreflood tests. 3. Corefloods demonstrated that the Simplified ASP procedure has the potential to be economic in terms of chemical costs, as well as being simpler to apply in an oil field 4. Corefloods showed that injecting small alkali and surfactant slugs during an ongoing polymer flood should greatly improve the oil recovery from the polymer flood. Acknowledgment The author wishes to thank R.S. Bryant for her guidance and advice in conducting this study. Special thanks go to C.B. Josephson for his assistance in all phases of the experimental work and to M.M. Chang for performing the computer simulations. Funding was provided by Fossil Energy I U.S. DOE contract DE-AC22-94PC91008. Nomenclature So; = Initial core oil saturation (after oil flood), % PV Sow! = Core oil saturation after waterflood, % PV So = Instantaneous core oil saturation, % PV

D = Diffusion/dispersion coefficient, ft2/daY References

Dauben, D. L., R. A. Easterly, and M. M. Western. An', Evaluation of the Alkaline Waterflooding Demonstration Project, Ranger Zone, Wilmington Field, CA. OOElBC/10830-5 NTIS Order No. DE87001240 (May 1987). 2. Lawson, J. B. and D. R. Thigpen. Staged PreformedSurfactant-Optimized Aqueous Alkaline Flood. U.S. Patent 4,502,541 (Mar. 1985). 3 . Peru, Deborah A. Aqueous Flooding Methods for Tertiary Oil Recovery. U.S. Patent 4,817,715 (Apr. 1989). 4. Falls, A. H., D. R. Thigpen, and R. C. Nelson. "A Field Test of Cosurfactant-Enhanced Alkaline Flooding," paper SPE/DOE 24117 presented at the SPEIDOE Eighth Symposium on Enhanced Oil Recovery held in Tulsa, OK. (Apr. 22-24, 1992). 5. Clark, S. R., M. J. Pitts, and S. M. Smith. "Design and Application of an Alkaline-Surfactant-Polymer Recovery System to the West Kiehl Field," paper SPE 17538 presented at the 1988 SPE Rocky Mountain Regional Meeting, Casper, WY, May 11-13. 6. Meyers, J. J., M. J. Pitts, and K. Wyatt. "AlkalineSurfactant-Polymer Flood of the West Kiehl, Minnelusa Unit," paper SPEIDOE 24144 presented at the SPEIDOE Eighth Symposium onEnhanced Oil Recovery held in Tulsa, OK. (Apr. 22-24, 1992). 7. Pitts, M. r., H. Surkalo, and K. Wyatt. "Design and Field Implementation of Alkaline-Surfactant Polymer Chemical Enhanced Oil Recovery," Proceedings of the Sixth UNIT~ International Conference on Heavy Crude and Tar Sands, v. (Feb. 1995) 795-803. 8. French, T. R. and C. B. Josephson. "Alkaline Flooding Injection Strategy," U.S. DOE Report No. NIPER-563. (Sept. 1991). 9. Jones, S. C. "Finding the Most Profitable Slug Size," JPT v. 24 (Aug. 1972) 993-94. 10. Holm, L. W. "Comments on Finding the Most Profitable Slug Size," JPT v, 25 (Jan. 1973) 77-79. 11. Franchi, J. R., K;: J. Harpole, and S. W. Bujnowski. "BOAST: A' Three-Dimensional, Three-Phase Black Oil Applied Simulation Tool (Version 1.1). Vol. I: Technical Description and Fortran Code," U.S. DOE Report No. OOEIBCIl0033. (Sept. 1982). 12. French, T. R. "Surfactant-Enhanced Alkaline Flooding Field Project," U.S. OOE Report No. NIPER-569. (Oct. 1991). 13. Chang, M. M. and H. W. Gao. "User's Guide and Documentation Manual for 'PC-GEL' Simulation," U.S. DOE Report No. NIPER-705. (Sept. 1993). 14. FY 1995 Annual Technical Progress Report. Management and Operating Contract for the Department of Energy's National Oil and Related Programs. NIPERlBDM-OI93. (Oct. 1995) 104-29. 15. Rosen, M. 1. and H. A. Goldsmith. Systematic Analysis of Surface Active Agents, Wiley-Interscience, NY (1972) 42324. 16. French, T. R. and C. B. Josephson. "The Effect of PolymerSurfactant Interaction on the Rheological Properties ! . Surfactant-Enhanced Alkaline Flooding Formulations," U;: OOE Report No. NIPER-635. (Oct. 1992). 1.

100

SPE35354

TROY R. FRENCH

Table 1 - Corefloods
Ch!i!rnical !OQWlctm1i2c, rng.!mL.

Ma~, ms
SurfactaDt 138 85 S3 73 0 85 0 70

Qil !i!atYm1i2c,~qE~ W$rflooc! 47.0 43.6 402 44.5 45.1

42.8

Coref!ood

Ici!i!Cti<m stratoov 2.0 PV waterflood, 0.05 PV surfactant, 0.05 PV alkali 0.6 PV polymer

Surfacta[)t :1S :1S 5

AIkaI.i
37.5 37.5 7.5 37.5 0 37.5 0 37.5 37.5 37.5 37.5 37.5 37.5

f2.OOIw
1.5 1.5 1.5 2.8 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5

Ch!i!rnical 24.4 172 23.8 21.0 42.8 16.1 41.5 202 18.7 16.8 242 31.1 19.7

2 3

2.0 PV waterflood, 0.05 PV alkali, 0.05 PV ~rfactant, 0.6 PV polymer 2.0 PV waterflood, 0.25 PValkaii/surfactant mixture, 0.6 PV polymer 2.0 PV waterflood, 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer 2.0 PV waterflood, 1.0 PV polymer 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer 2.0 PV waterflood, 1.0 PV polymer 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer 2.0 PV waterflood, 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer 2.0 PV waterflood, 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer 2.0 PV waterflood, 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer 2.0 PV waterflood, 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer 2.0 PV waterflood, 0.05 PV alkali, 0.05 PV surfactant, 0.6 PV polymer

4 5 5a 6 6a 7 8 9 10 11

25 0 :1S 0 25 20 15 5 2 10

44.7 41.5 46.6 45.4 46.5 43.4 432

6)

48 15 6 31

101

A METHOD FOR SIMPLIFYING FIELD APPLICATION OF ASP FLOODING

SPE35354

1.0

Surfactant

t=
~
W

-Alkali 0.8
0- •

Polymer

IZ

0.6

0
0
W

o
~ ~
0.4
_J

-c a:
0
0.2

FRACTIONAL DISTANCE
Ag. 1-COrefloodlng simulation for Injection of 0.10 PVof surfactant followed by 0.10 PV of alkali, and 0.60 PV of polymer, after 0.21 PV of fluid InJection.

1.0

---o------c..

IZ

-c a:

I-

, ,
\
\

--e- Surfactant _Alkali

\ \ \ \
0- •

0.8

Polymer '

o
0

0.6

\ \

o
0
W N

\
\

_.
~
c(

0.4

a:
0
Z
0.2

FRACTIONAL DISTANCE

Ag. 2-COrefloodlng simulation for Injection of 0.10 PV of surfactant followed by 0.10 PV of alkali, and 0.60 PVof polymer, after 0.60 PV of fluid InJection.

102

SPE35354

TROY R. FRENCH

1.0

...,.. -0_

I I I

\ \ \
\ \

s tZ Z
W
W

0 t-

-Surfactant _Alkali
0- -

0.8
I I I I I I I

o
0
Cl

0.6

o
~
0.4
I ...J

\ \ \ \ \ \ \ \
\

Polymer

:2
Z

-c

\
\

a:
0
0.2

\
\

\
\ \
0.

1.0

2.0

3.0

EFFLUENT VOLUME, PV

FIg. 3-SlmulaUon

of coreflood effluent for InJecUon of 0.10 PV of surfactant followed by 0.10 PV of alkali, and 0.60 PV of polymer, followed with chase brine.

1.0
\

b
\

-\

Surfactant
0- -

0 tZ
W

-Alkali
-

I-

a:

0.8

Polymer

U Z

0.6

o
C
W N 0.4

:i :2
Z

a:
0
0.2

\ \ \ \

b__
0.0 0.0 0.2 0.4 '0.6 0.8 1.0

FRACTIONAL

DISTANCE

Rg. 4-Corefloodlng simulation for InJecUon of 0.25 PV of a surfactant/alkali by 0.60 PV of polymer, after 0.40 PV of fluid InjecUon.

mixture followed

103