Professional Documents
Culture Documents
DOE/PC/9
1008-0374
OSTl_ID:
IN-SITU
COMBUSTION
HANDBOOK
- PRINCIPLES
3175
AND
PRACTICES
I
Final Report
November 1998
By
Partha S. Sarathi
January
Performed
(O@jnal
1999
Under Contract
No. DE-AC22-94PC91
008
BDM ,Petroleum
Technologies
BDM-Oklahoma,
Inc.
Bartlesville, Oklahoma
I
This report was prepared as an account of work sponsored by an agency of the United States
Government.
Neither the United States Government nor any agency thereof, nor any of their
employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility
for the accuracy, completeness, or usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any
specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise
does not necessarily constitute or imply its endorsement, recommendation, or favoring by,the United
States Govern ment or any agency thereof. The views and opinions of authors expressed herein do not
necessariIy state or reflect those of the United StatesGovernment.
Thisreporthasbeen reproducedc!irectlyfi-omthebestavailablecopy.
Av&lable to DOE and DOE contractors from the OffIce of Scientific and
Technical Information, P.O. Box 62, Oak Ridge, TN 3783 1; prices available horn
(615) 576-8401.
I
Available to the public from the NationalTechnicalInformationService, U.S. Departmentof
Commerce,5285 PortRoyal Rd., Springfield,VA 22161
I
DISCLAIMER
Portions of this document may be illegible
in electronic image products.
Images are
produced from the best available original
document.
DoE/Pc/91 008-0374
Distribution Category UC-122
By
Partha S. Sarathi
January 1999
NIPER/BDM-0374)
Prepared for
U.S. Department of Energy
Assistant Secretary for Fossil Energy
Prepared by:
BDM Petroleum Technologies
P.O. Box 2565
Bartlesville, OK 73005
008
Table of Contents
In-Situ Combustion
Abstract .....................................................................................................xv
Acknowledgments
. . . ... . . .. .. . ... . .. .. . . .. .. .. .. .. . ... . .. ... .. . ... .. ... ... ... . .... .. .. ... . . .... . .. .
xvi
CHAPTER 1 ...............................................................................................1
Introduction
.Background
.......................................................................l
Process .........2
........................................................6
Projects .... 7
Distribution
Activities .................................... 10
of U.S. Combustion
Project Activities
............................................. 17
Projects ...12
Process ............................. 18
Process .................................................... 18
Compared
........................................................................2O
Limitations of Combustion
Process .........................................................2O
References
..............................................................................................22
CHAPTER 2 Fundamentals
of Fireflooding .............*........*..........*..25
introduction ..............................................................................................25
In-Situ Combustion
Processes
................................................................25
of Temperature
..........................................................................................27
Wet Combustion
......................................................................................29
of Temperature
Wet Combustion
i
of Temperature
Process
of Temperature
of Saturation
Process
........................................................................... 32
Process
of Saturation
Process ............................................................................32
Reverse Combustion
Other Processes
References
.....................................................m..............3l
of Saturation
..........................................................................3l
.............m.........................................................m.......34
Variation ....................................................................... 35
.......s......................................................................................36
Studies ....................37
Oxidation
....................... 37
....................................................................38
of Dry Combustion
Temperature
on Oxygen
Oxidation ....................................................................42
Reactions .....................................................46
.............................................................................47
DTATechniques
.......................................................................48
for a Venezuelan
Mixture ..............................................................50
of Kinetic Parameters
from Thermogram
.,.......................51
Showing Various
Regime .................................................................................55
...............................................49
Oxidation
Showing Effect of
ii
Set-up ................................................56
to Evaluate
Rate Calorimeter
of Combustion
............................................................... 58
(ARC) ......................................................59
Tube
Rate Calorimeter
...................................................................................... 59
Experimental
Set. Up ........................................................67
Profile Showing
....................................................................................68
Oxidation
In-Situ
Profile Showing
...............................................69
introduction ..............................................................................................7O
FIGURE 3.16 Schematic of a Typical Combustion
Comments
About Combustion
Tubes ................................................................................... 74
Combustion
Description of Combustion
of Combustion
Tube Employed in
Laboratories
........................................... 75
Procedures ..............................................................................76
Interpretation
of Combustion
Profile as a Function of
Tube Run ......................................80
Schematic
...
111
Profiles ............................... 79
Temperature
Peak .................................8l
Analysis of Combustion
High Temperature
Schematic Temperature
Peak .................................82
Combustion
Stoichiometry
......................................... 83
Examples of Combustion
Parameter Calculation
from Typical Product Gas Composition ................................... ............ 91
Modifications
Combustion
to Illustrate Combustion
(mole
................................. 100
?40) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Conventional
Combustion
(mole Y~)........................................................... 09
Gas-Phase
Gas-Phase
(mole
Yo) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
09
Parameters ....................................................... 10
(mole
YO) . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
(mole
Parameters
0/O) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
11
....................................................... 12
for Combustion
............................................................................................ 25
References
............................................................................................l33
Geologic Characterization
..................................................................... 133
.............................................134
Thickness ..............................................................................l36
Competence
.....................................................s................l37
Reservoir Heterogeneities
............m......m.........................................m..m
....l37
........................................................................................... 39
Composition
............................................................................................ 144
CHAPTER 5 Engineering
In-Situ Combustion
Performance
of a Laboratory
Combustion
Tube,
................................................................................... 150
..............................................................l5O
....................................................................................., 151
for Combustion
..................................152
Front Movement as
.............................................................................. 156
......................... 163
..........................................................................................l78
In-Situ Combustion
...........................................................l82
........................................................................... 182
Air
Cumulative
Incremental
Oil
Tests) .............184
Incremental Oil vs Ak
Tests ........................................... 186
Field Cases
................................................................... 192
Incremental
Oil Production vs
Tests ..........................................l94
Field Performance
Prediction
............................................................................................l97
............................................................................2OO
..............................................................#...2ol
Pz_qSand) ......@
Project Production
Conclusions
Response ................................................................2O2
........................................................................................... 203
Facilities ...............................................................................2O7
Comments .............................................................................................2O8
TABLE 6.3 Mobil (General Crude) West New Port Fireflood ............209
Paris Valley Combinations
Project Well
Characteristics
Analysis ................................................................2l6
Project .......................................................2l8
FIGURE 6.2-
of Nacatoch
Analysis ................................................................222
General Obsewations
............................................................................224
References
............................................................................................227
Plant ..............................................229
introduction ............................................................................................229
FIGURE 7.1 Ak Compression
Compressor
Ranges of Compressor
of Centrifugal
and Reciprocating
Efficiencies .....................................................................233
Power Requirements
at Various
Ratios ...........................................................................233
of Various Compressor
Advantages
and Disadvantages
of a Centrifugal
Advantages
and Disadvantages
of a Reciprocating
Advantages
and Disadvantages
Compressor
in ISC Operation
Terminology
Basic Relationships
.........240
........... 241
.............242
.................... 243
...............................................................................245
vii
Principles of Compression
Compression
Theoretical
.....................................................................245
Horsepower .........................................................................250
Adiabatic Compression
.......................................................................... 251
Polytropic Compression
......................................................................... 253
Isothermal Compression
........................................................................255
Reciprocating
.................................................................... 256
Compressor
of Reciprocating
Compressor
........ 257
Ideal Reciprocating
Cycle ...........o..........................m....m$
................#.....m..........259
FIGURE 7.8A
FlGURE7.8B
Intake .........................................................................26O
.Compression
..............................................................26l
.......................................................................264
Compressor
Pedormance
...............................................264
Ratio vs Volumetric
Compressor
.........................................................................268
............................267
Discharge
Horsepower
Compressor
...273
Compressor
... 274
Compressor
Shaft Horsepower
Air Compressor
Packaged Compressors
Process Compressors
Reciprocating
........................................................................ 278
....................................................................... ... 278
Compression
...
Vlll
Definitions ..............................................................................................283
Centrifugal
Compressor
Characteristics
................................................285
Compressor
........................286
...................................................................... 287
Operating Characteristics
Compressor
Application
Characteristics
Compressor
of Centrifugal vs
Curves of a Centrifugal
Compressor,
Superposed
Compressor
Upon Demand-
Curves of a Centrifugal
Upon Demand-Load
Compressor
........................291
Compressor
Compressor
.............................................................................."297
.............................o.m
..........................o...................................3ol
Air Compressor
Characteristic
Compressor
..........................................., 302
Characteristic
Compressor ..............................................................3O8
......................................................................................3O8
Effect of Cooling.Water
Temperature
....................................................3O9
Compressor
Compression
Performance
on the
..............................................3o9
Establishing
Selection of Compressor
........................................................3l5
Station ..........................................................3l6
Compressed
Ak Distribution
for M Compressors
........................>...... 320
Plant ................#......................................32l
ignition .............................................................................326
Ignition ..............................................34O
............................................................................................342
Practices .....................................................................343
.................................................................347
............................................................................................. 347
Perforating
.............................................................................................348
Well Completion
and Workover
Fluids ...................................................348
..........................................................................349
Pack .................................................................................352
Treatment
........................................... 353
...................................................................................356
Screening ...............................................................................................36O
Suggested
Preservation
for Fireflood
at 250F ..... 365
FIGURE 9.7 Trend for H+3 Emissions from a Fireflood .................... 366
Operational
............................................................................................375
xi
and Disadvantages
of Oxygen/Enriched
Ak
Fireflooding .........................................................................................378
Economics of Oxygen Fireflood ............................................................. 380
Supply Option ........................................................................................38O
FIGURE 10.1 Schematic
Absorption
(Pressure Swing)
Oxygen ...............................................................................................385
Laboratory Studies .................................................................................385
Safety Consideration
.............................................................................387
General ..................................................................................................387
FIGURE 10.6 Propagation
Oxygen Concentration
FIGURE 10.7 Maximum
Oxygen Velocity in
. .................................m.............39l
Lines ..........m......m...m
.............................................................................395
Area ........................................................397
xii
............................................................................................ 402
...
Xlll
xiv
IN-SITU COMBUSTION
HANDBOOK
PRINCIPLES
AND PRACTICES
by
Partha S. Sarathi
ABSTRACT
For nearly 90 years in-situ combustion
improve recovery from oil reservoirs. Despite its long history and commercial success of many field
projects, the process has not found widespread acceptance among operators due to the excessive number
of failures of many early field trails. Most of these failures can be traced to the inappropriate
application of the process in poorer prospects. Analysis of the successful projects, however, indicates
that the process is applicable to a wide range of reservoirs, and the chances of failures can be minimized by careful selection of the reservoir and adopting prudent engineering practices.
The current work was undertaken to encourage operators to consider tiplern?nting
tiis
proven
oil recovery process in their reservoir by presenting the state-of-the-art knowledge and best practices
of fireflood in simple terms. The organization of the handbook emphasizes practical aspects of the
process. Unlike the steam process, no comprehensive
exclusively to fireflood. Hence, equations and detailed calculation procedures needed to engineer and
evaluate the performance of an in-situ combustion project are included.
The handbook includes ten chapters including the introduction chapter. The principles of, the
combustion process and its variations are discussed in Chapter 2. Since laboratory studies are an integd
part of a fireflood project, Chapter 3 is devoted to kinetic and combustion tube studies. The reser-
voir geological and site selection criteria for a fireflood project site are reviewed in Chapter 4. Chapter
5 presents a step by step approach for designing fireflood projects. Case history of selected U.S. and
non-U.S. combustion projects are presented in Chapter 6. The compressor and air plant requirements
for combustion projects are addressed in Chapter 7. Since ignition is a critical element of fireflood,
issues pertaining to the ignition of reservoirs are fully discussed in Chapter 8. The Chapter 9 covers
the well completion and operational issues pertaining to fireflood. The enriched air/oxygen fireflood
requirements are discussed in the final chapter.
xv
ACKNOWLEDGMENTS
This work was sponsored by the U.S. Department of Energy (DOE) under Work Authorization
Number AC/15054/Bc/42,
Rev.C. The work of this nature could not have been completed without the
help and cooperation of many individuals and organizations. Many past and current practitioners of
in-situ combustion have made significant contribution toward the preparation of this handbook. The
author in particular wishes to express his thanks to Dr. Gordon Moore and Dr. Raj Mehta of University of Calgary Canada, Dr. D. Yannirnaras of Amoco,
their input and permission to include portions of their work and figures in this monograph.
illustrations, plots and tables included in this monograph
Several
past fireflood practitioners. Since their origin is not known, no reference is cited for these in the text.
The author also wishes to thank the Gas Processors Association
pressed Air and Gas Institute of Cleveland, Ohio for permission to include many compressor related
text and charts in the monograph from their publications. Finally the author wishes to thank Mr. Thomas B. Reid of U.S. Department of Energys National Petroleum Technology
Olsen, Director UNIT~
xvi
CHAPTER
Introduction
In-situ combustion (ISC) or fireflooding,
thermal oil recovery
Background
as it is sometimes called, is the oldest m(
ators consider ISC as a high risk unpromising oil recovery method. This misconception
result of the excessive number of failures of many early field trials. Most of these fai
traced to the inappropriate application of the fireflooding process in poorer prospects. An
successful projects, however, indicates that the process is applicable to a wide range of ret
the chances of failure can be minimized by careful selection of the reservoir and adop
engineering practices. The operational problems
are no
severe than those encountered with other recovery techniques and are easily surmountable
Current work was undertaken to encourage operators to consider implementing tti
recovery process in their reservoir by presenting the state-of-the-art in fireflooding. In-s
tion has come a long way since its inception in the 1920s. By adopting the most current tt
planning, implementation, operation and managing firefloods, the operator can minimiz
failures. Both published materials and the authors knowledge of the subject are used as
this work. It is further supplemented by information garnered from the detailed discussit
current and past in-situ combustion practitioners.
Chapter 1 outlines the purpose, scope and organization of the handbook and discu
rent state-of-the-art in MC technology. Other sections focus on the history of the technok
eluding remarks.
Purpose
and Scope
of the Handbook
The purpose of this handbook is to present a state-of-the-art knowledge and best prz
ISC technology
Combustion kinetics.
Casehistory
ISCwell
ofpastprojects
to ISC.
Hence, equations and detailed calculation procedures needed to engineer and evaluate the performance of an ISC project are included. Important issues are clarified by providing examples drawn
from past projects. The handbook includes tables, graphs, and rules-of-thumb to assist in the design of
a project. Extensive references are provided on various topics to assist the reader in obtaining further
details.
Organization
of the Handbook
The handbook includes ten chapters including the introduction chapter. The next chapter (Chapter 2) describes the various MC processors.
studies. The reservoir geological
tube
and site selection criteria for an ISC project site are reviewed in
Chapter 4. Chapter 5 presents a step by step approach for designing ISC projects. Case history of
selected U.S. and non-U.S. combustion projects are presented in Chapter 6. Chapter 7 discusses in
considerable
detail the compressors and air plant requirements for a fireflood project. In Chapter 8
methods for igniting the reservoir are presented. The Chapter 9 covers the well completion practices.
This chapter also briefly touches upon the environmental issues. The last chapter (Chapter 10) considers the enriched air/oxygen fireflood requirements.
Early
History
and Development
of the In-Situ
Combustion
Process
In-situ combustion of reservoir crude probably occurred in the air injection projects carried out
in southeastern Ohio during the early part of this century. Lewis (1916) in an unpublished internal
U.S. Bureau of Mines memo contemplated the possible occurrence of combustion in projects where
warm air was injected into the formation to combat paraffin deposition problems.
According
to Lewis (1916) the Ohio operators originated the practice of warm air injection into
the producing formation to mitigate paraffin deposition problems, In an attempt to minimize paraffin
deposition and to increase the oil recovery from its Wood property lease (near Chesterhill in Morgan
county, Southeastern Ohio) Cumberland Oil Company began injecting 150,000 cu. ft. per day of hot
(160F) 40-psi compressor discharges into the shallow (600 ft.) first Cowrun sand in August 1911.
The injection was continued for 40 days. Analysis of the produced gas samples taken from wells in
the repressurized area showed that virtually all of the wells were making carbon dioxide (Lewis,
191 7). The percentage of carbon dioxide and oxygen in the produced samples ranged from 0.77 and
4.316.48 by volume, respectively. In retrospect it is clear that prolonged injection of warm air into
the reservoir resulted in the autoignition of the in-situ crude. Lewis (1917) attributed the presence of
COZ and low oxygen content of the produced gas to the chemical reaction between in-situ crude and
the oxygen of the air.
Osgood (1930) in his monumental work, Increasing the Recovery of Petroleum presented the
detailed case histories of several early day air injection projects (19151925).
Empire Companys (predecessor to Cities Service Company and Oxy-USA
In discussing The
air injection project (1925) he noted: It was found that only 60!Z0of the volume of air introduced was
ultimately recovered from the sand, and that this gas consisted of principally carbon dioxide and
nitrogen, the oxygen having been extracted from the air underground. This change was interpreted as
indicating a slow combustion. Similar results have been obtained in the Bradford District, Pennsylvania. The results obviously point toward oxidation underground, either of the air or the rock minerals
with which the air comes in contact.
The first planned successful in-situ combustion project in the U.S. (perhaps in the world) took
place in October 1920 near the town of Marietta in southern Ohio (Mills, 1923). The Smith-Dunn
Company used the in-situ combustion technique (actually cyclic combustion process) to melt parafiin
and increase production, In describing the process Mills noted: oil-soaked
dropped down a well,. where the Smith-Dunn compressed air process was being used. Air and natural
gas coming into the well made a combustible mixture, which ignited and burned for 24 hours. The fire
which was confined to the bottom part of the well was then extinguished by pouring in a few buckets
of water at the top of the well. The well was put to production by releasing the pressure. The rate of
production was doubled.
Following
in burning out
1920 (Mills, 1923). In this project, the well was bailed dry, tubing run to
the bottom of the oil sand, and a drum of gasoline poured down the hole. Compressed air was then
forced down the bottom of the hole through the tubing into the formation. Dropping down some burningoil soaked waste ignited the well. The fire was kept burning for several days by supplying com-
pressed air through the tubing. The fire was then extinguished and the well was put to production by
releasing the pressure. The well flowed oil along with gas. The rate of production was increased several times over what it was before the burning.
The Hope Natural Gas Company in 1922 used a similar, but more refined process to improve oil
production. This project was the first to use a combustion chamber to initiate ignition. The National
Petroleum News (1923) described the process in considerable
Mills (1923) noted that the well bottomhole temperature was raised to 2300F and this high temperature resulted in two or three joints of tubing being burned loose, but drilled up in short time. He ilso
noted that the intense heat fused the sand, but the fracturing and caving of the hole at these extreme
temperatures offset the damage. The process resulted in considerable
increase in production
in a
nearby well and is thus truly a combustion drive process. In order to avoid burning and parting of the
tubing, he recommended
The first patent setting out the in-situ combustion principle was issued to E.R. Walcott on June
5, 1923. This patent was applied for in 1920 few months prior to the actual beginning of the field
experimentation in the U.S.A second patent also applied for in 1920 was issued to F.A. Howard on
November 6, 1923.
In the summer of 1927, A.F. Melcher of Marland Oil Company (now Conoco) instigated a combustion oil recovery project in Ponca City, OK by injecting the hot exhaust gases (800F) produced
from the cracking stills of the adjacent refinery into a very shallow oil sand, which was found at a
depth of 20 ft. (Torrey, 1953). In this experiment, large manhole size holes were drilled to the top of
the sand and the hot still gases injected for several months. Oil recovery from the four offset producing wells, situated in a square pattern 50 ft. away (diagonal distance) from the injector increased considerably and the experiment deemed as a success. It was reported that the bottomhole temperature at
the injector was 700F and the gas injection ceased after a few months due to excessive backpressure.
Vrsual inspection of the sandface, after the hole has been cooled, revealed carbon deposition and clay
fusion. This experiment is the first know field implementation of combustion drive process in the U.S.
Following this successful experiment, Marland initiated another hot air injection pilot in the shallow
South Coffeyville
This experiment, however, was deemed unsuccessful because the operator was unable to maintain the
air injection due to plugging of the sandface by coke.
The first theoretical paper on oil recovery by injection of hot combustion gases from primary
depleted sands was published by U.S. Bureau of Mines in 1928 (Lindsly, 1928). It was estimated that
approximately one billion (10$) Btu of heat is needed to raise the temperature of 1-ac-ft. of oil sand to
700F. This paper, as for as the authors knowledge, was the first to present the engineering and economic calculations of a then experimental process.
The first known field experiment of in-situ combustion outside the U.S. occurred in Russia in
1935 (Sheinmann et al., 1938). By mid1940s, concept of burning a portion of the oil bearing formation for the purpose of enhancing oil production began receiving serious attention from major U.S. oil
companies. In 1947, Magnolia Petroleum Company (now Mobil) and Sinclair Oil and Gas Company
(now Arco) began investigating we process in the laboratory to learn about the parameters controlling
the process (Kuhn & Koch, 1953; Grand & Szasz, 1954). Both Magnolias
investigation cuhninated in field pilots in Oklahoma in 1950. Both pilots proved to be technically successful and paved way to the rapid development of this oil recovery technology.
Since the implementation of these first modern day pilots, more than 225 combustion projects
were undertaken in the U.S. A good many of these projects were economically
ers failed for various reasons including unfavorable reservoir and fluid characteristics, poor engineering and operational problem.
small experimental
projects
Current
Status
of In-Situ
Combustion
Global
ISC Activities
At the end of 1997, eight in-situ combustion projects are in operation in the U.S. producing
about 5,200 barrels of oil per day. Six of these were implemented in deep light oil reservoirs and the
rest in shallow heavy oil pools. Worldwide,
as of early 1998 that are producing
approximately
heavv. The known distribution of ISC rn-oiects in the world and their dailv oil txoduction are shown in
4
.,
Table 1.1.
COuntry
Project
Name
Operator
011
Grav
ity,
No.
No. of
;fw
PrO-
API tors
Current
Crrrrent
ProducAOR, tionbbl/D
Comments
Mscfl
bbl
U.S.A.
1
Bellvue,
Bayou
State
1970
wet
19
15
85
12
400
Continuationof
Oxys(CitiesService)Combustion
Moritis
(1998)
Midway
Sunset,
CA
Texaco
1982
dry
11.5
10
47
1,000
Combustionin a
preflwiol ~:.s:m-el
expansionia currentlyunderway
Moritis
Jw?&
Medicine
PoleHill
unit
Continental
Resources
1985
dry
39
17
7.2
725
Moritis
(1998),
Miller
(1994)
Buffalo,
SD
Continental
Resources
1979
dry
30
12
21
10.5
550
Moritis
(1998),
Miller
(1994)
W.Buffalo,SD
Continental
Resources
1987
dry
30
15
10.9
365
Lightoilcombustionin a tight(10
md)carbonate
reservoir
Moritis
(1998),
M]ller
(1994)
S.Buffalo,
SD
COnti&rerr- 1983
dry
30
19
40
9.3
1,420
Lightoilcombustionin a tight(10
md)carbonate
reservoir
Moritis
(1998),
Miller
(1994)
--
15.6
280
Combustionis a
meanstogenerate
fluegasiflsitu
snd displace
advancingwater
column.Oilis
recovered
throughgravity
drainage.
Gillhamet
al.(1997)
Combustionin a
steamfloodedreservoir
GlendtC.
(1995)
LA
Resources
(1998)
W.Hackberry,LA
Amoco
1995
dry
33
--
Mt.peso,
AERA
1997
dry
Horae
Creek
Field.ND
Total
Minatorne
1996
Dry
32.2
11
10
400
Lightoilcombus.
.
(lj~d~;r~o%;te
reservoir
Gerrnain
(1997)
Battrum
%akatche
Wmr
Mobil
Canada
1966
wet
18
15
94
10
3,700
Horizontalwells
arebeirr utilized
since?993to
improverecovery
Moritis
J115,9:)a1.
Battrum
Saakatche
wan
Mobil
Canada
1967
wet
18
35
10
1,200
Horizontalwells
arebein utifized
sincef 993to
improverecovery
Moritis
1998),
Arnesetal.
(1994)
Bat&urn
Saskatche
wan
Mobil
Canada
1965
wet
18
22
10
1,350
Horizontalwells
arebeii utilized
sincef 993to
improverecovery
Moritis
Wabaska
w Alberta
Amoco
Canada
1994
h&wzl-
14
--
260
Horizontalwell
cycliccombustion(Eh-essum
Up
Blowdown)
CA
Canada
1
Cyclic
combustion (dry)
(1994)
1998),
Amesetal.
(1994)
Faaeihiet
al.(1996)
try
Project
Name
Operator
Date
Jnitiated
CosnbUs- oil
tionType Grov
ity,
API
No.
of
[njec
tom
No. of
Pro-
Current
AOR,
Current
Production bbt/D
Comments
Reference
Mscf/
bbl
Ahania
Kasnice
Balakh-
.-
1973
dry
12
-.
1988
wet
16
..
Markoetal.
130
(1995)
Azerbaijan
1
35
--
600
Sabmmychi
with
alkalinewater
~FOn
combustion
Mamedov
et
al.(1995),
Torta(1995)
Chhsa
1
CNDc
1996
Dry
29
Demeni
Keet
-.
1976
wet
39
Balol
oil &
Nat1Gas
1990
wet
15.6
--
190
--
165
KerxingNemangu
Nemango
Hun?@V
1
14
Mamedov
et
al(1995),
270
Katz(1982)
India
1
Corp.
Prqectalikelyto
be ex anded
fieldwlx e in 1996
MOritis
(1998),Roy
Ch&t#ry
Lanwa
oil &
Nafl Gas
Corp.
1992
wet
13.5
Balol
Oil.$z
Nat1Gas
Corp.
1996
Dry
15.6
_MorMs
(1998)
SSrlthal
oil &
Nat1Gaa
Corp.
1996
Dry
17
.Moritis
(199s)
Bechraii
oil&
Nafl Gas
Corp.
1996
15.6
_Mor&s
(1998)
Karazharr
bas
--
1981
wet
22,3
78
364
7.2
4,150
Combustionwith
alkalinewater
injection
Sa Iacau
de i arcau
--
1964
wet
15.9
132
507
12.3
8,8(KI
Worldslargest
combustion
project
W.V]dele
--
Moritis
(1998)
Kaza-
kbatan
1
Romania
1
1980
dry
19
19
50
17
610
T;~tae;E&,
(1995)
T;ur~:l9~
(1995)
E.lldele
--
1979
dry
19
33
89
21
660
T;~ta&SJ
(1995)
W.Balaria
--
1975
dry
19
22
60
24.5
820
Turta
~lwwon
(1995)
cOmltry
Project
Name
Operator
Date
Idiated
Combustion Type
Od
Grav
ity,
No.
of
Injec
API
tors
No. of
PrOducers
Current
AOR,
Comments
Current
PrOductionbbl/D
Reference
bbl
Romania
(co~td
E.Balaria
1987
16
.-
17.4
15
47
22.5
550
:s),
Machedon
(1995)
Russia
1
Okha
.-
--
--
--
180
detailsonthis
pil:;:da;: is
?
No
Mamedov
et al.
(1992,1995)
According to the Oil and Gas Journals biannual EOR survey (Moritis, 1998) about 1.3 billion
b/d of oil is being produced worldwide using thermal methods, which is about 2% of the worlds current oil production. Steam and hot water account for 97.8 % of all thermally produced oil in the world.
Combustion accounts for the rest. In the U,S. combustion accounts for about 1.2% of the 446,430 barrels of thermally produced oil. In Canada excluding the bitumen production by steam, combustion
accounts for about 49~0 of the estimated 12,870 b/d of thermal oil production. Even though combustion accounts for less than 10% of the worlds thermal projects its share of thermally produced oil is
substantial.
The known distribution of ISC projects in the world and their oil production are shown in Table
1.1 Since the available information on the ISC projects outside of North America are limited, there is
a chance that more ISC projects maybe in operation in the world than those included in Table 1.1
Examination of this table shows that while the U.S. leads the world in the number of active ISC
projects, individual projects outside the U.S. produce more oil than all the U.S. projects combined.
For example, the Supalcu de Barcau project in Romania, the worlds largest active ISC project, produces nearly 1.7 times as much oil as all the U.S. projects combined. However, the per well production of U.S. projects (average 28 b/d per well) are much higher than those of most countries.
successful,
while another 54
are dubious). California, Texas, and Oklahoma accounted for the bulk of the combustion
projects implemented in the U.S. Texas also accounted for all the oxygen (enriched air) combustion
projects implemented in the U.S. Both the majors and independents were active ISC players in the
U.S. The U.S. in-situ combustion activities by decade are summarized in Table 1.2.
Decade
Total No.
Of
Projects
Economitally Successful
Projects
Technically
Successful
Projects
Unsuccessful Projects
Percent
of Economic
Success
Percent of
Technical
Successful
(including
economic)
Percent of
FaiIures
1950-59
42
10
26
14.2
38.1
61.9
1960-69
127
16
35
76
12.6
40.2
59.8
1970-79
33
12
15
36.4
54.5
45.5
1980-89
22
5,
12
22.7
45.5
54.5
1990-
--
--
--
--
Total
226
39
57
130
21.5
44.6
The breakdoti
55.4
of projects by state is depicted in Table 1.3. Both majors and independents were
active players in the U.S. A talley of projects initiated by majors and independents are shown in Table
1.4.
Examination of Table 1.2 shows a sharp decrease of new projects since 1970. However, the percent of the project considered successful (technically and/or economically)
successfid combustion projects. Texas and Louisiana accounted for nearly all of the combustion
10
projects conducted in this region. Texas also accounted for all the oxygen (enriched air) combustion
projects implemented in the U.S.
The following general observations can be made with regard to U.S. combustion projects.
assertion the suitability of the process to recover heavy oil from shaIlow reservoir.
Orily about 20% of the U.S. combustion pilots, undertaken prior to 1980 matured to become
commercial projects.
Projects undertaken by larger operators generally tend to be more successful then those initiated by smaller independents.
The most common causes of failure of earlier pilots have been lack of reservoir continuity,
channeling or bypassing of injected air, poor crude combustion characteristics, inadequate
compression capacity, and well failures.
The current emphasis is to implement combustion projects in deep light oil reservoirs, where
waterflood economics
means to generate flue gas in-situ and the thermal effects play lesser role in the production
of oil.
11
State
Economic
Success
Distribution
of U.S. Combustion
Projects
Technical
Success
Failed
Total
Projects
1950-59
1.
Arkansas
2,
California
10
3.
Colorado
4.
Illinois
5.
Kansas
6.
Kentucky
7.
Louisiana
8.
Missouri
9.
Montana
10.
Mississippi
11.
Nebraska
12.
New Mexico
13.
NewYork
14.
N. Dakota
15,
Oklahoma
16.
Pennsylvania
17.
S. Dakota
18.
Texas
19.
Utah
20.
W. Virginia
21.
Wyoming
Total
10
26
42
12
State
Economic
Success
Distribution
Technical
Success
Failed
Total
Projects
1960-69
1.
Arkansas
2.
California
20
26
49
3.
Colorado
4.
Illinois
5.
Kansas
6.
Kentucky
7.
Louisiana
8.
Missouri
9.
Montana
10.
Mississippi
11.
Nebraska
12.
New Mexico
13.
NewYork
14.
N. Dakota
15.
Oklahoma
10
13
16.
Pennsylvania
17.
S. Dakota
18.
Texas
24
39
19.
Utah
20.
W. Virginia
21.
Wyoming
Total
16
35
76
127
13
State
Economic
Success
Distribution
Technical
Success
Failed
Total
Projects
1970-79
1.
Arkansas
2.
California
3.
Colorado
4.
Illinois
5.
Kansas
6.
Kentucky
7.
Louisiana
8.
Missouri
9.
Montana
10.
Mississippi
11.
Nebraska
12.
New Mexico
13.
NewYork
14.
N. Dakota
15.
Oklahoma
16.
Pennsylvania
17.
S. Dakota
18.
Texas
19.
Utah
20.
W. Virginia
21.
Wyoming
Total
12
15
33
14
State
Distribution
Economic
Success
Technical
Success
Failed
Total
Projects
1980-89
1.
Arkansas
2.
California
3.
Colorado
4.
Illinois
5.
Kansas
6.
Kentucky
7.
Louisiana
8.
Missouri
9.
Montana
10.
Mississippi
11.
Nebraska
12.
New Mexico
13.
NewYork
14.
N. Dakota
15.
Oklahoma
16.
Pe&ylvania
17.
S. Dakota
18.
Texas
19.
Utah
20.
W. Virginia
21.
Wyoming
Total
13
22
15
Distribution
of U.S. Combustion
Proj.
Economic
Success
Technical
Success
Failed
Total
Projects
19901.
Arkansas
2.
California
3.
Colorado
4.
Illinois
5,
Kansas
6.
Kentucky
7
,.
Louisiana
8.
Missouri
9.
Montana
10.
Mississippi
11.
Nebraska
12.
New Mexico
13.
NewYork
14.
N. Dakota
15.
Oklahoma
16.
Pennsylvania
17.
S. Dakota
18.
Texas
19.
Utah
Wyoming
Total
Grand Total
39
56
131
226
20.
21.
W. Virginia
16
Decade
No. Of
Projects
Technically
& EconomiCally
Successful
Projects
Technically
Successful
Projects
Unsuccesslid Projects
Percent
Successful
Projects
(Technical&
Economical)
Percent Economically
Successful
Projects
Majors
1950-59
25
13
48.0
20.0
1960-69
88
13
28
47
46.6
14.8
1970-79
19
11
42.1
31.6
1980-89
22.2
1990-,
Total
142
24
40
78
145.1
116.9
Independents
1950-59
17
13
23.5
5.9
1960-69
39
31
20.5
5.1
1970-79
14
57.1
42.9
1980-89
13
46.2
30.8
199CL
Total
84
13
13
58
131.0
115.5
40+%/lA2=().451)
17
Assets
and Liabilities
of In-Situ
Combustion
Process
is a highly complex
was not well understood by most early day in-situ combustion operators.
This resulted in a high rate of project failures in the 1960s, and contributed to the misconception
that
ISC is a problem prone process with low probability of success. As a result, operators interest in the
process waned, as reflected by the number of new project implementations
1.2). The truth, however, is that ISC is an attractive oil recovery process and is capable of recovering a
high percentage of the oil-in-place, provided the process is designed correctly and implemented in the
right type of reservoir.
Like other oil recovery processes, ISC has its assets and liabilities and no general criteria can be
specified to assure its success. The probability of an ISC project failure, however, can be minimized
by recognizing
its limitations and designing the project accordingly. In this section we enunciate the
advantages and limitations of ISC process and examine critically the reasons cited for its declining
popularity.
Assets
of In-Situ
Combustion
Process
lSC is a unique oil recovery process. It can be viewed as a combination process. It encompasses
some aspects of nearly every known oil recovery method. These include steam distillation, steam displacement, COZ flood, hydrocarbon miscible flood, immiscible gas (NZ) displacement, and water (hot
and cold) flood.
Next to waterflooding,
ISC is perhaps the most widely applicable improved oil recovery tech-
It uses air, the least expensive and the most readily available fluid as injectant.
Though most combustion projects are implemented in heavy oil reservoirs, it is increasingly
being used to recover light oil from deep reservoirs. In the U. S., more combustion projects
are in operation in light oil reservoirs than in heavy oil reservoirs.
18
Economically,
successful
projects have been implemented in sandbodies ranging in thickness from 4-150 ft. The process, however, proved to be most effective in 10-50 ft. sandbodies.
Reservoir inhomogeneities
Reservoir pressure has no effect on the technical success of the process. The process has
been successfully implemented in reservoirs ranging in pressure from vacuum to 4,500 psig.
The formation permeability has minimal effect on the process. The process has been successfully implemented in formations whose permeability ranges from 5 md to 10,000 md.
Recovery efficiency is better than other oil recovery processes (see Table 1.5).
The process can be applied in reservoirs where waterflood and/or stearnflood is not effective.
For example, in deep reservoirs (greater than 10,000 ft.), steam injection is not effective due
to excessive heat losses and high lifting costs rendering waterflood economically
unattrac-
tive. In such reservoirs, combustion and gas are the only processes that can be applied to
recover oil economically.
Combustion projects permit the use of wider well spacing and can result in higher ultimate
oil recovery in comparison to steamflood.
19
TABLE 1.5 Recovery. Efficiency of In-Situ Combustion Compared to Other EOR Methods
(Hasiba and Wilson, 1975)
-
Process
(A)
Process
displacement
efficiency,
%
(B)
Areal
sweep
efficiency,
%
Vertical
sweep
efficiency,
%
Compound
recovery
efficiency,
%
In-SituCombustion
95
70
85
56
Stearnflood
65
70
85
39
Cyclic Steam
20
Micro-EmulsionFlood
90
70
80
50
COz-Waterllood
80
50
80
32
NaOH-Waterflood
35
70
80
20
(c)
m)
Limitations
of Combustion
Process
Like all other oil recovery methods, combustion process has its share of shortcomings. Most of
these limitations can be overcome at considerable expense. This has reduced its overall attractiveness.
Following are some less attractive features of ISC:
Though
The power
are high
enough that overall costs for delivering air to the reservoir can be substantial. Relative to
energy intensive steam injection operation, the costs for in-situ combustion are lower only
when the formation is less than 40 ft. in thickness. For thicker reservoirs, the heat losses during a steam chive are low enough to enable the heat to be delivered at a lower cost.
Operational problems
and require a
higher degree of technical sophistication to solve it. In comparison, steam injection operations are relatively problem free.
Unlike the steam injection process design of in-situ combustion processes must be preceded
by expensive laboratory investigations. This is needed to ascertain the burning characteris-
20
tics of the crude, fuel availability and air requirements. Thus, planning and design of a combustion project is more expensive.
Success and failure of combustion process implemented in reservoirs of similar and widely
differing characteristics preclude the development of satisfactory guidelines to screen reservoir for combustion application. Expensive pilot is the only satisfactory answer to judge the
feasibility of a particular project.
The complexity of the in-situ combustion process hinder the development of more sophisticated numerical simulators for complete performance prediction.
While considerable improvements are being made in the application of this technology, many
operators still view this technology
in the deep, extremely low permeability carbonate, and elastic reservoirs in the U.S. had made operators take a second look at this process. The success of horizontal well combustion technology in the
heavy oil fields of Canada have also contributed to revival of operators interest in this process. Currently several new combustion projects are on the drawing board, and one operator contemplates on
implementing this process in a deep offshore light oil reservoir.
It is likely that the coming decade may see important advances in the application of this process
in reservoirs found in hostile environments. It is also likely that the process will increasingly be
applied to recover light oil in the U.S. and elsewhere. Thus future potential for oil recovery by ISC is
very promising.
21
REFERENCES
1.
Ames, B.G., R.E. Grams, and F.N. Pebdani (1994). Improved Sweep Efficiency Through
the Application of Horizontal Well Technology
ings of the 1994 NIPER/DOE
2.
Anon (1923), National Petroleum News, Vol. 15, No. 34, (Aug 22): p. 91.
3.
of
Thermal Methods for Heavy Oil Production in Proceedings of 12mWorld Petroleum Congress, New York, John Wiley, Vol. 4, pp. 19-29.
4.
Gillham, T.H., B.W. Cerveny, E.A. Turek, ahd D.V. Yannirnaras (1997), Keys to Increasing
Production Via Air Injection in Gulf Coast Light Oil Reservoirs, Paper SPE 38848 presented at the 1997 SPE Annual Technical Conference
October, 58.
5.
6.
Technology
7.
Heath, W.A. and S.1. Betzer. 1992. Secondary Recovery of Oil in the Unites States, l Edition, American Petroleum Institute, Washington.
8.
Hoffman,
9.
Howard, F.A. Method of Operating Oil Wells, U.S. Patent Number 1,348,473. (filed August
9, 1920 issued Nov 6, 1923.
10. Kuhn, C.S. and R.L. Koch. 1953: In-Situ Combustion Newest Method of Increasing Oil
Recovery.
11. Lewis, J.O. 1916. Some Observation in Regarding Smith-Dunn Type Compressed Air Process. U.S. Bureau of Mines, Unpublished Internal Documents.
22
January 1995,
pp. 83-96.
15. Mamedov, Y.G. and A.A. Bacserman (1992), Application
U. S. S.R., Paper SPE/DOE
Symposium
in
on EOR,
on Heavy Crude
and Tar Sands, Houston, Texas, February 1217, Vol. 1, Sec. 1, pp. 1928.
19. Miller, R.J. (1994), Kochs Experience with Deep In-Situ Combustion in Williston Basin,
in Proceedings of 1994 NIPER/DOE Symposium on In-Situ Combustion, Tulsa, OK, April
2122. U.S. DOE Report No. NIPER/BDM-0086
20. Mills, R. Van A. 1923. The Paraffin Problems in Oil Wells. U.S. Bureau of Mines Report of
Investigation, RI 2550 (December) Government Printing Office, Washington, pp.1o--l 1.
21. Mortis, G. (1998), EOR Oil Production Up Slightly, 1998 Oil & Gas Journal EOR Survey,
Oil & Gas J. (April 20), pp. 4977.
22. Osgood, W.H. 1928. Increasing the Recovery of Petroleum. Vol. 1, ls Edition, p. 249, New
York McGraw Hill Book,
23
1938.
Gasification of Crude Oil in Reservoir Sands. The Petroleum Engineer Part 1. (December)
pp. 2730; Part 2 (February 1939) pp. 91-100.
26. Smith, R.L. and K.M. Watson. 1953. Oil Recovery Process. U.S. Patent Number 2,642,943
(Filed, May 20, 1949; issued June 23, 1953).
27. Torrey, P.D. 1953. Prospects for Improving Oil Recovery
in Proceed-
ing of 1994 NIPER/DOE Symposium on In-Situ Combustion, Tulsa, OK. April 2122. U.S.
DOE Report NIPER/BDM-0086
29. Wolcott, E.R. Method of Increasing the Yield of Oil Wells U.S. Patent Number 1,457,479
(Filed Jan 12,1920,
30. Zhdanov, S.A. and M.L. Surguchev (1991), The Status and Prediction of EOR Methods
Application in the USSR, in Proceedings of the 13 World Petroleum Congress, New York,
John Wiley, Vol. 2, pp. 529536.
CHAPTER
2 FUNDAMENTALS
OF FIREFLOODING
Introduction
In-situ combustion (ISC) is basically a gas injection oil recovery process. Unlike a conventional
gas injection process, in an ISC process, heat is used as an adjuvant to improve the recovery. The heat
is generated within the reservoir (in-situ) by burning a portion of the oil. Hence, the name in-situ combustion, The burning is sustained by injecting air or an oxygen rich gas into the formation. Often
times this process is also called a fireflood to comote
ervoir. The oil is driven toward the producer by a vigorous gas drive (combustion
In-Situ
Combustion
Processes
Based on the direction of the combustion front propagation in relation to the air flow, the process can be classified as forward combustion
combustion front advances in the general direction of air flow; whereas in reverse combustion, the
combustion front moves in a direction opposite to that of the air flow. Only forward combustion is
currently being practices in the field. The forward combustion is further categorized into dry forward
combustion
injected into the reservoir to sustain combustion. In the wet process, air and water are coinjected into
the formation through the injection well.
Dry Combustion
In this process, air (or enriched air) is first injected into an injection well, for a short time (few
days) and then, the oil in the formation is ignited. Ignition is usually induced using downhole gas
burners, electric heaters or through injection of a pyrophoric agent (such as linseed oil) or a hot fluid
25
such as the steam. In some cases, auto ignition of the in-situ crude occurs. For auto ignition to occur,
the reservoir temperature must be greater than 180F and the oil sufficiently reactive.
Once ignited, the combustion front is sustained by a continuous flow of air. The combustion or
fire front can be thought of as a smoldering glow passing through the reservoir rather than a raging
underground fire. As the burning front moves away from the injection well, several well characterized
zones are developed in the reservoir between the injector and producer. These zones are the result of
heat and mass transport and the chemical reactions that occur in a forward in-situ combustion process.
The locations of the various zones in relation to each other and the injector are shown in Figure 2.1.
The upper portion of this figure shows the temperature distribution and the fluid saturation from injection well to producer. The locations of the various zones are depicted in the lower portion of the figure.
1200
1000 ?
DEPOSITION OF COKE
. . . . . . .
INJECTION
WELL
PRODUCTION
WELL
Figure 2.1 is an idealized representation of a forward combustion process and developed based
on liner combustion tube experiments. In the field there are transitions between all the zones. The
concept depicted in Figure 2.1 is easier to visualize and provide much insight on combustion process.
26
Starting from the injection well, the zones represented in Figure 2.1 are:
1.
2.
3.
4.
5.
6.
7.
These zones move in the direction of air flow and are characterized as follows:
The zone adjacent to the injection well is the burned zone. As the name suggests, it is the area
where the combustion had already taken place. Unless the combustion is complete, which is usually
not the case in the field, the burned zone may contain some residual unburned organic solid, generally
referred to as coke. Analysis of cores taken from the burned portion in the field indicate as much as
2% coke and saturated with air. The color of the burned zone is typically off-white with streaks of
~ays, browns and reds. Since this zone is subjected to the highest temperature for a prolonged period,
they usually exhibit mineral alteration. Because of the continuous influx of ambient air, the temperature in the burned zone increases from formation temperature near the injector to near combustion
temperature in the vicinity of combustion zone.
DRY COMBUSTION
COMBUSTION ZONE
~
I
/
f
DISTANCE
Immediately ahead of the burned zone is the combustion zone. The combustion zone is where
reaction between oxygen and fuel takes place generating heat. The combustion zone is a very narrow
region (usually no more than a few inches thick) (see Figure 2.2) where high temperature oxidation
(burning) takes place to produce primarily water and combustion gases (carbon dioxide COZ and carbon monoxide CO). The fuel is predominantly coke, which is formed in the thermal cracking zone
just ahead of the combustion zone. Coke is not pure carbon, but a hydrogen deficient organic material
with an atomic hydrogen to carbon (H/C) ratio between 0.6 and 1.6, depending upon the thermal
decomposition
(coking) conditions. The temperature reached in this zone depends essentially on the
nature and quantity of fuel consumed per unit volume of the rock.
Just downstream of the combustion zone lies the cracking/vaporization
high temperature generated by the combustion process causes the lighter components of the crude to
vaporize and the heavier components to pyrolyze
transported downstream by combustion gases and are condensed and mixed with native crude. The
pyrolysis of the heavier end results in the production of COZ, hydrocarbon and organic gases and solid
organic residues. This residue, nominally defined as coke, is deposited on the rock and is the main
fuel source for the combustion process.
Adjacent to the cracking zone is the condensation zone. Since the pressure gradient within this
zone is usually low, the temperature within the zone is essentially flat (300-550F)
the partial pressure of the water in the vapor phase. Hence, the condensation zone is often referred to
as the steam plateau, Some of the hydrocarbon
the crude. Depending on the temperature, the oil may also undergo visbreaking
reducing its viscosity. Visbreaking is a mild form of thermal cracking. This region contains steam, oil,
water, and flue gases, as these fluids move toward the producing well. Field tests indicate that the
steam plateau extends from 10-30 ft. ahead of the burning front.
At the leading edge of the steam plateau where the temperature is lower than the condensation
temperature of steam, a hot water bank is formed. This bank is characterized by a water saturation
higher than original saturation. An oil bank proceeds the water bank. This zone contains all the oil
that has been displaced from upstream zones.
28
Beyond the oil bank lies the undisturbed zone which is yet to be affected by the combustion process, except for a possible increase in gas saturation due to flow of combustion gases (COZ, CO, and
NJ.
The overall fluid transport mechanism in a combustion process is a highly complex sequence of
gas drive (combustion gases), water drive (recondensed formation water and water of combustion),
steam drive, miscible gas and solvent drive.
Although the bank concept approach described above provides much insight on the combustion
process, it is not a true representation of the field behavior. In the field, various zones are not readily
identified and there are considerable overlaps between zones. Further, the relative locations of the various zones and the sequence in which they occur may also be different from that described previously.
This difference arises mainly because of the heterogeneous nature of the reservoir. Reservoir heterogeneity causes the fluid and heat fluxes to be different at various points of the combustion region.
The fluid distribution within each of these zone is influenced by the temperature profile as well
as the relative permeability characterization of the formation. The chemical properties of the oil that is
left behind by the steam bank determine the amount of coke that will be laid down, which in turn
determines the amount of air that must be injected to consume this coke.
Wet Combustion
In the dry forward combustion process, much of the heat generated during burning is stored in
the burned sand behind the burning front and is not used for oil displacement. The heat capacity of
dry air is low and, consequently, the injected air cannot transfer heat from the sand matrix as fast as it
is generated. Water, on the other hand, can absorb and transport heat many times more efficiently than
can air. If water is injected together with air, much of heat stored in the burned sand can be recovered
and transported forward. Injection of water simultaneously or intermittently with air is commonly
known as wet, partially quenched combustion. The ratio of the injected water rate to the air rate influences the rate of burning front advance and the oil displacement behavior.
The injected water absorbs heat from the burned zone, vaporizes into steam, passes through the
combustion front, and releases the heat as it condenses in the cooler sections of the reservoir. Thus,
the growth of the steam and water banks ahead of the burning front are accelerated, resulting in faster
heat movement and oil displacement. The size of these banks and the rate of oil recovery are dependent upon the amount of water injected.
29
.
!
COMBUSTION
ZONE
I
i
I
EVAPORATION
FRONT
1
I
CONDENSATION
I
i
i
DISTANCE
Ld
u
3
z
%
CL
C.OMWSTfON
I
ZONE
i
i
t
!
I
2s
w
t-
CONDENSATION
VAPOR\ZATION
I
DISTANCE
FIGURE 2.4 Schematic of Temperature Profile for a Normal Wet Combustion Process
Without Convective Heat Front (Courtesy of UNITAR Centre, Mehta and Moore, 1996)
30
I
I
COM13.HION
ZONE
I
1
i
I
I
I
i
1
CONDENSATION
FRONT
I---r
DISTANCE
FIGURE 2.5 Schematic of Temperature Profile for Super Wet Combustion Process
(Courtesy of UNITAR Centre, Mehta and Moore, 1996)
H!
~
uau
uJ&
600
Ill
Ii
Ill
400
200
g
l-
0
Iii
-1
WATER
WATER
COKE--
FIGURE 2.6 Schematic of Saturation Profile for the Incomplete Wet Combustion Process
31
SATURATION
SATUl?.~TION
TEMPERATURE
(C)
(%)
Naa
m
i
I
ii
N
9
00
000
:00
I I I t I
).
.
UJ
IQ
___
n ! .
0.
x
m
.
-.
-L
$0
Depending upon the injected water air ratio wet combustion process is classified
wet combustion (Figure 2.3) normal wet combustion
bustion (Figure 2.5), At low rates (incomplete
wet combustion),
superheated steam, as it moves toward the combustion front. In this case the injected water f
<
recover all the heat from the burned zone. At higher water injection rate (normal wet combustion), the
injected water will recuperate all the heat from the burned zone. At even higher water injection rate
(super wet combustion) the maximum temperature at the burning front declines. The operating pressure influences the temperature of the combustion zone during superwet combustion. The temperature
and the saturation profiles for the three modes of wet combustion are depicted in Figures 2.6-2.8
(Moore and Mehta, 1996).
Laboratory studies and field tests have also shown that water-assisted combustion reduces the
amount of oil burned as fuel. This behavior increases the amount of oil displaced but, more importantly, it decreases the quantity of air required to bum a specified volume of reservoir. The mechanism
causing the fuel deposit to be decreased during wet combustion is believed to be the increased availability of hydrogen in the combustion zone. A portion of the fuel deposit hydrogenates and becomes
mobilized,
lower than
required for burning at the trailing edge of the front causing the oxygen to pass through the region
unreacted. Therefore, a portion of the fuel deposit is bypassed, reducing air requirements.
Quenched combustion is more applicable in heavy oil reservoirs and less feasible for high gravity oils with low fuel deposits. The process has been patented and field tested by Amoco as the COFCAW method. Laboratory studies have shown that water rates from 500-1,000
bbl/million
scf air
result in quenching of the combustion zone and reduced air requirements. Insufficient field tests have
been performed to verify the technical merits of quenched combustion. Cities Services experience in
the Bodcau in-situ combustion project (Joseph et al., 1980) showed that a water/air ratio of 250 bbl/
million scf shows much improved burning characteristics compared to dry combustion.
33
However,
timum water rates are very difficult to determine because they are affected by reservoir inhomogeneities. Segregation of the fluids could result in extinguishing of the firefront and sacrificing of some
of the intended benefits of the process. Only operating experience in a particular reservoir will allow
selection of the best water/air ratio to maximize recovery and economics.
To summarize, the dry combustion process bums part of the crude and displaces the rest. The
heat utilization for displacement of oil is poor, In wet combustion processes, a smaller amount of
crude is burned as fuel, and heat utilization and oil displacement are more efficient, resulting in
improved recovery rates. The most important aspect of partially quenched combustion
is that the
burning front advances at a faster rate due to burning only part of the fuel deposit. This results in
shorter project life and reduced compressed air requirements compared to dry combustion.
Reverse
Combustion
In heavy oil, reservoir forward combustion is often plagued with infectivity problems because
the oil has to flow from the heated, stimulated region to cooler portions of the reservoir. Viscous oil
becomes less mobile and tends to create barriers to flow. This phenomena is especially prevalent in
very viscous oils and tar sands. A process called reverse combustion has been proposed and found
technically feasible in laboratory tests. The combustion zone is initiated in the production well and
moves toward the injecto~ counter current to fluid flow. The injected air has to travel through the reservoir to contact the combustion zone. The basic concept in reverse combustion is that the major portion of the heat remains between the production well and the oil when it is mobilized. Therefore, once
the oil begins to move, very little cooling occurs to immobilize the oil.
The operating principles of reverse combustion are not as well understood as those for the forward mode. Although, the combustion process is essentially the same, its movement is not controlled
by the rate of fuel burn-off but by the flow of heat. As explained in the section on dry in-situ combustion, the three things required for burning are oxygen, fuel, and elevated temperature. During reverse
burning, oxygen
throughout the formation. The factor which determines where the burning occurs is the high temperature which occurs at the producing well during ignition. As the heat generated during the burning elevates the reservoir temperature in the direction of the injector, the fire moves in that direction. The
combustion front cannot move toward the producer as long as all the oxygen is being consumed at the
fire front. Thus, the combustion process is seeking the oxygen sources but can move only as fast as the
heat can generate the elevated temperatures.
34
The portion of the oil burned by forward and reverse combustion is differer
tion bums only the cokelike residue, whereas the fuel burned in reverse combt
intermediate molecular weight hydrocarbon. This is because all of the mobile oil 1
the combustion zone. Therefore, reverse combustion consumes a greater percen
than forward combustion. However, the movement of oil through the high temper
considerably more cracking of the oil, improving its gravity. The upgrading pro{
bustion is very desirable for tar-like hydrocarbon deposits.
Although reverse combustion has been demonstrated in the laboratory, it ha!
the field (Trantham and Marx, 1966). The primary cause of failure has been the te
ous ignition near the injection well. However, projects in the tar sands are bein
attempt to use reverse combustion along fractures to preheat the fobation.
As the
injection well, the air rate is increased, and a normal forward fireflood is commen
Other
Processes
Variation
2.
3.
4.
5.
Steam-Oxygen
6.
7.
8.
co-injection
process.
Only the first four variations have been field-tested. Moore et al. (1994) has
variations and the interested reader is directed to this reference for additional deta
35
REFERENCES
1.
Moore, R.G., C.J. Laureshen, J.D.M Belgrave, M.G. Ursenbach, and S.A. Mehta. In-Situ
Combustion: New Ideas for an Old Process. Presented at the 1lti Annual Canadian Heavy
Oil and Oil Sands Symposium. March 2, 1994, Calgary, Alberta, Canada.
3.
Moore, R.G. and R.J. Mehta (1996), Heavy Crude: Energy Alternatives for Development
Heavy Oil Workshop Sponsored by UNITAR Centre for Heavy Crude and Tar Sands at
Campina, Romania, June 3-6, Workshop Volume 2, pp. 3-6 to 39.
4.
Trantham, J.C. and J.W. Marx. 1966. Bellamy Field Tests: Oil from Tar by Counterflow
Underground Burning Journal of Petroleum Technology.
36
CHAPTER
3 KINETICS
AND COMBUSTION
TUBE
STUDIES
Introduction
Unlike steam injection process, where the oil composition
impact on oil recovery, these parameters play a major role on the outcome of an in-situ combustion
(ISC) process. This is because, the ISC depends for its existence on the occurrence of chemical reactions between the crude oil and the injected air within the reservoir. The extant and nature of these
chemical reactions as well as the heating effects they induce depends on the features of the oil-matrix
system. The reservoir rock minerals and the clay contents of the reservoir are known to influence the
fuel formation reactions and their subsequent combustion.Hence
standing of in-situ combustion chemical reactions and their influence on the process is critical to the
design of the process and interpretation of the field performance.
The objective this chapter is to present an overview of the principal chemical reactions associated with the ISC process and to describe the various experimental techniques utilized to obtain process and design information
interest.
Chemical
Reactions
Associated
with In-Situ
Combustion
The chemical reactions associated with the in-situ combustion process are numerous and occur
over different temperature ranges. Generally, in order to simplify the studies, investigators grouped
these competing reactions into three classes: (1) low temperature oxidation (LTO), (2) intermediate
temperature, fuel formation reactions, and (3) high temperature oxidation (HTO) or combustion of the
solid hydrocarbon residue (coke).
The LTO reactions are heterogeneous (gas/liquid) and generally results in production of partially oxygenated compounds and little or no carbon oxides.
of hydrocarbons
which leads to the formation of coke (a heavy carbon rich, low volatility hydrocarbon fraction).
oil, fuel and the oxygenated compounds to give carbon oxides and
water.
37
Low Temperature
Oxidation
During in-situ combustion the hydrocarbons initially present in the oil undergo two types of
reaction with the oxygen (injected air) depending upon the prevailing temperature. Those reactions
which occur at temperatures below 400F are defined as the low temperature oxidation (LTO) and the
other being the high temperature oxidation (HTO). Unlike the HTO, which produces COZ, CO, and
water (HZO) as its primary reaction products, LTO yields water and partially oxygenated
bons such as carboxylic acids, aldehydes, ketones, alcohols, and hydroperoxides
hydrocar-
Thus LTO can be thought of as oxygen addition reactions. LTO occurs even at low reservoir temperature and is caused by the dissolution of oxygen in the crude oil. The degree of dissolution depends
upon the diffusion rate of oxygen molecules in the crude (Burger et al., 1972) at reservoir temperature. Light oils are more susceptible to LTO than heavy oils.
LTO reactions are highly complex and not well understood. However, it is believed that LTO
reactions consist of condensation of low molecular weight components to higher molecular weight
products. Compositionally
LTO has been found to increase the asphaltene content of the oil and to
(Alexander et al., 1962; Bousaid and Rarney, 1968; Severin et al.; Babu and Corrnack, 1984). It has
been shown that LTO reaction increases the amount of fuel available for combustion (Alexender et al.,
196; A1-Saadon, 1970) and causes a substantial decline in recoverable oil from the distillation and
cracking zones (Dabbous and Fulton, 1974),
Low air fluxes in the oxidation zone resulting from reservoir heterogeneities and oxygen channeling
promote LTO reactions. Poor combustion characteristics of the crude also tend to promote
when oxygen, rather than air, is injected into the reservoir. To rectify this situation some investigators
recommend
behind this suggestion is that the addition of steam to the oxidizing gas stream will lower the oxygen
partial pressure at the burning front and modify the kinetic reaction that creates the heat needed to
promote and sustain combustion.
38
LTO are generally believed to occur at temperatures of less than 600F, but this
temperature range is very oil dependent. It is very difficult to assign a temperature range to LTO region because the carbon oxide reactions (C-C bond cleavage) are begin to occur
at temperatures between
270F
reactions are evidenced by a rapid increase in the oxygen uptake rate as well as
the generation of carbon oxides, but their characteristics feature is that there is a
decline in the oxygen reaction rate at temperatures in the range of 450-540F.
This gives rise to the negative temperature
is a temperature interval over which the oxygen uptake rate decreases as the
temperature increases.
i
\
\
,
200
400
600
300
Temperature fC)
39
Failure of the reaction temperatures to transcend the negative temperature gradient region will
lead to a very low oil displacement efficiency, This is because the oxygen addition reactions cause
vapor phase to shrink significantly and also make the oil less mobile. Because the dominant end product of LTO reaction is coke, prolongation of LTO reaction for an extended period can cause the oil to
be permanently trapped in the pores.
The above observations were made based on LTO study of bitumens (Athabasca oil sand). Laboratory LTO studies by Fassihi et al. (1990) on four oils ranging from 31. 110. 1API and from 1454,300 cp dead oil viscosity indicate LTO has insignificant effects on either composition or viscosity
of light oil (31.1 API). LTO of light oils does not significantly affect either their mobility or recovery.
Yannimaras (1997) also observed similar trend in their laboratory air injection study of various light
crudes. However,
of heavier cmdes
asphaltenes and resin contents). For these oils LTO increases both the viscosity and density which in
turn affect their recovery.
Pre-oxidation
increases the fuel availability and subsequent air requirements for combustion. Thus conditions that
promote LTO reactions must be minimized during the in-situ combustion of heavy oils. Further, during LTO certain reactive species in the oil often produce unstable hydroperoxide
decomposition
intermediates. Their
releases much heat and can cause the oil to autoignite (Smith and Schweyer, 1967).
The Pyrolysis
Reactions
As the reservoir temperature raises, the oil undergo a chemical change called pyrolysis. Pyrolysis reactions (intermediate temperature oxidation reactions (ITO)) are often referred to as the fuel
deposition reactions in the ISC literatures, because these reactions are responsible for the deposition
of coke
molecules,
reac-
atoms untouched. In the cracking reactions, the carbon carbon bond of the heavier hydrocarbon
molecules are broken, resulting in the formation of lower carbon number (smaller) hydrocarbon molecules. In the case of condensation reactions, the number of carbon atoms in the molecules increases
leading to the formation of heavier carbon rich hydrocarbons.
of its constituent hydrocarbons determine the rate and extent of the different pyrolysis reactions.
40
donotundergo
and the larger molecules undergo cracking. Cracking reactions are usually initiated
reaction. The dehydrogenation molecules than recombine to form heavier molecules, eventually leading to the formation of coke.
C-H bonds of the ringed molecules are broken and replaced by a more stable C-C bonds and leads to
the formation of a less hydrogenated polyaromatic molecule. When subjected to further heating these
condensation products losses more of the hydrogen and recombines to form heavier carbon rich polymolecules, eventually leading to the formation of large graphite like macromolecules.
Laboratory pyrolysis studies on heavy (14-16API)
et al.,
1988) indicate that the pyrolysis of crude oil in porous media goes through three overlapping stages:
distillation, visbreaking, and coking. During distillation, the oil loses most of its light gravity and part
of its medium gravity fractions. At higher temperatures (400-540F),
breaking) occurs in which the hydrocarbon
branched compounds,
that are more stable and less viscous. At still higher temperatures, (above
550F) the oil remaining in the porous medium cracks into a volatile fraction and a non volatile carbon rich hydrogen poor residue often referred to as coke.
fraction of an oil and generally contains 80-90%
cracking reactions produce hydrogen gas and some light hydrocarbons in the gas phase. It is further
observed that distillation of crude oil at low temperatures plays an important role in shaping the
nature and extent of the cracking and coke formation reactions. High operating pressures generally
lead to the formation of more fuel that is leaner in hydrogen.
41
Researchers at the University of Calgary have for over 20 years studied various aspects of insitu combustion and they describe the bitumen pyrolysis reaction as:
Bitumen + Maltenes
Maltenes + Asphaltenes
Asphaltenes + Coke
Asphaltenes + Gas
Maltenes are crude oil fractions which are pentane and toluene soluble and maybe further separated into saturates, aromatics, and resins using liquid chromatography. The asphaltenes are toluene
soluble but pentane insoluble fraction of the bitumen. Coke is defined as the fraction insoluble in toluene. Thermal cracking of asphaltene to coke has a long induction period (initiation time). This
induction period decreases as the cracking temperature increases.
High
Temperature
Oxidation
The reaction between the oxygen in the injected air and the coke at temperatures above 650F
are often referred to as the high temperature oxidation (HTO) or combustion reactions in the ISC literatures. Carbon dioxide (COZ), carbon monoxide (CO), and water (HZO) are the principle products of
these reactions. HTO are heterogeneous (gas-solid and gas-liquid) reactions and are characterized by
consumption of all of the oxygen in the gas phase. The stoichiometry of the HTO reaction (chemical
equation) is given by:
cH;+[:;:)+;]02+[
&]co+[fi]co2+;H20
(3.1)
42
Studies indicate though, HTO is predominantly a heterogeneous flow reaction and the burning
process involve a number of transport phenomena. Combustion
(oxidation)
is a surface controlled
reaction and can be broken into the following steps (Scarborough and Cady, 1982):
1.
Diffusion of oxygen from the bulk gas stream to the fuel surface.
2.
3.
4.
5.
Diffusion of the products away from the surface and into bulk gas stream.
cess will be controlled by that step. In general chemical reactions (step 3) proceed at a much faster
rate than the diffusional processes. Therefore, the overall combustion rate likely to be difision
con-
trolled,
The issue whether the chemical reaction rate or the oxygen diffusion rate controls the combustion process is rather controversial. Some investigators (Dabbous et al., 1974; Lin et al., 1984) found
that the combustion reaction was diffusion controlled, while others found it to be chemical reaction
controlled (Hughes et al., 1987; Fassihi and Brigham, 1982; Burger, 1976). Islam et al., 1989, contend
that the laboratory experiments are almost always conducted under rather ideal conditions, and the
experiments conducted under such ideal situations permit oxygen interracial mass transfer rate to be
quite high and thus likely to mask the influence of diffusion on the combustion reactions in the reservoir. Other investigators contend that the overall combustion rate is dependent on the oxygen concentration (partial pressure) at or near the combustion
controlled and at low air fluxes the reaction is diflision
Reaction
Kinetics
Reaction kinetics can be defined as the study of the rate and extent of chemical transformation
of reactants to product. Though, simplistic this definition is accurate for this study. The study of reaction kinetics for the in-situ combustion process is undertaken for the following reasons:
1.
2.
To determine the conditions required to achieve ignition and or to determine if self ignition
will take place in the reservoir upon air injection,
43
3.
To gain insight into the nature of fuel formed and its impact on combustion.
4.
To establish parameter values for the kinetic (reaction rate) models used in the numerical
simulation of ISC processes.
Combustion of crude oil in porous media is not a simple reaction but follows several consecutive and competing reactions occurring through different temperature ranges (Fassihi et al., 1984).
Since crude oils are made up of hundreds of compounds,
~ of crude oil oxidation reaction would require an inordinately large number of kinetic expression.
However, this is not feasible because these compounds
described. This complexity
Many of
them contain several coexisting C-H bonds which can react successively or simultaneously and often
produce intramolecular reactions. Detailed models for hydrocarbon
only for the simplest hydrocarbon molecules and are made up of several reaction steps (equations).
Detailed hydrocarbon
mensional in-situ combustion simulators, because the computer size, speed, and cost requirements of
such a treatment would be too great. Detailed oxidation models have been developed and validated for
only the simplest fuel molecule and are not available for most practical fuels. However, very simple
models that approximate the oxidation reaction kinetics study of crude oils in porous media have
appeared in literature.
The simplest overall reaction representing the oxidation of a typical hydrocarbon fuel is
Fuel+ n102
-+-nz CO+n~COz+nl
(3.2)
HzO
where the stoichiometry coefficients { ni } are determined by the choice of fuel. This global reaction is
a convenient way of approximating the effects of many elementary reactions which actually occur in
the reservoir during the combustion process. Its rate must therefore represent an appropriate average
of all the individual reaction rates involved.
Most researchers describe the ISC oxidation reaction rates in terms of a simple reaction rate
model that assume functional dependency
dC~
Rc =
= kP;C;
dt
(3.3)
44
where
RC = combustion rate of crude oil,
C~ = instantaneous concentration of fuel,
k = rate constant,
Poz = partial pressure of oxygen,
a = order of reaction with respect to oxygen partial pressure,
b = order of reaction with respect to fuel concentration.
High temperature carbon and crude oil oxidation
1968) and others (Dabbous and Fulton, 1974) indicates first order reaction dependency on fuel concentration and 0.51.0 order dependency with respect to oxygen partial pressure; i.e., a = 1.0 and b
= 0.5 to 1.0.
The reaction rate constant k in Equation (3.3) is often a function of temperature and expressed
by
k.
exp(- /RT)
(3.4)
where
A = pre-exponential factor
E = activation energy
R = universal gas constant=
T = absolute temperature in OK
Equation (3.4) is often referred to as the Arrhenius rate equation and the constant A as Arrhenius rate constant in the ISC literature. The constants in Equation (3.4) vary with oil type, pressure,
and other parameters as well.
The units of the constant A, E, and R depends upon the unit selected for mass, pressure, temperature, and time. For example in the Iiterature the pre-exponentizd factor A is expressed in various
45
units such as day-l psi-l, see- atm.-l, see-l atm- , see-l Pa-l, hrl psi-n, etc. Similarly the activation
energy E expressed variously as cal/gm-mole,
Btu / lb.mole,
J/gin-mole
and E are functions of rock and crude and must be established experimentally using the crude and
porous medium of interest. ISC process simulators often require kinetic information to calculate the
firefront advancement rate and oil recovery. Experimental determination of kinetic parameters is
expensive and for a quick screening simulation the reported literature values are often used. Hence,
careful attention must be paid to the units when using these kinetic data.
Factors
Affecting
Oxidation
Reactions
Two of the most important factors in the in-situ combustion process are fuel formation and combustion. The physical and chemical processes that govern the ability of a crude to deposit fuel and its
subsequent combustion
(oxidation)
little fuel deposition may prevent the formation of a sustained, stable combustion front. Likewise, too
large a fuel deposition will result in uneconomically
agation of the combustion front and the air requirement depend on the extent of the exothennic oxidation reactions, which are controlled by the kinetics of these processes.
A substantial investigative effort has been made over the years in the laboratory to study the
many factors that affect the crude oil oxidation reactions in the reservoir. These investigations indicate
that the nature and composition
thermo-oxidative
of the reservoir rock and the characteristics of the oil influence the
characteristics of the reservoir crudes. The clay and metallic content of the rock, as
well its surface area has a profound influence on fuel deposition rate and its oxidation.
Clays and fine sands, have very high specific surface ~ea. Studies by Fassihi et al., 1984; Vossoughi et al., 1982; and Bardon and Gadelle, 1977, indicate presence of clays and fine sands in the
matrix favor increased rates of coke formation. Clays are solid acid catalysts and their catalytic activities are related to their acid site density and acid strength. Increased clay content increase the acid
site density and acid strength (Tanabe, 1970). Literature on catalytic cracking process (Tanabe, 1970;
Venuto et al., 1979) reveals that increased acid site density and acid strength lower activation energy
(smaller Arrhenius constant), and promote low temperature oxidation and coke formation reactions.
This is particularly advantageous in light oil reservoirs, where fuel deposition can be less than that
needed to sustain combustion.
waterflooded May Libby light oil reservoir (40 API), La is attributed to the influence of the rock
mineralogy (13.590 silt and 10.5910clay) on the fuel formation reaction (Hardy et al., 1972).
46
Metals and metallic additives also known to affect the nature and the amount of fuel formed.
Metals are used as catalysts in the petroleum refining and chemical process industries to accelerate
the hydrocarbon
coke formation and the catalytic effect of these metals was found to be ordered as follows: Cu < V <
Cr = Zn < Ni, with nickel about four to five times as active as vanadium (De 10SRios, 1988).
Studies on the effect of reservoir minerals on in-situ combustion indicate metals promote low
temperature oxidation and increase fuel deposition (Burger and Sahuquet, 1972; Fassihi, 1981; Drici
and Vossoughi,
1985). Studies conducted by the researchers at Stanford University (De 10S Rios,
1988; Shallcross, 1989; Baena, 199~ Holt, 1992) indicate that kinetics of all three oxidation reactions, (LTO, pyrolysis, and HTO) is affected by the presence of metallic additives in the crude oil.
These researchers also noted that the catalytic activity of a metal is highly dependent on the specific
composition of the crude. The benefits of metallic additives in promoting and sustaining combustion
in a light oil reservoir is documented by Racz (1985). The ability to initiate and propagate the combustion front in this Hungarian reservoir was attributed to the catalytic properties of the metallic additive which increased fuel concentration.
(TGA),
Differential
Therrno-
and Accelerating
Rate
In the quantitative technique a thin walled plug-flow type reactor cell containing
a sample of oil and sand is heated in a prescribed manner until the designated maximum temperature
is reached. Depending upon whether pyrolysis or oxidation was being studied, nitrogen or air flowed
through the sample. The kinetic parameters are calculated from the chemical analysis of the post test
core-oil samples and effluent gas.
Information
requirements, peak combustion temperature, oxygen utilization efficiency, etc., obtained from combustion tube tests.
47
Thermal
Analysis
Techniques
Thermal gravimetric analysis (TGA) and differential scanning calorimetry are the most widely
employed thermal analytical tools to study the oxidation kinetics of the combustion process. In the
recent years these also utilized as a screening tool for in-situ combustion. A more recent addition to
the family of thermal analysis tools is the accelerating rate calorimeter (ARC). ARC has been used in
the petroleum industry since the early 1990s mainly by Amoco
combustion application.
change in weight of the sample is recorded as a function of temperature or time, i.e., W = f (T or t). If
the temperature of the sample is increased as a linear function of time, the method is referred to as an
non-isothermal TGA. When pressure effects are important, the apparent weight changes are corrected
for buoyancy effects based on the blank run results (Bae, 1977). In differential thermal analysis
(DTA) technique the energy changes of a sample relative to a reference material are recorded as a
function of temperature or time. In practice, the sample temperature is compared continuously
with
48
M02003U)400!XW600
Tempera@w$%
-1
100
200
300
Temperature,C
400
500
600
In DTA technique the difference in temperature between the sample and the reference material
is plotted against the temperature and a baseline is established by connecting the extreme end of the
DTA curve by a straight line (Figure 3.3). The sample size affects the magnitade of the DTA curve
and the height of the curve is proportional to the weight of the sample.
-------
Weight
Weight derivative
-----
------
..-
---------------
---------, .
. .---.
_-.\,----
,.. ,
80.
0
100
200
300
,
400
Temperature, C
The TGA, DTA, and the DTG curves are known as the thermograms. Each crude oil produce a
characteristic thermogram that is quite distinctive from the others. Both the DTA and DTG thermogram produces a series of peaks. The areas under the DTG thermogram peaks (Figure 3.4) are proportional to the total weight change of the sample. The conventional thermal gravimetric analyzer can
plot on the same graph both the TGA and the DTG curves automatically. This DTG thermogram serve
as a complementary
piece of information and is easier to analyze for changes from one sample to
next. The DTG curves also tend to be more reproducible than the TGA thermograms and give accurate information of the beginning, the maximum, and end of weight change and is a reproducible fingerprint of the oil. When there are two changes close to each other in temperature, the DTG will
50
usually give two distinct sharp peaks. The TG, DTG, and DTA thermograms all can be used to determine the reaction kinetics. These are discussed briefly in the following paragraphs.
Determination
of Kinetic
Parameters
from
Thermogram
and Fatu, 1976; Reich and Stivala, 1978; Rock, 1978; Dharwadkar et al., 1978; Reich and Stivala,
1980; Dollimore,
from the TGA data. Here we briefly outline the classical Coats and Redfem, 1964, procedure to determine the kinetic parameters. For additional information the reader is directed to the original reference.
The rate of change of weight of a sample (rate of reaction) is related to the reaction rate as:
da/dt=k(l
(3.5)
-a)
where,
a = fractional weight change of the sample
=
(WO- w)/
(WO- Wm)
= t~e
(3.6)
da
-a)
~(1
dT
=/3
(3.7)
Substituting the expression for k from Equation (3.4) in Equation (3.7) we have
()
A x exp ~~
da
a=
(I-a)n
(3.8)
1 -(1
-a)
- ~ ART2
l-n
(3.9)
-[1
- %]exp(=)
Ornl
1 - (1 - a) -
In
[
(1 - n)T2
= In
AR
1 [-
(1
/? E\-~/-RT
2RT)
1-
for n ~ ,
(3.10)
(3.11)
Assuming n = 1, a plot of
-h[-h(:-a)]
slope -E /R. The value of E obtained graphically is substituted in Equation (3.11) to calculate the preexponential factor A.
If n is # 1 then a plot of
wia~;l
correct value of n .This trial and error procedure, though, appear straight forward is not that simple.
Reich and Stivala (1978) modified and simplified Coat and Redferns expression (Equation 3.9) and
presented a computer approach (1980) to determine kinetic parameters.
52
As an alternative to the above we also present the Segal and Fatus (1976) approach to the determination of the kinetic parameters from the TGA thermogram in the following.
Substituting the expression for k from Equation (3.4) in Equation (3.5) we have
da
(3)
dT
(3.12)
( ~ _ a)n
\RT)
(3.13)
da
log
()
constant, i.e.
dt
log
da
~
()
(3.14)
= log K
we get:
Iog(l
- a)
~5~5n
[*)
log K
+
n
1
(3.15)
()T
calculation of both the activation energy E and the reaction order n. The pre- exponential factor A
can be then calculated from Equation (3.4). For further details about this approach the reader is
directed to the original work.
Yoshiki and Phillips (1985) presented the following expressions for calculating kinetic parameters from the DTA data. These expressions are applicable to cylindrical cells and were obtained by
modifying
the original expressions presented by Bae (1972) for spherical cells. The relevant equa-
exp(a tf
)-
exp(a ta)
ATm
~
[a(t.
- %)]
(3.16)
(3.17)
n=l-e~-cz(tf-t~)]
53
(3.18)
atm
ln(~)(l-n)
B = AOexp -
(3.19)
(3.20)
RTO
-)
where:
n=
reaction
order
E = activation energy
AO = Arrhenius pre-exponential factor
~ = time of maximum reaction rate
~ = time of completion of reaction
t.= time of any point along DTA curve
ATM = AT at a maximum reaction rate
AT, = AT at t.
$ =
heating rate
1985; Jha and Verkoczy, 1986; Verkoczy and Jha, 1986). In a nonoxidative
below 500F is due primarily to distillation with thermal cracking becoming the predominant process
at higher temperatures. Only the carbon rich material remain above 1000F. In an oxidative environ-
54
ment the TGA and DSC curves are more complex. At low temperature, low temperature oxidation
competes with distillation and at intermediate temperature combustion competes with cracking.
1600
*
#lk 1400
A
/
n~ 1200
1-
z!
> 1000
g
~
z
3
800
600
n=l.6
Cp = 2.0 BTUAb. F
AMBlENT RESERVOIRTEMP = 80 %
s
400
@=1631bfi
3
MINIMUM FUEL CONTENT FOR VARIOUS FRONTAL
TEMPERATURES WITH LINEAR FRONTAL MOVEMENT
200
o
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
55
Heotrtg
coil
Metol
hose
F
s,
Somple
cell
ceJl
fkf%==-cepu,s
,___
. . . ..
Do
Pressure
pta$e
v
FI
~r----:--------l
Cyltrrdof PR
,
i
:
r6FT
NV
Cooling coil
PI
~ m.?
. -- J
TharmoCouplos
,---le.
---
~ ;
:+
l%-
,---
1
I
,
t
1
:
-1----y
#
t
I
~as
-. ----!
r----
Vent
Trap for
condensatas
Farnacc
c)
-sJlgiPIo
C*II
block
PR
Filter
FI
FiOW
i(rdicotor
Flow transmlttor
Nsedio
VOiVC
Prasmtrs indicator
; PR Pressura rt~ulatar
il
I IAT
cytindsr
cl
_A
Voltaga
rogula?ar
Ilov
60 Hx
56
The DSC is similar to DTA in operation and used to obtain enthalpy (heat of reaction) information and to delineate the various reaction regimes. In the DSC the energy required to keep a sample
isothermal whh a reference material is measured as a function of temperature. Like TGA, DSC also
employ a linear heating rate to simplify the data analysis. In the DSC experiment the amount of heat
released by either the exothermic combustion or the distillation/cracking
mined from the plot of temperature versus differential heat flow between the sample and the reference
(DSC thermogram).
DSC and TGA are complementary thermal analysis techniques and often used
In a typical DSC experiment a known mass of sample is weighted into the sample cell and then
placed on the sample platform of the DSC. A similar empty cell is placed on the reference platform of
the DSC (Figures 3.6-3.7).
mixture is flowed
through at the desired rate. The cells are heated at the desired heating rate over the temperature of
interest. The amount of heat caused by either exothermic combustion or cracking of the oil is determined from the thermogram. The amount of heat released or absorbed by the sample is proportional
to the area under the thermogram. The areas are converted to the corresponding
evolved or absorbed per unit weight of the sample by applying the appropriate instrument calibration
factor.
Shortcomings
of Using
to Evaluate
ISC Parameters
In the last decade procedures were developed to utilize TGA / DSC as an inexpensive and rapid
screening tool for in-situ combustion (Vossoughi et al., 1985; Jha and Verkoczy, 1986; Kharrat and
Vossoughi,
DSC based procedures to calculate rapidly such ISC parameters as: fuel laydown, percentage of crude
consumed
as fuel, combustion
required to sustain
oil saturation
fewer manpower to set-up and operate and the data analysis is less time consuming. The resource
requirements of combustion tube and the thermal analysis techniques are compared in Table 3.1
57
Description
Combustion lbbe
Capital($K)
Lab.Space(ftz)
OperatingExpense($K)
Core (Orams)
ManHours
I TGA
/ DSC
500-1,000
75-150
100-300
10-15
10-50
3,000-5,000
300-500
0.2- 0.5
I
I
2-3
30-50
The major limitation of this approach is that these procedures often employ experimental conditions that are dissimilar to those encountered in a reservoir or in the combustion
results are likely to be affected by experimental variables such as the heating rate, oxygen partial pressure, purge gas flow rate, sample size, etc. Further, TGA and DSC experiments lack the fluid flow
characteristics of the combustion tube test, Niclcle et al. (1987) investigated the effect of experimental
conditions on TGA / DSC generated ISC parameter values, in terms of heat/mass transfer limitations
and reaction kinetics and arrived at the following conclusions:
The heating rate employed in the TGA / DSC techniques significantly affects fuel laydown.
The fuel laydown increases with decreasing heating rate due to increased LTO at the lower
heating rates.
Kinetic parameters derived from TGA / DSC data using a first order kinetic model are far
from rigorous and are dependent on the sample heating rate. Hence the TGA / DSC heating
rate must be consistent with the thermal history of the system being simulated.
Since reservoir and trade oil minerals catalyze fuel deposition, TGA / DSC screening experiments must be performed using reservoir rock and crude,
The error in TGA sample temperature measurement (AT) increases with increase in: oil saturation, heating rate, sample size, purge gas flow rate, and oxygen concentration.
TGA analysis must be carefully designed to minimize AT.
58
Hence
parameters calculated
from
are
affected by experimental conditions, the conditions used in the TGA / DSC evaluations
should mimic as closely as possible, those found in the system being simulated.
Accelerating
Rate
Calorimeter
(ARC)
The traditional thermal analysis tools (TGA, DTA, DSC) are limited to low and medium pressure operation. The ARC is an instrument that has been developed in the early 1980s to study exotherrnic reaction kinetics at elevated pressure (to 10,000 psi) in the temperature range of interest (60
900F). The Arrhenius activation energy, pre-exponential factor, and order of reaction can all be calculated from the ARC data along with the main exotherm. Though both ARC and DSC are capable of
revealing LTO and HTO reactions, ARC is better suited for studying the interval between LTO and
HTO. The following
discussion of ARC is based on the works of Townsend and Tou (1980), Yanni-
maras and Tiffin (1995), and Zelenko and Solignac (1997), supplemented by additional information
made available to the author by Amoco Exploration and Production Technology
Pressure
Transducer
AH
:!
Heater
j
4:
Heaters
Top Zone
/0
Bomb
.
.
,.
~.....
~1/16P
Preasurizetion
/injection Ii ne
Bomb
C)\
ottom
Zone
~Bottom
Zone Thermem.ple
59
(ARC) Set-up
ARC consist of a small (1 in. diameter) spherical sample holder (bomb), into which the reactants are placed (Figure 3.8). The bomb is then placed inside a precisely heated calorimeter. The
power input to the calorimeters
heaters is controlled
between the bomb and the calorimeter walls to near zero. The temperatures are measured with thermocouples placed on the bomb and on the outer walls of the calorimeter which contains the heaters.
The heat flow toward the outside of the bomb varies according to the difference
of temperature
between the bomb and the calorimeter walls. This permit the tracking of rapid exothermic reactions
under strict adiabatic conditions. The key factor in the design and construction of the ARC is the
maintenance of near perfect adiabatic conditions. Once a self heat rate is achieved (limit is 0.02C /
minor higher) the sample completes its thermal spiral adiabatically, with automatic recording of time,
temperature, and pressure data until the exothermic reaction has gone to completion. A pressure transducer attached to the bomb head permit the pressure measurement inside the bomb.
Figure 3.8 shows a closed ARC set-up for kinetic study (Yannimaras, 1998). It consist of the
main unit with the pressure vessel, power unit, thermal control unit, and X-Y plotter. A computer is
used to control the operation of ARC. It also serves to collect and process data and obtain kinetic and
thermal values.
A crude oil sample is usually subjected to closed ARC system testing (i.e., fixed amount of air,
initially at reservoir pressure, over the oil sample and near perfect adiabatic operation). The runs starts
at the initial reservoir temperature. After 15 minutes, if the system does not detect a heat rate greater
than the detection threshold of the calorimeter (usually 0.03-O.04F per minute), the bomb temperature is increased by 10F. This procedure is repeated until a significant rise in temperature due to exothermic reactions between the oil and oxygen is noted.Then, the heat rate is adjusted to achieve. and
maintain adiabiticit y. Finally, when the heat rate drops to near zero, the bomb is heated up again to a
higher temperature to see if reactions start at higher temperature. This is done up to 900F.
The experimental results are generally presented as plots of log of the rate of exothermic heat
release (C/min)
adiabatically, no heat is
iillowed to cross the system boundaries once combustion is underway. Thus, in the ARC plots, the
presence of a trace over a temperature interval indicates a region of exothermic reaction, while the
absence of experimentally recorded points indicates no reaction (Yannimaras and Tiffin, 1995).
60
Compressed
Air Rsexvoir
atm
Gaseous
032 out
9(PF
&
.-
.,
. .
. .
,
-..
Hgh Pressure
!&
With q
Hgh Pressure
Low Fate Pump
~Ote:
R+sistive
SEQarator
CWiing
Heal ihg
TC (rnermOcOufie)
The ARC usually operates as a closed system, within its permitted pressure and temperature
range. This operating mode is satisfactory for obtaining most kinetic parameters associated with the
oxidation of oils. However, since in the reservoir the reaction kinetics are dynamic and affected by
oxygen flux, it is highly desirable to obtain kinetic parameters under flowing conditions. Hence ARC
set-up is often modified to operate it as a flowing system at any pressure up to the systems pressure
limit. A schematic of such a system is shown in Figure 3.9 (Yannimaras, 1998). Flowing ARC studies
indicate that the LTO / HTO kinetics from such a system tend to approach those of closed system at
near zero flow rates.
ARC
Theory
ARC assume a single reaction model. Starting with the Arrhenius rate expression (eqn. 3.5) and
applying the mass
~dc
\dt
the ARC and after algebraic manipulation, the following expression is obtained (Yannimaras, 1998):
dT
dt
@\c N-l(T,
- A P\fi,
o
TO)
[1
$~;
(3.22)
o
61
where
T = sample (bomb) temperature, at any time during an exotherm
t = time
A = pre-exponential factor
R = universal gas constant
CO= initial sample concentration, constant
n = reaction order
T~ = final exotherm temperature
TO= initial exotherm temperature
Defining
k =
k ql
=Aexp
-E
()RT
C~-l and mT ~
(3.23)
By plotting k versus T, on log scales, one obtain n as the slope of best fit straight line.
Also from the definition of k , we obtain
(3.24)
i.e., from the same plot, the activation energy E is determined as the slope of the best fit straight line.
62
Limitations
of ARC Tests
ARC tests permit the rapid determination of oxidation kinetic parameters. The results, however,
are only qualitative due to systems inherent shortcomings. In ARC tests the kinetic parameters (activation energy and order of reaction) are calculated assuming a single reaction model (usually a HTO
reaction model). A single reaction model, however, cannot realistically represent the crude oil oxidation because the oxygen is used not only to form carbon oxides and water (HTO reaction), but also to
form stable oxygenated compounds (LTO reaction). If both reactions occur simultaneously throughout the test, as usually is the case in ARC system, the kinetic parameters cannot be estimated quantitatively. In ARC the kinetic parameters are estimated based on data corresponding
to the highest
temperature (HTO reaction). The assumed model is then extrapolated to lower temperature and the
kinetic parameters for LTO are calculated based on differences in the heat rate between the two temperature regime.
Further, data obtained horn a closed ARC system is not useful for screening the candidate heavy
oil for its ISC potential. This is because in a closed system due to a rapid decrease in oxygen partial
pressure (mostly due to oxygen addition reactions) the reaction temperatures often fail to transcend
the negative temperature gradient region and unable to reach the HTO zone. A flowing ARC device
can avoid this drawback, because the oxygen pressure can be kept constant throughout the experiment. Flowing ARC device is also useful in asseising the role negative temperature gradient plays in
delineating the HTO mode from LTO mode during heavy oil combustion.
device can be used to determine qualitatively the conditions that give rise to negative temperature gradient region during an air injection process.
LTO reactions are not detrimental to the success of combustion process in light oil reservoir,
because in these reservoirs, the recorded combustion front temperature is often less than 600F (Germain and Geyelin, 1998; Fassihi and Gilham, 1994; Metwally, 1991; Pebdani et al., 1988; Struna and
Poettmann, 1986; Anthony et al., 1981). This still correspond to the LTO region. On the other hand, a
knowledge of the reactivity of oil and oxygen at reservoir temperature is essential to ascertain the auto
ignition tendency of the crude. Unfortunately due to their poor heat detection capability at low temperature, neither of the ARC system can answer this question with any degree of certainty (Zelenko
and Solignac, 1997).
Despite its shortcomings, the ARC remains a simple and useful tool to analyze qualitatively the
oxidation reactions of the crudes and to study the oil reactivity at low temperature. Since it track the
63
exothermic reaction adiabatically, ARC permits the rapid determination of kinetic parameters using
only the temperatures as the variable. However, its limitations must be borne in mind when using it to
screen the oils for combustion process.
Effluent
The EGA is a
Gas Analysis
(EGA)
Technique
kinetic parameters. I
Temperature Oxidatj
(RTO) tests. Ramped temperature oxidation tests involve the controlled heat-
ing of recombined o
.urated cores (or oil-sand mixture) in a flow reactor under a flowing stream of
lditions.
The experimer
;et up used for the EGA tests consisted of a flow type tubular reactor (often
e U. S.) with the appropriate instrumentation, control systems and flow loops.
tion details differs from laboratory to laboratory. The schematic of the kinetic
shown in Figures. 3.10 and 3.11 respectively. The schematic of Stanford Uni-
64
Thermocouple
(0.030 in. O.D.. Type f)
Thermoweit(14 in.L. x
1/16in. O.D. X 0.031 in. ID.)
Efowmgas
Manifoldandcross-over
assY.IK
(1/4 X
l/8X
Exterrsb
tube
lulF--Zi%
1/16 ti.)
V4in.~*
h 1-3/16
in.
5/16in. {4
---#-,
4
Coppergesket(1-1/2in.O.D.x
1-3{16
in.I.D.x 1/8in.Thick.) ~~
$i#;
SempIecup(?.I cmL.x
1
L2in.O.D.X 1.049 in. I.D.) ;.
,fi?i:
+
Coppergasket(1-1/2
in.O.D.x #
I-1/8in.I.D.X 1/8 in. f%iCk.)
.,,.
;$
1.2in.I.D.
1.9in.O.D.
T
Knifee6ge
Preheatercoil
(20ftL.x 1/8 in.O.D.)
Injected&
65
HEATERS
THERMOCOUPLE
VW
+EATING
Wxx
\
HEATMG
PLATE(2)
----%,.
iMw.ATlm
MATERIM. \
.MU3!2LE
HEATING
Ho&K
tamlm
@nOw
TWRMOCCWPL&
PORT
M1ODLE
?iEA-nt4G
Biocx
?301-roM
HEATINs3~
8LOCK
FIGURE 3.
66
1.AU
filter
2.Mmflowmntrdler
3.Pre$aure
trmsdacex
4.Kinetic
cell
s.Fdeztric
furnace
6.Combustion
tvbc
7.Prewmjacks!
8.SCWWOI
9.Wawpump
10.Condensw
11.Separator
12.Drierite
tube
13.Micro
fii\er
Camadm
14.GasSowmeter
15.Back-pm.sure
re.gdator w
16.Rotam&ef
relief wdve
17.Pressure
18. Acid wmhlxr
19. 0s.s attta-ssmpler
20. Gas chmtnato~mph
21. Car&m 6i0xidewIaly2u
22. C!artwn InOImxide analyza
Krw
23. Oxygen
mdy7iz
m
16
6
7
18
17
24,
ve
Z-way vafe
The reactor is packed with a pre-mixed oil and sand sample of interest and subjected to a prescribed heating schedule (20 C-400C
heating is continued at the desired rate until their termination or once the prescribed maximum temperature is reached and then held at that temperature for the remainder of the test. The flow of air is
continued until no change in the exit gas concentration could be seen.
Often a second cell is used to serve as a reference cell. The reference cell is packed only with
dry core and subjected to the same heating schedule as the test cell. During the course of a run the
temperatures in the corresponding zones of the active and reference cell are compared to isolate exotherrnic/endotherrnic oxidation events of the oil occurring in the active cell. TWical ramped temperature oxidation test temperature profiles illustrating various oxidation modes for the Athabasca oil
sands bitumen are depicted in Figures. 3.133.15 (Moore et al., 1995 b).
67
I amped Temperature
Oxidation Test
$
.........
... ..+
.. .,
..
Response
.. .;
...
..,,,,,,
.....................
.......
.............
.;.,,.
..,, ,,.,
.,.
.,
..
. .........
-
-- -.-
-.-::::1..
-:
:~~
0[
o
4
Runtlme (hours)
FIGURE 3.13 E:
Showing LTO
. . .
4
l?untime(hours)
.
....!....
FIGURE 3.14
((
68
,4
6
8
Fkmtlme(hours)
10
12
.
.........
FIGURE 3.15 Example of a RTO Temperature Profile Showing HTO Response and
Low Oil Recovery (Courtesy of UNITAR Center, Mehta, and Moore, 1996)
The kinetic parameters are determined from the effluent gas analysis (Mamora et al., 1993;
Burger et al., 1985; Fassihi and Brigham, 1981) as follows:
If mOis the mass of oil present in the sample, we can define a specific reaction rate K as:
Amo,
K= = Aexp
mOAt
.,
(~)
(po,~
(3.25)
where:
~oz
P02 = average
K = mass of the oxygen used up per unit time with respect to the mass of oil mO
R = gas constant
K can be calculated from the effluent gas data.
69
Assuming the o
er of reaction n = 1, a plot of in (K/p ~z) versus lfT will yield a straight line
Combustion
Tube
Tests
Introduction
Combustion tubes are the traditional tools employed in the laboratory to investigate the performance of in-situ combustion processes. A properly designed and operated combustion tube test can
provide much useful information about the test rock/oil systems combustion characteristics. Some of
this data can also be used for making a proper engineering and economic projection of a field tests
performance. Hence laboratory combustion tube studies are the necessary first step in the design of an
ISC project.
Moveable fhwnxzouple
(0.030in.O.D.,TyP?J
J)
ThermoweH(130 cm L. x
1/16 in, O.D. x 0.031 in. I.D.)
Stationary lbermocouples
(9x 0.020in. O.D,, ~
3)
Thennowell (13CIcm L. x
I Ill
Clesn
sad
Igxtiter
0.113
in.
I.D.)
II
-----g
coil ---4
~26.8
} 36.S
~46.8
56.8
Sanp!e
1/16 in, Swa@ok
tube fiuingandpbg
66.8
,~
76.8
Bouom fkmge (4-5/8 in. O.D.
x 0.812 in. Thick.)\
_
86.8
96.8
;hionu?y
IwmOcouples
hwace)%n
kmge
CQPWWsket@W6W~
x 3 ik LKI. x 3f32 in, Thick.)
:
top
{cm)
=-
Produced fluids
70
The combustion tube, shown in Figure 3.16 is a thin walled stainless steel (or some other corrosion resistant material such as Inconel) cylinder housed inside a pressure jacket and packed with
actual core, water and oil from the field under study. The fluid saturations in the tube are tailored to
approximate reservoir conditions.
advanced through the tube by continuous air injection, Temperature, pressure, air injection rate, oil,
water, and gas production and effluent gas composition data are collected and analyzed. From the data
analysis process variables are calculated.
Comments
About
Combustion
Tube Tests
Combustion tubes are unscaled elemental physical simulators and represents a piece of the reservoir simulated at full scale in the laboratory with the constraint of one-dimensional
Slater, 1986). Combustion tube tests permit the simulation of the nature of the propagating combustion front and the resulting dynamic chemical reactions in conditions closely approximating those in a
reservoir. Because the nature of the combustion front propagation is controlled by the chemical reactions and relative permeability characteristics of the reservoir system, combustion tube results can
provide a qualitative feeling for the field performance. It also serve as a mean for estimating the net
effect of the interrelationships between the various mechanisms which affect combustion.
Hence,
combustion tube tests have been accepted as a method of producing reliable data describing the insitu combustion process.
The main drawback of combustion tube tests is that they are unscaled experiments. Hence the
data obtained from these experiments cannot be scaled and direct correlation of combustion
tube
results to the corresponding reservoir is at best tenuous. Experience has, however, shown that as long
as the testing is done with actual reservoir rock and oil under proper operating conditions, the chemical reactions and the reaction stoichiometry in the tube will be the same as those occurring in the reservoir. This is because the stoichiometry of the reactions is controlled by the temperature, pressure,
and chemical characteristics of the oil. The burning behavior of the oil in the tube, however, will not
be the same as those in the reservoir, because it is affected by the flow rate and fluid saturations.
Further, combustion tube studies have little value in predicting bulk fluid movement. This is
because a combustion tube simulates only a differential element of a reservoir. A primary consideration of any recovery process involving fluid injection is sweep efficiency. Although several theoretical models have been proposed to predict the mobility and movements of fluids within a reservoir,
none have been completely successful in fireflooding. Due to the extreme temperature gradients, con-
71
stant changing of fluid phases, and reaction kinetics, a rigorous prediction of flow properties is difficult. Extrapolation
configuration
of combustion
Pilot test data remains the best source of information for sweep efficiencies.
In wet combustion process water and air are injected simultaneously. Three factors are associated with simultaneous air-water injection. These are:
1.
2.
3.
Questions concerning
how the air and. water will distribute within the burned zone and what
conditions will exist when they reach the combustion front cannot be answered from the combustion
tube studies. Gravity segregation is an important consideration and its effect is eliminated in the laboratory by performing
the tests with the tube in the vertical position. Laboratory tests are useful in
determining the optimal water-air ratio (WAR) for wet combustion. Caution, however, should be exercised in using this information for field design purposes because WAR is sensitive to the combustion
tubes heat transfer characteristics.
Although the heat losses from the combustion tube are considerably different from those of the
reservoir, a direct analogy exists between the two systems. The reservoir is naturally insulated by
overburden and underburden and heat losses are relatively small compared to the amount of heat generated. The laboratory tube has high heat losses due to its metal construction and limited amount of
heat generated. To decrease heat losses and sustain a combustion front, strip heaters are placed around
the tube. These heaters maintain an escalated isothermal condition which reduces the temperature
gradient with tube surrounding and lower heat losses. In other words, the heaters simply increase the
overall energy level of the system without interfering with process mechanism. The heat transfer
properties and the subsequent temperature profiles are not distorted by the presence of the heating
system. The heat flow through the packed tube does represent what will happen in the reservoir with
certain exceptions.
Interpreting data for wet combustion is completely dependent on the heat transfer characteristics
of the combustion tube, If heat losses from the tube are excessive and the combustion front is barely
sustaining itself, little advantage is realized by scavenging the heat behind the front. Injected water
72
has a cooling effect on the combustion zone, lowering its temperature and possibly quenching the process. Conversely, if high temperatures are maintained by the strip heaters (insulating system), sufficient heat is available behind the fire front to vaporize all the water injected. In this case, the energy
level in the combustion zone is barely decreased by water and high burning temperatures maintain.
Large quantities of water passing through the combustion zone as steam have no adverse effect on the
combustion reaction. Actual conditions in the reservoir will be somewhere between the two extremes.
The conditions maintained on the combustion tube are believed to be representative of those that will
prevail in the field. However, as mentioned previously, extrapolating water-air ratios from laboratory
to field conditions might be misleading.
The final consideration for air-water injection is if large quantities of water passing through the
combustion zone how it will affect burning efficiency. As previously discussed, combustion tube tests
accurately represent the reaction mechanisms. Therefore, laboratory results will satisfactorily predict
combustion zone behavior.
Experience has also reveals that the tube and field data (such as the hydrogen to carbon (H-C)
ratio, akfuel
produced water, etc.) often reasonably matches well when the projects are operated in the high temperature mode.
Information
ratio (OFR).
73
The last three piece of information, though, highly qualitative are valuable in assessing the tendency of the reservoir to burn in a stable or unstable mode.
Combustion
Description
of Combustion
Tubes
Tube
Test Set-up
Though, the design details of the combustion tube set-up vary from laboratory to laboratory,
they all include four basic components: a combustion tube assembly, heating and data acquisition system, flow control system and fluid analysis system.
Figure 3.12 shows the flow and instrumentation schematics of a typical combustion tube set-up
(Marnora et al., 1993).
The heart of the test facility is the combustion tube. The combustion tube assembly consist of a
thin walled corrosion resistant tube housed inside a pressure jacket. The wall thickness of the tube is
usually less than 0.062 inch to insure uniform heating and to minimize the heat conduction along the
walls of the tube. The pressure jacket is usually fabricated out of carbon steel and designed to withstand the desired operating pressure. The annular space between the tube and the pressure jacket is
filled with porous insulation and pressurized with insert gas to equalize the pressure between the inner
and outer wall of the combustion tube. The pressure difference
always inward.
Combustion tubes vary as to diameter and length. Tubes as large as 8 inches in diameter and 15
ft., in length have been used in the past to simulate the process (Parish and Craig, 1969). Most current
designs, however, employ smaller tubes to shorten the runtime. These tubes are generally 3-6 ft. long
and 2=4 inches in diameter. Table 3.2 show the dimensions of the tube used in some selected facilities.
74
Length (Ft.)
Diameter
(inches)
Wall Thickness
(inches)
Reference
U. of Calgary
0.042
Mehta(1997)
StanfordUniversity
0.016
Mamora(1993)
Amoco
N.A
Yannimaras
(1998)
Texaco
0.042
0.025
Facility
Bousaid
(1987)
Leaute andCollyer(1984)
Larger diameter tubes are preferred to smaller diameter tube because they maintain better adiabiticity and lower fluxes. Belgrave and Moore (1992) investigated the effect of tube design (diameter) on combustion tube performance
and conclude
condition can produce difference results due to differences in heat loss rate from the tube wall.
The laboratory combustion tubes described in the literature use one of two basic designs:
1.
Insulated tubes.
2.
With insulated tubes heat loss is minimized by insulation alone. Operation of insulated tubes
requires high air fluxes, of the order of 70-100
Weijdema, 1963). Such high air fluxes are needed to reduce the residence time of firefront in the ~be,
and hence the heat losses. .Such high fluxes are often difficult to sustain in the tube due high pressure
drops across the core. In addition at high fluxes, air requirement and fuel availability may increase
with increasing flux (Alexander et al., 1962).
Near adiabatic operation should allow combustion front to propagate at lower fluxes. To minimize the heat loss from the tube wall and to maintain adiabatic conditions in the tube, most current
designs employ tube assemblies with compensating heaters. These are narrow width (l2 inch wide)
band heaters mounted along the length of the tube at equally spaced interval and are designed to
match the wall (heater) temperature with the tubes center temperature at each heating zone. A major
concern with the use of heater is how to minimize the possibility of the heaters influencing the movement of the combustion front. Though the problem of thermal interaction between the heater and the
75
sand pack have been looked into by researchers (Leaute and Collyer, 1984; Belgrave et al., 1990w
Belgrave et al., 1990b) no satisfactory solution to the problem exist.
Combustion tube radial heat transfer studies by Belgrave and Moore (1992) indicate maintenance of near zero temperature difference between the sand pack and tube wall not necessarily imply
adiabatic operation conditions. Their analyze indicate combustion tubes of different design operated
under identical conditions can produce different results (temperature profiles) due to differences in
the magnitude of heat losses. Their study also revealed that combustion tube diameter and length of
heating zones play an important role in achieving adiabatic condition.
shorter heating zone length can result in lower heat losses and near adiabtic conditions.
It is necessary to control closely the power input to each of the heaters to minimize axial temperature gradient and lateral heat losses from individual sand section. In every heating zone a pair of
thermocouples,
one fixed to the outer wall of the tube and another inserted through the wall to the
tubes center is used to control the power input to the heater and maintain the wall and sand pack temperature nearly equal. Each heating zone wall temperature is usually maintained at 5F less than the
corresponding center temperature to minimize lateral heat losses and to ensure that the heaters are not
sustaining and/or driving the combustion front.
The tube is generally operated vertically so as to minimize gravity segregation effects. Itis documented that in vertically operated tubes at high pressures, distortion of the test results may arise due
to thermal convection
of gas in the annular region between the tube and the pressure jacket (Smith
and Perkins, 1973). Belgrave et al. (1990a), investigated this problem and found that high operating
pressure, high Rayleigh number annulus gases and use of high permeability annulus insulation likely
to intens~
annular convective circulation and increase heat losses. They recommend the use of low
molecular weight, low specific heat capacity (low Rayleigh number) gas such as helium and a low
permeability (less porous) insulating material in the annulus to minimize convection
induced heat
transfer.
Operating
Procedures
A typical combustion tube run is performed by first packing the tube with the native reservoir
rock or graded sand. If field core is to be used for the test, it is first crushed, homogenized
and mixed
with additional amount of oil and synthetic brine before being packed into the tube. All mixing and
packing is usually done in an enclosed area. A portion of the mixture packed into the tube and the
76
results from the fluid saturation measurements allow calculation of porosity and volumetric saturation
of each fluid phase.
In an alternative packing procedure adopted by some U.S. laboratories, the tube is packed with
crushed core material in 500-g increments. Each increment is packed to the same volume to ensure
uniform porosity. Alternating 500-g sand samples are saturated with methanol to facilitate packing.
Approximately
1/2 in. of 20-40 mesh silica sand is packed at each end of the tube to prevent the pro-
duction lines from plugging by fine migration. Following packing, the tube is saturated with methanol
and thermocouples
are inserted into place. The pack is then successively flushed with pentane and
nitrogen to clean and dry the sandpack. The pack is then saturated with synthetic reservoir brine and
the porosity and permeability established. Crude oils are then injected until immobile water saturation
is obtained. This technique known as the restored state procedure have shown to reproduce the fuel
lay down results better than other packing procedures.
The tube pack is then pressurized with nitrogen to the required operating pressure and checked
for leaks. The pressure jacket is also pressurized simultaneously with an inert gas such as helium.
After being pressure tested, the tube pack is then heated to the reservoir temperature. Nitrogen
once again is flowed through the tube to establish gas permeability and produce mobile fluids. The
produced gas composition is continuously monitored and the displaced liquids (if any) are collected.
After the whole tube had attained the reservoir temperature and no more liquids being produced, the
inlet end band heaters are activated. The inlet sand face temperature is gradually raised to 650F
(500F for wet combustion) while flowing nitrogen to facilitate combustion. The nitrogen flow rate is
gradually increased to the test design air flow rate. When the desired injection end sand face temperature is attained, nitrogen injection is stopped and air flow at the desired operating flux begins. Here the
flux is defined as the volumetric injection rate of air measured at standard condition of 14.7 psia and
60F divided by the cross sectional area of the tube. Ignition is usually observed immediately upon.
switching to air injection. The thermocouples
ture upon ignition. Confirmation is also obtained with effluent gas composition.
T~ical
air flux employed in a 4 in. tube is about 100 Scf/ft.2 - hr. (Oxygen flux 21 Scf/ft.2 - hr).
From time to time the flux is increased to help reduce heat losses, minimize LTO and improve the stability of the burning zone.
77
For wet combustion tests, after fie whole tube is at the reservoir temperature and no more liquids are being produced, the nitrogen flow rate is decreased a little and water injection into the tube is
started, The water flow rate is gradually increased until the Water-Nitrogen
desired Water-Air ratio (WAR) for the desired wet combustion mode (normal or superwet).
Next, with both water and nitrogen flowing, the inlet end heaters are turned on. When the
desired ignition temperature (usually 500F) is approached, the nitrogen flow is stopped and air injection commences. The flow rate of both air and water are then gradually adjusted to obtain the desired
WAR. For a normal wet combustion test, a WAR of 1.1 ft.3 water/Mscf air is recommended.
Data recorded during the run include air injection rates (also water injection rates for wet combustion), injection pressure, temperature at each thermocouple
duced oil, water and effluent gas volumes and rate and produced gas composition.
are analyzed for oxygen, carbon dioxide, carbon monoxide,
hydrogen, hydrocarbons,
hydrogen sul-
fide and nitrogen. The energy input to each heater is also recorded.
All the data (except the volumes) are recorded during the stabilized propagation of the front in
the tube after the initial ignition transient effects had settled down. The stabilized burning period in a
tube is that period during which the combustion front velocity, the air injection flux and the average
effluent gas composition
are approximately constant. This is necessary because the tube pack assume
to represent a reservoir element located at some distance downstream of the combustion front and is
unaffected by transient start-up effects.
After the combustion tube run, the following measurements are usually made to facilitate characterizing the run:
Post test core analysis to provide axial profile of extractable oil, coke, and water.
Composition
and asphaltenes.
78
Interpretation
TEMPERATURE;
800
100
of Combustion
Tube Data
FIRE FLOODING
- TYPE I
TEMPERATURE:F
800
100
FIRE-WATER
FLOODING
(LOW WATER-AIR
RATIO)
TYPE II
TEMPERATURE:F
600
100
FIRE-WATER
FLOODING
(HIGH WATER-AIR
TYPE Ill
DISTANCE
79
Temperature Profiles
RATIO)
1200
1000 2.49
r
o!
hrs
-10 0
20
40
60
80
100
120
t 1
140
FIGURE 3.18 Probe Temperature Profile as a Function of Time for a Dry In-Situ
Combustion Tube Run (After Bousaid, 1987)
Temperature measurements made during a tube run can be used to monitor the frontal advancement of the burn. Depending upon the combustion mode three different types of temperature profiles
are observed as shown in Figure 3.17. The Type I profile is for dry combustion (WAR=O); type II profile is observed where water is injected at relatively low values of the WAR (normal wet combustion)
and type III profile is observed when the tube is operated in super wet combustion mode.
Figure 3.18 shows the typical temperature profile for a dry combustion run in a clean silica sand
pack. The temperature history at 2.49 hour is characteristics of the profile that might be obtained
when an heater is used to obtain ignition. The profile at 5.99 hours is more nearly represent the temperature distribution obtained in a dry combustion run. The temperature rise from initial reservoir
temperature to the peak temperature at the combustion front and drops off sharply ahead of the combustion front to a value corresponding to the steam temperature at the model pressure. The steam zone
which is also known as a steam plateau is characterized by a flat temperature distribution. The temperature of the stream plateau is determined primarily by the air injection pressure. Since in dry combustion only air is injected, the size of the steam plateau developed is likely to be small and depends upon
80
the initial connate water saturation. Ahead of the steam plateau, temperature drops gradually to the
initial formation temperature.
I
\
*1
u;
I
I
,
1-
I
1
Tr
>
The frontal advancement of the combustion zone (combustion front velocity) can be determine
the slope of peak temperature (or a representative high temperature) location (measured from the
injection end) versus time curve. In Figure 3.19 a schematic of the temperature profile downstream
from the temperature peak, T~ is the peak temperature, lJ~ is the combustion front velocity, T, is the
steam plateau temperature, UCis the steam zone velocity and T~is the original reservoir temperature.
For dry combustion the combustion front velocity is a function of air flux and fuel concentration. Combustion front velocity is directly proportional to the amount of oxygen supplied (air flux)
and inversely proportional to the amount of fuel laydown at constant air flux. Thus the minimum rate
of advance is fixed by the amount of fuel that must be burned at the minimum air flux. The minimum
air flux in any given situation is the lowest flux that will sustain and advance the combustion front.
The minimum air flux and the corresponding
design and economics
81
the combustion
-----
[000
IPEAK:l?ONT
TEMPERATURE
.-.
1..--
800
;-n
i
~
L-J
I&
.
.\
&j 600
+4
5
Q
$ 400
FINAL TEMPERATURE
ii
*
Zotl
/
a
20
40
60
100
80
i20
FIGURE3.20
Effect of
Native reservoir rock contain clays and minerals and often contribute to increased fuel laydown.
Hence combustion front velocity in runs performed using native core materials are usually slower
than when performed using clean sand under identical run conditions. This is clearly evident from the
native core run temperature profiles (Figure 3.20) where the temperature drops less sharply compare
to
Figure 3.18.
+L
TB
I
I
I
Lib
I
II
I
I
I
I
I
I
Tr
e
x
In the case of normal wet combustion (type II profile) water is injected at relatively low values
of the WAR. The temperature profile (Figure 3.21) is similar to the dry combustion except that the
temperature decreases much more rapidly behind the combustion front and the steam plateau is elongated by the additional heat transported downstream. The steam front advance at a much slower but at
a uniform rate than the combustion front. In a stable wet combustion tes~ the combustion front usually progresses much more quickly than the vaporization front (trailing edge of the combustion zone
where the water vaporizes). The maximum temperatures attained in the combustion zone are nearly
independent of W~
maintained at the steam plateau temperature. In this profile the temperature increases from reservoir
conditions, levels off at steam temperature, and then decreases. No temperature above the steam plateau level are observed. In this mode both the combustion and the vaporization zones progresses at
nearly the same rate.
Analysis
of Combustion
Tube Data
As mentioned earlier, combustion tube tests provide many useful information about influence of
operating parameters on the process. Data obtained from a tube run include the combustion front temperature, the composition
and volume of the effluents, the amounts of oil and water produced, etc.
Process parameters such as the fuel deposition, air requirement, etc., are then calculated from these
data.
In the following,
procedures for analyzing the combustion tube data are detailed. These proce-
dures are made available to the author by Dr. Garden Moore and Dr. Raj Mehta of University of Calgary, Canada. The author wish to express his gratitude for their help.
High
Temperature
Combustion
Stoichiometry
There are many chemical reactions that occur in the tube during combustion. These include low
temperature oxidation of oil, thermal cracking or pyrolysis, and high temperature oxidation. Even
though significant LTO and pyrolysis reactions do occur in the tube, generally only the HTO reaction
is assumed to represent the process and used to analyses the combustion tube data. In this section we
present combustion tube data analysis based on HTO stoichiometry. In a subsequent section modification of the procedure to account for reactions other than the assumed HTO are presented.
83
The basic chemical expression that describes the combustion of coke can be written as:
CXHY+a Oz+RaNz
i.e.
R =
--=-bCOz+d
CO+ fOz+j
HzO+RaNz
(YIN, \
(3.25)
(3.26)
where R is the ratio of mole fraction of nitrogen to oxygen in the feed gas. (a, b, d, f, j are stoichiometric coefficients).
Performing an element balance we have,
[Carbon]
x=b+d
[Oxygen]
a=
&Iydrogen
y=2j
b+;
4 ab~f
(3.27)
+f+J
(3.28)
(3.29)
(3.30)
= ~
x
(3.31)
(b+ d)
2. The Oxygen-Fuel ratio is the minimum volume of oxygen required to burn a unit mass of fuel
which has an equivalent atomic H/C ratio given by Equation (3.30)
(a) x (moles
Oxygen /Fuel Ratio =
02)
(3.32)
moles fuel
(12.011 x + 1.008 y)
Where 12.011 = atomic mass of carbon and 1.008 = atomic mass of hydrogen
84
(3.33)
Assuming that the oxygen consumption is measured in standard volume and the fuel in kg. (or
lb~) per unit volume of formation we have:
02
F
23. 64a
=
(12011x
m3 (ST)
(3.33a)
/ g
1008y)
or
379 a
(12.011x
(3.33b)
1.008y)
fiin
In Equation (3.33a) the constant 23.64 refers to the volume (in standard cubic meter) occupied
by one mole of gas at the reference condition and the constant 379 (in Equation 3.33b) refers to the
volume (in standard cubic feet) occupied by one mole of the gas also at the reference condition.
3. Air/Fuel
Ratio: It is the volume of air required to bum a unit mass of fuel and is a function of
the amount of carbon and hydrogen in the fuel and the nitrogen-oxygen
23.64 (1 + R) x a
(12.011 x
379(1
r
(12.011x
1.008 y)
(3.34a)
3
ST) kg
+ R)xa
+
1.008y)
(3.34b)
Cf lbm
()
b-;
b2R
(3.35)
(a -f)
85
R=79/21
(3.36)
=3.76
Evaluation of Combustion
,,,
Assuming normalized gas composition the combustion parameters can be calculated in terms of
stoichiometric coefficients a, b, d, and f as follows:
The numerical values of a. b. d. and f are:
N,
[1
a=
f= [OJ
b = [COZ]
d= [CO]
where [ ] denotes normalized composition in mole percent.
Thus:
4(Y
-[co,]- q
(3.38)
c=
N,
4u
-[%1)
(w]
+ m)
co, -qo,]
[1
2
.-. H/C =
([co,]
-1-[co]
86
23.64
O,/Fuel
N2
L-1
379:
( R
[WI-[4
[1
23.64(1 + R)%
Air/Fuel
=
12.01 I([c02]
+ [co])
N,
+ 4.032 L-1
(
-co2-
[co] [02]]
[1
2-
,2
fa, where
_([co2~9
N,
,,
([4+!3
02= (9-[021)
N,
Fraction Oz utilized
Y=
[1
02
[1
[1N,
Excess Oxygen
(1 -Y)
Y
87
=1-R
02
N,
[1
CXHY
+CZ02
+[Iqaiv,
+ bco +dco +fo, +jH20+[qaN2
2.
[Rl~2
CHY,X + ~Oz +
x
3.
-&q
x
[R]YZN 2
+%O+Z02 +:11, o+y
x
-b
; =(b+d)=&
dd
=
X
(b+d)=(nz~l)
Jyn=
;=2X
~=b
xx
Where
(f\
l(d\
f
l(j\
+ ++
2\x)
x
21X)
is calculated from the overall oxygen utilization (Y) using the relationship:
\x)
y=(-f)
a
..~=ql-Y]
.
xx
88
lm
a
In
..=
.
x
y ( m+l+2(m+l)+Z
n\
=l[2m+l
y\2m+2
+ T)
+E(2m+1+2\N2+
Crfn
+( Zm+l+:)o
{2m+2
4)
m+l
R(2m+l
~\2m+2
Y\2m+2
41
~o+(l-Y)/2m+l+~)o
J%02 +
m+l
~flHo
\2m+2
4)
+?JN
4)
This form is the combustion equation originally proposed by Benham and Poeltmann (1958) is
useful for evaluating the product gas composition
379 ~
1. OJFuel Ratio =
(12.011 + j)08:)
379(2m+l
Y 12m +2
.-.O.-JFuel
=
2. Air/Fuel
+q
4) ~C~,lbm
(12.011+ 1.008n)
= (l+R)(O@el
Ratio)
89
(~+d/x\
L,, =
\x
2)
(f?_L\
\x
x)
J(
nz+l
1 \
2\m+l)
)
~(2m+l
+lz\*y
Y\2m+2
4)
(2m+l\
\2m+2)
fO,. =/2m+l
~?\
\lm+2
4)
4. Fractional OL utilization = Y
5. Excess Air=
[1 -Y]
Y
90
Examples
from
In the following
of Combustion
Parameter
Calculation
Typical Product
Gas Composition
culation from the produced gas for various combustion mode (dry, wet, superwet, etc.)
Consider the following product gas composition which corresponds to the average composition
during the stabilized portion of a dry combustion tube test on a Louisiana Heavy Oil Core.
Mole Percent
CO-2
11.2
11.3
co
3.8
3.8
02
4.4
4.4
N2
80.0
80.5
Hz
0.1
cl+
0.43
H@
0.07
Component
100.0
100.0
The Composition of the feed air was 21 % oxygen and 79% nitrogen, hence:
.2
0.79
E=
Y02
H/C Ratio
s 76
*
91
. 4
~[ll.3]@l-
4.4]
1
[(11.3)+(3.8)]
379
O#?uel
Ratio
UN,
R
N2
[co]
[1
12.011([coZ] + [CO]) + 4.032 ~
- [COZI - ~
[~zl
379[80.5]
376
[80.51
_ [~~.3]- y
12,011( [11.3] + [3.8]) + 4.032
3.76
[
=41 .25 scf/lbm
Air@uel Ratio
[1+ R]
[0, /FuelRatio]
92
- [4.41
f.,=
M-[*J
8051 _ [4.4
3.76
= 0.776
Oxygen Utilization:
uR -02
[1
N2
Y=
N,
[80.5]
44
1701
=376-
-pq-fi
3.76
= 0.795
[1 -Y]
=
Y
Excess Air
([w]+[a)
[co]
[1 - 0,795]=
0258
0.795
Ratio =
([11.3] +[3.8]) = s 97
[3.8]
Ratio
_ ([11.3]
([C02]+[COI)
Nz
[1
+[3.8]) = 0188
[80.5]
93
Oxygen
Nitrogen
78.00
Argon
1.00
21.00
3.71
UNNORMALIZED
3.60
3.65
02
0.00
0.00
N2
81.08
82.16
H2
0.20
0.00
H2S
0.92
0.00
0.20
0.00
100.OO
100.00
(Scf/lbm)
H/C Ratio
1.37
Ratio (m3(ST)/kg)
2.19
35.1
10.32
165.4
Air (%)
(C02+CO)/C0
Ratio
(C02+CO)/N2Ratio
I
TOTAL
Ofiuel
14.19
co
14.00
NORMALIZED
72.38
100.00
0.00
4.89
0.22
94
for a ATHABASBCA
5520 KPa
Oxygen
21.00
Nitrogen
78.00
Argon
1.00
3.71
Unnormalized
Normalized
CO2
8.07
8.09
co
2.10
2.10
02
0.05
0.05
N2
89.59
89.76
Hz
0.00
0.00
cl+
0.19
0.00
H2S
0.00
0.00
TCWAL
100.00
100.00
(Scfflbm)
H/C Ratio
5,88
3.13
50.2
14.74
236.3
37.81
Reacted 02Utilization
02Utilization
(%)
(%)
Ratio
(C02+CO)/N2Ratio
37.89
99.79
0.21
4.84
0.11
95
at
for a Athabasca
at 2760
Unnormalized
Normalized
co~
5.83
5.86
co
1.58
1.59
02
0.87
0.87
N2
91.25
91.68
H2
0.00
0.00
0.47
0.00
H2S
0.00
0.00
TOTAL
100.00
100.00
(Scfflbm)
H/C Ratio
9.22
3.68
59.0
17.35
278.1
26,95
ReactedOLUtilization
(%)
27.94
OzUtilization(%)
96.46
3.67
Ratio
4.69
(COz+CO)/NzRatio
0.08
(C02+CO)/C0
96
for a ATHABASCA
Run (Prior
to stall)
21.00
Nitrogen
78.00
Argon
1.00
3.71
J
Normtilzed
CO2
15.06
15.18
co
2.44
2.46
0.00
0.00
N2
81.72
82.36
Hz
0.00
0.00
Component
0.75
0.00
H2S
0.03
0.00
TOTAL
100.00
100.00
(Scfflbm)
H/C Ratio
1.31
2.23
35.7
10.51
168.5
02 Utilization
(?ZO)
Ratio
(C02+CO)/N2Ratio
73.99
73.99
100.00
0.00
7.17
0.21
97
~pical
Analysis)
21.00
Nitrogen
78.00
Argon
1.00
~,,
Unnormalized
Normalized
CO2
5.82
5.82
co
1.25
1.25
1.97
1.97
N2
90.96
90.96
Hz
0.00
0.00
cl+
0.00
0.00
H2S
0.00
0.00
TOTAL
100.00
100.00
I
I
(Scfflbm)
H/C Ratio
9.09
Oz/FuelRatio (m3(ST)/kg)
3.87
62.0
18.23
292.2
26.32
28.62
Reacted 02 Utilization
02 Utilization
3.71
(%)
91.96
(%)
8.75
21.00
Nitrogen
78.00
Argon
1.00
~,
3.71
Unnormalized
CO.2
Normalized
79.45
83.27
8.86
9.29
0.27
0.28
l----+-
6.83
7.16
0.80
0.00
3.38
0.00
0.41
0.00
100.00
100.00
co
t----=-
I
I
(Scf/ibm)
1.81
2.40
38.5
2.53
40.6
67.64
ReactedOzUtilization(%)
67.79
0~ Utilization(%)
99.78
0.22
(co~+co)/co
Ratio
(C02+CO)/N2Ratio
9.97
12.93
99
Modifications
of Equation
Other Than Assumed
High
In this section, we present combustion tube data analysis for reactions other than high temperature oxidation. These reactions other than high temperature oxidation. These reactions include: low
temperature oxidation (LTO), carbonate decomposition,
(a)
LTO Reactions
+(R)NZ + AO,a
where
~2
4(a(l-A0,
Hy
(b+d
E=;=
([
H4
=
+0-
1)
A2)-[co21-y
(b)
)-b-~-f)
[co] -[02]
([C02]+[CO])
Carbonate Decomposition:
ko~
Hy
4(a-b(l-ACO)-~-f)
b+d
E=;=
100
~
4(q-[c~211-Aco2)
=
c
([COJ+[CO])
(c)
-[~21)
4 a+~-b-~-f)
(b+ d)
HR
=
4([N2]) +j-[co2]-~
z
[co]
-[02]
([C02])+[CO]
Laboratory experiments conducted under isothermal conditions suggest that up to 0.3g O~g initial oil can be chemically
be exactly described, but it appears to be in an immobile and non-extractable fraction which we normally designate as coke.
The H/C ratio of the oxidized component is generally unknown, but for the purposes of illustration assume an H/C of the original oil.
Hence, for an oil having an original H/C of 1.5 but containing ~ oxygen content of 0.3g O~g
oil, based on 1 gram of original oil.
(lg)
Atoms: carbon
Atoms: hydrogen
lgatomC
()gmolfiel
= 0.0739
(12.01 1+ 1.5(l,008))g/gmol
lg)(l%::)
(12.01 1 + 1.5(1 .008))g/grnol
101
~o,,09
o.3g
Atoms: oxygen
. .2gamzs = 00188
gmol
32g/gmol
.~=1.5
.
x
Z
0.0188 = 025
; = 0.0739
When this term is added, it will cause the apparent the H/C ratio to increase.
102
Example
Calculation
to Illustrate
Combustion
of an Oxidized
Fuel
Basic Stoichiometry:
f +~H20+
~ ~COT + ~CO + Oz
xxx
Again letting:
n=
~
x
H. C. ratio
CO,
Co=z
=
x
m+l
1
=
m+l
J=l[y\
2\x)
x
L=
x
~(1-Y)
x
103
R@)N
\x)
2
Solving for
a
lm
1(
=.
_
x
Y ( m+l+~~m+l
n _Jz\
1
) z
a
1 2m+l
n _~(~\
=
x
Y ( 2m+2+Z
2(x)
Hence, per mole
(gmol
2\x)
Basic Relationshi~
Component
co~
m+l
co
1
m+l
(l-Y)
o~
2m+l
l(z~
( 2m+2+Z-?\;)
Nz
2m+l+~
)
~(z~
2m+2
4-2\x) -)
Total moles
m+l+(m+l)
)+{l-Y+R\
~
1+( lY+R\
\
2m+l
l[z)
) ( 2m+2+~-~\x)
2m+l
} ( 2m+2
-)
+fi?l~(z]
2ix)
104
\
ters which do not account for the oxidized nature of the fuel), consider the following examples:
Oxidized Components
Formula
1.
Formic Acid
HCOOH
2.
Acetic Acid
CH~COOH
3.
Perbenzoic Acid
CcH~COzOH
(Peroxybenzoic
4.
Acid)
Oxalic Acid
HOOCCOOH
Have assumed:
Y=% Oz utilization=
(w)
=~=10
2.
Y., = 20.946%
YN2
= 78.084%
YA,
= 0.934%
3.
m3(ST)
kmol
in air/fuel calculations.
105
100%
Product
Gas Composition
Combustion
and Conventional
Combustion
Parameters
Corresponding
to the
of an Oxidized Hydrocarbon
Example Calculation
Ratio
(mole %)
Oxygen
20.946
Nitrogen
78.084
Argon
0.934
100
10
Name
Formic Acid
Acetic Acid
PerbenzoicAcid
Oxalic Acid
Peracetic Acid
Formula
HCOOH
CH3COOH
C6H5C020H
HOOCCOOH
CH3C03H
Carbon
Hydrogen
Oxygen
H/c=y/x
0.857
o/c=z/x
0.429
1.5
Atoms
106
Component
Formic Acid
Moles
Acetic Acid
Moles.
PerbenzoicAcid
Moles
Oxalic Acid
Moles
Peracetic Acid
Moles
CO2
0.909
0.909
0.909
0.909
0.909
co
0.091
0.091
0.091
0.091
0.091
o~
0.000
0,000
0.000
0.000
0.000
N2
1.694
3.558
3.558
0.763
2.626
A,
0.020
0.043
0.043
0.009
0.031
Total
2.715
4.601
4.601
1.772
3.658
Composition
Component
Formic Acid
Mole %
Acetic Acid
Mole Yo
PerbenzoicAcid
Mole
Oxalic Acid
Mole
Peracetic Acid
Mole
33.49
19.76
19.76
51.31
24.85
co
3.35
1.98
1.98
5.13
2.49
o~
0.00
0.00
0.00
0.00
0.00
N2
62.42
77.34
77.34
43.04
71.80
A,
0.75
0.93
0.93
0.51
0.86
100.00
100.00
100.00
100.00
100.00
CO2
Total
$%0
107
90
Yo
Conventional
Combustion
Parameters
Apparent H/C
Ratio
-2.00
-3.000
-1.000
O#uel Ratio
1.075
0.538
1.514
5.084
2.544
7.158
(m3(st)/kg)
Air/FuelRatio
(m3(st)/kg)
InjectedOz to
Coxs (%)
210.000
ReactedOz to
Coxs (%)
210.000
0~ Utilization
466.667
135.484
466.667
135.484
100.000
100.000
(%)
100.000
100.000
100.000
0.000
0.000
0.000
(co~+co);co
11.000
11.000
11.000
Ratio
(co~+co)/N~
-i
0.590
1.311
Ratio
0.381
108
Oxygen
94.00
Nitrogen
6.00
0.00
Argon
0.06
Component
(mole Yo)
Unnormalized
(mole %)
Normalized
co~
75.76
76.70
co
14.24
14.42
o~
0.50
0,51
Na
8.27
8.37
H2
0.15
0.00
cl+
1.07
0.00
H#
0.00
0,00
Total
100.00
100.00
109
H/C Ratio
2,05
2.42
63.97
64.22
02 Utilization
99.61
(%)
0.39
(C02+CO)/C0 Ratio
6.32
(C02+CO)/N2 Ratio
10.88
110
(38.79 scf/lbm)
Abnormal
21.00
Oxygen
Nitrogen
I
I
Argon
R!,
78.00
I
I
Component
co~
(mole %)
Unnormalized
1.00
3.71
(mole Yo)
Normalized
30.00
30.15
co
1.50
1.51.
O*
0.00
0.00
N2
68.00
68.34
H2
0.00
0.00
cl+
0.40
0.00
H2S
0.10
0.00
Total
100.00
100.00
111
H/C Ratio
-1.58
1.32
AirlFuelRatio(m~(st)/kg)
InjectedOzto Coxs
167.96
(Yo)
167.96
0~ Utilization(%)
100.00
0.00
(C02+CO)/C0
Ratio
21.00
(C02+CO)/N2
Ratio
0.46
112
(21.16 scf/lbm)
Analysis
of Air and Fuel Requirements
For Combustion
Tube Tests
The following section provides actual combustion tube data from the University of Calgary Test
No. 115, AOSTRA
Test No. 29. This was a dry combustion tube test which was performed at a pres-
sure of 2760 kpa using normal air on Athabasca Oil Sands core.
It should be noted that the calculations presented are based on combustion tube data. The procedures outlined are the same as would be used for a field project for evaluating the combustion parameters and amount of fuel consumed. Direct calculation of the air requirement and fuel requirement for
a field project is not normally possible due to the lack of direct information on the volume of the
burned zone or on the velocity of combustion front. Post-bum coring and temperature observation
wells can provide this information, but these data are not normally obtained. Another difference
between field and laboratory calculations is the uncertainty associated with relating the air injected at
a given injector to the gas produced at a given production well which is in communication
with that
injector. White (1983) describes the calculations normally applied to the prorations of produced gas
back to the different air injectors.
The calculations presented relate to the following parameters:
1.
Overall Parameters.
2.
Stabilized Parameters.
a.
b.
on Stabilized Portion of
Test.
Storage of air in the swept zone or the chemical storage of oxygen in the hydrocarbon has been
neglected in the calculations presented. This is the normal procedure for normal air tests. Storage
effect however, must be accounted for in high oxygen concentration air tests, particularly at high pressures.
113
Analyzing Combustion
Thbe Data
22593g
5355g
669g
0.367 m3(ST)/h
Nitrogen
3.606
Oxygen
0.010
C02
0.670
co
0.177
4743 g
0.0143 m
Volume of swept
4743g
x 100 = 88.690
114
A.
Overall Parameters
Amount of Carbon Burned:
Moles C
Moles COZ
C).670M3~)
23.6445m
Moles CO
ST)
kmol
= 0.007kmol
(.).177m3~)
23.6445m
Moles C
= 0.028kmoZ
ST)
kmol
Moles COz+Moles CO
0.0325 kmol
12 kg/kmol
0.420 kg
Amount Oz Consumed:
Moles 02 consumed
[4.440x0.21 - 0.0101m3(ST)
23.6445 m3(sT)
kmol
0.039 ho]
115
Hydrogen Consumption
MolesCO
2
[0.039-0.028-
0.008 kmol
~]
Mass H2 reacted
0/016 kmol X -
2kg
kmol
= 0.032 kg
kg
2(0.0080) kmol X 18
kmol
0.288 kg
0.452 kg
0.016kmolHzx ~w~~~
Atomic H/C =
0.035kmolCx
116
Ikatom
= 0.91
lkmol
Alternate Calculation
Atomic H/C =
2.0kgHz
0.032x12
Atomic H/C
lkmolll,
Ikmol
kmolC x katoms
= 0.91
0.420
Fuel Requirements:
FR=
MassFuel
VolumeojSweptCore
Fuel Requirement =
0.452kg
0.0143m3
= 31.6 kg/ins
Air Requirements:
AR=
VolumeojSweptCore
4.440m3(ST)
0.0143m3
=
310 ms(ST)/ms
kg
100kgsand
117
On this basis:
Fuel Requirements
0.452kg
22.593kgsand
0.0200kgfuel
kgsand
Fuel Requirements
= 2.00
kgfiuel
100kgsand
B. Stabilized Parameters
Based on the rate of advance of the 500C leading edge [0.75 to 11.75 hours]:
Front Velocity
U~
0.144m/h
The air flux (as measured at the inlet to the tube) was:
AF
0.367m3(ST)/h
7.767 x10-3m2
47.3 ms(ST)/m2h
Hence the air requirements based on the air injection flux is:
47 ~ WZ3(ST)
Stabilized AR
mh
0.144m/h
= 32~m3(ST)
m3
which is slightly higher thabn the overall value for the air requirement, which was
~lo
?7Z3(ST)
m
118
Good agreement indicates that run was stable which is consistent with the long period of stabilizing
burning.
Two methods can be used to evaluate the stabilized air and fuel requirements and the stabilized combustion parameters.
Method 1:
This method is based on Numerical Average of Product Gas Composition and Injection Flux over the
Stabilized Period.
The average product gas composition during the stabilized period was:
Compound
Mole %
Normalized
(Mole Yo)
CO2
14.47
14.65
co
3.67
3.72
o~
0,21
0.21
N2
80.43
H2
0.27
cl+
0.85
H2S
81.42
100.0
100.0
.,
H/C Ratio
1.14
Ofluel
2.15 m3(ST)/kg
Air Requirement=328
Ratio
Air/Fuel Ratio
119
AirReq.
Air / FuelRatio
ms(ST/ms
02 Utilization=
328m3(ST)/m3
99.0%
10.11m3(ST)/kg
(co, + co)= ~ 94
co
32.1 kg/m3
The advantage of this method (based on average product gas analysis) is that it is independent of
product gas flow metering equipment. However, it can only be used if the product gas analysis is relatively constant.
Method 2:
This method is based on Incremental Production Rate:
For the time period from 0.75 to 11.75 h, cumulative production of product gases was:
Component
N2
3.123
o~
0.132
0.008
0.0003
co~
0.579
0.0245
co
0.148
Mass C
kmol(l)
Volume[m3(ST)]
0.00626
m3/kmol.
(Moles Carbon)
(12kg\
\kmol)
(0.0245 +0.00626)kmoles
120
(%)
(12kg\
~=)
Mass C
0.369 kg
Moles Oz Reacted
(Moles Oz Injected-
Moles Oz Produced)
pdesy
MolesOzProduced
0.132bnol
1(0.78)
- o.0003bnol
(0.21)
Moles Oz Reacted=
0.0352 krnol
(Mole@2reac~ed _ &folesco2 -
(
0.132kmol
- 0.0003kmol
( 0.78)
1-
\o.21)
2kg
= 0.0303 kg
kmol
Mass Fuel
(0.0303 + 0.369) kg
0.3993 kg
N+n(m))
(1 + !?2)
(0.79)
.. Air/Fuel Ratio
\3.71)
o.3993kg
121
0.21)
ol~cO)
(3.123rn3(ST))(1 + 3.71)
(3.71)(0.3993)
;3 nz3(sT)
kg
Volume of Core Swept
0.0123 m
(volu~e~J
(4.71)
(3.71)
0.0123m3
3.965nz3(SZ)
0.0123nz3
322 m (ST)/m3
MassFuel
Fuel Requirement
VolumeSwept
0.3993kg
0.0123m3
32.5 kg/m3
.,
Atomic H/C
i14assHz
12kg
lkmolC ~ Ikatondl
Stablized
(co, +co\
\co)
122
lkgH
(0.579nz3(ST) + 0.148nz3(ST))
0.148m3(ST)
4.91
v.hmeN~
Stablized Oz utilization
volu~.o
3.71
( VolumeN2 \
3.71
2,
(3.123 _ ~-oo8)m3(~~)
\ 3.71
/3123\m3(ST)
\ 3.71)
=
0.990
c~+~(2m+1+~\02+g(2m+1+~\N2
Y(2m+2
~co,
ln+l
Y\2m+2
41
2m+l+:\o
~o+(l-Y)
+J-
m+l
41
2m+2
4)
+: Ho+~(2m+l+:\N
22
Y(2m+2
(C02)
Since m =
(co)
m+ 1 =4.91;
4)
(C02
+co)
_l
co
m= 3.91
n = H/C = 0.98
Y = fraction Oz utilized = 0.990
(2m+l
\2m+2
n\
2)
123
1143
l(2m+l
~ =1.155
~(2m+2+;)
R ~ 2m + 1 + ~~ = (3.71)(1.155)=
?\2m+2
4)
m
3.91
= 0.796
m+l
=4.91
=4.91
m+l
4.284
= 0.204
l- Y(2m+l+~)_oo12
Y
=
n y
2
CHOg~+l .1550Z+4.284
\2m+2
4)-
= 0.49
NZ-0.796 COZ+0.204 CO
124
REFERENCES
1.
Abu-Khamsin,
S.A., W.E. Brigham, and H.J. Ramsey. 1988. Reaction Kinetics of Fuel For-
16.
2.
Adegbesan, K.O., J.K. Donnelly, R.G. Moore, and D.W. Bennion. 1987. Low TemperatureOxidation Kinetic Parameters for In-Situ Combustion Numerical Simulation. Sol. of Petroleum Engineers I, Vol. 2, No. 4, November, pp. 573-582.
3.
A1-Saadon, F.T. 1970. An Experimental and Statistical Study of Residual Oil Saturation
After Gas, Water, and Steam Drive and Fuel Availability for the In-Situ Combustion Process. Ph.D. Dissertation, University of Pittsburg, Pennsylvania.
4.
Alexander, J.D., W.L. Martin, and J.N. Dew. 1962. Factors Affecting Fuel Availability and
Composition During In-Situ Combustion. J. Petroleum Tech. Vol. 14, No. 10, October, pp.
1156-1164,
5.
Trans.
Anthony, M.J., T.D. Taylor, and B.J. Gallagher, 1981. Fireflooding a High Gravity Crude in
a Watered-Out West Texas Sandstones Paper SPE 9711. Presented at the 1981 Permian
Basin Oil and Gas Recovery
6.
Babu, D.R. and D.E. Cormack, 1984. Effect of Low Temperature Oxidation on the Composition of Athabasca Bitumen Fuel. Vol. 63, June. pp. 858861.
7.
Bae, J.H. 1972. Determination of the Kinetic Parameters from Differential Thermal Analysis. J. Thermal Analysis. Vol. 4, 1972. pp. 26169.
8.
Baena, C.J., L.NI. Castanier, and W.E. Brigham. 1990. Effect of Metallic Additives on InSitu Combustion
DOE/BC/14126-26,
9.
August.
Belgrave, J.D,M., R.G, Moore, and D.W. Bennion (1990a), The Thermal Behavior of Vertically Operated Near Adiabatic In-Situ Combustion Tubes, J. Pet. Sci. and Eng., Vol. 5 (l),
pp. 51-65.
10. Belgrave, J.D.M. and R.G. Moore (1992), A Model for Improved Analysis of In-Situ Combustion Tube Tests, J. of Petroleum Science and Engineering. Vol. 8 (2), pp. 7588.
125
11. Belgrave, J.D.M., R.G. Moore, M.G. Ursenbach, and D.W. Bennion, 1990b, A Comprehensive Approach to In-Situ Combustion Modeling.
orimetry Studies of Heavy Oils and Their Relevance to In-Situ Combustion. Paper No. 50,
Presented at the 4~ UNITAR Conference on Heavy Oil and Tar Sands. Proceedings, Vol. 4,
pp. 761-773.
13. Benham, A.L. and F.H. Poettmann, 1958. The Thermal Recovery Process An Analysis of
Laboratory Combustion Data. Pet. Trans of AIME, No. 213, pp. 406408,
14. Bousaid, 1.S. and H.J. Ramsey, Jr. 1968. Oxidation of Crude Oils in Porous Media.Soc.
Petroleum Engineer Journal, Vol. 8, No. 3, June, pp. 137148.
15. Bousaid, 1.S. 1987. Oil Recovery
Paper SPE
16739 Presented at the 62 Annual Technical Conference and Exhibition of the Society of
Pet. Engrs. Dallas, Texas, September 2730.
16. Burden, C. and C. Gadelle, 1977. Essaide de Laboraroire Pour L Etude de la Combustion
In-Situ. Paper Presented at the French Sovient Symposium
on Enhanced CM Recovery,
Moscow (May).
17. Burger, J.G. and B.C. Sahuquet. 1972. Chemical Aspects of In-Situ Combustion Heat of
Combustion and Kinetics. Sot. Pet. Eng. J. Vol. 12, No. 5, October. pp. 410-22.
18. Burger, J.G. 1976. Spontaneous Ignition in Oil Reservoirs. Sot. of Pet. Eng. J., Vol. 16, No.
2, April. pp. 73-81.
19. Burger, J.G., P. Sourieau, and M. Combaranous.
Data,
126
23. De 10SRios, C.F., W.E. Brigham, and L.M. Castamier, 1988. The Effect of Metallic Additives on the Kinetics of Oil Oxidation Reactions in In-Sits Combustion. U.S. DOE Report
DOE/BC/141264
1978. Evaluation of
Curves, Thermochimica
372-75.
25. Dietz, D.N. and J. Weijdema, 1963. Wet and Partially Quenched Combustion. J. Pet. Tech.
April. pp. 411-15.
26. Dollimore, D. 1980. Thermal Analysis in Chemical Processing in National Bureau of Standard (NBS) Special Publication No. 580. Proceedings of the Workshop on the State-of-theAt of Thermal Analysis. May 1979. Issued May 1980. pp. 131.
27, Drici, Q. and S. Vossoushi, 1985. Study of the Surface Area Effect on Crude Oil Combustion by Thermal Analysis Techniques. J. Pet. Tech. Vol. No. 4, April. pp. 73135.
28. Fassihi, M.R. 1981. Analysis of Fuel Oxidation on In-Situ Combustion Oil Recovery, Ph.D.
Dissertation, Stanford University, Stanford, California.
29. Fassihi, M.R., H.J. Ramsey, Jr., and W.E. Brigham, 1980. The Frontal Behavior of In-Situ
Combustion. SPE Paper 8907, Presented at the 50ti Annual California Regional Meeting of
the Sot. of Pet. Engr., Los Angeles, California, April 911, 1980.
30. Fassihi, M.R. and W.E. Brigham, 1982. Analysis of Fuel Oxidation in In-Situ Combustion
Oil Recovery. U.S. DOE Report, DOE/ET/12056-26.
August, p, 278.
31. Fassihi, M.R., W.E. Brigham, and H.J. Ramsey Jr. 1984. Reaction Kinetics of In-Situ Combustion: Part 1 Observation, Part 2 Modeling.
416.
32. Fassihi, M.R., K.O. Meyers, and P.F. Baslle. 1990, Low Temperature Oxidation of Viscous
Crude Oils. SPE Reservoir Engineering, Vol. No. November, pp. 60916.
33. Fassihi, M.R. and T.H. Gilham. 1994. The Use of Air Injection to Improve the Double Displacement Processes. Paper ISC-10 P. Sarathi and D. Olsen Editors Field Application of InSitu Combustion
Past Performance
Proceedings.
NIPER/DOE
Symposium on In-Situ Combustion, April 222. Tulsa, Oklahoma. pp. 143158. U.S. DOE
Report NIPER/BDM-0086,
CONF-940450
127
34. Freeman, E.S. and B. Carroll, 1958. The Application of l%errnoanalytical Techniques to
Reaction Kinetics; The Therrnogravirnetric Evaluation of the Kinetics of the Decomposition of Calcium Oxalate Monohydrate. J. Phys. Chem. Vol. 62, No. 4, April. pp. 394-397.
35. Germain, P and J.L. Geyelin, 1997. Air Injection Into a Light Oil Reservoiq The Horse
Creek Project. SPE Paper 37782, Presented at the SPE Middle East Oil Show, Manarna,
Babrim. March 1518.
36, Hardy, W.C., P.B. Fletcher, J.C. Shepard, E.W. Dittman, and D.W. Zadow. 1972. In-Situ
Combustion in a Thin Reservoir Containing High Gravity Oil.
37. Holt, R.J. 1992. In-Situ Combustion with Metallic Additives U.S. DOE Report No. DOE/
BC/1460@29
(DE92OO1O57) July.
38. Hughes, R., V.M. Kamatin, and D. Price. 1987. Kinetics of In-Situ Combustion for Oil
Recovery. Chem. Eng. Dev. Vol. 65, No. 1, January, pp. 2328.
39. Islam, M.R., A. Chakma, and S.M. Farouq Ali. 1989. State of the Art of In-Situ Combustion
Modeling
and Operations. SPE Paper 18755 Presented at the California Regional Meeting,
Conference,
Calgary, Alberta.
444 Lin, C.Y., W.H. Chen, and W.E. Culham, 1984. New Kinetic Models for Thermal Cracking
of Crude Oils in In-Situ Combustion Process. SPE Paper 13074, Presented at the SPES 59*
Annual Technical Conference and Exhibition, Houston, Texas. September 119, 1984.
128
45. Mamora, D.D., H.J. Ramsey, Jr., W.E. Brigham, and L.M. Castamier. 1993. Kinetics of InSitu Combustion. SOPRI TR-91,
(DE93000152)
ada. September.
47. Metwally, M. 1991. Recovery Mechanisms: Fireflooding a High Gravity Crude in a Waterflood Sandstone Reservoir, Countess Field Alberta. SPE Paper Presented at the 1s International Thermal Operations Symposium. Bakersfield, California. February 78.
48. Moore, R.G. 1993. New Strategies for In-Situ Combustion. J. Can Pet. Tech., Vol. 32, No.
10. (December). pp. 1113.
49. Moore, R.G., C.J. Laureshenj J.D.M. Belgraveland, M.G. Ursenbach, A.E. George,
and
K.N. Jha. 1995. A Comparison of the Laboratory In-Situ Combustion Behavior of Canadian
Oils. in R.F. Meyer Ed. Proceedings
Crude and Tar Sands, Houston, TX. Feb. 1217, 1995. Vol. 1, pp. 419432,
Report Conf-9502114
on Heavy
U.S. DOE
(DE95000188).
50. Moore, R.G., D.W. Bennion, J.P. Millour, M.G. Ursenbach, and D.N. Crie, 1986. Comparison of the Effect of Thermal Cracking and Low Temperature Oxidation on Fuel Deposition
During In-Situ Combustion in J.D. Westholf and L.C. Marchant, Editors Proceedings of the
1986 Tar Sands Symposium,
METC-87/6073
(CONF-860717)
pp. 270-283.
51. Moore, R.G., M.G. Ursenbach, C.J. Laureshen, J.D.M. Belgrave, and S.A. Mehta 1995b.
Ramped Temperature Oxidation Analysis of Athabasca Oil Sands Bitumen. Paper No. 95
23, Presented at the 46ti Annual Technical Meeting of the Petroleum Society of CIM, Banff,
Alberta, May 14-17. P. 19.
52. Moschopedia,
S.E. and J.G. Speight 1975. Oxidation of a Bitumen. Fuel Vol. 54, July. pp.
210-12.
53. Nickle, S.K., K.O. Meyers, and L.J. Nash, 1987, Shortcomings in the Use of TGA/DSC
Techniques to Evaluate In-Situ Combustion.
Annual Technical Conference
54. Parrish, D.R. and F.F. Craig, 1969. Laboratory Study of a Combination of Forward Combustion and Waterflood. The COFCAW Process, J. Pet. Tech. June, pp. 753-761.
129
55. Pebdami, F.N., R. Longoria, D.N. Wilkerson, and V.N. Venkatesan, 1988. Enhanced Oil
Recovery
by Wet In-Situ Combustion Esperson Dome Field. Liberty County, Texas. SPE
Paper 18072 Presented at the 6W Annual Technical Conference and Exhibition of the Society of Petroleum Engineers. Houston, Texas, October 25.
56. Prasad, R.S. and J.A. Slater. 1986. High Pressure Combustion Tube Tests. Paper SPE/DOE
14919 Presented at the 5 SPE/DOE Symposium on Enhanced Oil Recovery. Tulsa, Oklahoma, April 20-23.
and Application
of a Thermocatalytic
In-Situ Combustion
58. Reich, L. and S.S. Stivala, 1978, Kinetic Parameters fkom Therrno-Gravimetric
Curves.
Data. Thermo-Gravirnetric
60. Scarborough,
R.M. and G.V. Cady, 1982, The Design and Installation of an Oxygen
Supported In-Situ Combustion Project in Meyer, R.F., J.C. Wynn, and J.C. Olson, Editors:
The Future of Heavy Oil and Tar Sands. Proceedings of ~d UNITAR International Conference on Heavy Oil and Tar Sands. Maracaibo, Venezuela,
61. Segal, E. and D. Fatu, 1976. Some Variants of the Freeman Carroll Method, J. Thermal
Analysis, Vol. 9, pp. 659.
62. Severin, D., 0. Glinzer, H. Killesreiter, and H.J. Newman, 1983. The Quiet Oxidation of a
Crude Oil and the Effect on its Viscosity. Erdol and Kohl-Erdgas, Vol. 36, No. 3, March. pp.
127130.
Conference.
Syd-
64. Smith, F.W. and T.K. Perkin, 1973. Experimental and Numerical Simulation Studies of the
Wet Combustion Recovery Process, J. Can. Pet. Tech. Vol. 12, No. 3. May-June, pp. 44-54.
65. Struna, S.M. and F.H. Poettmann, 1986. In-Situ Combustion in the Lower Hospah Formation, McKinley County, New Mexico Paper SPE/DOE 14917, Presented at the SPE/DOE 5ti
Symposium on Enhanced Oil Recovery, Tulsa, Oklahoma, April 20-23.
130
66. Tanabe, K. 1970. Solids Acids and Bases Their Catalytic Properties. Academic Press,
New York. pp. 105127.
67. Townsend, D.I. and J.C. Tou, 1980. Thermal Hazard Evaluation by an Accelerating Rate
Calorimeter. Thermochimica Acts. Vol. 37, pp. 1-30.
68. Venuto, P,B. and E.T. Habib, Jr. 1979. Fluid Catalytic Cracking with Zeolite Catalysts.
Mercell Dekker Inc., New York, p. 121.
70. Verkoczy, B. and K.N. Jha, 1988. Behavior of Heavy Oil Core Material Under Simulated
Fireflood Conditions. ,Paper No. 44, Presented at the 4ti UNITAR/UNPP
International Con-
ference on Heavy Crude and Tar Sands, R.F. Meyers, and E.J. Wiggins Editors. August 712,
1988. Edmonton, Alberta, Canada.
71. Verkoczy, B. and K.N. Jha. 1985. Screening of Heavy Oil Reservoir for Enhanced Oil
Recovery by Thermal Methods. Paper No. 8536-32,
cal Meeting of the Petroleum Society of CIM, June 25, Edmonton, Alberta.
72. Vossoughi, S., G.P. Willhite, W.P. Kritikos, I.M. Guvernir, and Y. El Shoubary, 1982. Automation of an In-Situ Combustion Tube and Study of the Effect of Clay on the In-Situ Combustion Process. SPE 10320,56
face Area on the Process. Sot. of Pet. Eng. J. Vol. 25, No. 5, October, pp. 656-64.
74. Vossoughi, S., G.P. Willhite, W.P, Kritikos, I.M. Guvenin, and Y. E. Shoubary, 1981, Effect
of Clay on Crude Oil In-Situ Combustion Process. SPE 10320, Presented at 56ti Annual
Fall Technical Conference
and Exhibition
the
131
77. Yannimaras, D.V. and D.L Tiffin. 1995, Screening of Oils for In-Situ Combustion at Reservoir Conditions by Accelerating Rate Calorimetry. SPE Res. Eng., Vol. No. Feb. pp. 36-39.
78. Yannimaras, D.V. 1997. Personal Communication, Amoco Production Co., Exploration and
Technology
79. Yannimaras, D.V. 1998. Personal Communications, Amoco Production Co., Exploration&
Production Technology
Group. May
81. Yoshiki, K.S. and C.R. Phillips, 1985. Kinetics of the Thermoxidative and Thermal Cracking Reactions of Athabasca Bitumen Fuel., Vol. 64, November, No. 12, pp. 15911598.
82. Zelenko, V. and F. Solignac, 1997. Use of Accelerating Rate Calorimeter in Reservoir Oil
Selection for Air Injection Process. Paper No. 97166, Presented at the 7ti Petroleum Conference of the South Saskatchewan Section of the Petroleum Society of CIM, Regina, October 1922.
83. Zoaks, J. 1968. Kinetic Analysis of Thermogravimetric
July. pp. 2406-2411.
132
CHAPTER
4 EVALUATION
OF AN IN-SITU
COMBUSTION
PROSPECT
Introduction
Several factors must be taken into consideration when evaluating candidate reservoirs for in-situ
combustion application.These include site geology, reservoir rock and fluid properties, crude oil characteristics, and reservoir geometries.
This chapter discusses the geological,
tive importance to a reservoir engineer evaluating the properties for a particular in-situ combustion
projeet.
Geologic
Reservoir geological
Characterization
projects. Examination of the reservoir characteristics of the California, Oklahoma, and Texas fireflood
projects (states that account for more than 70% of the implemented US ISC projects) indicate that the
structure, lateral continuity, and physical characteristics of the individual sand layers within the reservoir as well as the reservoir heterogeneities played a significant role in the performance
of these
projects.
Lack of good sand continuity (due to complex lateral facies variations) and channeling have
been cited as one of the cause of failure of many California firefloods (Simm, 1967). Since ISC is an
interwell drive process good horizontal continuity is critical to the success of the project. Gaps in formation overburden or leaky interzonal seals in stratified reservoirs can allow fluid to leak into overlying strata and reduce the effectiveness of the injectant. Fractures and joint trends, however subtle, may
create preferential flow channels which influence recovery efficiency.Therefore,
geologic
a knowledge
of the
characteristic of the site is important for the proper evaluation of a prospect for ISC. Ear-
lougher et al. (1970) in analyzing the performance of the Fry in-situ combustion project, Illinois indicated that the reservoir geology played a prominent role in the outcome this project and stressed that
an understanding of the reservoir geology is essential to the design and successful operation of a combustion project.
The objective of a geological
of the qualitative and quantitative parameters of the reservoir so that the engineer can design a scheme
that most appropriately matches the reservoir conditions.
133
parameters to be considered
include: the degree and extent of lateral and vertical reservoir continuity, depth, thickness, structural
attitude and dip, overburden competence,
reservoir heterogeneities,
aquifer.
Lateral
and Vertical
Extent
of Reservoirs.
The continuity of individual sand layers within the producing formation, especially in thin, lenticular sands is a factor of major importance to the successful operation of flreflood. In-situ combustion require significantly more capital investment per unit of production than waterflood because of
the need for ancillary equipment, such as the air compressors, has high operating costs and is manpower intensive.This higher cost means that the volume of oil in place per unit area must be above a
certain minimum to make the project economically
ment of fluids. The general lack of areal continuity and poor understanding of the geology of the site
by early operators contributed to the failure of the process in these reservoirs. Many of these pilots
were unfortunate in site choice because they were done in properties that did not prove to be economic as primary producing cases.
134
Hence lateral and vertical extent of a reservoir is one of the key parameter to consider in the site
selection process for an ISC project.The degree and extent of lateral and vertical reservoir continuity
significantly affect the performance
of the ISC process, Clean, well sorted sands tend to have good
horizontal and vertical continuity. Reservoir continuity can be reduced by disseminated finer grains,
by the local occurrence of various types of shale interbeds, and cementation materials.l%e study of
setting in which the sands were deposited can give an approximation of the reservoir continuity in lateral and vertical perspective. A complete characterization of facies distribution would help to predict
how reservoir performance can be affected by flow barriers.
Vertical
Depth
Depth of the reservoir is not an handicap to the implementation of ISC process. Economically
successful projects have been implemented in reservoirs ranging in depth from 300-1,1500
ft. Depth,
however, is a factor in terms of temperature, pressure and well cost. Shallower depth (less than 200
ft.) would severely limit the pressure at which air could be injected. With increasing depth, air injection pressure generally increases with a corresponding
sor). Deeper reservoirs are usually hot enough, that spontaneous ignition of in-situ hydrocarbon
is
are
needed to meet the injection pressure requirements. Larger compressors are more expensive to purchase, operate and maintain. Depth also effects the fluid lifting costs, especially in wet combustion
process.Thus economic
order of 12,000-12,500
feet.
135
on the
Reservoir
Thickness
Sand thickness is one of the important parameters for the combustion process. The large difference in density between air and the reservoir fluids gives the air a tendency to override the oil column
and consequently bypass much of the oil if the reservoir exceed a critical thickness. A thin oil sand
tend to counter this override tendency and favor a more uniform displacement and vertical sweep. In a
thin heavy oil reservoir, rapid transfer of heat to the bottom of the sand will permit combustion front
to advance at the bottom more rapidly than it would be possible in a thick sand (Boberg, 1988). If the
sand, however, is too thin high overburden heat losses may drop the temperature below that necessary
to sustain a combustion front and can lead to low temperature oxidations and loss of recovery. Preferably pay thickness should at least be four feet and should not exceed 50 feet. It is preferable that very
thin reservoirs (less than 8 ft. thick pay) considered for fireflood be contain multiply stacked thin sand
separated by non communicating vertical barriers to take advantage of heat conduction in the vertical
direction.This can not only minimize the heat losses to the overburden but can also aid in promoting
and sustaining high temperature combustion mode in heavy oil reservoirs. The Fry in-situ combustion
project in Illinois (Hewitt and Morgan, 1965; Bleakley, 1971) is a prime example of an successfid insitu combustion project implemented in multiple thin sand (less than 5 ft. thick) reservoir.
Formation thickness is also an important consideration in reservoirs containing oil not readily
susceptible to auto-ignition. In such reservoirs the near well-bore area must be heated to a high temperature to initiate ignition. If the formation is very thick (>50 ft.) the amount of heat needed to raise
the well-bore vicinity above the oils auto-ignition temperature can be very large and expensive. Formations up to 60 ft. thick have been ignited using artificial ignition techniques.
Structural
Attitude
and Dip
Structural attitude and dip are important consideration in the location of wells for a combustion
project. Injected air and combustion front movement will be more rapid toward up dip wells than
toward wells low on the structure. In dipping reservoirs it is advisable to locate the air injectors downdip and production wells up the structure to compensate for the expected flow of air up dip. In steeply
dipping reservoirs some operators preferred injecting air at the top of the structure to take advantage
of gravity in the recovery of hot mobile trade affected by combustion (Gates and Skalar, 197 1). In the
steeply dipping Webster reservoir, the combustion project was initiated as a crestal drive in part, to
heat the oil at the top of the structure and promote migration of oil toward the flanks of the anticline
136
(Soustek, 1994). Dip and the result~g gravity dominance played a major role in the economic success
of the Santa Fe Energy Co.s (now part of Texaco) Midway Sunset Combustion project.
Turta (1995) recommend locating the ISC pilot at the uppermost part of the structure. The reason behind this recommendation is that the burned volume of the pilot located at the upper part of the
reservoir can be more accurately be determine and both the air-oil ratio (AOR) and the incremental oil
recove~
due to combustion can be estimated more reliably. Also by locating the pilot updip, resatura-
tion of the burned zone can be avoided in case the air injection is terminated due to compressor failure.
Overburden
Competence
The producing formation at the project site must site have sufficient and competent overburden
so as to confine the injected air within the pay zone. Gaps in oil sand overburden or leaky interzona.1
seals in stratified reservoirs can allow fluid leaks into overlying strata.
Reservoir
Heterogeneities
include perme-
ability barriers to lateral and vertical flow, natural fractures, high permeability thief zones, directional
permeability, presence of gas cap and aquifers.
Permeability barriers can have both positive and negative effect upon the in-situ combustion
process. As a positive effect, vertical permeability barriers can divide a thick reservoir into smaller
units, which may be more compatible with the in-situ combustion process. Vertical barriers can also
act as a seal to upward migration of injected air and may result in a more uniform burning in relatively
thick reservoirs. As a negative effect, horizontal permeability barriers can reduce the reservoir continuity and recovery.
Fractures and joints are secondary properties that may create preferential flow channels and
influence the recovery. A thin zone of high permeability at the top of the reservoir extending from one
well to another constitutes a hazard to fireflood by thiefing air and starving the fire front of needed
oxygen.
Directional permeability resulting from the anisotropic characteristics of the reservoir has a
major influence in the performance of many in-situ combustion project. These include the Iola fireflood in Kansas (Hardy and Raiford, 1975), Fry in-situ combustion project in Illinois (Earlougher et
137
al., 1970) and the Webster fireflood, in the Midway Sunset field, California (Soustek et al., 1994).
Grain size and its orientation contribute to the existence of directional permeability in a heterogeneous reservoir, Often the orientation of the medium to coarse grained sands establish the direction of
high permeability zone. Directional permeability can cause air to flow more freely in one direction
than in any other direction and result in uneven burn.Existence of directional permeability alone is not
a sufficient cause to reject a site for fireflood. By selectively locating the wells in the direction of permeability, the recovery can be maximize.
Presence of free gas caps or thin layer of high gas saturation at the top of the sand is not a desirable geological feature for fireflood operation because they can act as a thief zone for injected air and
promote uneven burning. Presence of bottom water leg or aquifer, though, not desirable from the
point of anisotropy is not an impediment to the success of a fireflood project. Many successful
projects as the Glen Hummel, Gloriana, Trix-Liz, N. government Wells fireflood in Texas were implemented in reservoirs with active aquifers. In these projects the aquifer not only provided pressure support to the reservoir but also acted as a conduit to transfer the heat ahead of combustion front.
Though, reservoir heterogeneity can have adverse effect on project performance
their impact
can be minimize through recognition of the distinctive architecture of the reservoir and tailoring the
project to accommodate
also often require unique reservoir management strategies to make the project economiciily
viable.
Some examples of successful fireflood projects where the combination of unique engineering design
and reservoir management strategies overcame the many conditions considered adverse to the success
of the process include Unocals
Brea-Olinda
fireflood
(Showalter,
Rock
Properties
The key rock properties of interest to an engineer evaluating a prospect for the application of insitu combustion process are: sand texture, permeability and its distribution, porosity, and composition
of rock matrix. In many firefloods, especially those implemented in light oil reservoirs rock composition is more important than oil properties in determining the amount of fuel available for combustion.
138
Sand
Uniformity
and Texture
flood performance. Actual grain size and grading, shape of grains, character and amount of cementing
material determines the physical characteristics and properties of the reservoir. The size, shape and
sorting of the grains determine the porosity
and permeability
and
uniformity exhibited in a profile more uniform the burn will be. However, there are many instances of
economically
1971; Soustek, 1994). In the Mobils Webster reservoir combustion project, where sand bodies are
lenticular and anisotropic, the success of the project was attributed to tailoring the operating policy to
suit the reservoir architecture. In this reservoir the bum front advanced more rapidly through the high
permeability medium to coarse-grained sands than through finer-grained, thin medium bedded sands,
reflecting the influence of anisotropy on the lateral rate of movement of bum. Proper reservoir surveillance and modification of production strategy based on the timely identification of heat breakthroughs
were cited as the keys to the success of the project (Soustek, 1994).
Thus from an oil recovery and sweep efficiency aspect the degree to which a particular profile
correlates from one well to another is more important than the exact shape or dimensions of that particular profile.
Permeability
The actual value of permeability has very little effect on the mechanics of combustion process.
Economically
successful firefloods have been implemented in less than 10 millidarcy carbonate light
oil reservoirs (Miller, 1995). The only requirement for permeability is that is must be adequate to permit air injection at a pressure compatible with overburden at an acceptable compression cost. In viscous heavy oil resemoirs too low permeability may fail to provide the minimum air flux needed for
sustained combustion. Low permeability also increases air injection pressure requirements and compression costs, and prolongs the operation. Low permeability in a viscous (greater than 100 cp.) shallow reservoir can limit the infectivity and promote low temperature oxidation. In such reservoirs a
permeability greater than 100 millidarcies would be necessary.
139
Porosity
High porosity is desirable, since it directly reflects the volume of hydrocarbons that the rock can
hold. In the U. S., economically
average porosity range from a low of 0.16 to high 0.38. As porosity decreases, the amount of heat
stored in the rock increases. A lower porosity will not have a significant impact on overall energy utilization in wet combustion process because part of the heat stored in the burned volume of the reservoirs will be recovered through scavenging operations. The main impact of porosity will be in its oil
content. The economic
success of a fireflood is dependent more on the actual value of the oil satura-
tion-porosity product (fSo) than on porosity. Porosity lower than 0.2 is acceptable only if the oil saturation is greater than 0.45.
Oil Saturation
A minimum oil content (the product of oil saturation and porosity) is necessary in order to offset
the consumption of oil as fuel in an in-situ combustion process. A widely accepted rule-of-thumb
in
the industry is that if fSo is less than 0.09 or 700 bbl/ac-ft. dry combustion should be eliminated from
further consideration.
This arbitrary cutoff simply implies that the reservoir should have enough
recoverable oil to cover the energy requirements of the process and supply additional production to
make the process economically
attractive. For wet combustion where the fuel laydown is lower some-
Composition
The economics
of Reservoir
Matrix
nature and amount of fuel formed in the reservoir. If sufficient fuel is not deposited the combustion
front will not be self-sustafig.
dismal economics
due to high air requirements, high power cost and low oil recovery rate. Consider-
able laboratory and some field evidence exists indicating that the mineralogical
reservoir rock and chemical composition
composition
of the
sustain combustion.
Actual laboratory measurements of fuel formed using reservoir rock and crude indicate rock
type is probably more important than the crude properties, particularly in light oil reservoirs in determining the amount of fuel deposition in the reservoir (Earlougher et al,, 1970). The clay and metallic
content of the rock, as well its surface area has a profound influence on fuel deposition rate and its
oxidation. Presence of clays and fine sands in the matrix favor increased rates of fuel formation.
140
Increase clay content particularly kaolonite and illite favor increased rates of fuel formation by favoring low temperature oxidation reaction. Rock minerals such as pyrite, calcite, and siderite also favor
fuel-forming
reactions. Low air fluxes resulting from reservoir hetrogeneneities and oxygen channel-
ing also promote low temperature oxidation and t%el formation reaction.
Results from the Fry in-situ combustion project showed fuel deposition varied with the lithologic characteristics of the rock. In the laboratory tests a very fine to fine grained sandstone containing
significant amount of pyrite and siderite deposited a greater amount of the fuel than medium grained
sandstone containing similar amount of pyrite. Similarly medium grained sandstone core containing
large amount of clay material yielded the largest amount of fuel (3.3 lb./cu.ft. rock).
Effect
of Well Spacing
Problems may arise in two ways when determining well spacing. If the well spacing is too close,
the combustion front may experience early gas breakthrough, while if the well spacing is too large,
the oil production rate will be slow, thus prolonging the life of the project and making the economic
unattractive. Hence, the well spacing should be in the optimum range to maximize oil recovery.
Geological
considerations
bodies in the U.S. where fireflood had been implement are not continuous sheets of sand, but are lenticular in shape. These sand bodies frequently exhibit anisotropy parallel to the bedding.The permeability of the sand in one direction therefore is greater than in another. In such anisotropic, lenticular
reservoirs, it is advisable from a stratigraphic standpoint drill injection wells at right angle to the
direction of high permeability trend and at closer spacing. The production wells can be drilled along
the trend on a wider spacing (600 ft. or greater). Such a flood pattern had been adopted in some of the
fireflood projects, implemented in the narrow shoestring pools of S.E. Kansas.
141
Prospect
Screening
Assessing the suitability of a prospect for an oil recovery process consists of (1) evaluating
available information about the reservoir, oil, rock, water, geology and previous performance, (2) supplementing available information with certain relevant laboratory screening tests, and selecting the
candidate reservoir that best match the process requirement.
In-situ combustion has been successfully implemented in reservoirs with widely differing rock,
fluid, and geological characteristics. This render the development of satisfactory guidelines to screen
reservoirs for combustion application difficult. Since the success of a combustion project depend to a
large degree on the geologic
ervoir for combustion application must include the rock, fluid, as well the geological
requirements.
An applicability criterion for combustion process is given in this section. It is intended as informational guide rather than absolute constraints. These are general guidelines that reflect current technology and economic
engineering judgement applied before a decision can be made to pilot test the reservoir.
Screening Criteria
Oil
Viscosity: Preferably less than 5,000 cp at reservoir condition.
Gravity: 1O-4OAPI
Composition:
low asphaltic, low heavy metal content crude. Heavy metal (Va, Ni, etc.) should
Water
Connate water properties are not critical.
Lithology
Heavy oil reservoir: Low clay content; low in minerals that promote increased fuel formation
such as pyrite, calcite, and siderite as well low in heavy metals.
Light oil reservoirs: Lithology that tends to promote fuel deposition is preferred.
142
Reservoir
Depth: 300 12500 ft.
Thickness: 5-50 ft.
Permeability: Not critical
Porosity: >0.18
Oil concentration: 700 bbl/ac-ft, fso >0.09
Thnsmissibility:
kh
~0 a 20 md-ft/cp
Favorable Factors
~
Extensive fractures
143
REFERENCES
1.
Metiods
of Oil Recove~,
Wdey&Sons,
p.
269.
2.
Buchwald, R.W., W.C. Hardy, and G.S. Neinast, 1973, Case Histories of Three In-Situ
Combustion Projects, J. Pet. Tech. Vol, 25, No. 7, July, pp. 784-792.
3.
Casey, T.J. 1971, A Field Test of the In-Situ Combustion Process in a Near-Depleted Water
Dirve Reservoir, J. Pet. Tech. Vol. 23, No. 2, February, pp. 153160.
4.
In-Situ Combustion
Project,
Meeting,
5.
Earlougher, R. C., J.R. Galloway, and R.W. Parsons, 1970. Performance of the Fry In-Situ
Combustion Project, J. Pet. Tech. Vol. 22, No. 5, May, pp. 55 157.
6.
Field, Duval County, Texas, Paper SPE 3776, Presented at the SPE Improved Oil Recovery
Symposium, Tulsa, Oklahoma, April 16-19.
7.
Gates,
Handy, W.C. and J.D. Raiford, 1975, In-Situ Combustion in a Bartlesville Sand Allen
County, KS, in Proceedings of Tertiary Oil Recovery Conference, Wichita, Kansas, October
2324, Institute of Mineral Resources Research, U. Of Kansas, Lawrence.
9.
Hewitt, C.H. and J.T. Morgan, 1965, The Fry In-Situ Combustion Test Reservoir Characteristics, J. Pet. Tech. Vol. 17, No. 3, March, pp. 337342.
10. Hoffman, S.J. 1998, Successful Application of In-Situ Combustion to a Dipping Heavy Oil
Reservoir in Midway-Sunset
of Petroleum Engineers
144
Improved
Oil
12. Turta, A. 1994. In-Situ Combustion From Pilot to Commercial Application, In Proceedings of 1994 NIPER/DOE Symposium on In-Situ Combustion, Tulsa, Oklahoma, April 21
22, U.S. DOE Report NIPER/BDM-O086,
13. Yannirnaras, D.V., A.H. Sufi, and M.R. Fassihi, 1991. The Case for Air Injection into Deep
Light Oil Reservoirs,
in Proceedings
145
on Improved Oil
146
CHAPTER
5 ENGINEERING
In-Situ
OF AN IN-SITU
Combustion
Performance
COMBUSTION
PROJECT
Parameters
Several variables affect the performance of an in-situ combustion process. The most important
parameters are fuel deposit, air requirement, air flux, air injection rate, air-oil ratio, injection pressure,
and oil recovery rate.
Fuel Deposit
The quantity and type of fuel deposit is an important variable and is expressed as pounds of fuel
per cubic foot of formation. The fuel burned per reservoir volume determines how much heat is generated. It also determines the amount of air required, the rate of burning front advance, the rate of oil
recovery, and project life.
Fuel deposition is a function of crude oil properties, oil saturation, formation permeability, and
temperature in the combustion zone. Fuel deposition can be determined experimentally using a combustion tube (Figure 5.1 ). Laboratory combustion tube tests have indicated that, in general, higher
fuel depositions are expected with heavier and more viscous crudes. Figure 5.2 shows a correlation of
fuel deposit with oil gravity obtained from laboratory studies. Based on laboratory combustion tests
in preserved or restored cores and from field data, Chu (1982) presented the following regression correlation (Equation 5,1 ) to calculate fuel content for field projects.
147
rA7 71NSUWTION o
1
STEEL JACKET
LLl
148
3.0
2.5
2.0
SHOWALTER -
1.5
1.0
0.5
E
o
20
10
40
30
API GRAVITY AT 60 P
M = -0.12+0.00262h+0.000114k
+ 2.23 So + 0.000242(kh/V)
- 0.000189z - 0.000652p
(5.1)
where M = fuel content (lb/ft.3), h = reservoir thickness (ft.), k = permeability (red), So = oil saturation (fraction), m = viscosity at reservoir temperature (cp), and z = depth (ft.).
The foregoing correlation can be used as a first approximation for the fuel deposition of a crude
for a prospective field project. Actual combustion tube tests can be conducted later to obtain a more
refined estimate.
149
1600
runua
dv
/0
u-
LL1400
1n~ 1200
1-
#
/
1. /
n=l.6
Cp = 2.0 BTU/lb.
400
ps=1631b/ft
200
o
0
0.2
0.4
0,6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
FIGURE 5.3 Minimum Fuel Content Required to Support a Fixed Frontal Temperature
Figure 5.3 shows graphically, the minimum fuel requirement to maintain a fixed combustion
front temperature as a function of porosity.
Air Requirements
The volume of air required to burn a unit volume of the reservoir based on stoichiometric analysis of the combustion gas produced from combustion tube is shown in Figure 5.4. In the absence of
laboratory data, this figure can be used for preliminary estimates of air requirements. The air requirements determine the compression
150
a function of the amount of carbon and hydrogen in the fuel and the ratio of carbon dioxide to carbon
monoxide produced by the combustion, as shown in Figure 5.5. The hydrogen-to-carbon
ratio of the
fuel deposited during a fireflood usually ranges between 0.1 and 0.15.
203
g&# 0
10
15
20
25
30
35
40
For most firefloods, a value of 180 scf air/lb fuel is often used when laboratory data are not
available. Based on field performance data, Chu (1982) presented the following
regression equation
(5.2)
where A = the air requirement (million scf/ac-ft.), Lacking sufficient laboratory data to calculate the
stoichiometric air requirements, the previous equation can be used to obtain a rough estimate of the
air requirement. The equation does not account for the crudes burning characteristics. Based on
reported data, estimated air requirement in million scf/ac-ft. is about ten times the fuel content in lb/
ft.3 of burned volume. For example, if the fuel content is 1.5 lb/ft.3, the air requirement will be about
15 million scf/ac-ft.
151
Qlco
co
200
100
0
0.3
0.2
0.1
H/C RATlO
50000
0.125
0.150
40000
0.175
0.200
0.225
0.250
20000
10000
FIGURE 5.6 The Teonetical Air Required to Move a Barrel of Oil in the
Reservoir is Shown as a Function of Fuel Content and Porosity
152
Air Flux
The combustion front can move only as fast as air is supplied to consume the amount of fuel
deposited, since all the fuel must be burned. Laboratory tests indicate that at relatively high air fluxes,
the combustion is quite vigorous, resulting in combustion temperatures of about 1,OOOFfor nominal
fuel deposition (1.5 lb/ft.3). However, lower air fluxes result in lower combustion temperatures. As air
flux is reduced farther, the process approaches a point where the heat losses exceed the rate of heat
generation, and the combustion front is extinguished. Thus, the minimum air flux is a function of both
the fuel deposition and heat losses. In a field situation, the air flux required to sustain combustion
increases with oil gravity and decreases with pay thickness (see Figure 5.7).
00
POROSITY = 3070
COMBUSTION EFFICIENCY = 100%
80
60
40
BURNING ZONE. .
THICKNESS, fT
20
0
>0
10
15
20
25
30
35
40
For Mobils
Midway-Sunset
combustion
153
0.5
1.0
1.5
2.0
2.5
0.10
.%O
3.5
008
006
1.1 0910-3(12+n)Y@
/
VF=
FRONT VELOCITY
17/HR
004
002
000
10
20
30
40
50
60
70
Air-Oil
Ratio
The air-oil ratio (AOR) is the single most important economic parameter in fireflooding. It is a
measure of the quantity of air that must be injected to recover a barrel of oil. It is used along with unit
air cost to determine the air injection cost per barrel of oil produced. The AOR is a function of the oil
in place and the fuel burned.
154
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
0.5
FUEL DEPOSIT,
1.0
POUNDS
1.5
2.0
2.5
3.0
Figure 5.9 shows the relationship between oil saturation, fuel content and theoretical ah-oil
ratio. The theoretical air-oil ratio is the volume of air injected per barrel of oil displaced from the
burned volume. The oil displaced is equivalent to the oil in place less the oil burned as fuel. In comparison, the produced ah-oil ratio is based on the air requirement of 180 scf/lb of fuel burned. For
example, if the oil in place were 1,000 bbl/ac-ft. and fuel consumption were 1.0 lb/ft.s of reservoir, the
AOR would be 7,850 scf/bbl for the swept region.
155
105
FUEL CONSUMPTION
0
i=_l
<m
Urn
&z
o ~
u
3
LB/FT 3
~
3.0
104
2.0
103
300
1.0
1000
4000
OIL IN PLACE
BBL/AC-FT
FIGURE 5.10 Air-Oil Ratio as a Function of Oil in Place and Fuel Consumption
The following regression correlation developed by Chu (1982) from the performance history of
field projects can be used (Equation 5.3) for estimating AOR (mcf/bbl) in terms of reservoir rock and
fluid properties:
AOR = 21.45 + 0.0222h + 0.001065k + 0.002645P 76.760S0
(5.3)
where @ = porosity.
The AOR is affected by the geometry of the oil field well pattern. A common practice in the field
projects is to convert four inverted 5spot patterns to one nine-spot pattern, thereby reducing by half
the air injection rate while burning the same reservoir volume with only a small reduction in productivity.
156
Injection
Pressure
A major expense in a combustion project is the cost of compressing air. The size of the compressor depends not only on the required air injection rate, but also on the discharge pressure required. Air
injection pressure is dependent primarily on the permeability of the formation to air, selected air
injection rate, well spacing, and formation depth.
The preferred way of determining the required air injection pressure is to run actual air injection
tests in the field. For design purposes, the values obtained from such tests should be increased by
some reasonable factor (such as 30Yo) to take care of any unexpected pressure increases during combustion operation. Nelson and McNeil (1961), on t4e basis of several fieldtests (for a large number of
adjacent 5spot pattern), suggested the following
sure:
=#+iaPa(T+460)
lW
0.703 k, h
(5.4)
[{+)-12381
well bottomhole
pressure
(W), k,= effective permeability to air (red), h = net pay thickness (ft.), a = well spacing (ft.), rW= production well radius (ft.), V~ = burning zone velocity (ft./day), and t = time to reach maximum air rate
(day).
Oil Recovery
Rate
In the laboratory, the oil recovery for an ISC recovery program usually ranges between 60% and
90% of oil in place. This high oil recovery usually results from burning mobile crudes in high-porosity sandpacks containing high initial oil saturation. Oil recoveries in the field are much lower than laboratory oil recoveries due to lower horizontal and vertical sweep efficiencies.
Graphical correlations
have appeared in the literature to estimate oil recovery (Brigham et al., 1980). These correlations may
be used in preliminary design work. However, they may not be valid outside the range of data used for
their development. These correlations are fully discussed in a later section.
157
In-Situ
Combustion
Project
Design
Several authors have presented procedures to engineer an in-situ combustion project (Nelson
and McNeil, 1961; Gates and Rarney, 1980; Brigham et al., 1980; Fassihi et al., 1981; Naji and Poettmann, 1991). These were developed based on reported field performance data. The calculations are
relatively simple and can be carried out using a spreadsheet program or a programmable calculator.
In the following
iVelson-iWciVeil
Method
Nelson and McNeil (1961) presented an engineering procedure to evaluate the performance of a
dry in-situ combustion project. Although a large number of assumptions were made, the method is
based on considerable field experience and may give reasonable estimates. They presented equations
to calculate:
1.
2.
3.
4.
The method, though, relatively simple require experirhental combustion tube data to calculate
fuel lay-down and air,requirement. A step-by-step procedure to calculate oil and water production is
described below.
158
Fuel consumption and air requirement are calculated based on laboratory experiments. In Equation 5.5 to 5.16 the following nomenclature is employed.
D = Inside diameter of the combustion tube, ft.
L = Length of the pack burned, ft.
@= Porosity, fraction
V~ = Volume of the produced gas, scf
Nz, = Volume fraction of nitrogen in injected air
Oz. = Volume fraction of oxygen in injected air
Nz~ = Volume fraction of nitrogen in produced gas
Oz~ = Volume fraction of oxygen in produced gas
COz~ = Volume fraction of carbon dioxide in produced gas
CO~ = Volume fraction of carbon monoxide in produced gas
It is assumed that the nitrogen is completely inert in the reaction, and all the injected nitrogen is
produced.
WW = 2[(Oxygen
(5.5)
x (18/379)
159
produced)] x (2/379)
(5,7)
(5.8)
(5.9)
(5.10)
(5.1 1)
The next step is to compute the toti air injected and the volume of the reservoir sand burned.
OL injected) scf
V,/W~ (scf/lb)
160
(5.12)
(5.13)
Air injected per cubic feet of reservoir sand burned= A = (V,/W~) x (W) x [(l-~)/(l-$P)]
(5.14)
Where F = (l-~)/(l-@P)
Assuming an areal sweep efficiency of 62.6%, the air required in MMscf per acre-ft in the 5spot pattern is computed as:
(5.15)
Let u be air flux or the volume of air required per square foot of burning front per day
(5.17)
u = Av (scf/day-sq ft.)
burned.
front, depending upon the relative location of the injector and the producers and the position of the
combustion front. If the air flux at a location is insufficient to support combustion, the fire goes out
and results in limited areal sweep. Nelson and McNeil (1961) for calculation purpose introduced a
dimensional flow term (i~) to calculate the air injection rate necessary to achieve a given sweep efficiency of the combustion zone.
161
i~ = i~/(uti~ ah
(5.18)
to
values of iD.
iD
3.39
I
I
50.0%
55.0%
4.77
I
I
6.06
UJ
57.5%
62.6%
Although the total air requirement for the 5spot pattern was calculated previously on the basis
of 62.6% sweep efficiency, it can be seen from the above table that infinitely high air rates would be
required to achieve this areal sweep efficiency in practice. Nelson and McNeil suggest that in the air
rate calculation one should use a sweep efficiency that will give a reasonable value of iD.
In the design method proposed by Nelson and McNeil, the air injection rate depends on the
desired rate of advance of the burning front. They found a satisfactory burning rate of 0.125-0.5
ft/
day. In the proposed method, a maximum air rate based on the minimum burning rate of 0.125 ft/day
and55% areal sweep (iD = 4.77) is first determined.
= 4.77 A (0.125) ah
162
(5.19)
Then a time schedule is chosen so that the air rate would increase gradually to the maximum
rate, hold at this rate for a definite period, and then reduce gradually to zero (Figure 5. 11).
2000
/
1800
TOTAL AIR SUPPLIED
\
1600
1400
1200
&v+
-0.5
P+
1000
~0,/
/
6.0
*
0.25
F~
5.0
4.0
800
\
600
3.0
2.0
400
/
/
200
~
o
1.0
[
100
200
300
400
500
600
TIME, DAYS
Field tests indicate that the movement of the burning front during the early stages of burning is
essentially radial and about 109o of the pattern area will have been swept during the radial displacement of the front. Nelson and McNeil contend that if the front velocity (vl) in the increasing rate
period exceeds the minimum burning velocity by a factor of three or greater, the constant maximum
injection rate will be reached before 10% of the pattern area has been swept regardless of the pattern
size.
Ifrl is the radial distance in feet at the end of the increasing air injection rate period, and r~is the
radial distance to the burning front (q = rl), the air flux rate in scf/day during the radial phase of the
displacement, can be calculated from the following
equation:
(5.20)
i~= 2m-fhAvl
163
Air injection rates, for a range of values of rf are calculated until the maximum air rate specified
by Equation (5.19) has been reached. Beyond this point, the air flux rates will began to decline, and
the burning front advance rate will be slowed.
The time in days (tl) required for the increasing rate period of the operation maybe calculated
as follows:
(5.21)
.,
The volume (Vl) of air injected during this period in MMscf is:
(5.22)
,- .,
V, = 0.5(t,
-. i.)
1
.-l-w
Next during the final stage of the burning operation decreasing air injection rate will be used and
the air rate is decreased linearly from the maximum rate i, to zero. For the purpose of balancing a
burning operation, Nelson and McNeil assumed that the volume of air Vq injected over time, t~during
the final stage of burning is identical to the volume injected during the increasing air injection rate
period. Thus, by this assumption:
(5.23>
\. ,
V. = V,~ and
t. = t,1
--.-J
If VT is total volume of air injected to burn the 5spot pattern in the field (Equation 5.16), then
the volume of air V, iniected at the constant (maximum) rate .ueriod will be
V2 =
(5.24)
MMscf
(5.25)
tz =Vz x 10G/i~
144
The total time required in days for the entire burning operation is
;=t~+t~+tq
(5.26)
In the design of an in-situ combustion, knowledge of air injection pressure is needed to size the
compression facilities. The most reliable method of determining the required air injection pressure is
to run actual air injection tests in the field. However, this is not possible or practical inmost situations.
For preliminary appraisals of in-situ combustion projects, Nelson and McNeil recommend the use of
following
the
steady state radial flow equation for a compressible fluid in a developed 5spot pattern.
P;
= p~+(%:h)
[&
-Z]
(5.27)
Where:
Pi~ = injection well bottom hole pressure, psia
PW= production well bottom hole pressure, psia
i,= maximum air injection rate, scf/day
Va= viscosity of air, centipoise
T~ = formation temperature, OR
a = well spacing, ft.
tl = time to reach maximum air rate, days
kg= effective permeability to air, md
h = formation thickness, ft.
rW= production well radius, ft.
In Equation (5.27) PWis usually low and can safely assumed to be atmospheric, because the producer will be kept drained at the time of maximum injection pressure. The effective permeability ~~is
usually not known and must be estimated. If no information is available on which to estimate kg, Nel-
165
son and McNeil recommend that a value of 5% of the specific permeability be used. T~ can be measured in the field and pa can be estimated from correlation at this temperature. Since additional
pressure beyond that indicated by Equation (5.27) will occur in the injection system and in the injection well, the compressor must be sized for pressure higher than indicated by eqn. 5.27.
The oil production mechanism in a combustion process is complex and the total amount of oil
displaced by the combustion front is dependent on the volumetric sweep of the front. Field data indicate that in addition to the oil displaced by the front to the producer, considerable volume of oil are
also produced as a result of depletion of unburned but heated regions adjacent to the burned zone. The
volumetric efficiency is a product of invasion efficiency EI and areal sweepo efficiency EA.
(5.28)
i.e., EV = E1 x EA
If the areal and invasion efficiencies
burned:
IVl = 43560
SOTR
5.615
WF
bbl
(5.29)
350 ) acft.
where:
SO= oil saturation in% pore space
350 = density of 10 API oil
Post burned cores taken from many combustion projects indicate that in heavy oil reservoirs,
more than half of the oil in the regions not contacted by the fire front may have been produced by a
combination
of gravity drainage and hot gas drive. Nelson and McNeil indicate that for preliminary
design purposes, 40% of the produced oil can be assumed to have come from the unburned region of
the reservoir. The equation for the oil displaced from the unburned region of the reservoir is:
166
The total oil recovery N~ from the burned and unburned region is given by:
N~ =Ev X NI + (1-~)
X Nz
(5.31)
N~ X 5.615
E~ =
(5.32)
43560 X SO X ~R
The total water production due to in-situ combustion is calculated as the sum of combustion
water plus the water originally present in the burned zone. This assumes that the water contained in
the unburned region is immobile and remains constant through out the burning process. Then the total
water produced in barrels per acre-feet of reservoir rock in the well pattern is calculated as:
43560 EV
Wp =
100
4WW F
w$R
+ 5.615
( 350n D2L
)
(5.33)
In-situ combustion field tests indicates that though the production-rate history vary widely from
one reservoir to mother, the oil production rates are low during the initial air injection phase and
increases rapidly as the heat of combustion approaches the producer. The peak production generally
occurs during the first half of the operating schedule. In the absence of specific information, for the
purpose of project design the oil production rate can be assumed to be proportional to air injection
rate. This implies a constant injected air to produced oil ratio and the oil producing rate curve will
have the same shape as the air injection rate curve.
The daily oil production NP is calculated by multiplying the average daily air injection rate by
the barrels of oil produced per MMscf of air injected.
167
Thus,
NP=
~v~l + (loo-Ev)N~
100
[ 100
106
bbl
x 43560X 0.626 X A
~~SCf
(5.34)
If the initial water saturation is immobile, no water will be produced during the initial air injection phase. As the combustion front progresses toward the producer, water production will increase
due to the formation of a water bank. The water production rate usually will accelerate once the flush
production from the oil bank ceases. To facilitate the estimation of water production rate, Nelson and
McNeil recommend that the produced water oil ratio to remain constant. This implies that the barrels
of water produced per MMscf of air injected will be a constan~
Wp =
Ev x
4WW F
106
0.626A
(350)n
In the following
~Z ~
sw@R
bbl. water
5.615 MMsc~
air
(5.35)
cedure.
Using the data given (Table 5.4), calculate the following:
1.
The total air requirement per acre-ft. for a 5spot pattern assuming an areal sweep efficiency of 62.670.
2.
3.
4.
The maximum air rate for the field pattern at a frontal advance rate of 0.5 ft,/day.
5.
6.
7.
8.
The duration of the constant rate period and the total time for the entire operation.
9.
168
horsepower
assuming three stages of compression and a compressor horsepower requirement of 80 million scf/day/stage.
11. The oil displaced from the burned reservoir, the oil displaced from the unburned reservoir,
and the total oil recovery and over-all oil-recovery
efficiency.
Solution
Laboratory Data: The combustion tube experiments showed that the fuel consumes per acre-ft.
of reservoir burned was 87,120 lb/acre-ft. and the air requirement was 388 scf/ft?.
5 ac
Pattern area
I ~istancebetween
injection mdproduction
wells
330 ft
Formation thickness
30 ft
Formation temperature
85F
production
14.7 psia
bottornhole pressure
35%
Porosity
500 md
Specificpermeability
Oil saturation
55%
I Water saturation
40%
30970
0.276 ft
well radius
169
Solution
The laboratory data gives the fuel burned as 87,120 lb/acre-ft. For on oil gravity of 1.0, the fuel
burned becomes:
87,1201bs/acre - ft
= 248.91 B/acre - ft
3501bs/B
Each item of the method is calculated below.
1.
The air requirement for the laboratory experiment was 388 scf/ft.3. This can be converted to
acre-ft. as
For a 5spot pattern with areal sweep efficiency of 62.6%, the total air requirement is
(~9
~ . ;)(0.626)
2.
= 10.58 mscf
acre - ft
Total Air =
( 388Scf \
[
ft)(
Total Air
3.
= (388~)(Oo626)(2)(330)2(
30)10-b Scf
Scf
= 1,587 mmscf
The air flux for a burning front advance rate of 0.125 ft./day is
170
4.
= (48.5)(4.77)(330)
x 1(H
(30)1W
2.29
Time =
,cj Scf
10
D
2~h (388@
(
ft, )(k)
~ Scf
2.29 X 10
tl
= tie
6.
= 2Z(30)(388)(~5)2
= 1253 ays
(125.3 days)(2.29m~f)
VI =
2
VI = 143.4 million scf
7.
The volume of air injected during the constant rate period is (See Figure 5.12)
Vz =V~o~ - 2(143.4)
Vz = 1587-
2(143.4)
171
8.
1300mmscf
t2 =
2.29
mmscf
567.75 days
t~~ = t + 2t~
9.
2.24 years
where
Pw = 14.7 psia
9, = 2.29scf/D
pa = 0.0186 Cp
kg = 5%(ksp)
= (0.05)(500)
h=30ft
a = 330 ft
172
= 24md
rx = 0.276 ft
VI =
0.5 ft/D
t =125.3days
(2.29)(10)(0.0186)(545)
P2 =(14.7)2+
lW
(0.703)(25)(30)
(330)2
{[
n (0.276)(0.5)(125
1}
.3) 1238
[(
bhp=(4)
sequence is"givenby
2.29 ~cf)(qstages)(ggmm::tage)
is
173
N,=(43,560g)[(so@/5.61
xlo4)-~]
ZVl = (43,560)
;:::$
1=
1.246X 103
-~
[.
B
acre - ft
where SO= oil saturation, $ = porosity, and WF = fuel consumed per ft.3.
lb
~
1
WF = (87,120
= 2.Olb / ft3
acre ft) ( 43, 560ft2 / acre )
\
(43,560)(5
6~lo,
w)
N,=
where E~v~= average efficiency of recovery from the unburned portion of reservoir.
Let E,v~ = 40%. Then,
= 43,560(55)(35)(0.4)
2
5.61
= 598 ~ B/acre
.
104
_
3-
JWl
+ (100
100
q~z
100
174
_ ~t
N, =~(1.246
x103)+
(100 - 30)(598)
100
N~ = 792.3 B/ acre - ft
The overall recovery efficiency is given by
_ N,(5.61 x1 O)
R-
43,560SO~
(792.3 B/acre
-ft)(5.61
x 10)
(43.560)(55)(35)
ER = 53%
12. The oil recovered per million scf of air injected is
oil
792.3 B/acre
- ft
175
= 74.9 B/mmscf
In-Situ
Combustion:
Oil Volume
Burned
Method
Gates and Rarney (1980) presented an engineering method for calculating ati-oil ratios and oil
recovery as a function of volume of reservoir burned. Their oil recovery-volume
based on laboratory data and pilot and field data from Mobils
burned method is
provides a means for making engineering and economic evaluations for the design and monitoring of
ISC projects. The reliability of the method is presently limited to reservoirs with characteristics similar to South Belridge;
100
90
INITIAL GAS
SATURATION
%
/
80
70
60
/
50
40
30
20
10
0
10
20
30
40
VOLUME
50
60
BURNED,
70
80
90
100
Yo
The basis for the design method is the observed relationship between oil displaced and the bulk
reservoir swept by the burning front. (Figure 5.12) If the initial oil and the fuel per unit volume
burned remained constant, the relationship between displaced oil and volume burned would be linear.
176
In actual cases, the oil recovered is greater than predicted because of other recovery mechanisms
(e.g., hot water drive, steam drive, hot gas drive, miscible phase displacement, or expansion and gravity drainage) acting on the oil zihead of the burning front. Gates and Ramey (1980) showed that several factors influenced the oil recovery-volume
fuel concentration. The Gates and Ramey method is discussed in much more detail in the following
section. Here, the salient features of this design methodology
is summarized.
The fuel concentration is the oil which remains and is burned as the burning front progresses. It
depends upon the oil and formation characteristics and the burning conditions. The fuel concentration
is the largest factor by far in determining the air requirement. Gates and Ramey (1980) present five
separate approaches to determine the fuel concentration. The fuel concentration normally increases as
oil gravity decreases. Hence for low gravitates, percent oil recovery is reduced and the air required is
increased. Often, however, this is more than offset by the high oil content of heavy oil reservoirs. The
cumulative ti-oil
ratio for the Gates and Ramey method is less than required for pure frontal dis-
placement.
When the leading edge of the burning front arrives at the producing wells, excess air which has
b~assed
duction-well
operating techniques. For example, production wells producing gases with substantial
Initial oil-in-place.
data.
operational data) presented by Gates and Ramey (1980) can be used to estimate the air-oil ratio,
instantaneous oil rate, and cumulative oil production. These are fully discussed in the following.
177
Methodology
To determine the amount of displaced oil, initial oil and gas saturations must be determined
using conventional well logging, coring, material balance or tracer techniques. The displaced oil is the
initial oil minus the final oil minus the burned oil. Fuel concentration (C~) is another important parameter in evaluating an in-situ combustion project. Fluid properties, lithology of formation and operating
conditions all affect the value of C&
Several methods can be used to estimate CP They are: a) coring the reservoir as the combustion
progresses; b) measuring the water cut and correlating with C~ by material balance; c) averaging the
value of C~ obtained from combustion tube runs with natural core; d) history-matching
the combus-
tion behavior in the field using a numerical simulation; and e) using the burning velocity/air flux correlation. In t.ke absence of other data, engineering calculations can be made using correlations of fuel
concentration vs oil gravity to determine Cfi (Alexander et al., 1962; Showalter, 1963)
To compute the cost of air compression, the value of the combustion air requirement should be
determined. This can be calculated if oxygen utilization (U) and the volume of air needed to bum a
unit weight of fuel (AFR) are known. AFR can be from the combustion chemistry. After the burning
front breakthrough, more air must be injected to compensate for the air produced because of channeling. Knowing the AOR, the oil production rate can be determined if the air compressor capacity is
known. The data for air requirement, fuel concentration, initial oil and gas saturations, and oxygen
utilization can be combined with an oil-recovery/volume-bumed
178
/ c=
co,
1.209 *
c.
(5.36)
+ co
10-3q~[C02
-.
- 0.5C0j
+ C0][12
.
+ H/C]
(5.37)
479.7 N,
AFR =
(C02 +CO)[2
Cf
B=.
Pf
(5.38)
+H
k)
43560
350
(5.39)
R= SOi-B
(5.40)
(A/p %)(R)(A)(H)
(5.41)
100
0.9(NP %) 15.85
Excel Air
(5.42)
100
V.(0)
~=
(5.43)
(5.44)
=0.147143 S- + 0.010714s!
v, -v,(o)
(5.45)
Y=
Deviation
Maximum Deviation
= 6.775267-
= 6.77526X
31.895588X+
100
M.D.
dY
100 ~ (o)+
100-v
(0)dx
0.46787S~
-15.947794X2
48.561561X2
(5.46)
+16.187187X3
-28.058636X3
-7.014659X4
(5.47)
(5.48)
(5.49)
179
ASR
Current AOR =
(5.50)
m
N,%
_ (Np
(5.51)
=1OOX + (Y)(M.D.)
%)(R)(A)(H)
(5.52)
100
Air Required =
Cum. AOR =
Time =
(AsR)(.)(@B
100
(5.53)
Air Required
(5.54)
Np
Air Required
(5.55)
Excess Air =
0.9 (,NPYo}15.85
(5.56)
100
Total AOR = Current AOR (1 + Excess Air)
(5.57)
AFR=
ASR =
B=
Cf =
co~ =
co=
180
H=
Thickness, ft.
H/c =
N2
Np =
Np% =
o~ =
q=
qg =
rt =
R=
Sg =
S~i=
VB =
Vf =
rf =
181
Satman
Brigham
Correlations
Satman and Brigham (Satman et al., 1981; Brigham et al., 1980) correlated the injection production history from 12 dry combustion projects and presented two correlation to predict the field wide
oil recovery of dry in-situ combustion process. They developed an analytical model describing the
heat transfer processes and movement of the steam plateau ahead of the combustion zone. The results
of laboratory combustion tube runs were used to verify the model results. The model was then applied
to field data to develop correlation to predict field scale recovery of dry in-situ combustion processes.
They used a combination of engineering and statistical approach (multiple linear regression analysis)
to develop the correlation. The correlation work is summarized in the following.
Correlation
Technique
The first step in the development of the oil recovery correlation was to plot the cumulative incremental oil production (CIOP) versus the cumulative air injected (CAI) for the 12 field wide tests as
shown in Figure 5.13. Here cumulative incremental oil is only that recovery caused by the combustion
process itself. In Figure 5.14, both coordinates were normalized for field size by dividing the abscissa
by the oil in place at the start of combustion (OIP) and the ordinate by dividing by the original oil in
place (OOIP) to yield the fractional recovery.
182
10
20
30
CAI, MMSCF
1. Glen Hummel
7. Brea Olinda
2. Moco
8. Notth Tisdale
3. Heidelberg
4. Trix Liz
5. Gloriana
@ 11. MIGA
6. West Newport
10. Fry
12.
N. Government
14
12
10
0.2
0.4
0.6
CM
~
..
0.8
MSCF/BBL
1. Glen Hummel
2. Moco
8. North Tisdale
3. Heidelberg
9. West
4. Trix LiZ
10. Fry
5. Gloriana
11. MIGA
6. West Newport
12.
if
1.2
Casa Blanca
N. Government
It must be recognized that the air injection is really a measure of the amount of heat added to the
reservoir, because in all combustion operations a given volume of air supplies nearly a fixed amount
of heat from combustion.
Since a large percentage of the heat added to the reservoir is stored in the
rock and left behind by the combustion front rock volume is an important parameter in the oil recovery calculation. The rock volume is calculated as:
Rock Volume =
:(I-@)
(5.58)
Since, in combustion
injected is less than the actual volume injected. Hence the next step in the correlation development
technique is to divide the abscissa by the rock volume and multiply it by the oxygen utilization, factor
(o#t).
Thus
rw.
Ww-?l
(5.59)
Since some of the oil in place is burned and therefore not available for recovery, the ordinate
must be modified to account for this fact. The ordinate is modified by adding the fuel burned (FB) to
the CIOP. Thus
CIOP + FB
Ordinate =
(5.60)
00IP
The results shown in Figure 5.15, using these parameters is clear improvement.
185
14
12
6
4
2
0
0
0.05
CAI
0.2
0.15
0.1
0.25
0.3
Glen Hummel
8. North Tisdale
Moco
Trix Liz
10.
Gloriana
11. MIGA
West Newport
12.
Fry
N. Government
Brea Olinda
L
FIGURE 5.15 Effects of Fuel Burned, Rock Volume, and Oxygen Utilization on Cumulative
Incremental Oil vs Air Injection for Fieldwide Combustion Tests (After Satman et al., 19S1)
186
The correlation of Figure 5.15 was further improved using the multiple linear regression analysis to include additional parameters expected to be of important, but not accounted for in Figure 5.15.
These include, the formation thickness (h), oil saturation (SO), and oil visocity (me). The ordinate in
Figure 5.15 was correlated against the abscissa as a linear function of oil saturation, thickness, and oil
viscosity to yield.
WhereY
(?io) =
CIOP + FB
00IP
~d
x (MSCF
bbl
h -0.0082
~) X
(5.61)
x 100
(5.62)
CAI(02U*
)
) = (omv@.$J(l
-~)
air
=
.
oil atlo
187
(5.63)
14
12
0
8
6
4
2
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
1. Glen Hummel
8. North Tisdale
2. MOCO
H 4. Trix Liz
M 10. Fry
5. Goriana
6. West Newport o
7. Brea Olinda
11. MIGA
12. N. Government
FIGURE 5.16 Multiple Linear Regression Analysis and Data on Figure 5.16
(After Satman et al., 1981)
188
12
10
8
$?
0
0
0.05
0.1
0.15
0.2
0.25
0.3
FIGURE 5.17 Fkst Correlation Curve for Dry In-Situ Combustion Field Cases
(After Satman et al., 1981)
Figure 5.16 shows the actual values of the recovery function (Y) versus the valves on the righthand side of the correlation and indicates considerable improvement in the resulting correlation. A
smooth curve drawn through the data Figure 5.16 yields the general recovery correlation curve shown
in Figure 5,17. The correlation in Fig 5.17 can be used to predicts dry combustion oil recovery as a
function of air injected, given the values of SO,h, ~, & oxygen utilization (Oz UJ, and fuel content.
The first four of these are readily available, while the last two are normally must be obtained from
laboratory combustion tube data.
189
Caution must be exercised on the use of this correlation. This correlation may not be valid if any
of the parameters are outside the range of data used to develop it. The ranges of oil saturation, oil viscosity, and reservoir thickness used to develop the correlation were:
0.36 <SO <0.79
lo<~<700cp
4.4 <h c 150 ft.
It should be noted that the above relationship is linear with viscosity, with a negative coefficient
mid thus will predict a lower recovery if the in-situ visocity is of the order of several thousand centipoises. To rectify this deficiency, Satman et al. modified the right-hand side of the equation to yield
the following correlation.
025
Y = 47.0
[
Y (%)=
01
CIOP + FB
OIP
(5.64)
(5.65)
x 100
And
MSCF
[1,/21
al x(O,ut)
(5.66)
(oly/,o )(H)
Figure 5.18 shows the actual values of the recovery function (Y) versus the values on the right
hand side of Eq. 5.64 for each of the 12 fields. Comparison of figure 5.16 shows that the second correlation (Eq. 5.64) fits the field data considerably better than does the first (Eq. 5.61).
190
14
,Yfj
,d N-
12
o I
;i f
,/
/
1 l-w
10
F~?;/
l\
0.05
0:1
0.15
2.196 ~
(Po)
0:2
~1
0.25
CAI X(Percent ~
0:3
Utilization)
](olF/@so)(l-O)
1. Glen Hummel
&
7. Brea Olinda
2.
B.
3. Heidelberg
4. Trix Liz
5. Gloriana
(j
11. MIGA
6. West Newport
12. N. Government
MOCO
North Tisdale
191
0.05
0:1
((
0.427S0 - 0.00135h + 2.19 ~
())]w
~~
o.i5
0:2
0.25
0:3
0)
.FIGURE 5.19 Second Correlation Curve for Dry In-Situ Combustion Field Cases
(After Satman et al., 1981)
Although the second correlation (Figure 5.19) is better, it is also limited in its applicability to
the same ranges of parameter
recovery
better
for rese~oirs
oil
10 cp or less the first correlation should be used to predict recovery. Further whenever there has been
considerable recovery prior to the onset of in-situ combustion, the second correlation is preferable to
the first one. In general, for most application, the use of second correlation is recommended.
Application
of Correlation
To test the validity of the developed correlation, Satman et al. (1981) used the correlation to calculate the performance of eight dry in-situ combustion pilots. In Figure 5.20, the field data were plotted. This figure can be used to determine the air injected/oil produced ratios. The Figure 5.20 is then
re-plotted in Figure 5.21 by taking into consideration the rock volume, fuel burned, and oxygen utilization. Finally the second correlation was used to improve the results of Figure 5.20 as shown in Figure 5.22. In general, this plot confirms the validity of the developed correlation.
192
120
r
1
00
0:4
0:6
C 1
1.2
1.4
1.6
1 8
CAI, MMMSCF
1.
f= 4. North Tisdale
5. West Newport
6. Fry (A-1 )
7. S. Belridge
8. Bellevue - Getty
FIGURE 5.20 Cumulative Incremental Oil Production vs Cumulative Air Iniection for Pilot
Dry Combustion Tests (After Satman et al., 1981)
70
60
50
*.
40
f$
+ &
g 5
u
30
20
10
0:5
1:5
2;5
5. West Newport
6. Fry (A-1)
7.
4. North Tisdale
8. Bellevue - Gettv
South Belridge
FIGURE 5.21 Effects of Fuel, Rock Volume, and Oxygen Utilization for
Pilot Dry CombustionTests (After Satman et~., 1981)
194
/u
60
30
20
10
0
0
..
0.2
r0.427S.
0.4
0.6
0.8
()
- 0.00163h + 2.196 ~
1.2
1.4
1.6
1.8
.25
-0)
$J 5. West Newport
6. Fry (A-1)
7. South Beldridge
4. North Tisdale
8. Bellevue - Getty
FIGURE 5.22 Dry Combustion Field Performance Prediction Using Second Correlation
195
The Thermal Task Group of the National Petroleum Council (NPC) developed an In-Situ Combustion predictive model (ICPM) for use in the 19821984
recovery potential. They adopted the Satman-Brigham
ICPM contains an extensive set of default equations to calculate the non-critical reservoir and fluid
properties, and economic
was added to
ICPM by NPC based on the work of Garon and Wygal (1974). The U.S. Department of Energy (DOE)
supported the NPC study and has maintained the model since the NPC study was completed. ICPM
can be used to predict the recovery performance and economics
Office (NPTO) in Tulsa, Oklahoma. A ready to run version of the model can be down-
196
REFERENCES
1.
Alexander, J.D., W.L. Martin, and J.N. Dew (1962), Factor Affecting Fuel Availability and
Composition During In-Situ Combustion, J. Pet. Tech. (October), pp. 1154-1164.
2.
Combustion Field Projects and Application to Combustion Pilot, J. Pet. Tech. @ecember),
pp. 21322138.
3.
Chu, C. (1982) State of the Art Review of Fireflood Projects, J. Pet. Tech. (January), pp.
1936.
4. Fassihi, M.R., B.D. Gobran, and H.J. Rarney, Jr. (1981), An Algorithm for Computing InSitu Combustion Oil Recovery Performance, U.S. Dept. of Energy Report DOE/ET/1205625 (DE8103O34O), (October), p. 25.
5.
Garon, A.M. and R.J. Wygal Jr. (1974), A Laboratory Investigation of Fire-Waterflooding,
Sot. Pet. Eng. J. (December), pp. 537544.
6.
Gates, C.G. and H.J. Ramey, Jr. (1980), A Method for Engineering In-Situ Combustion Oil
Recovery Projects, J. Pet. Tech. (February), pp. 285294.
7.
8.
(DE86000284),
(December).
Project,
9.
Satman, A., W.E. Brigham, and H.J. Ramey, Jr. (1981), In-Situ Combustion Models for the
Steam Plateau and for Fieldwide Oil Recovery,
12056-11,
(June), p. 112.
10. Showalter, W.E. (1963), Combustion Drive Tests, Sot. Pet. Eng. J. (March-April),
58.
pp. 53
198
CHAPTER
COMBUSTION
CASE
6 IN-SITU
AND PERFORMANCE
ANALYSIS
HISTORIES
Since early 1950, more than 270 field tests have been conducted in the U.S. and elsewhere in the
world. A large number of the field tests were conducted in reservoir situations that were not suited to
firefiooding
and therefore the results reflect only where the method should not be attempted. Oil
industry economics
now. However, all costs have increased along with the price of oil, and the availability of fuel has
changed considerably, which makes it necessary to make a complete reevaluation of each prospect.
In ihe following
five firefloods will be reviewed in detail to present the latitude to which this
method has been used, from the very shallow low pressure reservoir of the General Crude, New Port
Beach, California fireflood to Gulfs deep Heidelberg Cotton Valley Air Injection Pressure Maintenance Project.
Miga
Fireflood
In 1964, Mene Grande Oil Company started a fireflood in the Miga Pz_~Sand reservoir. (Terwilliger et al., 1974). This test had been preceded by two small volume fireflood tests in the nearby Melones area.
There is a tremendous volume of heavy oil at moderate depths (2,0004,000
of the Eastern Venezuela tertiary basin. Development
ply of lighter gravity blend oil was available from nearby fields. However, the recovery from the reservoirs containing 1O-15API oils was very low. Solution gas drive recovery was of the order of 5%
of the oil-in-place,
and because of the poor mobility ratio between water and the very viscous oil,
water drive recoveries were not much better and in some cases were even worse. Even though the
industry experience with fireflooding was rather limited in 1958, it appeared that this method could be
used to increase the recovery from the Eastern Venezuela heavy oil reservoirs from a few percent to
more than half of the oil-in-place.
The Miga fireflood was designed to furnish basic data on the applicability of the fireflooding
process to these heavy oil reservoirs. Although the reservoir oil viscosity is a few hundred centipoises,
the very high permeability causes the per well productivity to be in the range of 100-300 B/D. The
199
Miga Pz.q reservoir was chosen because it was nearly depleted, was small compared to other reservoirs in the area, had only a few completions in deeper reservoirs, and was completely closed with a
limited aquifer at one end. With these features it was expected that response to the fireflood would be
rather rapid, and injection versus production could be closely controlled.
The test has been very successful from a technical standpoint. However, it has been economically unsuccessful because of the depressed price of heavy oil.
Table 6.1 is a summary of the reservoir properties and project performance.
22.670 is much lower than was originally calculated for this unconsolidated
The porosity of
were based on routine core analyses of conventional cores, and more recent values are from rubber
sleeve cores and density logs that have been run in new wells. The estimated ultimate primary recovery was determined from individual well decline curves and a comparison with other similar reservoirs.
Reservoir
Description
ela. It is one of the several Pz_~reservoirs scattered throughout both Miga and the neighboring Oleos
Fields. The updip seal is a combination
tighter as it thins, therefore the 10-foot isopack is considered to be the lateral limit. The downdip limit
is formed by the original oil/water contact and a fault. There may have been a very small gas cap at
the updip limit of the reservoir.
4,050 ft.
Q Depth
1525 ft.
Thickness
Permeability
Porosity
22.6%
Gravity
13API
Oil Viscosity
280 Cp
Temperature
146F
4,396,162 bbl
3,624,383 bbl
3,236,162 bbl
110 Darcys
39,948 MMScf
11,023 Scflbbl
Q Cumulative AOR
Oil Prod. (Prim.)
771,779 bbl
The reservoir was developed in 1958 and was soon producing 1,000 B/D. After about a year the
production began to decline. An attempt was made to arrest the decline by injecting natural gas in the
most updip Well MG525,
the same well that was later to be used for air injection. The production
rate was hampered rather than helped by the gas injection, as inferred from the dip in production rate
in 1961 and 1962. The reservoir was then shut-in while the fireflood facilities were being installed.
The increase in production rate in 1964 was the combined effect of the shut-in period and the fireflood. The decline in production in late 1964 was a deliberate effort to curtail liquid production and
encourage gas production.
201
by the fh-eflood was more than twice the estimated ultimate primary
recovery. No other method of secondary or enhanced recovery would have been as successful as fireflooding in this reservoir.
Project
Production
Response
and was increased to 15 MMCfd in late 1964.
in 1969 when part of the air plant capacity was used to start a
fireflood in another reservoir. The air-oil ratio increased for a short period of time in 1965 when oil
production was being curtailed to permit equilibrium liquid saturation conditions to be established.
The early breakthrough to the producing wells indicates that oil displacement by the gas drive is
practically @l. The combustion reaction moving through the reservoir is required to displace the oil.
The composition
of the fireflood flue gas stabilized very early, indicating complete utilization of the
injected air. The composition of the gas has been very constant throughout the life of the project, indicating that there has been no deterioration of the fireflood.
In a fireflood there seems to be no one well that is typical; therefore, Well 524, which was close
to the original air injection well, and Well 817, which was at the other end of the reservoir, were used
as examples. The oil production at Well 524 declined from 300 BOPD to 150 BOPD before the injection point was moved to the other end of the reservoir. The decline is possibly a result of asphalt precipitation around the wellbore caused by the solvent action of the carbon dioxide from the fireflood
flue gas. A rapid rise in the water production in 1970 and 1971 is a result of injecting water in the
original air injection well. This was done to recover some of the heat that was left in the reservoir
around the air injection well.
The gas-oil ratio was erratic at Well 524 in the early life of the fireflood. This was in part caused
by the deliberate curtailment of the liquid production. However, it leveled out and was relatively constant until the injection point was changed.
The gas composition
stabilized rather quickly at Well 524, even though it was 3,000 ft. from the
202
rate declined slightly until the injection point was changed from a distance of 7,500 ft. away to only
2,000 ft. away.
Although Well 817 was one of the better producers, its gas-oil ratio was poorer (higher) than the
average of 11,300 cu ft./bbl.
The early values of carbon dioxide exceeding the fireflood flue gas carbon dioxide was probably
a result of the oil in the reservoir around the newly drilled well having a large amount of carbon dioxide in solution. When the well was put on production and the local reservoir pressure lowered, the carbon dioxide came out of solution. The nitrogen production from this well was more erratic than the
average for no apparent reason.
Conclusions
Fireflooding
Cotton
Chevrons
Project
(formerly Gulf) Cotton Valley air injection project (West Heidelberg Field, Jasper
County, Mississippi) began in December 1971, with the start of air injection into the 11,000 ft. Cotton
Valley No. 5 sand. Like the Miga project, it is another example of how successful fireflooding can be
when applied in a suitable reservoir.
The Cotton Valley No. 5 sand (CV5) is one of 14 sands in the Cotton Valley formation which are
identifiable. Eight of the sands are productive, but only the No. 5 and No. 4 sands are of appreciable
significance. Both of these sands are extensively developed and cover about 400 surface acres on the
western flank of the field.
The CV5 sand occurs at an average depth of 10,850 ft. subsea. Average sand thickness is 35 ft.
The trap is a monocline having a dip of about 8 from east to west. The eastern updip limit of the sand
203
is bounded by a salt intrusion and the downdip limit by a zone of thick asphaltic oil which effectively
seals the Cotton Valley formation from any active water drive.
Core analyses for the CV5 sand indicated 17.8% porosity, 6080 md permeability and 15%
interstitial water saturation. Original reservoir pressure was 5,045 psia. Reservoir fluid analyses indicated a bubblepoint pressure of 930 psia and an oil viscosity of 6 cp at the original reservoir pressure
and reservoir temperature of 220F. Oil gravity was 23API.
As of June, 1981 (no data published beyond this date), cumulative oil production from the CV5
sand by air injection was 2.07 MMstb or 16% of the estimated 13 MMstb of oil originally in place.
The oil production rate at that time for the CV5 sand was 1,100 Stb/D from eight producers compared
to 60 Stb/D at the start of air injection. Peak oil production rates have been as high as 2,000 Stb/D.
The air injection averages at that time was 5 MMscf/D
to 5 MMscf/D.
producers would be recompleted in the next higher No. 4 sand. The No. 4 sand would then be produced by pressure maintenance by reinfecting flue gas produced from the No. 5 sand back into the
No. 4 sand.
Project oil production from the No. 4 sand began in December
1977, followed
by the start of
flue gas injection in June 1978. In January 1980, a dual air injection well was completed to inject air
separately, through isolated tubing strings, into both the No. 4 and No. 5 sands.
As of June 1981, cumulative project oil production from the No. 4 sand was 186,000 Stb or
about 290 of the 12 MMstb of oil originally in place. The injection rates at that time for the No. 4 sand
were 2.5 MMscf/D
the No. 4 sand pending repairs on the dual air injection well. Oil production at that time from the No.
4 sand was 560 StblD from 5 producers. The cumulative air-oil ratio was 2,505 Scf/Stb.
204
The known status of the Cotton Valley air injection project is summarized in Table 6.2.
TABLE 6.2 June 1981 Status of the Cotton Valley Air Injection Project
CV5
CV4
2.07
0.186
16
1,100
560
2,662
2,505
2.5
3,250
2,850
13
12
1.4
1.4
West Newport
Fire@ood
The West Newport Iireflood was started in 1958 in the center of Mobils
Kadane & Sons) (and later General Crude) Banning lease. the fireflood was essentially a series of
irregular inverted 5spot patterns. The major reservoir being burned was the B sand with a net
thickness of about 500 ft. consisting of a highly porous and permeable oil sand with scattered shale
stringers which had little continuity. Depth of the B sand was from 1,4002,000 ft. and contained a
15API crude with a viscosity of 3,700 cp at reservoir temperature. At that time there were a total of
205
245 producers, some of which produced from the A and C sand reservoirs. There were 35 air
injectors in the B sand and 1 in an A sand pilot. Production was 4,000 BOPD with 18,000
13WPD. The B sand was productive over 590 acres. Original oil in place was estimated at
120,000,000 barrels. Cumulative oil production at ll76
injection plant capacity was about 24.0 MMscfpd with the injection rate at that time being 11.0
MMscfpd at 550 psi. The ti-oil
ratio (AOR) ranged from 2-4 Mcf per barrel with the ratio at that
Wells
Producing wells were completed by drilling a 97/8 in. hole to the top of the B sand where 7
in. casing was set and cemented to the surface with Class G cement containing 4% gel and 35% silica flour. A 6-1/8 in. hole was drilled to the base of the B sand and underreamed to 12 in. A 5 in.
liner was set and gravel packed with no liner hanger being used. All wells were produced by beamtype pumping units. Producing rates were kept up by cyclic steaming all but hot wells every six
months with 6,0008,000
by use of the gravel packed liners. Each well was equipped with a small gas scrubber and orifice
meter to measure gas flow rate and temperature from the annulus. Downhole temperature profiles and
gas sampling were periodically carried out.
Hot wells were producing wells that were being approached by the firefront which was indicated by increasing temperatures and volumes of flue gas. When a well was about to enter the hot
stage, the normal sucker rod string was replaced by a hollow rod string (1 in. pipe), but was pumped
up the tubing in a normal manner. The hollow rod string was utilized as a conduit for a thermocouple
circuit with the temperature sensing element being placed opposite the hottest section in the wellbore.
The thermocouple was used to measure downhole temperature and automatically controlled a water
injection system used for cooling the downhole producing string and also controlled backpressure on
the annulus by restricting flue gas flow. If downhole temperature increased above 275F, water injection was automatically started and maintained at a rate sufficient to keep temperature at or below that
level. At the same time, the rate of flue gas flow was restricted to a selected optimum dependent on
fluid production rate. In this manner, the fireflood front was slowed and diverted away from the hot
well allowing the well to be produced much longer. This technique appeared to be successful and
allowed hot wells to be produced for several years longer, substantially increasing oil recovery and
206
reducing corrosion problems. Mobil (General Crude) believed that this technique increased sweep
efficiency particularly in a reservoir of this thickness.
Injection
Wells
Air injection wells were completed by drilling a 77/8 in. hole to the base of the B sand and
cementing to the surface with the same high temperature cement mix used on producing wells. An
interval in the bottom third of the B sand was jet perforated and 23/8 in. tubing run. Steam injection was carried for at least 24 hours and then air injection started at low rates. Ignition was spontaneous and occurred in one hour to one week. Mobil (General Crude) utilized a portable, high pressure,
low volume compressor for this stage of air injection. Ignition was evidenced by a 300500 psi injection pressure drop and sometimes was felt physically. Air injection was continued for about two
months when the tubing string was replaced by plastic lined 2-3/8 in. tubing with a packer. Injection
rates were very gradually increased to a preselected rate determined by pattern size, producer proximity and injection well performance. The average air injection rate at that time was about 3,120 Mcfpd
per well. Mobil (General Crude) operating and engineering personnel believed strongly in using low
air injection rates, particularly on start-up. Attempts at backflowing
Production
Facilities
Production was gathered at a central battery where it came in at about 120F. A reverse emulsion breaker, oil soluble wax inhibitor and defoamer were added at a freewater knockout with steam
coils where the temperature was raised to about 150F. Additional emulsion breaker was added and
the fluid went through two large horizontal heater-treaters which raised the temperature to 215F.
..
Fluid production then went to settling tanks for retention time and oil was split off to sales ;torage.
Water was sent to a nearby sewage plant for disposal. This plant furnished Mobil (General Crude)
with a volume of 670 BTU/Mcf
207
about 275F. The heat generated by this process was scavenged by a radiant heat section and generated steam for producing operations. Operation of the incinerators had been satisfactory from an ecological
viewpoint.
These
by Heater Technology,
Comments
This project was apparently an economical operation. Mobil (General Crude) had learned much
about iireflooding in the 18 years the project had been under way. The post project analysis indicated
good sweep efficiency for a reservoir of this thickness. This may have been due mainly to operations
policy of fireflood control and low air injection rates. On an overall basis, Mobil (General Crude) tried
to maintain a voidage rate 14% greater than the air injection rate.
The personnel of Mobil (General Crude) were highly involved in their project and had developed some excellent equipment and techniques to improve recovery and efficiency of the fireflood
process.
Table 6.3 summarized the available general information regarding this fireflood. These were
made available to the author by an engineer formerly associated with this project.
208
Reservoir Properties
West Newport Field-Banning Lease
Reservoir
Depth
800-1,200
1,400-1,800
2,200 ft.
Porosity
high
Permeability
Viscosity
3,5004,500
API
15
Reservoir Temperature
110F
Sulfur
15%
Sand Thickness
Dip
Acreage
590
Pattern
5spot
209
cp @ 110F
ft.
ft.
I
TABLE 6.3 (cont.) Mobil (General Crude) West New Port Fireflood
I
I
O~eratinE Parameters
Number of wells
Injectors
36
Production
245
Oil Production
4,000 B/I)
Water Production
18,000 BID
Steam used
Air Dressure
Steam rxessure
at
500 Psi
generator
at wellhead
Natural gas some from sewage plant
4.5
Best well
140 B/D
Air Compressors
2 stage, 500 psi output
22x 10 CFD capability-using
No. of units:
1 2,000 HP
2600HP
3 1,000HP
11 x 106 CFD
3.000 CF/B
TABLE 6.3 (cont.) Mobil {General Crude) West New Port Fireflood
5 units are used to generate steam (each use 1.5x 106 CFD).
The flame temperature is monitored and natural gas is used to enrich the fireflood flue gas
automatically.
Steam capacity 580 BID, 60% quality
Exhaust temperature kept above 270F to eliminate sulfi.nic acid formation.
Generate 14% more fireflood flue zas than air iniected.
.
Ignition of Wells
Warm up with steam-then
change to air
on surface
Casing
7-7/8 in. hole, 5-1/2 in. casing to TD
Injectors
Producers
2 company owned
Rig cost
$55-60/hr
drilling
$40/hr workover
Hole Cost
211
Paris
Valley
Combinations
Thermal
Drive
Husky Oil Company, with the support from U.S. DOE (then ERDA), initiated a wet in-situ combustion pilot augmented with cyclic
County, California, in March 1975. The purpose of this test was to assess the technical and economic
feasibility of these thermal recovery techniques within an unconsolidated
never produced economic quantities of oil due to the very viscous nature of the crude.
Wet combustion, in lieu of other thermal techniques, was selected as the oil recovery process in
Paris Valley field, because it was thought the heterogeneous permeability profile would permit heat to
breakthrough to the producers and improve the vertical sweep while the well produces at an elevated
temperature. Also, to match the production with displacement prior to the heat breakthrough stimulation of producers by steam was proposed.
The pilot site was in the southwest part of the Paris Valley field in T21S-R9E, Monterey County,
California, about 160 miles south of San Francisco. Geologically,
unconsolidated, oil bearing miocene sands that were deposited along the ancient shoreline of the Salinas basin--commonly
referred to as the Gabilan shelf. The pilot was in the Ansberry sand, which is
found at an average depth of 800 ft. from the surface. The Ansberry sand is separated into three distinct zones, referred to as the Upper, Middle, and Lower Lobes. The Middle Lobe is thin and contained insignificant amounts of oil. The total net oil sand thickness in the pilot area varies from 4-84
ft., while Upper Lobe net oil sand thickness varies from 4-24 ft. and the Lower Lobe from 958 ft.
The pilot was designed to operate in five staggered line drive patterns. Eighteen producers and
five air injection wells were drilled in the pilot area (Figure 6.1). Nine wells (1, 8,9,
16, and 18) were completed in the Upper Lobe, six were completed in the Lower Lobe, and eight
wells in the full interval of the Ansberry sand, including the Upper, Middle, and Lower Lobes. The air
injectors were completed down-dip from the center of each pattern in an effort to compensate for the
expected directional flow of air up-dip. Two temperature observation wells were also drilled for monitoring and data gathering. All the wells were cored and logged to characterize the formation. The
core materials from Well 3 were utilized to evaluate the combustion characteristics of the Ansberry
sand. The pertinent reservoir and combustion characteristics are presented in Table 6.4
212
&
UPPER ZONE
LOWER ZONE
o
I
500
I
?000
I
FEET
FIGURE 6.1 Paris Valley In-Situ Combustion Project Well Pattern Map
of Ansberry Sand
800
Porosity, % PV
32.2
Permeability, mD
3,748
Oil saturation, % PV
63.7
Water saturation, % PV
36.3
1,801
2.6
58
235
87
Reservoir temperature, F
10.5
227,000
Lower Lobe
23,000
1.0
2.33
295
Bbl/ac-ft.
417
18,165
Mscf/ac-ft.
1,296
214
Initial combustion test was initiated in January 1976 using a rental compressor, due to operational problem with the main compressor equipment. The initial testing was concluded in September
1976 after injecting 60 MMscf of air. The purpose of the test was to establish the formation air infectivity and to determine the air injection rate. The first well was ignited utilizing a downhole
gas
burner, which was set immediately above the perforation. The use of downhole burner was, however,
discontinued in later ignition operation because it was found that the formation could be autoignited
by injecting steam to supply the heat requirement for ignition.
After completing several repairs to the main compressor air injection was resumed in May 1977
in Wells 6, 8, 12, 15, and 18. It became quickly apparent that the injectivities in wells completed in the
Upper Lobe were significantly lower than Well 6, completed in the Lower Lobe. Air injection and
burning appeared to have been retained in the Lower Lobe of the Ansberry sand. Leading edge of the
heat zone reached updip wells completed in the Lower Lobe, but did not reach the upstructute wells.
High injection pressure became necessary to inject air into the Upper Lobe, which contained the most
viscous oil. High injection pressure resulted in severe channeling of air upstructure into Well 21. Air
injection was halted in August 1977 due to first stage piston failure. After repairing the compressor,
air injection was resumed in January 1978 in Wells 6, 12, 15 and 22. Air injection into Well 18 was
discontinued due to mechanical problems with the surrounding wells. Air injection was terminated in
February 1979, due to operations problems.
Well 21 was the only well that produced incremental oil from combustion operation. After heat
broke through in the well, a cooling system was installed in the well to prevent the bottornhole temperature from exceeding 350 F. However, failure of the cooling pump caused the bottomhole temperatures in Well 21 to exceed 700 F and darnage the liner. Attempts to replace the damaged liner were
unsuccessful and eliminated any further chance of producing the well. Heat also broke through in several other wells (Nos. 4, 11, 14, and 17), but combustion gas and high water production prevented
these wells from being kept pumped off and producing the oil.
Only 61% of the injected air was recovered in the producing wells. The balance was probably
flowed outside the pilot area. Tracer tests indicated the migration of unrecoverable air outside the pilot
area. Significantly high operating costs were incurred during the operation of the pilot. The electric
power costs for air compression increases by 233%. Required well work was impeded by high casing
pressure in the producing wells. Damage to producing wells from severe channeling of the combustion gas resulted in expensive workovers. Oil production rates were not sufficient to offset the high
215
operating costs and continue the project. In view of the problem prone operations and poor oil production, the pilot was terminated in March 1979 as uneconomical.
pilot were $3.317 million or $23.89/bbl.
The pilot did not generate any profit at 1978 oil price of
$24.bbl.
Project
Performance
Analysis
In spite of possessing many of the desirable attributes (thicker pay, high oil saturation, good lateral continuity, high permeability and porosity) that one would like to see in a candidate reservoir for
combustion, the Paris Valley combustion project was unsuccessful and failed to produce significant
incremental oil. While there are several interrelated reasons for failure, two factors adversely affected
the project performance: (1) high oil viscosity in the Upper Lobe and (2) operational problems resulting from poor planning. The vast difference that existed in the viscosity of oil produced from the
Upper and Lower Lobe had a very significant effect on the project performance. The Upper Lobe produced an oil that was ten times as viscous as that from the Lower Lobe. The viscous oil in the Upper
Lobe caused a viscous oil block to form and much higher air injection pressure was needed than for
the lower viscosity Lower Lobe. This resulted in a pressure gradient sufficient to allow the injected air
to break into the Lower Lobe. This channeling of air into the Lower Lobe caused much of the injected
air to bypass the upper zone and starve the Upper Lobe combustion front of oxygen.
Combustion
front stalling appeared to occur as characterized by a low static temperature profile in the Upper Lobe.
The maximum observed temperature in the observation wells completed
never exceeded 500F indicating the occurrence of low temperature oxidation (LTO) reactions. Since
this temperature was within the negative temperature gradient range (See Chapter 3) it is suspected
that he failure of the reaction temperature to transcend the negative temperature gradient region could
partly explain the poor oil recovery from the Upper Lobe. Air injection into the Lower Lobe was no
problem and the lower zone exhibited good burning characteristics as evidenced by high observation
well temperature (greater than 750 F). However, the channeling of air from the Upper Lobe into the
Lower Lobe affected efficiency of downhole pumps in the Lower Lobe.
Frequent compressor failures also hastened the demise of the project. When air injection was
interrupted due to compressor failure, backflow occurs. Due to the unconsolidated
mation, the backflow resulted in severe sanding at the injection wells and costly workover. Further,
when air injection was interrupted, vertical drainage resaturated the burned zones. This resaturated
rock was hot and must be burned again before the burning front can proceed. This increased the over-
216
all air requirement of the project. In at least one instant, the backiiow of the combustion gases into an
injection well resulted in an explosive
resumed.
In the final analysis, the project failed due to poor selection of test site and planning. The combustion tube tests indicated an air-oil ration (AOR) requirement of 18.2 Mscf/bbl
values reported for the economically
and economic
project wells gave reliable results for porosity, permeability, and oil saturation, but did not indicate the
vast difference in the viscosity of oil produced from the Upper and Lower Lobe, A production test
from a well open to both zones ,was used for the productivity calculation and average viscosity. The
result was very misleading, and the project was designed based on this information. This resulted in
the purchase of an undersized compressor, which later proved to be totally inadequate for delivering
air at the desired injection pressure. The situation could have been avoided, had each zone been tested
separately and oil viscosity from individual zones determined prior to the design of the project.
While the effective permeability in both zones were almost equal, injectivities were widely different. This was not recognized
during the planning and design stage. The viscous oil in the Upper
Lobe caused a viscous oil block to form which necessitated a much higher air injection pressure that
exceeded
the compressor
217
Bodcau
In-Situ
Combustion
Project
1994)
The Bodcau fh-eflood was a DOE-industry cost-shared in-situ combustion project that was both
technically and economically
successful. In 1971, Cities Service (now Oxy USA) began a pilot com-
bustion test in the Bellevue Field, located in northwestern Louisiana, which was later expanded to a
leasewide operation. The success of this in-situ combustion project prompted Cities Service Company
to enter into a cost-sharing contract with the U.S. Department of Energy (then ERDA) to demonstrate
the efficiency
niques for increasing vertical sweep efficiency while reducing overall project time.
The Bodcau in-situ combustion project was conducted in the Bellevue Field, located about 18
miles northeast of Shreveport, LA, on the eastern edge of Bossier Parish. The demonstration site,
located in the southwest quarter of Section 11T19NR1 lW (Figure 6.2) was part of Cities Service
Oil Companys (now OXY-USA)
400 ft. depth, is the main producing sand in the field. The demonstration site was selected based on
the data from five evaluation wells. The log and core data obtained from these wells were utilized in
mapping the structure and pay thickness, as well in determining the pattern size and configuration.
The patterns were about 4 acres in size and arranged in an elongated inverted 9spot (Figure 6.3). The
injection well was located down-structure to compensate for the movement of air up-structure. The
patterns were elongated up-structure to provide for optimum sweep efficiency in the patterns. Five
patterns were developed due east of Cities original fireflood project. The reservoir and fluid characteristics are shown in Table 6.5.
218
L -2
4-3
4-4
2-3
oo
[2-3
/2-2
i-3
12-7
I G?-/
41
I
0
2-4
2-5
12-6
2-s
o *o
12-4
6-4
IS-S
15-
!4-7
13-7
0 1
I ,3-5
3-6
14-6
00
o
&6
15-s
I 1s-5
f6-6
.1 L
-oooq
200
40
600
219
Depth, ft.
450
19
54
Average porosity, 70
Average permeability, rnD
33.9
700
Water saturation, % PV
27.4
72.6
1,909
40
Reservoir temperature, F
75
19
676
4.5
220
using electric heaters in August and September of 1976. During the first six months of operation, air
was injected down the casing for the dry burn phase. After the injection rates stabilized at their maximum, the injectors were reperforated in the top 10-12 feet to allow water injection into the upper section of the Nacatoch sand simultaneously with air injection into the base of the zone. A limestone
interval provided partial separation of the two injected fluids. The purpose of water injection was to
improve vertical sweep efficiency by forcing the combustion to expand farther out in the lower section
of the reservoir before rising to the top, thereby heating a large volume of the reservoir.
During the combustion phase, the producing wells experienced hot gas breakthrough or even
burnout as combustion front approached the producers. When this occurred, the wells became impossible to operate due to sand production or tubing leaks resulting from the blowtorch effect of hot gases
on the tubing. When a producing well became difficult to operate, the tubing was pulled, and the
depths of the markings on the tubing were noted as an indication of the interval through which the
sand laden hot gasses were flowing. The producing interval was squeeze cemented and only the sections not causing the well problems were reopened for production. This process was repeated several
times during the combustion phase.
In spite of considerable care in design and operation, an explosion occurred in the air compressiordinjection
system, destroying
one compressor.
Buildup of a lubricant film on the inside walls of the air injection line was determined as being the
cause of the explosion. The distribution lines were repaired and put back into service within 36 hours,
and preventive actions such as periodic washing of the air distribution system with 5% nitrox solution
(mixture of sodium hydroxide
valves and cylinders for buildup during maintenance, were taken to prevent future explosions,
Air injection was terminated in late 1980 after 50% of the pattern volume was burned and heat
scavenging water injection was initiated. Because of the gravity segregation effects, much of the
heated oil remained in the lower portion of the thicker pay after the termination of air injection. This
heated oil was displaced and recovered by the heat scavenging water.
During the six years of operation (duration of the DOE contract), the project produced 667,609
bbl of oil, compared with 700,000 bbl predicted. It was anticipated that the project would ultimately
produce more than the predicted 700,000 bbl of oil, if the economics permitted it. Whether this was
221
the case is not known, because no project-related production data were released since the expiration
of the DOE contract in 1982. The project was terminated in 1990.
The project paid out and yielded attractive economics
climate in which it
was conducted. During the DOE contract period the actual expenditures were $8.79 million compared
to the initial estimated expenditure of $8.23 million. Additional work beyond the initial scope of work
for the project, including drilling nine additional producing wells and four postburn core wells and
also repairing the damaged compressor, added to the unplanned costs.
Project
Performance
Analysis
Many of the operating problems were predicted and backup provided. Due to the extremely low reservoir pressure (40 psig), it became necessary to operate the producing wells at minimum possible bottomhole pressure to maximize pattern oil recovery. This operating procedure required that a static
liquid level be maintained below the producing formation at all times. This was accomplished by setting subsurface pumps at a total depth and keeping the well continually pumped off. This was also
necessary to minimize sand production. The operator installed a gas collection system to control the
venting of hot produced gases and maintain the bottomhole pressure at the desired value. The extreme
downhole temperature and the velocity of the gas caused condensed water to float inside the casing
just above the high temperature zone. This resulted in pressure buildup and fluid production stoppage.
The problem was solved by periodic cycling of vented gas from zero to maximum flow. This cycling
allowed the condensed water to drop to the bottom of the well, where it was removed by the subsurface pump, reducing bottomhole pressure and also partially cooling the hot section. Hardened and
honed pump barrels were utilized to minimize pump erosion due to sand production. Pressurized
heaters treaters were used to break emulsions and lower bottom sediment and water (BS&W).
buildup in the casingkubing
operation, an explosion
Coke
occurred in the air injection system, destroying the distribution lines and
severely damaging one of the three main compressors. The explosion was attributed to the buildup of
a lubricant film on the inside walls of the air injection lines.
In spite of all the efforts to minimize sanding problems, sand production and the handling of hot
combustion gases remained the major operating problems during project life. Of the eight original
producing
wells, only the one farthest from the injector did not require remedial work. Remedial
cement squeeze jobs were performed in producers at frequent intervals to shut off perforations at
222
which hot combustion gases carrying sand eroded the perforations. Remedial work at wells along the
north line of the project area became so frequent and difficult that these wells were shut off and not
returned to producing status until long after termination of air injection in October 1980.
To better evaluate the performance of the simultaneous air and water combustion process, four
wells were drilled, cored and logged. Results of the program indicated extensive burning outside of
the pattern area. Based on the interpretation of the core and log results, Cities Services arrived at the
following
1.
conclusions:
The geological
while water moved down, causing variations in the situ combustion process ranging from
dry to quenched combustion at various positions.
2.
Several hard lime streaks kept the injected fluids separated behind the leading edge of the
waterflood. Ahead of the water, they did not impede the migration of air to the very top of
the zone.
3.
Residual oil saturation in the burned intervals was zero, indicating a highly efficient process.
4.
Significant reductions in oil saturation down-dip of the injector, in intervals above the hard
lime streak, was an indication of quenched combustion
5.
The simultaneous air and water injection process with partial vertical separation is not well
suited to reservoirs with appreciable structure when applied in a pattern type development.
In summary, the Bodcau in-situ combustion project was both technically and economically
suc-
cessful, paying out quickly, and provided an attractive return on the invested capital. The capture efficiency, however, was low. Though 7290 of the oil in place was displaced, only 42% was captured.
Extensive migration of oil outside the lease line occurred as a result of excessive air injection. The
operator should be commended for having devoted considerable effort to the design and operation of
the project and for thoroughly documenting the project operations. The project performance indicates
that very little was left to chance, and that virtually all difficulties were foreseen and every effort made
to deal with them.
Since the oil recovery was close to the forecasted level and the project costs were close to the
projected amount, it is clear that the operator fulfilled the set objective and demonstrated that a care-
223
success.
General
Observations
On the basis of published information, and the unpublished results of some of the tests, firefloods may be divided into several classifications. The combustion reaction was essentially the same
regardless of the reservoir configuration or the method of application, but the manner in which it was
used could be considerably different. Firefloods can be divided into the following types or classes. In
each of these classes there is at least one example of a technical success and in some there are many
examples.
A.
2)
B.
C.
Specialty Applications
1) Air-water fireflooding wet combustion COFCAW.
224
2)
3)
Heat Wave.
Delaware Childers Sinclti
4)
Reverse combustion.
Kentucky Gulf
Bellamy, Missouri -
5)
Phillips
Tar Sands.
Kyrock Gulf
Athabasca Amoco, Texaco
Mobil Sun
6)
were probably
described earlier a fireflood moves through a sand packed and oil saturated tube or an oil reservoir and
does several things. The oxygen in the air reacts with the fuel that has been deposited on the sand and
creates a temperature ranging from 500-1000F.
driven at this temperature so aIl of the liquids move out except for a small portion of the oil which is
essentially solidified by the high temperature and gas drive and remains behind as fuel. The liquids
and vapors are driven ahead by the gas drive and eventually condense as the cold formation ahead of
the firefront cools the fluids. The air volumes required to support the combustion process are tremendous. For example, at Miga the air-oil ratio is 11,000 cu fthbl. The pressure at the firefront is 2,200
psia and the estimated temperature is 800F; therefore, about 30 reservoir barrels of gas are flowing
through the combustion zone for every barrel of oil that is flowing. The main function of the air is to
furnish oxygen for the combustion reaction but the flue gas that is generated also serves other fimctions. It helps vaporize and move oil and water ahead of the combustion zone and at the producing
wells it provides gas lift for the oil and water. A fireflood therefore most closely resembles a cycling
project in a gas condensate reservoir where dry gas is supplied continuously from an outside source.
225
The volume of gas involved greatly outweighs the oil and water. Literally, for every barrel (300 lb) of
oil produced there is also a barrel of water (350 lb) and in the case of Miga there is also 900 Ib of air
injected and approximately
flue gas are produced for each pound of oil that is produces.
Back to the task of explaining the difference in uses of fireflooding. If one had a circular reservoir with an infinite number of producing wells around the perimeter so as to form a ditch and an
injection well at the center, one could fireflood the reservoir and collect all of the oil that was displaced which would be about 7585% of the oil originally in place. This would be ideal but it is also
non-existent, so we have to take what does exist and do the best we can with it. Complete liquid displacement or type A requires that sufficient producing wells exist to produce all of the oil and water
that is being displaced by the combustion front and all of the flue gas that is being generated. Once the
flue gas has passed through the firefiont it is free to move about the reservoir in the same manner as
free natural gas would move, i.e., gravity segregate and occupy only the top fraction of the reservoir,
move updip and form a secondary gas cap, etc. However, before the air moves through the combustion zone it performs as if the air and combustion zone had a 1:1 mobility ratio. The movement of the
combustion zone is controlled mainly by the fuel that is deposited and the heat that is conducted forward through the r6ck; therefore, it is not affected to any great degree by changes in permeability, liquid saturations, porosity,
impermeable barriers, the shape of the combustion zone will obviously be influenced. The path the
hot fluids take ahead of the combustion zone will also have an influence, but since at field rates most
of the heat is carried forward by conduction through the rock, the heat camied forward by the fluids is
less significant. Inadequate injection rates may also allow gravity segregation to take place between
the air and the fluids being displaced. In reservoirs of moderate thickness the gravity override can
probably be avoided by using adequate air injection rates. In thick reservoirs it may be impossible to
avoid.
226
REFERENCES
1.
2.
Huffman, G.A., J.P. Benton, A.E. E1-Messidi, and K.M. Riley, (1983
nance by In-Situ Combustion, West Heidelberg Unit, Jasper County,
Tech. (October), pp. 1877-1883.
3.
4.
Sarathi, P.S. and D.K. Olsen (1994), DOE Cost Shared In-Situ Combus
ited, Paper ISC4
Conference
on 1
5.
E.
227
228
CHAPTER
7 AIR COMPRESSION
PLANT
Introduction
vide clean, dry, high-pressure air at the required rate to support and sustain combustion, while meeting environmental and safety requirements. A compression plant consists essentially of one or more
compressors with a power source (driver), control system, intake air filter, inter and after coolers, separators, filters, dryers, fuel and lube oil storage tanks, interconnections piping, exhaust emission control equipment, and a distribution system to carry air to injection wells.
A schematic of a typical air compression plant for an in-situ combustion project is shown in Figure 7.1. The air compressor is the heart of the compressed air system and the selection of proper compressor system is critical to a trouble free operation. Compressors vary widely in design or type, each
with a fixed set of operating characteristics. It is the duty of the project engineer to best speci@ the
compressor type to the projeet need. Air compressor is an expensive and rather a complicated piece of
equipment to specify, purchase, operate and maintain. To achieve the expected performance,
care
must be exercised in selecting equipment, laying out and installing it correctly, and then maintaining
the system in good working order.
The intent of this chapter is to provide guidance to the ISC projeet engineer or operator in the
selection, purchase, operation and maintenance of air compressors and related equipment.
229
AIR COMPRESSION
AR ~ ATMCEFHE31C
DQ-MJ.Jsr
FRESURE
*
lNT/wE
FILTE3
EMS90NS
CONTROL
.
{
-AR
COMPRESSOR
>
u++%
T?&
I
I
A::~
L___
___
_)
INTBKXIOLER
FUR
$
n
w
m
0
0
LU9EOIL
i90R=E
WATER
FILTE3
-i
LU=OIL
REOVERY
COMPRESSED AIR
R-
Ull
FIGURE 7.1 Air Compression Equipment for Fireflooding
230
This chapter also includes numerous charts and a step by step procedure for com
Compressor
Types
2.
flow macl
and rotary. C
machines are also subdivided into two basic groups: dynamic and thermal. The pri
compressors are shown in figure 7.2. Not all types of compressors are made in all pI
ranges. The typical operation ranges of various types of compressors are shown in Fig
231
Compressors
Positive
Dkplacement Type
I
I==x-l
Rotary
Single-Stage
Mutti-Stage
internal GasEngine Driven
Seperable
BalancecUOpposed
Thermal
Type
Dynamic
Type
Straight Lobe
Helical Lobe
(Screw)
Slidin Vane
Liqul5Ring
Dzid
Ejectors
P
Multi-Sta e
Fixed Sta?or
Vanes
Variable
Stator Vanes
Sin le-Stage
Mu7
tl-Stage
Horizontally Split
Vertically Split
(Barrelltegral
pl
200
~
\
1
102
\\
; 03
; 05
104
FLOW
RATE,
106
cfm
232
Sin Ie+tage
Mu ti-Stage
86
9-
Oe
DESIGN
POINT
00
N/
84
DESIGN
POINT
82
0
/
0
80
/
0
CENTRIFUGAL
I
78
1.1
---
1.2
1.3
RECIPROCATING
1.4
1.5
1. 7
1.6
COMPRESSION RATIO, C
FIGURE
7.4 Comparison
Compressor Efficiencies
35
1
---
30
---
- RECIPROCATING
9
-/
1
I
&
= 60F, 520R
P: = 14.73
,/
psia
k=l.26
T1
= 60F, 520R
disch
= 850
psia
1.2
1.6
1.4
COMPRESSION
RATlO, C
233
1.8
The principal compressor types used in ISC projects are either reciprocating or centrifugal. In
addition in many projects rotary screw compressors are used as auxiliary units either to boost the discharge pressure of the main unit or to serve as the first stages of a larger reciprocating system. In at
least two projects, reciprocating compressors are used as the upper stages of a centrifugal system to
achieve the desired discharge pressure. Table 7.1 presents the type of compressors employed in some
past and current U.S. in-situ combustion projects. This table indicates that the installation most often
found in the field is the gas engine driven multistage reciprocating compressor.
234
Projectand
Duration
Rateaad
Pressure
Compressor~e
and
Manufacturer
MobilOil
12 Stage(8000
27 MIUScfld at
CompanyMOCO-T
BHP) Cooper
Bessemer Centrifugal Compressor.
First (1961)
installation was
a 350 HP (angle
type) Ingersol
Rand Reciprocating Compressor.
675 Psig
Lease Project, CA
(1961-1992)
2.5 MMScf/d at
600 Psig
Drivers
~
Gas Turbine(8000
HP). Later
replacecl by Synchronous motor.
Natural gas fired
integrated engine.
Clark Brothers 6
Stage Horizontally Opposed
Double Acting
Reciprocating
Compressor.
6.7 MMScf/d at
4125 Psig
Four, 3 Stage
Horizontally
Opposed IngersolRand Reciprocaling Compressors
with air cooled
intercoolers.
Three, 6 Stage,
Ingersol-Rand
Centrifugal Compressers
20 MMScf/d each
at 125 Psig
Each compressor
is rated for 1.3
MMScf7d at 1000
Psig and 230 BHP
300 HP Turbo
charged gas
Engine
235
Sulf Oil Co
[Chevron)
CottonValley
Field,Heidelberg,
MSCombustion
Project
{1971-1992)
Rate and
Pressure
Compressorme
and
Manufacturer
Drivers
Originallydriven
by a 1000 HP
WaukeshaNatural Gas Engine.
Laterreplacedbya
Compressor# 1
3-StageIngersol
RandModel
4 RDS-3 ReciprocatingCompressor withWater
cooled intercooler
3.18 MMScffdat
Psigand742
BHPat800 rpm.
Compressor# 2
3-StageIngersol
RandModel
4 RDS-3
Reciprocating
Compressorwith
Watercooledinter
cooler.
5.74 MMSct7dat
315 Psigand 1170
BHP at 885 rpm
1250 HP General
Electric Induction
Motorrunningat
900 rpm.
Compressor# 3
3-StageIngersol
RandModel
4 HHE-3-VG
Reciprocating
Compressorwith
Watercooledinter
cooler
12 MMScf7d at
280 Psig and2800
3000 HP General
ElectricSynchronousmotorrunningat327 rpm.
315
3 ClarkBros.
Reciprocating
Compressors.One
6-StageUnitand
Two5-Stage
Units.
6.2 MMScf7dtotal
capacityat 5000
psig. 1.2 MMScf/d
at 5000 psig
2.5 MMScfYdeach
at 3000 psig.
Each 5-stageunit
wasequippedat
the downstream
endwitha single
stage Sullairoil
floodedRotary
ScrewCompressor to boostthe
pressurefrom
3000 to 5000 psi.
Boosterscrew
compressorrated
at 2.5 MMScf7d
witha pressure
ratioof 1.67
236
2500 HP natural
gas firedengine
Duration
Ikxrico
~addoPineIsland,
La,HarrellFee
LeaseIn-situcommstionproject
[1980-1983)
hlonterey
Resources
:nowpartof Texico)
MidwaySunset
Fireflood
:1982KochExploration
[Continental
Resources)BufhiloRed River
LJnitAir Injection
Project S. Dakota
:1979-
Rateand
Pressure
Compressor~e
and
Manufacturer
Drivers
First Stage: 50 HP
440 V 3- Phase
InductionMotor.
SecondStage: 300
HP 440 V 3-Phase
InductionMotor
One2-stage
Reeiproeating
Compressor.
1 MMScf7dat 900
psig.
OneSingleStage
IntegratedReciproeatingCompressor
IngersolRand
CENIACII CentrifugalCompressor witha
Reciprocating
BoosterCompressor at theupper
stage.
5 MMScf/dat 250
Psig
1500 HP Synchronousmotor
CentrifugalCompressorfor first
fourstages(CooperW. Superior)
13 MMScf7dat
250 Psig and3600
rpm
3000 HPInduction
Motor
Reciprocating
Compressorfor
last 3 stages(W.
SuperiorModel
13 MMScf/dat
5000 Psig
230 HP Integrated
GasEngine
3000 HP SynchronousMotor.
w 74)
LJ.S.Dept.of
Energy
NavalPetrolwrn
Reserve#3 (NPR3) Tea Pot Dome
FieldFireflood
Casper,Wyoming
5 StageHorizontallyOpposed
DoubleActing
IngersolRand
Reciprocating
Compressorwith
watercooledintercoolers,
3.2 MMScf/dat
80QPsig.
Turbocharged
WaukeshaGas
Engine(2900 HP)
KochExploration
(Continental
Resources)
MedicinePole
HillsUnit
@@~,
Bowman
County,N. Dakota
Air Injection
Project
7-StageW. SuperiorModelMW 68
Reciprocating
Compressorwith
watercooledintercoolers(2 units)
5.2 MMScf7dat
5000 Psig
Turbochargedgas
engine(2650 HP)
237
TMLE
Projectand
Rateand
Pressure
CompressorType
and
Manufacturer
Duration
AlnOCO
ProductionCo.
WestHackbeny,
La Air Injection
Project
(1994 -
TOTALMinatome
Corporation
HorseCreek
FieldAir Injection
Project, Bowman
County,N. Dakota
(1996-
Drivers
WaukeshaModel
5108 GL Lean
Burn gas engine
WaukeshaModel
9390 GL Lean
Burngas engine.
8 ThrOW / 7
Stage Reciprocat-
Turbocharged
integralgas engine
(2650 BHP at 900
rpm)
iWO
ing Compressor
withWatercooled
intercoolers(2
units)
Before presenting the details of the principal compressor types employed in ISC projects, a discussion on the relative merits of these compressors and the reasons for the popularity of reciprocating
compressors in ISC projects are in order.
Relative
Proper compressor
Comparison
of Various
Compressor
Types
selection for ISC service is very difficult and several factors must be taken
into consideration before finalizing the selection. These include peak project air requirements, desired
injection pressure, initial cost, power requirements, operation and maintenance cost etc. For small
pilot or experimental projects, where air requirements are small, skid mounted, packaged reciprocating compressors
are preferred because they are less expensive to purchase and operate. In a large
operation, the choice is not that obvious. For many operations, both reciprocating
and centrifugal
compressors can meet the flow and pressure requirements: Both these compressors have advantages
and disadvantages for a particular application and must be taken into consideration during the selection process.
238
Advantages
and Disadvantages
of a Centrifugal
Compressor
operating and maintenance expenses (shut down required only once every three
efficiency
compression efficiency is the ratio of theoretical work to actual work required to compress a
given quantity of gas.
Minimal operational attention requirement when operated within the design envelope.
Adaptability to high speed, low maintenance cost drivers.
The principal disadvantages of using a centrifugal compressor in an ISC project are:
Because
they are dynamic, the flow is much more sensitive to operating environments,
External maintenance is mandatory; only the manufacturer could trouble shoot and repair
the compressor.
If major mechanical failure of compressor internals (rotor) occurs, due to unstable operation.
it will take longer to repair or overhaul a centrifugal unit (than a reciprocating one) unless a
complete spare rotor is available.
Carrying a spare rotor, though ensure lower downtime, can increase the compressor installed
cost by about 50~0.
Require more skilled manpower to maintain and operate due to relatively narrow range of
stable operating regime.
239
Advantages
and Disadvantages
of a Reciprocating
Compressor
Superior compression efficiency at compression ratios above 1.3. This advantage increases
dramatically as the compression ratio increases (Figure 7.4.)
(Figure 7.5).
Capacity (output volume) can be easily be varied to meet the demand by using appropriate
capacity controls.
Can deliver air at any pressure up to the limit of its mechanical strength and drive capability.
Require less skilled manpower to operate, since the machine is less sensitive to variation in
the operating environment.
Major mechanical problems probably will not arise. If major problems do arise, it is less
expensive to repair.
Lower capacity.
Foundation requirements are much more massive than that for an equivalent centrifugal
machine.
Has more and heavier moving parts and, therefore, lower mechanical efficiencies
trifigal compressors.
240
than cen-
,.
Require more careful design of the discharge piping system (piping between compressor
and intercooler and at the compressor outlet) to avoid vibration and pulsation. Centrifugal
do not have this problem.
Require higher maintenance (shutdown and overhaul required once every 9 months) due to
large number of moving parts and corresponding wear.
Advantages
and Disadvantages
has following
of Rotary
Screw
Compressors
pressers:
Can develop 7.3 times the head of a centrifugal compressor operating at the same speed.
Operation and maintenance costs compare favorably with the centrifugal compressors which
are about one third the cost of maintaining reciprocating machines.
Lighter in weight than the reciprocating machines and does not exhibit the shaking forces of
the reciprocating compressor, thus making the foundation requirements less stringent.
Flooded screw compressors can realize pressure ratios (ratio of discharge pressure to intake
pressure) as high as 21 to 1, making it an ideal machine to boost the discharge pressure of
the centrifugal.
Disadvantages:
Q Overall efficiency (70%) is less than the reciprocating or centrifugal machine and decreases
with increase in compression ratios.
Has much higher internal fluid leakage losses than the centrifugal or reciprocating machines
and this impedes its attractiveness as a mainstream compressor. Higher fluid leakage losses
also prevent the machine from realizing higher compression ratios.
Typical flooded single stage screw compressors are restricted to 4 tol compression ratio.
241
loss.
Reasons
of Reciprocators
in ISC Operation
Despite its lower capital, operation, and maintenance costs compared to the reciprocators, centrifugal compressors
have not found widespread acceptance among most ISC operators. Following
In-situ combustion operations generally involves significant amount of compression and the
service require considerable flexibility in operating pressure and throughput. T~ically
at
the start of an ISC operation, the injection pressures are higher than the projects average
injection pressure and the injection rates are well below the average. Once ignited and the
combustion
front began to move away from the injector, the injection pressure drops and
the injection rate increases. The reciprocating compressors best provide this operating condition flexibility. Reciprocating compressors are the only compressor types whose capacity
and pressure may be varied in accordance with the requirements, without sacrificing the
efficiency.
2.
Most ISC projects implemented in the U.S. in the 1950s and 1960s were either small pilot
or experimental projects. The volume of air required in these projects were relatively small
(often much less than and outside the practical limits of most centrifugal compressors).
Though specially designed centrifugal compressors can handle such low volumes, they are
not cost effective compared to reciprocators of similar throughput. This preclude the use of
centrifugal compressor in such application. The centrifugal compressor has been applied in
the approxima~e range of 1,000 scfhnin to 150,000 scfhn.in (14.4 MM scf/d to 216 MM scf/
d). Below 6MM scf/d flow rate the operation of centrifugal compressors becomes unstable,
leading to mechanical failure of the machine.
3.
Availability of natural gas in most field location and the suitability of natural gas fueled gas
engines as drivers for the reciprocators (centrifugal are high speed machines and require
either gas turbines on electrical motors as drivers) also can be cited as a reason for the popularity of reciprocating compressors in ISC services.
4.
Other reasons for the widespread use of reciprocating compressors in ISC projects include:
familiarity and ease of repair on site, less skilled manpower requirements to operate and
maintain, less costly down time due to ready availability of spares, low power requirements
and superior compression efficiency compared to centrifugal.
242
5.
performance inthepast
operators tend to stick with the tradition and unwilling to consider other compressor types
for ISC operations.
Basic
Terms
and Definitions
of Compressor
Terminology
An engineer or operator in charge of selecting or specifying compressors for an in-situ combustion project must have a good understanding of various compressor related terms. This is needed to
avoid confusion and to prevent miscommunication
with vendor.
In this section an attempt is made to explain selected terms specific to compressor related calculations. These terms apply to all compressors regardless of type.
Suction Pressure: This term refer to the air (gas) pressure existing on the inlet or suction,
side of the compressor. This is expressed in pound force per square inch (lb#iN*) gauge
(psig) or lb#N2 absolute (psia). The exact point of measurement of suction pressure is
important in the final sizing and selection of the compressor.
Discharge Pressure: The air (gas) pressure existing at the discharge or outlet end of the compressor. This is also expressed in units of psig on psia. As with suction pressure the exact
point of measurement is important in the final sizing and selection of the compressor.
Compresswn
the absolute suction pressure. It is usually applied to a single stage of compression, but may
be applied to a complete multi stage, compressor as well.
Capacity: It is the quantity (volume) of air (gas) actually delivered when operating between
specified suction and discharge pressure. In the compressor literature, it is expressed in various units. These include: cubic feet per minute (cfm), inlet cubic feet per minute (icfm),
actual cubic feet per minute (acfm). Standard cubic feet per minute (scfm), millions standard cubic feet per day (MMscf/d),
All these units expresses the volume delivered at certain conditions of pressure, temperature gas
composition
Cubicfeetper
minute (cfm): It is the volume rate per minute measured at the specified or known
inlet condition (usually ambient) of pressure and temperature. The pressure usually the barometric
pressure expressed in psig and the temperature is the prevailing air temperature. It is best to specify
the pressure and temperature at which the volume is to be expressed.
243
Actual cubic feet per minute (acfm): This term originated in the chemical process industry and
expresses the volume
at process condition.
Like cfm,
expressed at the prevailing compressor room pressure and temperature conditions. Often in compressor literature the term cfm and acfrn are used interchangeably.
Inlet cubic feet per minute (icfm): Same as cfm and refer to the volume at the prevailing suction conditions.
scfm
means cfm
at standard condition.
However,
standard vary and some care is necessary. In the United States, the usual standard is 14.696 psia and
6F. In Europe, the usual standard is latm and OC. The U.S. natural gas industry adopted 14.4 psia
and prevailing inlet temperature (usually 65F or 70F) as its standard. Hence the engineer must specify the standard conditions to the compressor vendor.
(standard) inlet condition. It is not the same for all people. Like scfm,
the standard adopted. In the natural gas pipeline industry where this term is most frequently used, it is
refer to the volume at 14.4 psia and inlet temperature.
adopted by ISC operators to express the compressor capacity. Again, the engineer is advised to speci& the standard conditions, so as to leave no doubt in the compressor vendors mind.
Horsepower
refer to the time rate of work required to compress and deliver a given quantity of
gas (air) in accordance with a specified process. The total required horsepower is a function of capacity, compression ratio and compressor efficiency. Like volume, the compressor literature refers to several different horsepower and an explanation is in order.
Theoretical Horsepower
Gas Horsepoweic
(GHP) is the actual work required to compress and deliver a given quantity of
gas. This includes all losses (thermodynamic, leakage and fluid friction) except mechanical losses.
Indicated Horsepower
244
Break Horsepower
and deliver a given quantity of gas. This includes gas horsepower and all mechanical losses. It is also
sometime referred to as the shaft horsepower. It is also the power developed and delivered at the compressor shaft by the prime mover.
Peak Horsepower
ating at a (1) constant discharge pressure with variable suction (intake) pressure or (2) constant suction pressure with variable discharge pressure.
Basic
Relationships
This section discusses briefly some basic thermodynamics and mathematical relationships necessary to determine the size of a compressor and the power required to compress a given volume of
air. To keep the discussion simple, no attempt is made to derive various gas compression equations.
Only the final expression is presented. For a detailed discussion on the thermodynamic fundamentals
and the equations derivation, the reader is referred to other publications.
Principles
Pressure, temperature, ad
of Compression
volume are the three variables that influence the status of a gas. A
Pv ~= Constant
/
(7.1)
=R
Where
R is the universal gas constant and its value depends upon the units chosen for P, V and
T .
When
P is in lbf /iNz abs
V is in cu. ft./lb. mole
and T is in R (F+ 460), the value of R is 10.729
245
Similarly R has a value of 1545 when P is in lbs/sq.fi., V is in cu. ft./ lb. mole and T is in R.
All gases deviate from ideal gas laws to some extent. It is therefore, necessary that this deviation
from ideality be accounted for in compressor calculations to prevent compressor and driver sizes from
being greatly in error. This deviation from ideality is accounted for by multiplying the right hand side
of Equation (7.1) by the gas deviation factor (or compressibility factor) Z.
Thus the real gas law that applies to all gases can be written as
PV = ZRT
(7.2)
The value of Z depends upon the system pressure, temperature and type of gas.
Another important gas property is the specific heat ratio k. This ratio is given by
k=cp
C
/
C
v
C~986
(7.3)
Where
Cv= specific heat at constant volume, Btu/ lb~,- mole - F
CP = specific heat at constant pressure, Btu / lb~ - mole - F
In compressor calculation, the compressor capacity is a volumetric value based on flowing conditions of pressure, temperature, and relative humidity (if moisture is present). The flow &its are
given as inlet cubic feet per minute (icfm) or actual cubic feet per minute (acfm). In oil field, the flow
*
value is stated as st&ia.rd cubic feet per minute (scfm). Unfortunately, the definition of standard condition is not universal. The ASME standard uses 68F and 14.7 Psia at 36% relative humidity. The
API standard is 60F and 14.7 Psia. The natural gas industry uses 14.4 Psia and 60F as its standard.
To avoid conilsion,
it is always advisable to check the reference condition whenever the scf abbrevia-
tion is encountered. In the S1 unit, there is only one set of reference conditions, 101.325 kpa and 15C
as established by the international standard 1S0 5024. The recalculation of volumes from one set of
standard conditions to another is rather straight forward and can be accomplished
(7.2) as follows.
246
using Equation
Given the volume occupied by dry air at 60F temperature and 14.7 Psia pressure as 100 cubic
feet. What is the corresponding
volume at (a) 14.7 Psia and 68F and (b) 14.4 Psia and 60F. Gas
V*= ?
%4=
2%22
Hence
v=
<V,
T2Z,
TIZl x Pz
V2=
14.7 X 100.0
520 X LO x
528x 1.0
14.7
= 10154 CubicFeet
Similarly for the conditions (b), the corresponding volume is 102.1 cubic feet.
Compression
There are three compression
Cycles
cycles or thermodynamic
Isothermal compression.
2.
Adiabatic compression.
3.
Polytropic compression. Only isothermal and adiabatic compression are pure basic thermodynamic cycles.
The polytropic
process is a modification
of adiabatic compression
involving an efficiency
to
more nearly represent actual condition and as such is not a true basic cycle. Although, true isothermal
and adiabatic compressions
nevertheless,
7.6. When a gas is compressed heat is generated. The manner in which this heat of compression
dealt with distinguishes the compression
is
occurs
when the temperature of the gas being compressed is held constant. This requires continuous remowd
of the heat of compression. This compression is described by the formula.
(7.4)
Adiabatic compression: also known as the isentropic (constant entropy) compression is obtained
when no heat is added to, or removed from the gas during compression. This process is reversible
when no friction exists. For an ideal gas the compression follows the formula
Where k is the ratio of specific heat= CP / Cv
and
(7.5)
(7.6)
(7.7)
=(RC~
Where Rc = compression ratio = ratio of discharge pressure to inlet pressure and the subscripts
1 and 2 in Equation (7.7) denote inlet and discharge conditions.
248
<vlr = P2V;
(7.8)
And
(7.9)
approached with most reciprocators and used as the basis for these compressor design.
Polytropic compression cycle is used as the basis for centrifugal compressor design and follow
the relation
PW = constant
(7.10)
When Equation (7. 10) is expressed between the initial (1) and final (2) conditions we have
(7.11)
Where n = polytropic exponent and
(7.12)
n*lorn*k
n1
T2
=
TI
P2
()~
(7.13)
lower or higher
than the adiabatic exponent k. In reciprocating and internally cooled centrifugal compressors, n is
usually less than k. In un-cooled centrifuge, n is usually higher than k due to internal gas friction.
Thermodynamically,
ible. For real gases, below a compression ratio of 2, the polytropic exponent n for the gas can be calculated from the following relation
nl
n
k-l
1
knP
()
(7.14)
tion) and discharge pressure and temperature are known. The following formula maybe used.
n-1
1
I
,
I
(7.15)
I
pd (T)
1
[)F
na = -,
4-
(7.16)
,k_l\
I
The adiabatic compression efficiency q, is defined in terms of the polytropic efficiency qP by the
following
In the above equations, the subscript d and s refer to the discharge and suction conditions
respectively.
Polytropic efficiency q~ can be obtained from (7.14) by rewriting the equation as
(7.17)
Theoretical
Horsepower
.
Theoretical work (horsepower) is the heart of the compressor design and the engineer must have
some basic understanding
250
Adiabatic
Compression
requirement is given by
HPT =
OOYQ[w-l(azJ
(7.18)
Where:
Q= gas flow rate in standard cubic feet per minute (14.7 Psia and 60F)
P,, P~ = absolute pressure at suction and discharge, respectively, Psia
m=k-1/k
k= C~Cv (molal specific heat ratio)
units of Cp and Cv are in Btu/lb-mole F
T,= absolute temperature at suction, R
Z~= (Z, + Z~) / 2 = mean compressibility factor
Z~= compressibility factor at suction
Z~= compressibility factor at discharge
Although Equation (7. 18) is derived based on ideal gas with a constant specific heat, it is
equally valid for real gases provided the specific heat variation is not large. Several efficiencies are
defined in order to apply Equation (7.18) to real world compressors. These efficiencies are defined as
follows:
Ta = adiabatic efficiency:
It is defined theoretical isentropic horsepower (Eqn 7.18) divided by the actual or gas horse
power (GHP) delivered to the gas
(7.19)
=HPT/GHP
251
(7.20)
developed
Tao ~T
13HP
*a*
(7.21)
*m
:Q
[W-l(az%l
(7.22)
a= (%)(W$FI(ZJ
(7.23)
Where
M = gas molecular weight
R = gas constant=
252
Since work is force times distance or in this case, weight times head, the compressor horsepower (GHP on BHP) is related to the adiabatic head by the following:
GHP =
BHP=
w H,
(7.24)
33,000?7=
wHa
33,000 qao
where
w = mass flow rate of gas in lb~ / rnin
For a single stage of compression,
(7.25)
neglecting any changes in potential or kinetic energy, the
temperature change from inlet to outlet, can be obtained from the first law of thermodynamics (energy
balance) as
At=t~-t,
(7.26)
Q C,
Where
t,=
suction temperature, F
td = discharge temperature, F
Polytropic
A polytropic
compression
Compression
Equation (7. 10) is a constant that depends on gas properties, amount of cooling supplied to the compressor and energy losses due to friction and irreversibility.
253
(Equation 7.18) except that the isentropic exponent k is replaced by the polytropic exponent n and
the a~labatic efficiency q~ by polytropic efficiency VP.
Thus the power equations for a one stage polytropic compression are
YQ[(k:er-l(TYJf).)(+)
(7.27)
Where
m=
nl
()
n
p=(RTRion)(i)[rc;erl[TYiY)(.)
(7.28)
The outlet and inlet temperature for polytropic compression are related by the expression given
by Equation (7.13) where
Tz = discharge temperature, R
TI = suction temperature, R
As in the case of adiabatic compression, the polytropic formulas must be applied separately to
each stage in a multistage system.
For a given speed and inlet capacity, the centrifugal (or axial) compressor develop a constant
polytropic head (in ft.-lbf / lbm) regardless of the nature of the gas, its inlet temperature or whether it
254
is cooled or not during compression. The polytropic head, theoretically remains the same, regardless
of whether the gas compressed is C02 (gravity = 1.528) or air (gravity = 1.0) or natural gas.
Although the polytropic head developed by a dynamic machine (such as the centrifugal compressor) at a given speed and capacity remains constant regardless of the gas, the pressure rise generated is not. The pressure rise vary approximately
centrhlgal
compressor operating at a given speed will compress a given inlet volume of COZ to a
greater pressure, than the same inlet volume of air. This fact must be kept in mind while selecting a
compressor for in-situ combustion service, because oil field compressor vendors often rate the compressor for natural gas, than for air. In such cases, the compressor discharge pressure, when used for
air compression, is likely to be lower than that quoted by the vendor.
Isothermal
Compression
The theoretical horsepower for a reversible isothermal compression is given by the expression
HP~
QTZm In
= 8110
P~
(7.29)
()~
Where
P~
is the natural logrithrn of compression ratio
rn~
() s
Equation 7.29 assumes that the heat of compression is fully removed by cooling.
In practice,
this is not feasible and the actual power developed is less than the theoretical power predicted by
Equation 7.29.
The performance of a real world compressor can be evaluated by the following:
(7.30)
(7.31)
Where
qt = isothermal efficiency
255
Reciprocating
Compressor
Introduction
The reciprocating compressor is probably the best known and the most widely used compressor
type in the oil patch. More than 80% of all the compressor systems in use in the oil field are reciprocators. Its widespread popularity among oil field operators can be attributed its superior compression
efficiency and greater flexibility in capacity and pressure range at a lower power cost. Reciprocating
compressors had been employed in the oil field since the early 1900 when the practice of reservoir
pressure maintenance through well head gas re-injection began (J.O. Lewis, 1917).
Oil field reciprocating compressors vary in size from small 5 horsepower units to larger power
units for several thousand horsepower. Majority of oil field reciprocating compressors are moderate
horsepower, heavy duty packaged units, and direct connected to a gas engine driver. These units are
usually designed to meet the minimum requirements of the API specification 11P (SPEC 11P) developed to meet the oil and gas fields need for a compact, semi-portable, well balanced, skid mounted
compressor.
The packaged reciprocating compressors are adequate for light to medium duty services such
the compression of well head gas for discharge into trunk lines, gas lift operations, reservoir pressure
maintenance activities and pilot gas injection EOR processes such as the in-situ combustion and flue
gas flood. In large EOR projects such as the commercial scale in-situ combustion, COZ and hydrocarbon floods, large multistage reciprocating compressors with interstage coolers are employed.
Irrespective of the project size an operator or the engineer in charge of specifying and purchasing compressors
for an in-situ combustion project must have a good understanding of the specific
characteristics for different compressor types and the many factors that must be considered in the
compressor selection process. These issues are not only essential to the proper formulation of the
256
compressor specification sheets for vendor quotation, but also to the selection of right type of compressor for the intended service.
In the following pages various facets of reciprocating compressors are reviewed. Detail description of the compressor components and their design features are beyond the scope of this work and
not presented. A brief overview of major components of the reciprocating compressor is, however,
presented to aid in the development of compressor specification sheets.
Description
As mentioned earlier, a reciprocating compressor is a positive displacement machine that utilize
the motion of a piston in a cylinder to compress the gas. The basic construction of a reciprocating
compressor is depicted in Figure 7.7. It consists of a frame (housing) to which are attached the necessary compressor
(where
SUCTION VALVE
\
CROSSHEAD
PISTON
ROD
PISTON
I
CYLINDER
HEAD
II
~
CONNECTING
ROD
1,
;\J!
1111II c1
-b
\_
Wi=u
WRIST PIN
li$1
1-
CRANKSHAFT
~
DISCHARGE
VALVE
of Reciprocating
Compressor
heads, suction and discharge valves, and the parts necessmy to convert rotary motion of the driver to
257
reciprocating motion of the compressor piston (connecting rod, crosshead, crosshead pins, and piston
rod)
A reciprocating
Almost all of the compressors employed in the oil field are double-acting.
gas (air) is compressed alternatively on each end of the piston. Because of the cylinder area occupied
by the piston rod, the available area for compression on the crank end (piston rod side) is less than on
the head end of the piston.
The reciprocating compressor uses automatic spring loaded valves to admit the gas into and out
of the cylinder. These valves open only when the proper differential pressure exists across the valve.
The inlet valve opens when the pressure inside the cylinder is slightly below the intake pressure. Discharge valve open when the pressure in the cylinder is slightly above the discharge pressure. The discharge valve closes once the pressure in the cylinder drops below the discharge line pressure, thus
preventing flow reversal.
All reciprocating compressors have a space between the end of the piston and the cylinder head.
The volume formed by this space plus the volume formed by the valve cavities is called the fixed
clearance. This clearance, plus any variable clearance (addition or subtraction of clearance volume
through the use of clearance pockets), constitutes the total clemmce
258
volume.
P2=reciever pressure
Pl=inlet pressure
Clearance
Volume
Compression Stages:
1=Start
2=compression
m
P2 .
3
;
P1
14
. inlet compressor
Stroke
valve
piston ~
discharge compressor
valve
I
II
II
II
II
II
II
4
II
II
II
II
II
II
II
fl
Compressor Cycle
Figure 7.8 shows the ideal basic cycle of a reciprocating compressor on a pressure volume (PV)
diagram. The events during this cycle are as follows.
RECEIVER PRESSURE
P2 w
~w
z~
$Q
P,
INL13 PRESSURE
i ~1
STROKE
------d
Figure 7.8A, shows the basic element with the cylinder Ml of atmospheric air. On the theoretical PV diagram, point 1 is the start of compression. Both valves are closed.
260
-..
DISCHARGE
Figure 7.8B shows the compression stroke. Here the piston moved to the left, thus reducing
the original volume of air with an accompanying rise in pressure. With the suction and discharge valves closed, the piston moves from point 1 to point 2 and the air is compressed
isentropically (no heat transferred in or out of the air, and no frictional losses) until the pressure in the cylinder reaches the discharge pressure.
LSTRO.J
DISCHARGE
Figure 7.8C shows the discharge stroke. At a point just beyond point 2, the discharge valve
opens and remains open until the piston reaches the end of the stroke at point 3. The compressed air flows out through the discharge valve into the discharge line.
262
P*
. RECEIVER
PRESSURE.
y
Z3
~~
$9
-1
P,
INLH PRESSURE
1
114
Dl@4R
n
INLET
EXPANSION
Figure 7.8D shows the expansion stroke.When the piston reaches point 3, the discharge
valve closes; leaving the clearance space filled with air at discharge pressure. During the
expansion stroke, both the inlet and discharge valves remain closed and the trapped air in the
clearance space expands isentropically from discharge pressure at point 3 to point 4, at
which point the pressure drops below the inlet pressure. During this cycle the piston moves
to the right.
263
P2 u.!
P,
d. RECEIVERPRESSU
2
3
$$
d
INLH PRESSU~E
F
INTAKE
Figure 7.8E shows the intake or suction stroke. At point 4, the inlet valve opens and the air
flows into the cylinder as the piston moves from point 4 to point 1. At point 1 on the PV dia~~
the iflet v~ve closes ~d the cycle repeats on the next revolution of the crank.
Reciprocating
Compressor
Performance
To determine the quantity of air that a specific compressor can compress and discharge, the dis-
placement,
Displacement
expressed in cubic feet per minute, or ft. %n.in. It is a function of the area of the piston face, the length
of the stroke, and the number of strokes per unit time. Most compressor manufacturers publish data
listing displacement
at full compressor speed for each size (bore, stroke, rod size) cylinder that is
manufactured.
264
% Clearance
Clearance
P2
Clearance Volume
Stroke
(loo)
+-
PI
Actual Capacity
Clearance volume is the volume remaining in the compressor cylinder at the end of a discharge
stroke (Figure 7.9). The clearance volume is usually set by the compressor manufacturer and is set to
match the specified compressor cylinder capacity.
Compressor cylinder capacity is the intake volume of air expressed in cubic feet per unit time at
suction conditions of temperature and pressure. This volume is calculated as the piston displacement
(expressed in ft. %nin.) multiplied by the volumetric efficiency.
Volumetric ejj!iciency in a reciprocating compressor is the ratio of actual cylinder capacity to piston displacement,
compression ratio, and the ratio of specific heat k. The theoretical volumetric efficiency maybe cal.
culated as follows:
E, = 100 c
(7.32)
(Rc)x 1
[
Where:
E,= volumetric efficiency expressed as a percentage
C = cylinder clearance expressed as a percentage of cylinder volume
RC= the ratio of compression
discharge pressure,
psia
suction pressure,
!1
psia
efficiency
efficiency
is
always lower than that calculated from Equation (7.32) due to the neglect of factors such as the inlet
gas preheating, pressure drop across the valves, internal leakage, gas friction, etc. A more practical
formula used by compressor manufacturers to calculate actual volumetric efficiency is given by:
=96-L-C(R)
&..-,]
(7.33)
Where:
Z,, Z~ = gas compressibility factors at suction and discharge respectively.
= C an arbitrary adjustment factor to account for losses due to inlet preheating, pressure drop
across the valves etc.
266
100
90
30
70
60
50
40
30
20
OMPRE.SS(
LEARANa
= 1.36k
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.s
267
Discharge
Temperature
Gas temperature increase as gas is compressed. Since air support combustion, the oil vapors
present in the cylinder can ignite and explode if the cylinder temperature is too high. Further, at elevated temperature the lubricating oil may breakdown and prone to carbonization. Also the life of the
cylinder packers is shortened by the dual requirement to seal both high pressure and high temperature.
Hence the cylinder discharge temperature of an air compressor must be limited to a safe value.
The temperature of the gas discharged from the cylinder is calculated from Equation (7.7) as:
T~ =T,
(7.34)
(RC ~
Where:
T~ = absolute discharge temperature, R or K
T,
268
TheoreticalDishcageTemperatures
Single State Compression
Read rto kto~toti
360 1 I
au
40
r=6.O
60.
1.1
1.2
1.3
1.4
1.5
~k4)k
1.6
Figure 7.11 is a chart that can be used to solve Equation 7.34. The discharge temperature determined from either Equation 7.34 or from figure 7.11 is the theoretical value. Since the Equation 7.34
neglect the heat from friction, irreversibility effects etc., the discharge temperature obtained from the
chart or equation is always lower than the measured value. Discharge temperature is often controlled
by circulating water through cylinder jacket. For a water cooled reciprocating compressor operating at
a volumetric efficiency
greater than 20%, the value calculated from Equation 7.34 is close to the
actual value.
To reduce carbonization of the lubricating oil and the danger of fires when compressing air, it is
desirable to maintain discharge temperatures below 250F.
Multi
Staging
cylinder is limited by the desirable cylinder discharge temperature. Further, a high compression
ratio
will mean a low volumetric efficiency and require a larger cylinder to produce the same capacity. For
this reason multi staging is used in reciprocating compressors.
If NO is the overall compression ratio then for n stages the compression ratio per stage is given
by:
(7.35)
Where &is
For example:
= z R,O
The per stage compression ratio obtained from Equation (7.35) is a theoretical value and does
not account for the pressure losses between stages. However, it is a useful first approximation for estimating the required number of stages. The actual interstage pressures are defined by the compressor
manufacturer, who will take into account the intercooler and aftercooler pressure drops. Many com-
270
pressor vendors recommend that the per stage compression ratio be kept 4 or less when compressing
air.
Compressor
Horsepower
Estimation
The total horsepower required to compress the air from the inlet pressure to the desired final discharge pressure determines the compressor size and drive requirement. This is needed to prepare the
compressor inquiry data sheet. Gas processors Association recommend the use of Equation (7.36) to
obtain a quick and reasonable estimate of compressor horsepower.
BHP = (20) x (ratio per stage) x (No. of stages) x (MMcfd) x (F)
(7.36)
Where:
MMcfd = Compressor capacity referred to 14.4 psia and suction temperature
F = An allowance for interstage pressure drop
Use 1.0 for single stage compression
Use 1.08 for two-stage compression
Use 1.10 for three-stage compression
Curves are available which permit easy estimation of approximate compression
horsepower
requirements. Figure 7.12 is typical of these curves. The horsepower required is determined by entering the x-axis at the total compression ratio and going vertically to the appropriate curve, and then
horizontally to the y-axis to determine the break horsepower requirement per MMcfd.
Horsepower
obtained using Equation (7.36) or Figure 7.12 are based on average values and
more accurate values are needed for accurate sizing of the compressor. Detailed compressor horsepower calculations can be made using the BHP/ MMcfd curves (Figures 7.13 through 7.16). These
curves published by compressor manufacturers provide net horsepower and include mechanical efficiency and other losses. The proper use of these charts should provide the user with reasonably correct horsepower requirements for inquiry.
271
1.7
72
/
/
70
68
66
64
}
62
//
//
1.2
,
/
, 1.1
60
r
UI
58
//
/.
56
54
52
50
48
46
44
42
40
38
36
CURVE A
34
32
30
28
26
/y
24
/
22
20
2.62.7
2.8 2.93.0
18
16
1.41.5
1.6 1.71.8
1.92.0
COMPRESSION RATIO
FIGURE 7.12 Horsepower Curves for Reciprocating Compressor for Different K
(Compression Ratio 23) (Loomis, 1980)
272
110
1.7
108
106
1.6
104
102
1.5
100
98
1.4
96
94
1.3
92
90
1.2
88
86
84
1.1
82
80
78
76
74
72
70
68
66
64
62
60
58
56
%!.9 3.03.1
3,23.33
.43.53.63.73.8
3.94.04.14
.24.3
4.44.5
COMPRESSIONRATIO
273
ul
b)
o-l
to
0)
1.14
Ii I
1.13 \ \l
1.6
1.7
3.2
3,3
3.4
\ \
\
1.11 \.
1.12
1.10
1.09
1.08
1.07
1.06
1.05
1.04
1.03
1.02
1.01
1.00
1.4
1.5
1.81.9
2.02,1
2.22.32
.42.5
COMPRESSION
2.62.7
2.82.9
3.0
3.1
RATIO
FIGURE 7.15 Correction Factor Curves for Low Intake Pressure (Loomis, 1980)
1075
10000
8000
.5
6000
4000
.5
3000
2000
n
I
m
1* 1000
&
800
600
400
300
200
100
20
3456810
30
40
60 80100
Estimation
Curves
Note that Q in these charts are in MMcfd at prevailing suction temperature and 14.4 psia rather
than at 14.7 psia and 60F (scf) the standard used in oil field.
The above horsepower charts provide the power requirement per ,stage. When using the horsepower charts, each stage of compression must be calculated separately and the totals added together
to obtain the total brake horsepower requirement. Volume to be handled at each stage must be corrected to the actual temperature and moisture content at the inlet to that stage.
When multi stages are used, allowance must be made for interstage pressure drop. Inter-stage
pressures may be obtained as follows:
1.
2.
This procedure should give the absolute discharge pressure of a stage and the calculated intake
pressure to the next stage. The calculated upper stage intake pressure is then corrected for intercooler
pressure drop by reducing the pressure by 35 psi.
The per stage horsepower for compression
m=
(:::.)(H)(2)(Z.V,(MT)
276
(7.37)
MMScf/d = Millions of standard cubic feet of air (rate per day) measured the standard condition of 14.7 psia pressure and 60F (520F).
avg=
r+:),
The horsepower requirement for multistage compression is decreased by intercooling between
stages. If intercooling is not employed, the heat of compression in the first stage will increase the volume to be compressed in the second stage with the resultant increase in horsepower requirement. The
horsepower required by the second stage is in direct proportion to the absolute inlet temperature of
the first stage. Hence, the second stage horsepower requirement will be equal to the first stage horsepower multiplied by the ratio of temperature increase across the first stage. The temperature ratio may
be expressed by (see Equation 7.34):
k
c=
(7.38)
script 1 refers to the first stage and subscript 2 to the second stage.
(7.39)
If intercooling is used, the total horsepower required is roughly equal to the horsepower requirement of the first stage multiplied by the number of stages. This assume that the inlet temperature and
compression ratio per stage is the same for all stages.
Reciprocating
Air Compressor
Both packaged and process type reciprocating compressors are used in ISC projects.
277
Packaged
Compressors
Packaged type compressors probably account for the majority of installed units in the past U.S.
ISC projects. The vast majority of past U.S. ISC projects were small pilots, implemented by smaller
operators with limited technical man power resources. A packaged system is ideally suited for such
situation, because it offers a completely
site. This relieve the operator or his technical staff from the need for the sizing and specification of
compressor and its auxiliaries.
A packaged compressor maybe
to provide a controlled compressed
included in a package, API Spec. 11P recommend that such a system be include compressor, driver
(engine or motor), compressor cooling system, exhaust silencer, emission control equipment, inlet air
filter, scrubbers, control panel, piping, pulsation bottle, etc., all mounted and aligned on a skid as an
operable unit with a minimum of field construction. The skid must be designed to withstand the loading and the moves from one job site to the next. The ideal and most desirable package for the operator
is one in which only the required utilities of power and cooling water are brought to the package, with
the end product being clean dry air.
Packaged type compressor for oil field service are generally offered in sizes ranging from a minimum of approximately 3007,000 BHP. Discharge pressure range from 125 psig to up to about several thousands psig. These systems are ideal for projects where the injection volume and pressure are
moderate.
Process
Process compressor
Compressors
to the packaged
driver, power transmission equipment and basic control system. Procurement of all the other necessary components
such as the lubricating system, instrumentation, inter and after coolers, pulsation
suppression system, piping, emission control equipment etc. is the responsibility of the purchaser.
The compressor vendor is responsible for the satisfactory performance of the secondary equipment
only if he supplies them. Otherwise, the compressor vendor is responsible only for the proper performance of compressor train which includes compressor, driver, power transmission components
and
the associated equipment. In the event the auxiliary equipment are purchased from other sources, the
purchaser should coordinate these purchases with the compressor vendor to ensure smooth integration
278
of components and for realizing low maintenance, trouble free operation under the specified service
condition.
Process type air compressor is a heavy duty low to moderate speed (300-600
rpm) machine
designed for a minimum service life of 20 years and at the minimum 3 yeas of uninterrupted operation. Process type compressors are best suited for use in large scale commercial air injection projects
whose volume throughput and injection pressures are high. All the commercial air injection projects
in the U.S. utilize process type compressor in their operation.
Reciprocating
Compression
Selection
2.
Decide whether the compressor is to have one, two, or more stages of compression. The
ratio of compression permissible in one stage is usually limited by the discharge temperature. A sufficient number of stages, with cooling between them, must be used to prevent the
discharge temperature from exceeding the practical limit. As a first approximation, the temperature of the air discharged from a cylinder (stage) can be estimated from the Equation
7.34. When multistage operation is involved, equal ratios of compression
commonly
used. For three-stage compression, the compression ration per stage is approxi-
Establish the required capacity in standard cubic feetiday and convert it to the actual volumetric quantity of air at the inlet condition of the respective stages on a per minute basis
(cubic feet per minute, cfm). The desired capacity rate must be explicit and the conditions
of pressure and temperature at which it is measured must be given.
4.
Establish the horsepower requirements. For the purpose generalized charts provided by the
compressor manufacturer (Figures 7.12 through 7.16) can be used. The total required compressor horsepower (sum of the horsepower per stage) determines the required frame size.
A frame can be considered as a platform to which the necessary compressor cylinders, driv-
279
interconnecting
piping,
valves,
etc., are
mounted.
5.
Determine the approximate compressor cylinder sizes. The required capacity of each has
already been established (Step 3). From appropriate charts or formulas, the volumetric efficiency of each stage of compression must be established to calculate cylinder size.
6.
Now check the frame loadings to be sure that all are within the design limits established for
the selected frame. If all are too high, a heavier frame must be used.
7.
Assuming the frame requirements have been met, the selection so far as stroke, speed, number of stages and sizes of cylinders is now complete.
requirements more accurately. Table 7.2 depicts a typical reciprocating compressor inquiry
form recommended by compressed air institute. It maybe used as a guide to the information
that should be submitted with an inquiry about a new reciprocating air compressor.
280
Yes
No
tion).
( ) Description of Application:
( ) Duty Cycle
( ) Gas Handled
~ry
()
minutes
Contaminated with
Clean
Wet
()
()
Barometer
()
Intake pressure
psia or Altitude
psig
Intake Temp.
()
Disch. Press
()
Capacity required
..
psig.
F Rel. Humidity
psig to
psig
psig to
psig at
()
()
()
()
Type of Drive
()
Electrical Conditions
()
TWe of Mounting
Cfm to
intake pressure
psig Cut-out
psig Cut-out
psig
F dry or sat.
cfm
discharge pressure
psig
psig
Steam Conditions
281
hotorcold
mbient;
space.
( ) Cooling Water: Temp
Special accessories
Filter
Clean
Dirty
Salt Water
Corrosive?
Fresh Water
()
F.
required:
Aftercooler
Receiver
()
()
()
()
Type Starter
Shipment needed
282
Belt Guards
Centrifugal
Compressors
Centrifugal compressors are machines in which air is compressed by the mechanical action of
rotating vanes or impellers imparting velocity and pressure to the air. In a centrifugal compressor the
flow is in a radial direction. The following is adopted from the Compressed Air and Gas Handbook
with permission from the Compressed Air and Gas Institute, Cleveland, Ohio.
De@nitions
The following definitions will be helpful in understanding the construction and application of
centrifugal-type compressors:
Base pzate is a metal structure on which the compressor is mounted and after the driver as
well.
Capacity is the rated maximum flow through a compressor at its rated inlet and outlet temperature, pressure and humidity. Capacity is often taken to mean the volume flow at standard inlet
temperature, pressure and humidity.
Casing is the pressure-containing stationary element that encloses the rotor and associated
internal components, and it includes integral inlet and discharge connections (nozzles).
Diaphragm is a stationary element between the stages of a multistage centrifugal compressor.
It may include guide vanes for directing the air or gas to the impeller of the succeeding stage.
In conjunction with an adjacent diaphragm, it forms the diffuser surrounding the impeller.
Diaphragm cooling is a method of removing heat from the air or gas by circulation of a coolant in passages built into the diaphragm.
Eficiency
by a qualifying statement that identifies the efficiency under consideration. (See Sec. 7) for
definitions of adiabatic, polytropic, etc.)
Flange connection (inlet or discharge) is a means of connecting the casing to the inlet or discharge piping by means of bolted rims (flanges).
Guide vane is a stationary element that maybe adjustable and that directs the flowing medium
to the inlet of an impeller.
Impeller is the part of the rotating element that imparts energy to the following medium by
means of centrifugal force. It consists of a number of blades mounted so as to rotate with the
shaft. Impellers may be classified as follows:
a. Open face without enclosing cover, may be cast in one piece, milled from a solid forging, or built up from castings, forgings, or plates.
283
b.
Closed ~pewiti
wtichmay
becasth
onepieceor
butit
up from castings, forgings, or plates. Blades maybe attached to the enclosing cover and
hub disk with separate rivets, with machined integral with the blades or by welding.
Impellers are further classified with respect to blade form, as follows:
a.
b. Backward
Performance curve is a plot of expected operating characteristics (e.g., discharge pressure versus inlet volume flow, or shaft horsepower versus inlet volume flow; see Figure 7.18)
Rotor is the rotating element and is composed of the impeller or impellers and shaft and may
include shaft sleeves, thrust bearing collar, and a thrust balancing device.
Seals are devices used between rotating and stationary parts to separate and minimize leakage
between areas of unequal pressures. Basic types include clearance-type metallic labyrinths:
single and multiple, injection-type labyrinths, eductor-type labyrinths, or a combination. Dry
contact seals include carbon ring and synthetic materials such as Teflon. Liquid-injection
types use water or oil seals.
Shajl is that part of the rotating element on which the rotating parts are mounted and by means
of which energy is transmitted from the prime mover.
Shafi sleeves are devices that maybe used to position the impeller or to protect the shaft.
Single-stage
centrifugal compressors
classified as follows:
a. Single-fhw.
b.
Double-flow.
Stability or percentage stability is 100 minus the surge limit at rated discharge pressure, where
the surge limit is expressed in percentage of rated capacity.
Surge limit (pulsation point) is the volume flow below which partial or complete cyclic flow
reversal occurs, resulting in unstable aerodynamic operation.
Thrust balancing
device (balance piston or drum) is the part of the rotating element that
284
Centrifugal
Compressor
Characteristics
285
INLET DIAPHRAGM
IMPELLER
OUTER
INTERSTAGE
DIAPHRAGM
EXIT GUIDE
SPACER
UIR ~
ROTOR
SPACER
5%2
VANE
DISCHARGE
BEARING
AND SEAL
AND SEAL
HOUSING
STUB SHAH
by multiple impellers that act in series and are mounted on the same shaft. Flow of gas between stages
is guided by the interstage diaphragms from the discharge of one impeller into the inlet of the next
impeller. Sealing between stages is done by the labyrinth rings, which impose restriction on the flow
between impellers at the shaft, at the impeller eye, and at the balancing drum.
286
Operating
Characteristics
Demand
Load
Regardless of the actual service to which a centrifugal unit maybe applied, the general nature of
the demand load may be divided into three classifications as follows:
1.
Frictional resistance.
2.
3.
The frictional resistance load is that which would be typically encountered in ISC application. It
is the pressure necessary to overcome
equipment.
The fixed-head or pressure load is that which is required to overcome a controlled back pressure.
A third type of load, which is a combination of the above, is by far the most common. Viiually
all loads are to a certain degree a combination. An example of this type of load is that of ISC prospect
for which the majority of the pressure requirement is to overcome frictional resistance, but in addition, some is required to maintain a pressure at the sand face to prevent back flow.
In Figure 7.18, the three types of loads are shown graphically. Curve CD represents the purely
frictional load, curve AB represents the fixed head, and curve AE represents the combination.
287
c
Capacity
Application
As previously
of Centrifugal
to Loud
variable-
pressure unit. This is shown in Figure 7.19 line JK. Actually, because of the decrease in volumetric
efficiency
288
LK
J
Capacity ~
of Centrifugal vs Reciprocating
1989)
Compressor
It would be possible to select a centrifugal or a reciprocating compressor for the same flow and
pressure as indicated by point H, and the characteristic curves are shown in Figure 7.20 as FG for the
centrifugal and as LJ for the reciprocating units.
To explain the application of a centrifugal unit to any given service, the performance
curve,
shown by line FG, is superimposed on a demand or load curve, AE. With this combination of com~
pressor and load, the capacity handled will fall at the intersection of the two curves at point H. This is
only point at which the compressor will operate at that given demand requirement.
289
J
Capacity ~
For many applications, the demand curve will change, and this may result from varying the frictional resistance. This is shown by lines AE and AE, in which case the flow handled by the compressor is at the intersection of these two curves, or point H.
Controlling
Pressure
or Capacity
The use of a valve as a means of controlling pressure or flow at any given value is the simplest
form of control. It is not efficient, however, since this artificially created resistance represents an irrecoverable loss of power. A more efficient way to control the unit for any given pressure or flow is to
vary the speed. This creates a family of curves like those shown in Figure 7.21 by curves FG, FG,
~?G,,
Yand so on. By varying the speed, it is possible to set the intersection of the compressor charac-
teristic and the demand curve to any given required pressure or flow within the operating limits of the
compressor and driver.
290
Pumping
Limit
\
E
I
F
F
Fl,l
/
G
I
G
\
G,,,
l
A
Capacity ~
FIGU~
driver. For steam-turbine drive, the normal method of control would be vary the speed, permitting
efficient operation and a wide range of control. In the case of motor drive, the picture becomes more
involved, since the most commonly
drive, the following
1.
Speed variation maybe obtained by using hydraulic coupling between the motor and the
compressor.
the compressor.
291
Pressure or flow
variation
may be obtained
by means
of
a butterfly
valve
or
equivalent installed near the compressor inlet or near the compressor discharge, preferably the former.
vanes.
Static-type, variable-frequency
Wound-rotor induction moto~ Speed variation maybe obtained by varying the resistance in
the rotor or secondary circuit using either a liquid rheostat or a step resistance (rarely used).
Briefly, a centrifugal compressor can be operated to meet a given pressure and flow requirement
by varying either the demand curve or the compressor characteristic curve so that the intersection of
these curves will beat the required point.
Selection
The preceding
of Unit
emphasizes two points in connection with its proper application. They are:
1.
The importance of investigating the operating conditions to be sure that the compressor is
large enough to meet the job requirements.
2.
The necessity .of a means of controlling the centrifugal compressor when variations in the
operating conditions are such that a pressure would be developed in excess of the actual
requirements.
The following
cally capable of meeting the requirements under the most adverse operating conditions. Also,
knowledge of the range of these conditions is needed to ensure that a driver of sufficient size is furnished. Attention is called to Table 7.3 which shows a typical centrifugal inquiry form.
1.
2.
3.
4.
Maximum k value.
292
5.
6.
7.
8.
Supercompressibility
9.
In selecting a machine for a given application, the preceding characteristics must be carefully
considered. The practical lower limit of capacity for which either compressor may be designed is
determined by the specific requirements. In connection
with low-capacity
also be given to an economic comparison with reciprocating and rotary compressors capable of doing
the same job, including cost and installation. In some instances, it may be economical to use a centrifugal compressor for a portion of the required pressure range, followed by a reciprocating or rotary
compressor for the remainder of the pressure range.
Since the size of the required compressor and the horsepower of the driver are direct functions
of inlet capacity, extreme care must be exercised in establishing actual inlet capacity. Standard capacity (scfm) is frequently used in specifications and, since there are several standards in common use, it
is necessary to establish the particular standard involved. Air has been frequently specified at 14.7
psia (1 .013 bars), 60F (15.5C) and dry. For all applications, the standard capacity must be corrected
to the actual inlet pressure, inlet temperature, and moisture content to arrive at actual inlet capacity. In
consideration of accepted metric equivalents, the CAGI (Compressed Air and Gas Instituti) Standards
Committee has settled on 14.5 psia (1 bar), 68F (20C) and dry as a standard.
.,
Approximate
Although the following
rule-of-thumb
Selections
Limitations
brief survey of the following points will serve to guide the selection of the type of compressor and, in
some cases, will eliminate centrifugal units from consideration where conditions make it inherently
not suited.
There is no define minimum inlet capacity. However, single-stage and multistage compressors
have not been built for inlet capacities lower than 250 cfm (425 cmh).
293
Numerous
commercial
centrifugal
compressors
25,000 psig (172 bars). For other applications, centrifugal compressors have been built for discharge
pressures up to about 10,000 psig (690 bars).
There are frequent applications involving an inlet capacity within the practical limits of a centrifugal compressor, but because of the high pressure ratio, the inlet capacity to the last stage maybe
so low as to preclude the use of a centrifugal compressor.
Because of thermal stresses and problems of alignment, a limit of 450F (232C) is normally set
as the maximum discharge temperature at design pressure and capacity for centrifugal compressors of
standard design using standard materials.
Where discharge temperatures must exceed this value, special designs and materials are used.
Centrifugal compressors have been built for discharge temperatures up to about 1000F (538C). In
cases where the pressure ratio would result in a discharge temperature in excess of 450F (232C) and
where this temperature is neither required nor desired, several alternatives may be used:
1.
Multisection compressors with external coolers between sections of the same compressor.
2.
Multicasing
compressors
If necessary, each
Internal cooling
densed vapor into the gas stream in the diffuser sections of the compressor, or indirectly as
with coolers mounted within the compressor casing or external to the compressor.
294
CompressorData
Normal
Maximum
1. nlet flow*
analysis
%volume
Gas
Constituent
2.
1Mass flow
3.
1Molecular weight
4.
I\verage compressibility
5. Ikverage
Minimum
Guar.
(Zave])
K (CP /~)
6.
1Relative humidity
7.
13arometer
8.
1Met temperature
9.
1Inlet pressure
Flow
Pressure
Temperature
15. Indicate guarantee
* If inlet flow
Temperature
is
point
listedSCFM (standmdcol litions) indicate conditions under which flow was measured, i.e.
, Pressure
% R.H.
295
Mt.
ElectricMotorData
Enclosuretype:
Sitedata:
Area
TEWAC
TEFC
Dlv
Group
Class
Weatherproteeted
TEIEF (using_gas)
Force ventilated
Nonhazardous
Open-dripproof
Open
Exp. Proofi
Fumes
Jnusualconditions
Dust
%ltitude
Ambienttemperature
Jlotor Type:
Basic Data:
Squirrel
Synchronous
Group
Class
Brushless
volts
phase
Hertz
Slip ring)
Wound rotor
Nameplatepower_
Direct current
ServiceFactor
Synchronousrmp
Insulationclass
Type
Temperaturerise _C
above_C
Unloaded
Voltagedip
by _
starting:
Full voltage
Reducedvoltage
Loaded
296
Sizing
Consideration
Centrifugal air compressors normally are specified on the basis of required air flow volume.
However, there are several ways to calculate volume, and serious problems can result unless both user
and manufacturer use the same method. At the very least, the user can have trouble comparing bids
from competing manufacturers. At worst, he may choose the wrong compressor.
These problems can be avoided by specifying
and by understanding how compressor capacity is affected by variable ambient conditions, such as
inlet pressure, temperature, and relative humidity. Also, factors such as cooling-water
temperature
and motor load must be considered before a compressor and its drive motor can be sized.
The main item that must be specified is inlet air volume. Standard cubic feet per minute (scfm)
is the common unit of measure for compressor inlet capacity; however, several definitions of this unit
exist. One definition of scfm fixes the variable at 14.7 psia (1.013 bars), 60F (15.5C), and 0910relative humidity. Another, adopted by the ASME, specific scfm at 14.7 psia (1.013 bars), 60F (15.5C),
and 3690 relative humidity. A third (metric) definition specifics a standard (or normal) cubic meter of
air at 1 atmosphere, 32F (OC), and 0910relative humidity.
To avoid the confusion caused by these variable standards, some users have adopted a simpler
unit that expresses inlet volume in terms of the actual inlet pressure, temperature, and humidity. This
inlet cubic feet per minute (icfm) indicates the actual volume entering the first stage of a multistage
compressor
at the expected
operating conditions.
This volume,
design, nozzle diameter, and casing size that provides the most efficient operation.
Inlet pressure is determined by taking the barometric pressure and subtracthg a reasonable loss
for the inlet air filter and piping. A typical value for filter and piping loss is 0.3 psig (0.02 bar).
The need to determine inlet pressure at the compressor flange accurately is particularly critical
in high-altitude installations. Because barometric pressure varies with altitude, a change in altitude of
more than a few hundred feet can greatly reduce compressor capacity. Often, the lost capacity can be
restored by using larger-diameter impellers, but occasionally
a different-sized
compressor must be
used.
The other variables that influence compressor capacity, temperature, and relative humidity of
the inlet air must be considered over the range of conditions expected in service. Air volume is lowest
at the highest expected
297
designed to deliver the required flow at the highest temperature expected. This guideline also applies
to the temperature of the cooling
stage following
an intercooler.
Relative humidity also affects the useful volume of air available at the compressor inlet. The
higher the humidity, the less is the effective air volume available: thus the impellers must be sized for
the highest humidity expected.
(7.40)
Air density (weight of a cubic foot of air) is inversely proportional to its absolute temperature.
Thus, the higher the temperature, the less weight flow in each cubic foot. The weight flow delivered in
summer is less than in winter. Therefore, the specification for a compressor should provide for the
required weight flow to be delivered on a hot summer day. A slightly larger compressor
will be
required if air temperature is 90F (32C) rather than 68F (20C). If the manufacturers rating (based
on air at 68F or 20C) is accepted, the compressor will not deliver the same weight of air per minute
at 90F (32C).
Another reason that the volume flow by itself must be qualified is that air density also depends
directly on air pressure. Because atmospheric air pressure depends on altitude, a compressor installed
at a higher elevation (above sea level) gets less weight of air in each cubic foot of intake air than the
same compressor installed at sea level. This change in weight flow due to differences in barometric
pressure can be significmt. For example, because of the lower atmospheric pressure, a compressor in
Kansas City will deliver nearly 5% less air than the same compressor installed in Miami.
298
= 144P
P
(7.41)
RT
p=
144(Pb AP)
(7.42)
RmT
where pb = absolute barometric air pressure, psia
~ = volume flow at the given air density, cfm
~
Density of moist air at various temperature can be estimated from Figure 7.22.
299
MIXTURE
1.000
10
14
18
TEMPERATURE,
22
26
30
34
T
0.998
0.996
40
38
-F
-1o.
0.994
0.992
0.990
_3(h
0.988
40\
0.986
T-
50,
0.984
P~
0.982
\ \
I
0.980
I
0.978
0.976
SPECIFIC
0.074
I
;
1
I
0.972
0.970
50
60
70
MIXTURE
80
TEMPERATURE,
90
100
100
110
FIGURE 7.22 Density of Moist Air as Function of Temperature (After Rollins, 1985)
300
Humidity
Another variable that often causes confusion
is humidity. Atmo-
spheric air always contains water vapor. As a result, the compressor takes in a mixture of air and water
vapor. This affects compressor operation and performance because the higher the pressure of the air,
the less water vapor it can hold. And what it can no longer hold condenses in the intercoolers and
aftercooler and is drained as water. So, once again, the weight of the cubic foot in is not the same as
the weight out. The compressor must be sized slightly larger to allow for the water vapor loss, which,
although it is part of the inlet flow, is not part of the delivered weight flow.
20
10
40
100
200
300
500700
3000
o
2000
1500
1000
0
0
700
500
400
0
0
300
1 I ..
3
200
0
150
100
5
70
50
40
30 !
It
1X
II 141111
PERCENT AIR
BY WEIGHT
20
15
1(-)
-0.2
I 1111111
.3
.4.5
.7
1.0
II II
1.5
2.
4 5
IIHI
7
10
I I
20
30
A portion of the inlet volume is attributable to water vapor. This depends on the relative humidity of the intake air (See Figure 7.23).
Specify
Ambient
Conditions
In order to be sure to have enough air, the compressor buyer must be careful in specifying ambient conditions. These should tend toward the minimal conditions, that is, high air temperature, normal
barometric pressure, and high humidity. This does not mean that the specification should be based on
the maximum air temperature on record. The result would be an unnecessarily large compressor.
A related item to be considered in selecting a compressor of the correet capacity is coolingwater temperature, since most air compressors are intercooled. Water temperature has much the same
effect as air temperature. This is easy to understand because the water cools the air before it enters the
next compression
be. Therefore,
stage. The warmer the cooling water, the warmer the air, and the less dense it will
wise, the compressor will deliver less air than expected. When cooling-water
temperature is lower
Centrifugal
Centrifugal air compressor
Air Compressor
performance
Characteristic
charge pressure versus flow. This is a continuously rising curve from right to left (Figure 7.24). The
effect of environment on performance requires understanding of two phenomena associated with this
curve: choke (stonewall) and surge.
302
120
/
/t
115
/
~
,+
\7
<.9
&
Guaranteed Design
Point
/
/
/
Stonewall
75
65
70
75
80
85
90
95
100
105
Curve
increased volume of air. As the system pressure continues to deerease, the air delivery from the compressor continues to increase until the air velocity somewhere in the compressor reaches the speed of
sound. At this point, the flow is said to be choked because further reduction in system pressure does
not result in additional air delivery by the compressor.
On the other hand, the maximum discharge pressure of the centrifugal compressor is a function
of the iuterseetion of the surge line and the sloping performance curve. When the compressed-air system pressure increases, the compressor furnishes less air as higher pressures are encountered until the
system resistance is matched. This relationship may continue until the compressor is unable to maintain a steady flow of air into the system.
303
When the compressor cannot maintain a steady flow of air, backflow from the system through
the compressor occurs until a momentary equilibrium is established between the compressor and the
system. This backflow is commonly
is roughly equivalent to
the stalled condition of an airfoil. Under this condition, compressor operation moves from surge to
some point below the operating point shown on the performance curve. When the compressor continues to operate against sustained excessive system pressure, compressor operation moves up the curve
and surge occurs again.
Neither of these conditions is desirable and both should be avoided. Control systems that allow
the compressor to function without reaching the choked or surge condition must be based on prevailing environmental conditions. It is, therefore, helpful to examine individual environmental factors that
can affect compressor performance.
Characteristic centrifugal compressor performance curves (Figure 7.24) rise from right to left,
showing a reduction in delivery at increasing pressures. Stable operation occurs on the portion of the
curve that lies between the design point and the surge line.
Weight
or Volume
Flow
The compressed-air system is in reality a vessel that stores energy in the compressed air, energy
that can be transferred to the formation. When a portion of this stored energy is withdrawn from the
system, it must be replenished by the compressor.
The performance of the compressed-air system is measured by the pressure of the air in the system. Air pressure for a steady air usage and relatively constant system temperature depends on the
weight of the air in thes ystem. Boyles law states that, for a constant gas temperature,
(7.43)
where v = specific volume, ft.3 /lb
Plvl
Therefore,
WI
P2v~
=
(7.44)
W2
304
PI
=
WI
P2
W2
(7.46)
or
P1W2
P2=
(7.47)
WI
As air is withdrawn from the compressed-air system, the weight of the air (if not replaced) and
the air pressure decrease. The performance of the compressed-air
weight of the air delivered by the compressor. Because each centrifugal compressor has a fixed volume design capacity, in cubic feet per minute or cubic meters per hour, the weight flow capacity is
determined by the pressure temperature and relative humidity of the air entering the compressor.
The compressor characteristics curve and work input are related to flow as a function determined by physical geometry, blade angle, speed of rotation, molecular weight of gas, and other factors to a minor degree. Once this characteristic has been established for a compressor,
affected by inlet air temperature, inlet air pressure, and cooling-water
Effect
of Inlet
it can be
temperature.
Air Temperature
The relationships (Equation 7.7) can be used to explore the effects of inlet air temperature.
Aerodynamic
(work) per
pound of air or, more simply, as feet of head. Power is then obtained by multiplying head times total
weight flow and considering mechanical losses and efficiency. The polytropic head is obtained from
Equation 7.28.
For a fixed geometry and constant speed, air compressor head per stage is constant. The only
variables are the inlet temperature and the pressure ratio. Therefore, if inlet pressure is constant and
305
inlet temperature is increased, discharge pressure must necessarily drop to maintain the equality. Conversely, when inlet temperature decreased, discharge pressure must increase.
Inlet air temperature also affects the weight flow through all types of compressors:
~-QP=
RT
(7,48)
constant x ~ = constant x p
Weight flow through a centrifugal compressor is proportional to inlet volume and inlet pressure
and indirectly proportional to inlet air temperature. Because P l/RT 1 is weight density, another way of
stating this relationship is that weight flow is proportional to density. As inlet temperature decreases
weight flow through the compressor increases, and vice versa, although volume flow remains constant.
Often the compressor manufacturer states the capacity for a standard air temperature, which
may be as low as 60F (15 .5C). Then, when the compressor operates with 90F (32C) inlet air, for
example, the weight flow is reduced by the ratio of the absolute temperatures:
(7.49)
flows or 5.5%, other factors remaining unchanged. In S1 units, capacity stated in terms of a standard
temperature of 68F (20C) with an actual inlet temperature of 86F (30C) would mean a weight
reduction of
(7.50)
Proper performance of the compressed-air system requires that the compressor rating be guaranteed for the summertime air inlet temperature
or that a weight flow ratimz
.
. be .mmranteed at the same
conditions.
306
The effect of wintertime air temperature on air density must also be considered. A compressor
rated for 90F (32C) summertime inlet air temperature, for example, will have a 17% higher weight
flow when operating with a 10F (-12C) wintertime inlet air temperature:
w~
90+460
_ ~17
WI
10+460
(7.51)
This increased weight flow will not impair the pressure performance of the compressed-air system because the resultant increase in pressure can be relieved through a relief valve or through more
frequent cycling of the compressor. But this increase weight flow will increase the cost of compressing the air because of the increased power required:
(7.52)
P=c~xw
where: P is the power required. Therefore, for this example, the power required by the compressor will increase 1770. Alternative expressions for power are
P=
C2X;
P=c~xw
Figure 7.25 shows the effect of inlet air temperature. Increasing temperature means decreasing
flow and power requirement, and decreasing temperature means increasing flow and power requirement. The implications to the buyer are twofold. First, the compressor must be rated at a sufficiently
..
high temperature so that the plant does not run short of air on a hot day--perhaps not the highest temperature of the year, but a mean temperature based on a reasonable confidence level. Second, controls
must be provided to prevent the compressor from drawing excessive additional power when the air is
cooler.
307
D sign Point
/=,
.
-i,
Surge Line
+ig~j:er
Weight
100
Flow, Percent
Effect
of Inlet
Air Pressure
A change in the inlet pressure does not affect the established pressured ratio, but the discharge
pressure varies directly with changes in the inlet pressure (Figure 7.26). Reducing the inlet pressure
also reduces the weight flow through a compressor, but the volume flow remains the same. Because
weight flow is reduced, the power requirement is also lower.
Decrease in
inlet pressure
reduces discharge
pressure
reduces power
requirement
100
308
If a given discharge pressure is required, a higher pressure ratio is required when the inlet pressure is lower, which, in turn, causes a higher work input. This factor should be considered when the
compressor operates at high elevations.
Under normal operating conditions, the daily change in inlet air pressure is relatively small,
except when the inlet air filter becomes dirty and needs cleaning.
Effect
First-stage performance
of Cooling-Water
Temperature
temperature. However, all successive
temperatures directly affect the temperature of the air entering the second
Colder Water
Increases Volume
and Discharge
Pressure
-\-\
\
l\
Surge Line
~t
100
Warmer Water
Decreases Volume
and Discharge
Pressure
Colder Water
Increases Power
Requirement
Warmer Water
Decreases Power
Requirement
A reduction in cooling-water
power consumption.
309
Most significant are the combined effects of simultaneous changes in water and air temperatures
because in most plants they tend to increase or decrease together. On a summer day, higher air and
water temperatures are normal, whether the water comes from cooling towers, a public supply, a river
or the sea, although changes in air temperatures are more extreme than changes in water temperature
because of the moderating effect of heat storage capacity in water. The combined effect of higher air
and water temperatures is to depress the compressor characteristic, resulting in lower discharge pressure, lower weight flow, and lower power consumption. Colder temperatures increase weight flow and
power requirements. This discussion is related to an uncontrolled compressor. It also describes what
happens to an inskilled compressor selected for rated ambient conditons at a give site.
Designing
the Compressed
Air System
Air compression plant for an in-situ combustion project (See Figure 7. 1) it consists essentially
of one or more compressors, drivers, air intake filter, aftercooler and separators, air dryer, interconnecting piping, and a distribution systems to carry the compresses air to the injection wells. For illustration, the compressor in Figure 7.1 is shown as a two-stage compressor.
Air at atmospheric pressure is filtered before entering the system to remove dirt that would harm
the compressor. After each stage of compression the resulting heat is removed in the coolers, followed
by removal of condensed water in the separator. The compressed air is given a final filtering to remove
any material that might block the formation. Remaining water vapor is removed in a final dryer, and
then, if a dry bed dryer was used, the air is filtered once more. The blowdown line provides for pressure release when the compressors are shut down. The fuel source shown in the Figure 7.1 could be
natural gas line, an electric line, or diesel fuel storage tank. In most areas, environmental regulations
may require the treatment of engine exhaust (if natural gas or diesel fueled engine is used to drive the
compressor)
to remove NO= and other pollutants. The compressor and the engine use fire-resistant
310
The most important consideration in the design of compressed air system for fireflood project is
the availability of some air for the project at all times. This is very important for a number of reasons:
1.
When air injection stops, in most cases there is a small backflow into the injection well. In
unconsolidated formations, this means sand-fill in the injection well and costly workover.
2.
In thick formations,
vertical
drainage may resaturate the burned zone. This resaturated rock must be burned again before
the burning front can proceed. This increases the overall air requirement for the project.
3.
resumed.
Therefore, consideration should be given to the use of smaller multiple units in the project. If
multiple units are used this will ensure the availability of some air all the time. Economics
is the main
reason why many operators shy away from the use of multiple units in ISC projects.
There are many and equally important considerations in designing a compressed air system for
in-situ combustion services. These include:
1.
Establishing the peak and average air demand and injection pressure.
2.
3.
4.
5.
6.
7.
8.
311
Establishing
Injection
Rate
and Pressure
In the design of the compressed air system for in-situ combustion operation the major governing
factors are tie daily peak air injection rates and pressure required to conduct the field operation. This
is needed for determining of the compressor plant capacity and the compressor type to be used. In dry
combustion operation air need to be supplied at a rate commensurate with the desired rate of advancement of the burning front. For a given fuel concentration, higher the injection rate farther the combustion front will advance and faster the oil will be produced.
designed based on maximum possible air rate that will be used on a constant basis during most of the
burning operation. The compressor must be sized not only to meet the peak air requirements, but be
flexible enough to accommodate a reduction in capacity due to changes in operating conditions.
Estimation of peak and average injection pressure is also very important because together with
the daily output they influence the type of compressor to be used. As noted previously, typically at the
start of a combustion project the injection pressures are higher than the projects average injection
pressure. The injection pressure drops once the combustion front move away from the injector. This
requires operating pressure flexibility. Compressor
Selection
of Compressor
and Prime
Movers
Ideally a compressor considered for ISC service is required to satisfy the following
criteria:
Must lend itself to outdoor installation with minimum of weather protection required in
more severe climates.
Must have the ability to operate over a wide range of capacity and pressure.
Must have the ability to operate over a wide range of volumetric efficiency.
Must have the ability to deliver the desired volume without loss of efficiency.
312
Selection of compressors for a specific fireflood project must consider all conditions anticipated
through the life of the project. Three primary elements that determine the type of compressor to be
used in a project are:
1.
2.
Dependability of service (compressor must have a high degree of operating reliability with
plant availability to be 90% or better).
3.
These three items usually overshadow the first cost making it a distance fourth. Large projects
requiring large volumes of air at moderate pressure (less than 2,000 psi) tend to use centrifugal compressors. Their flow capacity ranges from 5144 MMscfd and can provide long periods of relatively
maintenance free operation.
Multistage reciprocating
compressors
pressure (greater than 2,000 psi). These compressors can have discharge pressures to 8,000 psi and
are normally limited to 7 MMscfd
requirements are small use of skid mounted packaged reciprocating compressor systems are recommended because they are less expensive to purchase and operate. If both centrifugal and reciprocating
compressors can meet the flow and pressure requirements of a project it is suggested that an individual assessment of the advantages and disadvantages of each type as they related to users needs be
made prior to selection.
Another type of positive displacement
rotary air compressor. Rotary compressors
ranges from 0.52 MMscfd with discharge pressure usually limited to less than 200 psi. In large fireflood projects they are generally employed
as a pre-compression
spheric pressure to 200 psi and to eliminate stages from a reciprocating unit.
The prime mover or driver of a compressor is the main power source that provides the energy to
drive the compressor. It requires proper selection and matching to the compressor to ensure satisfactory compressor
313
should be
1.
Service requirements of the compressor (in-situ combustion compressor must operate continuously 24 hours a day, 7 days a week, the year around, and the driver should be selected
with this in mind)
2.
3.
Site conditions.
4.
Economic
1.
2.
3.
Maintenance cost.
4.
Installation cost (especially when comparing natural gas driven engine versus an electric
motor).
5.
Environmental consideration.
It is preferable to have the compressor vendor furnish the compressor, the driver and accessories
so that all elements are properly coordinated and assume responsibility for the complete package.
In in-situ combustion operation, low-speed natural gas engine is often the prime mover of
choice. Since engine manufacturers rate their engines for maximum, intermittent, or continuous service care should be exercised when selecting a gas engine driver to be certain that the horsepower ratings meets the required service. Since the recent federal and state air pollution regulations are very
restrictive with regard to engine emissions, it is advisable to review the emission characteristics
of the
engine prior to purchase. Low emission engines or catalytic converters are two potential solutions to
exhaust emission requirements.
If gas is not readily available at the site or gas engines cannot be used due to environmental consideration, electric motors may be used to drive the compressors at higher cost, but with less maintenance
expense.
compressors
Electric
motors
Electric
drive reciprocating
are constant speed machines and hence no rate adjustment is possible by speed control
314
i
problems.
such as is possible with a natural gas engine drive. There are three basic types of motors available that
apply to oil field compressors.
Proper motor selection is a very critical decision and involves several key factors. Some of these
included:
1.
A proper motor-to-compressor
speed match.
2.
3.
4.
Current restrictions.
5.
6.,
It is important to evaluate energy costs over the project life because it is the major part of the
operational expense.
Package
or Process
Compressors
Many combustion projects are undertaken to investigate the technical and economic viability of
the process in a given field and as such do not require the large elaborate compressor installations
found in large fireflood projects or gas transmission projects. Minimal initial investments and lower
operating expenses over the life of the project are the general requirements of such projects. Smaller
projects are also generally undertaken by organizations with limited technical manpower resources.
Smaller combustion projects requiring moderate volumes of air should consider installing packaged air compressors.
These factory-engineered
because they are less expensive to purchase (or leased) and install, have a high degree of reliability
and are easily relocatable. Operating and maintenance expense, excluding fuel are also lower than the
process type units found in larger projects. Since these systems are offered as a ready to install operable unit with a minimum of field construction, they relieve the operator or his technical staff from the
need for the sizing and specification of compressor and its auxiliaries.
The ideal and most desirable package for the operator is one in which only the required utilities
of power and cooling water brought to the package, with the end product being clean dry air.
315
Process compressor systems found in larger projects are similar in configuration to the packaged
system, but are designed for permanent installation. Unlike the packaged unit which include all the
components necessary to provide a controlled compressed air source the process system include only
the compressor, power transmission equipment and basic control system. Procurement of all the other
necessary components
equipment, etc., is the responsibility of the purchaser. Process compressor systems are designed for
high volume, high pressure unattended operations and have a high degree of reliability, with plant
availability to 95% or better. The system is designed to be highly flexible and capable of automatically adjusting to changing pressure and volume requirements inherent in combustion projects.
Process type compressor
operate and maintain than packaged units. In larger projects their use is justified due to their reliability, operational flexibility and low unit compression and maintenance cost.
Locating
the Compressor
Station
In selecting the site for compressor station location, items to be considered include its convenience for inspection and maintenance by technicians working on the lease, proximity to the water
supply, its nearness to lease road, surface topography, and in some cases the distance to the near est
electric powerlines. If feasible, the station must be located near the center of the lease to minimize the
length of the injection well piping network. In hilly country, the compressor station should be located
at highest possible elevation to allow adequate drainage in bad weather. The selected site should allow
for maximum exposure to prevailing winds in warm climates, but shelter the station in cold areas. The
compressor station must be located at sufficient distance from other facilities to minimize potential
noise problems or darnage due to explosion.
316
Control
and Safety
Systems
The typically oil field compressor has a basic protection and monitoring control system. These
include sensors to monitor process pressures, temperature, and vibration levels of major components.
Since reliable operation of the air compressor is critical to the success of combustion project, consideration should be given to equip the compressed air system with a highly dependable automatic warning and shutdown system.
A chief reason for employing an automatic warning and shutdown system is to free maintenance
personnel for duties other than monitoring compressors. The first cost of the system depends upon the
type and the degree of sophistication sought. Automatic warning and shutdown systems are available
from the simplest, designed to. shut compressor
too
high, without any indicating device, to the most sophisticated systems containing visual displays to
indicate both impending danger warning points and the exact cause of shutdown for more serious
conditons.
Since the first cost of the system depend upon the degree of sophistication sought, the question
of how much sophistication is warranted for a given machine is often raised. There is no simple
answer to this question. A trouble free air injection plant operation is vital to the economics
being of a combustion project. Economic
well
The
first cost of a surveillance system should be balanced against the potential loss of revenue in the event
of a compressor shut down.
Ancillary
Equipment
Ancillary equipment such as the air filters, air dryers, compressor lubrication system, interstage
coolers, interstage scrubber, etc., are vital to the trouble free operation of a air compression plant and
care must be exercised in their specification and selection. Completely pre-engineered units in addition to the basic compressor and driver usually include all the basic ancillary equipment from the air
filter to the aftercooler, to exhaust silencer.
Compressor vendor often offers a customized packaged ancillary system for his compressor. It
is recommended
that the operator give serious consideration to purchase the system from the vendor.
317
In any event, it is suggested that compressor vendors service be utilized during the ancillary equipment selection process.
Compressed
Air Piping
For compressors to be functional, it must be connected to the system with piping. Since the user
is responsible for sizing and furnishing the compressor suction and discharge pipings ystem, he must
at least have a rudimentary understanding of the compressor inlet and outlet pip@g requirements. The
inlet and outlet piping requirements are somewhat different for the reciprocating and centrifugal compressors and these are briefly addresses in this section.
A clean, cool, dry air supply is essential to the satisfactory operation of a compressor. The air
inlet must always be located far enough from prime mover exhaust pipe (if a fossil fuel engine is used
as the drive) to ensure that air will be free from dust, dirt, or moisture and contamination by exhaust
gases.
The inlet piping should be as short and direct as possible, with long-radius elbows where bends
are necessary. If the inlet pipe is extremely long, a larger size should be used. A good inlet piping system should be self-supporting. The compressor should not be subject to the pipe strain associated with
long runs of heavy pipe.
Compressors:
Since steady flow does not exist in either the inlet or discharge of a reciprocating compressor
cylinder, compressor piping should not be sized based on pressure drop under steady flow conditions.
Sizing based on constant airflow from the compressor will result in the selection of pipe size, which
are too small. If the pipe leading to and from a compressor cylinder are too small the pressure drop
through lines will decrease the capacity of the compressor cylinder and increase the power required to
compress a unit volume of air.
Due to the cyclic intake of air into the compressor cylinder, the air inlet flow to a reciprocating
air compressor will be pulsating flow and cause inlet and discharge pipe to pulsate. Uncontrolled piping pulsation can affect compressor performance, and even cause piping failure. If the pulses generated by the cylinder are in step with the natural frequency of the air column within the pipe, the pulse
amplitudes will peak and the system is said to be resonant. Resonance
undue pressure losses, reduced capacity, after-cooler
method to avoid resonance is through proper piping design and avoiding lengths that can cause pipe
318
of
accumulated dirt and oil. Discharge piping should be connected to the compressor with flanged fittings to permit removal of the cylinder at any time without disturbing the piping. Provision must be
provided to allow the expansion of disch&-ge pipe carrying hot air. All discharge pipes should be suitably anchored to minimize vibration, which results from the pulsating discharge of the compressor.
Compressed
Air Distribution
System
Piping
The compressed air distribution systems piping is one of the most important elements of a fireflood air injection plant, The air from the compressor is distributed to various injection wells through
a system of pipelines, which are either buried or supported close to the ground level. Normally a carbon steel pipe with no internal coating is used for the air injection lines. In many projects, however,
internally coated piping is being used to avoid corrosion. Since any pressure loss between the compressor and the injection well is an unrecoverable
general guidelines
The air injection distribution line must be designed for a working pressure preferably 25%
greater than the maximum expected wellhead injection pressure.
319
2.
manifold and the injection well head will not exceed 5 psi per 100 feet of pipe. Pressure
drops must be determined based on maximum airilow rates through individual piping section. Pressure drops through valve, fitting, etc., should be pegged at 5 psi total for each
expansion section.
3.
Air injection lines from the manifold to injection wells should contain flow control valves
and meters to control the flow of air to each injection well. In cold climates, these lines
should be buried to minimize heat losses and to prevent any moisture from freezing and
plugging the line.
4.
A positive air flux must be maintained at all time from the compressor to the formation to
achieve high level of safety.
Lubricating
Oil Requirements
Compressors are lubricated to prevent wear and to seal clearance around pistons, valves, and
rotor elements, Both mineral and synthetic oils are used to lubricate air compressors. Since mineral
oil based lubricants began to breakdown and deposit carbon about 325F and can cause explosion,
lubricant oils have to be completely excluded from the air stream. During normal operation of the air
compressors, some lubricant is always carried into the air discharge lines as a suspended mist. Since
oil filtering equipment suitable for high-pressure air injection systems is sophisticated and expensive,
use of mineral oil in air compression system is not recommended.
when both fuel and oxygen are mixed in the proper proportions in the presence of an ignition source
possessing sufficient energy to ignite the combustible vapors. Early day fireflood explosions in compressor and distribution systems is often attributed to the use of mineral oil based lubricants in compressor cylinder and the presence of an oxygen enriched atmosphere in the piping.
Synthetic lubricants do not readily breakdown and have the advantages of lower volatility and
flammability
when compared
to conventional
employed in the fireflood projects are designed to provide high temperature stability and long term
lubrication with minimal deposit than conventional mineral oils.
A characteristic of compressed air system is that the auto-ignition temperature of air/fuel mixtures is decreased as air pressure increases. The auto-ignition
when the air stream contains rust particles, which act as a catalyst promoting the oxidation of the
320
lubricant. Whereas the auto-ignition temperature of Andersol 500 (a widely used synthetic lubricant
in fireflood projects) in air is 770F at Opsig, it drops to 615F at 200 psig air pressure and to 5 15F at
2,000 psig air pressure. Despite the substantial drop in the auto-ignition temperature of the synthetic
lubricant, the normal temperature of the discharged compressed
than the depressed auto-ignition temperature of the lubricant at elevated pressure. The auto-ignition
temperature is lowered even further if rust particles are present in the air stream. It is the responsibility of the operator to keep the air stream free of contamination to reduce the risk of explosion.
Since synthetic lubricant tend to swell many types of packing and gasket elastomers, it is necessary to make certain that packings and gaskets are made of a material not affected by the synthetic
lubricant. It is recommended
Explosion
in Air Compression
Plant
Although explosions in compressed air systems have happened in the past and are uncommon in
present day operation, they continue to happen. No in-situ combustion project should be undertaken
without first conducting a careful study of potential hazards and a preventive maintenance program to
eliminate or reduce the possibility of such explosions must be undertaken. A brief of description of an
in-situ combustion project air system explosion
(Hardin, 198 1)
Operating under a U.S. Department of Energy Contract, Cities Service Oil Company conducted
a cost-shared commercial scale demonstration in-situ combustion project in the Bodcau Fee lease of
the Bellevue Nacatoch Reservoir near Shreveport, Louisiana. This 19acre, 5pattern project was
started in July 1976 and terminated in 1982 (See Chapter 6 for more information on this project).
This project was carefully designed and operated in an expert manner and was an economic success. The operator employed three reciprocating
pressor cylinders were lubricated with synthetic lubricant to minimize explosion hazards. In spite of
considerable care in design and operation of the air compression plant, an explosion occurred in the
air compression/injection
system.
Analysis of the explosion indicated that the explosion have resulted from the auto-ignition of
lubricant carried over from the cylinders of the #2 compressor. The lubricant which had been accumu-
321
lating in the low spots of compressor #2 cylinder discharge valves was carried over in the air stream
into #3 compressor cylinders. As the lubricant passed through the #3 compressors
cylinder discharge
valves, the air turbulence transformed the liquid lubricant into a finely divided oil mist. This mist
formed an air/fuel mixture, which was within the flammable range for the lubricant. The mixture was
ignited at the #3 compressor discharge valve by the heat of oxidation resulting from the swiftly moving oxygen enriched airflow. The ignition caused a low order explosion and the resulting shock wave
was sufficient to propagate a larger oil film detonation, which then traveled throughout the air discharge piping system.
In order to prevent similar problems in the future, the operator adopted the following preventive
program.
1.
Periodic washing of the air distribution system with 5% (by weight) solution of sodium
hydroxide and sodium nitrate, with frequent checks of possible lubricant collection points
made between washes. The chemical wash was designed to destroy any oxidized residues
in the lines.
2.
Careful checks for buildup of carbonaceous residues on discharge valves and cylinders during routine maintenance.
3.
Settings of high temperature shut down switches were lowered to activate at 300F and discharge pressure held at minimum levels.
4.
322
REFERENCES
1.
American Petroleum Institute, (1989), Specification for Packaged Reciprocating Compressors for Oil and Gas Production Services, API Specification
2.
Bresler, S.A. and J.H. Smith (1970), Guide to Trouble Free Compressors, Chemical Engineering (July), pp. 161170.
3.
Coker, A.K. (1994), Selecting and Sizing Process Compressors, Hydrocarbon Processing,
(July), pp. 39-47.
4.
Gas Processors Suppliers Association, Tulsa, Oklahoma, Engineering Data Book, 10fi Edition, Section 13.
5.
Govern-
6.
Loomis, A.W. (1980), Compressed Air and Gas Data, Ingersoll Rand Co., Woodcliff
Lake,
Rollins, J.P, (1989), Editor, Compressed Air and Gas Handbook, 5fi Edition, Compressed
Air and Gas Institute, Cleveland, Ohio, P T R Prentice Hall, Englewood Cliffs, New Jersey,
Chapters 2,3,4,
323
CHAPTER
8 IGNITION
Introduction
Afireflood
ignition could not be sustained for an extended period of time the project may not be successful.
Many past firefloods especially in heavy oil reservoirs have failed an account of poor ignition attempt.
Hence ignition is one of the critical aspects of combustion recovery. The ignition is generally initiated
by supplying large amounts of heat, sufficient to elevate the formation temperature around the injection well to a value higher than the crudes ignition temperature. Once this temperature is reached, in
the presence of an oxidant, usually air, ignition is achieved. A sharp increase in the sand-face temperature at constant air injection and heat input rates is indicative of ignition. TWical values for ignition
temperature range from 250500F.
Several different techniques are employed in the field to achieve ignition. These techniques can
be divided into two categories, spontaneous and artificial. If the reservoir temperature is high enough
and the formation contain sufficiently reactive crude than it can be auto ignited after a period of air
injection without the need for any artificial ignition device. Alternatively, the formation can be ignited
using one or more of the many artificial techniques developed for this purpose such as the electrical
downhole
means.
This chapter briefly review the various techniques and devices employed in the field to initiate
ignition in the field. The equipment details and operation procedure for the gas fired burners and electrical heaters, the two widely used ignition devices in fireflood operation is also presented. For additional information, the reader is directed to an in-depth review report on ignition by Shallcross (1989).
This DOE publication review the various ignition techniques and devices appeared in the patent liter:
ature including their unique features, limitation and reliability.
325
Spontaneous
Ignition
Crude oils oxidizes (low temperature oxidation reaction) when exposed to air and releases heat.
If the heat release rate is rapid enough to overcome heat losses, ignition can occur spontaneously,
without the need for an external source of heat. The time required to elevate the formation to the
crudes ignition temperature depend upon the crude reactivity and initial reservoir temperature. In
general, higher the initial reservoir temperature lesser is the time required to reach the ignition temperature. The oils that are more reactive often generate larger quantities of heat and thus can heat their
surroundings more rapidly and reduce the time required to achieve self ignition. Since heat loss rates
to the over and underburden zones are greater in thinner than in thicker reservoirs spontaneous ignitions are more likely in deeper thicker reservoirs containing moderately reactive crudes.
An estimate of the time required (measured from the start of air injection) to achieve spontaneous ignition can be made from the following modified expression (Prats, 1982) of Tadema and Weijdema (1979), if it can be assumed that all the heat of reaction is consumed in heating the formation
and the heat losses to the surrounding are negligible.
T, [ 1 + 2T
++-)
,-.
(8.1)
where:
~w = ignition time, day
(pC)* = equivalent heat capacity per unit volume of the porous medium, Btu ft.- F-
TO= initial reservoir temperature, R (F +460)
R = universal gas constant, (1 .9869 Btu/lb-mole R)
E = activation energy, Btuflb-mole
@ = porosity, fraction
So= oil saturation, fraction
pO= oil density, lbrn/ft.
326
AH= heat of low temperature oxidation of the crude, Btu/scf of oxygen reacted
(= 100 Btu/scf oxygen)
A = oils low temperature oxidation reaction pre exponential factor, see-l atm-n
Poz = partial pressure of oxygen, atm
n = reaction order with respect to oxygen
From the above expression it is readily apparent that the time necessary to achieve ignition
decreases as the reservoir temperature (TO) or crude reactivity increases (smaller E or larger A).
The effect of increasing oxygen partial pressure (higher reservoir pressure) is also to shorten the
autoignition time.
Equation (8. 1) is very sensitive to temperature. Even an error of lF in the estimation of reservoir temperature would result in an error of several percent in the estimated ignition time as indicated
by the following hypothetical example.
Example:
Estimate the spontaneous ignition time assuming the reservoir temperature to be 90 and 91F
and the following data.
Air injection pressure
350 psig
110 Btu/Scf
Heat of oxidation
Porosity
0.34
Oil saturation
0.62
Oil density
59.8 lb#ft.3
39 Btu/ft. F
0.6
E/R
16,000R
Case 1: t = 90F
327
14.7)/14.7=
5.21 atm,
(4.31 x 102)}
10)(3090)(16000)(5.21)0]
=60 days
similarly for t = 91 F
TO= 91 + 460= 561R
ZW = { (2.04
10-7)(39)(561)2
+ [(0.34)(0.62)(59.8)(1
[1 + 2(561)/16,000:
10)(3090)(16000)(5
(4.084
1012)}
.21)O]
=57 days
Thus a 1F difference in the reservoir temperature result in 5% error in the estimated ignition
time.
Since Equation (8.1) neglects the heat losses, the self ignition time computed using the above
expression is always shorter than would be expected in the field. However, the ignition time computed
using this simple expression does give an indication whether it is feasible to achieve autoignition in a
reasonable amount of time. If, for example, the calculated autoignition time is only a few hours or
days, than spontaneous ifition
is economically
a calculated ignition time of months or years points to the need for techniques that can lower the ignition time greatly. A knowledge
voir rock is useful in ascertaining whether it is feasible to autoignite the oil in reasonable time.
Burger (1976), using a numerical simulator, investigated the effect of heat transfer (in the direction of flow) and flow geometry on ignition time and concluded that unless initial reservoir temperature is very low, Equation (8.1) yield an acceptable value for ignition time, Burgers calculated
ignition times are only a few percent greater than those given by Equation (8.1) except for low temperature reservoirs, where the differences are as much as 100%.
32s
Spontaneous ignition has been utilized to initiate fireflood in many fields. Spontaneous ignition
can often be achieved with in a reasonable time in formations deeper than 4,000 ft. due to high reservoir pressure and temperatures. For example, in the Medicine Pole Hills Unit, N. Dakota and the Buffalo Red River Unit, S. Dakota in-situ combustion projects, ignition occurred in as little as five hours
due to high initial reservoir temperature (230F) (Miller, 1995). It also appears that in many shallow
California reservoirs autoignition can be achieved in a matter of weeks due to high reservoir temperature (s 100F) and presence of fairly reactivity crude.
Though, the initial oxidation rate and consequent heat generation rate are highest in the immediate
vicinity of injector, autoignition does not occurs near the well. This is because the injected air
transfers heat away from the well, and consequently ignition occurs at a point some distance away
from the injector. This distance increases with the air flow rate. Since the region between the ignition
point and the injector is saturated with oil and oxygen, after ignition the burning front moves backwards toward the injector, prior to moving in the direction of air flow. During this backburning period
temperature around the injector can rise to a very high value and damage the well casing if the free oil
entering the hole (backflow)
of
cemented completions. Also, by maintaining a high enough air injection rate and by ensuring that no
interruption of air flow occurs, backflow of reservoir fluids in to the wellbore can avoided, thus minimizing the chances of wellbore damage.
A rapid auto-oxidation,
of low temperature oxidation products. Low temperature oxidation tend to promote the formation of
unstable peroxides and buildup of coke near the wellbore. When autoignition does occur, the backburn can ignite the coke and the resulting high temperature can darnage the downhole tubulars.
Artificial
Ignition
As noted previously autoignition cannot be achieved in reasonable time in low temperature reservoirs or in reservoirs containing low reactivky crudes. Ignition in such reservoirs can be achieved
by heating the oil-bearing formation to the oils ignition temperature. Several artificial ignition techniques are available for this purpose. These include: downhole electric heaters, gas-fired downhole
burners, catalytic heaters, hot fluid injection technique, chemical methods, and fluid pack methods.
329
The electrical heater, gas burner and hot fluid (steam) injection are the most widely used artificial ignition methods and are described in detail in the following
techniques are only briefly dealt with. For a more completed discussion of these lesser used techniques the reader is referred to the DOEs report on in-situ combustion ignition devices (Shellcross,
1989).
The amount of heat needed to initiate ignition depends upon the heating rate and on the ignition
temperature of the crude. Though these conditions vary considerably from one field to another rough
estimates can be made using empirical correlations. As a first approximation, the energy which must
be supplied to rise the temperature of unit thickness of the formation (Q/h) at a distance r from the
well from the reservoir temperature T. to imition temperature T,- is given by (Burner, 1985):
I
(8.2)
where:
r = heated radius, ft.
rW= wellbore radius, ft.
(p C)r = volumetric heat capacity of the formation, Btu ft; F-I
T,= initial reservoir temperature, F
330
Gas Fired
Burners
Historically gas fired burners are the most widely used device to initiate ignition in a fireflood
project. The heat required to ignite the formation is generated downhole by burning a combustible
mixture of air and hydrocarbon gas such as the natural gas or propane in a specially designed combustion chamber (burner) housed inside a heat shield. The heat shield is a stainless steel cylindrical vessel
(pipe) lined with a refractory material and serve to protect the casing and downhole tubulars goods
from direct exposure to the high combustion temperature.
The principal advantages of employing gas burners in combustion projects are:
1.
2.
3.
It has been used successfully to ignite formations at depths as shallow as 100 ft. and as deep as
6,200 ft. and pressures to 3,500 psi. Reported air injection rates during ignition vary an order of magnitude from 100 Mscf/D to 960 Mscf/D.
A downhole burner system requires an independent means of introducing fuel and air to the bottom of the hole for combustion, means to initiate the fire at the burner, and surface metering and control equipment for the fuel and air streams. Stoichiometric
combustion temperatures (of the order of 1200-2000F),
formation and are destructive to downhole equipment and materials. For these reasons, a secondary
supply of air is utilized in burner assembly to reduce effluent temperature and to minimize possible
well darnage.
A variety of downhole
burner system have been patented in the U.S. since the early 1950s
(Shallcross, 1989) and a few are described in the petroleum engineering literature (DePriester et al.,
1963; Brandt et al., 1965; Smith, 1967; Crevolin, 1984). These designs differs in mechanical details,
configuration utilized to supply fuel gas and air to the combustion chamber, the method employed to
ignite the air-fuel mixture and the technique followed to reduce the effluent temperature.
One design (Smith, 1967) employ a single tube string other than the casing to supply fuel and
air to the burner. In this arrangement, the fuel is supplied down the inner string, and air is forced down
the annular area. The air is split into two streams (primary and secondary)
burner chamber. With this arrangement, the ratio of primary to secondmy air is fixed once the burner
331
is run into the hole. The effluent gas temperature is reduced by mixing with the secondary air prior to
injection into the formation
In an alternate configuration (DePriester et al., 1963; Brandt, 1965), the fuel gas and the air are
introduce into the combustion
down the casing tubing anmdus. While this arrangement permit the easy adjustment of the primary
air-fuel ratio and the ratio of primary air to secondary air at the surface, they are more cumbersome to
work with due to the difficulty of simultaneous lowering and raising of two parallel strings within the
well.
As pointed out previously, the combustion of air-fuel mixture often result in temperatures much
higher than that needed for ignition hnd must be cooled prior to transferring the heat to the formation.
This is accomplished
gases are
cooled by transferring some of the heat of combustion to a secondary stream of air flowing around
outside of the combustion
gases are
transfer the heat of combustion indirectly to the secondary air, the exhaust gases are vented to the surface (Brandt, 1965). The relatively cool air-effluent gas mix or the heated air is then discharged over a,
fairly limited section of the pay to initiate and sustain the combustion.
In a typical downhole installation system, a tubing string with a pump seat at the bottom is first
lowered into the cased ignition well. The seat is positioned few feet above the top of the perforation,
so that it is away from the heat generated by combustion within the formation. The burner is then lowered inside the tubing string on a second, smaller tubing, until it rests on the seat. The fuel gas is then
supplied down the inner tubing, while the air is forced down the anmdus between the two strings. A
number of small conduits in the burner just above the combustion chamber allow the air to flow into
the ignition chamber and mix with the fuel gas. The ah-fuel mixture is ignited by some means such as
by the use of a sparkplug or chemical compound. The secondary air that is injected down the casing
anmdus serves a number of useful roles. Firstly, the air cools the burner unit as it flows around it and
helps to control the temperature within the well. More importantly, it is this supply of air that is used
to oxidize the oil within the formation. Finally, a constant flow of air into the casing annulus prevents
the hot combustion gases from flowing back up the well instead of into the formation. Most burners in
use today incorporate thermocouples
332
Oil-bearing formation often do not ignite on the first attempt and must be reignited. Further, for
a variety of reasons, the combustion front dies before it is well established. Hence it is important that
the downhole burner system be equipped with a reusable ignition device so that it can be reignited if
necessary without the need for having to pull the burner from the hole.
Gas fired burners are ignited using an electrical ignition device or by chemical means.
Electrical igniters include spark plugs, glow plugs and resistance heaters. Several igniter designs
have appeared in the patent literature over the years and are described by ShallCross (1989). Prior to
1965 electrical igniters were built into the burner as an integral component and hence are not reusable
due to damage incurred during the ignition of the gases. To overcome this drawback, the later igniters
were designed to operate as a stand alone battery or generator powered unit. The unit is lowered to the
burner on a wireline within the fuel gas tubing string. When in place, the spark gap electrodes or the
resistance heaters protrude into the burners combustion chamber where they may ignite the gases.
To ignite the burner, the igniter is lowered into place, guided by the location stops. When the
igniter is in place, tension on the wireline and the weight of the batteries bears down on a pressure
switch located in the mid-section of the unit. The shifted weight closes the electrical circuit and the
electrodes spark across the gap. Once ignition is achieved, the igniter is withdrawn from the burner. In
an alternate design, the igniter electric circuit incorporate a bimetallic element that permit the electrical igniter to operate only when it is cool. Thus, once the fuel gas has been ignited and the igniter is
hot, the spark plug or resistance wire will not operate. Thus, should the burner flame is put out, this
design permit the automatic reignition by the igniter once the bimetallic element has cooled sufficiently.
A major disadvantage of using electrical ignition system is that its heat generating capacity is
limited and as such it may fail to ignite the gas if the amount of gas reaching the igniter is excessive.
Further, since power losses increases with depth, electrical igniters cannot be employed in deeper
wells. Also operational problems such as the electrical shorts, poor connections, heater element burnoutetc. can preclude the satisfactory functioning of these igniters.
Burner gases can also be ignited chemically using pyrophoric
chemicals.such
as the calcium
phosphide, triethylborane (TEB), silane and disilane. Linseed oil containing catalysts such as cobalt
naphtheneate and dimethyl aniline have also been used as igniter. The chemical ignition system is
extremely easy to use and is the preferred choice of many operators. Most operators prefer TEB over
333
other pyrophoric chemicals because it does not react with water or hydrocarbons,
Description
and Operation
of a Popular
Gas Fired
Ignition
System
Ignition operation is a complex and non-trivial task and certain procedures must be followed to
ignite the formation successfully. A variety of gas fired ignition system is employed in the oil field
and the operating procedures vary the system. Nevertheless, with all the system the injection well
must satisfy the following conditions prior to the installation of the burner:
The well must be in good mechanical condition (no casing leaks; wellhead and casing have
satisfacto~ pressure rating, etc.)
The tubing must have been drifted and found free of obstructions.
Both the well casing and tubing must be thoroughly cleaned and free of any hydrocarbon
residue.
A schematic of a popular downhole burner ignition system is shown in Figure 9.1. This system
developed
widely under licence by the petroleum industry includes surface equipment for installing the burner
and for supplying metered quantities of air and fuel, the thermocouple-fuel
assembly.
334
FUEL LINE
Y-BLOCK
THERMOCOUPLE
SINKER BAR
LIN
BLOWOUT
PREVENTERS
0-,
IF
1/
BURNER ASSEMBLY
\/
The basic burner assembly (Figure 8.1) consists of a burner and burner stop, fuel check valve, Yblock, and thermocouple.
assembly is designed for use inside tubing with a minimum internal diameter of 2 and 3/8 inches.
The basic surface equipment also shown in Figure 8.1 includes a simple wellhead lubricator
hookup and a mast and hoist for installing the burner. In addition the surface facility include air and
fuel supply, metering facilities, and temperature recording and safety control instruments.
An air compressor with sufficient capacity (usually in the range of 0.5-4 MMScf/d)
and capable
of uninterrupted operations at pressures up to 1 psi per foot of depth is utilized to supply air. The air is
supplied to the tubing and casing through a parallel set of control valves and orifice meter runs,
thereby permitting the control and measurement of the air flow into the tubing and casing.
335
A metering pump is utilized to supply metered quantity of fuel to the burner. Although the
burner can burn any liquid paraffin hydrocarbon more volatile than, gasoline, propane is generally the
preferred fuel owing to its wide availability and low cost.
A downhole temperature limit switch is utilized to ensure safe operation of the burner. This
switch automatically stops the fuel injection to the burner in the event an excessive temperature develops owing to air supply failure, wellbore fire, etc.
Additional information
who
described the system in considerable detail. Here we describe the step by step procedure for the installation and operation of the burner.
1.
Install and pressure-test all injection lines (at 11/2 times the operating pressure).
2,
3.
4.
5.
Assemble the lubricators with the blowout preventer open and line wiper packing removed.
The lubricator is rested on the ground.
6.
Assemble the burner with an appropriate sinker bar weight and thermocouple.
7.
String the fuel tubing with Y-block connections to the burner assembly.
8.
Insert the burner assembly through the line wiper and into the lubricator.
9.
Fuel line is filled with propane and the check valve operation tested. The fuel
ine is prei-
tubing is permitted during the installation (i.e., the assembly should not be raised by the
fuel tubing).
1. Raise the burner assembly from the top of the gate valve by one foot.
2. Assemble the line wiper packing hand-tight.
3. All tubing head valves are opened slowly.
14. Remove all the handles.
15. Lower the burner assembly inside the well tubing to the burner seat,
336
16. The blowout preventer tid line wiper are firmly tightened.
17. Recheck fuel line pressure and check valve operation. The pressure should approximate the
casing pressure, plus 350 psig, less the weight of the column of propane in the fuel tubing.
18. Recheck temperature recorder and air flow meter readout. The recorder should read bottomhole temperature.
19. Discontinue fuel pumping.
20. The fuel metering pump is adjusted to the proper rate to produce a predetermined burner
exhaust temperature.
21. Install TEB container on the fuel bypass manifold.
22. All fuel lines leading to the TEB container are purged with propane before tightening.
23. Activate the fuel pump to raise the fuel line pressure to the proper value.
24. Open the upper valve of the TEB container slowly.
25. When the pressure .of the TEB container reaches the fuel tubing pressure, open the lower
valve of the TEB container. The fuel bypass valve is then closed.
26. men
the TEB enters the air stream, an abrupt temperature increase will be noted on the
temperature recorder.
27. Burner operation is usually carried out for a period of 35 days to ignite most formations.
337
Electrical
Ignition
Many operators prefer electrical igniters over gas fired igniters due to their relatively simple
operations and low cost, The major advantages of electrical igniters include: (1) ease of operation; (2)
excellent temperature and heat input rate control; (3) flexibility in changing heater location to permit
the exposure of all parts of the formation to the ignition temperature; and (4) flexibility over the duration of heater operation to ensure formation ignition (if feasible). Further unlike the gas burner (see
section 8.3.1 1) the electrical ignition system does not require elaborate surface and downhole facilities and tools to install the igniter downhole and initiate ignition.
lllNlllON?!111E(!IIIPUENT
5&HMAllt
//--.,
TEMPERATURES
AIR RATE
RECOROER
41R
ONE
INCH
THEQMOWELL
THERMOCOUPLE
TOP
OF
FORMATfON
ELECTRIC
HE&TER
CO M@UST(ON
BASE
OF P4Y ZONE
COMBUSTION
PERFORATIONS
BURNED
GASES
ZONE
ZONE
FIGURE 8.2 Schematic of an Electrical Ignition System for Fireflood Injection Well
338
While electrical igniters are safer and easier to control than downhole burners they cannot be
deployed in deeper wells owing to excessive line power losses. A significant portion of the power supplied at the wellhead can be dissipated as heat within the cable due to the cables electrical resistance.
The amount of electrical power lost depends upon the cable material chosen, the cable diameter, the
depth of operation, the line voltage, and whether the power is supplied by single-or three-phase transmission.
Basically, the electrical ignition systems consist of
A high-pressure lubricator.
An air-injection system,
Power supply for the ignition system generally is rated for 220, 440, or 880 volts AC. Electric
power can be supplied by portable generator units or from normal field power sources through suitable transformers. Downhole
ignition systems are designed for a specific voltage and power phase.
Load requirements range between 20 and 40 kw. The power supply must not only be compatible with
the heater design, but also it must be capable of the maximum load delivery. Schematic of an electrical ignition system for Iireflood is shown in Figure 8.2.
Two general types of electrical downhole heaters are in use for ignition services. In the simpler
and most common design, a series of nichrome
stainless steel case constitute the heater. This heater is lowered into the well on an armored cable. The
cable not only bears the weight of the heater, but also supplies the electric power and may carry signals from any thermocouples
at
1500F. The electrical resistance wires inside the housing is insulated with magnesium oxide. Air is
injected down the well and the heat is irnptied to the air as it passes around and along the electric elements. In the second design, the heater is made up of a number of commercial
alloy steel-sheathed
heating elements connected to a suitable head arrangement and mounted on the end of a tubing string.
The end of the string incorporates baffles and jackets so that air must pass over the heated section several times before passing into the well. Often times the combination of conductors, electrical insula-
339
tion, and armor makes the size of the cable rather large and makes it difficult to lower the heater into
the well under high pressure.
Proper operation of the igniter is indicated by current and voltage measurements. Heater temperature is controlled by varying the power to the heater either by adjusting the voltage, or by a
simple on-off control of the heating cycle. Temperatures may also be controlled by proper adjustment
of the air-injection rate.
Hot-Fluid
Injection
and Chemical
Ignition
In shallow formations containing sufficiently reactive crude, ignition can be achieved by injecting surface-heated fluids such as steam. This technique is more applicable to cases where the oil is
sufficiently reactive but low reservoir temperatures prolong the ignition time to such an extent that
autoignition is not achieved. Autoignition can be achieved in reasonable time by raising the formation
temperature by hot-fluid injection. Once a significant volume of the formation has been heated to
ignition temperature, the injection of heated fluid is terminated and normal air injection is begun.
The formation temperature can also be elevated above the ignition temperature by inducing and
controlling an exothermic reaction within the wellbore or formation proper. Difficulties involved in
chemical ignition include hazards associated with the handling of chemicals and the problem
of
gas
(such as air) is then injected into the formation, and the autoignitable fluid (the tung oil or linseed oil)
ignites. The heat generated within the formation is sufficient to ignite the crude. The purpose of the
injection of the second fluid is to push away from the wellbore the highly reactive autoignitable fluid
so that the injection well is not subjected to the severe temperature. To accomplish this, the second
fluid must be less reactive and miscible with the autoignitable fluid. Light crude oil, kerosene, or diesel oil can all be used as the displacing fluid.
340
Detecting
Ignition
Since the cost of injecting air is substantial, it is important to ascertain early during the operation whether ignition has been achieved. This knowledge is important because it can prevent unnecessary incurring of expenses for igniter operation if the formation fails to ignite in a reasonable time.
In artificial ignition methods, temperature measurements and gas analyses are useful in detecting ignition.
In electrical ignition,
thermocouples
being operated opposite the formation face, a rapid rise in temperature at constant heat input rate is
often construed as evidence of ignition. Gas analyses at surrounding producers also can provide an
indication of ignition. In many @stances, gas breakthrough to offset producers occurs within a matter
of days after initiation of air injection. Initially, the oxygen content of the produced gas streams may
be higher but subsequently drops, indicating ignition. The produced gas may also contain CO and
341
REFERENCES
1.
Brandt, H.W., G. Poynter, and J.D. Hurnmell, (1965), Stimulating Heavy Oil Reservoirs
with Down-Hole Air-Gas Burners, World Oil, (September), pp. 9195.
2.
Burger, J.G. (1976), Spontaneous Ignition in Oil Reservoirs, Sot. Pet. Engr. J, Vol. 16, (JanUary), pp. 7381.
3.
5.
DePriester, C.L. and A.J. Pantaleo (1963), Well Stimulation by Downhole Gas-Air Burner,
J. Pet. Tech. (December), pp. 12971302
6.
Miller, R.J., (1994), Kochs Experience with Deep In-Situ Combustion in Williston Basin,
Keynote Address presented at the 1994 U.S. DOE/NIPER
tion, April 2122, Tulsa Oklahoma, U.S. Dept. of Energy Report NIPER/BDM-OK-O086Conf-940450
7.
Prats, M. (1982), Thermal Recovery, SPE Monograph, Volume 7, Sot. of Pet. Engr., Dallas,
Texas, pp. 128132.
8.
ShallCross, D.C. (1989), Devices and Methods for In-Situ Combustion Ignition, U.S. Dept.
of Energy Report DOE/Bc/14126-12
9.
(DE89000766),
(October) p. 34.
(March), pp.79-82
10. Tadema, H.J. and J. Weijdema (1970), Spontaneous Ignition of Oils, Oil and Gas J, Vol. 68,
No. 50, (December 14), pp. 7780
342
CHAPTER
COMBUSTION
9 IN-SITU
AND PRODUCTION
WELL DESIGN,
PRACTICES
COMPLETION,
Introduction
Since potential explosive situation can develop rapidly in in-situ combustion projects due to the
presence of hydrocarbons in compressors (lubricants), air injection lines, air injector, etc., safety is of
prime concern in fireflood operations. Design and engineering of fireflood surface and subsurface system are more involved and require careful planning and monitoring.
In this chapter, the fireflood injection and production well design and completion procedures are
detailed. Fireflood operational problems and project monitoring techniques are also discussed.
Well Completion
Practices
The design of fireflood well completions requires special consideration of several troublesome
factors that are amplified by the fireflood process, which include high temperature, corrosion environment and proper selection of corrosion-resistant materials, preparation of the injection well for safe
injection of air by assuring that the surface equipment and downhole
hydrocarbons, and sand and clay control. Figures 9.1 and 9.2 show schematics of typical injection and
production wells, respectively.
343
T7
/
FULL PORT
BALL VALVES
BALL VALVE
F
~;yz::%r;
l-x4-
344
CHECK VALVE
COMBUSTION
VENT GAS
IF
+
x+
89-LB FLOWLINE
\
BALL VALVE
I
CONT~OLVALVE
II
PLUG VALVE
+)?
PLUG
345
aoyo
1.
2.
type,
4.
All tubing, work string and injection string should be cleaned and hydrocarbon-free,
5.
to unconsolidated
sand as
fluid is produced can be and generally is a production problem in most fireflood producers. The sand
production mechanism can be extremely complex
operations
ranging from first bit penetration to the start of production or injection. The problems of sand control
in thermal recovery wells are especially troublesome because of high production rates of reservoir fluids as a result of the high volume of fluid injected in these processes. The greatest single problem of
any enhanced recovery process is that of making lasting, high-productivity
well completions
which
346
Drilling
Successful preparation of a well for sand control requires that each well completion procedure
be designed and executed properly. Any failures in drilling, underreaming, casing, cementing, perforating, or the use of completion or workover fluids could be detrimental to well productivity.
Drilling
Fluids
Both productivity and infectivity can be adversely affected by washouts through the production
interval and by reduction in permeability due to fresh water invasion into the formation which causes
swelling of water sensitive clays. Drilling fluids, therefore, have to be inhibited against clay hydration, generally with 110% brine or KC1 solutions and/or other clay stabilizing compounds,
such as
hydroxy-aluminum.
Cementing
Poor cementing techniques could cause channeling and may contribute to excessive water or gas
production in certain wells. Formation damage from cement particle invasion into the production
zone could be a severe problem in certain enhanced recovery wells. For this reason open hole completions or a cement puddling technique designed to minimize column height or head while cementing
casing through the production
through and cemented with high silica flour cement. For an open hole completion operation, the well
is drilled to gauge to the top of the producing formation and then cemented in accordance with the
best available cementing practices; i.e., the cement should possess low-fluid loss, and the casing
string should be centralized, equipped with scratchers, and rotated or reciprocated during the cementing operation. A high-silica flour cement is recommended
ture.
During the cement puddling
operation, cement with proper water loss and setting time addi-
tives is spotted in open hole through the production interval while pumping through the drill string
equipped with scratchers. The drill string is rotated and withdrawn S1OW1
y through the production
zone during the cementing operation to remove, or scratch mud cake from the side of the wellbore
and to prevent the formation of pockets of mud in the cement. After the cement has been puddled in,
the drill string is removed from the well; centralized casing is run to the top of the cement pool and
then slowly through the cement to the bottom of the well; and the cement is then permitted to set prior
to perforating.
347
Perforating
Jet perforating is a piercing and pulverizing process that creates a flow path through the casing,
cement sheath, and into the formation. However, cement, cooper, lead and carbon from the jet charge
can penetrate and mix with crushed formation material, sometimes causing significant perforation
darnage and restricted productivity and infectivity.
Perforations should be gravel packed with properly sized sand to prevent the tunnel from
being filled with formation fines when the well is placed on production. If gravel is not sized properly,
sand will enter the perforated area, become trapped in the gravel, and cause a prohibitively high pressure drop across the perforations.
Completion fluids across the interval at the time of casing perforating should be free of solids
because they can be driven into the perforation by the force of the charge. Perforation debris removal
methods include
underbalance,
backflow,
perforation
washing,
Underbalance, backflow, and perforation washing, for the most part, are effective remedial measures,
but are not always 100% successful. Acid stimulation under certain conditions could be responsible
for reduced permeability and compressive strength of the formation, sometimes resulting in abnormal
production
excessive sand removal and the uncertainty that all perforations have been properly cleaned.
Well Completion
Colloids,
and Workover
Fluids
consist of calcium lignosulfonate, hydroxyethyl cellulose and other polymers. In addition, special fluids, such as oil base, inverted emulsion and chalk emulsion systems, have been used successfully to
protect sensitive clay bearing formations from fresh water. The chalk emulsion is actually a water-inoil emulsion containing acid soluble bridging and weighting materials which have been stabilized by
finely divided acid soluble chalk solids.
Stable performed foam has replaced mud, oil, and water as a circulating medium in well completion and remedial operations in many shallow, low pressure reservoirs of California. The foam
with a density as low as 0.3 ppg is performed at the surface to a consistency of aerosol shaving cream
by blending detergent, fresh or salt water, compressed air or gas, and sometimes other special additives such as surfactants. The foam has been used in California wells to core the completion interval,
348
stabilize severe sanding and shale sloughing problems, retrieve open hole liners, recomplete old, low
pressure wells by drilling in new liners, and recover sanded-up equipment and cement liners in lost
circulation zones. Advantages include low hydrostatic head, elimination of formation darnage, high
carrying capacity for cuttings removal, low circulation pressures, and simple installation.
Open Hole
Completion
The main advantage of open hole completion is the increased productivity over that of perforated completions. For example, theoretically it would require four perfect l/2-in. diameter holes per
foot, penetrating the casing, cement sheath, and 8 in. of the formation to equal open hole productivity.
In addition, these openings in unconsolidated
high-temperature, low pH fireflood producers incorporating cased hole completions wiI also be discussed.
Screens
Slotted liners, wire wrapped screens and pre-packs, have been used effectively in certain areas
to control sand production. However, it is not recommended
own pack around a screen placed across the producing interval. This procedure will generally result in
a disturbance of the formation, possibly causing excessive fine sand production or plugging before the
screen or pack stabilizes. All screens, therefore, should be used in conjunction with gravel packs to
retain gravel packing material which, in turn, supports the formation sand.
The all-welded wrapped-on-pipe
screen has the highest flow capacity. Wire and pipe are nor-
Incoloy are available, such as Iconel 625 and Incoly 825. Some laboratory corrosion tests have shown
that Inconel 625 has a better high temperature corrosion and oxidation resistance than Incoloy 825. A
schematic of a recommended completion for a fireflood production well is shown in Figure 9.2
A slot or wire width opening of two times the 10 percentile diameter of the formation sand has
been proved effective in certain areas containing sand with broad size distribution. In areas with relatively uniform sand size, slots equal to the 10-15 percentile formation sand diameter have been used
349
under that of
parallel sided slots in which particles 1/2 the slot width tend to wedge and plug.
Ike-packs are designed to bridge fine formation sands without the need for gravel packing. They
are fabricated with an outer cylinder of either unconsolidated or resin-consolidated
ported by a slotted mandrel. Since pre-packs necessarily must be somewhat smaller than the wellbore
into which they are positioned, the formation may subsequently shift and cause productivity impairment.
Pre-packs may plug when:
1.
2.
3.
4.
advantages:
1.
2.
3.
4.
5.
6.
darcys).
To reduce the chances of formation movement around the pre-pack and the risk of plugging with
formation fines or clay, this device should be used in conjunction
350
Open
Hole
Gravel
Packing
Open hole gravel packs should provide higher productivity than inside gravel packs because
restrictive casing perforations are eliminated and the underreamed wellbore improves radial flow into
the well. Open hole gravel packing in stratas of clay and shale are preferred over other completion
techniques since the gravel should prevent migration of the shale or clay to the screen and sealing it off.
6-1 2 ~
PROPERLY SIZED GRAVEL
PACK (KARPOFFS RULE)
SPECIAL ALLOYS
HASTELLOY C-276
INCONEL 625, MP35N
NON-DAMAGING
COMPLETION
FLUID
FIGURE 9.3 Schematic of Open Hole Gravel Packing for Sand Control in Producer
Gravel packing (Figure 9.3) consists of underreaming the wellbore 6-12 inches larger than the
casing diameteq running and hanging the wire wrapped liners; packing with gravel using a viscous,
high gravel concentriited (15 lb/gal) slurry at low circulation rates to prevent intermixing of the sandpack with formation sand washing to remove gravel bridges; and then repacking, if necessary. During
the foregoing operations, a nondamaging completion fluid should be used as suggested earlier. Also,
pre-flushes of clay stabilizing material should be used prior to both underreaming and gravel packing
and will be discussed later. In addition, the hole should be enlarged slowly to prevent corkscrewing, a.
caliper log should always be run in the underreamed hole to determine if sloughing has occurred, and
debris should be removed from the well prior to running the liner.
The ratio of gravel size to formation sand size should be estimated from Karpoffs rule which
states that gravel size should range from 51 Otimes the 50 percentile or median formation sand diameter. Sand movement should be properly restrained under these conditions since the gravel pore openings, theoretically, are properly sized relative to the sand particle diameter.
351
Effectiveness
of gravel packs c~
tracer methods,
Gravel packs in open hole injection wells sometimes tend to slump soon after the start of injection,
and logging may indicate that repacking is required.
Consolidated
Pack
The consolidated pack (Figure 9.4) is a slurry consisting of a carrier fluid, resin (or other type of
bonding agent), coupling
agent, sand, and catalyst, which is blended at the surface and pumped
behind a liner or in open hole where the resin or bonding agent is permitted to cure. This process is
designed to achieve gravel packing and sand consolidation in a single injection step. If a liner is used,
any consolidated material remaining in the wellbore is simply drilled out. In most open hole completions, no liner is used and the consolidated pack is redrilled through the entire producing zone with a
regular bit. All the same rules and precautions associated with conventional open hole gravel packing
and drilling apply to this particular technique.
FIGURE 9.4 Sand Control (Slotted Liners and Wire Wrapped Screens)
352
of the sand is
assured in all areas where the sand laden slurry can be pressure packed. The pack possesses excellent
strength (5,000 psi) and permeability (60 darcys) and is stable in most well environments up to 500F,
The process works well with clay bearing formations, and excess resin can be used to consolidate a
portion of the formation behind the pack to stabilize both sand and clay when a special clay stabilizing compound, such as Halliburtons Claylok or Claysta, is injected ahead of the resin.
Cased
Cased hole completions
Hole
in fireflooding
Completion
completion cannot be satisfactorily performed, or is not desired for some reason, then setting the casing with the best cementing technology along with the best well completion, perforating, gravel packing and/or chemical sand consolidation methods available should be implemented.
In unconsolidated
as previously discussed, either with a sandpack or with chemical or artificial consolidate. Another
technique is to gravel pack a wire wrapped liner inside and across the perforated interval of the long
string of casing. One such method of chemical consolidation designed for high temperature, low pH
atmosphere is discussed below.
Solder
Glass
Sand
Consolidate
Treatment
The solder glass method for sand consolidation is patented by Gulf (new Chevron) and uses a
devitrified glass to bind together sand particles.
The consolidated
mass possesses
high internal
strength and resistance to hot gases, hot water, steam, acidic environment, and high temperature. The
consolidated formation also retains permeability and compressive strength when subjected to a variety of well production and stimulation conditions. For a fireflood or steamflood reservoir with sand
production problems, it is a promising technique to control sand production.
In addition to sand control, the solder glass method of sand consolidation promotes a certain
amount of production stimulation which is unrelated to oil viscosity. Increased production can be
achieved through a decrease of wellbore darnage caused by clay swelling and emulsion blocking. Formation permeability in dirty sands is increased by shrinkage and permanent dehydration of clays during a pre-bum period prior to the sand consolidation treatment.
353
An eight-hour period of air injection at 1200F to burn the sand clean and to shrink and
clays present.
2.
3.
The well selected for this treatment should be one in which the zone to be treated is +50 feet
from a gas cap or aquifer and has an initial gas saturation greater than 10%. To maximize control over
the process, a single point entry is mandatory in the well to be treated. Although not a preferred
method if the selected well has been previously completed,
existing perforations with a low-water loss, heat-resistant
it is necessary
to squeeze-cement
the
with a single perforation. If the well has produced a large amount of sand, it may also be desirable to
repack the formation with clean sand.
If more than one perforation
treated individually; i.e., start from the bottom of the production zone and back-fill the well with sand
to a depth of at least 2 feet above each previously treated perforation before perforating and treating in
a higher plane.
The air rate to be used is determined by the flow rate of air required to sufficiently cool the
burner but not to exceed a flow to cause flame-out. Optimum air requirements for any given bottomhole pressure are based on a burner slot actual velocity of 14 ft./see. The downhole burner used in this
treatment was discussed in Chapter 8. A tentative program for the solder glass sand consolidation
method follows:
1.
Install suitable 3,000 psi W.P. wellhead on casing. Pressure test casing to 3,000 psi with
fresh water.
2.
detergent solution, until returns are clean. Displace cleaning solution with fresh water, or
4% KCI solution, or clay-stabilizing
3.
solution, if required.
Pull tubing and perforate casing with one hole using a suitable jet shot (5/83/4
354
4.
Run 2-7/8 in. O.D. EUE tubing with a burner stop on bottom to a depth 5 ft. above the perforation. Install Christmas tree.
5,
Install and pressure-test to 3,000 psi all injection lines and Christmas tree. Install the mast,
lubricator burner assembly, fuel tubing hoist, and the fuel injection system.
6.
Unload water from hole to a tank or pit by circulating air down the casing and up the tubing. Maintain surface casing pressure 200 psi above the reservoir pressure. When all fluid
has been displaced from annulus, the remainder of fluid in tubing may be displaced into
formation.
7.
When essentially all of the water has been displaced from the tubing, continue injecting air
down tubing and casing. Increase the injection pressure gradually until the desired injection
rates are obtained in both tubing and casing. Casing rate should be only 10% of total injection rate.
8.
Lower the downhole burner to the seat. Adjust the tubing and casing air rates and ignite the
burner at a fuel rate such that the temperature of the gas entering the formation is at 1,200F
to clean the sand to be treated.
9.
After eight hours of operation at 1200F, pull burner out of seat and continue cold air injection for one hour. Pull burner out of hole into lubricator.
10. Inject 15 bbl of fresh water, followed by 10 bbl of 0.5~0 caustic solution, down the casing to
further cool the formation while injecting all air into tubing. Then displace fluids into formation with air down both tubing and casing.
11. When the cooling liquids have been displaced into the formation, direct all air into tubing
and inject 5 bbl of cementing chemicals (solder glass). Resume air injection down casing to
displace the cementing material into the formation.
12. Lower the downhole burner to the seat. Adjust the air and fuel rates such that, after burner
ignition, the temperature of the gas entering the formation is 800F.
13. Increase the burner exhaust temperature at a rate of 50F/hr to 1200F.
14. After 8 hours of operation at 1200F, shut off the burner and retrieve the burner assembly
horn the tubing.
15. Continue cold air injection for one hour. Inject 25 bbl of fresh water to cool the wellbore
and formation.
16. Stop air injection.
355
17. Slowly inject water down the casing and bleed the casing and tubing to a tank until there is
no oxygen produced.
18. When the produced gas has no oxygen content, stop water injection. Put well on production
by producing the tubing to a separator.
Clay
Stabilization
Serious reductions in permeability of the formation can occur when clay minerals restrict flow
by either expanding to fill the pore volume or dispersing and lodging in restrictions. The expanding
force acting on clays is caused by the affinity of exchangeable interlayer cations and surfaces for fresh
water. The inherent negative charge on almost all clay minerals is neutralized by adsorption of cations, producing a positive surface charge with a negative inner charge. These clay particles with an
electric double layer repel each other, and, thus disperse.
Chemicals are now available which stabilize many water-sensitive formations. One such compound, a positively charged hydroxy-aluminum
num chloride and sodium hydroxide in a high shear device, is absorbed and polymerized
on the
negatively charged clay mineral surface to form a stabilizing network that prevents further migration.
This chemical, Claylok, a trademark of Chevron, has been successfully used to improve permeability
following acid treatments, stabilize formation fines, desensitize clay bearing formations prior to steam
injection, and reduce formation damage during perforating and gravel packing. Most field brines and
fresh water will not remove the treatment, but mixtures having pH values below 4 or above 9 will
damage its effectiveness.
Halliburton has a process that uses an organic, oil-soluble chemical,
to
shrink water-swollen clays and then surround them with a hard furan-type resin which increases permeability and stabilizes clay in some dirty formations. A new Halliburton product, Claysta, is organic
in nature and stabilizes clay in a similar fashion as Claylok, but has the added advantage that it is stable to subsequent acid or caustic treatments.
It appears that these clay stabilizing compounds have a possible application in open hole as well
as in cased hole completions.
may eliminate or significantly reduce formation damage caused by clay swelling. Incorporating these
inhibitive chemicals into the gravel pack slurry, drilling fluid and/or other completion fluids may also
prove to be advantageous. All possibilities should be evaluated in the laboratory before implementation.
356
Assumptions
1. Well has not been previously steamed
2, Sand production problem
3. Formation sensitive to fresh water
General Procedure
1. Bail out and clean up well.
2. Run injectivhy
*** Injected Claylok should be 0.3 Molar and, therefore, must be diluted 20-fold with 1(ZO
KC1
water. Always add 1 To KCl to concentrate.
357
B. Case 2
Assumptions
1. Well has not been previously steamed.
2. Sand production problems.
3. Formation sensitive to fresh water.
General Procedure
1. Bail out and clean up well, and acidize, if necessary (Steps 14
in Case l).
2. If well is perforated, shoot additional perforations (8 shots per foot) under 2% KC1
solution.
3.
Install gravel pack in hydroxy aluminum while keeping well full with hydroxy
aluminum.
Pack fluid should incorporate hydroxy aluminum, or 2% KC1 solution when possible.
4. If open hole, remove old liner, installnew liner, and gravel pack in hydroxy aluminum.
358
C. Case 3
Assumptions
1. Well has not been previously steamed.
General Procedure
1. Bail out and cleanup well.
2. Cool well with filtered (1 micron) 2% KC1 solution until temperature of formation around
wellbore is less than 200F.*
Temperatures greater than 200F will cause aluminum chloride to precipitate in formation.
359
Screening
Certain preliminary tests should be run in the laboratory to determine which completion technique is most likely to be successful. The following is an outline of proposed tests which should be
run for the purpose of designing the best or optimum well completion:
10
2.
3.
Grain size distribution to be used for the proper selection of slotted liner size and/or conventional or consolidated grain size.
4.
5.
Fluid penetration studies determine depth of fluid penetration in core and resultant damage from conventional mud, extremely low water loss mud, invert emulsion, chalk emulsion, and fluids containing bridging material of calcium carbonate, oil soluble resins, and
HEc.
6.
Degree of clay or fine sand migration in core samples during fluid or steam injection.
7.
Effectiveness
8.
9.
Chemical consolidation and compressive strength of consolidated native cores and compare
with clean sand chemical consolidation.
10. Design and run combustion tube and autoclave corrosion tests* in a more severely corrosive, erosive, oxidizing,
past. In this manner, the best material available for any given fireflood well completion can
be selected along with other auxiliary techniques which may be required to lengthen the
useful life of downhole equipmen~ i.e., downhole cooling, corrosion and oxidation inhibitors, special coatings, etc.
360
Suggested
The following
Drilling
Procedures
procedures is recommended
for air
injection and producing wells. These procedures are meant only to be a guide for the engineer in
designing such well completions.
Am Injection Well (Figure 9.1)
1.
Drill the well with best low-water loss drilling mud using current drilling practices for
that area.
2.
Incoloy Series, on bottom of casing and set the casing with a high-silica flour cement, using
centralizers and scratchers, while rotating or reciprocating
and pockets of mud from annulus during the cementing operation. A synthetic lubricant
should always be used on threads while making up casing.
3.
4.
Install
5.
Run neutron, gamma ray, CCL, and other desired well surveys.
6.
Run open-ended 2-7/8 in. O.D. EUE tubing,* equipped with casing scraper, to plug-back
and reverse-circulate
solution until returns are clean. Displace cleaning solution with fresh water followed by 5~0
Nitrox**
7.
Pull tubing and perforate special alloy casing with one hole using suitable jet shot (3/4-in.
diam, hole and 911 in. penetration) at a depth determined from electric log. To prevent
flowback of sand in unconsolidated regions, perforating should be done with pressure differential from wellbore
into
solution.
8.
Run infectivity and pressure fall-off tests with 0.2% Nitrox solution. lnjectivity should be at
least 1 BPM at pressure less than fracture pressure. Pressure fall-off should be at least 500
psi/rein.
9.
If infectivity testis unsatisfactory, either reperforate at same depth or clean up existing perforation with mud acid. Before reperforating
make certain there is no sand fill. Repeat infectivity and pressure fall-off tests.
361
10. Repeat Steps g to i until adequate infectivity and fall-off are attained.
11. Run 27/8 in. O.D. EUE tubing with burner stop, pup joint(s), crossover, packer, and telescoping union (Figure 9.1 ). Space out and set packer to have burner seat at 5 feet above perforation.
Packer
perforation.
12. Remove BOP and install wellhead.
13. Purge all surface lines and equipment with 5% and 0.2% Nitrox solution. Install and pressure test to 3,000 psi all injection lines and Christmas tree.
14. Close well in and observe pressure decline. A pressure fall-off rate of 4 times the stabilized
time (4x 30 minutes) indicates adequate infectivity with minimum formation darnage.
15. Commence injecting air at 23 MMscfd and continue for 3 days.
16. Install the mast, lubricator, burner assembly, fuel tubing hoist, and fuel injection systems.
17. After 3 days, reduce air injection rate to 1 MMscfd for one day.
18. Lower the downhole burner to the seat. Adjust the air and fuel rates such that after burner
ignition the temperature of the gas entering the formation is 800F.
19. Lower the downhole burner to the seat. Adjust the air and fuel rates such that after burner
ignition the temperature of the gas entering the formation is 800F.
20, Air injection will continue over the life of the project at various increased rates, depending
on the size of the fireflood.
**
362
For liner type completions, drill to the top of the producing formation with standard drilling
mud using current drilling practices for that area. For set through completions, skip Steps
a, c, and e.
2.
Set the casing with standard low-water loss cement, using centralizers and scratchers, while
rotating or reciprocating
the casing to remove mud cake and pockets of mud from armulus
during the cementing operation. The bottom two joints of casing are fabricated from special
alloy material.
3.
Drill through the production sand with the best completion fluid as determined from laboratory screening tests. Clay stabilizing agents should be incorporated into the completion
fluid, if needed.
4.
5.
Underream formation 6-12 in. using best completion fluid to circulate out cuttings.
6.
Run caliper log and other desired logs with best completion fluid.
7.
Run properly designed prepacked liner or wire wrapped peflorated pipe and perform tailored gravel pack or consolidation pack with clay stabilized completion fluid.
8.
Perform properly tailored open hole consolidation pack with clay stabilized completion
fluid. Permit bonding agent to cure sufficiently, and drill through with standard bit, using
best completion fluid. A slotted liner used only to properly guide the bit could be incorporated into this process.
9.
Lower tubing to bottom and reverse circulate clean with 5% KC1 solutiow POH with tubing.
10. Set retrievable bridge plug, install steam quality expansion type casing head, install BOP,
and recover bridge plug. In place of expansion type casing head, a telescoping union or
expansion joint could be run above packer if packer is used. If no packer is used, no expan-
sion joints are required, i.e., downhole cooling may be required during production of hot
well fluids.
11. Run 2-7/8 in. O.Il. tubing, packer (if required), 2 joints special alloy tubing, expansion joint
(if required), and seating nipple.
12. If packer is used, space out to set packer inside special alloy casing at least 20 feet above
pay zone, gravel pack, or consolidated pack. If no packer is used, position bottom of tubing
at top of pay. Tubing should be lined with straight phenolic coating. All materials below
363
packer and including packershould be fabricated from special alloy material. If no packer
is used, the bottom eight joints of tubing should be special alloy,
13. If packer is used, displace anmdus fluid with salt water containing 80-100
lyzed sodium sulfide oxygen scavenger and Tretolite KW12
Preservation
corrosion resistance.
of Hot Production
Wells
During the fireflood process, the combustion zone is propagated through the reservoir by burning a portion of the oil, with the remainder of the oil and connate water moving ahead towards the
producing well. Water and oil are vaporized near the combustion zone and transported to cooler
regions in the formation
decreases ahead of the combustion front and levels off at between 200-600F
The temperature
ahead of which is located still another zone of warm fluids. The combustion zone moves through the
reservoir at a speed which is a function of both the air rate and the amount of oil burned and travels at
a speed less than the advance of the steam plateau. At the time of steam breakthrough in the production wells, model studies have shown that about 17910of the reservoir has been burned and about 50%
of the original-oil-in-place
Relatively high temperatures after steam breakthrough can accelerate corrosion and/or failure of
downhole equipment and cause premature shutdown or loss of production wells. Significant additional oil recovery can be obtained from hot wells with the implementation
conjunction with proper material selection.
364
of downhole cooling in
Estimated cooling water requirements, as a function of oil production rate and formation flow-
k
o
300
t%
250
200
~
8
150
WELLDpTH4000 ~
SURFAEWAW?T6LIP.6 O F
100
WAIRINN@UJLUS
50
100
200
300
400
500
600
700
800
900
365
I
100
I
I
I
I
I
I
I
I
3
I
<4
,- .,,-ltiNl
1k
TIME, YEARS
100
3
}
IGNITE
TIME
366
Operational
Problems
Each fireflood is unique and the operational problems encountered are very much project dependent. These were discussed in detail in Chapter six. The most common problem plaguing many combustion projects include sand production, corrosion, emulsion, well failures, compressor failures, and
explosion. These and other operational problems are discussed under individual projects in Chapter
six.
Project
Various measurements
fireflood. Measurements
ents
Monitoring
required to monitor the process are: injected and produced fluid measurem-
produced gas analysis, downhole temperature measurement and flowline temperature measure-
1982)
1.
Combustion gases from producing wells were tested monthly for volumes and content. The
volumes were determined for material balance calculations.
cated the efficiency of combustion and whether the combustion front had reached the producer.
2.
Producing wells were tested twice a month for oil and water production rates.
3.
Flow line temperatures were checked daily for indications of thermal stimulation or downhole problems. The volume of quench water used at the producers was based on observed
flowline temperature as well as gas production rate.
4.
Downhole temperature measurements at the observation wells were utilized to calculate the
size of the burned zone and location of the burn front.
367
Waste
Fluid
Disposal
Introduction
Consideration of environmental impact is an important part of planning a fireflood clearly; a
fireflood can produce large amounts of pollutants by way of the flue gas and from the air compression
and oil handling equipment. This will have a major impact on project design in terms of equipment
selection, operating procedures, and cost.
This section offers an introduction to the pollution control methods applicable to a fireflood
project. Topics covered include:
1.
2.
3.
Waste Liquid
and Their
Disposal
The water produced from a fireflood is generally similar to the water obtained from primary production. In that case standard oil field water disposai procedures are applicable, and this waste maybe
combined water from primary production. Two exceptions are noteworthy:
1.
The water may contain dissolved HJ3. If so, special handling procedures will be required to
assure operator safety.
2.
The water will contain dissolved C02, which may warrant special equipment to avoid corrosion.
Some pollutant removal system, such as scrubbers, will produce toxic liquid wastes requiring
special handling. These wastes may contain caustic, mixtures of (HCOJ-,
2.
Blend with a large amount of production water and inject into a water disposal zone.
3.
Since EPA and local pollution disposal regulations depends upon the toxicity and nature of the
pollutant and often changes, it is advisable to consult the locaJ and state pollution control authorities
prior to formulating a disposal plan.
368
Waste
Gases
General
The possible waste gases released from a fireflood operation and their sources are listed in Table
9.3. All of these pollutants would not be produced from a fireflood at the same time. Their occurrence
depends on the flue gas composition, the operation of the fireflood, and the equipment used for the
project. Although flue gas is the key source of pollutants, pollution control equipment can also release
pollutants as process by-products or through system inefficiency. The typical amounts are estimates
based on the most probable behavior of a fireflood producing 8 MMscf/d of flue gas and requiring
2,535 kW (3400 hp) for air injection. Engine exhaust emissions are based on EPA guidelines and on
data obtained for a 746 kW (1 ,000 hp) turbo charged reciprocating engine burning natural gas at a rate
of 97.3 scf/min. Emissions from scrubber systems are based on 90-9590 removal efficiency.
Pollutant
Sources
TypicalAmount
Cone.Range
Hydrocarbons
FlueGas(Flare)
CrudeOil Storage
EngineExhaust
1-5%
o-3%
0.3-0.6%
200,000Scfld
9 Scf/bbloil (estimate)
14,400Scf/d
H#
FlueGas(Flare)
H# Scrubber
o-1%
0-0.170
80
IncineratorExhaust
S02 Scrubber
0-0.06%
0-0.06%
1700 Scfld
80 Scfld
o-3%
0-0.5%
0-0.5%
0-0.6%
16,000
32,000
10,000
30,000
32,000
0.04-0.38%
4,200 Scf/d
EngineExhaust
Incinerator
Flame
Catalytic
13,200 ScfJd
10 300 ppm
c 10 ppm
600 Scfld
60 ScUd
<50 ppm
200 Scf}d
so~
co
,,
NOX
Odor(frommiscellaneousS-gasesand
hydrocarbons)
369
1,600Scfld
Scfld
Scf/d
Scfld
Scfjd
Scfld
ScfJd
Flue
Gas
The hydrocarbon composition of flue gas changes with time according to the trend illustrated in
Figure 9.6, Actual times will depend on the operation of the fireflood.
The first flue gas produced is the natural gas in the reservoir. As C02 from combustion and nitrogen from injected air sweep through the reservoir the hydrocarbons
until, after some months, their concentration reaches about 3% with a heating value of 50-80
Btu/Scf,
Larger concentration changes can occur on account of shutting in production wells, changing flue gas
production rates, or changing air injection rates.
Depending on the reservoir, flue gas can contain H.Jl and trace amounts of other sulfur gases
such as COS, CS2, mercaptans, and disulfides. The concentration of H2S in flue gas can be expected to
follow the trend shown in Figure 9.7.
Since HJ3 is highly soluble in most crude oils, the concentration in flue gas will be very low initially. When the crude oil becomes saturated with H2S, the concentration in the flue gas will increase
dramatically. This generally occurs when the flame front is near the production well.
Flue gas typically contain 0.2 & 0.1%
may be indications
of
incomplete combustion resulting from air breakthrough or an excessive air injection rate.
Most flue gases have a burnt gas odor resulting from trace amounts of sulfur gases and higher
hydrocarbons. If the odor of the flue gas is highly obnoxious., gas treatment maybe required to eliminate a public nuisance.
Pollution
Control
Equipment
General
Table 9.4 summarizes the different pollution control systems suitable for a fireflood project and
their recommended
natives
370
Equipment
Gas Treated
Method
Pollutant
Removed
Suggested
Application
FlareStack
Fluegas
None
Ventto
atmosphere
Fluegas meetsair
qualityregulations
FlareStackwithFlame
Burner
Fluegas
H/C1),CO(l),
S-gases(l2)
B~
Fluegaswithenough
H/c to support
H/c,co,
Burn
Bum
Bum
Chemical
Reaction
Thermal
Incinerator
Flue gas
Catalytic Incinerator
(Excess Air)
Flue Gas
Catalytic Incinerator
(Substoichiometric)
Flue Gas
H2S Scrubber
Flue gas
combustion
(>200 Btu/Scf)
S-gases(2)
H/c, co,
S-gases(2)
H/C, CO(3),
S-gases(2)
H2S(4)
amounts of other
S gases
S02 Scrubber
Incinerator
Exhaust
S02(4)
1. Removal Efficiency
may be poor
2. S-gases are converted to S02
3. May increase
amount of CO for
flue gas with heat
values >70 Btu/Scf
4. T~ical removal efficiency is 9G95 70.
371
Chemical
Reaction
Flare
Stack
The most common means of flue gas disposal has been through a flare stack. The simplest and
lowest cost system available.
A flare stack may consist ofi
1.
2.
A pipe that releases gas inside a vented tank for mist containment.
3.
Combustion
Gases
of combustibles,
some assistance is
Oxidation
in thermal incinerators
is accomplished
to
dence provide the necessary heat energy time results from slow flow through a long insulated flow section.
2.
Catalytic incinerators use a combustion catalyst to reduce the activation energy required for
hydrocarbon oxidation.
Similar to thermal incinerators, combustion efficiency increases with increased residence time
over the catalyst and increased operating temperature.
Thermal
Incinerators
an exhaust stack.
Heat recovery equipment can also be included. This may be worth considering at installations
where heat treaters, buildings, steam generators,
372
The major disadvantage of all thermal incinerators is the cost of fuel for the flame burner. Some
designs offset this somewhat by using the heat released from burning the pollutants to supplement
flame heat. Even so, the cost of fuel can be significant and should be considered carefully when purchasing such a unit. Thermal incinerators are most suitable for flue gases having heat values between
85 and 200 Btu/Scf.
Catalytic
Incinerators
A catalytic incinerator uses flameless combustion on a combustion catalyst to destroy gas pollutants. Such units are more expensive than thermal incinerators, but they offer much lower operating
costs because no supplemental fuel is required for operation. The major components of a catalytic
incinerator are: the catalyst beds, a heat exchanger, air blowers, a preheat burner, and the stack. The
preheat burner is used only for start-ups to heat the flue gas, and in turn the S 1 catalyst to a temperature above the ignition temperature for the waste gas. Most flue gases will ignite at 350-450F.
Once
the first stage exhaust temperature is high enough to assure self-sustaining operation, the burner is
shutoff. A flame shield is recommended
mixer shown after the heat exchanger is recommended to assure good fuel and air mixing.
Combustion in a catalytic incinerator is controlled by the amount of air added to the waste gas
before each catalyst bed. Two types of operation are possible. The first involves substoichiometric
(fuel rich) combustion in both stages. This is especially suitable for flue gases having heat values less
than 85 Btu/Scf. In this mode enough air is added to bum 25 Btu/Scf per stage or about 50 Btu/Scf of
the total flue gas heat value resulting in a final exhaust temperature of about 1600F. Any excess heat
value will exit the system as unburned hydrocarbons in the exhaust. Fortunately catalytic combustion
follows the standard combustion hierarchy where larger molecular weight compounds are preferentially burned. Thus, the unburned hydrocarbons are mostly methane, which is an acceptable emission.
The second mode of operation involves substoichiometric
combustion in the last stage. This extends the upper limit of burnable flue gas heat value and it provides for complete combustion. Suggested applications would be waste gases having heat values less
than 200 Btu/Scf. or projects requiring complete removal of the pollutants in flue gas.
373
Scrubbers
Scrubbers remove gaseous pollutants by contact with a solution that will react with the pollutants and thereby absorb them. Scrubbers can also be designed for particulate removal, but this is not
neeessary for the particulate free gases produced by a fireflood. In a fireflood application scrubbers
would be used
to
remove H2S from the flue gas or S02 from the exhaust of a system which burns flue
gas. Generally, removal of H2S should be preferable because smaller gas volumes and gas at lower
temperature are involved. However, incineration of the flue gas followed by S02 scrubbing of the
exhaust may be required for flue gases containing large amounts of miscellaneous
removable by an H2S scrubber. Removal efficiencies are typically 90-95%
bing.
Caustic scrubbing is the most common means of removing H2S from a waste gas such as fu-eflood flue gas.
There are many reactions that will effectively remove S02. The most common are:
1.
2.
3.
Scrubbing with NaOh is probably the most suitable for the size system required for typical fireflood projects. It is highly efficient and generally allows for a smaller, simpler scrubbing system. The
other approaches have merit for very large scrubbing systems or for locations where the scrubbing
agent is available at very low cost. Two alternatives to caustic scrubbing are noteworthy: dry scrubbing and scrubbing with oil field produced water. Dry scrubbing involves flashing a concentrated
solution of NaHC03
and Na2C03 in the scrubber followed by filtration of the effluent to remove par-
ticulate solids. Production water may be usable for S02 scrubbing if it contains large amounts of carbonates and bicarbonates.
374
REFERENCES
1.
MReview
of Fkeflood Rejects,
pp.
CHAPTER
10 OXYGEN
/ ENRICHED
AIR FIREFLOOD
Introduction
Use of pure oxygen or oxygen enriched air instead of air as injectant in in-situ combustion process has several advantages and has been practiced since the early 1980s. Greenwich Oil Corporation
was the first operator (Hvizdos et al., 1982) to successfully employ oxygen-enriched
air in a combus-
tion project. Laboratory and field studies with pure oxygen or oxygen enriched air have suggested that
the performance of an in-situ combustion process can significantly be improved over that of conventional air injection. Oxygen or enriched air combustion projects increase the percentage of oxygen
being injected to greater than 21 %. In heavy oil reservoirs, use of enriched air often enhances the high
temperature oxidation reactions and improves recovery. The use of oxygen or oxygen enriched air in
fireflood to improve performance
enriched air in fireflood could provide the desired oxygen flux at lower injection rate and pressure.
Oxygen fireflooding could minimize many operational problems encountered in the traditional air
iireflooding such as the high gas vent rate, tubular erosion, excessive sand production, etc., increase
oil productivity and improve the economics.
the laboratory, and practiced in the field since the 1980, the process is still considered as an unproven
emerging technology. Assuming that field is amenable to oxygen fireflood, the decision to implement
the process in the reservoir will depend primarily on economic,
377
air
Potential
Advantages
and Disadvantages
of Oxygen/Enriched
Air Firejlooding
A major advantage of using oxygen or oxygen enriched air in a fireflood is that the volume of
gas necessary to move the combustion front at a given velocity is about one-fifth that required for air
injection (Garon et al., 1986).
injection pressure than is possible with normal air. This often translates to accelerated oil production
due to faster propagation of the combustion front through the reservoir. In addition use of pure oxygen or oxygen enriched air in fireflood will result in a significant reduction in the volume of produced
gas and generates a high COZ concentration in the exhaust gas.
Moore et al. (1984)
Under conditions of elevated pressure and high oxygen flow rate, oxygen is cheaper than air
(Howard, 1982).
2.
Surface facilities for small scale (pilot operations) are relatively simple and should present
less severe operating problems than are normally experienced with conventional air injection operations.
3.
Lower injection rate resulting from the use of enriched air could make some tight reservoirs
amenable to fireflooding that otherwise would not be suitable and could enable larger well
spacings to be used.
4.
In enriched air fireflooding, the concentration of C02 in the produced gas could be in excess
of 90% (Persico et al., 1983), and, consequently, the partial pressure of C02 would be much
higher. Since the volubility of C02 in crude oil increases with pressure, high concentration
of C02 can result in miscible displacement of the crude oil.
5.
In heavy, viscous oil reservoirs, swelling and viscosity effects associated with C02 volubility in the oil will contribute to accelerated oil recovery. The extent of the benefits due to
C02 will depend to a large degree on the extent of reservoir contacted by the carbon dioxide.
6.
Oxygen enriched air combustion projects often exhibit better combustion efficiency than
the normal air provided the oil display non zero order reaction rate with respect to oxygen
partial pressure (Moore et al., 1984).
7.
In oxygen firefloods, spontaneous ignition can occur much more quickly due to greater
reaction rates.
378
8.
The fuel availability and the oxygen requirements per unit volume of reservoir are generally
lower in enriched air firefloods than the normal air projects. Moore et al. (1984)
attribute
this to lower gas partial pressure that result in higher steam bank temperature and quality.
9.
The use of oxygen or oxygen enriched air as injectant in firefloods could result in improved
sweep efficiency. The reasons for this improvement are: (a) improved mobility ratios resulting from the increased volubility of C02 in the crude, and (b) improved stability resulting
from a lowering of produced to injected gas ratios.
10. The use of oxygen instead of air can lower the produced gas-oil ratios by a factor of about
five and would result in higher oil production. The benefits of lower produced gas-liquid
ratios are: improved bottom hole pump performance,
expensive gas handling equipment, and lower sand production due to lower gas velocities at
the perforations.
11. Oxygen fireflood can alleviate some of the more common operational problems associate
with the normal air injection projects such as the poor infectivity, ignition failures, sand
migration, erosion, gas leakage to other sand, emulsion, corrosion and explosion hazards.
Problems associated with the reservoir heterogeneities and anisotropices, however, can only
be resolved through field tests in specific reservoirs.
The use of oxygen or oxygen enriched air as injectant in firefloods offer nMrly advantages, the
process also posses many disadvantages. The inherent disad vantages associated with the process are
as follows:
1.
when considering
or even
enriched air for oil recovery. In oxygen service everything is fuel. Conventional oil field
carbon steel tubing and casing will burn in an oxygen rich environment when subjected to
an ignition source. The use of oxygen in systems contaminated with hydrocarbons can be
particularly
which
might occur,. and to eliminate all hazardous situations. Velocities in the wellhead and tubing
must be constraint to prevent creation of sparks due to impingement. Though, field experience with oxygen fireflooding is limited, pure oxygen can be injected into an oil reservoir
as demonstrated by the success of Forest Hill oxygen fireflood operation.
2.
The amount of oxygen stored behind the combustion front is much greater in oxygen fireflood than in normal air injection fireflood. The fraction of the injected oxygen retained
behind the combustion front will increase significantly with increasing operating pressure.
379
The volume of oxygen stored behind the front, however, can be lowered by co-injecting
water with oxygen or by injecting oxygen and water alternatively.
3.
Corrosion in oxygen fireflood operation is likely to be more severe than in air fireflood due
to the elevated concentrations of oxygen and COZin the vent gas.
4.
The operation of a pilot test conducted to ascertain the suitability of a reservoir for oxygen
fireflood is often more expensive than the normal air fireflood. This is because the oxygen
fireflood pilot is usually performed
with expensive
liquid oxygen
delivered by tanker
operation,
also
Economics
of Oxygen
Supply
Fire#ood
Option
Once fireflooding has been identified as the preferred technique there is the question of whether
to
use air or oxygen. Though enriched air offers much greater flexibility than normal air and likely to
the decision whether to use air or oxygen (or oxygen enriched air) as
should include both the cost of supplying oxygen to a fireflood project as well its effect on the overall
fireflood performance.
wider well spacings, lower gas to oil ratios, and produced gas handling and disposal costs (Shahani
and Gunardson, 1994).
There are three options for supplying oxygen to a fireflood project. In the first option (Figure
10. 1), liquid is trucked from a suppliers air separation plant and stored on the site in vacuum insulated cryogenic tanks. These tanks normally would hold a weeks supply of oxygen. The liquid is
pumped from the tanks to a vaporizer at the required pressure and then vaporized and stored until
needed. In the adsorption processes (Figure 10.2), air is passed through a vessel containing adsorbent
(an inorganic crystalline material known as a zeolite molecular sieve), that selectively adsorbed nitrogen and carbon dioxide molecules from air on its surface and let the oxygen passes through. The gas-
380
eous oxygen with about 90% purity (entirely satisfactory for fireflood applications) at low pressure is
then compressed to the injection pressure. In the third option (Figure 10.3) a cryogenic air separation
plant is employed to separate oxygen from other constituents of the air. Cryogenic separation is the
most energy efficient and lowest cost method to produce large volumes of oxygen for a large commercial scale oxygen fireflood. Selection of the oxygen supply mode depends on the conditions of the
field project including the project life expectancy
1982).
n
~smus
Oxwm
LIQUID
STORAGE
OXWBJ
STOR4GE
C%smvs
FLowmrfrm
VAPWIZOR
LIUJID
(XYGEN
UJMP
381
yJSU V>-%g=i+fjw
LIC147 OXYGS4
PUMP
DISTILLATION
TruckedLiquid
AdsorptionP1ant
CryogenicPlant
<5
>5
>10
0.2-1.4
Operatinglife (yr.)
Capacity (MMscf)
0.2-2
Capital cost
Low
Operating cost
Lead time (months)
cost ($/Mscf)
The oxygen
supply companies
Medium
High
High
Medium
Low
14-16
18-22
90-95
95-1oo
-1oo
Purity (%)
0.2-50+
3.0-5.0
,.
2.5-4.0
1.2-3.0
include:
1.
2.
3.
These companies can provide support to operators in the design of oxygen injection facilities,
well completion design, and train field personnel in the safe handling of oxygen and in the selection
of oxygen compatible materials.
Economics
Many factors affect the economics of oxygen fireflood. The factor, which has the greatest effect
on the commercial
scale operation, is the cost of oxygen compared with the cost of air. To inject oxy-
gen into a well it is necessary to compress the gas to an elevated pressure. A given volume of oxygen
can be compressed to the desired pressure using a smaller compressor at lower compression cost, than
the air containing the same volume of oxygen. This advantage in capital investment and operating
cost for compression, however, is offset by the cost of producing pure oxygen, The oxygen production
cost is a strong function of the plant size and decreases
increases due to economic of scale. The larger the quantity of oxygen the lower is its unit cost. Therefore, the difference in cost between supplying oxygen or air at elevated pressure is a function of the
quantity of gas supplied and the pressure at which it is delivered. The cost of oxygen is only slightly
affected by the desired purity. The cost of producing 80% purity oxygen is only about 5% lower than
the cost of producing 99% purity oxygen (Hvizdos, 1978).
60
-:
40
~ mwh
hwwk iniedingoxygenvs. in@ing
mmw~ a,,- anoxysm
quir%nmni
d 10MMs.cidat
ap==eoftsoopm
Air is cheaper to
lWM
Answw
Appmxi!nmteiy
compress and inject
20
A ~~f
Breskeverr
40 ~
Oxygenie chesper
to compress and
inject
60
80 ~
56
10
20
30
383
Injection of pure oxygen can be cheaper than the equivalent amount of oxygen in air at elevated
pressure. Hvizdos et al. (1983)
showed that oxygen could be less expensive than air at high pressures
and high flow rates. In Figure 10.4 the difference in cost between supplying oxygen or air at elevated
pressure for a range of quantities and delivery pressure is depicted (Shahani and Gunardson, 1994). In
this plot the quantity of contained oxygen is shown on the x-axis. The differential cost between compressing oxygen and air in cents per thousand cubic feet of contained oxygen is shown on the y-axis.
This differential cost takes into account the capital and operating cost of the compressor(s)
the cost of producing the specified quantity of 99.5%
as well as
project. The economic life of the project is assumed to be 15 years. Both the power cost and the rate
of return on investment are also assumed to be same for both the systems. The analysis shown for
three different deliveries pressures. The discontinuity in the curves for 100 psia and 3,000 psia are a
result of switching from a combination of reciprocating
at lower flow
pure oxygen.
advances into analysis will more likely further improve the economics
Incorporating
technology
these
have
attractive.
For small experimental projects, the cheapest source of supply would probably be oxygen delivered to the site as liquid oxygen and vaporized as needed. The current price of liquid oxygen delivered
to the site range from $110-$140
gen plant. Since most commercial
per ton depending upon the delivery distance from the liquid oxyprojects will require large volumes of oxygen, it may be best to
erect an on site oxygen plant. The best way to supply oxygen for a commercial
complex and the service of an oxygen supplier must be enlisted to arrive at the best plan.
384
----
SIR
OXYGEN
......
I 1 1 1 I 1t
tO
100
OELKUERY
PRESSURE
1 atw
(PSIfl)
Pete et al. (1988) prepared economic projections to compare oxygen with air injection for the
production of heavy oil in Texas. This analysis indicate that when the projects oxygen requirements
are greater than 4 MMscf/d and the injection pressure is greater than 180 psi, use of pure oxygen is
more cost effective than air (see Figure
0.5)4
Laboratory
Studies
Combustion tubes are the principal physical models used by most organizations to evaluate and
compare the combustion performance of oxygen/enriched air versus normal air and to obtain design
parameters for oxygen/enriched
air flooding. Combustion tubes are linear models and cannot provide
information on sweep efficiency. Three dimensional scaled model experiments of fireflooding and
numerical simulation can provide useful information on fireflood sweep efficiency and one such study
(Garon et al., 1984) has been reported in the literature. This section summarize the results of selected
laboratory oxygen/emiched
air studies
385
Pusch (1977) carried out the earliest reported tube tests using oxygenx. Results of the combus-
tion tube tests carried out by Moss and Cady (1982) using oxygen indicated that the peak temperatures and oxygen utilization were almost identical for both air and oxygen. This study also indicated
the oxygen requirement was 35 and 20% higher than the equivalent air in dry and wet combustion.
* This study was the first to document the importance of safety in using oxygen in a hydrocarbon environment.
A more comprehensive
(from 21-95%)
and light oil (31 API) indicated only very small differences between runs with regard
to fuel consumption,
apparent hydrogen-carbon
utilization. Shahani and Hansel carried out further work on light (32~1),
heavy (13API)
combustion characteristics
from 2195%.
medium (2123AP1),
and
tend to improve with oxygen additions, and that the chief benefit of oxy-
gen enrichment in light and medium gravity oils was that the combustion could be maintained for
conditions at which the use of air was unsatisfactory. They also found that for heavy oils, the fuel
availability was significantly reduced at high oxygen concentrations,
ratio. Oxygen combustion tube studies carried out by Garon and Wygal
(1974), however, indicated no significant change in fuel deposition with a change in the injected oxygen concentration.
Results of the combustion tube experiments performed by Moore and his co-workers at the University of Calgary, Canada (1987)
suggest that the degree of oxygen enrichment of the air can affect
the bum performance. Their tube tests performed on Athabasca oil sand core indicates that the overall
oxygen and fuel requirements
because of the direct effect of total pressure and total gas injection flux on oxygen storage, the oxygen
flux at the bum front is likely to be lower for enriched air combustion than for normal air combustion.
Moore (1987) also observed significant oxygen storage in the unswept region downstream of the main
combustion zone,
386
Results of oxygen enriched a+ combustion tube studies by Petit (1987) on heavy (15.8 API) and
light (25.7 API) oils indicated that irrespective
are
higher at the front of the combustion front than for normal air combustion. The oxygen requirements
are slightly higher for heavier oil than for the lighter oil.
Enriched air wet combustion tube studies by Venkatesan et al. (1990) indicate that the fuel consumption and oxygen requirements generally decrease with increasing water oxygen (WOJ
ratio
and remain constant during the transition from normal wet to super wet combustion.
Three dimensional scaled model tests by Garon et al. (1984)
Safety
Consideration
General
Beyond the economic and process issues, safety is a prime issue in oxygen fireflood. Oxygen,
especially at high pressure, is more hazardous than air. Many metals including the conventional oil
field carbon steel tubing and casing will bum in high-pressure oxygen enriched environment (common to most oxygen firefloods), if ignited. In the fireflood process all three components of uncontrolled combustion namely, oxidant, fuel and a potential source of ignition are present. Hence
special precautions must be taken in the operation of an oxygen or enriched air fireflood. Technical
cooperation between the operator and the oxygen supplier is a must to ensure safe operation of the
oxygen/enriched
air fireflood. Service of the oxygen supplier must be enlisted in the development of
operating procedures for the safe operation of oxygen injection system and in the design of injection
wells, and production wells.
387
Pressure, psia
4,000
2,000
1,000
800
600
400
Flame did not
progrogate
200
100
0
20
40
60
80
100
% Oxygen in nitrogen
In an oxygen rich environment, the combustibility of metals depends on the oxygen concentration, initial pressure, temperature, geometry, heat sink potential, and direction of bum (Shahani and
Gunardson, 1994). Flame propagation in carbon steel is a strong function of the oxygen concentration
and pressure. Laboratory tests indicate that at a given pressure, carbon steel tube would continue to
bum upon ignition, provided the oxygen concentration
value the tube will not bum. In Figure 10.6 the propagation
(Merington,
1983). In oxygen
injection well, flame propagation can be controlled or eliminated by incorporating firebreak sections
in the tubing and casing. The firebreak is a length of casing or tubing section that is made from metals
whose melting point is well below their ignition point and hence do not bum in oxygen. Nickel- and
copper-based alloys are the materials of choice for firebreak in oxygen fireflood application. The use
of firebreak in oxygen fireflood is very cost effective since it eliminate the need for expensive fire
resistance materials in large quantities.
388
lCarbon
Velocity, fps
steel pipe I
200
150
100
50
0
Internal pressure, psig
Ignition may occur if the velocity of oxygen in the piping system is such that it led to generation
of spark due to particulate impacting metal surface. The maximum velocity of oxygen in the piping
system must always below a certain threshold value. The threshold value is a function of the prevailing temperature and pressure. Figure 10.7 shows the effect of internal pressure on the maximum
velocity for straight runs of carbon steel pipe (Merington, 1983). For bends and elbows, the maximum
allowable velocity should be reduced by 50%.
Injection
Well
In oxygen fireflood, the greatest safety concerns are at the injection well. Design of the wellhead
and wellbore, including injection tubulars, packers, ignition equipment, burners, chokes, and support
equipment is governed by the same velocity and material constraints outlined in the previous section.
Tubing and casing need not be made of exotic metals provided rigid adherence to velocity threshold is
maintained. The injector should be initially cleaned using Nitrox solution to minimize presence of
hydrocarbons. A packer must be installed to maintain the casing-tubular armulus oil free. Packer seals
389
gaskets, valve packing and thread lubricants must be made of oxygen compatible material such as the
Teflon.
Injection tubing must be free of rust or other loose particles, which can create spark by impingement upon other surfaces. Injected gases must also be free of such particles. Use of premium thread
designs which eliminate internal gaps between connecting collars is recommended to avoid disrupting
oxygen flow and to prevent dirt, oil, and other particles from lodging during cleaning operations
(Merington, 1983). Premium connection tubulars should be cleaned and redoped with Teflon sealant.
The well should be equipped with bottomhole temperature monitoring equipment and purging
facility. If the temperatures exceed predetermined limits then the well should be purged with nitrogen
or quenched with clean water. As mentioned earlier, the oxygen injection well completion
design
should incorporate a firebrealc section in the casing and tubing to prevent upward migration of a metal
burn should one occur.
An overbalance injection pressure should be maintained at all times to prevent backflow. If the
primary injection system shuts down for any reason, a backup system must be available.
Oxygen enriched operations should always be start with nitrogen diluted synthetic aiq then
gradually increase oxygen concentration over a period of few weeks to reduce potential for high temperatures near injection well-bores (Duncan and Khallad, 1996). If oxygen injection is to be terminated for an extended period, Duncan
(1996)
recommend
that an oxygen
compatible
wireline
blanking plug be installed in a downhole profile. The tubing must be flushed with water or nitrogen
prior to installing the plug. The slickline unit must be cleaned and relubricated with synthetic oil.
Producing
Well
Completion of producers for oxygen fireflood application is no different from those normal air
fireflood producers. In oxygen fireflood increased corrosion can be expected at producing well due to
the presence of higher concentration
the producers. At atmospheric pressure oxygen and natural gas form an explosive mixture if the oxygen level exceeds
exceeds 5% by volume. At
higher pressures, the oxygen natural gas mixture can explode at lower oxygen concentration.
For
instance, at 500 psig the required oxygen concentration is reduced to 9%. Hence the oxygen concen390
tration in the produced gas must be measured frequently (preferably twice a day) and corrective
action tziken to avoid a potential explosive situation. Production wells should be shut-in upon detection of high oxygen levels, and pressures kept low to minimize damage in the event of an explosion.
If the well is shut in due to high oxygen level in vent gases and if it become necessary to vent the
annulus gases, then special precaution must be taken. The venting valves must be blanketed with
nitrogen prior to opening and the impinging velocities must be limited to avoid creating sparks. During shut-in it is advisable to circulate the wells with inhibiting fluids and avoid taking acoustic fluid
level measurements.
Oxygen
Distribution
Lines
The delivery of pure oxygen to the wellhead is not much different from piping oxygen to a reactor in the chemical process industry. By adopting existing technology and safety procedures oxygen
can be safely delivered to the wellhead. The pipeline system must be sized to satisfy for both allowable pressure drop and velocity constraints. Elbows, tees, valves, orifices, and other pipeline components must be sized based on the velocity constraint. Fairlield (1985) recommended that in the design
of oxygen deliverys ystem for fireflood consideration be given to the following:
Since impingement of loose particles on other surfaces can be a potential ignition source,
maintain oxygen velocities at low values.
material.
Avoid conditions that can result in rapid compression of gases during the valve opening and
closing operation.
391
Field
Projects
known to have been in operation in North America during the 1980s. These include 4 in the USA and
6 in Canada. The U.S. Oxygen Fireflood include: Greenwich Oils Forest Hills field fireflood in
Texas, Arcos Holt sand unit oxygen fireflood in Montague county, Texas, Mobils Esperson Dome
Oxygen
fireflood, Liberty
. County, Wyoming.
The Canadian fireflood include: Husky Oils Golden Lake Waseca fireflood in Saskatchewan,
DOME Petroleums Lindberg fireflood in Alberta, BP Canadas Marguerite Lake fireflood in Alberta,
Gulf Canadas Pelican Lake project in Alberta, Petro Canadas Kinsella B fireflood in Alberta, and
Canadian Occidentals Morgan Area fireflood in Alberta.
Fairfield (1985) has presented the salient features of these operations. In the following the operational history of Greenwich Oils fireflood is presented
operation aspects of a commercially
to
Forest
Hill
Oxygen
Project
Fire$ood
History
The Forest Hill Field is located in Wood County Texas, near the town of Quitrnan. The formation is the Harris sand, Eagleford in age. The field is three miles long (northeast-southwest
direetion)
and two and one-half miles wide. The crude in the Harris formation is a low gravity (l OOAPI), high
viscosity, asphaltic sour crude. The field was discovered in 1954 and by 1964, 26 wells were drilled
and completed. Field development was curtailed in 1966 due to low crude prices and a rapid decline
in production. From 1966-72
392
ary 1980 and terminated in 1982. This two year field test was the first known successful oxygen fireflood in the world and demonstrated that high concentration
80% purity oxygen was injected into injection well at an average rate of 250,000 scf/d at an
injection pressure of about 2,200 psig over a two-year period. A total of 0.22 MMM scf of 8090 oxygen was injected into the formation at an oxygen cost of $6.6/bbl (1980 dollars). The wells in the oxygen pattern showed a 55% increase in oil production over those in the air injection patterns. The oil
production increased from 10,200 bbl/month (average air injection pattern production) to 15230 bbl/
month an increase of 5,030 bbl
The oxygen fireflood test was terminated in .January 1982 due to falling oil price. Following the
termination of oxygen injection, Greenwich started injecting water in March 1982, to recoup the heat
generated in the reservoir during the fireflood and to maintain reservoir pressure. The oxygen fireflood
project results were very favorable and Greenwich planned on resuming oxygen injection in 1985, but
later abandoned the plan due to oil price collapse.
Project
Description
The Forest Hill oxygen fireflood operating facilities were divided into two sets of subsystems:
injection, and production. The components of these subsystems are described in the following sections. The properties of the test reservoir site is shown in Table 10.2
393
TABLE 10.2 Forest Hill Oxygen Fireflood Reservoir and Fluid Properties
Description
Value
Formation
Harris Sand
Depth (ft.)
4,800
27.7
626
36
1,362
28,568
1.01
10
1,002 @ 18F
185
31
Injection
Subsystems
Injection pipelines.
Injection
Gas Supply
System
02NZ supply/blending system where it is cleaned, dried, and mixed with oxygen to provide a blend
up injection gas with oxygen purities ranging from 2 190 mole percent. A sidestream of clean air is
mixed with oxygen from the liquid supply system to provide a stream of 0.9 mole fraction oxygen.
Nitrogen for purging is supplied from liquid nitrogen storage, pumping, and vaporization facility.
Oxygen is supplied from a liquid oxygen tank. All three products (Nz, blend up gas, and 90% 02) are
supplied to separate manifolds at a pressure of 3,000 psig.
The injection gas supply system is owned and operated by the oxygen supplier. The injection
gas supply system sports a safety system to prevent uncontrolled
break downstream. Also, gaseous nitrogen is stored at an elevated pressure to be made available for
line and well purging during power outage. Plant emergency trip stations are incorporated into the
system to allow manual shutdown of the system in an emergency.
Flow
Control
Skid
The function of the flow control skid is to provide the capability for taking blend gas (21<90%
02), oxygen, and nitrogen from the supply system headers and distribute the required gas or blend to
individual wells. Each skid is designed to service six or seven injection well.
395
Individual mass flow meters are utilized to provide a measurement of the volume of gas injected
into individual well. The operating flow range for each well is O-45,000
included in each skid to allow the operator to set the oxygen purity during the blend up phase. Feedback from the injection gas supply system is provided by an oxygen purity read-out.
For personal protection, valves and indicators (gas flow, line pressure, manifold temperature,
pressure drop across the strainer, Nz purge flow) are mounted on panels which shield the operator
from the piping valve system. Similarly, instrumentation, which may need calibration (pressure, flow
and temperature transmitters),
maintain gas flow for well protection in emergency situations and protect against overpressurization.
The skids are designed for 3,550 psig, and are protected by relief valves set at 3,300 psig.
The flow control skids are also designed to interrupt the flow of oxygen in case of a line break
downstream and to provide for automatic line and well purge during a power failure or activating of a
pkmt emergency trip station at the injection gas supply system.
Injection
Pipelines
The injection pipelines carry nitrogen, air, air-oxygen blends, and oxygen from the flow control
skid to the wellhead area of each injection well. The lines are welded and run underground.
Each pipeline consists of 2.375 O.D x 0.218 wall externally coated carbon steel carrier pipe
and terminated at each end with a 1 schedule 80 monel tie-in piece topped with an insulating connector. For safety, all ca.tbon steel piping is buried at a minimum depth of four feet. Bright yellow marker
..
posts bearing pipeline signs identify the pipeline route. These posts are located directly over the pipeline and each can be seen easily from the next.
Pipelines are equipped with cathodic protection to monitor corrosion. Cathodic protection consists of magnesium anodes located at intervals not exceeding 1,000 feet along the pipeline. For each
anode location, there is a test station, which consists of a junction box on top of one of the yellow
marker posts. The insulating connectors at each end of each oxygen pipeline are parts of the cathodic
protection scheme and are protected by Kirk cells from internal arcing during lighting strikes.
396
The pipelines are all made of welded construction, are hydrotested, and then cleaned for oxygen
service. Relief valves mounted on the flow control skid are utilized to prevent the pipeline from overpressure.
High pressures and high oxygen concentrations pose potential combustion risks to the pipeline
system. Careful pipeline cleaning, gas velocity limitations, proper equipment and material selection,
and monitoring of the oxygen supply can limit the risk of pipeline contamination
and subsequent
Injection
Wells
and Wellhead
Area
The wellhead area subsystem is an above ground arrangement of piping, valves, and controls
that connects the injection piping to the Christmas tree. Nitrogen, air, air-oxygen blends, and oxygen
are carried through the wellhead area piping and valves into the injection well.
All wellhead equipment uses conventional technology with provisions for oxygen compatibility,
resulting in high reliability and control. Safety systems are fail-safe with no electrical power required
for operation. There are no unusual safety or operating problems imposed by temporary equipment
failure, curtailment of oxygen supply, or power interruptions, so long as pressure is maintained on the
wellhead. The gas supply system equipment redundancy and backup system and injection well design
ensure this,
The wellhead area controls basically consist ofi
A check valve to prevent injection gas back-flow from the well and minimize fluid backflow from the formation into the bottomhole area.
Pneumatic pressure sensors which isolate the well in case of either a loss of nitrogen pressure in the annulus or a loss of feed gas pressure to the injection well.
Pressure transmitters which allow monitoring of the annulus and the injection gas pressure.
A nitrogen pumper connection which allows purging of the wellhead and well with nitrogen
in case of a gas supply interruption or a well workover.
397
Production
Subsystem
Production wells.
Diluent oil is distributed to each well and is blended downhole with the viscous crude to facilitate pumping. The oil produced from each well is collected and transported to the treatment system.
Produced oil, first processed in a gas oil separator to remove entrained gas, flows to a heatertreater where the oil-water emulsion is broken and the water is removed. The water is pumped to
wastewater storage tanks prior to final disposal using well injection. The treated oil is pumped to storage tanks prior to sale. Produced gas collected from the field and from the oil treatment system is
passed through a gas-oil separator prior to final processing in a gas incinerator. A flare is provided for
start-up and as backup to the incinerator. Waste heat, recovered from the incinerator flue gas, is used
to reheat the oil feed to the heater-treater.
Production
Wells
The production wells employ dual-string completions with downhole gas separation. Diluent oil
flows down one string and produced liquids, with a small amount of entrained gas, are pumped up the
other tubing string. Most of the produced gas flows up the annulus. The gas production rate is controlled at the wellhead to prevent blowout of downstream equipment and to influence oil production
rates.
398
Diluent
Oil Distribution
Diluent oil, which enables production of the highly viscous crude, is distributed to each well
through m underground pipeline network. A 40API gravhy cmde is used as the diluent. The diluent
oil is pumped from the storage at about 600 psig into the pipeline network. The pressure drops to
about 300 psig at the wellhead.
Since the pipeline is underground, the effects of an energy release due to pipeline breakage
would be mitigated. The piping is fitted with flow control valves so that a section of leaking or
blocked pipe can be isolated. If sufficient surface observable oil leakage is noticed, special soil samples
are collected to define the components of the oil leakage and verify whether the oil is similar to
production oil or the diluent oil. This may or may not be of importance from an environmental perspective and will depend upon whether the two oils are hazardous or have different hazardous classification.
Produced
Oil Handling
The produced oil handling system collect and treat the produced oil to meet sales specifications.
This system consists of pipelines, test stations, tank batteries, and sales tanks.
Produced fluid flow through an underground piping network under the wellhead pressure of
approximately
100 psig. The tank line leading to each battery is 4 in diameter. A separate. gas gather-
399
The warmed oil flows into a heater-treater, where additional heat is added via a firetube burning
natural gas. T configuration of vessel internals promotes the breaking of emulsions. Chemicals may
be added to further reduce the surface tension of water droplets and promote their coalescence.
Sepa-
rated gas taken from the top of the heater-treater flows to the gas gathering system. The water collects
at the bottom of the vessel where it is drawn off. Treated crude is drawn from the side. A weighted
float determines the level of the oil-water interface.
Treated crude flows to a sales tank for storage. Each tank is 18 feet in diameter and can hold 500
bbls of liquid. Since the highly viscous crude has very low vapor pressure, vapor recovery from the
storage is not required.
Crude is transferred from the storage tanks to the sales tanks via a LACT unit, which meters and
analyzes the oil. The oil is first pumped through the BS &W mofitor. If the water content is too high,
the bad oil is diverted back to a spare sales tank for reprocessing through the heater-treater. If the oil
is within specifications, it passes through a precise liquid flow meter, which corrects volume measurements for temperature and pressures. The LACT unit also collects time-composited
samples of the
Produced
The produced gas (off-gas)
Gas Handling
System
eliminate hydrocarbon pollutants from the vented stream. This system consists of produced gas gathering and produced gas treating.
A pipeline system carries the produced
source is the produced gas from the production wells. Gas overheads from the heater-treaters and various tanks also contribute to the gas stream. Gas pressure at the wellhead is regulated to about 18 psig.
so gas entering the pipeline downstream can do so under its own pressure. A gas-oil separator, or
knock-out
vessel allows entrained or condensed liquids to settle out of the gas stream. The liquids
400
The gas stream is monitored for composition changes, which could lead to corrosive or flammable conditions. Multiple pipelines with interconnections provide flexibility, while fail-safe
controls,
safety valves, check valves and metering equipment limit the effect of pipeline problems and act as
environmental controls to prevent hydrocarbon
content of the produced gas are oxidized in a gas incinerator. Natural gas is
minimum combustion temperature of 1500F. At this temperature, over 99% of the hydrocarbons in
the produced gas will be oxidized.
Heat is recovered from the flue gases using a closed loop organic fluid system. The heater working fluid passes through the preheater,
A gas flare is utilized to provide backup for gas incineration and to facilitate start-up of the system. The flare is 50 feet high with a lo-inch
Wastewater
Disposal
Produced water, separated in the treatment battery, is pumped to the wastewater storage tanks
prior to disposal. During storage, some additional oil coalesces. Periodically, this oil is skimmed off
and stored in the skim oil tanks from which it is slowly recycled to the inlet of the heater-treaters. The
substantially oil-free wastewater is injected into the wastewater disposal well. If the wastewater injection operation is down for any reason, the wastewater is trucked to an off-site wastewater injection
facility for disposal. Trucking wastewater to an off-site injection facility is a common practice within
the oil patch.
401
REFERENCES
1.
Duncan, G., A. Khallad, and P. Stemler, (1996) Enhanced Recovery Engineering World
Oil, Vol. 217, No.1 (January) pp.6572.
2.
Paper Presented at
the Pacific Oil and Gas Show (Petroleum Abstract No. 364821).
3.
Petroleum Engineering
Interna-
4.
Garon, A.M., M. Kumar, and G.C. Gala, (1986) The State of the Art of Oxygen Fireflooding, In-Situ, Vol. 10, No. 1 pp. 126.
5.
Garon, A.M., M. Kumar, K.K. Lau, and M.D. Sherman, (1984) A Laboratory Investigation
of Sweep During Oxygen and Air Fireflooding, Paper SPE/DOE-12676
6.
Garon, A.M. and R.J. Wygal, JR, (1974) A Laboratory Investigation of Fire-Water Flooding
Sot. Pet. Engr. Journal, pp. 537544.
7.
Hansel, J.G., M.A. Benning, and J.M. Fembacher, (1984) Oxygen In-Situ Combustion for
Oil Recovery:
8.
Fireflood,
Pasha Publications
9.
Hvizdos, L,J. (1978) Oxygen Supply to In-Situ Coal Gasification Projects, Paper Presented
at the 4* Underground Coal Conversion
Symposium,
pp.
165168.
10. Hvizdos, L.J., J.V. Howard, and G.W. Roberts, (1983) Enhanced Oil Recovery
Through
Oxygen Enriched In-Situ Combustion: Test Results from the Forest Hill Field in Texas, J.
Pet. Tech. (June), pp. 10611070.
Oxygen Fireflooding,
12. Moore, R.G., D.W. Bennion, J.D.M. Belgrave, D.N. Gie, and M.G. Ursenbach, (1987) New
Insights into Enriched Air In-Situ Combustion,
Paper SPE16740
402
13. Moore, R. G., D.W. Bennion, and J.P. Millour, (1984) Comparison of Enriched Air and Normal Air In-Situ Combustion, in F.A. Curtis, ed. Energy Development:
able, Conservation: Proceedings
of ENERGEX
15. Perisco, P.J., L.B. Wetherington, and L.J. Hvizdos, (1983) COZ Recovery
from Oxygen
16. Pete, B., G. Edmonds, and P.B. Crawford, (1987) Use of Oxygen Enriched Air to Produce
the Heavy Oil and Tar Sands of Texas, Paper No. 160 Presented at the 4ti UNITAR Conference, Edmonton, Canada, pp. 793805.
17. Petit, H. (1987) In-Situ Combustion
Hamburg, Germany,
In-Shu
Combustion
with Oxygen
Combined
(ISCOW) A New Process for Tertiary Oil Recovery: Erdol and Kohle-Erdgas-Petrochemie Verinight mit Brennst Off-Chemie,
19. Ramey, H.J., Jr. (1954) Discussion on Development of an Underground Heat Wave for Oil
Recovery, J. Pet. Tech. (January), pp. 3233, Transactions, AIME Vol. 201.
20. Shahani, G.H. and H.H. Gunardson, (1994) Oxygen Enriched Fireflooding,
Presented at the DOE/NIPER
Paper ISC2
Conf. 940450
Determination
of Maximum Water/Oxygen Ratio for Normal Wet Combustion, Paper No. CIM/SPE 90-62
Presented at the Petroleum Society of CIM, Calgary, Canada, June 10-13.
403