UNDERGROUND COMBUSTION 01L-REC0VER%

EXPERIMENT IN THE VENANGO FIRST SAND,

,

WARREN COUNTY, PA.

By N. A, Caspexo,

W. T. Wertaq

and W, E. Eckard

UNITED

STATES DEPARTMENT
OF THE INTERIOR
StewartL.Udall,Secretary
BUREAU OF MINES
MarlingJ.Ankeny,Director

Theworkuponwhichthis reportis basedwasdoneundera cooperativeagreement

betweenthe Bureauof Mines,
US, Department
of the Interior,andthe QuakerStateOii RefiningCorp,,Bradford,Pa

This publicationhas been cataloged as follows:

,

Caspero,
N
A
Underground
combustion
oil-recovery
experiment
in the
VenarqoFirst
sand,
WomenCounty,
Pa.,hy N. A.Caspero,
W. T.Wertman,
andW.E.Eckard.
[Washington]
U.S.Dept.
oftheInterior,
Bureau
ofMines[1963]
39 p. tables.
tigationa 6320)

29 cm.

(U. S. Bureau

of Mines.

Report of inve a-

Cooperative
agreement between Bureau of Mines, U. S. Dept. of
the Interior and Quaker State Oil Refining Corp., Bradford, Pa. for
this work.
Bibliography

p. 38-39.

1. Petroleum
industry and trade-Pennsylvania.
corwervation. 3. Combustion.
L Title.
(Series)

TN23,U7 no.6320 622,06173

U. S. Dept. of the Int. Library

.
_.. -.

.. . .

2.

Permkurn

CONTENTS

Abstract.................................................................
Introduction.
.,..........................................................
Acknowledgments
..........................................................
General geology, field development,and productionhistory...............
Preparationfor the experiment
...........................................
Performanceof the experiment..,
.........................................
Ignitionand gas injection..........................................
Subsurfacegas-tracertests.........................................
Flow tests in experimentalwell 1 ... ...... ..** .... .* ...., ..
Reservoircharacteristicsfrom the observationwell......................
Cortngand core analysis............................................
Comparisonof core-analysisresults, experimentalwells 2 and 3.....
Carbon-hydrogentests...............................................
Injectedand producedgas material balance...............................
Discussionof results....................................................
Conclusions
...........................................................
....
References..........................................~....................

1
2
3
3
5
13
13
21
26
28
28
29
32
34
36
37
38

ILLUSTRATIONS
&
1. Columnar sectionof geologic formationsof the Titusville
quadrangle.......................................................
2. GoodwillHill oilfield,SouthwestTownship,Warren County, Pa...... :
3. Oil-productionhistory and well-developmentrates of the Hunter,
Campbell,and Stanton leases..................................... 5
4. Structuralcontourmsp of the top of the effectiveVenango First
6
sand in the vicinity of the combustionexperiment.
....
7
well pattern..;....................
5. Combustion-experiment
6. Core analyses,Experimentalwells 1, 2, and 3...........
s. 9
10
7. Well logs for J$xperimental
wells 1 and 2................. *9*=.*.**
8. Air- and natural gas-injecttonsystem................... .**m.**** 13
9. Air- and natural gas-injectionhistory of the combustion
experiment,Experimentalwell 299 *. ** .. ..* . ... ..*****..* 15
10 Gas productionand oxygen and carbon dioxide content of produced
16
gas from we114A ........................... .W*****9***
*.*.*****
11. Gas productionand oxygen and carbon dioxide content of produced
gas from well 17........................... ****8**..**..*****.* 17
12 Gas productionand oxygen and carbon d~oxide content of produced
gas from we1120A ......................... ma9 ....vs*.s .*.-.**** 18
13. Gas productim and pxygeh and carbon dioxide content of produced
gas from we1121A ................................................ 19
14...Gasaproductionian&oxygerlerl
and.carbon-dioxide-contenc.of.produced-.
.
gas from well R,................. ................................
15. Results of radioactivekrypton 85 gas-tracertest..................
16.. Results of helium gas-tracertest..................................

ii

Comparisonof permeabilityand porosity,Experimentalwells

2 and 3..........................*.4.*...**a******~****~********
Comparisonof porosity exponent,formstion-resistivity
factor,
bulk density,and effectivesand-graindensity,Experimental
wells 2 and 3....**********
.*,*#***.*~#**..**....****....*..~*~.
19. Gas-oil relativepermeabilityratio--totalliquid saturation
relationships,Experimentalwells l? 29and 3.*.~~*,...*00.0.*.*
20. Formation-resistivity
factor--porosityrelationships~
Experimentalwells 2 and 3.......................................

TABIXS
1. Results of air-injectionflow surveys,Experimentalwell 2.,......*
2. Analysis of crude oil from Hunter well 17;.........................
3. Gravity and viscosity of crude oil from thermalexperiment
pattern wells.........i..................d.......................
4. Backgroundgamma radiation and helium content of produced gas.,....
5. Krypton 85 and helium tracer results...............................
6. Gas productionand gas-tracerrecovery.............................
7. Calculatedvolumes of air dilution from wells outside the
experimentalpattern.C.ocOC-Cc.g-C.O....gO,~Ot.c....-...*.Dcoo*
8. Results of gas-productiontests in Experimentalwell 1.............
9. Reservoirtemperaturesdeterminedby petrographicthin-section
analyses$.Experimentalwell 3c.9a*.***a**.*
.D-ta~.Be9SO*9*
..*=...
10. Carbon-hydrogenanalyses,Experimentalwell 3.......*0cCSCOC.CSC0.0
11. Gross calorificvalues of core samples,Experimentalwell 3........

.
. -.

___

-.

.- ..-

.. ...

UNDERGROUND
IN

THE

COMBUSTION

VENANGO

FIRST

OIL-RECOVERY
SAND, WARREN

EXPERIMENT
COUNTY,

PA.

by
N. A, Caspero,~

W. T. Wertrnanr2

and

W. E. Eckard3

ABSTRACT
A thermal oil-recoveryexperimentwas conductedin the Venango First
sand, GoodwillHill oilfield,Warren County, Pa., about 10 miles northeast of
Titusville. The Quaker State Oil Refining Corp., Bradford, Pa., performed the
experimentwith assistancefrom SinclairResearch,Inc., Tulsa, Okla., and the
Federal Bureau of Mines.
The GoodwillHill field was discoveredin 1885. Primary-production.
operations continueduntil 1930 when an afr-gas tn~ectionproject,whtch fs still
in operation,was started. At the locationof the combustionexperiment,the
reservoiris 400 feet deep and 30 feet.thick,with wide permeabilityand
porosityvariations. The crude oil is typicalPennsylvaniaGrade paraffinbase oil with a gravity of 43.9 API and a viscosityof 4.55 CP at 60 F.
The well pattern for the experimentconsistedof 5 producingwells,
enclosingan area of 1.28 acres, and a central ignition-injection
well. An
air-nakuralgas mixture was ignited in the ignitionwell on July29, 1961.
Ignitionwas followedby injectionof air and natural gas at various rates.
The experimentwas suspendedJanuary 5, 1962, when it became evident that
economicoil productioncould not be achieved.
There was no conclusiveevidencethat self-sustainedcombustionwas
achievedor that combustionincreasedoil and water production. Aslight
increase in liquidproductionwas attributedto stripping&ctton caused by the
air-injectionrates, which were higher than those normallyused during air-gas
injection.
lSupervisinggeneral engineer,Federal Power Commission,Washington,D.C.,
formerlywith the MorgantownPetroleumResearch laboratory,Bureau of
Mines, Morgantown,W.Va.
2Project leader,MorgantownPetroleumResearch laboratory,Bureau of Mines,
Morgantown,W.Va.
Project coordinator,BartlesvillePetroleumResearchCenter, Bureau of Mines,
BartleSvill@;Okla.; for-rnerly-assfstantchieCof-the-Bureaus
llorgantom ------
PetroleumResearch Laboratory,Morgantown,W.Va.
Work on manuscriptcompletedJuly 1963.

I -

.-

.,

Additionalexperimentsusing the undergroundcombustionprocess in highgravity,low-viscositycrude-oilreservoirsis consideredwarranted.

INTRODUCTION
During 1958, the Bureauof Mines startedlaboratoryexperimentsto determine if the in situ combustionmethod of oil recoverycould be appliedto the
high-gravity,low-viscositycrude oils found in Appalachianarea reservoirs.
In situ combustionis a thermaloil-recoverytechniqueby which movement
(or production)of r~sidualhydrocarbonmeterial, usually consideredunrecoverableby currentproductionmethods, is achievedby a seriesofcomplexdi.splacementprocesses. The drivingmechanismsderive their energy in part from
water of combustionand the vaporizationvf water and oil that occurs inanediately ahead of a high-temperature
zone. Some crude oil Is consumedin the
presence.
of free oxygen. Oxygen to supportcombustionof the residualfuel is
suppliedby injectingair into the reservoirthroughthe ignitionwell. The
thermochemical
reactionwill continueas long as requirementsfor combustion
are met,

Appalachianarea secondary-recovery
productionmethods (4)4 afford several favorableconditionsfor the applicationof in situ combustion. Several ,
areas that have been developedintensivelyfor air-gas injectionoperations
have high-capacityair compressorsavailablefor operationat relativelylow
pressures(200 to 300 psig). Closewell spacingon propertiespreviouslysubjectedto air-gas injectionreducesnecessarydevelopmentcosts. The.relatively shallowdepths of the oil-bearingformations(300 to 800 feet) also
reduce developmentcosts. However,this shallownesswould limitair-injection
pressureand thus could preventthe injectionof sufficientquantitiesof air.
The price of PennsylvaniaMiddleDistrict and otherAppalachianarea crude
oils providesadditionalincentivefor increasingproduction.
Considerablelaboratory(~, 10, 13) and field (~, 14, ~) investigative
work has aided Ln the evaluationof the in situ combustionprocessand the
determinationof variousoptimumrequirements. Previousefforthas been made
.to developtechniquesfor recoveringcrude oils of much greaterviscosityand
lessmobilitythan the PennsylvaniaGrade high-gravityparaffin-basecrude oils.
This field experimentwas made in the shallowVenangoFirst sand reservoir on the Hunter lease,GoodwillHill field,Warren County,Pa. The Quaker
StateOil RefiningCorp.,Bradford,Pa., owned and operatedthe propertywhich
is locatedabout 10 miles northeastof Titqsvtlle, Quaker State conducted.
the
experimentwith assistancefrom SinclairResearch,Inc., Tulsa,Okla., and the
Bureau of Mines.
The primaryobjectiveof the experimentwas to determineif a heat waye
could be ini~iated. The secondaryobjectivewas to determinethe advancement
,-.
__
-rate-of--the-heat
wave and the-optimum-requlretnent-sfm
suseain-combustion----- ----1.

k-

Underlinednumbersin parenthesesrefer to items in

. the list of referencesat
the end of this report,
~

.-.

II
I

3
The experimentdescribedin this report and other preliminaryreports
(~,~,~)
representsone of the more recent attempts to study the effectsof
combustionin a PennsylvaniaGrade crude oil reservoir. A previous test was
reported in October 1959 (~). Additionaltestingand evaluationis required
before the feasibilityof the in situ combustionmethod can be determined.
ACKNOWLEDGMENTS

The cooperationof the officialsand employeesof the Quaker State Oil

Refining Corp.~ SinclairResearch,Inc., and the Bartlesville(Okla.) Petroleum Research Center of the Bureau of-Mines is gratefullyacknowledged. t
Special thanks are extendedto Q. E. Wood and C. V, Gray, Quaker State
Oil Refining Corp.; L. W. Emery and H. Wolcott, SinclairResearch,Inc., and
F. E. Armstrongand W. D.
Howell,
BartlesvillePetroAVERAGE
ROCK
USTOMARY
leum
Research
Center,
SYSTEM
SERIES
GROUP
THICKNESS,
COLUMN
iANDNAME
feet
Bureau of Mines, for their
. . .. . . . . .
significantcontributions
Berea
:;:.:..
30
Corry sand
-,,., . . . .. .:.:
.. .
during various phases of
. .. .. . . . . .. . ...
.1.. . .. . .... .....
..
. :
this experiment.
.. . . .
,-. . ,, ..,. .:: 4.
Conewango

::.::.-.:; ::
., ..,:.,..,:.
..,..,..,..
.
. . . .. . . .. .

90

Cussewago
sand

95

Riceville
shale

..
:.,
., ,. ...~.,....

Riceville

.. .. ... .. .. ... ... ... .,:

.

~;<:~~
Venango
85

E4 .:. :,...,.,
EC-g

-g%

...........

1st
sand

:.

,? . . .. . ..O
,
. .
. ..

:,
.::...
..,.
,,:.
... ,. .,.
:

Saegerstown
shale

50
t.

. . .. .. .... .,..
:.,. ,,. .... .

k?=
?s~
3

:::
.. .....:..:
.... . ,,.. - ....
: ...-.... :.:... .

Red Velley
send

25
25

Shela

25

~dng;g;
n

6G

Amity
shala

.. :.. ... .....,,

0
~-g

l-)g!~

,=.

g--

: . . . .. :. ....
: ::.:! :..

%.

.;:. . :..

::,;:;;::;::
. . .. .. . . . . . . .
...., ,. ; ..: ,,.., . .

: : !. ..,. ::...,.

FIGURE

Venango
Id Stray san

40

35----

l. -Columnar Section of Geologic

of the Titusville
Quadrangle.

Venango.
3dWnd-

Formations

GENERAL GEOLOGY, FIELD

DEVELOPMENT,AND
PRODUCTIONHISTORY

The First sand of the

Venango group (hereafter
referred to as the First
sand) is of Devonianage
(fig, 1). The Venango
group of oil pools is generally classedas stratigraphic and characterized
as sand lenses embedded in
less permeable sandstone
and shale (~). In the
GoodwillHill field, Southwest Township,Warren
County, Pa., 90 to 100 feet
of Ricevilleshale with
occasionalthin silty sandstone beds overlays the
sand section. Below the
First sand there is about
50 feet of dark Saegerstown
shale (~). About 200 feet
be~dwthe-rirst-~and-ia
the-...
Venango Third,Stray! or
upper intervalof the
Venango Third sand.

!,
......

. ... . .

.. .

----

@9

;:;::, ;.::?.
+,3

To Tltusville

,,.::.:$-~
,:.;:>

. ..

,-

FIGURE

Enterprise
..--

2. -Goodwill

Hill

Warren County,

--
VENANQOCOUNTY
1
-2

ToPleeeentville

Oilfield,
Pa.

~
let sand productive limits
=3deandpmductlve
limb

Seek, miles

Southwest Township,

About 37 feet of the

80 to 85 feet of First sand
present is consideredpay
sand, and the average depth
to the First sand is 400
feet. The First sand lies
in a northeast-southwest
trend about 50 miles long
and 10 miles wide. The
interbeddedshales, sandy
shales,and tightly
cementedpebbly zones are
essentiallynonproductive.
When present, the pebbles
are usually in the top and
bottom of the formation;
however, fine-to-coarsegrained,pebbly sandstone
sectionsare known to exist
at differentintervalsof
the pay sand throughoutthe
field. The First sand is
mostly light-graysandstone
with occasionalthin and
discontinuousstreaksof
shale. The highest oilsaturatedsectionsof the
First sand usually appear
in the lower part of the
formation. These factors
and the omissionof the
First sand from many
drillerslogs make correlationsdifficult.

Figure 2 shows the irregularly shaped area comprisingthe GoodwillHill

field. Drillingrecords to 1885 show that the First sand was fgnor+dwhile
the lower Third Stray was being developed. That developmentcontinuedinto
the early 1900s,at which time the average well densitywas one well to 5
acres. Vacuum was applied to the field for about 10 years and effecteda
slight increase in oil production, In the early 1930s,the field was drweloped for air-gas injectionon a slightlymedifiedseven-spotpattern. A substantialincreasein productionresulted and was attributedto accidental
injectioninto the First sand.
Figure 3 shows the oil-produc~ionhistory and well-developmentrate for
..the..Hu&r..and
Campbell..leases..
anChehe adj~ining Stanton lease... Oil .produc=..._ .
tion before 1902 was estimatedfrom initialproductionrecords after the wells
has been shot with nitroglycerinand by the well-developmentrate. Oil and
water produced from the First and Third Stray sands was allowed to mix; consequently, individualproductionrecordsare not available. Total oil

60-I, rlllgul l qllf n ll

1[

- 160

50 .

InjectiOn w@ls

-140

40

120

I
/

30 -

Producingwells
~-d

-. /

>-------

I-%-f

\
\ --

20

I
I

I
;

\ -----

,.--/

- 100

- 80

/
- 60

~-~

Air-gas injection
Estimated

started

564,000 bbl by
air-gas injectkm

Oil production vacuum operations

20

- 0

v
i~

Estimated
----

430,000bbl byprimary and vacuum

o~

1930
1920
1900
191O
1890
FIGURE

40

; /\L\
$/
10 -/

----

/(
/

3. -Oi[-Production
History ond Well=Development
Campbell, and Stanton Leases.
.

1940

prima~decline.
-.

,
1 1 1 Ill,,
1950
1960

Rates of the Hunter,

productionfrom the Campbell,Hunter, and ~inton leases to 1960 was approximately 994,000 barrels. Estimatedtotal oil,productionfrom 1885 to 1960 by
primary and vacuum-secondaryrecovery operationswas 430,000 barrels,or ayut
additionalbarrels of
43 percent of the total recovery. Approximately564,000
oil were producedby air-gas injection.
.
Total oil in place, on January 1, 1942, for the Colorado-GoodwillHill- Grand Valley field has been estimatedby Lytle (~). He reports that 2,750
acres of First sand contained17 million barrels of oil, or about 6,200 barrels per acre. Using an average sand thicknessof 30 feet, about 206 barrels
of oil was contained
in each acre-footof sand at the beginningof 1942. The
results of core analysis,from Experimentalwell 1, which was cored with air
during November 1960, show 243 barrels of oil per acre-foot.

.,

PREPARATIONFOR THE EXPERIMENT

..
.
Air-gas injectionand production-wellcharacteristicsindicateda highly
permeablereservoir sectionextendingabout 4 well-locationswide (900 feet)

..

FIGURE

4. = Structural

Contour Map of the Top of the Effective

Sand in the Vicinity

Venango First

of the Combustion Experiment.

through the property in a northwest-southeast,

trend; the test site wes located
in the approximatecenter of that trend. Figure 4 is a structuralcontour map
showingthe top of the First sand in the vicinity of the experiment. In November 1960, the proposed ignitionwell, Experimentalwell 1, was
drilledwith cable tools to a depth of 394 feet. The First sand and adjacent
formationswere rotary diamond-coredwith air to 444 feet and 7-inch casi;g
was set at 411 feet. The top of the First sand was locatedat 406.1 feet; the
bottom was at 436.6 feet. After the casing was cemented,an air-injectionsurvey showed that air was leavfngthe boreholeat the casing seat. Rather than
attempt to rework this well, a second ignitionwell (Experimentalwell 2) was
drilledabout 225 feet southwestof Experimentalwell 1 (figs. 4 and 5).
Experimentalwell2 was drilled with cable tools to 375 feet, then a
cable-toolcore was cut with water-baseaquagelmud to 436.5 feet. A 5wl/2inch string of carbon steel casing was set at about 390 feet and cemented to
the surfacewi~h a ~ernpergture-~esi.stant
cement (50-50mixture of Pozmix A
and regular cementswith an admixtureand accelerator),
TWO sp~nner-typeair-injectionflow survey8were run in Experimentalwell
2 with an inflatablepacker that could be set at l-foot intervals. Some

Experimental1
* Elev 1544,8

20A
Elev
1543,3

I
/
/

Experimental 2 \/

Elev
1529,3 ~F

Experimental 3

Elev
1531.9

/b\J&
,

,/@

Elev
#38,25
@

\
-

11.
Elev~$25,5k
\

8>
&

/
\

. Producing oil well

@

Air-gas iniection well

Observation well

Ignfion

LEGEND
A

R
Elev
1530.4

\\

well

&

~ 25

50

100

Scale, feet

Elev
%21.7
FIGURE

5.-

Combustion-Experiment

Well Pattern.

bypassingoccurredaround the fnflatablepacker during run 1; however, the

bypassingwas evaluatedand the results of the two runs were used to calculate
the injectivities,=flux, and percent of total intake for various reservoir
intervals. Metered air-injectionrates and pressuresand an average reservoir
pressureof 30 psig were used for the calculations. The results of 24-hour
pressure-builduptests on the pattern producingwells and the ignitionwell
were used to determinethe average reservoirpressurebefore ignition. Table
1 shows the calculatedresults for each flow survey.
Severalpacker settlngewere made below 403.5 feet, and the air injectivity was determinedto be zero. About 19 percent of the injectedair entered
the formationbetween the casing seat at 390 and 395 feet, and 81 percent
entered the formationbetween 395 and 403.5 feet.
Standardcubic-fee,t-ofair injected;per,day
divtded-bythe-differencein-the-squarasof the injectionand average producing-wellbottom-holepressures.

.,,

.
eb

8
TABLE 1. - Results of air-iniectionflow survevs%
Experimentalwell 2
Reservoit
interval,
feet
390-391.....
391-393.....
393-395***,*
393-399..*..
395-397.....
397-403.5...
399-403.5...

Run 1
Run 2
,160,000 scf/davat 124 psig
83.000 scf/iiayat 110 Psig
Injectivity, Flux, Percent Injectivity, Flux, Percent
scfldayl
scfldayl
scf./hr/of total
scf/hr/ of total
intake
feetz
feet2
A (PSi)2
intake
A (psz)2
.
.
0.80
182.5
4*3
104.2
5.0
117,1
10.6
.91
1.41
.
.
88.0
8.1
L*O5
.
6.93
265.0
38.1
,

.
.
162.5
15,1
1.95
.
.
.
66.2
221.0
8.60
.
9*53
486.0 .52.6

Results of core arialysisfrom Experimentalwells 1, 2, and 3 (observation

well cored after completionof the experiment;locationshown infig. 5) are
shown in figure 6. Electric and nuclear logs for wells 1 and 2 are shown in
figure 7.
Average water saturation
calculated
from the electric logs run in Experimental well 1 was 30 percent less than that obtainedfrom core analysisa,nd
decreasedtoward the bottom of the sand.
Average oil saturationfrom core analysiswas 23.4 percent of the pore
volume, or 243 barrels per acre-foot. lhg interpretationsof water saturation
and free-gasspace indicatedthat average oil saturationapproaches35 percent
pore volume below a depth of 410.5 feet.
J!he
top of the l?irstsand in Experimentalwell 2 was establishedat 391.3
feet and the bottom at 421,2 feet. Air permeabilityranged from 0.1 md to
246.7 md, and averaged 70.8 md; porosityranged from 1.7 to 29.5 percent,and
averaged 19.1 percent. Porosityof 10 percent, determinedfrom the dualspacedneutron log, probably resulted from.insufficientliquid saturation.
Below the depth of 403 feet, porosityvalues obtainedby each techniquewere
in reasonableagreement. From 403.5 feet to the bottom of the sand a 50percentwater saturationwas calculatedfrom eJ.ectriclogs. This value was 20
percent lower than the results from core analysis. Water saturationcalculated from electric logs was higher for Experimentalwell 2 than well 1. Oil
saturationfor Experimentalwell 2 did not vary considerablyfrom the value
determinedfor Experimentalwell 1.
An atmosphericdisti.llationibythe Bureau of Mines,method,was made on a
sample 0$ producedcrude oil from well 170 At the final boil%ng point of
527 F, 54.3 percent of the fractionshad been distilled,leaving45.3percent
--residuum;-the
balamce-was-loss.---!lhe
results-of.tk--analysis
ar.e.
showr.in........ ..
table 2. Severalother produced crude oil sampleswere tested at 527 Fby
the ASTM distillationprocedure,and the residutim
varied between 40.5 and 53.0
percent. The paraffinpoint for crude oil samples from the pattern producing
wells varied from 43 to 54,5 F.

.. ....
,.
4,- ..--.
.. ........:...._:._
:.= .....-.
_.
..=____
. .._
-.-,.
-.;.,.

.
,

,--

f
*S

11%

11ss

I-1 I

1 I

?5

32

w.a

1-

4Wd
-

Sa9.s

4%6 -

W3

3933

1199

41OJ -

35.2

41zd -

397.3

414B -

3s23

t
llm

In

1136

416s -

401.s

41SJ -

MM

con -

1124

4i28

11s2

kg

11*

424s

111s

426.9

1116

426.6

1114

4*R

1 !
,.,
..,,
... .
g~==6
%%-..

1112

4926

1110

4%*

l!m

43M

C--459

a,=

Sn!m Vebbles

mob

413.3

bfuk

wt. Slwly,an&
Sl@hltytill,.
CCWs

417.3

.:

419,:

00
a.

pnncs
MM.-a4213
Wdysnam

4233

tl
H,rd. mmml.

1 I , i@f J
1 I I I I I I 3340~M;
70S0
mvdmwce.t
SoUwsmw ihcl,
Uwwdml

11

::?

~
6

ti3040t030709Q
w?J3SWUIUTW3,

30 #?~
F+ftsnt

L0

m!%
milt

GPHlmentsl Well 1
4 disiwmhwy
smd with air

.. . . .

,...,

------

,.:

..

-FIGURE -6

- -.

1 11 I

I[I

..
![

-E!

1 1.

Eiii

3s9.9

I I I 1

..

.,.

S9noytill

.. . .. .. .

,.

Tepdund.llnn
Calmo

r4n4,

.... .

391.9

mu

393.9

::0.:.,:.
. .. C.... D

<

Kj&tw#rw

34S9

... . .
.,:, ......... .. .

Nmmd.

finq.hm.
uay wd.

WY

dnh*O4W
M

md

d w

NotV und

~~.::.::
.: .,,:. .
,...;. .. ::,:.. .,
. .. . . . . . ..
. .. .. ,. ..,.:
.,.: .,:. ,.,.
. .. . . . .
.,.
.

397,9

~Q

.....
....
....
..
.::...,
.
..,,
:.
..,.,:
.,.
,.,AO1.9
,,., .: ..,.::
,, . . . . .

... .: :.. .,,

rH..-.,
,.:.,,
.,......,.

,.,

aEt9 -

. . . .........
.. .. . ,,.,.
. ~,:o..,.
IWO, 999rS8#,8nc4VW Im4, ,0...,
.,
..,-., ..*:
*,* $+bEut *M Sh,ll
..e; :::4...
,C#mlcrll
Malvnl to Cwrlqmk,
VW
WM. mbbln mrcJ1rW md
W& *h
SOiqw$,&
cd
F@ 10IIWlm.pmti,
#$xY

gnr

c08rManirM WM. Ih!lv am

mbbk
rlmd

4C5,9 &

!!!!?
r
,.. ,b..
..O::.
; ... ,,.
4a9
..,..,,:g;a
... . ... .. .

:*,.;

.,

0.
~
: ,. :,..

411,9 .

..

WV -m, ncauut

41S.V -

. .

tuq.imd,
ctun. gw urd,
LW40iilhlm
Shtv WM. t!m hkm
9uRttrJ @4kS

..
.:.i:::,: ::
.:.,,,.,..,
; : ,.
,., . . .. . ..

41s.9 -

W.9

419.9 -

mulldrd
vu ud, hw
9m9uPlW9$, Iad&l)..m

nOum Ofurd.11N9.l
W5ylmlt
IuIellmk
Ir#!.JIW,l
9mk9n 9hIh
.w&m@5.

1 (,,,
10 #&
P9rMat

15

30

lm
UR KRMt48%W,
mlhdmyl

w
Iuu
9alm:L$a4awtamm20
wATERSATUFATION,
9.4.4,~l~N,
Fu.rdum FWml

h9f5, my WM.

.:. .,.;,,.:.... ..
:::.:: .g.:
.. :.;:.,.:

U1.9 .

.
,..., ..,:.

41a9

Et
.,, :.:...
... . ,.,.,

4259

I_ur
,alalw

10 &~

k% *

I
Q

I
lm

I
m

ul
J
imoloa)o

UR F&&Ll19,
EQU&!llmlk#4,

PI

Wc$nt

Expwimental W&ll 2 (Ignition Well]

C5bbtml

FIGURE 6.,. ,. ..--_. .,_

~m+llmlW9113
4 di9mwlQ$oMry

cored wilh wa19r.lwm mud

Core Analyses,

LLJl
loba
.92
W&f*mm w Ml

W&z

Experimental

Wells-1,

2, and 3.

.. -.

WI*

(obW
WI

1 III

I 1

II

S19.9

k/

$?1.9
Sm

39%9

f&vHnm.

till M.m

,..
:.,,
1:
r
,.

.,

,..

to mwmwmt

*ml W*, m

1 tm

11-n

ml

. m

k!
,,,
.

Fw,m+d
lQhl@lnl Urd
rmfmnta
lot&mm.
Iwm
mwd mu Ud%f.+,
h II 61F0Mil

mw

llza

.:- i
.? i
..,,

..

..
,.,.
.

11?2
~
:

. :

Mm
llM

. .
.. .

--

. .. .

,IIIJ
1114

...

..

.,,

...:

ml

1110

..:

Ilm

IIM

*
Pmmn-7.

I
m
m

F&&Y.

X9

muli$mql

Al

w~l~~

ml

Olomolozommmo
GiL

lW

MATI!JWURATICU.

smm?ccd%

@m@
Enp9fimM181 well3 @JiWnuc+l
Wtlll
4 dimxdfc.hryCM VIMI
vmWIM4mud

. .

,.

...

120

Ma

MCI.

10

1155y

399.E
f-

1150

394.s-

1145-

399.0 -

1140 -

4MS

1135 -

40ss

1124-

414,8-

Ii!
d 1125-

~ 419.s-

<

1120_i

Rm ahmt Mz/M

M 2S F - 1.S4

424.s-

i?
S
d 1115-

4s9s-

1110-

434.s-

1100-

439a-

Ilco -

444.s-

1099-

449.s-

low-

4n.s-

.
I

(MMM;;!$.;

~
m
NEUTRON LOG, cdikl
1s Slmcing

16
100,

-)
~J

,~

-H
SPONIANEOIJS
FQTENTIAL L&?,
IIUNiwoiis

l,W
units

NORMAL LOO,
ohms M2/M

w
6UA$! LO13;
ohms M2/M

Espxlmentol Well 1

FIGURE
.-.

-.-.

..-.

..

-.

.,

~--l

..$
,:

:
j

(
1,
~.,.
*

*-... .-

353(

.
)

Rmf, ohms M2/M tt S2 F -4.70

R-ohms
M2/M at 52 F -MO
Rm, tires M2/M m 53 F -5.20

:i
*#t#**

(
<- %

,,
!
\
:,

1
I

540
NORA&~ Lo@,
ohms M2/M

I
w

QUA;: LOO,
ohms M2/M

L-J_J474849s
lEMf&+l#RC

700

12

@14MA 6WY LW,

bR/hr

CWWRWLW,

.
I

W SMclm

&Q._i
----

-k%+

1
0

tloRhiLp#

2FQNTANEOUS
FOTEm~TELO&

*ms

M ./M

600

24S $@cing

NEUTRON LOO, cali5ratsd unils

Eqw@ntd
FIGURE
. .

..

7,.

Well Logs for Experimental Welk 1 and 2.

_.. .

_.

....- ... .

-.

-. 3

374

~-

- 379.3
- w.

339.3

394.2

39s.3

404.3
3

~
4&!,38

- 414.3

- 419.s

4243

42s.3

- 424.2

-w+

3FQNTANEOUS
FWWENLOG,

+.&&
.....

o~

o~.

9U4RD p
C4m14 M /w

N~OR.!#~

6054

4ss.2

62

TEMFfRATURt
LOQ F

34. ,@g

NEuTRON L03, cWEfated unik

Ex@mentalWell

.-

...

_- _.-

. .. . --

. . .. ...

1155

1
1130

1145

1135

1
I130

1125

1120

1115

1110

!
1103

1100

1035

X96

11
TABLE 2. - Analysis of crude oil from Hunter well 17
GENERAL CHARACTERISTICS
Gravity, specific,0.812
0 API, 42,8
BS&W,z volume-percent,4,0
Color, green
Pour point, * F, below -5

60 F,
70 F,
80 F,
90 F,
100 F,

ViBCOSftY~t

Gravity,
,

6.16 Cs
<5.35CS
4.68cB
4.14 Cs
3.69 CS

5.00 Cp
4.32 t?+
3.76 Cp
3.31 Cp
2.74cp

DISTILLATION,BUREAU OF MiNES ROUTINEMETHOD

First drop, 183 F

Distillationat atmosphericpressure,745 mmHg

Fraction

Cut temp.,

Pertent

1.*..******
2..........
3 m.,**m**a
4 ****o**m*
5 *...9****
6 *a*0e*s9a
7 .**.****.
894*bamm***
9..*.......
2109..0. maae

122
167
212
257
302
347
392.
437
482
527

.
.
0,3
2.4
1.5.6
10.2
, 5*3
6.1
6.5
7s9

Sum,
Percent
.
.
0.3
2.7
18.3
28,5
33.8
39.9
46.4
54.3

Sp gr,
60/60 F
.

0;;16
.717
.730
.751
.766
.778
.790
808

0 API
60 F

.
66.2
65.9
62.4
57.0
53.1
50.3
47.6
43.6

Correlation
index

11.0
9.4
13
14
14

14
18

APPROXIMATESUMMARY
, Percent
Yield
Light gasoline....................................
0.3
Total gasolineand naphtha.......+..... .........
33.8
Kerosinedistillate
................. .............
20.5
.
Gas oil .**********
.m*m9****9**9,*m*a********8*
Residuum.........................................
45.3
Distillationloss................................
.4
lBasic sedimentand water,
aFinal boilingpoint.

SP m
0.716
.741
793
.
.864
-

0 An
66.2
59.5
46.6
.
32.4

The gravityand viscosityof crude oil samplesfrom pattern test wells

were determinedat various temperatures. The results of the tests are shown
in table 3. Crude oil produced from well R apparentlyhad not been affected
by air-gas injection.
The air- and natural gas-injectionsystem to the ignitionwellhead
(Experimental
well 2) is shown in figure 8. The i.gnition
well was equipped ~
with a bottom-holeburner (f3)and concentricstringsof tubing for natural
gas and air injection.
!.
... .. . ________ . . .--_._.
. . .
.
.
..-

..

12
TABLE 3. - Gravity and viscosityof crude oil from
thermal experimentpattern wells

mTi-20A

2U

17

Exp. 2

Temperature,0 F

viscositY,

Cp

SpecificLravitv

42
50
60
81
93

23.77
7.15
.
4.33
3.87

0.819
.815
.812
.804
.799

43
50
60
81
93

11.80
6.14

3.86
3.54

.813
..810
806
.798
,794

34
42 50
60
81
,94

11.40
5.10
4.18
.
2.64
2.29

,799
.796
.792
.788
.780
.775

42
50
60
81
94

18.09
7.80

.817
.814
.810
.802
.788

26
29
36
40
42
44
46
48
51
54
56
60
64

.
.
.
20.80
17.78
14.80
,11.82
9.20
4.69
4.59
4.49
4.31
3.80

4:24
4.05

.817
.815
.810
.807
.805
.804
.803
.803
.802
.800
.799
.798

,Each offset producingwell was reconditionedby settinga bridge at the

top of the Third Stray and pluggingback with gravel approximately30 to 50
feet below the bottom of the First sand. Each producingwell was shot with 1
quart-of
liquid.n$tr.oglycerin.
per .foo~of,pay.sand.to r.ernava.paraffin
dep~siks.- .......
-r-.
The wells were cleaned out andequippedwith 4-inch carbon steel casingseton
a hookwallpacker at the top of the First sand. Well 4A wasconverted from an
..
k.

13
6+YI, 1.~$#:2Er.hp

~,,
~,,

4.eyl, IzR~XVg: 150 hp

.

F
~,,

TI
4-

8*YI. 1.R+XVVe350+P

Cooling ceils

I.R 80.hp single-stage

4$ x 10 booster

,
I.R Mage
Sllf$;e:lly.

I.R 10.hp single-stage

2*x
4,, au~i~a~

2
1 control bypeaa

booster

~,,

compressor

cg~g

.>

~,,

l ga;nrfpply

Standby
compressor

Control
bypass
2

*,,
1 gealine
$

Radioactk
7

tracer.~
+

gniJm

~F

--+1+WJ
2 secondary alr to tubing.
casing annulus

FIGURE

8, = Air-and

air-injectionwell to a producing well and had not been

shot previously,while the
other 4 producingwells had
been @hot two or three times.
Individualelectrifiedpumping
units were set at each producfng well.

Natural Gas=lniection System.

Downstreamfrom each
producing-wellorifice-meter
run, produced gas was diverted
through a 55-gallondrum.
Fill and vent lines and a
geiger-tubecounting device
were attached to the drum.
When the gas-tracerstudy was
made, using the method
describedbyArmstrong and his
coworkers(~); the calibrated
system served as a nonpressurized containerto continuously
monitor produced gas for
radioactivity.
The daily a~erage production per well before ignitiw
was 0.25 barrel of oil and
0.30 barrel of water. The gas
that was produced from pattern
wells and secondlineoffset
wells prior to ignitionwas
analyzed for oxygen, carbon
dioxide,and carbon monoxide.
The averagevalues,.in volume,percent,were as follows:

Oxygen...........y.......... 20.3 (range 19.1 to 20.9)

.6 (range 0.3 to 1.4)
Carbon dioxide..............
.20 (range 0.05 to
0.3)
Carbon monoxide.............
The gas sampleswere collectedand liquid productionwas gaged at the same
time that air wasbeing injectedinto adjacentpressurewells outside the
experimentalpattern.
PERFORMANCEOF THE EXPERIMENT
-.

_-

Ignitionand Gas Xtiection

--- -..

..___
. --..-..-.>..
~.,
.--.... .

For a period of 36 hours before the injectedair-naturalgas mixture was

ignited,the total air- and natural gas-injectionrate was 231,000 scfd at.185

14
psig wellhead pressure. The heating value per cubic foot of mixture injected
was 15 Btu. The natural gas volume was 2 to 3 percent of the total gas
injected. Natural gas was analyzedperiodicallywith a combustionOrsat apparatusp and air contaminationof the natural gas was detectedoccasionally.
Laboratoryanalysis of an uncontaminatednatural gas sample was:
Volume, percent

Component:

coZ**mm*..,
..***
02....................
C2H6........**.*......
Cq ...................
CO...*..,.....**.*..*.
N~0..*.*b******.*..**

0.1
.3
26.4
72.0
.0
1.2

Each cubic foot of the natural gas would require 2.35 cu ft of oxygen for
completecombustionand would generate 1.25 cu ft of C02, Heating value of
the natural gas varied from 825 to 1~180 Btu per cu ft at 60 F and 30 in. of
mercury,and generallyaveraged 825 to 1,000 Btu per cubic foot.

I
;.
,

i
1
I
1$
i
!
z

The air-naturalgas mixture was ignited chemically(~) on July 29, 1961,

by personnelof SinclairResearch, Inc. The temperatureat the burner shell
increasedfrom 60 to 460 F in 25 minutes. About 12 hours after ignition,
the injectionpressurebegan to increaseslowly and reached 500 psig in 57
hours, then graduallydecreasedto 370 psig in 12 more hours. During the
gradual pressure increase;the air- and natural gas-injectionrates varied
from144,000 to 240,000scfd and 3,360 to 6,000 scfd, respectively,while the
heating value per cubic foot of injectedair-naturalgas mixture wa$ maintained at about 25 Btu. Figure 9 shows the air- and natural gas-injection
rates and pressuresthroughoutthe experiment.

The air injectivitydecreasedfrom 8.7 to 1.2 after burning natural gas

for one day, and further decreasedtoO.86 in 3.5 days. When the injection
pressure declinedto 370 psig, the injectivityincreasedto 2.0 and remained
at that value for the period that air and natural gas were burned in the
borehole.
Air and natural gas injectionrates were maintainedrelativelyconstant
during the next 9 days at 360,000and 6,000 scfd, respectively,and at an average injectionpressure of 420 Psi.gm Two weeks after tgnition,the produced
oxygen had decreasedto 16.4 percentbyvolume and the carbon dioxidehad
increasedto 2.3 percent by volume. The air-injectionrate was increasedto
about 330,000 scfd during the secondweek. The injectionpressure continually
decreasedfrom 500 to 400psig. Natural-gasinjectionrate was maintainedat
about 2 percent of the air-injectionrate.
During the 2-week period that natural gas was burned in the wellbore, the
volumes of air.and natural gas injectedwere 4 million and 83,000 scf,,respectiively.Complet-e
combustionofthenatura,l-gas
wouldgenerate99-million
Btu-, --
or 3.,25million Btu per vertical foot of First san,d. The averageheat-release
rate was 290,000 Btu per hour,
,

15
900

I
I
Transferred
fromwell
bore to formation
Period of flashbacks

800 -

~
m
8

[it
Average injection pressure

- 550 a

500g
8
450 g

~a

700

600

lKlll&/

400g
.Q

Natural gas-injection rate

.+

24

~K

l--v .
2fJLy30
AUG

SEPT

OCT

1961
FIGURE

9.-

Air- and Natural Gas-lnlection

Experimenta[Wel12.

NOV

DEC

JAN
1962

History of the Combustion Experiment,

.
,
I

1
I
1
,.
,

pressure before ignitionwas 3@ p=ig, Recompletion

The average reeervoi.r
of well R and conversionof well 4A from an air-injectionwell to a producing
well was completed3and 5 days~ respectively,after Ignition. Wells 17, 21A
and 20A were each producingat a rate of 10,000 ecfd, or about 5 percent of
the total injectionrate, with a casing-headpreesureof 18 to 37 psig. When
Ehe casing-headpressure of the wells was reduced and reconditioningof wells
R and 4A was completed,total.gasproductionwas 107 percent of the 330,000
scfd injectedvolume.

wells was discontinuedapproximately4 days after ignition:

scfd of air had been injectedinto the 3 wells at pressuresbetween 65 and 150

16
psig. Wnen air injectioninto offset injectionwells was discontinuedand the
wellhead pressuresreduced to atmosphericpressure,the volume of produced gas.
was 110 percent of the total gas injectedin the ignitionwell and increased
to 133 percent after 15 days.
The oxygen and carbon dioxide content in the produced gas 2 days afterignitionwas 19.0 and 0.65 percent, respectively. Reducing the back pressure
on the producingwells, increasinginjectionrates, or stoppingair injection
into the offset pressurewells, all of which occurredwithin the same day, did
not change the producedvolume of oxygen or carbon dioxide.
Producedgas was analyzedat least twice a day with Orsat apparatusor a
chromatograph$canalyzer. Figures 10, 11, 12, 13, and 14 show the daily average oxygen and carbon dioxide content of the produced gas and the gas-production rate of wells 4A, 17, 20A, 21A, and R, respectively.
Well 21.Adid not respond to the increasedinjectionrate as rapidly as
the other wells. The gas-productionrates of individualwells correlatedwihh
changes in air and natural gas injectionrates and pressuresthroughoutthe
life of the test. Wells 17 and R showed the greatestresponse.
,>

FIGURE
.

10. . Gas Production and Oxygen and Carbon Dioxide Content

.of.Produced Gas From..Well.4A . ...
_
. . .
.. .
. .

..

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

17
240

24

well
17

220-

- 22

4
\*\
/ -%.--+~<- -

- 20

200\

160
-

- 18
1 - 16

160
I

140-

120
- gg

*&
100

I
I
I
I

-H1

p10

1
1-

Tranaferrad from well bore to formation

80 -

e\
:
2

60 40 -

20

1
I

JULY

AUG

SEPT

9CT

Nav

DEC

1961

FIGURE

12

I
i

I
~:

+-.
~, Qas pr~u~iorlrate

- 14

\-

8
6

JAN
1962

11. - Gas Production and Oxygen and Carbon Dioxide Content

of Produced Gas From Wel I 17.

The oxygen or carbon dioxide content of the producedgas did not indtcate
efficientcombustionof residual carbonaceousmaterial; therefore,combufition
was transferredto the formation.
Natural gas injectionwas stopped for about 4 hours (not shown in ff.g.s)~~~
and the air-injectionrate was increasedabout 30 percent. Natural-gasinjection was resumed and the air-injectionrate maintainedto give a heating value
of 15 Btu per cu ft for the injected
mixture. The temperaturein the wellbore
declinedto about 70 F. The air-injectionrate increasedfrom about 330,000
to 625,000 scfd after transfer,while the injectionpressure decreasedfrom
400 to 360 psig. The calculatedinjectivitywas 4.5 after transfer. The
higher air-injectionrate was maintainedfor about4 days. Injectednatural
gas was,maintainedat about 2 percent of the injectedair and averaged 10,000
scfd. The increasedinjectionrate did not reduce the volume of producedcarbon dioxide.

,-.

The air-injectionrate was variedbetween120,000and 530,000 scfd for 11

more days. Natural gas-injectionrates varied between 2,400 and 24,000 scfd.
e8~~rerr8nge~e$rO~ZOQZOQ
.p+%&$.02?0.ps3g~.?~~.!h?
fnj~c?iy=_...._.
---~~e..~$z-.%nj@~$iOpr
ity decreasedfrom 4.2 at the lowe6tair-injectionrate to 2.5 at the maximum
injectionrate. The average produced o~gen waa 14.5 percent by volume and
,,

18
220
I

30
JULY

30
AUG.

OCT

SEW

NW

DEC

1961

FIGURE

22

JAN
1%2

12. - Gas Production and Oxygam and Carbon Dioxide Content

of Produced Gas From Wel 120A.

the average carbon dioxide ~s 3.7 percent by volume. During the period when
24,000 scfd of natural gas was injected,the producedmethane increasedfrom a
backgroundof about 0.25 percent to 0.55 percent of the total gas produced.
!l?he
higher percentagecorrespondedto 18 percent of the total volume of methane injected. When the natural gas-injectionrate was decreased,produced
methane decreased. This iqdicatedthat injectedgases were bypassingthe
high-tefieraturezone. However, the volume of injectedgas passing unreactively through the reservoircould not be quantitativelydeterminedbecause
of air dilutionfrom outside the pattern.
On three occasions the burning air-naturalgas mixture was flashedback
into the wellborefrom the formationby $ncreasingthe natural gas-injection
rate. The thermocouplein the wellbore indicateda rapid increase in temperature. The flashbackproved that there was sufficienttemperatureat some
reservoir,
int.erval,
if not throughoutthe entire section,to ignite the
injectedair-naturalgas mixture. The temperaturerequired to ignite the
mixture was about 1,000 F. When the wellbore temperaturebegan to increase
.
rapidly,nattral gas injectionwas stopped. Xhe wellborewas allowed to cool
before natural gas injectionwas resumed. The first flashbackoccurredabout
30 hours after increasednatural gas injection, The second flashbackoccurred
I
i...
._....an.
a.bFu.F.fiO!%!!%!
af.!@r:.n
a?Ura?.@as-t?jec!iOnas eswedf After the second

i9

22

220

r~

Trarrsferrad from well bore

to formation

200

1s0

Well 21A

160
@
- ..-

20

02

#. ,4
\
I,_ I

i
18~
E
16$
,*4

FIGURE

13, - Gas Production and Oxygen and Carbon Dioxide Content

of Produced Gos From Wel I 21A.

flashback,an air-gasmixture of 15 Btu per cu ft was injectedfor 1 day.

This volume of natural gas was injectedto determineif there was an increase
in producedmethane. Natural gas injectionwas then increasedand a third
flashbackoccurredabout 4 hours later.
It was estimatedfrom the length of time for the first flashbackto occur
that combustionof the injectedair-gas mixture was taking place within a 2to 3-foot radius from Ehe wellbore. When the flashbackoccurred~the indicated tempe~aturein the wellbore increasedimmediatelyto 1+350 F. Less
time was required for each subsequentflashback,indicatingthe relative
closenessof the burning air-gas mixture to the wellbore. During the flash-
backs, the average produced oxygen was 14.6 percent and the carbon dioxide was
3.8 percent. When the large volume of natural gas was injected,the oxygen
content from well 17 decreasedto 13 percent and th~ carbon dioxide increased
to 5 percent. However,when the natural gar+-injection
rate was increased,
bypassingwas indicatedby increasedmethane production.
Natural gas injectioncontinuedfor 95 days after combustionwas transferred from the wellbore to the formation. For 56 days, air- and natural gas-

,,

20.

130

3 8.
E

50

Transferred

;
I

30

26

from well bore to formation

40

Period of flashbacks

it

Gas traceraInjected

20 -

10 -

- 4

ON

/
- 2
1
0

20
30
JULY

AUG

SEPT

OCT

NOV

DEC

1961

FIGURE

JAN
1962

14. -Gas Production and Oxygen and Carbon Dioxide Candent

of Produced Gas From WeII R.

changed from 9,600 t6 18,000 to 32,500 scfd, with an average air-injection

rate of 480,000 scfd and an injectionpressure of about 400 psig. As the
natural gas-injectionrate increased,the quantity of producedmethane rapidly
increased;and at peak natural gas injection,about 88 percentwas being
produced.
The total volume of natural gas injectedwas 1.2 MMscf, which was capable
of generatingl,400millionBtu upon completecombustion. After combustionof
,
the injectedair-naturalgas mixture was transferredto the formation,1.1
MMscf of natural gas was injected. Total air injectedduring the natural gas
injectionperiod was 49.8 MMscf, of which 45.8 MMscf was injectedafter combustionwas transferredto the formation.
Prior to,termin~tionof natural gas injection,the air-injectionrate was
decreasedto 120,000 scfd (the heating value of the mixture was maintainedat
20 Btu per cu ft) in an attempt t.oimprovecombustionefficiencyby increasing

21
decrease,respectively. These changes in compositionof the producedgas were
due to increaseddilutionby air entering the pattern producingwells from the
reservoirbeyond the experimentalpattern and were not caused by a decrease in
combustionin the reservoirat the lower injectionrate. It was apparentat
, this time that the amount of outside dilutionwas very dependentupon injec tion rates and pressures. The air-injectionrate was then increasedto
485,000 scfd. The average injectionpressureat the low injectionrate was
225 psig and increasedto 375 psfg at the hiih injectionrate. These rapid
changes are not shown on figure 8.
Natural-gasinjectionwas discontinuedbecause of the apparent low,cornbustion efficiency,poor oxygen utilization,and bypassingof injectedmethane.
The oxygen and carbon dioxide content of the produced gas did not significantly change with increasedair injectionin the latter stages of the experiment. Because continuedair injectiondid not Improveconditions,the operation was terminated.
A total of 38 MMscf of air was injectedafter natural-gasLn.jection
was
terminated. During the experiment,air injectionwas 87.8 MMscf, total,gas
(air and natural,gas)injectedwas 88.9 MMscf, and a total of 95MMscf of gas
was produced.
When air injectionwas discontinued,the ignitionwellhead pressure
deer.eased,from
540 to 250 psig in 1.5 hours and to 100 psig in 24 hours, after
which the decline to 5 psig was quiterapid. Three hours after shutdown,
shutin pressuresfor wells 17, R, 4A, 21A, and 20A were 78, 64, 53, 98, and 54
psig, respectively. The rapid decline of the i~ition wellhead pressure @dicates that high-permeabilityzones.existed in the reservoir. The producing
wells were opened after wellhead pressuremeasurementswere completed. The
ignttionwell remained shutin to minimize backflowand caving of the formation.
During the experiment,oil and water productioneach averaged 0.75 barrel
per well per day for the pattern test wells. The slight increasein production probablycannot be attributedto the effects of combustion. The airinjectionrate for the durationof the experimentwas considerablyhigher than
the 50,000 to 75,000 scfd usuallyinjectedduring the air-gas injectionoperation and more likelyresulted in the increasedfluid production.
SubsurfaceGas-TracerTests

The gas tracersand-theirfunctionswere (1) krypton 85, to determine

travel time between injectionand productionwells; (2) tritium, to determine
vertical extent of the thermal front and relative degree of conibustion
in the
vicinity of the injectionwellbore;and (3) helium~ to determineareal sweep.
efficiencyand the volume of air being produced from outside theexperimental
well pattern.
(

-uAir-was-Injectedat--arate-of-485;O006cfd-at
-37>-psigfar2-weeksbefore--:the tracerswere injecteduntil gas productionfrom the pattern test wells
stabilizedat 560,000 scfd. Of the total producedgas, well 17 produced29 .

22
percent;well R, 16 percent;well 4A, 22 percent;well 21A, 16 percent;and
well 20A, 17 percent. The average oxygen content of the producedgas.was 16.5
percent and the average carbon dioxide contentwas 2.6 percent.
fie gas tracerswere injected44 days after ignition. Natural gas injection was etopped6 hours before the tracerswere injectedto allow the tritium
to enter the formationwithout being decomposed. Natural gas injectionwas
resumed immediatelyafter krypton injectionwas completed, Before the tracer
was injected,severalbackgroundmeasurementswere made to determinenatural
gamma radiation.andhelium content of the produced gas. Average results are
listed in table %
TABLE 4. - Backgroundmums radiationand helium
content of producedgas

Well
21A...........
BOA...........
17 **********
R.~*..........
4A............

Backgroundgamma I
radiatfon
cpm
I tic/ft*
0.011
55.0
49.2
.010
37.8
.008
37.1
.007
36.0
007

Backgroundhelium
concentration,
ppm
50
40
t60
60
50

About 1 curie (A2Q percent) of radioactivekrypton 85 and 27 curies of

tritium in a gaseous phase were injectedin about 15 seconds. Helium was concurrentlyinjectedat an averagerateof 108 scfh, or 0.5 volume-percentof
the air-injectionrate. Helium was continuouslyinjectedat that rate for 7
days with only a few minor interruptions.
Countingapparatuswas installedat each pattern productngwell for continuouskryptonmonitoring. Samples from each well were also collectedautomatically.in evacuatedcylinderswhen a predeterminedkrypton 85 count-level
occurred. An ion-chamberdetectorwas used in the laboratoryto determine
tritium content. A helium leak detector (mass-spectrometertype) was adapted
to quantitativelydeterminehelium concentrationin the producedgas 2 or 3
times each day for 3 weeks, until helium concentrationwas almost to background. Gas productionfrom secondlineoffset wells was testedwithout tracer
detection. The results of the radioactiveand inert gas-tracer
testsare
.
shown in figures 15 and 16.

.-

The helium-detectionequipmentand samplingtechniqueindicatedwide

variationsin recordedhelium concentrationin the produced gas. The use of
this equipmentat a field locationnecessitateddifficultstandardizationprocedures. The radioactivitymonitoringsystem gave similarbu~ less severe
probl~s. The accuracy of the ratemeterswas on the order of k5 percent.
This accuracy,along with a better samplingtechnique,providedmore reliable
results.
.. _..
.._._.
.____________ ,._ .._:... .=- ,.........

Radioactivekrypton f3-rstappeared 6 hours after injection-atwell 17,

which was 129 feet from the injectionwell. Breakthroughtimes for the
*
,.

23

d-

Well 17

,.

Well R

2.0

1.6
1

121
I \(
Well 4A

8 -

o
Well 21A

0
Well 20A
:
12 8

-..

1
.

13-

TIME AITER INJECTION,daye

FIGURE

15. - Results of Radioactive Krypton 85 Gas-Tracer Test.

>.

24
6,000

(J

)0

..7:

4,000

.OO

17

Well

2,000
-d
/
I I I I I I I
o~
6,000 I 1 I I ! I I I I

4,000
-

Well R
,,

0
0
~-

o&o

.
&o

2,000

.x

oI

I 0

u~

Well 4A

6,000
o

4,000
-

>-$=;>
Xo

2,000
-

00

6,000 , ,

01

Well 21A

o
L.--J

4,000
>:

0 0

o~y

2,000
-

/
o~~
I

I
o

6,000

.
I

Well 20A

0
-=-y

4,000 -

;O

0<0

r ,

\.

2,000
- o/
\J 00
/
L
b&+T
- .<* -J 1. I I I I 1 I
2
4
68
10
12

14

1 I
16

TIME AtiFR INJECTION, days

FIGURE
<

16. - Results of Helium Gas-Tracer Test.

18

20

+
22- - r
----: -

pattern wells and the average rates of travel for both helium and krypton
tracers are shown.in table 5.
TABLE5. - -ton

85 and helium tracer results

Second breakthrough
First,breakthrouxh
Distance
Helium
from
Kr ton
inject-ion Time, Rate, I Time,
Time, Rate,
well, ft
hr
hr
ft hr
I ft/hr l..hr I ftlh~
17.....*....*
129
11
11.7
6
21.5
6.5
20
188
R............
51
3.7
179
42
4.2
4A...........
VIA..........
129
2.4
20A..........I
178
3.8
:;
Y
Well

Although the helium tracer was not intendedto determine travel time, it
is interesthg to note the brealtthrough
time for each gas tracer. The average
helium-tracerrate of travel to each producingwell was considerablymore uniform than for krypton. The initialbreakthroughof krypton was consideredto
be through the highly permeablereservoir intervalbetween 391.5 and 403.5
feet. The average travel rate for the secondkrypton breakthroughindicated
that travel was through a more uniform and lower permeabilityzone between
403.5 feet and the bottom of the reservoir.
The use of more sensitiveequipmentand continuousmonitoringfor krypton
detectionprovideda more exact breakthrough-timeindicationthan.thehelium
samplingprocedures. However, the large differencebetween mobility and diffusivityof the two gases was the most importantfactor in determiningbreakthrough time. The much smaller-heliumatom, with its correspondinglygreater
diffusivityand Wobility, encountereda much larger effecttvepore volume than
the largerand much heavier,krypton atom. The net effect wasthat the breakthroughpeak for krypton was much sharpey. The combinationof these two factorswas an apparent lag in breakthroughfor helium, although the breakthrough
times were essentiallythe same.
About 95 percent of the 16,750 scf of helium injectedwas recovered. An
assay was not made of the amount of krypton injected;therefore,the krypton
recoverycannot be established,Table 6 gives the total gas, helium, and
radioactive-gasproductionfor each well and the percentagethat each well
produced.
The recwery of gas, helium and krypton frdm t%e individualwells, as
percentagesof the total recovery of each, are in reasonablygood agreement.
When air-injectionrates were increasedas the experimen~progressed.the
individualwells produced gas in the same percentagesas shown in table 6.
Six gas samples from each pattern producingwell were analyzedfor tritium content. Tritlumwas detectedIn the produced gas from all exceptwell
.4A. .Gas sawples fi.orn.wells
17,and20A$collected48 hours after tracer injechad-thesatiekl?kiti-content.
-@s sarnple$
frornwells-R
an-d21A~-coI--
iion-,
lec~ed 24 hours after $racer injection,had 2.5 and 7times, respectively,the

..>

26
amount of tritiumthat was present in wells 17 and 20A. The analyticalresults
enough to indicatecombustion%n the reservoiror vertical
conformanceof a high-temperaturezone.

were not conclusive

TABIJ36. - Gas productionand gas-tracerrecovery

Helium production KrYuton 85 production
Production=
Scf
Percent , Scf
Percent
Curies Percent
of total.
of total
of total
17.,.......... 163,000
31.7
0.164
28.4
29,0
5,050
15.5
12.9
.039
6.7
R............. S7,000
2,050
19.8
,158
3,150
4A............ 123,000
21.8
27.4
15.8
15.8
2,520
.085
14.7
2U9*9*6****** 89,000
;131
17.9
3,150
19.8
22.8
20A........... .101.000
15,920
.577
Total..., 563,000
lDaily averagefor tracer-testperiod.
Gas

Well

A significantincreasein the carbon dioxidecontentof the producedgas

occurredduring the gas-tracertests,from about 2 to 5 percent in 10 days;
however,the producedoxygen.
remainedabout 16 percent. Air. and natural gasinjectionrates and individual-well
gas-productionr
ates were constant. Each
well showedabout the sainepercentageincreasein the carbon dioxideproduced.
Producedcarbon d$oxidefrom wells 17, R, and 4A decreasedslightlyafter
reachinga peak. Carbon dioxideproductionfrom wells 21A and 20A increased
and remainedconstant. The interruptionbf natural ga_sinjectionat the time
the tracerswere injectedmight have caused the change in carbon dioxide
production.
,
Areal sweep efficiencyfor the reservoirintervalbetween391 and 403.5 ,
feet was 42 percentsas calculatedfrom the helium-tracerresults. The calculationwas based on an assumedpressuredistribut~on,a uniformporosity,the
elapsedtime for the tracerto travela certaindistancealong a line between
the injection
and each producingwell assuminga relativelyuniformrate of
advance,and the total gas injected during that elapsedtime.
The volume of air producedfrom the reservoiroutsidethe experimental
patternwas also calculatedfor each well from the results of,theheliumtracertest. The volume of helium producedbyeachwell as a percentof the
volume injectedrepresentedthe volume of injectedgas and combustionproducts
that each weil would produce. The calculationsassumeradial uniformityof
combustionin the reservoirand that.1 cubic foot of combustiongas is produced from 1 cubic foot of o~gen-natural gas combustionmixture. Table 7 ,.
shows the calculatedresultsof the individualwell dilut~onvolumes.
Flow Tests in ExperimentalWell 1
Gas-productiontestswere made in Experimentalwell 1 on November2 to
determinethe compositionand volume of gas producedfrom differentreservoh
intervals. A hookwallpackerwas set on 2-inch tubing to isolategas production frorstwo.separatereservoir.inte~vals.
. Four packer-settingdepthswere

,.

27
Experimentalwell 2. Gas productionfrom above and below the packer was stabilized at each de~th before oRen flow was measured and sampleswere collected.
Results of the tests are shown in table 8.
TABLE 7. - Calculatedvolumes of air dilutionfrom wells
outside dhe experimentalpattern
..
Helium produc,ed,
percent of
I inlectedhelih I
30.1
17......... I
Re,o.ee.e..
4A.........
21A...,.,..
BOA........ .
18.8
I
Total.
94.9

Percent
dilutlon

Well

scfd
163,000

$l%!!l

101,000

563,000

scfd
144,500

18,500

E!?RE

90,000
455,000

11,000
108,000

26.8
19.1
10.9

I
b

TABLE 8. - Results of gas-productiontests in Experimentalwell 1

Packer
setting
depth, ft
416,0
421.9
425.8
429.9

Equivalent
depth, ft,
Exp. well 2
400 5
J406.4~
410.3
414.4

Open,flow, scfd
Above c Below
packer packer
1,100
2,110
2,170
2,130

1,510
910
860
805

Oxygen,
volume-percent
Below
Above
uacker packer
17.2
10.2
17.1
8.4
16.4
7.2
17.0
6.6

Carbon dioxide,
volume-Percent
Above
Below
packer packer
2.6
7.4
3.1
8.2
10.3
3*5
3.6
10.8

Below the packer,oxygencontentdecreasedand carbon dioxide increased,

generally,in proportionto the gas production. At 416 feet, the gas production was less than at other depths,possiblybecause part of the formation
face was sea:ed over by the packer rubber. Before the packer tests were made,
produced gas from Experimentalwell 1 had an average oxygen content of 14.2
and carbon dioxide coqtent of 4.3 percent. The significanceof the results of
the tests was the differencein compositionof the produced gas from above and
below the packer toward the bottom of the sand. Even though the gas production from Experimentalwell 1 was only a small percentageof the total gas
productionfrom the pattern wells, the results indicatedbypassingthrough the
high permeabilitysectionat the top of the reservoir. At each lower packersettingdepth the decreasingoxygen and increasingcarbon,dioxidecontent of
the gas produced from below the packer indicatethat gas productionwas influ-
enced to alesser degree by the air bypassingthrough the upper highpermeabtlitysection.

-.

After the packer tests were completed,the packer was permanentlyset at

416 feet and gas sampleswere taken daily from abcwe and belowthe packer.
During the latter stage of natural gas injection,when the rate was 32,000
scfd and air injection,was,500,000scfd, the average oxygen content from above
-and-be-lmw-the
-packerwas--l5.5
and 9:3--percenta
rid--thecarbon.dioxide..was..4.4.
.........
and 4.8 percent, respectively. When natural gas injectionwas stopped,the
o~gen content below the packer decreasedto 7.4 percent and the carbon
!>

28

dioxidecontentin the gas producedfrom abwe the packer decreasedto 1.0

percent. Before the experimentwas discontinued,the air-injectionrate
increasedto 715,000scfd,theoxygenbelow the packer decreasedto 4.1 percent, and the carbon dioxideincreasedto 9.0 percent. The oxygenand carbon
dioxidecontentof the gas producedfrom above the packerwas 15.8 and 1.5
percent,respectively. In January 1962,2 weeks after the piojectwas shut
down, the oxygenand carbon dioxidecontentof gas producedfrom below the
packerwas 3.1 and 10.0 percent,
respectively;above the packer the oxygenwas
10.8 and the carbon dioxidewas 1.1 percent. These dataindicatethat combustion of residualhydrocarbonsmay have been occurring@ the lower intervalof
the reservoir.
RESERVOIRCHARACTERISTICS
FROM THE OBSERVATIONWELL

Coringand Core Analysis

After air injectionwas discont@ued in January 1962, the reservoirpressure was allowed to dissipate. An observationwell, Experimentalwell 3 (fig.
5), was then drilledanda rotary core taken of the sandsectionbsing waterbase mud. The locationof Experimentalwell 3, on a line 14 feet from Experimental well 2 betweenwells 17 and Experimentalwell 1, was based on the calculatedburned-outradius and the compositionof gas producedfrom those
wells, which indicatedmore efficientcombustionin that direction.

Resultsof the core analysisare shown in figure 6. The top of the First
sand was determinedto be at 391,0 feet and the bottom at 423.7 feet. me
shaleand interbedded,fine-grained,thin sandstonesec,tions
above the First
sand showeda color transitionfrom bluish gray to.lightbrown. Only a fraction of an inch of theshale adjacentto the First sand apparentlywas
affectedbythe high temperatureswhich indicatesthe value of overlyingshale
as a heat barrier, The top 2.5 feet of First sand consistedof alternating
layersof coarse-grainedand well-cementedand fine-grainedporous sandstone
with crossbedding. A carbonaceouswhite crystallinesolidwas noted on the
surfaceof the coarse-grainedsandstoneafter coringwater evaporated. Qualitative laboratorytests indicatedthe pzesenceof carbonatesAn that reservoir
intervalofExperimentalwell 2. Belowthatsectionwas about 0.5 foot of
natural-coloredsandy shale. At 394.1 feet the sandstonebecame,uniform
with
some interbeddedshale. From 395 to 403.8 feetj the sandstonecore was brickred and containedconsiderablefree salt on the surface. The equivalentsodium
chloridecontentof rock samples,taken from the centerof the core after
removalof the exteriorsurface,was less than 13000ppm,

i-

At 404 feet, the brick-redsandstonegraded into a lightsbrownish-gray

sandstonedistinguishablefrom the natural-graysandstone. Some free salt was
presenton the surfaceof the core materialdown to 419 feet, but wae considerably less than in the 395- to 404-footinterval. At 419 feet, natural-gray
sandstonebecame prevalentwith thin brown streaksirregularlyspacedh.+tparallel t? the beddingplane. From 422 feet to the bottom of the reservoirat
423.7 feet, a very strongodor of oxidt=edhydrocarbonswas detected. Core
anagys.ia
apecisena.-~~lectsd
.f.rgv
tlw.twerv+l-.yeye
.Uffls.!tlt
tO-.?lecn-bx_sOk:,.....
.--.-...>,
.....
.....
vent extraction. The bottom one-halffoot ofthereservoirsand showeda

,,,

29

slight oil bleed when removed from she core-barrel. The core was almostcompletely devoid of crude oil except in the bottom 2 to 3 feet of sand, Also,
the bottom 1 foot of the reservoir sand and adjacentunderlyingshale was
natural in color and appearance,
A formtion-temperaturemeasurementmade with coring fluid in the hole
showed that the temperaturewas greater than 230 F. After the coring fluid
was bailed from the wells a weighted thermocouplecould not be loweredbelow
407 feet, either because of caving of the formationor dehydrationof the
coring fluid. After thewell was cleaned out to 444 feet (28 days after
~
coring),a maximum temperatureof 160 F was recordedat 420 feet. The temof the sand to 75 F at 444 feet.
perature.decreasedfrom 152 F at the bottora
The temperatureat the top of the sand was 140 F. The measured temperatures
are not the reservoirtemperaturebefore coring because the foruwtionwas
cooled by the coring fluid and the backflow of gas from the formation.
Petrographicstudiesof sampleswere mde to determinethe approximate
temperatureto which various reservoir intervalshad been elevated. Table 9
presents the results of these tests.
TABIX 9. - Reservoirtemperaturesdeterminedby petro~ra~hic
thin-sectionanalyses, Experimentalwell 3
Estimated
temperature,F

Depth, feet
391.6

1,100

395.6
399.6
403.6
407.6

1,200
1,300
1,300
1,300

Depth, feet
411.6
415.0
417.6
421.6

Estimated
temperature,0F
1,300
>400
>400
>400but <700

COMP arisen of

Core-AnalysisResults%
ExperimentalWells 2 and 3

Core-analysisresults for Experimentalwell 3 are comparedwith those

determinedfor Experimentalwell 2 in figures 17 and 18. The differences
could be due to eithernatural changes in the reservoiror to elevated
temperature.
Figure 19 shows average relativepermeabilityratios (relativepermeabilto gasto the relativepermeabilityto oil - Krg/Kro) at different liquid .
=
saturationsas a percentof the pore volume for Experimentalwells 1, 2, and
.)
3. For Experimentalwell 2, the ratio rapidly approachesinfinityas the
liquid saturationapproaches40 percent; whereas, the sand,i,n
well 1 with
lower average air permeabilityapproacheshigh values of Kr &O
at lower
liquidsaturations. The average curve for well 3 unexpecteiily fell on the
average curve for well 2$ but it representsthe portion of the re~?~oir that
the
was
affectedby high temperature. Since the ratio is an indicatfon:,of
_. __,.
_=
_-.
_.
gas-oi~--r~tio,
itwould,-bs:~~$ct~thtit-foa
givenen Itquid-satmatio,n-,-t
he-- -----:-=
reservoirat ExperimentalwellQ would permit a greater flow of gas per unit
volume of oil and water than the reservoirat Experimentalwell.1. The
ity

30

1140

3s9,3

1135

334.3

----------- _- . -.z===-

------

---.-

391.9

.=---- ----

%-

- 396,9

-=z?=f==

1130
t

Exlses,lyel

---

3sPerer3ritsl

399.3~...
----

\
------

i.

- 401,9

d-

4C4.3-----

-----

--=-------4M3 -

1110

419,3-

1105

424.3

I
md-

1
429.3!

--------------------,.
-------

------

--------a
-.
----

P -/-..

I
100
PERMWBILITY,

FIGURE

----------~-----

I
200

5~431g

mfllidarcys

17. - Comparison of Permeability

30
IWROSIW,

ond Porosity,

percent

Experimental Wells 2 and 3.

reservoirbehind the combustiongone would permit an infinitevolume of gas to

flow without liquid rnoyement
at a liquid saturationof 25 percent of -thepore
volume. Ahead of the combustionzone, where products of combustionwould be
condensingand mobile reservoir liquids collecting,the averageKrg/Kro curve
for Experimentalwell 2fndicates a decreasingvolume of gas flowing per volume of reservoir liquidsas the liquid saturationincreases. If liquid accumulation ahead of the advancinghigh-temperaturezonehad occurred in the
field operation,a deczease in the gas-liquidratio shouldhave been, but was
not, observed. The quantitiesof ltquidwhich could have been moved ahead of
the front may have been too small to affect the ratio with the well spacingat
this loca~ion.
;

F5gure 20 shows the relationshipof formation-resistivity

factor and
),
t
porosity for Experimentalwells 2 and 3.

The bulk clensityand effectiveporosity for each core sample from Experimental wells 2 and 3 were used to calculatethe sand-grainor apparent density
well 2
.
(mineral specificgravity). The p~ofi.leof the data for Experime.Btal
was erratic. Values of sand-graindensitiesfor individualsamplesvaried
from a minimum of 2.48 to a maximum of 2.90 grams per cu cmand averaged 2.74
grams.percu cm for the Cotal sand thickness. For the total reservo~~,thickneski
in'"~@eritintal:wel-l-3-;:-the'walv&~f"'effe@&iVe.sand~graZ-n-d~,s2titie
were=----------=.
-.
more uniform,ranging from.2.64 to 2.86 grams per cu cm and averag@g,.2.71
grams per cu cm. From the top of the san~ to 407 feet in.Experimei@lwell 3,

i
I
,
,

rT_-=--

---1

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

r-----l=

/
\
,
:
----

,2

~
~-.

,,ol!u,~l
1.3

1.5
1.7
1.9
POROSITY
EXPONENT

21

lllllllllflll

10

20
N..4O5O6O7C
FORMATION-RES!STIWIY
FACTOR

-----\

-II

1 1
I
)
I
I
-u--&L&l:
431.9
26
2.5
26
27
28
29
20
EFFECTIVE
SANDGRAINDENSITY,
@ cm
81JLKDENSITY,
@u cm

,.
1

i$.
(.:
,.

FIGURE

18.-Comparison
of Porosity Exponent, Formation-Resistivity
Factor, Bulk
Density, and Effective Sand-Grain Density, Experimental Wells 2 and 3.

32

1,000
!.-

~
> 1~ &
g
E
s
g
A

:
10:
Ks,
md
o Well 1
Well 2
A Well 3
1

30

FIGURE

20.6
68.2

M43

C!arbon-Hydr6gen
Tests ,

70
60
6a
40
TOTAL LIQUID SATURATION,porewolume percent

19.

the effectivesand-graindensity averaged 2.68 grams per

cu cm comparedwith 2,74 grams
per cu cm for Che same interval in Experimentalwell 2.
Sand-graindensity below 407
feet to the sand bottom in
Experimentalwell 3, and in
the same interval in Experimental well 2, averaged 2.74
grams per cu cm. The combustion process should decrease
the bulk denflitywhile increasing th(?effectiveporosity.
As the porosity increases,the
volume of reservoirframework
decreasesand the sand-grain
density should decrease.

80

Gas-Oil Relative Permeability Ratio-Total

Liquid Saturation Relationshipps, ExperimentatWells 1,2, and3.

Samplesof core material

at 2-foot intervalsthroughout
he and ection
rem Xperi
mental, well 3 were analyzed
for total
carbon and hydrogen

by the stand&rdASTM method

(D 271-58). Table 10 shows,the results of the analyses. Inorganiccarbonate
and water of hydrationof silicates,if present, in addition to hydrocarbon
mterial are includedin the determinations. The weight-percentof each element is based,on.the absolute sand-graindensity because the analyticaldeterminationswere made for crushed samplesof core material.
Structurally,PennsylvaniaGrade crude oil has an atomic hydrogen-carbon
ratio of 2:1. On a weight basis, there is about a 6:1 ratio of carbon to
hydrogen. A correlation
of these basic relationshipsand the analytical
results indicatedthat the ratio of carbon to hydrogenhad been alteredby
elevatedtemperatureand/or that quantitiesof inorganiccompoundswere
present. The gross heat of combustionof the sampleswas determinedby
aseumingthat inorganiccompoundswere present. The analyticaltechniquewae
in accordancewith ASIMmethod D 271-58 for determiningcaloriffcvalues using
benzoicacid as the standardcombustionsample in a Parr adiabaticoxygen calorimeter. By this method, only the organic carbon and hydrogen in the sample
contributeto the heating value. Samples from three depths were analyzedand
the results are shown in table 11. Based on the heating Value of First sand
crude oil (19,S00 Btu per lb),
the percentagesof crude oil necessary to produce the measuredheating values were calculated. Next the percentagesof
carbon and.hydrogen,that contribute to the gross heating value were determii~dti---tuwasanecessaryry
toassme- tha$--carbon-a
nd~hydrogen-existed.-in
-the--.
same ratio as in the crude oil. A comparisonof these results with the tombustlon atilysesare shown in table 11.

.-...=-..

33
I

0
0

Experimerttal Well 2
Average cementation
factor = 1.76

Average cementation
factor = 1,71

~
10

..

FIGURE

20

30

50

20. . Formation-Resistivity
Wells 2 and 3.

100 7
POROSITY,
percent
Factor-Porosity

20

Relationships,

30

50

100

Experimental

It appears that the assumptionupon which the percentagesof hydrogen and

carbonwerecalculatedfrom the h,eatingvalues was invalid because the percentage of carbon determinedfrom the heating value is greater than the value
,
determinedby combustionfor each depth. ,
,..
An additionalcalculationwas made by assumingthe percentageof carbon
determinedby combustionanalysesrepresentedonly organic carbon and did not
includecarbonates. The percentageofhydrogenwas then calculated..using
the
gross heating valueof the samplesdeterminedanalyticallyand the calorific
values per pound of carbon and hydrogen. ,Forthesample from 407.7 feet, the
percentageof hydrogenwas determinedto be 0.085 comparedwith the combustion
.analysisof-:O~l10;5.and:.for--tfie-.sample..
represeri$ingz4-17.~7feeta-the.-hydr.ogen:
--.!:----value was calculatedto be..O.Ol8
percent comparedwith 0.188 percent determined by combustionanalysis. The carbon-hydrogenratio was calculatedtobe
2.7 and 2.4 for the respectivedepths. The same procedurecould not be used
.,.

for the samplerepresenting395.5 feet where the weight-percentof hydrogen

based on the gross heatingvalue of the samplewas greaterthan the weight of
carbon determinedby combustionanalysis.
analyses,Experimentalwell 3
TABLE 10. - Carbon-hydrogen

--m--

Hydr{ en
Moistures
Ca on
weighLWeightLb/cu ft Weightweightpercent -Jercent
percentl of sari@ ~ercent of sand~
0.520
0.22
0.087
o 144
391.7
.......... 97.39
.091
.055
3.564
.26
2.160
393.6......... 92.54
99 14
,147
.114
.18
.089
.069
395.5.........
.158
14
122
.096
.074
397.6......... 99.24
,086
.18
.087
.052
.046
399.5......... 99.36
.078
.08
.084
139
047
401.6......... 99:24
.03
.182
.110
.153
.093
403.6......... 98;97
.
.611
.079
.130
405.5......... 97.77
.370
.611
.182
.08
.110
370
407.7......... 97.74
.173
.08
105
.087
409.7....s.... 99.84
.144
.140
.04
.141
411.7...,..... 98.74
.231~
.233
.07
413.6......... 98.32
.095
.128
.211
153
,10
,160
.093
415.6......... 98.60
.264
.310
.26
.603
.188
.995
417.7......... 97.98
.932
.28
,200
.330
.565
419.6.......i. 97.80
.805
1.328
.36
.140
.231
421.7......... 97.38
1,010
.29
423.5......... 97.73
.254
.419
Based on removalof free moisture.
2Calculatedon the basis of an average sand-graindensityof 2.65gramsper
cu cm.
Depth,
ft

I
1

/TABLE 11. f-Gross calorificvalues of core sarsples~

Experimentalwell 3

Heating value
Combustion
, Btu
Oil saturation,
analyses
per Percentpore volume
calculation
lb Calculated Core
Percent Pertent Percent Percent
analysis hvdro~en carbon hydrogen carbon
395:5........ 66
4.2
1.5
0.069
0.020
0.369
0.089
033
.110
.370
.592
407.7 . . . . . . . .
106

8.9.
.5
188
.603
13.4
11.0
.049
,896
417.7 . . . . . . . .
160

Depth,
ft

INJECTEDAND PRODUCEDGAS MATliRIALBALANGE

A completematerialbalancecould not be made betweenthe injected-and
/
produced-gasvolumesbecauseof the unknownquantities,
for (1) air dilution
from outsidethe pattern,(2) sweep efficiencyof the injected-gasvolumes,
(3) verticalconformenceof the injectivityprofile,.(4)degree of combustion
of injectedand residualhydrocarbon and (5) volumes of oxygenused and carbon dioxidegeneratedby combustionof theresidualreservoirhydrocarbons,
Th&par~ia-lly
ox%d%ze~state
of-the-resetioir
oil:,caimed=by-prevfous
air-gal+-------injectionoperations,and the presente
of organicand inorganiccarboncompounds complicatedthe evaluationofthermochemicalreact$onsin the reservoir.
..

..,
.

?
~.

,,,

. .

..-

,!

Air injectioninto the First sand, one well locationbeyond the test pattern, was discontinued4 days after ignition;therefore~gas-productionrates
from the pattern producingwells before ignitionwere not known. The gasproduc~ionrates from each,wellwere plotted against total gas-injectionrates
for differentperiods after ignition. Anextrapolationof the curve for each
well indicateda probkble productionrate at zero injection. The combined
initialproductionrate was 150,000 scfd (estimateddilutionvolume).

~~

A.nitrogenbalance between the injectedand produced gas volumes during

the latter stage of burning gas in the wellbore itidicated
that about 120,000
scfd of air could be enteringthe pattern producingwells from the reservoir
beyond the experimentalpatiern-assumingthat none of the injectedair or gas
generatedby combustionIeftthe pattern. Oxygen consumptionwould then
amount to about 60 percent of the volume injected;completecombustionof the
injectednatural gas would.have resulted in using only about 38 percent of the
oxygen consumed.
A mterial-balance calculationwas made using the helium-tracerresults
and average injectionand productionrates. About 485,000 scfd of air and
9,000 scfd of natural gas was injectedduring the tracer-testperiod; therefore, some 97,000 scfd of oxygen was being injected. Completecombustionof
the injectedna~ural gas would require about 21,000 scfd of oxygen and would
ecovery indicatedthat
generate 14,000 scfd of carbon dioxide. Helium-tracerr
5.1 percent (or about 5,100 scfd) of the availablec~gen was unrecoverable.
Assuming that loss occurredand that in situ combustionwas not achieved, some
76,000 scfd of the oxygen injectedshould have been recovered;

Completecombustionof,ttheinjectednatural gas would have liberated

11,000 scfd of carbon dioxide and the average productionwas 24,000 scfd, thus
additionaloxygen was utilized to generate the additionalcarbon dioxide produced. The sourceof the additionalca~bondioxideproducedmay.havebeen
from one or a combinationof the following: (1) Combustionofresidual
hydrocarbon,(2) oxygenationof partiallyoxidizedcrude oil left in the
reservoirafter the long-termair-repressuringproject, or (3) natural
carbonatesand bicarbonatesin the,reservotr.
A nitrogenmaterial-balancecalculationwas made to deterinine
air dilution,,assumingthat recoveredvolume of injectedgases and combustionproducts
was proportionalto the volumeof injectedhelium recovered. Approximately
383,000 scfd of the inert gas was injected;but consideringthe helium-tracer
test results,about 19,500 scfd of that amount was not recovered.The difference between ~he 445,000 scfd of actual nitrogenproductionand the 363,,500
scfd of the injectedvolume calculatedto have been recovered shows some
81,500 scfd a~!thevolumeof nitrogendilution. .Kfair was the diluting.gas,
some 21,500 scf~of oxygen would accompanythe 81,500 scf of nitrogen,or
103,000scfd ofair,would be produced from outside the pattern.
, .. fwbin!w
@!evo@IE.of.d~Wiw _...Omp
and the.76@0...Scf6?$th?
:...,....
i~jectedvolume not used ~nthereactionwiththe-naku~al
gas; 97-~500scfdof
oxygen shouldhave been produced. However, daily average oxygen production
was 93,000scf, leavingonly some 3,500 scf,.available
to generatetki balance
,
J
,
-,
J
/.
f

of the carbon dioxide produced from a source other Chan c~ustion

injectednatural gas.

of the

IJ%is
rather inconclusiveinterpretationapplies only to the,timeduqing
the experimentwhen the tracer test was conducted. Inaddition,the soltibility of combustionproductswas discountedin the evaluationbecause of raptd
travel times.

DISCUSSIONOF RESULTS
The experimentwas suspendedJanuaey 5, 1962, when it becatieevident that
economicoil productioncould not be achieved. However, an importantoriginal
objectivewas achievedby establishingthat a heat wave could be initiatedin
the vicinity of the wellbore and then propagatedradially throughChe reservoir rock toward the oil-producingwells.

...

There was no conclusiveevidencefrom injecttonand productionresults

that self-sustainedcombustionwas achieved or that combustionwas responsible
for increasingQil and water production. A slight liquid-productionincrease
was attributedto strtppingaction caused by the air-injectionrates, which
were higher than those normallyused during air-gas repressuring. There was
no indicationthat liquidbanks had formed in the reservoir.
Severalreasons can be given for the lack of success or low o~gen utilization#nilinabilityto supporta sulf-sustainingheat wave. The reservoir
wa$ excessivelyheterogeneous,whit~permitted
bypassingof the injectedair
thxough a highly permeable,1ow-.$1u~-saturatedreservoir interval. This condition would not permit sufficientoxygen to enter the m~re uniform lowerpermea%ilityintervalwhere the oil saturationwas higher. However, good
verttcal conformancewas showh by.the,c~re
taken from a post-experiment
~observation
welldrilled
14 feet from the ignitionwell. Core samples indicated that much of the sand was completelydevoid of hydrocarbonsand had
been heated to temperacuresin the range of .l,100Ftol,200degrees F.
,,
POOr
combustioncould also be attributedto.the characteristicsor properties of this crude oil and/or low oil saturation.Testdata did qot indicate whether insufficientfuel depositionor short contact time between the
injectedair and fuel caused,the unsatisfactoryresults. The effects of airgas injectionon the oil (that is, low-temperatureoxidation)and/or,for~tion
are unknownwith respect to the experiment.Adverse airflow patternscould
have been establishedprior to the experiment. Core analysisresults obtained
befotiethe tesh, however, indicated,
sufficientoil saturationand a satisfactory range of permeability. Probable fuel depositionalsowasbelieved to be
adequatefor combustion-wavepropagation.

The test was discontinuedwhen it became evident that aky high-tempera=,

ture heat wave going through the.reservoirwas being transferredby convection ,
and m.ightrequireup.to_10 years--for
definit%tegvglua~+on.

,,

37
CONCLUSIONS
1. There was no conclusiveevidetice
from injectionand productiondata
that self-sustainedcombustionwas achieved..
-.
2. Bypassingof the injectedair througha highly permeable, low-fluidsaturatedreservoir interval.
may not haveperurittedsufficientoxygen to enter
the more uniform, lower permeabilityreservoirintervalwhere the oil saturation was higher.
;
3. Combustionwas not responsibleforincreasingoil and water producwas no apparent indicationthat liquid banks had formed in the
tion. lhere
reservoir.
4. The rate of advance of the combustionfront and ~ptimum air requirements were not estab:ished because of the inabilityto attain self-sustained
combustion.
5. Sweep effic:ency and the effect of heterogeneitycould not be evaluated from the single observationwell drilled.afterthe experiment,
6. The ignition-wellcompletionmethod was satisfactoryto withstand the
conditionsof pressure,andtemperatureduring ignition.
I
/
7. Interpretationof the test data was greatly complicatedbecause the
experimentwas conductedin a reservoirthat previouslyhad been subjectedto
air-gas injection.
;,
I
8. Additionalexperimentationwith the process in high-gravity,lowviscositycruile?oil
reservoirsis warranted.

.,

.,,

. . -.... ..
,,
..-

,.,
:...

- -

,--

38

REFERENCES
Alexander~John D,r W. L. Martin, and John N. Dew. Factors Affecting
Fuel Availabilityand CompositionDuring In-Situ Combustion.Pres. at
AIME Prod. Res. Symp., Apr. 12-13,.1962,
lulsa~Okla., 6 pp.

1.

Armstrong,F. E., W. D. Howell, and J. Wade Watkins..RadioactiveInert

Gases as Tracers for PetroleumReservoirStudies. BuMines Rept. of
Inv. 5733, 1961, 15 pp.
*
3.: Bureau of Mines, Morgantown (W. Vs.) PetroleumResearchLaboratoryStaff.
AppalachianRegion OilfieldReservoirInvestigations,Venango Group
First Sand, Goodwill Hill-GrandValley Field, SouthwestTownship,
Warren County, Pennsylvania. ProducersMonthly, v. 25, No. 7, July
1961, pp. 2:5.

2.

4* Caspero,N, A., W. T. Wertman,and Q. E. Wood. Thermal Oil Recovery

Experimentin the PennsylvaniaMiddle DistrictProducingArea.
ProducersMonthly, v. 25, No. 2, February 1961, pp. 27-28.
Quadrangle,Pennsylvania.
5* Dickey, Parke A. Oil Geology of the T.itusville
l?ennsylvania
Geol. Survey Bull. M22, 4th ser., 1941,,87 pp.
6. Dickey, Parke A., and R. B; Bossier. Oil Recoveryby Air Drive inthe
Venango Fields,of Pennsylvania. Proc., 21st Annual Meeting,,API,
Prod. Bull. No. 226, v. 21~ sectionIV, Chicago,Ill., Nov. 11-15,
1940, pp. 11-25.

7* Emery, L. W. Ignitionof Fuel for Lightinga Well. (Assignedto

SinclairOil and Gas:Co.) U.S. Pat. .2,941,595,June 21, 1960.

8..

Bottom Hole Burner. (Assignedto,SinclairOil and Gas Co.)

U.S. Pat. 2,985,240,May 23, 1961.

9*

. ResultsFrom a Multi-WellTherrnal-Recovery~Test
in Southeastern
Kansas. J. Petrol. Technol.,v. 14, No. 6.,June 1962, pp. 671-678.

10,

Fuchida,T. Laboratoryand Field Experimentson Fire-FloodRecovery

Method. Pres. atFifth World Petrol. Cong., New York, N.Y,,May31June 5, 1959, sec. 11, Paper 212 12 pp.;ProducersMonthly> V. 23$
No. 9, July 1959, pp. 30-35.

11. Holbrook~G. W., ~. E. Hungerfordsand M. R. Rankin. A Fire-Flooding ,:

Experimentin the Bradford-AlleganyArea. ProducersMonthlySv.232
No. 12, October 1959, pp. .26-28.

IL:.

. .

..

12, Lytle, W, S. Crude Oil ReservesofPennsylvania~.Pennsylvania

Topographicand Geul.--Sur,.BulllM32:,2:,-1950.,--P.
+226.T-----==
--=----.:------.-I;;=Z
. .

. . . .

.. .

I
,

l-
!
[,,

?.

,:

.,

.,

.
,.

39
.
13,

14. Mosst J. T., P. D. White, and J. S, McNeil, Jr. Zn-Situ Combustion

Process - Results of a Five-WellField Experimentin SouthernOklahoma.
J. Petrol. Technol.,v. 11, No. 4, April1959, pp. 55-64; Trans. AIMES
(Petrol.Devel. and Technol.),v:216, 1959, pp. 55-64.

,,

Martin, W. L., J. D. Alexander,and J. N. Dew. Process Variablesof

In-Situ Combustion. Trans. AIME, (Petrol.Devel. and Technol.),v. 213,
1958, pp. 28-35; J. Petrol. Technol.,v. 10, No. 2, February 1958,
pp. 28-35.

15. Parrish,D. R., R. W. Rousch, K. W. Beaver, and H. W. Wood. Underground

Combustionin the ShannonPool, Wyoming. J. PeRrol. Technol.,v. 14,
No. 2, February 1962, pp. 197-205.
.,
16.

ProducersMonthly. PennsylvaniaUndergroundCombustionExperiment.
V. 25, No. 10; October 1961s pp. 11, 21.

17.

Sherrill,R. E., Parke A. Dickey, and L. S. Matteson. Types of Stratigraphic Oil Pools in Venango Sands of NorthwesternPennsylvania.
Ch. in StratigraphicType Oil F3elds, ed. byA. I, Levorsen. American
Associationo~ Petroleu&Geologists,Tulsa,-Okla.,1941, pp. 507-538.

r
,.

,,
/

/.

,
. -,
,

-,.-

-.

-----

i.

lNT,

.-

cBu.0~

. ...

. . .

MI ME S, POHi,

PA.

. . ..

5958

,
-,

,.
,f
,,. . .

.. . ..!J.

.
. .. .

.,. _,

,.