MISCELLANEOUS

This section contains two (2) miscellaneous papers, as follows.

"A Pressure-Volume-Temperature Correlation for Mixtures of California
Oils and Gases"
By M. B.Standing,
Standard Oil Company of California, La Habre; Calif.
(Presented a t Pacific Coast District, Los Angeles, Calif., May 1947)
I
"Formation and Operation of Unit Projects in Secondary Recovery"
By K. E. Beall,
Phillips Petroleum Company, Bartlesville, Okla
(Presented at Mid Continent District Meeting, Amarillo, Texas, May
1947)
A PRESSURE-VOLUME-TEMPERATURE CORRELATION FOR
MIXTURES OF CALIFORNIA OILS AND GASES t
ABSTRACT
The solution of reservoir-performance problems re-
clulres t hat t he physical properties of t he reservoir
fluids be known These propert ~es may be determined
The paper presents correlat~o~is of bubble-polut pres-
sures, for~natlon volumes of bubble-po~nt Iigu~ds, and
forn~atlon volun~es of gas plus liqu~d phases as em-
pir~cal f ul ~ct ~ol ~s of gas-011 ratlo, gas gravlty, 011 gravlty,
pressure, and temperature. Although the correlat~o~ls
in t he laboratory either from bottom-hole samples or
froin proper recoinb~nation of surface t r ap samples If
t he results of laboratory t est s a r e not available, how-
are on Calrforll~a crude 011s and gases, comparrsons are
made for the varlous crudes reported by Katz. In order
to fac~l~tate the use of the data, the results of the cor-
r e l at ~o~~s are prese~itecl 111 the form of calculat~~ig charts.
ever, t he physical properties must be estinlated from
field i neasuren~ent s The purpose of t hl s paper is t o
give the results of several correlations between t he
variables normally measured in t he field and t he phys-
ical properties necessary for t he solution of reservoir-
perfornlance problems Other correlations of t hi s t ype
have been reported by Gosline and Dodson,l" and by
Katz Sage and Ol ds4 have recently reported a n es-
cellent correlation of formation volumes of condensate
systeins
The accuracy of t he following P-V-T correlations is
restricted by two fact ors 1 The varied and con]-
plex multi-component hydrocarbon systeins which ar e
dealt with ar e defined by only t hree siinple paramet ers
gas gravity, oil gravity, and gas-oil rat i o, and, 2 These
parameters themselves depend upon t he process by
which t he oil and gas ar e separated The method used
by Sage and Reamer2 i n t he Rio Bravo Field studies of
specifying hesanes and heavier mat eri al a s "oil" and al l
pentanes and lighter material as "gas" \vould overcome
these difficulties However, t hi s approach does not lend
: Itself t o field usage
A furt her aid to prepari ng correlations would be
to make use of a st andard procedure i n separat i ng t he
oil and gas when gas-oil ratios a r e determined How-
ever, a s t he P-V-T dat a ar e prepared f or individual
field conditions, it is now~al l y not posslble to use such
a method
The gas-oil ratios, gas gravities, oil gravities, and
formation volumes presented in t hi s paper ar e labora-
t ory values They ar e t he result of a 2-stage flash
separation a t 100 deg F-the first st age norinally being
within t he pressure range of 250 psi t o 450 psi and
* Stanc1,ird Or1 Co of Cnlrfor~ira, Ln t1.1lrri1, Cnl ~f , r e~nor ed,
1'34; t o Cn11fornr.r Research Cabrli. La Habm. Ca l ~f
t Presented a t tlie sprrng meetlug of tlre Pacrtic Coast D~s t r l c t ,
Dl \ l s ~o n of Proi111ctron. Los Augeles, Cal l , B1a.r 15, 1947 pre-
srdlng, E V Wat t s. General Petroleurn Corp . Los Bugeles. 'Calrf
a F~g n r e s refer t o REFERENCES on 1, 279
second st age a t atmospheric pressure This procedure
is considered t o approximate t he average California
field practices
Bubble-POIII~ Pressure Correlat~o~~s
One of t he inp port ant functions of P-V-T dat a 1s to
indicate whether t he reservoir oil is undersaturated or
saturated, or whether free gas i s being produced from
t he sand This requires a knowledge of t he gas solu-
bility-bubble-point-pressure relationship of t he oil and
gases associated 111 t he reservoir
I n considering t he manner 111 which t he several
variables affect t he bubble-point pressure of a mixture
of an 011 and a gas, ~t seems reasonable to postulate a
correlation of t he form
P, = *(GOR, y,, T, API ) (1)
P b = bubble-point pressure, 11~1, absolute
GOR = gas-oil ratio, cu f t per bbl
7, = gravi t y of dissolved gas (ai r = 1)
API = gravi t y of t ank oil, deg API
T = temperature, deg F.
9 = a function of
I n developing t he specific equation relating t he bubble-
point pressure to t he variables on t he n g h t side of equa-
tion ( I ) , t he general relationship between t he variables
was used to suggest graphical methods of determining
specific rel at ~onshi ps For example, t he bubble-point
pressure normally increases with an increase i n gas-oil
rat i o This suggests t hat Pa = +, (GOR)' or P,, =
a2(.4) L " Likewise t he bubble-point pressure increases
with an increase in temperature, but decreases with an
Increase in oil gravi t y (deg API) or gas gravl t y
(ai r = 1)
Aft er a number of at t empt s it was found t hat a plot.
of log(?) vs log Pa resulted in a series of st rai ght
lines with a n average slope of 0 83 Mathematically,
t hi s gave t he relationship
[ P b l GOR
T Al . 1
Predrctlon of Bubble-Polnt Pressure from Gas-011 Ratlo, Solution-Gas Gravity, Tank-011 Grau~ty, and Temperature.
FIG. 1
A second plot of log VS T S O re-
sulted in straight llnes of almost constant slope, or,
expressed mathemat~cally
Flnally, a third correlatlon was made to determine
the effect of 011 gr av~t y, and the following specific
relationship was obtalned
To obtaln the relationships expressed In equatlon
(4) it was necessary to have tests on numerous mix-
tures of oil and gas a t a variety of temperatures
Fl g 1 shows the results of plottlng 105 experllnentally
determined bubble-polnt pressures on 22 different crude-
oil-natural-gas mlxtures The range of the data was as
follows
Bubble-polnt pressures 130 to' 7,000 psi, absolute
Temperature 100 to 258 deg F
Gas-011 ratlos 20 to 1,425 cu f t per bbl
Tank-011 gravities 16 5 to 63 8 deg API
Gas gravities 0 59 to 0 95 (air = 1)
The agreement to be expected from the foregoing
correlatlon can be estimated from the curves shown In
Fi g 2 The upper curve shows t hat 58 per cent of the
I00
80
6 0
40
2 0
0
0 100 COO 300
PRESSURE DISAGREEMENT - PSI
- 14- 12- 10 - 8 - 6 - 4 - 2 0 2 4 6 8 M 12 14 16
DISAGREEMENT - PER CENT
Frequency D~str~but~on of Bubble-Po~nt Pressure
Correlat~on.
FIG. 2
points d~ffered froln the correlatlon by less than 100 psi,
and t hat only 12 per cent were further than 200 psi
from the correlatlon The lower curve gives the fre-
quency distribution of the errors resulting from esti-
matmg the bubble-polnt pressure froln the correlation
More than half of the experimental points were wlthln
5 per cent of the correlatlon The arithmetic average
error was 4 8 per cent and 106 psl
The data on 53 crudes reported by Kat z3 do not,
when plotted as lndlcated in Fl g 1, give as good a
correlation as the data reported In thls paper A llne
drawn approximately 150 psi hlgher than t hat shown
in Flg 1 gave the best correlatlon of the Katz data.
However, in terms of the present correlation, 52 per
cent of the Katz data fell withln 200 psl of the correla-
tlon as compared to 88 per cent of the California data
The better correlation obtalned In the case of the
California crudes is posslbly explained by the fact t hat
the crudes reported by Katz were from a larger varlety
of sources Differences in laboratory methods, however,
lnlght account for part of the dlsagreement
Formation Volumes of Bubble-Point Lquids
A second factor requlred in reservoir calculations IS
the formation volume of the saturated llquid phase
This factor is used to compute the shrinkage of the
reservoir oil when ~t IS processed to the stock tanks.
The fornlatlon volumes of the 105 bubble-point llquids
used In the prevlous section were correlated In terms
of the gas-011 ratio, gas gravlty, tank-oil gravity (spe-
cific gravlty), and temperature The correlatlng equa-
tion finally selected was
Vb = forlnatlon volume of bubble-po~nt hquld, bbl
per bbl of tank oil
GOR = gas-011 ratio, cu f t per bbl
-ys = gravity of dissolved gas (air = 1)
yo = speclfic gravlty of tank 011 a t 60 deg F
T = temperature, deg F
.f. = a functlon of
Fi g 3 shows the results of plotting the experlinental
format~on volumes agalnst the function shown in equa-
tlon (5) To glve an idea of the nlagnitude of the
errors Involved in the correlatlon, llnes of 5 per cent
d~sagreement are shown
The frequency dlstributlon of the errors is shown in
Flg 4 I t wlll be noted t hat 45 per cent of the points
fit the correlatlon with errors less than 0.5 per cent,
and that no errors were greater than 7 per cent The
arithnletlc average of the errors IS 1 17 per cent
The shrinkage data presented by Katz, when con-
verted to formation volumes, showed an arithmetic
average error of 120 per cent when correlated by equa-
t ~ o n (3) These close agreements indicate t hat the cor-
relation of bubble-golnt formation volume IS more
general than the bubble-polnt pressure, and t hat the
correlations can safely be used for estlinates on a wider
variety of crude 011s and gases.
Predletion of Formation Volulne of Bubble-Po~nt Liquld from Gas-011 Ratlo, Solut~on-Gas Gravrty, Tank-Oil Gravity,
and Temperature.
FIG. 3
1 1 CORRELATING EQUATION 1 I
Frequency Distribution of Bubble-Po~nt Formation-
Volun~e Correlation.
FIG. 4
Two-Phase Forn~at~on Volumes
The formation-volume dat a of t he gas plus liquid
phases cover a much wider range of gas-011 ratlos t han
do t he bubble-polnt pressure or bubble-point formation-
volume correlations Th1s. l ~ because bubble-point llqulirs
rarel y have gas-oil ratios in excess of 2,000 cu f t per
bbl, whereas 2-phase format~on-volume dat a a r e often
required f or mlxtures havlng rat i os a s hlgh a s 100,000
cu f t per bbl
A correlatlon based on t he equatlon
where
VF = formatlon volume of gas plus llquld phases, bbl
per bbl of t ank 011
P = pressure, psi, absolute
GOR = gas-011 ratio, cu f t per bbl
T = temperature, deg F
ys = gas gravity ( a ~ r = 1)
yo = speclfic gravi t y of t ank oil a t 60 deg F
+ = a function of
Predi ct ~on of formatlon volume of gas plus llquid
phases is shown In Fl g 5 Thl s correlation contams
387 experimental polnts, 92 per cent of which ar e within
5 per cent of t he correlation The ranges of t he dat a
ar e as follows
Pressure 400 t o 5,000 psl, absolute
Gas-oil rat i o 75 to 37,000 cu f t per bbl
Temperat ure 100 to 258 deg. F.
Gas gravl t y 0 59 to 0 95 ( ai r = 1)
Tank-011 gravl t y 16 5 t o 63 8 deg API
The raising of t he oil gravi t y t o a power which In
itself is a function of gas-011 rat l o IS a necessary feat ure
of t hl s correlatlon Thl s causes t he tank-oil gravi t y to
become relatively unimportant a t rat l os ~n excess of
2,000 cu f t per bbl, wh ~ c h IS 111 accordance with act ual
behavlor of hydrocarbon systems
Fig 6 shows t he agreement between esperimentally
determined formatlon volumes and t he values obtained
from t he correlation
At t he tlme t he correlations were prepared, t he dat a
of Sage and Ol dsL on hlgh gas-oil-ratio mixtures were
not available and, therefore, could not be used ~n pre-
pari ng t he correlations A recent check with t he
Sage and Olds dat a showed t hat , above 160 deg F,
t he correlations reproduced 58 experimental observa-
t ~ o n s w ~ t h a maximum error of 5 4 per cent and a n
arithmetic average error of 1 5 7 per cent At 100 deg F
t he errors amounted t o a s much a s 11 per cent, t he
great est error being noted In t he case of t he 5,000 psi,
absolute, vol un~es
Use of the Correlat~ons
The evaluation of bubble-polnt piessure, bubble-point
formatlon volume, and.2-phase formatlon vol un~es from
t he correlations presented In Fl g 1, 3, and 5 IS
somewhat laborious To facilitate t he use of t he
col-relations, t he calculating chart s shown la Fig 7, 8,
and 9 were constiucted Esanlples of t he use of each
of t he chart s a r e , ~ndi cat ed on t he chart s From these
chart s i t IS possible qulckly t o obtain engineering
est ~mat es of t he physical properties of multi-com-
ponent hydrocarbon ' systems under pressure and tem-
perat ure conditions encoul~tered in 011- and gas-pro-
ducing reservoirs
ACKNOWLEDGMENT
The aut hor IS grat eful to t he nlanagelnent of t he
St andard 011 Company of Cahfornia f or permission
t o publlsh t hi s paper
REFERENCES
' J E Gosllne and (L' R Dodson, " Sol nl >~l ~t s Relations and
Vol rl ~nrt rl c Br hnr l or of Tl ~ r r p (.;r:rrltles of C r n d ~ s a nd Assoc~:lted
C:ases." Llrrllrrrg n1t17 Proclrrrtro~i Pr ~t ct l cc. 43.1 (1938)
and Gas from
-1fct E~r g r s 14
3 n I. Kn t s
t he Rln Bm\ o Field." Trolls dill. Il l st ~l f l ~l r l r o
2, 179 (1941)
- 4 . " P r e ~ l ~ c t ~ ~ n of the_Shr~r!li;l~ge of Crude O~l s , "
~ r r c r ~ r ~ o ~ t d Prc
4 R A S:IBP
)drrctron Prnct rcr 1'37 (1942)
.- and R H ~ l d s , ' " Vol ~~r ne t r ~c Rehnrl or of 011
and- as from Several San J oaql ~l n Valley Fields," Tml rs Awl
I ns t MI I ~ I I L ~ Net Elrgrs 17n 15G 119-1SL . . , . A . . . . , A . . . ,
of Hs~l r ocar bon Gases and TTapors,"
DISCUSSION
B H Sage (California Institute of Technology, Pasa-
dena, Calif) The aut hor has presented a n ~nt er est r ng
correlatlon of t he volumetric and phase behavlor of
nat ural l y occurring lnlstures of 011 and gas of low
and ~nt e r me d~a t e gas-oil ratlo The analytical expres-
slons proposed f or t he relationship of bubble-polnt pres-
sures and formatlon volumes to gas-oil ratios ar e some-
Prediction of Formation Volume of Gas Plus Liquid Phases from Gas-Oil Ratlo, Total Gas Gravity, Tank-Oil Gravity,
Temperature, and Pressure.
FIG. 5
P
R
O
D
U
C

T
l
O
N

T
E
C
H
h
O
L
O
G
Y

L
A
B
O
R
A

T
O
R
Y

C
o
p
y
r
r
g
h
t

I
9
4
7

C
h
a
r
1

f
o
r

C
a
l
c
u
l
n
t
~
n
g
B
u
b
b
l
e
-
P
o
i
n
t

P
r
e
s
s
u
r
e

o
r

S
o
l
u
b
~
l
~
t
?
.

(

R
r
p
r
n
,
l
u
r
,

d

h
v

p
e
r
r
n
r
s
s
,
,
,
,
,

n
j

m
p
v
r
r
p
h
t

n
u
i
r
r
r
r
)

F
I
G

7

P
R
O
P
E
R

T
I
E
S

O
F

N
A
T
U
R
A
L

I
f

Y
D
R
O
C
A
R
B
O
N

M
I
X
T
U
R
E
S

O
F

G
A
S

A
N
D

L
I
Q
U
I
D

F
O
R
M
A
T
I
O
N

V
O
L
U
M
E

o
f

B
U
B
B
L
E

P
O
I
N

T

L
I
Q
U
I
D
S

E
X
A
M
P
L
E

R
E
Q
U
I
R
E
D

F
o
r
m
a
f
t
o
n

v
o
/
u
m
e

a
t

Z
O
O
'F

o
f

a

b
u
b
b
l
e

p
o
r
n
t

/
l
q
u
r
d

h
o
v
m
q

a

p
a
s
-
o
t
l

r
a
h
o

o
f

3
5
0

C
F
B
,

a

q
a
s

q
r
o
v
r
f
y

o
f

0

7
5
,

a
n
d

a

f
o
n
k

o
r
/

q
r
o
v
r
f
y

o
f

J
O

'
A
P
I

P
R
O
C
E
D
U
R
E

s
f
a
r
t
r
n
g

a
f

M
e

/
e
f
t

s
~
d
e

o
f

t
h
e

c
h
a
r
t
,

p
r
o
c
e
e
d

h
o
r
r
z
o
n
t
a
I
/
y

a
l
o
n
g

f
h
e

3
5
0

C
F
B

/
m
e

t
o

a

g
a
s

q
r
a
v
t
f
y

o
f

0

7
5

F
r
o
m

t
h
i
s

p
o
r
n
f

d
r
o
p

v
e
r
f
i
c
a
I
/
y

t
o

t
h
e

3
0
2
P
I

h
n
e

P
r
o
c
e
e
d

h
o
r
r
z
o
n
f
a
I
I
y

f
r
o
m

M
e

t
a
n
k

a
/

g
r
a
v
r
f
y

s
c
a
l
e

t
o

t
h
e

2
O
O
.f
h
n
e

T
h
e

r
e
q
u
i
r
e
d

I
b
r
m
a
h
o
n

v
o
l
u
m
e

IS

6
u
n
d

t
o

b
e

I

2
2

b
a
r
r
e
l

p
e
r

b
a
r
r
e
l

o
f

t
a
n
k

O
I
I

C
o
p
y
r
~
q
h
t

I
9
4
7

C
h
a
r
t

f
o
r

C
a
l
c
~
~
l
a
t
~
l
~
g

F
o
r
n
i
a
t
~
o
~
~

V
o
l
u
l
i
i
e

o
f

B
u
b
b
l
e
-
P
O
I
I
I
~

L
~
q
u
~
d
s
.
.

(
R
e
p
r
o
d
u
r
r
d

b
v

p
e
r
m
n
s
s
z
o
r
l

o
f

r
o
p
v
r
l
e
h
r

o
a
~
n
r
r
)

F
I
G

8

.
I
F
O
R
N
I
A

O
R
A
T
O
R
Y

C
h
a
r
t

f
o
r

C
a
l
c
u
l
a
t
i
~
l
~

F
o
r
n
~
a
t
i
o
n

V
o
l
u
m
e

o
f

G
a
s

P
l
u
s

L
~
q
u
~
d

P
h
a
s
e
s
.

(
R
e
p
r
o
d
n
c
c
d

b
y

p
e
r
r
n
z
s
s
a
o
n

o
f

c
o
p
y
r
c
g
h
t

o
w
n
e
r
)

F
I
G
.

9

what complex In order to peri n~t inore direct ut111z.a-
tion of the data, the last three figures of t h s paper
present graphical solut~ons for the equat~ons From
these it is a relatively s~lnple matter to estimate the
pressure and fornlation volume a t bubble-point and the
formation volume in the heterogeneous reglon froin
knowledge of the pressure, temperature, gas-oil raho,
and the gravities of the oil and gas
The correlat~ons apparently are empirical and, there-
fore, it is improbable t hat they can be applied with
known accuracy to condit~ons widely &berent from
those covered by the primary data upon which the
correlations were based It IS belleved t hat the ~nf or -
mation subni~tted in Fig. 7, 8, and 9 affords a useful
means of estiniating the pressure and formahon volume
Temperature
a t bubble point as well as the 2-phase formation volume
for a wide variety of m~xt ures of 011 and gas from
Cal~fornia fields
In Table 1 (Sage) of this discussion is presented a
comparison of predicted and exper~mental bubble-point
pressures for 3 fields wh~ch were not involved in the
paper under d~scussion. The calculated values were
froin 8 per cent below to 21 per cent above the observed
bubble-point pressures This large variation shows the
uncertainty that may be realized in using the correla-
tion for materials involv~ng ail oil of relatively h ~ g h
gravity Table 2 (Sage) indicates the agreement be-
tween the observed and calculated values of the forma-
tion volume a t bubble point In this instance calcu-
lated values were from 7 6 per cent below to 2 3 per
TABLE 1 (SAGE)
Bubble-Po~nt Pressure
F~el d "A" Field "B" Field "C"
Deg F
u.
100 190 250 100 190 250 100 190 250
Observed value
Calculated value
Difference
Pressure. pountls per square Inch
D~R'erence expressed as per cent
TABLE 2 (SAGE)
Fornlat~on Volun~e at Bubble Polnt
F~el d "A" F~e l d "B" Field "C"
Deg F
Temperature 100 190 250 100 190 250 100 190 250
Observed value
Calculated value
Difference
= Ditference expressed us per cent
TABLE 3 (SAGE)
Format1011 Volun~e in the Two-Phase Reg~on
F~e l d "A" Field "B"
Deg F
Field "C"
Temperature . . 100 190 250 100 190 250 100 190 250
Pressure, pounds per square inch,
absolute 1,000 1,000 1,000 800 800 1,000 1,000 1,000 2,000
Observed value 1 7 8 2 11 2 33 272 335 316 239 3 1 3 1 9 9
Calculated value " 1 8 8 2 30 2 55 2 96 3 60 3 30 2 58 3 28 2 30
Difference 5 6 ' 9 0 9 4 8 8 7 5 4 4 7 9 4 8 1 5 6
Calculated value 1 7 6 204 2 2 3 282 342 3 3 1
Difference -11' - 3 3 - 4 3 3 7 2 1 4 7
a = Calculated fro111 Stunding's correlation
= Calculated from equation (1) of thi s d~scussi on
= Difference expressed as per cent
cent above t he observed fonnat l on voluine a t bubble
point Agaln, t he l argest discrepancy was found f or
mi st ures involving an oil of a relatively high gravi t y
This is not unexpected, lilasinuch a s t he aut hor indi-
cated t hat only a llinlted amount of dat a lncludlng such
materials was available
The formation voluine in t he heterogeneous r e gon
f or a pressure of 1,000 psl has been conlputed f or 2
ml st ures of oil and gas on t he basis of t he correlation
shown i n t he paper under dlscusslon A comparison
of t he observed and calculated values 1s presented In
Table 3 (Sage) I11 t hi s Instance t he calculated forma-
tlon volumes ar e from 4 4 t o 15 6 per cent l arger t han
t he observed values As a11 alternative procedure, t he
formation voluine i n t he two-phase region was computed
uslng t he following espressi onz
TZ
V = Vb+O 005062 F;- (r-rb)
(1)
1
I n applylng equation (1) t he experimentally observed
formation volunle and gas-011 rat l o a t bubble point
were employed f or st at es corresponding t o t he pres-
sure and t emperat ure of i nt erest The compressibility
fact or also was obtained from e ~p e ~l i n e n t a l dat a How-
ever, these dat a could have been estimated from avail-
able correlations based upon t he pseudo-reduced st at e
and t he lnforlnatlon submmtted In t he St andmg paper
The results of t he application of equat ~on (1) a r e in-
cluded in Table 3 (Sage)
R H Smi t h (Signal Oil and Gas Company, Los
Angeles, Calif ) The chart s prepared by Mr Stand-
i ng will reduce a cumbersome laboratory or calculating
process to a slide-rule type of operation I t hi nk it
may be st at ed t hat t hl s information incorporates t he
best broad definition of P-V-T behavior of California
crude 011s yet ava~l abl e i n published form Because
it does represent definite progress i n t he calculation
method, t he question of choice between laboratory de-
t ermnat i on or derivation by reference t o systeinatlzed
esperlence dat a is agai n raised Without at t empt i ng t o
revlew all t he fact ors involved i n making t he choice, I
would rat her confine my comment t o one factor, vi z,
t he iinportance of which IS frequently overlooked
The accuracy and usefulness of all P-V-T dat a a r e
dependent upon a properly carried out sampllng opera-
tlon, as well a s upon t he precision with which t he
laboratory can work Under many circumstances en-
countered 111 t he field, a sample representative of sub-
surface composltlon is difficult t o obtaln As nat ural l y
disposed in a con~pl es reservolr, or a reservoir of l arge
closure, hydrocarbon fluids frequently display a marked
degree of compositional variation which must be 111-
vestigated t o Insure correct t reat ment in working out
t he subsurface inechaiucs The best guarant ee t hat
sampllng methods have been adequate would be t he
adoption of a program of multiple sampling designed
t o nlinlmize or explore t he uncertainties.
The employment of correlated experience dat a, on
t he other hand, 1s a practlce which neatly avoids t he
v~cissltudes of sampling, i e , it replaces t he element
of sampling with a n equivalent operation over wh ~ c h
a great er degree of control can be eserclsed I n t he
assembly of P-V-T dat a as background material, only
those variables subject to definltlon in t he laboratory
ar e deal t mlth No assumptions regarding t he fidelity
achieved In reprocluclng reseilroir fluid con~positioil a r e
made a t t hl s p o ~n t I t t hen remains t o qualify and
adapt t he einplrical P-V-T composit~on relation t o meet
speclfic need A general survey of t he mass of surface
and subsurface information which has been accumu-
lated 111 t he field t hrough t he pertinent ranges of time,
zone, or location will afford t he perspective necessary
to accomplish t he adaptation The sacrifice i n assured
accuracy inherent in obtalnlng pressure and vol un~e
factors from general correlations is frequently offset
by t he inore hscreet t reat ment which may be given t he
varlous portions of t he pool This approach has meri t
when lack of ~nfori nat i on or t he conlplesity of t he
reservolr problem dictates a recourse t o trial-and-error
methods
The dat a t he autllor has presented may be used t o
advantage m coinbinatlon with specific laboratory de-
terminations The chart s have use i n calculatlng t he
effect of snlall changes In any of t he variables (pres-
sure, temperature, deilslty of t he 011 or gas, and t he
gas-oil rat i o) oil preclse laboratory measurements
when t he investigation was not c a r r ~e d into t he range
of lmnlediate interest
As Mr Standlnrr has nointed out, t he accuracv at t al n-
-
able 111 applylng this type of correlation 1s limited by
t he degree to whlch co~npositioil of nat ui al l y occurring
systems can be specified by t he statement of gas-011
ratio, gas gravity, and oil gravi t y If any method, such
a s t hat of Sage and Reamer defi n~ng t he gas a s al l
t he pentanes and lighter fraction of t he composite, can
be employed with improvement 111 general apphcability,
t hen lt 1s t o be hoped t hat t he fund of baslc dat a which
was drawn upon f or t he correlations herein presented
ultimately will be restated In t he inore precise form
Hydrocarbon analyses ar e perfonned wlth facility and
could be made available In inany instances when added
precision is desired
Mr St andi ng Mr Sinit11 has a good p o ~ n t One
difficulty t hat I have stressed is t he effect of t he inethod
of separabon of t he oil and gas a t t he surface on t he
resulting gas-oil rat i o As no doubt many of you will
recall, In Sage and Lacey's Rio Bravo report an at -
t empt was made t o get around t hl s difficulty by speclfy-
1ng pentanes and heavier mat eri al a s oil, and butanes
and hght er materlal a s gas Thi s method is a step in t he
ri ght direction However, it is confusing t o t he field man
and, as yet, ~t has not caught on t o t he extent I should
like t o see ~ t . do
IV Teinpelaar Lietz (Shell Oil Company, Inc , Los
Angeles, Cahf ) Mr St andi ng 1s t o be congratulated
on a very constructive and interesting paper Obvi-
ously, a trenlendous amount of work was required t o
obtain and correlate t he dat a One outstanding use f or
t he calculatlng chart s will be to give us some idea of
orlginal leservoir cond~t ~ons In t he older fields 011 whlch
no P-V-T dat a a r e available However, we should hke t o
polnt out t hat , on comparing t he calculated bubble-point
pressures with 21 experimental detern~lnations, devia-
tions of from 700 p s ~ , gage, to -415 psi, gage, a r e
found, wlth an average of 120 psi, gage
One of t he lnajor uses of P-V-T dat a i s i n carryi ng
out material-balance calculat~ons, especially ~n t he earl y
life of t he field, in order to choose t he most desirable
development schelne However, a t such a n earl y date,
pressure drops ar e small, and, ~f an error such as
700 psl, gage, or -415 p s ~ , gage, were introduced, any
conclus~ons drawn from t he balance calculation would
not be valld On t he other hand, on colnparing t he
calculated forl nat ~on voluines of bubble-polnt liqulds
wlth act ual determlnations, it appears t hat 18 out of
t he 24 determlnatlons check very closely, t he reinalnlng
6 havlng a d e v ~a t ~o n of froin 4 t o -9 per cent
E C Babson (Peerless Pacific Company, Portland,
Ore ) * Mr St andl ng has glven us a set of correla-
tlons from whlch ~t IS possible t o estimate t he bubble
polnts and t he formation-volume fact ors f or nllxtures
of Cal i forn~a oils and gases under a wide range of
pressures, temperatures, and 011 gravities If al l Call-
forma 011s and gases behave in t he same manner a s
t he samples studled In t hl s ~nvest i gat i on, one could feel
fai rl y confident t hat bubble polnts estimated by t hl s
method would be within 10 per cent and formation-
volume fact ors wl t h ~n 4 per cent of t he t r ue values
Although ~t 1s obv~ous t hat any correlations wh ~ c h
would permit such accuracy a r e of gr eat value, a dis-
cusslon of methods of u s ~ n g t he dat a niay t hrow some
hght on t he possiblllty of u s ~ n g these correlations In-
stead of making laboratory P-V-T ~nvest i gat ~ons
P-V-T dat a ar e used prl ~l cl pal l y In material-balance
cal cul at ~ons which range from sllnple estlmates of t he
or1 orlg~rlally In place to det a~l ed studles of reservolr
perforniance The results of such calculatlons can be
used f or a varlety of purposes such a s
1 Estimation of reserves
2 Est ~mat i on of s ~ z e of reservoir
3 Evaluation of st rengt h of wat er drlve
4 Pr e d~c t ~on of f ut ur e performance of a pool
5 Evaluation of a proposed production program
Sometlines a rough es t ~mat e 1s all t hat 1s justified
by t he circun~stances, and it IS obvlous t hat t he corre-
l at ~ons i n t hl s paper ~ 1 1 1 be h~ghl y satisfactory f or
such work The real question regardi ng thew appl ~ca-
tion arises when t he h~ghes t precision attainable is
really needed, and t hl s s~t uat i on often arlses i n material-
balance work
Inasmuch a s it IS unwise t o requlre great er accu-
racy In t he P-V-T dat a t han In t he ot her fact ors enter-
l ng into t he calculat~ons, ~t seems pertlneiit t o examine
sonle of these ot her factors from t hi s standpoint
I n most material-balance calculations ~t i s necessary
t o know t he volume ava~l abl e f or 011 and gas In t he
reservolr I n order t o arrl ve a t t hl s figure, it IS neces-
sar y to estimate.
* Presented by Jan Lam, consultant, Los Angeles, Culff
[PERATURE CORRELATION 287
1 The hulk sand volume
2 The average poiosity
3 The lnterstihlal-water saturation
Considering t he uncertainties enterlng into each of
these estlmates, ~t seeins unlikely t hat t he uncertainty
111 estimating t he reservoir volume will be less t han 5
to 10 per cent except under most favorable concll-
tions I n many cases, of course, t he uncertainty wlll
be even great er t han t h ~ s
A second fact or t hat ent ers into al l material-balance
calculatlons is t he composition of t he reservoir hydro-
carbons, usually expressed a s a n ~n-pl ace gas-011 rat l o
Unfortunately, careful investigation seems to ~ndl cat e
t hat t hl s fact or often vanes materially fro111 point t o
point with111 a reservoir I t IS not uncommon f or t he
h~ghes t gas-oil rat i o In a new field t o be double t he
lowest gas-011 ratlo, wlthout any evldence of f r ee gas
In t he reservolr Under these conditions it would be
necessary t o conduct a long and eqens l ve laboratory
i nvest l gat ~on t o develop representative P-V-T dat a
The t hl rd fact or t hat is necessary III all material-
balance cal cul at ~ons is t he reservoir pressui e Sub-
surface pressures In wells can be measured t o almost
any desired degree of accuracy if enough t est s a r e
lnacle wlth sufficient s k~l l Unfortunately, t he pressure
measured in t he wells may not be representatlve of
t he pressure In t he reservolr Unless t he permeability
of t he sand is h ~ g h enough to permit pressure equahza-
tioil In a reasonable period of t ~me , ~t 1s very difficult
t o obtaln representatlve reservoir pressures Furt her-
Inore, ~f pressures vary ~nat erl al l y from urell t o well,
it IS clifficult to con~put e a represei l t at ~ve average
Also, i n many matenal-balance calculat~ons, t he
critlcal fact or 1s t he relation bet ween-t he-fonnat ~on-
volume fact or and pressure, rat her t han t he absolute
value of t he format~on-volume fact or a t any given pres-
sure If t he slope of t he f ~~mat i on- vol ume fact or vs
pressure curve IS represent at ~ve of t he conditions In
t he reservolr, satisfactory material-balance cal cul at ~ons
can be made even though t here nlay be some discrepancy
In t he absolute values
With these fact ors 111 mind, it seems to me t hat Mr
Standlng' s ~ o ~ r e l a t ~ o n could well be used In place of
laboratory P-V-T lnvestigatlons In engineering work
011 pools h a v ~ n g
1 Few wells
2 Irregul ar sands
3 Low permeahihties
4 Llttle or no wat er drive
On t he other hand, it seeins t o me t hat laboratory
P-V-T lnvestlgations w11 be advisable 111 fields of
l arge slze havl ng consistent sands of relatively h ~ g h
permeability and soine reasonable expectation of a
st rong wat er drlve Laboratory dat a would be particu-
l arl y needed a t pressures above t he bubble polnt, a s
Mr Standlng' s correlations do not cover t he compres-
slbllity of undersaturated liquld
It IS obvious fro111 t he last two paragraphs t hat I
would cons~der these ' correlations t o be entirely ade-
quat e f or t he maj orl t y of t he oil fields In Ca l ~f o r ~u a