PERMEABILITY AS RELATED TO PRODUCTIVITY INDEX +

NORRIS * AND JOHNE. SHERBORNE~

JOHNSTON
ABSTRACT
The variables affect~ngthe radial-flow formula are well behavlor for high-permeability sands approxlnlates
d~scussed quantitatively. The importance of reservolr the theoretical, there 1s an increas~ngd e v ~ a t ~ obetween
n
v~scos~ty of 011 is shown to be equal to that of sand actual perfornlance and theoretical w ~ t hdecreasing per-
pernleabll~ty ~tself In determin~ng flow. A modified meabll~ty. The spread of individual welJ performance
speclfic product~vity-indexfunct~onis developed wh~ch from the average Cal~fornia SPI-k curve may be ac-
bears a slmple theoretical relat~onshipto sand permea- counted for by deviat~ons between alr and water per-
b111ty. Pert~nentdata are tabulated on 141 Cal~fornia meab~l~ties, var~ablespread of permeab~l~ty values, by
011 wells on whlch thorough core-analys~sand produc- varylng stages of reservoir deplet~on, and by varying
e x were available. The modrfied spec~fic
t ~ u ~ t y - ~ n ddata efficacy of well eomplet~on. It is expected that these
product~v~ty index functlon IS studied graph~cally for variables wlll be studled further as more data become
these wells. The conclus~on1s reached that, although the available.

Durlng the past several years attention has been the condition of the fluld occurring durlng the flow
focused with Increasing Intensity upon the problems process
involved In the flow of oil, gas, and water in the porous 2 The empirical correlation of well productivity in-
media which comprise our oil pools d ~ c e swith homogeneous-fluid perrneabillt~es
I t matters llttle whether the petroleum englneer IS Although this paper is primarily concerned with the
concerned wlth problems relatlng to well spaang, in- second method of evaluating well performance, it can
creasing potential productlon, increasing ultimate re- be shown that both methods are closely related and t h a t
covery or, In fact, almost any problem involving the factors which are influential in one method are also
productlon from a petroleum reservolr; finally he is equally effective In the other For this reasoil it IS
confronted by a n Inadequate amount of lnformatlon per- deemed advisable to discuss bnefly these factors before
talnlng to the mechanics of flow in the reservoir This discussing the empirical relationshp studied by the
arlses not from the scarcity of information which has writers
been pubhshed In the field, for since the origlnal work
of Darcy '"In 1856 and the noteworthy contributions of
Fundamentals
King and S l l c h t e r i 9 n 1899, many hydrologlsts and
physlclsts and, in recent years, many petroleum engi- I t IS well known that, for the flow of hon~ogeneous
neers have investigated fluld flow in porous media In- fluid through a porous mediun~undcr steady-state con-
stead, the d~fficultyarlses from the fact that by f a r the ditlons, the volume rate Q a t which llquid wlll flow radi-
greater part of the work has been confined to homogene- ally into a well bore IS given by the formula
ous flow, whereas only a few investigators have pub-
lished material relatlve to heterogeneous flow In porous Q=-- 2 ~ k h A P- 2 ~ k h ( P . - P w ) (1)
re re
media p log. -
r w
p log, -
rw
Experllnental work wlth heterogeneous systems is dlf-
ficult to perform, and to date ~t has not been found "here I; 1s the ~erineabllityof the porous body, 11. IS ~ t s
practicable to institute routine permeablhty measure- thickness, p Is the vls~osityof the fluld in the reservolr ,
ments ii~volvingheterogeneous flow For these reasons AP lS the pressure between the and
It has beell found necessary to attelnpt to relate infor- the parts of the reservoir relativelv unaffected by the
matlon obtained, f o r the most part, from homogeneous well, a distance from the well given by y c , and rluis the
flow studies to the actual behavlor of wells alld well radius The foregoing equation, whlch will be
reservoirs recognized a s an expression of Darcy's law, assumes
The present-day effortsto evaluate reservoir behavior l n c o m ~ r e s s l b i l i tof
~ the llc]uld If the fluid were one
fall, primarily, into two classes which would essentially obey the perfect gas law, the
1 The application of Darcy's law for homogeneous rate of flow Q into the well would be given by the
flow, w t h modifications for the heterogeneity of the
exPresslOn
rkh(Pe2-PW2)
system and with suitable adjustments for variations in Q= (2)
* General Fetrolel~mC'orp Los Sngeles C a l ~ f
rw
P log.
re

t Umon 011 Co of ~a11fo;nta Los ~ n g e l e s C n l ~ f

$ Prebented, hy hfr ~herhornc:a t Twenty-f&rth Annual Meet- Inasmuch as the writers have ~nsufficlent data a t
Ing. Chlcngo. 111, Nov 1043
Ftgures refer to REREREXCES on p 60 present to evaluate gas reservoirs emplncally, the dls-
cusslon wlll be confined to flow of liquids and, conse- crude oil the SPZ will be more closely approached by
quently, primarily wlll lnvolve equatloil (1) the relation
Wells a r e class~fiedby means of the prodlictrvzty wt- 27rk
dex 10. n which 1s equal to the number of barrels per SPI = Q= -
r (6)
day of gross llquid * produced per pound per square l'(AP) p ( F V F ) log. 2
r w
inch pressure dlfferentlal existlng between t h e reservolr
and the well-bore The f o r m a t ~ o nvoluine of the tank oil v a n e s from place
to place 111 the reservolr a s the pressure and tempera-
t u r e vary throughout the system, and a mean value
n u s t be assumed As a result of the researches of
It 1s evldent from equation (1) that, everything else Sage, Lacey, and coworkers,*' and Katz and coworkers,"
b e ~ n gconstant, the amount of productlon to be expected ~t is possible to predict reasonably well the formation
from a well wlll be a direct function of the thickness
of the productive zone F o r this reason, In attempting
to use well data a s a means of evaluat~ilgthe inechanlcs
of oil productlon, t h e use of a factor lnvolvlllg thickness
IS des~rable The use of the term speczfic p ~ o d l t c t ~ v z t y
zyzdex has been estabhshed 111 prevlous publications lo " 10
The speclfic p r o d u c t ~ v ~ t~y n d e xIS expressed a s the
X
number of barrels per day of gross liquid produced per W
0
pound per square ~ n c hpressure differentlal per foot of z
sand, and by the equatlon ,l o
k
S P I = Q=
h(AP) h(Pe-Pw)
Q , (4)
3
I-
U
The speclfic productivity ~ n d e sf o r a n honlogeneous in- 2lo-'
coinpresslble fluid may be related to equation (1) a s
follows
Ba
Q 27rk u
SPI = 1
LL
310'
h(Pe-Pw) re (5)
EL loge W
a
cn
F o r the case to which equations (1) through (6)
apply, it 1s assumed t h a t gross liquld and tank 011 a r e
lo-'
identical Where a n apprec~ablepercentage of t h e pro-
duction IS water, these equations lnust be inodlfied
appropr~ately Thls wlll be discussed In more detail
hereinafter
10 lo4
PERMEABILITY - MILLIDARCYS
Factors wh~chAffect the Spec~ficProductivity Index
Var~at~on
In Specific Productirr~tyIndes w ~ t hFormat~on
T1z.e fo~-~~aatton .~loljclnef a c t o ~ It will be noted t h a t
Volun~eFactor.
the SPZ IS glven In terms of gross liquid a t A P I standard
conclltlons" It 1s well recognized t h a t t h e voluine occu- FIG. 1
pied by tank oil under standard cond~tionsdoes not
represent the volume occupled by this oil when assocl- volume factor f o r a crude oil over a wlde range of
ated wlth ~ t sproduced gas a t reservolr conditions pressure, temperature, and gas-oil ratio, with a mlnl-
Rather, 111 general, ~t wlll occupy 111 the reservolr froin muin of data I n F l g 1 a r e presented 2 theoretical
one to two t ~ m e sa s much space a s ~t does a t standard curves showlng the relatlonsh~p between speclfic pro-
condltlons The voluine to which a barrel of t a n k 011 ductlvlty ~ n d e xand permeabihty over the range of for-
1s increased by associating it wlth ~ t g s a s under equi- nlatlon voluine factors generally encountered
l ~ b r l u l nreservolr cond~tlons1s known .as the fonnatzon It can be seen In F l g 1 that, wlthan this range,
volztn~e f a c t o , Under some conditions free g a s will changes in the formatloll voluine factor ( F V F ) have a
e x ~ s t111 equ~llbnuniwlth the liquid petroleuni in the sinall effect upon the productlvlty-permeability (SPZ-k)
reservolr Inasmuch a s thls paper deals wlth the flow relationship, relatlve to t h a t caused by v~scosity
of llquids, t h e volume of the saturated reservoir hquld D,raz?~c~geT ( L ~ L L L SThe SPI IS affected by the slze of
h a s been used instead of the forinatloll volume factor the well bore and the effective dramage radius a s shown
f o r these cases It follows, then, t h a t f o r the normal i n equation (5) It 1s evldent t h a t lrregularitles In the
cyllndrlcal surfaces defined by these radil wlll affect
* Terln* used In tlus paper a r e a s defil~etl In t h e "Glossary the SPZ. The magnitude of these lrregularitles and, in
R e l a t ~ n - to rieserrolr Behi~rlor," ~ > n b l ~ s h ei nd t e u t n t ~ r eforlu i u
API ~ r % d ! t c t r o tBull
~ A70 208, p 86 (1941) , t o be presented in ' fact, the magnitude of the dralnage radius a r e seldom
rerlsed form before t h e A P I T o p ~ c a lC o m m ~ t t e eon P r o d u c t ~ o n
Tecl~nologg,N o r 1043, C I ~ I C B ~I11 O, known in practice Indeed, the drainage radius 1s
known not to be a constant, but varles throughout the I by the varylng compos~tlonsof dlfferent crudes More
productive life of the well '' Fortunately, however, rela- recent data indlcate t h a t such variations do esist ''
tlvely large v a r ~ a t i o n sIn these radll have little effect However there is such a scarclty of published d a t a on
upon the SPZ The effect t h a t changes In the ratio of the vlscoslty of multi-component systems t h a t the au-
drainage radius to well-bore radius have on the produc- thors belleve f o r t h ~ sstudy the use of the inethod of
tivity-penneab~htyrelationship is shown In F l g 2 pred~ctlon referred to previously herein is a t present
Vzscoszty Vlscoslty h a s a most profound effect upon justified Because of t h e utlhty of this method of pre-
the flow of 011 through the reservolr Viscosities of Cali- dlctlon in the application of the data of this paper to
fornia crude oils under reservoir conditions vary from reservolr flow problems, ~t1s repeated here I n this pro-
30,000 centipoises to a s low a s 0 1 centipolse If con- cedure the viscosity of the tank 011 1s deternllned a t
densate reservoirs were ~ncluded,the range would be atnlospheric pressure and 100 deg F If v~scosityd a t a
extended to about 0 02 centlpolse There a r e plotted in a r e not ava~lable,a reasonably rellable value may be
Flg 3 curves whlch show the theoretical productivity- obta~ned from Fig 4 F o r the calculation of flow

10 loa lo3 lo4 10 1oz 1o3 10'

PERMEABILITY - MILLIDARCYS PERMEABILITY - MILL1 DARCYS
V a r ~ a t ~ oIn
n Specific Product~v~ty
Index w ~ t hV ~ s c o s ~ t y .
FIG. 3
FIG. 2
I condltlons in the reservolr, this vlscoslty must be re-
perlneablllty relationship for dlfferent reservolr vlscosl-
duced to t h e vlscoslty of t h e reservoir 011 To do thls
t ~ e s It may be seen from these curves t h a t the viscoslty
the appropriate inultiplylng factor is obtained from
exerts a s g r e a t a n Influence upon the flow of 011 In t h e
F i g 5 o r Fig 6, depending upon the nature of the oil
reservolr, a s does the permeability itself
F i g 5 and 6 represent reservolr vlscoslty expressed a s
The v~scosityof the oil flowing 111 the reservoir may per cent of tank-oil viscos~tyf o r Dominguez and S a n t a
differ many fold from t h a t of the tank 011 Tempera- Marla Valley oils when associated with their respective
ture, pressure, a n d d~ssolved g a s contribute to this gases a t various pressures and temperatures These
variation A number of petroleunl systems have been curves a r e constructed from d a t a deter~nlnedby Sage
investigated '-3 ' 3 34-ai to determine the effect of the solu- and Lacey and coworkers 3' ' When no direct viscoslty
t ~ o nof g a s a t elevated temperatures and pressures On data a r e available, F i g 5 1s used for crude oils hav-
the basis of d a t a thus obtained, efforts have been made ing gravities of 20 deg A P I and h ~ g h e r ,and F i g 6 1s
to estabhsh methods of predlct1ng the vlscos~tyof crude- used f o r crude 011s havlng g r a v ~ t ~ below
es 20 deg A P I
oil systems under reservolr conditions '' It w a s recog- The r e s u l t ~ n gvalues a r e used a s the reservoir viscoslty,
nlzed t h a t such a inethod mlght have limited applica- p , In the equations
tion because of the effects which mlght be ~ntroduced Flzizcl heteroyene~ty As Indicated previously, t h e
 relatively small amount of experimental investigation gations have been confined with very few excep-
of the flow of heterogeneous fluids reveals that the rate tions=* to essentially homogeneous porous solids.
of flow of any one phase in the presence of appreciable Very few formations are homogeneous, either laterally
quantities of other phases is reduced materially below or vertically, for more than very short distances. I t is
that for an homogeneous fluid. The permeability to true that, for a structure composed of relatively thin
flow of a single fluid phase in the presence of other laminae separated by impervious barriers, the total
fluid phases is defined a s the effective permeability.
, The effective permeability to oil is designated by k,.* '

z
8
E 0 500 1000 1500 2000
SATURATION PRESSURE-PSI
2590 3000

50 Viscosity Reduction with Gas Saturation for Dominguee

> Crude Oil.
k FIG. 5
cn
8
cn
> 10
LJ
I-
30 5

1
10 20 30 40 50 60
GRAVITY AT 60 DEG F - A P I DEG
Average Viscosity as a Function of Gravity for
California Oils.
FIG. 4 SATURATION PRESSURE-PSI
Viscosity Reduction with Gas Saturation for Santa Maria
I t has been shown further that, for any given system, Crude Oil.
the effective permeability k. is largely determined by
FIG. 6
the relative quantities of the various phases p r e ~ e n t . ~
I t is axiomatic that for any system containing hetero-
geneous fluids the effective permeability k, must replace rate of flow is merely the sum of the rates for the indi-
k in equation (6). vidual strata, each of which may be considered homo-
geneous. However, for the condition in which no bar-
2rk. riers separate the pervious layers, the mathematical
SPI =
p ( F V F ) log. -5 analysis of the mechanics of the resulting oblique flow
rr has appeared so complex a s to have prohibited the work-
Formation heterogeneity: Insofar as the writers ing out of a general solution. Among these problems
know, previous theoretical and experimental investi- only a special case for the flow of a single fluid through
two adjacent discs of differing permeability has been
Nomenclature conforms to that of A P I permeability code solved.""Therefore, it has been found necessary to
No. W.'
resort to a simple average of the available permeabillty 3 To enhance our knowledge of the mechanics of
data a s obtalned In the laboratory Sllnple arithlnetlc fluid flow In reservolrs
and weighted averages have been t r ~ e don many wells, The prev~ouslypubhshed curves showlng t h ~ srela-
and ~t h a s been found that, where frequent samphng tionship f o r Callfornla wells have been questlolled by
has occurred, the a r l t h m e t ~ caverage is a s satisfactory investigators '8 ?3 with experience ~n Eastern and
a s a welghted average It 1s possible t h a t the applica- Mid Continent areas F i g 8 is presented f o r compari-
tion of statistical methods to the analysls of perme- son of the d ~ r e c td a t a with mod~fiedd a t a given hereln-
ability data a s recently presented by Law Is may prove after, and also to Illustrate that, f o r the more diversified
fruitful However this may be, it does not appear pos- condlt~onsexlsting in the reservoirs here represented,
sible a t present to evaluate quantitatively the magnl- such a slnlple correlat~on 1s not adaptable without
tude of the effect t h a t the non-homogeneity of the porous modification I t will be noted, however, that, wlth t h e
body has upon the flow relationship exception of a few wells of hlgh permeablllty sand
which also happened to have hlgh-vlscoslty oil, the data
Rap~dDeterm~nationof SPI from Pern~eabilityData represent essentially the same curves a s those prevl-
ously published for Cahfornia wells a s mentioned above
If one wlshes to express the specific productivity In- Since t h e publ~cationof these curves, data have be-
dex In terms of barrels per day per pound per square come available which perinit a partial evaluatlon of
inch pressure d~fferentialper foot of sand, equation (6) some of the modifying factors It is the present pur-
becomes pose to exalnlne the productlv~ty-permeabil~ty correla-
3 073k tion In the light of these factors Thls is most r e a d ~ l y
SF1 = (8)
p ( F V F ) 10g.o 2 done by multiplyng the speclfic productivity index by
the forlnatlon volume factor, t h e reservolr v~scosltyof
I n this equation k 1s the average permeab~htyIn darcys, the 011, the log of the ratlo of radil and, when neces-
p 1s the reservolr vlscoslty in centlpolses, F V F 1s the sary, by a factor for the water cut of the well, a s shown
f o r n ~ a t ~ ovolume
n factor, I.. and r , a r e drainage and by the follow~ngequation
well radn, respectively, in feet T h ~ equat~on
s lends it-
S P I ( F V F ) (p) (loglO-fi) (1-C)
self to graphical solution by means of a n allgninent r w
chart, such a s t h a t shown in F i g 7 By means of this
@ (SPI) =-
3 073 (9)
graph ~t 1s posslble to determ~net h e spec~ficproduk-
1u1~el.e+ ( S P I ) 1s the lnodlfied speclfic productivlty m-
tivlty index from the permeability-provlded the vls-
dex, and C 1s the fraction of gross.production which 1s
coslty, fornlation volunle factor, and dralnage radlus
water I t is recognized t h a t the factor (1-C) does not
a r e known If the effectwe permeablhty IS known, ~t
necessarily represent the proper correction f o r the
may be subst~tutedf o r the permeability-whlch will, of
presence of water However, ~t 1s assumed to be very
course, ~ n c r e a s ethe accuracy of the result
unhkely t h a t oil and water a r e produced sllnultaneously
from the same portion of the formation If this assump-
Well Productlv~tyas Related to Core Pern~eabll~ty tion is accepted, and ~f there is included In the thick-
Data from 141 wells, representing more t h a n 12,700 ness (IL) only the amount of sand assoelated w ~ t hoil
individual permeability tests, a r e shown In Table 1 production, then the cut factor ( 1 4 ) reduces the gross
Although nearly all of the values given In Table 1 SPI to a functlon of the net 011 production F o r the
were experimentally determined or obtained a s pre- relat~velyfew wells which produced appreciable quantl-
viously described, it h a s been necessary to estimate tles of water Included In t h ~ sstudy (24 above 10 per
one factor f o r a few wells when the other data were cent), the above procedure h a s been applled Of the
rellable Such wells a r e grouped a r b i t r a r ~ l yIn the low- wells studied, 74 had cuts of less than 1 per cent, and
est "quahty" classlficat~on The data of Table 1 a r e the great l n a ~ o r i t yof cuts were less than 10 per cent
plotted In Fig 8 Here the spec~ficproductivlty Index Wells f o r which good d a t a were available, but havlng
1s plotted a s a dlrect function of the a l r permeabillty lugh cuts, were Included i n order t h a t the effect of the
of'cores from each well, a s was originally done by cut may be observed Wlth t h e esceptlon of one well
Pyle and Sherborne," and later amplified by Sher- which had a cut of 76 per cent, but w h ~ c h1s belleved
to have been otherwise poorly completed, the polnts f o r
borne 38 and Johnston '' It has been recognized by these
the wells with hlgh cuts a r e thoroughly Interspersed
authors t h a t no such s ~ i n p l erelatlon a s this can be
with those from d r y wells
expected accurately to account for the performance of
F o r the t h ~ c k ,lammated structures encountered so
any given well Nevertheless, the development of such
generally In California, ~t is virtually ~ n ~ p o s s i b lto
e
a r e l a t ~ o n s h ~ peven
, though restricted III its scope to
asslgn a value w h ~ c hwlll be representative of the aver-
wells In reservolrs of s l n ~ l l a nature,
r was felt to be justi- age dralnage radius Fortunately, however, a s is ~ l l u s -
fied for the following reasons trated in F l g 2, the cho~ceof a value of 2,000 f o r the
1 To ald in the pred~ctlonof well perforinance p n o r ratlo of the dralnage radius to t h a t of the well bore In
to colnpletlon inost of the cases studied introduces less than 5 per
2 To ald in the evaluatlon of the quahty of the well cent error Consequently, the value of 2,000, w h ~ c his
completion believed to be reasonably representative of the rat10 of
 FVF
rw

Al~gnmentChart for Compr~t~ng

Well Prodnctivity.
FIG. 7
 TABLE 1
SPI-k

".&

Sand Grav Forma-

Qual- Th~ck- Perme- I ~ Y Cut Temper Pres- t~on
Well ~ t of
y ness ab111ty (Deg (Per ature Gae 011 Volume P
No F ~ e l d Data (Feet) E SPI t PI S API) Cent) (Deg F) Ratlo Factor (Cent~po~ses) @(SPI) M(SPI)
1 A d 39 8,500 0 0119 0 465 14 4 8 100 271 40 112 516 6 33 243
2 A d 46 7,000 0 00296 0 137 14 5 14 100 430 65 112 390 111 51
3 A d 109 6,000 0 00300 0 325 14 7 25 100 330 50 112 400 101 110
4 A d 87 6,000 0 00261 0 227 14 5 42 100 407 60 112 403 0 68 59
5 A d 85 5,000 0 00458 0 385 14 6 4 100 407 60 112 384 190 162
6 B c 95 3,820 0 0035 0330 14 0 6 95 76 14 103 981 3 45 328
7 C b 45 3,000 0 0162 0 73 15 3 1 100 400 15 108 390 6 72 302
8 B c 90 3,000 0 0067 0 60 14 0 7 95 253 40 103 765 4 95 445
9 B c 90 2,500 0 0087 0 78 138 30 95 301 50 103 840 5 16 696
10 B c 135 2,215 0 0081 109 135 45 95 296 50 103 1,040 4 84 650
11 A a 81 2,120 0 0054 0437 15 18 100 166 30 112 410 2 01 162
12 D b 100 1,385 0 119 11 9 27 8 0 215 3,613 600 134 0 90 0 144 14 4
13 E c 47 1,012 0 51 24 2 40 2 0 250 4,670 1,200 170 0 107 0 093 4 37
14 D a 286 875 0 164 47 32 5 0 205 3,260 620 136 0 46 0 010 - 2 86
15 E a 13 785 0 048 0 63 38 5 3 250 4,155 740 141 0 217 0 014 185
16 F a 62 745 0 030 185 33 0 22 146 2,425 590 133 106 0 033 2 05
17 G c 220 715 0 0040 0 88 28 8 1 186 , 878 160 113 3 87 0 017 3 74
18 H b 246 696 0 0104 2 57 28 8 2 152 488 90 108 3 78 0 092 22 6
19 E d 113 694 100 113 41 2 0 250 4,640 1,200 170 0 105 0 178 202 '
20 E b 37 657 0 81 30 39 9 1 251 4,780 855 147 0 172 0 203 7 51
21 J c 50 657 0 00388 0 194 29 4 1 207 1,900 420 120 135 0 0062 0 300
22 D a 518 620 0 232 120 32 3 0 205 3,172 650 136 0 58 0 183 94 9
23 D a 572 575 0 320 183 32 2 0 205 3,220 600 140 0 64 0 286 164
24 E c 36 562 0 17 6 13 39 5 0 250 4,466 1,100 160 0 118 0 302 108
25 D d 663 545 0 302 (200) * 30 1 0 205 3,280 640 140 0 78 0 329 218
26 D a 445 545 0 145 64 29 6 0 215 3,510 - 600 133 0 71 0 136 44 5
27 D d 164 527 0 060 98 26 7 0 215 3,680 600 134 11 3 0 091 14 9
28 D a 491 514 0 079 39 31 8 0 205 3,155 655 136 0 62 0 067 32 9
29 D a 553 451 0 157 87 32 2 0 205 3,110 725 136 0 55 0 118 55
'30 D a 576 439 0 100 57 32 2 0 205 3,190 620 136 0 62 0 084 48 5
81 D d 492 .426 0 203 (100) * 29 1 80 205 3,365 640 136 0 85 0 231 114.
32 E b 47 421 0 27 12 6 41 0 0 250 4,526 1,100 160 0 127 0 055 2 31
33 E b 45 383 - 0 111 50 39 6 1 250 4,820 1,100 170 0 138 0 026 120
34 D a 573 366 0 122 70 31 7 0 205 3,200 615 136 0 65 0 108 61 1
35 K c 80 350 0 040 32 37 3 3 211 3,135 960 150 0 365 0 021 170
By "qual~tyof data" (rat~ngsa, b, c, d) 16 meant the frequency or adequacy of permeab~l~ty tests R a t ~ n ga means more than 1 a m p l e for every 2 f t of sand, b means 1 sample
for 2 f t to 4 f t , c means 1 sample for 4 f t to 8 f t , and d means that one sample represents nlore than 8 ft of sand, or that an estimate of some k ~ n dwas necessary
f Speclfic product~vity Index
$ Product~vity index
 TABLE I-Continued
Reservoir
2-
Sand Forma-
Qual- Thick Perme- Temper Pres- tlon
\!'ell ~ t of
g iiess ability ature sure Gas 011 Volume
No Data ' (Feet) li (Deg F) (PSI) Ratio Factor
350 211 3,000 700 139
346 170 970 150 113
340 250 4,115 800 145
334 178 2,005 540 127
333 215 3,680 600 1 34
330 250 4,225 800 145
323 210 1,400 600 137
309 178 2,035 540 126
301 250 4,120 900 .. 1 4 5
295 142 965 175 1 12
277 210 850 250 113
270 205 3,280 640 1 36
260 211 2,985 880 143
260 210 3,125 900 144
233 162 3,345 450 126
228 211 3,118 880 143
214 209 1,397 230 1 18
211 205 3,245 570 136
210 211 3,260 1,000 155
200 200 1,505 275 119
199 210 1,479 260 1 20
194 160 473 100 108
191 220 1,920 450 128
189 165 1,021 300 126
182 250 4,439 1,100 170
180 211 3,215 980 149
180 212 3,345 880 144
169 252 4,680 1,100 170
165 211 3,185 980 143
164 210 1,200 250 1 16,
161 172 1,355 250 1 16
161 235 4,850 1,230 168
158 211 3,181 890 144
153 240 4,540 1,100 170
151 210 2,720 670 143
 TABLE

Sand Forma-
Qunl- Thick- Cut Temper- Pres tlon
Well ~ t of
p ness (Per ature 6Urf: Volume P
No Fleld Data (Feet) Cent) (Deg F) (PSI) Factor (Centlpoises)
141 0 158 462
403 0 210 2,810
72 7 211 2,897
68 0 240 4,070
124 12 211 3,110
45 3 211 3,500
363 12 170 860
222 0 228 4,012
174 3 211 3,139
293 1 207 2,850
34 1 210 975
180 2 163 1,050
467 10 210 3,000
32 3 211 2,875
35 9 207 1,620
306 0 210 2,390
308 1 216 2,260
75 8 150 1,545
72 0 210 2,295
271 5 210 2,645
325 9 228 2,775
370 2 210 2,760
109 8 210 1,250
530 12 210 2,520
310 0 210 2,840
316 0 200 2,420
61 0 203 2,000
310 0 210 2,960
75 4 210 2,760
123 0 180 2,115
100 40 150 1,573
86 0 210 1,975
320 3 200 1,350
196 1 210 2,840
398 7 170 1,980
". O G O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O
the drainage radius to well radius, has been used in The array of poiiits in Fig 9 is such that a reason-
evaluating @ ( S P I ) Thus equation (9) IS reduced to able curve can be drawn through the trend. The devla-
the following expression tions of the indivldual @ ( S P I ) values from the curve
have been determined for each value of permeability. To
facilitate inspection of the data, these deviations have
This function, a s related to perlneability In millidarcys, been expressed for points above the cullre a s the ratio
is plotted in Fig 9 There IS also plotted In Fig 9 a of the actual +(SPZ) to that of the curve for the same
lliie representing the theoretical relationship between permeability For points below the curve the reciprocal
SPZ and k, in inillidarcys, for the cond~tionsgiven by of this ratio has been taken The deviations thus ob-
the equation tained a r e given in geometrical progression in Table 2-
SPI=O 000931k (11) the points above the curve being shown a s positive
In this equation a viscos~tyof 1 centipo~se,a formatlon values, and those below a s negative For this method
volume factor of 1 0 , a cut of 0 per cent, and a ratio of presentation, a deviation ratio of 1 0 represents per-
of radii of 2,000 have been assumed fect agreement with the curve The average mean devi-
ation of 2 98 is a clear indicabon that factors not repre-
sented 111 equatlon (10) have an Important effect upon
the relationshlp In fact, ~tcan be seen from the maxi-

TABLE 2
Dev~atlonof Indiv~dualValues of + ( S P I )
Fro111Actual Curve
Number of Deviations
Range of A

Dev~ation Positive Negative

Total
Mean deviation + 273 - 3 27
Masiinuin deviation +17 2 -17 8
Average mean deviation & 298

mum devlations, which reach a magnitude of 17 8-fold,

that these factors ( a s yet unexplained) are of para-
mount importance The effect of some of these factors
such a s fluid heterogeneity, well completion, e t c , will
be discussed below
Speclfic Product~v~ty
Iudex as Related to Pernieabil~ty. The deviation of the individual points from the theo-
FIG. 8 retical curve have also been determined for each value
of permeability, and have been expressed a s the ratio
Interpretat~onof Results of the theoretical +(SPZ) to that of the indivldual
values These data are grouped in geometrical progres-
An inspection of Fig 9 reveals that the application sion 111 Table 3 The mean deviation froin the theoretical
of all of the quantitatively known factors to the simple so calculated is 30 9, whereas the maximum devlation
relationshlp between k and SPZ does surprisingly little IS 443 The magnitude of these devlations merlts coii-
to alter the spread of the points shown in Fig 8, ex- siderable discussion, which wlll be given later.
cept for those polnts corresponding to high permeabili- The deviatloii of the average California @ (SPZ) curve
tles and low-gravlty high-v~scosityoil As can be seen, from the theoretical is shown in F i g 10 It is note-
these particular values have been shifted to a posltlon worthy that the deviat~onof the average curve from
very close to the theoretical curve This is believed to the theoretical curve increases uniformly with a de-
result from the fact that the wells represented by these crease in permeability within the range of the data
points are 111 reservoirs in whicli there is little inter- studled For perineabilities ranging from 10 to 20 milli-
stltial water and essentially no free gas Flow under darcys, the average curve yields values amounting to
these conditions should approslmate that p v e n by only one-fift~ethof the theoretical
Darcy's law In referring again to Fig 9, it is apparent that some
of the points he above the theoretlcal curve Thls 1s can be seen in Fig 9 that such a correction has done
belleved to result from one or more of the following little to narrow the general array of points I t is true,
c~rcumstances of course, that actually known values of reservoir
1. The polnts represent very hlgh-permeability un- viscos~tywere avallable for a relatlvely small number
consol~datedsands, and the measured values after re- of the wells shown and that, a s a consequence, most
compaction may not be representative of the v~scoslt~es were obtained somewhat a r b ~ t r a n l y ,
2 Portlons of the core representing the more perme- a s described previously I t 1s quite likely, therefore,
able strata nlay not have been recovered that for the great range in the nature of the crude oils
3 The productivity Index becomes increasingly diffi- under consideration (13 to 44 deg API gravity), closer
cult to measure accurately In the hlgher ranges agreement will occur when better viscosity data are
There can be no doubt that, In spite of the fact that available
the data studled In this investigation represent only the So f a r ~thas not been quant~tat~vely
posslble to a d ~ u s t
most conlplete and accurate m a t e r ~ a lavallable, experl- the productivity-pern1eabillty relation for the effect
mental error and non-representativeness of data give upon flow introduced by heterogeneity of the fluld sys-
rise to some of the varlatlons which occur However, tem I t has been esperlmentally shown that, for hlgh-
the magnitude of thls effect 1s d~fficultto evaluate for

TABLE 3
Deviat~onof lnd~vidualValues of @(SPZ) from
Theoretical Curve
Range of Number of
Devlatlons Samples
0-1-fold 3
1-4-fold 16
4-16-fold 56
16-64-fold . 50
64-256-fold 12
256-fold+ 2

Total PERMEABILITY -MlLLlDARCYS

Mean deviatlon Deviat~o~i

of Spec~fic-Product~v~ty-Index
Functioii from
Theoret~ealHomogeneous-Flow Curve.
Maxlnlum deviatlon
FIG. 10
data comlng from so many different sources a s have
those reported in this paper perlneabillty unconsolidated sands '". 'O and for one con-
The effect of variations In the formation volume factor solidated sand having a perineabllity of approximately
upon the average + ( S P I ) curve has been shown In 500 n1illida'cys,3 both lnte~stitialwater and free gas
Flg 1 to be of relatively small magnitude For the may have a profound effect upon permeability. These
range of FVF from 1 0 2 to 177, whlch covers the data data indicate also that, for less-permeable more-con-
hereln reported, a maximum additional deviation ratio solldated sands, the prepence of water and gas has a
of less than 1 7 5 would occur ~f the FVF were com- greater effect upon the permeabihty than has been noted
pletely neglected However, this value has been ob- for more-permeable unconsolidated porous media Thls
tained and applied dlrectly whenever possible Where 1s in keeplng with the observed field data and labora-
none was available, it has been estimated by a method tory data obtalned for California formations To date,
of predict1011 simllar to that recently suggested by practically nothlng 1s known esperlmentally about flow
Katzlc based upon a s yet unpublished data for Call- In the piesence of these fluld phases for sands contain-
fornia crude 011s lng relatlvely large quantities of argillaceous o r siml-
Unlike the relatively unimportant effect upon the lar material There is every lnd~cationthat such ma-
average @ ( S P I ) -k relatlonsh~pexerted by the forma- terial may be of considerable llnportance In lts effect
tlon voluine factor, viscoslty can influence the correla- upon the fluids in the reservoir. This effect 1s believed
tion a s dominantly a s permeablllty, a s has been ~ n d i - to be respons~bleto a great extent for the large devia-
cated In Flg. 2 I t was bel~eved earlier that, when tions dep~ctedIn Fig 9 and 10.
rellable v~scosltydata were avallable, the productlvlty- I t llkewise has been lnlpossible to show for the ma-
permeability relabonship could be made to conform jorlty of cases the quantitative effect that free gas has
much more closely to the theoretlcal curve Although had upon the data However, for one well, indicated in
such is the case for some of the points, a s was noted Flg 9 by a triangle, data are available for reservoir
previously herem, and although correct~onfpr viscoslty condit~onsunder which relatively large quantities of
results In a thorough rearrangement of the points, ~t gas had heen-freed fi*oln solution I t 1s believed that
thls fact esplalns In part, a t least, the very low produc- adjustment for still another major factor Any remain-
t ~ v l t yvalue ing divergence then would be largely ascribable to vari-
Besldes the factors mentioned previously, others of atlons In well-complet~onpractice.
lnajor ~mportance,but a s yet not suscept~bleto quantita- In order to present the data In a manner s l m l a r to
tive evaluation, are those attending well completion that published for some fields outslde of Californ~a,"
wh~chinclude the effect of drilling mud, drilling-mud both the speclfic productivity index function and the
filtrate, cement, the llner itself, the nature of the liner perineabil~tyhave been multiplied by the thickness of
perforations, the amount of plugglng or the degree of 011-productive format~oaopen to the well The result-
effectiveness of the perforations, and others Although ing correlation, which will be recognized as a funct~on
some of these factors have been Investigated to a limited of the product~vity Index @ ( P I ) ,has been plotted In
extent.'5. " " much inore ~ n f o ~ r n a t ~must
on be made Flg 12 Here, again, ~t w ~ l lbe seen that the slope of
available before their quantitative effects can be evalu- the trend deviates from the theoretical of 45 deg, the
ated Values for wells which are definitely known to
be poor completions have been circled In Flg 9. I t can
be seen that these are among those havlng the greatest
d~vergencefrom the theoretical productlvlty curve.
There IS one point in t h ~ sconnect~on,however, which
IS worthy of note I t IS that the general technique em-

II 1 W E L L WAS KILLED
WITH WATER
AT T I M E A I 1

.-
TlME - YEARS
Effect of Water upon Well Productiv~ty.
FIG. 11

ployed In well completion ln Callfornla cannot dlffer PRODUCTIVE CAPACITY -MILLIDARCY FEET
greatly froin that practiced elsewhere, and, therefore,
Modlfied Product~vityIndex as a F u ~ ~ c t i oof
n
the wide divergences found for the California data must
Productive C a p a c ~ t ~ .
result not so inuch froin the lnethods of completion em-
ployed a s from some fundamental difference in the be- FIG. 12
havior of the reservoir dralned by the wells It has
been found that cores conta~nlng~nterstitialand drill- steepness of the slope indicating larger divergences a t
ing water have lower permeability, not only to air but ?ow values of perineablllty than a t h ~ g hvalues
also to water, than do the same cores to alr after dry-
Ing This effect 1s apparently more marked for Call- SUMMARY AND CONCLUSIONS
fornia than has been reported for cores from other
areas Furthermore, ~t has been shown by Travers3' Data from 141 wells in Cahfornla have been analyzed,
that, for some pools, the productivity Index of a well and a correlation between specific productlvlty index
decreases the longer the drilllng fluld has stood agalnst and permeabil~tyhas been developed for those wells.
the formation The loss of well product~vltythat re- Factors whlch influence this relatlonshlp have been dls-
sulted when a well was k~lledw ~ t hwater IS illustrated cussed and, wherever possible, have been quantltatlvely
in Fig 11 evaluated When insufficient speclfic information was
I t would be desirable to have data s i m ~ l a rto that of available, methods and charts have been developed and
Leverett and Lewis2' for multl-phase fluld flow for the used to predict the value deslred These and other
reservoirs hereln discussed I t then would be posslble charts useful In coinputlng productivity values are de-
to evaluate the effect~vepermeabillty In the manner de- scribed herembefore Possible differences between Cali-
veloped by Babson ' Such an analysls would permit the f o r n ~ asands and forinat~onsoccurring elsewhere have
 been discussed, and a n effort h a s been made to indicate 16 D L I<atz "Predlctlon of t h e Shrinkage of Crude Oils,"
Drillzng a n d Prodtrction Practzce, 137-47 (1942)
t h a t these might be Influential In bringing about the " B P Kantzer a n d E G Trostel. "Oil-Well Performance-
A D~scussiona n d Proposed Terminology," Drtlltng a n d Prodtrc-
relat~velywlde d~vergencefrom the theoretical homo- tion Practice. 118-127 (19371
geneous-flow curve found f o r the C a h f o r n ~ awells a s
compared to the smaller d~vergencesreported upon f o r
wells from regions o u t s ~ d eof Callfornla
ncterlstlcs of Reservoir Snnds "- presented before a n n u a l mert-
, I t 1s felt that, even though the foregoing data com- Ing, Petroleum Div , dnz I n s t 'Mint?zg Met E n g r s , Los Angeles.
prlse a tremendous number of individual ~nvestlga- Cnllf , Oct (1043)
tions, they are a s yet incomplete because of the lack of
" hI C Leverett:, "Flow of 011-Water &fixtures through Un-
consolidated Snnds, T r a n s Am I n s t Mzning Afct E n g r s 132,
the necessary quantitative value f o r some of t h e factors 149-69 (1939)
-'&I C L e r e r e t t a n d W B Lewls, "Steady Flow of Gas-
whlch a r e belleved to be of major ~ n ~ p o r t a n c eNever- Water-Oil hllxtures through Unconsolidated Sands." T r a n s Ant
I n s t Mzrung Met E n g r s 142, 107-116 (1041)
theless, the correlation thus f a r developed 1s thought to 2 2 3 1 C Lererett D' B Lewls a n d hI E T r u e , "Dlmenslonal
blodel Studles of 011-Field ~ e h a ' v l o r , " T r a n s Ant I n s t Mznzng
have considerable practical utillty The wrlters belleve
t h a t .not untll these factors a r e more definitely estab- 23 <
B e t Etigrs 146 175-93 (19421
A ~ e w l s discussion
' following "Core-Analrsls Interpreta-
tlon, by N ~ d h n s t o n ,Drtlllng a n d ~ r o d ~ t c t i oPractice,
i 192
hshed wlll t h e fundamental mechanics of multl-phase (1241)
J A Lewls. W L Horner, a n d hlnrlon Stekoll, "Produc-
fluid flow be sufficiently understood t i v ~ t y I n d e s a n d hfeasurnble Reservolr C h a r a c t e r ~ s t ~ c s , ".4)n
I n s t Mz~ilngMet Cngrs Petrolelon Teclr Mar (1942)
?5 F G h l l l l ~ r ."Pres?pre D ~ s t r i b u t ~ onboiit
n a Slotted Llner in
ACKNOWLEDGMENT a Producing 011 R'ell, T r a n s d m I n s t Afzni~ig Met Eiigrs
142, 137-51 (1941)
The wr~tei-sa r e Indebted to the Cahfornla Distrlct S T V hloore, dlscusslon follomlng "Core-Analysis Interpreta-
tlo-n.- by N Johnston, Drzllt~ig a ~ i dProdrictton P r a c t ~ c e , 192
Subcommittee on Core Analysls and Electrlc Logglng, ( 1341)
"T V Rloore. R J S c h ~ l t h u i s .a n d \\'~'llli~rnH o r s t , "Deter-
under whose ausplces the data were collected, and In nllnnt1on of Permeability from Field D a t a , dPI Prodrtctzon
Bztll No 811 (1033)
particular M r W. L Jarvls of The Texas Company, hluskat, d~scusslon following "Core-Analysis Interpreta-
z8 ?! b r N Johnston. Drtlk?io a n d P r o d ~ i c t z o ~Practtce.
tlon. c 193
chalrman of the committee They also acknowledge all
those companies whose cooperation In submlttlng the
data studled made thls report posslble Particularly In-
debted a r e they also to Mr H N Marsh, General Petro- ' O'Brlen a n d J A P
leum Corporatlon, Major H C Pyle, C E , U S Army, zontal S t e a d s Flow III Porous hledia." Am I n s t Mtnrna Met
E y ? Petroiettm Tech J u l y (1941)
and to Mr E C Babson, Unlon Oil Company of Call-
fornla, from whom, In the course of a number of pleasur- .
E Olds. B H gage, a n d nr N Lacev "Volumetric a n d
Vlscos:ty Studies of Gas a n d 011 f r o m t h e 9a' n t a illaria Valley
Fleld, d m I n s t Mtnztzg Met E n g r s Petrolcum T e c h , hlar
able dlscusslons, the wrlters have gained many helpful 1144.1\
,----,
m H C Fyle a n d J E Sherborne. "Core Analysis," T r a n s
thoughts and suggesbons d m l?zst Minzng Met E ~ c g 132, 33-61 (1939)
W B I3 Raze and W N L ~ c e a ."Formntlon V o l ~ i m ea n d Vis-
coslts Studies f o r Dominruez Iheld." Drzlllna a n d Prodrtctzon
~ r a & t ~ c 141-7
e. (1935) -
REFERENCES 85 B H Sage, W hlendenhall a n d W N U c e y "Vlseos~tyof
I I y d r o c a r b o ~ , S o l u t i o n s ~ o l o t l c hof~ F o u r H d r o c ~ r b o nGases In
' A P I Code No R "Rtandard P r o c e d ~ t r e f o r Determtnlng a Crude 011 d P I P r o t l ~ t c t ~ oBull n No 216 &ov (1935)
Permeabllltu of Porous Medra.." Apr (1942) " B H sage, J E Sherborne, a n d W ' N Lacey, l'V~scosity
? E C :?bsou, "Fredictlon of Reservolr Beharlor from Labora- of ,yydrocarbon S o l u t ~ o n s E t h a n e a n d tiButane i n Crystal
tory Data, Am I n s t M z ~ i ~ nMet
g E n g r s , Petrolezinc techno log^ A P I Prodztctton Brill No 216, Nov (1035)
Ton
"I-\L
IlO dd)
"I-l O1:7B A Sage, J E Sherborne, a n d W N Lacey "V~scositv
a H G Botset "Flow of Gas-Llqu~d Mlstures through Con- of IIydrocarbon Solutions Methane a n d ~ r o p a n e ' l n Crystal
solldnted Sands," Trnns Ant I n s t M r n ~ n g Met E n g r s 136, 011." I n d Erig Chem 27, 954-6 (1936)
91-ina 11940) as J E Sherborne ( c h a l r m a n ) "Con~parison of Speclfic Sari!
4 ~ o s e ' p hChalmers. D B Tallaferro J r a n d E L Rawlins P r o d u c t ~ v l t rwlth Arerage Alr P e r m e a b ~ l l t yof Core Samples,
"Flow of Air a n d Gas through ~ o r o u s ' ~ e d l a ,T"r a n s Am ~ n s t '
A P I ~ r o d v E t t o nBull No 226, 129-31 (lC140)
Mznzlrg Met E ~ r g r s98, 3i5-94 (1032) aDW J T??vers, J r , "Complet~on Practices Related t o IVell
= H e n r y Darcy, Les F o n t a & ~ i eprtbltqnes
Vlctor Dalmont. P a n s (1858)
s de la Vtlle de Digon,
I'roductlrltr. dea I n s t Mz?cr~ioMet E ~ t o r s P e t r o l e n ~ ~Tecli
hlnr (1942) '
c .
H H Evinger a n d M AIusliat. "Calculatlons of F'roductlv~ty 40 R D n'yckoff a n d H G Botset." Th?, Flow of Gas-Llrl~ild
F a c t o r s for 011-Gas-Water Sybtems IU the Steady State," T r a n s l l l s t u r e s through Unconsohdated Sands. P h ~ s t c s 7, 325-45
d n i I?rst Afi~ilngMct E n g r s 146, 194-203 (1941) (1936)
H H Evlnger, a n d hi Muskat. "Calculatlons of Tlieoretlcal
P r o d u c t l v ~ t y Factor," T r a n s Anc I?ist Mtnzng Met E n g r s
146, 126-39 (1942) DISCUSSION
s G H Fanchpr and J A Lewls. "Flow of Slolple Fluids
through Porous hledla," Itrd Etcg Claenr 25, 1139-47 (?,033) Parke A Dickey (Forest 011 Corporatlon, Bradford,
s G H Fnncher J A Lewls, a n d 6 B Barnes Some
Physical Character;stlcs of 011 Sands." P o r n S t a t e Coll 'Mtncral P a ) Mr Sherborne mentioned connate water a s one
I n d Ex]) S t a B ~ a l l 12, G5-167 (1033)
'0 hI L Haider. "Prodnctl\,~tg Index." D ~ t l l l n ga n d Prodnctron of the probable causes f o r the devlatlon, and I thlnk
Practzcc, 181-90. American Petroleum I n s t ~ t u t e , New Tork thls 1s particularly llkely t o be true In vlew of the f a c t
119'ifii
\----,

R 1' Hlgglns "F'roductivitr of 011 Wells and,,Inherent I n - that, a s the permeability decreased, the deviation in-
fluence of Gas-011 R a t ~ o sa n d 'Water Saturation, U S B u r
Mzties Teclr Pzibl 5657 (1942) creased rather remarkably It seems to be t r u e i n Call-
* C; R Hocott a n d S E Buckley, " h l e a s y e m e n t of t h e VIS- fornla, and I belleve there is evidence from Pennsylvanls
cosltles of Oils under Reserrolr Condlt~ons, T r a n s Am I n s t
t h a t the connate-water saturation Increases \nth de-
Mtnrno Met E n o r s 142. 1:-:1-6 (1941)
la .
T- W Johnson a n d D B T n l i a f e ~ r o ,J r "Flow of Alr a n d
Natural Gas through Porous hledla, U S B ~ i r M l ~ i e s Tech creasing pern~eability
P v h l 591 11 ')RS\
,----,
Norrls Jo1i;ston (chairman) "Glossary Relatlng t o Reser-
M r Sherborne I a m incllned to agree t h a t t h e water
volr Behavior, published ~n t e n t a t w e form A P I Productwrr. h a s a profound effect upon t h e deviation Unfortu-
B ~ t l lA70 9%. $6.96. Nor 1941. a n d nresented in revised form a s
p a r t of the -annual report o f ' Callf6rnia ~ ~ s h c t - f i p ~ c a l - ~ o & nately, we do not have a n y way a t present of evaluating
mittee on Production Technology. Chicago, I l l , Nov 1943 ~ t effect
s However, we belleve t h a t t h e dev~atlonarlses
l5 Norrls Johnston "Core-Analys~s Interpretat~on." Drzlltng
a n d P r o d ~ t c t ~ oPra.ctlcc,
n 180-199 (1041) not only from the lnteractlon between water and other
fluids present, but also from the interactions between f o r the 141 wells would not be warranted Consequently,
water arid c e r t a ~ nconstituei~tsof the s o l ~ dmatter In t h e the tabulated values of k a r e "dr~lledsand thlcknesses."
forination We agree with Muskate ' (see bibliography, The effect of obliqueness of wells d r ~ l l e dthrough a
p 80) t h a t the former probably will not account f o r sand body on the productivity ~ n d e x1s not so simple a
the large divergences encountered However, we be- matter a s the effect on It The obl~quenessaffects three
lieve the latter to be amply sufficient to do so factors 111 the Darcy's law equation, 11, r., and +, The
effects on these factors become important in the equa-
S T Yuster (The Pennsvlvaina State College, State tion only when the angle ls large, the sand is very thick
College, P a ) I wonder if M r Sherborne has con- compared with the well spacing, the sand 1s massive
sidered the possibility of calculating the v~scosityof the (non-laminar), and has good vertical permeability com-
reservolr crude from a knowledge of the composit~on pared w ~ t ht h a t parallel t o the bedding planes
of the gas and ~ t condensed
s v~scosity,the viscos~tyof An attempt h a s been made a t a rough evaluation of
the crude, and t h e use of the ideal mixing laws the changes in the three var~ables,a s a f u n c t ~ o nof the
Mr Sherborne No D r Yuster's suggestion a s to angle A between the well bore and t h e normal to the
an'alternat~vemethod of est~rliatingthe v~scosityof the sand bedd~ngplanes, assulnlng a uniform isotropic sand
reservolr crude is very worthwhile When t ~ n i epermits, body of g r e a t thickness (commensurate with t h e well
such a n investigation would be very advantageous not s p a c ~ n g ) The variables must be considered in two
only a s a n a ~ dIn the present study, but also f o r com- direct~ons along the dlp, and along a contour Thls
parlson with experimentally determined v~scosities Un- rough evaluation shows t h a t the deviation between
fortunately, p n o r to presentation of this paper, time well potentla1 calculated by Darcy's radial-flow formula
was not available f o r elaborat~onof means for e s t ~ m a - and a formula modified for a n oblique well appears to
t1011 of reservoir v ~ s c o s i t ~ e s be less than 10 per cent for a11 angle .4 of 30 deg F o r
lower angles the dev~atlonapproaches zero rapidly, and
S E Buckley (Humble Oil and Refining Company, for greater angles lt Increases rapidly
Houston, Texas) I should like to take thls opportunity Although the formulation is admittedly only approxi-
to thank the Califorma Subcommittee on Electrical mate, it is thought t h a t the order of magnitude of the
Logging and Core Analysis, of w h ~ c hMr Sherborne r e s u l t ~ n gdeviations is correct A rigorous mathematical
has been cha~rnmanf o r several years, f o r making t h ~ s treatment a t some later date would undoubtedly be
paper ava~lableand to congratulate t h e committee on worthwh~lefrom the s t a n d p o ~ n tof well spacing alone,
~ t sexcellent work You can see t h a t i t required a and would likely have other useful appl~cations
tremendous ainount of effort to compile thls ~ n f o r ~ n a - In the process of approx~matingthe effectwe h, con-
t ~ o n a, s well a s to interpret ~t I hope t h a t this excellent s~deration was given to oblique flow Into the well
work may be taken u p and followed t h ~ o u g hin other Obl~queflow In petroleum reservoirs h a s appealed to us
areas of the United States a s a very important but usually neglected phenomenon
The average permeability of a lain~nated-sands e d ~ m e n t
E V Watts (General Petroleunl Corporation, Los depends greatly on the thickness of the streaks o r layers
Angeles, Calif (written)) I n regard to the net sand of d~fferentp e r i n e a b ~ l ~ tand
y , on the lnagnitude of these
thickness, 11, do the tabulated values represent t h e diffel.ei~ces Where the laminae a r e separated by im-
drilled sand thickness or thickness normal to t h e bed- pervious rock, and each lamina is t h ~ n there , 1s no ap-
ding planes? F o r structural d ~ p su p to say 30 deg, ~t prec~ableoblique flow effect Where la~ninaea r e ' t h ~ n ,
would seem t h a t the productivity index of a well might and vertical pernleabihty equals the hor~zontal, the
be almost the same regardless of whether the hole 1s effective average permeability may differ radically from
drilled vertlcally, or normal to the bedding planes If the numerical average The ratio of effective average
so, the normal sand thickness should properly be used per~neabilityto t h e numerical average cannot be h ~ g h e r
in the calculation of the specific productivity ~ n d e s than 2 nor lower than 1 It approaches 1 a s laminae
However, l t would be interesting to calculate the actual becoine thick and a s t h e changes in permeability be-
differences in PI t o be espected f o r varlous assumed tween adjacent layers become small I t approaches 2
reseivoir configuratlons and well spaclng a s laminae become thin, and the changes in permeability
Mr Johnston (written) Mr Watts h a s brought out between adjacent l a m ~ n a ebecome large
a n ~ m p o r t a n tmatter which was not discussed a t suffi- The obvious application of these observations IS in
cient length 111 the paper The effect on IL of slant holes obtaining a more accurate figure for effective average
o r of vertical holes through steeply d i p p ~ n gformations permeab~lityfor estimating well potentla1 This appli-
IS quite obv~ous D u n n g the accuniulat~onof the data cation is beset w ~ t hdifficulties because of lack of com-
given 111 Table 1, thls matter was neglected, and a f t e r plete information on permeability and degree of inter-
some discussion ~t was declded t h a t the probable effect communication of thin l a m ~ n a e I n general, ~t may be
of using "drilled sand thickness" 111 place of the more safely assumed t h a t n u m e r ~ c a averages
l a r e more nearly
correct "normal sand th~ckness" would be so small, In correct in massive sands, and less accurate In laminar
most cases, compared with errors In viscosity estimates, sed~ments This feature of core-analys~sinterpretation
t h a t the work of finding the well slant and dip angles deserves considelably more study