Вы находитесь на странице: 1из 7

Economics of Hydraulic Fracturing Using Wall-building Additives t

F. J. SHELL* AND 0. K. BODINE*


ABSTRACT
Many irnprovelllents have been made in the technique
of stimulating wells by hydraulic fracturing. One of .
these is an engineiring approach whereby t he areal
extent of t he fract ure can be calculated. With t hi s
knowledge t he cost of hydraulic horsepower can be bal-
anced against t he fracturing-fluid cost t o obtain t he.
most econonlical well treatment.
Frequently, t he most econonlical job will result from
using a fluid having a higher cost. Fract uri ng with a
hydraulic fluid having a low fracturing-fluid coefficient
permits t he use of less hydraulic fluid. This can, i n turn,
result in t he use of less hydraulic horsepower. Both
hydraulic fluid and hydraulic horsepower cost money
and must be balanced against each other f or t he great-
est economy.
Particular attention must be directed toward obtain-
i ng efficient fract uri ng fluids. Specific t est s must be
made to determine an economical t reat ment f or each oil
considered f or a fract uri ng fluid.
.
INTRODUCTION
The oil industry was given t he idea of well stimula-
tion by hydraulic fract uri ng some 12 years ago. I t was
Introduced in a paper by J. B. Clark.' A number of-steps
were made in t he improvement of t hi s original idea of
"hydraulic fracturing." These include t he use of fluid-
loss control a c l d ~t i v e s , ~. ~~- l ~~ the use of wat er as a frac-
t uri ng fluid,fi,7."nd specialized eclu~pment developed
f or t he purpose of fracturing."10.11~12~13 A number of
investigators have developed - methods f or determining
t he optimunl size f or a fracture.l-lJ5 Still others have
discussed fract ure orientation.16,17 No at t empt will be
made in this particular paper to discuss t he methods
used in arriving a t t he optimum. size of a fract ure nor
its orientation. Rather, t hi s paper will be primarily
devoted t o techniques of obtaining a particular size
of fracture.
,
Even with t he l arge nulllber of improvements noted
over t he first eight years of-fract wi ng, t he actual pro-
cedures were largely t ri al and error. In 1957, G. C.
Howard and C. R. Fast presented an engineering
appr oach t o f r a c t ur i ng t reat ment s. l B Thi s paper
Included an equation developed by R. D. Cart er which
presented the Idea of a fracturing-fluid coefficient. The
extent of a fract ure can be measured by t he volume of
fluid pumped down t he well bore less t he amount of fluid
t hat leaks into t he formation. Aft er making some very
general assumptions, Carter was able t o develop a
mathematical expression f or t he amount of fluid-which
would leak off into t he formation. This leak-off i s -a
function of a fracturing-fluid coefficient.
'Drilling Specialties Co.. Bartlesville, Okla.
?Presented by F. J. Shell at the sprlng meeting of the Mid-Continent
D~stri ct. Dl v ~ s ~ o n of Production. Wi ch~ta, Kans . March 30. 31.
April 1, 1960.
'References are at the end of the paper. ,
A number of these service companies have t aken
these engineering principles advanced by Howard and
Fast and simpl~fied them with graphs and nonlographs
for t he ease of calculating t he fract ure ext ent and sand
requirements t o pack t hat fracture.
THEORY
This presentation is based on an evaluation of t he
formula of Cart er f or estimating t he ext ent of t he
fractured area in t erms of t he t reat i ng conditions.18
Wherein :
x =OCV; ; - TIIY
A =ar ea of one face of t he fract ure, sq ft .
t =pumpi ng time, min.
Q =const ant injection rat e, cu f t per min.
W = constant fract ure width, f t .
C = fracturing-fluid coefficient.
The fracturing-fluid coefficient is a measure of t he
flow resistance of t he fluid leaking off into t he forma-
tion during t he fract uri ng treatment. This flow resist-
ance may be a function of t he hydraulic fluid or depend
upon t he characteristics of t he formation and fluids
comprising t he reservoir, or upon both factors. To
designate between these t hree mechanisms of leak-off
control, subscripts have been used. In an actual frac-
t uri ng t reat ment all t hree controlling factors may be
present but one i s norn~al l y predominant.
The fracturing-fluid coefficient depending upon t he
characteristics of both t he reservoir and t he hydraulic
f l u~d IS designated a s CI and expressed as:
C, = 10.0469 (dkAP+Ipf)l ( 2)
146 F. J. SHELL AND 0. K. BODINE
Thus C, can be determined once t he absolute viscosity
of t he fract uri ng fluid, p,, the formation permeability,
k, t he forni at ~on porosity, ,$, and t he different pressure
across t he f ~ a c t u r e face, AP, ar e known.
If in t he foregoing case the viscosity of t he fract ur-
i ng fluid is low, t he forliiation characteristics alone
may be controlling. In this case t he viscosity of t he
reservoir fluid, pr, and i t s con~pressibility coefficient,
C,,, can be expressed:
CI I fO.OS74 AP ( d K4 Cr l p r ) I (3)
A t hi r d f or m of t he f r act ur i ng- f l ui d coefficient
depends only upon t he characteristics of t he fluid. This
must be a specla] fluid with wall-building characteris-
tics, however. For these fluids the loss through a porous
medium is determined experimentally. When t he fluid
loss is plotted agai nst t he square root of time, t he
result is a st rai ght line of t he form:
V = ,,n + VSP (4)
Tt 7 l l e ~e i ~: V is t he fluid loss, t is time, ,tiz is the slope
of the Ilne, and llgp is t he y-intercept or "spurt loss."
From this equation i t is possible t o develop an expres-
sion for C,,,, the fracturing-fluid coefficient for a wall-
building fluid :
C,,, = 0.0378 (nt /.?a) (5)
T Vl j e ~ e i ~ ~ : (1, is the cross-sectional area of t he filter
medium. The area of t he fract ure can be corrected for
spul t loss, VSp, by including this value a s an increased
w~dt h where:
1.I" = TV + (V, qp/ 15. , 04(~) (6)
THEORETICAL PROBLEM
To help understand t he importance of fracturing-
fluid coefficlent, C, a problem has been chosen consisting
of three part s, viz.: Pa r t A where C equals 10 x 10-3; .
Par t B where C equals 5 x lo-" and Par t C where C
ecjuals 1 x 10-3. For thls problem i t was assumed t hat
a fl.acturing-fluid coefficlent of 10 x 10-3 could be
obtained f or a cost of $0.01 per gallon. This $0.01 per
gallon would represent t he cost of storing and handling
t he fract uri ng fluid. In all three cases a fract ure of
cons t ant si ze, packed t o t h e sanl e extent, will be
assumed. Therefore, t he amount of sand necessary will
be a constant and has been neglected. I t was assumed
f or this problem t hat t he fract uri ng fluid could be
treated f or $0.02 per gallon so as to improve i t suffi-
ciently to give a fracturing-fluid coefficient of 5 x 10-3,
and t hat an acldit~onal $0.02 per gallon cheln~cal t reat -
ment of t he hydraulic fluid would reduce t he fracturing-
fluid coefficient to 1 x This would make t he total
cost of t he fluid in case B (C = 5 X 10-3) $0.03 per
gallon and the hydraulic fluid in case C (C = I x l os 3)
cost a total of $0.05 per gallon. The fract ures were
assu~ned to be initiated and extended with 3,000 psi
surface pressure a t a cost of $1.00 per hydraulic horse-
power. The fract ure width was assumed t o be 1/ 10 in.
throughout t he fracture. Fig. 1 is a plot of area as a
function of t he size of t he job a t various C values.
These plots were made a t three different injection rates.
120
I I '
-
o I NJ RATEZ3O BBLIMII4
o 2 3 4 5 6 7 8 9 1 0 20 30 4 0 50
I I 1 1 1 1 1 , I . 1 1
AREA X I O - ~ SQ FT
Fig. 1 -Area Vs. Treatment Size for Various
C Values and Injection' Rate
Inspection of this graph shows innnediately t hat , f or
areas great er t han 100,000 sq f t , C needs to be less t han
5 x lo-s to be in an economical range.
Let us consider now t he case f or a fract ure having
an area of 50,000 sq f t . Assunling t he fract ure is con-
centric around t he well bore in a horizontal plane, a
50,000 sq f t fract ure area will extend 130 f t (Fi g. 2) .
An injection rat e of 10 bbl per nlin a t t he surface
pressure of 3,000 psi would require 735 hhp. Hydraulic
horsepower is computed by t he equation:
HHP=0.0.245 P, Q (7)
I Vhe~ei n: Ps is surface pressure, in psi; and Q 1s
injection rate, in barrels per minute.
With a fracturing-fluid coefficient of 10 x 10-3 i t
would be virtually impossible a t an injection rat e of
10 bbl per min to obtain a fract ure area of 50,000 sq ft .
A 100,000-gal t reat ment under these conditions would
result in a fract ure of only 27,000 sq ft . If C is 5 x lo-=,
t he 50,000 sq f t area can be obtained with a t reat ment
of 88,000 gal of hydraulic fluid. This hydraulic fluid, a t
a cost of $0.03 per gallon, would amount t o $2,640.
This, added t o t he cost of hydraulic horsepower of
$735, equals $3,375 (Table 1). If t he fracturing-fluid
coefficient were only 1 x a n 8,000-gal t reat ment
100- -, -;., , , ;; ,
80 -
40 -
20
-
RADI US
-
1 0 -
x r o - ! ~ ~
AREA x 10- 5 sa FT
Fig. 2 -Radius Vs. Area of Circle
Table 1
Fracturing-treatment Cost Comparisons for Fluids of Various
Fracturing-fluid Coefficients Pumped at Different Injection Rates
Injection Rate,
Bbl per Min C*
Treatment, Hydraulic Horsepower,? ~r a c t u r i n g Fluid,$
Gallons Dollars Dollars
Total.
Dollars
For a 50,000 Sq Ft Fractures
Impossible
88,000
8,000
I m~ossi bl e
Impossible
Impossible
15.000
30 1 X 10-3
-
*Assume $0 01 for fluid wrth C = 10
$0. 03 C = 5
$0 05 C = 1
f 3 000 psi surface pressure and $1. 00
3Fractorlng-flu~d cost.
SAverage w~ d' h of 0.1 in
X lo-'
X l o- $
X lo-'
I per hydraulic horsepower.
would result in a 50,000 sq f t fract ure area. At a
price of $0.05 per gallon, this hydraulic fract uri ng
fluid would cost $400. Added t o t he cost of the hydraulic
horsepower of $735, t he fract uri ng t reat ment would
cost $1,135. This i s about one-third of t he cost of
t he t reat ment with a fluid having 5 x coefficient;
The same line of reasoning was used for injection
rat es of 20 and 30 bbl per min. The injection r at e must
be 30 bbl per lnin before i t is possible t o create a
fract ure with an area of 50,000 sq f t using a fluid
having a C of 10 x lo-? This job would be slightly less
expensive t han a job consisting of an injection rat e of
10 bbl per min and a fluid having a coefficient of
5 x lo-:'. I t would still be almost t hree times a s
expensive a s the efficient fluid a t an injection rat e of
10 bbl per min. As t he injection rat e is increased t he
cost of hydraulic horsepower becomes a major portion
of the total fract uri ng cost; whereas, a t low injection
rates, t he cost of t he fluid itself may assume a major
portion of t he cost of t he fract uri ng job.
These calculations were repeated f or a job having a
fract ure area of 100,000 sq f t , or a radlus of 180 ft .
This area could not be obtained with a fract uri ng fluid
having a coefficient of 10 x 10-3 a t injection rat es up
to 30 bbl per min. If t he fluid had a coefficient of
5 x 10-3, t he fract ure could not be obtained a t an
injection rat e less t han 30 bbl per win. At an injection
rat e of 10 bbl per nlin and a C value of 1 x 10-3, t he
total cost of the fract uri ng fluid plus t he hydraulic
horsepower would be less t han in any case for 50,000
sq f t fract ure area except t he case of an injection rat e
of 10 bbl per min and a C value of 1 x 10-3. If a
fract ure of 100,000 sq f t can be created for less cost
t han one with an area of 50,000 sq f t , a great er ret urn
can be expected on t he dollar.
' EXPERIMENTAL RESULTS AND DISCUSSION
The question then arises, "Can a fract uri ng fluid be
treated to obtain t he assumed C values for the assumed
cost?" Using commercially available materials, i t is
not possible a t t he present time to get C values 1 x 10-3
using water. I t is possible t o get coefficients of 1 x 10-3
and even lower using treated oil. I t should be remem-
bered, however, t hat from a physical standpoint wat er
fract uri ng t r eat ment s follow t he same rules as those
using 011-only the economics will be altered.
I t has been only during t he past year t o a year and a
half t hat chemical additives have been available t hat
will reduce t he fracturing-fluid coefficients of oils t o
values as low a s 1 x 10-3. Of course, t here may have
been selected oils t hat earlier additives could have made
this efficient. Present additives ar e capable of producing
a hydraulic fluid of t hi s efficiency from almost any oil,
providing sufficient concentration of t he proper additive
is used. Not all crudes respond t he same to different
additives. A group of selected t est s ar e presented in
Table 2 to demonstrate t he varying degrees of effective-
ness of different additives in oil or t he variations
possible with t he same additive in different oils. These
t est s also demonstrate t hat i t is possible to get C
values of 5 X 10" and 1 X 10-%t costs of $0.03 and
$0.05 per gallon, respectively.
Oil
1
1
Table 2
Effective Fracturing-fluid Coefficient for Various Chemical Additives in ~i f f er ent Oils
Cost per Gallon,
Additive Dollars cIII x l o-3
B $0.03 3.06
B 0.0425 3.46
No at t empt was made to include all possible fluid-
loss-control additives in this table and only a small
sampling of various oils i s used. The particular addi-
tives presented were selected because of t he similarity
in t he results they gi ve' i n various oils. They ar e also
five of t he most effectwe additives available.
For t he dat a in Table 2, Cl l I was obtained by adding
t he fluid-loss-control chemicals t o oil and st i rri ng for
3 ~ n i n with a Wari ng Blendor. The oil was placed in a
Baroid No. 387 high-temperature filter press with t hree
filter papers. The filter press was placed in an inverted
posi t ~on in t he heating cell f or 15 min while t he oil
canle t o a constant t emperat ure of 125 F. The press
SLOPE
Fig. 3 - Fraction-fluid Coefficient Vs.
Slope of Water-loss Curve '
was shaken vigorously and inverted. The oil was filtered
with a differential pressure of 1,000 psi. Fluid-loss
values were read a t 1, 4, 9, 16, 25, and 36 min. These
values were plotted a s functions of t he square root of
time. The slopes of t he st rai ght lines were used t o get
CI I 1 values (Fig. 3) . The straight-llne portion of t he
curve was extrapolated to zero t ~n l e t o give t he
spur t loss.
A number of graphs with curves of fract ure area a s
a function of fracturing-fluid coefficient were prepared.
Each of these graphs contained curves for various job
sizes from 5,000 to 100,000 gal. Graphs were made f or
various injection rates, Q, and f or various spurt-loss
values a t those injection rates. The fract ure width a t
zero spurt of 0.1 in. (0.00833 f t ) was used. Only two
of these graphs ar e included herein as esamples (Fl g.
4 and 5).
The effect of spurt on . t he fract ure size i s actually
a function of t he job size, t he injection rate, and CI I 1.
Other factors being constant, t he effect of spurt
decreases with an increase in t he size of t he job. Con-
versely, an increase in t he injection rat e increases t he
effect of spurt. This i s because an increase in t he
injection rat e decreases t he tiine to pump a given
volume, hence decreases t he tinle a t which filtration is
a controlling factor. This increases t he relative impor-
tance of spurt. In other words, a relatively great er
portion of t he fluid is lost while laying down the initial
cake rat her than in subsequent filtration through t he
cake. For t he same injection rate, however, an increase
in t he size of t he job increases t he time for t he job
which increases t he relative importance of t he filtration
SPURT = 0 ML
I
I
I
FRACTURE AREA x SQ FT
Fig. 4 -Area Vs. C
10
8
6
--100,000 GALLONS
4
2
103 1
0 6
0 4
0 ?
SPURT 4 ML
0 1 5 6 7 8 9 1 0 2 3 4 5 6 7 8 9 1 0 0 2 3 4 5 6 0 0
FIi ASTUI:E AREA A 1 0 - 3 , SQ FT
Fig. 5 -Area \Is. C
as t he controlling factor. To demonstrate t he relation-
ship between CIll and spurt , Fig. 6 has been prepared
showing area as a function of Cllr a t different spurt s
f or a constant i nj ect ~on rat e of 10 bbl per min and a
constant job size of 25,000 gal. Inspection of this graph
shows t hat t he relative importance of spurt increases
rapidly as Cl l l decreases.
The C values used in t he theoretical problem were
not 'complicated by spurt loss. To get the dat a from
filtration t est s on a comparable basis an effective
fracturing-fluid coefficient, Ceff, was chosen. This was
equivalent to a C w ~ t h zero spurt . Realizing t he compli-
cation of correcting C f or spurt , Ccff was defined by
limiting t he cond~t i ons to a job of 25,000 gal a t a n
injection rat e of 10 bbl per min. The method of obtain-
ing Ceff can best be shown by an example. Take a n 011
with a CIIl value of 2;8 X 10" and a spurt loss of
4 ml. Using Fig. 5 (Q =10 and VSP =4) , find a C of
cmx
AREA X I O- ?SQ FT
Fig. 6- Effect of Spurt on Area
2.8 X 10-3, proceed to a job of 25,000 gal, and thence
to a fract ure area of 39,500 sq ft . Go to Fig. 4 (Q = 10
and ITSP= 0) a t an area of 39,500 sq ft , follow this
line t o a job of 25,000 gal, and thence to t he fractul-ing-
fluid coefficient axi s where the Celf is found to be
3.25 X 10-3.
For oil No. 1, additive B gave a Ceff of t he ordei. of
5 X a t a cost of $0.03 per gallon. The costs in
Table 2 represent t he cost of the additive plus $0.01
per gallon handling charge. When t he concentration of
additive B was increased until t he cost of t he oil was
$0.075 per gallon, Crff was still considerably great er
t han 1 X However, in t he case of additive A an
effective coefficient of 3.2 X 10-3 was possible a t a cost
of $0.03 per gallon and a CCrf of less t han 1 X
was possible a t a cost of $0.05 per gallon. I n oil No. 2,
additive B was f a r more effective. Effective fracturing-
fluid coefficients of t he order of 1 X 10-3 were possible
a t costs of both $0.03 and $0.0425 per gallon. I n fact ,
in this oil additive B was more effective t han either
additive D or E. Additive D decreased in effectiveness
with an increase in concentration. This has been noted
in a number of oils with various additives. To guard
agai nst this, additives t o be used should be pilot-checked
in t he particular oil t o be used in t he field. This i s less
important in a case of chenlical control of water, but
should not be neglected because wat ers from various
areas do change in composition. All five additives were
checked in oil No. 4. Because additive B did not control
t he fluid loss of oil No. 4, these dat a ar e not included.
This particular oil responded quite well to t he other
four additives. These dat a indicate t hat various oils
respond differently to various fluid-loss additives, but
t hat a particular additive can be found f or almost any
Table 3
Cost of Creating 50,000* Sq Ft Fracture
Co'st and Oil Volume a t Q =
h
I ,
10 Bbl per Min 30 Bbl per Min
A
r > ,
h
,
Hydraulic Hydraulic
Cost per Oil Oil Horsepower Total Oil Oil Horsepower Total
Gallon, Volume, Cost, Cost, Cost, Volume, ' Cost, Cost, Cost,
Dollars Gal Dollars Dollars Dollars Gal Dollars Dollars Dollars
Part A:Addi t i ve A in Oil If
Part B: Additive B in Oil 2f
-
*Assume fracture width of 0 1 In. and surface pressure of 3,000 psi.
+From Table 2.
oil t hat will reduce the fracturing-fluid coefficient t o a
value approaching 1 X 10-3. The'se dat a also indicate
t he importance of checking t he various additives in
question in t he oil to be used in t he field.
.
The importance of running t he fluid-loss t est to
determine t he economics of a job i s emphasized from
t he dat a in Table 3. Consider t he case of additive A
in oil No. 1. If we were t o use this fract uri ng fluid to
create a 50,000 sq f t fract ure a t a surface pressure of
3,000 psi and an injection rat e of 10 bbl per min, the
oil volume would be reduced from 31,000 gal t o 8,700
gal by doubling t he additive concentration. This would
reduce t he total cost of $1,665 to $1,170 even though
t he unit cost of t he hydraulic fluid increased from
$0.03 to $0.05 per gallon. To do this job a t 30 bbl per
min would require less oil but t he hydraulic horsepower
increases necessary would almost double t he cost of t he
fract uri ng treatment. If this same job were done with
additive B in oil No. 2, t he same type of savings could
be realized going from an injection rat e of 10 bbl t o
30 bbl per min. However, in this case an increase in
t he concentration of t he additive cannot be justified.
Even though slightly less oil is used with t he l arger
concentration, i t does not offset t he increase in t he unit
cost. This again points out t he necessity of pilot-testing
t he individual additives in t he oil under consideration
f or any fract uri ng treatment.
CONCLUSIONS
Hydraulic fract uri ng a s a method for well stimulation
is a progressive science. One of t he l at est advances has
been the introduction of an engineering approach t o
obtaining opt i n~un~ fract ure sizes. ~ h e s e -principles.
although based upon some very general a ~~u mp t i o n s ,
give t he operator a tool whereby he can design more
economical fract uri ng treatments.
Frequently, t he most economical job will result from
using a fluid having a higher cost. Fract uri ng with a
hydraulic fluid having a low fracturing-fluid coefficient
permits t he use of less hydraulic fluid. This,can, in turn.
result in t he use of less hydraulic horsepower. Both
hydraulic fluid and hydraulic horsepower cost money
and must be balanced agai nst each other for t he
great est economy. One point not covered by this paper
is t hat less production will be lost if less oil i s used
f or t he same fract ure size. This becomes increasingly
niore important for wells of low productivity.
ACKNOWLEDGMENT
The aut hors wish to express their appreciation t o
Drilling Specialties Company for permission to prepare
and present this paper, and to J. R. Bruns of Phillips
Petroleuni Company for his assistance in t he solution of
t he equation of R. D. Carter.
REFERENCES
' Clark, J. B: A Hydraulic Process for Increasing t he
Productivity of Wells, Trans. Awl. Inst. Mi~zilzg Met .
Engrs. (Petrolelint Developnze?~t nwl Tecl ~nol ogy), 186,
1 (1950).
ZReynolds, Jack J ; Bocquet, P. E; and Coffer, H. F:
Conversion of Crude Oils t o Low-fluid-loss Fract uri ng
Fluids, Drillilly am.? Prodt~ction P~ci ct i ce, 111 (1955).
3Davis, J. G; Reynolds, Jack J ; and Coffer, H. F:
Control That Fluid Loss, 011 Gas J.. 51 [61 I 102,
Jul y 2 (1956).
4Brown, J. L. and Landers, M. M: U. S. Pat ent No.
2,779,735, Jan. 29, 1957.
SCoulter, A. W: Chemical Additives Improve Frac-
turing, World Oil, 144 [21 148, Feb. 1 (1957).
=Gibbon, A: Fresh Wat er I s Becoming Favorite Frac-
t uri ng Fluld, ,World Oil, 145 [ l l 76, Jul y (1957).
:Harmon, J. A.: The Chemistry of Fresh-water Frac-
turing, Drilli~zg and Production Practice, 50 (1957).
BWalsh, E. N: Wat er Frac in t he San Juan Basin,
Paper No. 906-2-A, presented a t t he spri ng meeting
of t he Southwestern District, API Division of Produc-
tion, Dallas, Texas, March 1957; Oil Gas J. (title: How
They' re Reducing San J uan Basin Fract uri ng Costs)
55 1211 146, May 27 (1957).
gClark, J. B; Fast , C. R; and Howard, 'G. C: A
Multiple-fracturing Process for Increasing t he Produc-
tivity of Wells, Drilling and Production Practice, 104
(1952).
"'Paul. J. R. and Muir, P. D: Temporary Plugging
Agents Improve Results from Fract uri ng Treatment,
Pclrolcztrt~ E?t gmeer, 26 [61 530, June (1954).
' ]Fract ure Aid Unveiled, Oil Gas J., 54 [711 86,
Sept. 10 (1956).
"Rubber Tanks Cut Fract uri ng Costs, Oil Gas J.,
54 1721 128, Sept. 17 (1956).
l"astrop, J.' E: Formation Fract ures in One Con-
trolled Plane, Pet r ol et ~m E?zyi7zeer, 29 [91 B-21, Aug.
(1957).
"Crittendon, B. C : The Mechanics of Design and
Interpretation of Hydraulic Fract uri ng Treatment,
J. Pet r. Tech. , 11 ClOl 21, Oct. (1959).
ISHoward, G. C; Flickinger, D. H; and Fast , C. R:
Hydraulic Fracturing, Oil Gns J.. 56 [211 81, May 26
(1958).
"Xrawford, P. B. and Landrum, B. L: Do Horizontal
Fractures Affect Production Capaci t y?, Worlcl Oi l , 142
[61 196, May (1956).
'TClark, R. C. and Reynolds, J. J: Vertical Hydraulic
Fracturing, Paper No. 851-28-5, presented a t t he spri ng
meeting of t he Mid-Continent District, API Division of
Production, Oklahoma City, Okla., March 1954.
'RHoward, ,G. C. and Fast , C. R: Optimum Fluid
Characteristics f or Fract ure Extension, Drillilzg ctnd
Production Prcictice, 261 (1957). .
DISCUSSION
H. C. Walther (Continental Oil Co., Ponca City,
Okla.) (wri t t en): This paper shows how significant
savings can be obtained by t he application of engi-
neering principles to fract ure design. In particular,
t he authors use a number of specific examples to
demonstrate t he importance of fluid-loss control.
Their conclusions can be verified from examination of
t he Carter equation. To determine t he importance of
C , the Cart er equation can be written as:
I. Vl j e, ~c. , ~: K , and K2 ar e constants; and F( x ) is a
' funct ~on of x. For l arge x, F( s ) is proportional to x:
and, hence, A is proportional to (VQ)'/z/C. As r
decreases, F(x) becomes proportlonal to x to a hlgher
power.
Generally, for large t reat ment s where treatment
des ~gn is most cr~t l cal , x IS large. For example, t he
values of x in Table 1 range from 0.5 to over 30, wlth
t he minimum t reat ment cost occurring in both cases
a t x great er t han 1.0. When x is l arger t han 10, A is
inversely proportional to C, and a t z=1.0, A is
~nversel y proportional to (C)M or:
Thus, for l arge treatments, t he area, A, will usually
be proportlonal to C t o t he -0.5 t o -1.0 power; t o'
treatment volume, 17, to t he 0.75 to 0.5 power; and t he
rate, Q, to the 0.25 to 0.5 power. For t he case where
r i s above 10, increasing t he fluid-loss coefficient by a
fact or of 10 would require a 100~fol d increase of either
V or Q if t he area is to remain constant.
For Table 1 i t was assumed t hat t he hydraulic horse-
power required was directly proportional t o t he pump
rate. This is not t rue if t he frictional losses ar e
significant, as in t he case of pumping down t he tubing.
Then, doubling t he pump rat e will more t han double
t he required horsepower. The resulting increase in
pumping costs makes t he use of additives even more
hi port ant .
Untreated crudes usually have fluid-loss coefficients
great er than 10 X 10-3. Recent improvements in fluid-
loss additives make i t possible to obtain fluid-loss
coefficients of about 1.0 X 1 0 - h t a t reat i ng cost of
only 2.5 cents per gallon, which, when added t o t he
fluid cost of 1 cent per gallon, results in a total fluid
costing only about 3.5 cents per gallon in contrast to
t he 5 cents per gallon used in Table 1.
In conclusion, we fully concur with t he aut hbrs on
t he importance of fluid-loss additives. Based on t he
assumptions of t he Howard and Fast equation, i t can
readily be shown t hat fluid-loss additives can usually
reduce fract uri ng costs.

Вам также может понравиться