Você está na página 1de 6

Rheological Design of Cementing Operations

HALLIBURTON CO.
KNOX A. SLAGLE
DUNCAN, OKLA.

ABSTRACT is not possible to define their rheological or fluid proper-


ties by the single factor of viscosity and then make cal-
Hydraulic analysis of the wellbore has become increas- culations for the quantities just described. Because the
ingly important for designing cementing operations and shear stress-shear rate ratio is not constant, it becomes
selecting equipment, materials and techniques to comple- necessary to establish at least two parameters for adequate
ment modern well-completion practices. Non-Newtonian fluid-flow calculations. It is not the purpose of this paper
fiuid technology has advanced beyond the point where to delve into the mathematical development of non-New-
former empirical methods of analysis adequately define the tonian technology, nor to discuss the arbitrary classifica-
hydraulic system and fluid properties. tion system under which a single fluid may resemble two
In view of these factors, this paper describes a series or three different classes depending upon experimental
of rheological calculations which have been found prac- conditions. Rather, the intention is to present a useful
tical, through field usage, for assistance in selecting a series of calculations based on a concept applicable to
cementing program. A relatively simple laboratory method both Newtonian fluids and to the preponderance of non-
using standard viscometric equipment is suggested for de- Newtonian fluids encountered in the oil-producing industry.
termination of the rheological properties of slurries, and Development of this approach was begun some 32 years
data are presented on some of the more common cement- ago; and has most recently been brought to fruition by
ing compositions. A criterion for divergence from laminar- Metzner'-' and his co-workers at the U. of Deleware.
flow characteristics has been proposed. Usefulness of the Some non-Newtonian fluids encountered in the petroleum
calculations is indicated by examples of cementing opera- industry, other than cementing slurries, have also had
tions where they have been used. the benefit of this method of analysis."" The two para-
meters required to define the fluid are usually denoted
INTRODUCTION by the symbols n' and K' and, for the purposes of this
discussion, are called "flow behavior index" and "consis-
With the changing aspects of well-completion practices tency index", respectively. These two slurry properties per-
during the past few years, it has been increasingly im- mit calculation of the Reynolds' number and the "critical"
portant to have a relatively simple method of analyzing velocity, or the velocity at which departure from laminar
the flow conditions existing in the well during cementing flow begins.
operations. This is particularly true in view of the im-
proved economics toward which most of the changes have
been directed. Rheological characteristics of slurries used EXPERIMENTAL DETERMINATIONS
for cementing should be a major consideration in the trend The two principal instruments used for rheological
toward smaller casing sizes, either single or multiple strings. studies are the pipeline (capillary-tube) viscometer and the
Receiving increased attention is the practice advocated rotational viscometer. When conveniently possible, a capil-
in 1948 by Howard and Clark' of attaining turbulent flow lary-tube viscometer (where the pressure drop and flow
with the fluids circulated during a primary cementing rate of the material can be measured) is the better method
operation. While there may still be a difference of opinion for rigorous determination of the flow behavior index
concerning this technique, most available information and consistency index for non-Newtonian fluids. With
indicates that superior primary-cementing results are gen- pressure-drop data at various flow rates, it is then possible
erally obtained when high displacement rates are em- to prepare a logarithmic plot of shear rate as the abscissa-
ployed. Fluid properties of the slurry to be used must shear stress as the ordinate. For fluids which do not exhibit
be available, as well as calculation methods, to determine time-dependency, these data will usually produce a straight
what flow rates should be attained and the probable con- line. The flow behavior index n' represents the slope of
sequences in terms of frictional pressure and horsepower this line, while the consistency index K' becomes the
utilization. It would certainly be inappropriate to attempt intercept of this line at unity shear rate in accordance with
high displacement velocities if sufficient pressure might be the mathematical derivations associated with this concept
developed to create lost circulation. of rheology.
Since cementing slurries are non-Newtonian fluids, it Due to the difficulties anticipated in maintaining a uni-
form, pumpable cement slurry for the time interval re-
Original manuscript received in Society of Petroleum Engineers office quired to obtain measurements from the pipe viscometer,
Sept. 9, 1961. Revised manuscript received Jan. 9, 1962. Paper presented
at 36th Annual Fall Meeting of SPE, Oct. 8-11, 1961, in Dallas.
the n' and K' data reported herein were obtained using
'References given at end of paper. a direct-indicating rotational viscometer (Fig. 2). The

MARCH, 1962 SPE 152 323


instrument was of the type described by Savins and Roper," TABLE I-SLURRY PROPERTIES OF API CLASS A CEMENT
but was equipped to operate at six rotational speeds (600, P
Additive n' K' (lb/gai)
300, 200, 100, 6 and 3 rpm). With this instrument, the None' 0.30 0.195 15.6
shear stress is obtained by dividing the dial reading by Calcium Ligno5Ulfonate* 0.33 0.104 15.6
Bentonite-4o;.* 0.10 0.95 14.1
100; the shear rate is a function of the speed of rotation Bentonite-8o;.' 0.10 0.90 13.1
Bentonite-12o;.' 0.10 0.76 12.6
and the dimensions of the rotor and bob. For the standard Bentonite--12 % plus Calcium lignosulfonate--O.6 % * 0.14 0.124 ~2.6
instrument, shear rate equals 1.703 X rpm. As with the Saturated Salt Water 0.20 0.180 16.2
Diatomaceous Earth-20o;.' 0.20 0.312 12.4
pipe viscometer, a logarithmic plot of shear rate vs shear Latex Cement 0.44 0.039 14.5
Pozzolan X Cement-O% Bentonite* 0.30 0.166 15.2
stress yields essentially a straight line where the slope and Pozzolan X Cement-2% Bentonite* 0.10 0.81 14.1
intercept at unity shear rate again are the desired slurry Pozzolan X Cement plus Calcium lignosulfonafe 0.23 0.155 14.1
Pozzolan X-lime 0.35 0.120 14.0
properties. Pozzolan X-lime plus Calcium ligno5ulfonate 0.43 0.057 14.0
Complexity of the chemical reactions occurring when *Average data for 17 sources of API Class A cement.

cement and water are initially mixed makes it extremely API Class C-Sulfate Resistant Cement
None 0.43 0.03 14.1
important that slurry preparation methods be standardized Benton ite--4 % 0.10 0.30 13.0
in order that reasonably reproducible data be obtained Bentonite--8 % 0.10 0.30 12.5

with the rotational viscometer. The following steps have


been found most satisfactory.
merical pipe, while both types of fluids are accommodated
1. Mix the slurry initially with a high-speed mixer, as in the laminar region by f = 16/NRe • The extensive data
described in API RP lOB, "Recommended Practice for represented by the non-Newtonian curve were obtained
Testing Oil-Well Cements and Cement Additives". using 1,000 ft of 2-in. tubing which was approximately
2. Transfer slurry to atmospheric-pressure thickening- one-half test section and the remainder calming sections.
time Tester A of API RP lOB for 5 to 15 minutes of slow- Three triplex plunger pumps commonly used in the oil
speed mixing. field were employed to provide power for measurements in
3. Pour the slurry into the sample cup of the viscometer, the turbulent region approaching a Reynolds' number of
raise the cup into the operating position and start the 200,000. It should be noted here that these data were
instrument at 600 rpm. obtained with bentonite-water suspensions and with emul-
sions, both of which are discreet-particle systems as dif-
The slow-speed mixing period is required to minimize ferentiated from gelatinous fluids, for which an entirely
the initial gelling characteristics of freshly mixed slurries different correlation is necessary. The not unreasonable
which complicate viscometric measurements. assumption must be made that cement and the predominant
Each step in handling the slurry should involve a mini- number of additives will produce suspensions of indepen-
mum time delay since thixotropic or gelling tendencies dent particles which are adequately described by this cor-
of a much lower magnitude still exist and could interfere relation. Only one cement slurry has been tested in this
with measurements. Another part of the procedure which manner and through a rather limited range of Reynolds'
also tends to depress thixotropic interference is the practice numbers. Excellent agreement was obtained both with the
of taking the initial viscometer reading after 60 seconds friction-factor correlation and with the comparison of
at the highest speed (600 rpm), then decreasing speed viscometric and pipeline slurry properties. Metzner and
(300, 200 and 100 rpm) and taking resultant dial readings Reed' also reported essentially the same friction factor-
at 20-second intervals. These data are used to prepare the Reynolds' number plot in their resume of published data
flow curve on logarithmic co-ordinates as described earlier on various non-Newtonian fluids.
to obtain n' and K'.
The readings at 6 and 3 rpm are not normally used in VARIATIONS IN CEMENTS
this analysis. It is believed that erratic readings are obtained
at these rotational speeds because of the extreme shear-rate One of the major sources of deviation which can be
gap between 100 and 6 rpm. This was apparently sub- introduced into any system of rheology dealing with ce-
stantiated by brief experimental data obtained on a rota- menting slurries is that of the variable properties of
tional viscometer capable of infinitely variable speed. It cement. This study involved the measurement of n'
was found that all points below 600 rpm, including the and K' for the API Class A cement from 17 sources (each
readings at very low shear rates, produced a basically cement being mixed in 20 different slurry compositions, or
straight line on a logarthmic shear rate-shear stress plot.
The data presented in Table 1 were obtained using this
procedure at room temperature. This table includes some
of the average n' and K' values discussed later for 17
sources of cement, as well as the values for several slurries
,"' I"'
,
I
I"
i
;
prepared with a cement of median viscous properties.
The most recent addition to cement slurry rheology liter-
~
I
I . . 2:N I ,
ature has been presented by Ish-Shalom and Greenberg" '"~ .0 I
I
N£WT01W):;;-
This work, done at the Portland Cement Assn. lab, utilized l! I
z j----. -i--
a different rotational viscometer. When comparable slur- o ---i{
;:: +LAMINAR FLOW
r-- IT,-t+'
f----
ries were tested, re-evaluation of their viscometric data <.)
ii: I-- -hREGI?NI
I NON-NEWTONIAN
u.. I
in terms of n' and K' shows very similar slurry properties II
'"Z I
~tl
Z I
to those of Table 1 within the limits imposed when test- z
I

I
ing cements from different manufacturing plants. l! !!
iii
I I

The Fanning friction factor-Reynolds' number correla- .00 I i i i'i Iii I


I
I
I
10' 10' 10' 10'
tion proposed for cementing slurries in the turbulent region
REYNOLDS NUMBER - N
is that marked "non-Newtonian" in Fig. 1. The Newtonian R,

curve is the standard correlation for such fluids in com- FIG. I-REYNOLDS' NUMBER-FRICTION FACTOR CORRELATION.

324 JOURNAL OF PETROLEUM TECHNOLOGY


a total of 340 fluids), always tested in duplicate and often The friction pressure-drop formula
- in triplicate. Geographical distribution of the cements in- 0.039 LpV'f
cluded four from the Southwest, two from the Midwest, !:::..P! = --D-=-'-- (4)
six from the Gulf Coast, three from California and two
from the Rocky Mountain area. Generalization of the is the usual equation associated with this calculatIon, with
cement characteristics indicated that Gulf Coast cements a constant being inserted to correct for the dimensional
were more viscous than average, cements from the South- units used.
west and Midwest had median viscous properties, while One extremely interesting finding during establishment
California and Rocky Mountain cements were less viscous. of the friction factor-Reynolds' number correlation was
It was found that no apparent correlation existed between the rather pronounced divergence from laminar flow which
viscous properties and surface area or fineness of the occurred at N Re = 2,100. Although this does not represent
cements tested. an abrupt change to turbulent flow and in reality the tran-
Further generalization showed the more-viscous neat sition region probably is quite broad, these disperse fluids
cements to have minimum n' values and maximum K' did not yield the fluctuations often obtained in the transi-
values, while the less-viscous cements tend toward maxi- tion region with Newtonian fluids. In view of this, a
mum n' values and minimum K' values. These factors also minimum criterion for turbulent flow was established at
influence the calculations described later, with the more- this Reynolds' number and rearrangement of Eq. 1 yields
viscous cements yielding higher "critical" velocities and a minimum "critical" velocity.
frictional pressure drop. V,"" = 2,100 K' (96/D)n. 1,129 K' (96/D)n.
Actual ranges in rheological properties for different ce- 1.86 p p
ments mixed with the API recommended water ratio (46 (5)
per cent) were n' from 0.19 to 0.42 and K' from 0.638 to Also of use is the expression for hydraulic horsepower.
0.049, as compared to the average values of 0.30 and HHP = 0.0245 PwQ (6)
0.195, respectively. The average values reported in Table
1 were obtained from a shear stress-shear rate plot after P w = !:::..P, + Ph. - Ph' • (7)
averaging the basic viscometric measurements. The ex- Calculation of hydrostatic pressures can be determined
tremes of n' and K' caused fluctuations in calculated "crit- rather easily when fluid densities are available in the
ical" velocity of approximately ± 35 per cent. Somewhat form of pressure gradients as recommended by APr for
higher variations existed in Reynolds' number calculations, reporting mud weight (Table 2).
particularly in the region beyond laminar flow. Friction- One of the major problems of wellbore hydraulic analy-
pressure drop calculations were more seriously affected in sis which has not been discussed is that concerning calcu-
the laminar-flow region, while in the transition-turbulent lations in annular sections. There presently exists only
region the deviation between cements was about ± 15 limited information for this type of flow channel and,
per cent. while methods have been proposed for this situation,
Additives such as pozzolans, bentonite and diatomaceous they are largely unsupported by published experimental
earth generally tended to minimize variations in cements proof or data. In view of this inadequacy, it becomes
due to the decrease in cement content of the slurry. With contingent upon us, as on many previous authors, to use
these additives, n' and K' ranges were not as widespread the old standby-hydraulic radius or "equivalent diame-
and "critical" velocity deviations were usually ± 20 per ter". D. is described as four times the cross-sectional area
cent or less. Chemical additives such as calcium lignosul- of the channel divided by the wetted perimeter of the
fonate usually caused the opposite effect and created channel; for a concentric circular annulus, this reduces to
larger deviations. An extreme variation of 100 per cent D. = Do - DI (8)
in calculated "critical" velocity occurred with the ligno-
sulfonate, but was attributable primarily to two cements Admittedly, there is a certain undefined error introduced
which reacted at marked variance to the other 15 cements. by use of this term, particularly in thin annuli, but correc-
tion must await the development of a more practical
FLOW EQUATIONS analysis than now exists. Also for many of the pipe size-
hole size relationships encountered in well-completion
There is no particular complication involved in calcu- practice, the matter may be rather academic since the
lations for this system of rheology when compared to usual effect of the annulus often appears to be minimal in the
Newtonian formulas. The fluid properties n' and K' appear over-all hydraulic analysis, with the exception of calcu-
only in the Reynolds' number formula. This equation, with
the constant required to permit use of standard oiI-produc-
tion-industry dimensional units and with rearrangement of TABLE 2-PRESSURE GRADIENT
terms for easiest calculation, becomes Density Gradient
Ib/gal Ib/eu ft (psi/l,OOO ftl
N _ 1.86V'-·' p
8.33 62.4 433
Re - K' (96/D)n. (1) 8.5 63.6 442
9.0 67.3 468
For Newtonian fluids n' 1 and K' is proportional to 9.5 71.0 494
10.0 74.8 520
viscosity so that this relationship reduces to 10.5 78.5 546
11.0 82.3 572
NRC = 927.6 DV p (2) 11.5
12.0
86.0
89.8
598
624
JL 12.5 93.5 650
13.0 97.2 676
Another useful expression for some phases of these calcu- 13.5 101.0 702
lations is the relationship between Fanning friction factor 14.0 104.7 728
14.5 108.5 754
and Reynolds' number in the transition and turbulent 15.0 112.2 780
15.5 115.9 806
region for the non-Newtonian correlation of Fig. 1. 16.0 119.7 832
f = 0.00454 + 0.645 (NRe)-'<O (3)
16.5
17.0
123.4
127.2
858
884

MARCH, 1962 325


lating velocities for approaching turbulent flow. Even in Frictional Pressure Inside Tubing @ 27.9 bbl/min:
this latter situation, the error may not be acute because 27.9 (5.61)
of the almost equal reported efficiency' of upper laminar V = 60 (0.0325 ) = 80.3 ft/ second.
and transitional or turbulent flow in removing circulatable
mud from the hole. 1.86 (80.3)' " (12.6)
N Re = 0.28 (96/2.441)014 = 177,610.
FIELD EXAMPLES From Fig. 1, f = 0.0044,
P _ 0.039(3,800) (12.6) (80.3) 2(0.0044 )
The following are not hypothetical field situations; they Cl f - 2.441
represent some of the actual calculations which have been
used to set up cementing operations. Where data were = 21,704 psi.
available on performance of the job, they have been Obviously, it is impractical to consider annular turbulence
included. as a possibility.
NUMBER 1 Solution of the second phase was most simply achieved
Set 2V2 -in. tubing in 8¥! -in. hole at 3,800 ft. Use 12 per by trial-and-error methods . Eq. 1 using the known factors
cent gel cement with 0.3 per cent calcium lignosulfonate reduced to
to bring top of cement to 1,800 ft. N Re = 50.0 V UH •
Problem: Can turbulence be approached in the annulus? Substitution of this in Eq. 3 gave
What flow rate can be obtained with 3,000-psi pressure 1= 0.00454 + 0.0416 V ' 3
limitation? What hydraulic horsepower could be utilized? which, when introduced into Eq. 4, yielded
Slurry Properties: n' = 0.14, K' = 0.28 , and p = 12.6 ClPf = 3.473 V' + 31.82 y O' .
Ib/ gal.
Assuming hydrostatic heads and frictional pressure in the
Flow Channels: annulus would tend to balance each other .during the
2%·in. Tubing Annulus major portion of the mixing and displacement tim.e, espe-
D = 2.441 in. D, = 8.75 - 2.875 = 5.875 in . cially since the mud properties are not known, then ClP f
Area = 0.0325 sq ft Area = 0.372 sq ft = P lY = 3,000 psi, the maximum permissible pressure.
"Critical" Velocity in Annulus: The trial-and-error solution under this assumption gave
V ' .86 = 1,129(96/ 5.875) °14 (0.28) V = 27.75 ft/ second in the tUbing.
c 12.6 = 37.1 , Q = 27.75 (6~.~iO.0325) = 9.65 bbl/min.
V c = 7.00 ft/second.
_ 7.00 (60) (0.3725) HHP = 0.0245 (3,000) (9.65) = 709 hp.
Q -----:~;---'- = 27.9 bbl/ min.
5.61 NUI\1BER 2
Set 4Vz -in., 9.5-lb casing in TVs -in. hole at 3,307 ft. Use
Pozzolan X cement with 0 per cent bentonite (104 cu ft).
Calculate: Displacement rate to approach turbulent flow
in annulus. Hydraulic analysis for comparison with pres-
sures recorded during the actual operation.
Slurry Properties:
Fl ui d Properties Slurry Drilling Mud '
n' 0.30 0.29
K' 0.166 0.066
p 15.2 Ib/gol 9.7lb/gal
Pressure Gradient 790 psi /1 ,000 ft 505 psi/1 ,000 ft
' Properties measured on location while circulating prior
to cementing.

Flow Channels:
4'h·in. Casing Annulus
D = 4.09 in. De = 3.375 in.
Area = 0.0912 sq ft Area = 0.2278 sq ft

"Critical" Velocity lor Slurry in Annulus:


1.10 = 1,129(96/3.375)°30(0.166)
Vc 15.2 = 33.7.
V c = 7.91 ft/ second.
7.91 (60)(0.2278)
Q = 5.61 = 19.3 bbl/min.
Actual displacement rate during job = 16.5 bbl/min ,
which calculates to
V (Pipe ) 16.5 (5.61) = 16.92 ft/second.
60 (0.0912)
16.5 (5 .61)
V (Annulus) =6.77 ft/second .
FIG. 2-ROTATIONAL VISCOM ET ER. 60 (0.2278)

3·2 6 JO(!R NA L OF PETROLE U M TECHNOL OG Y


Frictional Pressure-Drop Calculations: Cement in pipe, Problem: Calculate rate to approach annular turbulence.
N - 1.86 (16.92)" (15.2) - 5 Slurry Properties: n' = 0.24, K' = 0.25, and p = 13.9
R. - 0.166 (96/4.09)0'0 - 8,0 0, lb/gal.
/ = 0.0057,
Annular Flow Channel:
e:..P = 0.039 (1,000) (15.2) (16.92)' (0.0057)
Hole Size 0, Area
, 4.09 (in.) (in.) (sq ft)
= 236.6 psi/1 ,000 ft. 8% 1.75 0.1503
Cement in annulus, 9% 2.75 0.2512
11 4.00 0.3927
N _ 1.86 (6.77)'7 (15.2)
1,610, Annular Calculations for "Critical" Velocity:
Rc - 0.166 (96/3.375)030
/ = 0.0099, Hole Size V. Q
(in.) (ft/second) (bbl/min)
e:..p, = -0.039 (1,000) (15.2) (6.77)' (0.0099)
-------,,---c==--------
8% 9.55 15.3
3.375 9.00 24.2
9%
= 79.7 psi/1,000 ft. 11 8.52 35.8
Drilling mud in pipe, Actual displacement rate during job 21.5 bbl/min.
N Re = 13,650, Attainment of a higher rate was not attempted due to
/ = 0.0054, the limited number of 2-in. lines possible between the
e:..Pf = 143.0 psi/1,000 ft. cementing units and the wellhead, and due to the proba-
Drilling mud in annulus, bility of lost circulation being induced at the extreme rates.
N Re = 2,710, Calculated Values at 21.5 bbl/min in Annulus:
/ = .0071, Hole Size V llP,
e:..P, = 36.5 psi/1 ,000 ft. (in.) (ft / second) N"c (psi /1 ,000 ftl
8% 13.37 3,790 365
Hydraulic Analysis with All Slurry Inside Casing: 8.00 1,720 117
9%
104 cu ft of slurry in 41h-in., 9.5-lb casing. 11 5.12 855 66
Height of cement column = 104/0.0912 = 1,140 ft. These data were for the first stage of a two-stage job
Height of mud in pipe = 3,307 - 1,140 = 2,167 ft. with the stage collar located at 10,727 ft. Indicated cement
Total llP, top from a sonic bond log was 10,800 ft. Upper stage
1.140 X 236.6 = 270 psi was cemented at 24 bbl/min using a similar slurry except
2.167 X 143.0 = 310psi that calcium lignosulfonate percentage was reduced to 0.2.
3.307 X 36.5 = 121 psi This well was successfully completed in two formations
701 psi and perforated in two other formations without communi-
cation between zones. Indicative of the value of high
1.140 X 790 = 901 psi 3.307 X 505 = 1,670 displacement rates was the correlation between caliper log
2.167 X 505 = 1,094 psi and sonic bond log, where the lower sound-transmission
1,995 psi values were recorded in the section of the hole nearest
Pw = 701 - 1,995 + 1,670 = 376 psi. bit size. Similar comparisons have also been obtained on
Actual P IV on pressure recorder at this stage of the job other wells, particularly when displacement rates ap-
was 350 psi. proached turbulence.
These examples represent the predominant usage of the
Hydraulic Analysis Just Be/ore Bumping Top Plug: techniques and data discussed herein; further elaboration
28 cu ft of slurry in 4Yz -in casing (L = 307 ft) . for other jobs on which these calculations have been used,
76 cu ft of slurry in annulus (L = 334 ft). and for which data are available, would be largely repeti-
tion. Problems encountered using high displacement rates
Total llP f for cementing casing have mainly been caused by restric-
0.307 X 236.6 = 73 psi tions to the flow. The principal of these restrictions has
3.000 X 143.0 = 429 psi been the manifolding arrangement available between ce-
0.334 X 79.7 = 27 psi menting units and the top of the casing. Under normal
2.973 X 36.5 = 109 psi
operating conditions this represents some 500 to 1,000
638 psi
equivalent ft of 2-in. tubing, and relatively high llP, may
be encountered here. This situation has been combated on
0.307 X 790 = 243 psi 0.334 X 790 = 264 psi at least one occasion by bringing the discharge line of
3.000 X 505 = 1,515 psi 2.973 X 505 = 1,501 psi each cementing pump into a manifold and proceeding
1,758 psi 1,765 psi from this manifold to the 13%-in. cementing head with
Pw = 638 - 1,758 + 1,765 = 645 psi. two 4-in. lines.
Recorded P w was 650 psi before top plug landed. Generally, the restriction in floating equipment has not
apparently been significant because its equivalent length
NUMBER 3 represents a very small percentage of the total depth of
Set 7-in. casing in 8*-in. hole at 13,150 ft. Caliper casing. It is quite obvious that advanced planning by the
showed portions of the hole washed-out to 9* and 11 in. well owner is necessary to organize a job of this type to
Use Pozzolan Y cement, 0.4 per cent calcium lignosulfo- insure sufficient tankage, water supply and proper mani-
nate and 121h Ib of gilsonite / sack (total 350 sacks), with folding of equipment, as well as to establish what type
1,000 gal of chemical wash ahead of cement slurry. of slurry could be used to the greatest advantage.

MARCH, 1962 327


CONCLUSIONS fl.P, = frictional pressure drop, psi
L = length of flow channel, ft
A relatively simple method has been suggested to permit Vc = bulk critical velocity, ft/sec
hydraulic well bore analysis for cementing operations, HHP = hydraulic horsepower
whereby it is possible to calculate several factors which P IV = wellhead pressure, psi
influence primary cementing practices. Using these criteria, Q = bulk flow rate, bbIjmin
it is then possible for the well owner to select the most P". = hydrostatic pressure-annulus, psi
satisfactory arrangement of equipment and (to a lesser P he = hydrostatic pressure-casing, psi
degree) the cementing composition to obtain a pre- De = equivalent diameter-annulus, in.
conceived performance during the operation within reason- D. = outside diameter-annulus, in.
able limits of engineering accuracy. The essential data not D I = inner diameter-annulus, in.
usually available are the slurry rheological properties
which affect flow performance. A substantial number of REFERENCES
more-common slurries have been measured and these
properties determined, in some cases on API Class A 1. Howard G. C. and Clark, J. B.: "Factors to be Considered
in Obt~ining Proper Cementing of Casing", Oil and Gas
cement from 17 different manufacturing plants. A simpli- Jour., (Nov. 11, 1948).
fied method using standard cementing-laboratory facilities 2. Rabinowitsch, B.: "Uber die Viskositat and Elastizitat von
will yield these data on other cementing compositions Solen", Phys. Chem. (1929) 145A, No.2, 1.
rather rapidly. 3. Metzner, A. B. and Reed, J. C.: "Flow of Non-Newtonian
As with other non-Newtonian fluid systems, work of this Fluids-Correlation of the Laminar, Transition and Turbulent
Flow Regions", AIChE Jour. (1955) I, No.4.
nature on cementing slurries is continuing. It is anticipated
4. Metzner, A. B.: "Recent Developments in the Engineering
that improvements will be made in all areas of rheological Aspects of Rheology", Rheologica Acta (1958) No. 2-3.
analysis, but there is no indication thus far that gross 5. Dodge, D. W.: "Turbulent Flow of Non-Newtonian Fluids in
inaccuracies exist in this sl'ggested simplified analysis. Smooth Round Tubes", PhD thesis, U. of Delaware (June,
1958) .
NOMENCLATURE 6. Melton, L. 1. and Saunders, C. D.: "Rheological Measure-
ments of Non-Newtonian Fluids", Trans., AIME (1957) 210,
196.
N Re = Reynolds' number, dimensionless
7. Savins, J. G.: "Generalized Newtonian (Pseudoplastic) Flow
V = average or bulk velocity, ft/second in Stationary Pipes and Annulus", Trans., AI ME (1958)
volumetric flow rate/cross-sectional area of 213,316.
flow channel 8. Savins, J. G. and Roper, W. F.: "A Direct Indicating Vis-
p = fluid density, lb/gal cometer for Drilling Fluids", Paper No. 90l-30-A presented
n' = fluid-flow behavior index, dimensionless to API Div. of Production (1954).
K' = fluid consistency index, lb-sec'" /sq ft 9. Ish-Shalom, Moshe and Greenberg, S. A.: "The Rheology of
Fresh Portland Cement Pastes", Paper presented at Fourth
D = diameter of flow channel, in. Int. Symp. on Chern. of Cement (Oct. 2-7, 1960) in Wash-
fJ. = Newtonian fluid viscosity, cp ington, D. C.
f = Fanning friction factor, dimensionless 10. "Mud Pressure Gradient", Pet. Eng. (Jan.,1958). ***

328 JOURNAL OF PETROLEUM TECHNOLOGY

Você também pode gostar