Sie sind auf Seite 1von 7

SPE 9324

SPE
SocIety of PatroIetm EngIneers of AIME

AN EMPIRICAL CORRELATION TO PREDICT DIAMOND BIT DRILLING RATES

by Clare P. Duklet, Christensen, Inc.;


Thomas R. Bates, Shell Oil Co.

@Copyright 1980, American Institute of Mining, Metallurgical, and Petroleum Engineers, Inc.
This paper was presented at the 55th Annual Fall Technical Conference and Exhibition of the Society of Petroleum Engineers of AIME, held in Dallas, Texas, September 21-24, 1980.
The material is subject to correction by the author. Permission to copy IS restricted to an abstract of not more than 300 words. Write: 6200 N. Central Expwy., Dallas, Texas 75206.

have been formulated to extend the correlation to each


ABSTRACT unique set of drilling conditions and various bit de-
signs and sizes. The empirical equation can be easily
Diamond bit drilling tests have been performed applied to most field operations, with reasonable
at the Drilling Research Laboratory (DRL) in Salt Lake accuracy.
City, Utah. These tests were sponsored and supervised
by the Shell Development Company. The tests simulated
down hole drilling pressures on a typical diamond TEST FACILITY
drill bit and oilfield formation. Variations of
typical field operating variables were tested for Drilling tests were conducted in the DRL well
penetration rate response. An empirical equation has bore simulator. DRL test facility details and cap-
been derived from the test results to predict diamond abilities have been described previously by Tibbitts 3
bit drilling rates. Essentially, the DRL well bore simulator accomodates
a sizable rock specimen under simulated down hole
pressures for full scale drilling tests. Authenti-
INTRODUCTION cally reproduced overburden, confining, and bore hole
pressures or stresses are the key features of the fac-
Oil well profitability is determined by drilling ility. Although the overburden and confining press-
costs. Penetration rate is a substantial factor in ures have only a weak relationship to drillability,
determining well drilling costs. Diamond drill bits the capability to vary the bore hole pressure is of
are used to economic advantage in many drilling situa- upmost importance to penetration rate response.
tions. Design, performance, and application of dia-
mond bits has been greatly improved the past lQ to 15 The DRL facility is capable of utilizing any mud
years. However, only a small percentage of oil well system, at full scale flow rates. A computer re-
footage is drilled by diamond bits. If the penetra- trieval system allows for an accurate and instantan-
tion rate could be easily and accurately predicted, eous record of all the pertinent test variables.
more economic applications for diamond bits may be
found.
TEST PROCEDURE
Existing methods for predicting diamond bit per-
formance range from simple charts to sophisticated Three rock specimens of Mancos shale were used
computer inalysis. Simpler methods, as cited by in the drilling tests. The rock properties and de-
Striegler , do not account for all the pertinent scription are presented in Table 1. The same diamond
drilling variables and have been found to error up to drill bit was used on each rock sample. Diamond bit
200%. Sophisticated methods, developed by Appl and design and specifications are detailed in Table 2.
Walker2 , consider all the drilling and bit design Each rock specimen was drilled using a different field
variables. Accurate prediction of penetration rate typical mud system. Mud system description and pro-
can be obtained using sophisticated computer programs, perties are given in Table 3.
if appropriate bit design and field data are avail-
able. Two different values of weight on bit, rotary
speed, flow rate, and bore hole pressure were tested
A new empirical correlation to predict diamond on each rock specimen and mud system. These values
bit drill rates has been derived from full scale test of the test variables were combined to provide sixteen
results at simulated down hole conditions. The de- unique test conditions or data sets. Tests #1 and #2
rived empirical equation involves most of the perti- were completed for the sixteen test conditions. Test
nent drilling and bit design variables. The variables #3, however, achieved only twelve test conditions when
References and tables at end of paper.
2 AN EMPIRICAL CORRELATION TO PREDICT DIAMOND BIT DRILLING RATES SPE 9324

the rock specimen collapsed. All of the test vari- The correlation variables WDI, HSI, and EFS are
ables were within the range of typical oil field specifically constituted variables to generalize the
operating conditions. correlation to other diamond sizes, bit sizes, rock
formations, hydrostatic pressures, and hydraulic
conditions. These generalized variables require fur-
TEST RESULTS ther explanation, as follows:

The test results are presented in Table 4. Each


data line or test condition represents approximately WDI VARIABLE
two inches of rock drilled. The DRL computer prints
out the average variable values for five second drill- The WDI variable is defined as the average weight
ing intervals. Five to ten intervals are recorded per cutting diamond per projected area of the diamond.
for each test condition. The interval readings are The projected area of the diamond is based on the
also averaged to obtain each line of data presented diamond diameter. The WDI variable can be calculated
in Table 4. from the following expression:
The most significant testing error results from
variations in rock specimen strength. From small
cores taken from each test rock, laboratory compres- WDI WOB - .942 PSI (D-l! . (2)
sive strength tests were performed. A compressive
strength variation of +10% was determined, under .026 TCW DPC 28
confining pressure. -
New variables introduced in Equation (2) are
MOst of the test variable values, in Table 4, defined in the Nomencl.a.ture. The numerator, an the
are straight forward instrument recordings. The bit right side of Equation (2), is composed of the meas-
pressure drop, however, was determined by the differ- ured weight on bit (WOB)minus the pump off force. The
ence between the standpipe and bore hole pressure pump off force is approximated by an empirical express
while drilling minus the same difference when off ion derived from previous Christensen tests. The pump
bottom. This calculation gives the pressure loss off force can be a SUbstantial hydraulic force created
generated on the face of the bit while drilling. by the differential pressures on the bit, due to the
bit face pressure drop. This force tends to unload
the cutting diamonds and is subtracted from the meas-
CORRELATION FORM ured load to obtain the actual drilling weight. The
actual drilling weight is then divided by the number
The objective of correlating the test results of face cutting diamonds and the projected diamond
was to obtain an expression for the penetration rate area. The expression in the denominator of Equation
as a function of the pertinent test variables. Also, (2) is derived in Appendix A and yields the number of
a fairly simple equation was desired to the cutting diamonds times diamond projected area. The
gap between existing, simplified and complicated WDI variable generalizes or compensates for bit size,
prediction methods. From the results of many previous diamond size or sharpness, and number of cutting
experimenters and present practical objectives, the diamonds.
following mathematical form was assumed for analysis:

ROP HSI VARIABLE


The correlation variables are defined in the The HSI variable is defined as the hydraulic
Nomenclature and explained in detail later. Equation horsepower per square inch of projected bit area. HSI
(1) simply states that the rate of penetration is can be expressed as follows:
equal to a constant times the product of the funda-
mental variables each raised to some power. Equation
(1) is an assumed form and its validity depends on PSI GPM
HSI ( 3)
how well it correlates with the test data.
Applying the correlation to field operations assumes 1346 D2
that the drilling mechanism in the laboratory tests
is similar to actual field drilling. Unusual field Equation (3) is simply a calculation of the hydraulic
conditions, such as fluid jetting or cutting of the horsepower expended at the bit face and divided by
formation, extremely flat or worn diamonds, etc. may the hole or projected bit area. This variable has
render the correlation inappropriate. been used by Christensen for many years to relate the
cleaning, cooling, and drilling efficiency attribut-
able to hydraulic factors. The HSI expression gener-
CORRELATION VARIABLES alizes the hydraulic variables for any bit size.

The ROP, RPM, and AFL variables are taken direct-


ly from the test data and are defined in the Nomen- EFS VARIABLE
clature section. The constants a, b, c, e, and f
are derived from the data analysis and are dimension- The EF$ variable or effective formation strength
less values. The constant CST is also derived from is defined as the sum of the unconfined compressive
the data analysis and can be considered to have the strength and the hydrostatic pressure below the bit.
proper dimensions to render Equation (1) dimensionally The EFS variable is expressed as follows:
consistent.
EFS UCS + .052 PPG DEP . . . . (4)
SPE 9324 C.P. DUKLET AND T.R. BATES 3

The unconfined compressive strength (USC) is the aver- isiana and two bits in the Wasatch trend of Utah. The
age of standard laboratory test results. The hydro- field data in Table 5 represents a good variation of
static pressure is calculated by the product of a unit weight on bit, well depth, and diamond size. The
conversion constant, mud weight, and well depth. The Tuscaloosa bit runs were in a relaxed fluid loss oil
EFS variable is expressed in psi and is analogous to base mud and the Wasatch runs in a water base mud6
the down hole compressive strength. For the test The Wasatch bit runs were those reported by Primm .
data, the expression for hydrostatic pressure is re-
placed by the regulated bore hoZe pressure. Investi- In Table 6, Equation (5) was used to calculate
gators, such as Appl and Rowley, have demonstrated the predicted penetration rate for the nine field runs
that diamond bit drilling performance is proportional in Table 5. Using the average laboratory tested com-
to strength calculations similar to Equation (4). pressive strength and field reported data, the pre-
Field experience indicates that penetration rates are dicted penetration rates compare to the actual rates
strongly related to formation strength, mud weight, with an average error of 15.3% and a maximum e~ror of
and well depth. 40.5%. Using Equation (6), the compressive strength
was calculated from the field data and is shown in
Table 6.
CORRELATION RESULTS
In Table 7, 280 field run diamond bits have been
Given the assumed correlation form and the de- analyzed for formation strength and predicted drill
fined correlation variables, the test data in Table rates. Six formation trends are represented from five
4 can be analyzed. In Appendix B, Equation (1) is state locations. Using the actual drill rates and
transformed into the mathematical form of Equation reported field data, the unconfined compressive streng
B-9. Equation B-9 is in the required form for multi- th was calculated from Equation (6). The minimum,
ple linear regression analysis. One of the standard average, and maximum calculated strength is presented
computer routines was used for the regression analy- for each formation trend. Using the average calcul-
sis. Based on the 44 test conditions of Table 4, the ated strength, the predicted penetration rates were
resulting co~relation equation was obtained: calculated from Equation (5). The average error of
predicted versus actual drill rates was 27% for the
WDI. 62 RPM. 60.HSI.26AFL.14j 280 diamond bit runs.
ROP = 1,220,000 . (5)
( EFS 2 1 CONCLUSIONS
The empirical correlation expressed by Equation An empirical correlation to predict diamond bit
(5) fits the test data of Table 4 to within an aver- drilling rates has been derived from full scale labor-
age error of 6.6% and a maximum error of 19.5%. The atory tests at simulated down hole conditions. The
computer analysis had a standard error of 3.5%, based derived correlation relates accurately to the test
on the logarithmic test data. This degree of acc- data and considers most of the fundamental drilling,
uracy appears somewhat remarkable, considering the formation, and bit design variables. The correlation
number of variables, extent of test conditions, and variables have been uniquely formulated to extend the
rock specimen inhomogeneity. It appears that the correlation to other rock formations, bit designs, and
assumptions of Equations (1), (2), (3), and (4) are drilling conditions. Applying the correlation equat-
sufficiently validated by the correlation result and ions to actual field data reveals an acceptable de-
the test data. gree of accuracy with a minimum amount of calculation.
It appears that the correlation results can be applied
Other than using Equation (5) to predict diamond to field drilling operations to improve planning, bit
bit drilling rates, the equation can be solved to selection, operating parameters, drill rates, and
predict the unconfined compressive strength. economics.
WDI' 30RPM 29HSI ' 12AFL' 07)
UCS 790 NOMENCLATURE
( ROp 48
A,B,C - Designates test mud system.
- .052 PPG DEP . . . (6) a,b,c,e,f - Exponential constants, dimensionless.
AFL - A.P.I. fluid loss, cc/30 min.
Equation (6) can be applied to past performance data CST - Proportionality constant, dimensions
to obtain an approximation for field formation uncon- consistent with Equation (5).
fined compressive strength. This strength approxi- d - Diamond diameter, in.
mation can be used for future penetration rate pre- D - Drill bit diameter, in.
dictions, using Equation (5). DEP - Average drilling depth, ft~
DPA - Diamond projected area, in
The results of Equations (5) and (6) are some- DPC - Diamonds per carat, dia/carat.
what similar to the results Cunningham) obtained for EFS - Effective formation strength, psi.
roller bits. GPM - Flow rate, gal/min.
HSI - Hydraulic horsepower per square inch,
hp/in2 .
FIELD APPL M - Constant, dimensionless.
NCD - Number of cutting diamonds.
Nine diamond bit field runs are presented in PSI - Diamond bit pressure drop, psi.
Table 5, where laboratory tested compressive strength PPG - Mud weight or density, Ib/gal.
had been determined from core samples. Seven of the ROP - Rate of penetration, ft/hour.
diamonds bits were run in the Tuscaloosa trend of Lou- RPM - Rotary speed, rev/min.
4 AN EMPIRICAL CORRELATION TO PREDICT DIAMOND BIT DRILLING RATES SPE 9324

T -
Designates Tuscaloosa formation.
TCW -
Total bit carat weight, carats. DPA (A-2)
UCS -
Unconfined compressive strength, psi.
W -
Designates Wasatch formation. 4
WDI Average weight or load ~er diamond per
- The diamond diameter is related to the diamond
inch squared, lb/dia-in size variable DPC, as follows:
WOB - Weight on bit, lb.
X - Independent variable designation.
Y - Dependent variable designation. d
.193 . (A-3)
DPC 36
ACKNOWLEDGEMENTS
Combining Equation A-I, A-2, and A-3, the num-
We would like to thank the staff and management ber of cutting diamonds times the diamond projected
of the Shell Development Company and Christensen, area is ob tained.
Incorporated for their cooperation, assistance, and
permission to publish the test data. NCD DPA = .026 TCW DPc 28 (A-4)
Equation A-4 is the expression used in Equation
REFERENCES (2) of the text.
l. Streigler, J.H.: "Proper Use of Diamond Bits
Reduces Drilling Costs," World Oil (July 1979), APPENDIX B
109-116.
Multiple linear regression analysis requires
2. MOore, N.B., Walker, B.H., and Appl, F.C.: tlA that relationships or equations be of a certain mathe-
MOdel of Performance and Life of Diamond Drill matical form. Equation (1) of the text is not the
Bi ts, n Journal of Pressure Vessel Technology exact form required. However, if the natural logar-
(May 1978), 164-171. ithm is taken on both sides of Equation (1), we
obtain:
3. Tibbitts, G.A., Sandstrom, J.S., Black, A.D., and
Green, S.J.: tiThe Effects of Bit Hydraulics on In( ROP) = 1n( CST WDIWHSl cAFL eEFSf)
Full Scale Laboratory Drilled Shale, n paper
SPE 8439 presented at SPE 54th Annual Fall Meet- Or:
ing, Las Vegas, Nevada, September 23-26, 1979.
In( ROP) = In( CST )+aln(WDI )+bln( RPM)
4. Appl, F.C. and Rowley, D.S.: "Analysis of the
Cutting Action of a Single Diamond," Society of
+cln(HSI)+eln(AFL)+f1n(EFS) (B-1)
Petroleum Engineers Journal, (September 1968),
268-280. Defining: Y 1n(ROP) (B-2)
5. Ctmningham, R.A.: nAn Empirical Approach to
Relate Drilling Parameters," paper SPE 6715 pre- M" 1n( CST) (B- 3)
sented at SPE 52nd Annual Fall Meeting, Denver,
X = In(WDI) .( B-4)
Colorado, October 9-12, 1977. 1

Primm, L.A.: "Diamond Drilling at Altamont, X = In(RPM) (B-5)


6. 2
Utah," paper SPE 5073 presented at SPE 49th Annual
Fall Meeting, Houston, Texas, October 6-9, 1974. X3 = 1n(HSI) .( B-6)
X In(AFL) (B-7)
4
APPENDIX A
X5 = In(EFS) (B-8)
The number of cutting diamonds on a bit can be
calculated from the following expression: Substituting the definitions of Equations B-2
through B-8 into Equation B-1, we obtain:
NCD = .9 TOW DPC (A-I)
Y = M+ a~
+ bX + cX +ex + fX .(B-9)
2 3 4 5
The 90% factor in Equation A-I approximates the
carat weight of the face cutting stones. Ten percent Equation B-9 is in the required mathematical
of the total carat weight is used for gage stones or form for multiple linear regression analysis. The
stabilization and does not sustain a vertical load. constants M, a, b, c, e, f can be determined from
Multiplying the carat weight of the face cutting dia- a standard computer routine for the best fit accord-
monds times the diamonds per carat yields the number ing to Equation B-9. Constants a, b, c, e, f are ob-
of face drilling stones. tained directly and CST is equal to the natural logar-
ithm base e raised to the M power.
The projected area of a diamond, based on the
diamond diameter, is straight forward as follows:
TABLE 1 TABLE 2

TEST ROCK TEST BIT

Name Mancos Shale Diameter 6 1/8 inch

Description Shale-Sandstone Style MD 262

Compressive Strength - 7500 psi Total Carat Weight 317

Grain Density 2.67 gms/cm3 Diamonds Per Carat 3

Porosity 5.2% Total Flow Area

Water Saturation 75% Diamond Quality SP Grade

Coring Location Central Utah Manufacturer Christensen Inc.

TABLE 3

TEST MUD SYSTEMS

Letter General Specific Mud A.P.I.


Designation Type Characteristic Fluid Loss
00/30 min

A Water Base Non Dispersed 10.3 7.8

B Oil Base Relaxed Fluid Loss 12.5 4.2

C Oil Base Conventional 12.4 .1


TABLE 4

TEST DATA

Test Weight Rotary Flow Bi t Pressure Bore Hole Drill


Condition On Bit Speed Rate Drop Pressure Rate
Designation Ibs rpm ~ psi psi ft/hr

Test #1: Mud System-A, Overburden Pressure-4000 psi, Confining Pressure-3200 psi

A-I 12,004 100 160 427 1991 9.06


A-2 12,018 140 160 441 1998 10.37
A-3 19,890 100 160 590 1965 12.36
A-4 19,936 140 160 642 1973 14.41
A-5 11,969 100 200 533 1959 9.51
A-6 12,003 140 200 552 1962 10.93
A-7 19,804 100 200 681 1954 13.37
A-8 19,875 140 200 689 1968 :l5.55
A-9 11,929 100 160 432 469 10.88
A-1O 11,959 140 16o 439 454 14.78
A-ll 19,834 100 160 538 449 16.62
A-12 19,853 140 160 545 465 20.86
A-1] 11,988 100 200 600 488 12.71
A-14 11,980 140 200 586 509 14.02
A-15 1$),892 100 200 769 490 16.95
A-16 19,904 140 200 740 514 23.04

B-1 ll,850 102 160 497 2051 8.26


B-2 12,032 144 160 513 1934 9.23
B-3 19,961 101 160 595 1928 12.97
B-4 20,030 144 160 604 1898 16.90
B-5 ll,995 104 200 690 1956 8.21
B-6 12,021 144 200 691 1941 10.16
B-7 19,966 104 200 802 1940 13.71
B-8 20,086 146 200 832 1916 17.82
B-9 11,818 104 160 .501 524 11.67
B-I0 11,715 145 160 539 544 13.64
B-ll 19,501 100 160 670 443 15.98
B-12 19,608 144 160 702 522 18.28
B-13 ll,592 101 200 970 515 12.63
B-14 11,644 145 200 953 466 16.54
B-15 19,553 101 200 ll34 586 17.49
B-16 19,637 144 200 1178 598 22.76

Test #3: Mud System-C, Overburden Pressure-7200 psi, Confining Pressure-6400 psi

C-l 12,238 101 160 600 2246 3.75


C-2 12,165 142 160 599 2164 4.91
0-3 20,472 102 160 766 2066 6.97
C-4 20,103 145 160 777 1922 9.24
C-5 12,266 102 200 756 2074 4.69
0-6 12,315 142 200 809 2092 5.29
0-7 20,267 102 200 958 2174 7.35
0-8 20,520 145 200 962 2082 9.91
0-9 12,114 103 160 569 492 7.03
0-10 12,089 144 160 614 538 8.29
O-ll 20,013 103 160 735 540 11.35
0-12 20,242 145 160 787 531 14.02
TABLE 5

FIELD DATA - 6-1/8 INCH DIAMOND BITS

Bit A.P.I. Average Total Diamonds Actual


Bit Weight Rotary Flow Pressure Mud Fluid Well Carat Per Drill
Run On Bit Speed Rate Weight Loss Depth Carat Rate
Designation ..lE.!!.... ...!!?!!!.- ~ .J2ElL cc/30 min ft dia/cts ft/hr

T-l 14,000 100 183 600 14.5 3.8 18,230 298 3.0 ).64
T-2 16,000 110 200 600 15.0 3.2 18,705 298 3.0 2.67
T-3 16,000 110 190 600 15.4 3.2 19,0)0 298 3.0 2.60
T-4 17,000 100 200 500 15.0 3.0 18,262 298. 3.0 4.00
T-5 12,000 110 160 500 15.0 3.4 18,660 298 3.0 1.71
T-6 12,000 110 160 500 15.0 3.4 18,904 193 6.0 1.90
T-7 12,000 110 160 500 15.0 3.4 19,140 277 1.5 2.06

W-l 22,000 108 175 900 15.9 4.8 11,632 385 2.5 5.72
W-2 32,500 120 160 900 15.9 4.8 12,636 484 1.5 10.00

TABLE 6

EMPIRICAL PREDICTION COMPARED TO FIELD DATA

Actual Predicted Prediction Laboratory Prediction


Bit Drill Drill Percent Tested Predicted Percent
Run Rste Rste * Error* UCS UCS
** **
Designation' f't/hr ft/hr psi psi

T-1 3.64 2.25 -38.3 5000 1150 -77.0


T-2 2.67 2.51 - 5.8 5000 4460 -10.8
T-3 2.60 2.22 -14.8 5000 3520 -29.6
T-4 4.00 2.38 -40.5 5000 800 -84.0
T-5 1. 71 1.80 + 5.4 5000 5500 +10.0
T-6 1.90 1.99 + 4.5 5000 5420 + 8.4
T-7 2.06 2.07 + .7 5000 5070 + 1.4
W-l 5.72 6.08 + 6.3 4000 4390 + 9.8
W-2 10.00 7.25 -27.5 4000 1960 -51.0

Based on equation (5)


* Based on equation (6)
**
TABLE 7

UNCONFINED COMPRESSIVE STRENGTH DERIVED FROM


EMPIRICAL EQUATION AND ACTUAL FIELD BIT DATA

Number Unconfined Compressi ve Strength Drill Rate


Formation Formation General Bit Runs Predicted + Prediction
Name Type Location Used In Minimum Maximum Average Error
Analysis psi ~ ++

Lewis-Almond Sand-Shale Wyoming 50 2100 6100 9600 22


Miocene Sand-Shale Louisiana 45 930 3800 9300 33
Morrow-Springer Sand-Shale Oklahoma 65 2400 4800 7600 20
Vicksburg Sand-Shale Texas 40 1000 5200 8900 30
Wasatch Sand-Shale Utah 25 2000 6900 9400 18
Wilcox Sand-Shale Texas 55 1000 5100 10,000 35
+ Based on equation (6)
++ Based on Average Predicted UCS and equation (5)

Das könnte Ihnen auch gefallen