Beruflich Dokumente
Kultur Dokumente
by C.M. Pearson
Now with ARCO
q
Gu/f R&D Co., and N.B. Nguyen, Gulf Oil E&P Co.
SPE Members
Copyright 1985, Society of Petroleum Engineers This paper was preeented et the SPE 1985 ProductionOparetiona Symposiumheld in Oklahoma City, Oklahome, March 10-12, 1885. The material ie eubjectto correctionby the author, Perrniaaionto copy ie restrictedto an abefractof not more than S00 words. Write SPE, P.O. Box 8SS8S6, Riihardaon, Texaa 75083-S8S6. Talex 7S028S SPE DAL.
1,2
Ifi
~h~
-----present
work
results are included from a field research program set up similar to the program described by other engineers studying the appl~cation of massive hydraulic fracture treatments. Three West Texas formations were chosen for study during the first phase of this project. The criteria used in field/formation selection are listed in Table 1. Formation A is a shallow (2500-3000 ft deep) oil reservoir with typical net and gross pay zones of 150 and 300 ft respectively. It is characterized by gradual structural changes and no well defined cap and base rocks. Formation B is a moderately deep (6500 ft) predominantly gas reservoir with defined shale barriers. Formation C is a deep (11-15,000 ft), tight oil and gas formation characterized by a high degree of secondary permeability and poorly
A,36<. ..A 1....4--. Ucl Illeu Uall 1=1>.
on the created fracture and the required fracturing pressure. This information was then used in the field to optimize limited size treatments in particular formations.
INTROOUCTIOH The technique of hydraulically fracturing formations to increase production rates and available reserves is ccmnnon practice within the petroleum industry. The placement of an optimal fracture to act as a high conductivity flw channel requires both correct design and correct field procedures. The measurement, interpretation and understanding of previous treatments together with on-the-job analysis and redesign play an important role in improving treatment performance. This paper presents a sensitivity analysis of input parameters on fracture design, and results of using this information to improve hydraulic fracture treatments. Similar analysis has been used in past studies to understand the effect of
A number of different tests and measurements were selectively carried out in conjunction with the 35 hydraulic fracture treatments considered part of this project. It is not the purpose of this paper to present all of the data from each test conducted in this study. Material will be limited to showing how an analysis of input parameters affecting fracture design identified certain critical parameters, and how field measurements of these values were attempted. FRACTURE DESIGN Two fracture simulators were used for design and post-fracture analysis reported in this paper. They are referred to as the GULF-1 and GULF-2 simulators. The GULF-1 hydraulic fracture simulator is an analytical model for the design and evaluation of treatments= It is ~ two-djmensi~n~l @QrjQ~f~r.~h~ case of constant height vertical fractures. It consists of a fracture propagation model, a proppant transport model, and a reservoir production model. These nmdels incorporate several improve.AA..1 . ~~f~e ==1.,++.-31 .?.+e+.lm nu.+e ,,K,, ~,:~~ b= UllUIJblbGll lIWUCIS GAsa(.mlly retaining the simplicity of analytical solution in
77
PARAMETERS AFFECTING FRACTURE DESIGN AND THEIR MEASUREMENT CURING TREATMENTS closed form. This allows for rapid and inexpensive fracture designs and sensitivity studies of design parameters. The fracture propagation model predicts fracture length, width, bottomhole pressure, ~nd~ fracture -SZ2-:----- c-- rerKlns-Kern-Noragren 3 - rot- -- -ell Ictency ~elllptical fracture width) and Geertsma-DeKlerk (constant fracture width) type fracture geometries. The formulation incorporates fracture continuity and pressure dependent fluid leakoff (with spurt loss) to the formation.
SPE 13798
laboratory core tests. It is therefore important to knw to what degree an error in the value of each of these parameters will affect the fracture design. The required accuracy of input data may then be determined. --- . A rdfigeOf iiive~tig~ti~nfCIr each input parameter was determined. Although somewhat arbitrary, ranges were determined to reflect likely maximum and minimun values that could be encountered in The ranges determined for the this formation. mechanical and fluid flow properties of Formation A are given in Table 3. Fracture simulations were then run using the standard set of treatment data listed in Table 2. The resulting variations in fracture length, width, and job efficiency are shown in Table 3. A variation coefficient (V) was used for comparisons in this analysis. This is defined as:
v =
The proppant transport model was developed for variable width, constant height fractures, and considers proppant transport by non-Newtonian pseudoplastic (power law) fracturing fluids. The . ~,c~e? .w,a~... ifi 1 Ull the sensitivity analysis with specified batches of slurry with a given proppant size, density, concentration and settling velocitY. The production model, which was used in the study for post-stimulation production history matching, was developed for a finite conductivity fracture in an infinite reservoir. The reservoir can be simulated as producing under either constant rate or constant pressure conditions. Details of the production model were reported by Lee and Brockenbrough. The GULF-2 hydraulic fracture simulator is an analytical model for the design and evaluation of vertical hydraulic fractures with variable height. It is based on the pseudo JhreeA:.----,--7 ulmerl>lunal --* moae I deveiopea by Ciearyanci modified to account for variable injection rate and closure. The simulator has only a fracture propagation model. This model allows for injection, shut-in, reinfection and closure. This model was used in the sensitivity analysis to show the effect of confining stress contrast. More detailed information on the use of these models for fractu$e design has been presented by Abou-Sayed et al.
(ye;nman )
100%
(Values of V of 10% or greater in fracture length, width and treatment efficiency have been printed in boldface type in Tables 3-7).
.-.=,J-,Analvcie of ~~~
mnt-h..+..l ,Im,,,, , w ,
prcper~~e~
.hA_ alluwa
+h.+ L.llab
neither values of Poissons ratio (0.15-0.3) nor the minimum in-situ (1500-2500 PSi ) stress resulted in a variation of 10% or more for the range of values used. B# contrast the range of _..-Ioungs modulus (3-7 x i~- psi) resuiteci in a iZ variation in average fracture width$ and the ranae in total compressibility of 1 to 100 x 10-5 psi=l (reflecting the difference between oi1 and gas saturated formations) resulted in significant length and efficiency variations. The dominant effect of change in fracture height (100-400 ft) is shown in the resulting variations of 30 and 37% in fracture length and efficiency. A similar result of variations in input reservoir parameters having different effects was also found. The ranges of porosity, permeability, reservoir fluid viscosity and reservoir pressure all resulted in less than a 10% variation in output characteristics. The effect of varying the pay zone height from 50 to 200 ft (and hence the leakoff area) resulted in a 19% change in fracture length and 25% change in job efficiency. The variations in maximum bottomhole pressure for the ranges of these input parameters are shown in Table 4. The results are particularly important because of the wide use made of analyzing the fracture treating pressure to predict the nature of fracture propagation. Table 4 shows that and variations in minimum in-situ stress (S3) fracture height dominate the fracturing pressure. A 1000 psi change in S3 results in nearly 1000 psi difference in bottomhole treating pressure; and a change from 100 to 400 ft in fracture height gives a 700 psi variation. The inference of this is that if all other parameters are known, and either fracture height or minimun stress can be measured in the field, the other may be determined from analysis of the fracturing pressure.
SENSITIVITY
ANALYSIS OF FRACTUREDESIGN
Table 2 lists a typical set of values used for fracture design in Formation A. For the work presented in this paper the Perkins-Kern-Nordgren type fracture geometry was used exclusively. This geometry had been determined to be nmre,appropriate for the reservoir conditions being studied, and enabled a common geometry to be used by both the constant height and variable height fracture simulators. Input parameters gories:
q
are
divided
into
three
cate-
Mechanical properties of the formation, Properties affecting flu d flow in the reservoir, Treatment data.
Often mechanical and fluid properties of the formation are not accurately known, but are calculated or measured indirectly from well logs or
u.
rii
II.
DC AD CrlN r Lnndus,
CT LB
&l t-w
Analysis of the effect of in-situ stress contrast In on fracture height can be made.1,11,12 this study a pseudo 3-D model, referred to as the Table 5 shows the GULF-2 simulator was used. results of such an analysis for variations in conTtris fining stress contrast up to 1000 psi. clearly shows its importance on fracture geomeKnowledge of the mechanical properties of try. cap and base rocks which can be used to determine the in-situ stress in these zones is crucial if fracture geometry is to be modeled accurately. A sensitivity analysis may also be performed on the treatment data. For the purposes of this analysis the same prappant schedule was used since only changes in the created fracture were investiTables 6 and 7 show the effect on the gated. created fracture and maximun fracturing pressure. The ranges chosen for injection rate, power law index and consistency index reflect the ranges of treatment rates and fluid properties that could be used to carry out the treatment. These tables show that decisions made concerning the rate and type of fluid to be used are of great importance, giving the engineer who designs the treatment a reasonable degree of control. Perhaps of greater importance is an understanding of the effect of treatment parameters which are a function of the formation properties. Tables 6 and 7 show that for the range of values investigated leaked-off fluid viscosity and spurt loss had a minimal effect on the fracture geometry and treating pressure. The leaked-off fluid viscosity is defined as the viscosity of the fracturing fluid after it has leaked into the formation, and is us~~lly ~~~ q~~~ ~Q ~~~ ~~g~~~~a ~~~~ Fl,i.id ,,, viscosity. However, the wall building leakoff -------varied between 0.0001 wen and cOeTTcenl 1/ 9 0.005 ft/mln resulted in variation factors of 36% in fracture length and 46% in the job efficiency. This is of particular importance because its effect on the treating pressure over this range was only 85 psi as shown in Table 7. Unlike the case of fracture height, which was,discussed earlier as having a significant effect on both the created fracture geometry and the fracturing pressure, a fluid leakoff coefficient different to that used in the design would cause a large change in the created fracture yet would not be evident from analysis of the treatment pressure profile during injection. This is because its effect is dominated by the increase in volume of the fracture during injection. However, it should be realized that changes in the leakoff coefficient ~ff~~~ ~h~ pr~~~~r~ fallnff rlata nnt-n will areatlv =. __-.* .. 0.. ,,wthe well has been shut-in. This is because leakoff then becomes the primary process of fluid movement.
altering the treatment procedure to include preand post-stimulation tests, and improved data collection and analysis during the job. Bottomhole Treating Pressure
The analysis of bottomhole treating pressure has long been recognized as n effective method of A, 14 The bottomhole fracture interpretation. treating pressure is usually calculated from the surface treating pressure with appropriate adjustments for the hydrostatic head and tubing and perThe arlvant foration friction los~g~e ~--,m Cf ~~rn~~. terized data gathering and calculation of bottomhole pressures has, to some degree, provided the means for on-the-job analysis based on surface pressure measurement. 15 However, the analysis of treating pressure profiles obtained during this stu~ showed that calculated and measured bottomhole pressure profiles were rarely similar when injection was down tubing. Smoother profiles were observed in treatments where injection was down casing. Figure 1 shows the surface treating pressure profile obtained when fracturing well A-48. Also shown on this figure is the computer predicted pressure profile calculated by the GULF-1 simulator. The minimum in-situ stress had been measured by a minifrac to be 2720 psi and fracture height was assumed to be the perforated gross pay height. Since the sensitivity analysis presented earlier had shown that fracture height and minimum in-situ stress (S3) caused the greatest changes in pressure, and S3 had been measured, the model was rerun with different fracture heights. The r:..-..9 -ac,,l+e ~f $~e~e p~fi~ ~~e ~~aw,fiin r Iyul-e ,==,$,= c, together with the A-48 pressure data. This shows that pressures as high as 3800 psi could have been generated if the fracture height had been contained under 100 ft. However, the variations experienced in the job log indicated that unless the fracture height selectively .opened and closed during the treatment such pressure traces as in Figure 2 could not be interpreted solely on the basis of fracture height growth. Figure 3 shows the bottomhole pressure traces from treating well A-65. This well was the first of a series of wells which were treated down production tubing without a packer in the annulus, enabling both the surface treating pressure (tubing) and surface annulus pressure to be measured. The calculated tubing and annulus bottomhole pressures are shown with the conputer predicted bottomhole anwfi-mfin+ 4. FYFallnn+ f~~g~~~jng pp~~$~p~. -.s, ,s,,., -Y, ==,,-=,,,= obtained between the annulus calculated values and theoretical pressures. The bottomhole pressure calculated from the surface treating (tubing) pressure was as much as 1800 psi higher. No correlation could be determined between calculated bottomhole pressure peaks and dips antivariation in the base gel viscosity. Jobs in which bottomhole pressure bombs were run for post-job comparison with the calculated bottomhole pressure (from the surface treating pressure) showed similar results. Figure 4 shows the treatment record for well B-12, together with the calculated and -----.. ----..--4 \uuwnnule /A-..-L-l-L-L\ pressures. measureu uumD) The discrepancy between :calculated and measured bottomhole pressures is again very significant.
FIELD MEASUREMENTS The field measurement program was designed to obtain the maximum amount of information possible without incurring substantial additional costs on each well. Table 8 contains a list of the work which could be accomplished within the cost guideA. . ..AA..A . .. . ..TI...--1. ------A-A L--..,?, ~i~e~ ueLlueu Irle wurK present.eu nere WI I I Upurl. only be concerned with information obtained by
SPE 13798
inability to determine the leakoff coefficient from fracturing pressures, as discussed in the sensitivity analysis presented on Formation A. The field techniques which were used to measur ;n~ ~;;;;d;
the lnaknff
Th$$e methods have been reviewed in the literature and other researchers have found both to give credible results.
22:---
cnmff+rian+
wni-n
nrmcan+aA
-= =-=-==
hu
Nnl+a
2?
= =
Five attempts were made to analyze pressure falloff data according to Noltes technique of plotting dimensionless pressure drop against Whilst realistic leakoff dimensionless time. coefficients were calculated in some cases, in no instance was a good fit obtained with the published type curves. The reasons for this included: c Pressure rising during the shut-in period, due to heating of the injected fluid.
q
Knowledge of the minimum in-situ stress is nece ?: sary to determine the net fracturing pressure. The sensitivity analysis presented in Table 4 also shows the direct correlation between S3 and maximum treating pressure. Several authors have presented data on the use of minifracs conducted prior to the treat n stre#t,\7,M minimum in-situ obtain the The method adopted in this program was to change the job procedure to immediately shut down once fracture initiation had occurred. This Instantaneous Shut-In Pressure (ISIP) was then used for S3 4. .al.,e4.. A.A.. j.~a~ .m.+m,s.+ na.?e,,a In lwle+-+cml.,,,G,,b plcaaulc allalJala ~al.-b, dependent effects could be minimized. cases when a definite breakdown was not observed from the pressure trace a volume of 5000 gals was pumped. The well was then shut-in and pressure fal1off data were plotted against the square root of shut-in time. The inflection point observed in the plotted data was taken for Ss. Figures 6 and 7 show results of two such tests on wel1 B-30. The similar inflection points obtained after each injection period demonstrates the repeatabi1ity of this technique.
time
Pressure falloff being too slow to generate a sufficient drop in pressure during the shut-in period. The shape of the calculated curves from the shut-in data was significantly different from the type curves.
Tests made using the Nierode analysis technique consisted of pumping pre-pads of crossliked fluid ISIPS with different fluid-loss additives. obtained before and after injection were then used to calculate the total leakoff coefficient from the published eq~~tion for Perkins-Kern-Nordgren shaped fractures:
=19
*n
c=
J(tz1- tl)
[k(--)o-o
F1 uid
Leakoff
Coefficient
Table 9 1ists the treatments in which pre-pads were pumped and 1eakoff coefficients were calculated in the field. Based on the degree of agreement between the calCU1ated and design values the need for additional fluid-loss additives was The 1imited success of using this determined. technique can be seen in the outcome of treatments listed in Table 9. Fracture Nei ght Ds+termi nation
The sensitivity analysis for Formation A showed the dependence of the created fracture on the wal1 building leakoff coefficient, Cw. The total leakoff coefficient (C) is calculated in }~e GULF-1 simulator according to the relationship:
c= c~cvcw Cvcw + Cccv + Cccw
Post stimulation logs were run after all treatThese ments to determine fracture height. consisted of repeated temperature logging over a 2-3 hour period, and measurement of the increased ganma ray activity resulting from logging radioactive tracer material pumped with the proppant. Tkne iiterature contains studies of the different techn~~u$~ #ed for fracture height determinaResults obtained during this tion. * 9 -c *. .4....--J-A ----AI.. l-L_ ummmr L-**_- UT *L- *-_ -*..-_ Im me Traccure
var-ieu yr-ear. Iy.
where Cc and Cv, the compressibi1ity and fluid invasion functions, are calculated internally. Screenout which had been occurring during jobs in Formations B & C necessitated attempts to measure fluid leakoff coefficients in the field. .~ure 8 shows the computer predicted pressure plots for well C-6 using wall building lgak~ff ~~~ffi~i~n~~ The of 0.0038, 0.001 and 0.00038 ft/min 1*. closeness of the curves to one another shows the
>Luuy
was usually interpreted from the depth at which successive temperature 1ogs converged or crossed. The top of the fracture was 1ess easily identified and depended on interpreting ganmna ray intensity and changes in gradient of the temperature profile. Figure 9 shows an example of a poorly
Ful --
SPE 13798
C. M. PEARSON ET AL CONCLUSIONS The work covered in this paper showed the importance of understanding the effect of input parameters on fracture design and the fracturing pressure. A sensitivity analysis using the GULF-1 and GULF-2 fracture simulators was presented with input data from a West Texas formation. Changes in Poissons ratio, minimum in-situ stress, . . ...-....-:.. S1..:A ..4-.-..-:*.. nr.ne+,, ..,...,.SIS4141 I .!+.. 4 pulvall.y, pclmeaul [by, ICaelvull Ilulu Vl>cu>ll.y, reservoir pressure, leaked-off fluid viscosity, and spurt loss resulted in minimal variation of Fracture height predicted fracture dimensions. and minimum in-situ stress were shown to particularly affect the required fracturing pressure. Field measurements and analyses were presented which showed the importance of measuring bottomhole pressure and the results of different methods to determine minimum in-situ stress, fluid leakoff
Cnnff<r+nn+ .,=,, ,W,w,, ~~~ fp~~~~p~ hninh+ ,,=,=,1-. rhammc m,t,yGa ~~
defined fracture. Figure 10 shows an example of a clearly defined fracture height which could be correlated to bounding shale barriers identified from the base gamma iog. Ps discussed, in the work presented on bottomhole pressure measurement, fracture height could also be determined from analysis of the fracturing pressure. Figure 3 showed an excellent agreement
ha+wann ----,, tha -,,~ hn++nmhnla -V--VWWI~
r, =~-uI
nvacc,,-a
~ UIIU
anrl
+ha
QIIG WAI=AIU~=U
r.1.,,1.+cd
design pressure for a 270 ft high fracture - the gross pay height. Good agreemnts were repeatably obtained between bottomhole pressure profiles and calculated pressure data using credible fracture heights.
DISCUSSION
OF RESULTS
Oata presented from this study showed how the lDf~~d~~ign of the ~r~ated f~~~t~r~ ~~ h+nhlv ,..=...J enced by a few input parameters. No statistical analysis was attempted on the results of the sensitivity because of the complex analysis relationships between input parameters. Results of attempts to measure particularly significant input parameters were presented. The authors do not contend that the techniques used are necessarily the only or optimal methods. All too often field work which can be done to better understand the hydraulic fracturing process is limited by time and cost considerations. Results of in-situ stress measurements and postfracture analysis of bottomhole pressure profiles to obtain fracture height were found to be very satisfactory. The values obtained could be used in the design of similar treatments in the same formation. However, further investigation and understanding of leakoff behavior and the accuracy of fracture height measurement is needed. This study showed that much information can be obtained at little additional cost by revising the completion/stimulation procedure. The required measurements can be easily made if advance planning is conducted to cover:
q
+ha l,m#G
completion/stimulation procedure were suggested which enable significantly better field measurements and analysis to be made at little additional cost to the operator. NOMENCLATURE Ct = Total system compressibility (psi-l) Combined
(f~/mini/2\ . . . . ... . . . J
c= cc
=
fluid
leakoff
coefficient
Fluid loss coefficient, effects of fracturing fluid vis osity and relative !/2, permeability (ft/min Fluid loss coefficient, effects of ~;;;;ib;i;;~ :~~;i~$~
Cv
Cw
E
=
.
Fiuia 10ss coef-i ient, wail buiiaing 1,$ effects (ft/min Youngs modulus of elasticity (psi) Fracture pressure gradient, at time t (psiIft) Fracture height (ft) Pay Height (ft) Permeability (red) Consistency index (lbf.sn/ft2) Flow behavior index Reservoir pressure (psi) Spurt loss (gal/ft2) Minimum in-situ stress (ps Injection rate (bpm)
FG(t) =
Completion Procedure - Change tubing/packer arrangements to enable a direct (downhole bomb) or indirect (surface annulus) bottomhole pressure to be measured. Acid Treatment for Formation Breakdown (If Used) - Conduct low rate injection/flowback tests to determine fracture opening/closure stress. - Obtain accurate ISIP at end of job. Frac Procedure Conduct pre-frac base temperature and gamma 1Ogs. Shut-in well to obtain an ISIP after initial breakdown. If the effect of fluid-loss additives is to be investigated pump pre-pads of cross-linked fluid and monitor the changes in ISIP or pressure decay. ?--- ..-A ---* c--- A--_-_-A..-_ -. . . . 4
bunuuc~ pusc-Trac temperature surveys (at,
Hf Hp k K n P res sp
S3
= = . = = = .
=
Q= tl tz
= Time of first shut-in pres!ure (rein) = Time of second shut-in pressure (rein)
6 Pff =
PARAMETERS AFFECTING FRACTURE DESIGN AND THEIR MEASUREMENT DURING TREATMENTS Viscosity of leaked-off fracturing fluid (Cp) Viscosity of reservoir fluid (cp)
SPE 13798
11.
Prf v $
= = =
12.
Poissons ratio Formation porosity (%)
ACKNOWLEKEUENTS The authors wish to thank Gulf management for permission to publish this paper. The assistance of other Gulf personnel involved in this project acknowledged: J.R. Carlson, gratefully ~~H. Current, H. Cypher, J.C. Fair, T.H. Le, J. Munroe and K. Sahr. Comments made by B. B. McGlothlin in reviewing the original manuscript were much appreciated.
13. 14.
15.
REFERENCES
1.
2.
3.
5.
6.
7.
8.
0 .7.
10.
Daneshy, A.A., Williams, J.R., Jr., and Effect of Treatment ParaTinsley, J.M.: meters on the Geometry of a Hydraulic Fracture, paper SPE 3507 presented at the 1971 Annltal Mnat+pn ~~~ ~p~~a~s, ~C~e ~=~o SPE ..,..,-, ,,---, ,3s Lowe, D.K., McGlothlin, B.B. and Huitt, J.L.: Program A Computerized Designing for Fracturing Treatments (Horizontal Case), paper SPE 1541 presented att he 1966 SPE Annual Meeting, Dallas, Oct. 2-5. Veatch, R.W. Jr. and Crowell, R.F.: Joint Programs Research/Operations Accelerate Massive Hydraulic Fracturing Technology, J. Pet. Tech. (Oec. 1982) 2763-2775. Perkins. T.K. and Kern. L.R.: Widths of Hydraul{c Fractures, i. Pet. Tech. (Sept. 1961) 937-49. , Propagation of a Vertical Nordgren, R.P.: Hydraulic Fracture, Sot. Pet. Eng. J. (Aug. 1972) 306-14. Geertsma, J. and OeKlerk, F.: A Rapid Method of Predicting Width and Extent of Hydraulically Induced Fractures, J. Pet. Tech. (Oec. 1969) 1571-81. Lee, S.T. and 8rockenbrough, J.: A New Analytic Solution for Finite Conductivity Vertical Fractures with Real Time and Laplace Space Parameter Estimation, SPE paper 12013 presented at the 1983 SPE Annual Technical Conference and Exhibition, San Francisco, Oct. 5-8. Cleary, M.P.: Comprehensive Design Formulae for Hydraulic Fracturing, paper SPE 9259 presented at the 1980 SPE Annual Technical Conference and Exhibition, Dallas, Sept. 21-24. Aboii-Sayed, 1.S., Pearson, C.ki.ad Siiir,K.; Hydraulic Fracture Simulators Aid in Fracture Treatment Design, presented at the 1985 Computer Technology Symposium, Texas Tech. SPE, Lubbock, Texas, Feb. 28 - March 1. Rosepiler, M.H.: Determination of Principal Stresses and the Confinement of Hydraulic Fractures in Cotton Valley, paper SPE 8405 presented at the 1979 SPE Annual Technical Conference and Exhibition, Las Vegas, Sept. 23-26.
16.
Simonson, E.R., Abou-Sayed, A.S. and Clifton, Containment of Massive Hydraulic R.J.: Fractures, Sot. Pet. Eng. J. (Feb. 1978) 27-32. Thiercelin, M. and Lemanczyk, R.: The Effect of Stress Gradient on the Height of Vertical Hydraulic Fractures, paper SPE presented at 11626 1983 the SPE/OOE Low-Permeability Gas Reservoirs Symposium, Denver, March 13-16. Nolte, K.G. and Smith, M.B.: Interpretation of Fracturing Pressures,: J. Pet. Tech. (Sept. 1981) 1767-75. Nolte, K.G.: Fracture Design Considerations Based on Pressure Analysis, paper SPE 10911 presented at the 1982 SPE Cotton Valley Symposium, Tyler, Texas, May 20. Hannah, R.R., Barrington, L.J. and Lance, L.C.: The Real-Time Calculation of Accurate Bottomhole Fracturing Pressure From Surface Measurements Using Measured Pressures as a Base, paper SPE 12062 presented at the 1983 Annual Technical Conference and SPE Exhibition, San Francisco, Oct. 5-8. Abou-Sayed, A.S., Brechtel, C.E. and Clifton, R.J.: In-Situ Stress Determination by Hydrofracturing A Fracture Mechanics Approach, J. Geophysical Research (June 10,
1070! !>10) 09 C@, 8-6, 28 5 1-62.
17.
18.
19.
20.
21.
22.
23.
24.
Hydraulic Zoback, M., and Pollard, D.: Fracture Propagation and Interpretation of Pressure-Time Records for In-Situ Stress Determinations, Proc. 19th Symposium on Rock Mechanics, Stateline, NV, May 1-3, 1978, 14-22. Daneshy, A.A., Slusher, G.L., Chisholm, P.T. and Magee, D.A.: In-Situ Stress Measurements During Drilling, paper SPE 13227 presented at the 1983 SPE Annual Technical Conference and Exhibition, Houston, Sept. 16-19. and Roegiers, J.C.: How McLennan, J.D. Instantaneous Instantaneous are Shut-In Pressures?, paper SPE 11064 presented at the 1982 SPE Annual Technical Conference and Exhibition, New Orleans, Sept. 26-29. Settari, A., A New General Model of Fluid Loss in Hydraulic Fracturing, paper SPE 11625, presented at the 1983 SPE/OOE Low Reservoirs Permeability Gas Symposium, Denver, March 13-16. Determination of Fracturing Nolte, K.G.: Parameters frcnnFracturing Pressure Decline, paper SPE 8341 presented at the 1979 SPE Annual Technical Conference and Exhibition, Las Vegas, Sept. 23-26. Nierode, D.E.: Comparison of Hydraulic Fracture Design Methods to Observed Field Results, paper SPE 12059 presented at the 1983 SPE Annual Technical Conference and Exhibition, San Francisco, Oct. 5-8. Cooper, G.D., Nelson, S.G. and Schopper, M.D.: Comparison of Methods for Determining In-Situ Leakoff Rate Based on Analysis With an On-Site Computer, paper SPE 13223 presented at the 1983 SPE Annual Technical Conference and Exhibition, Houston, Sept. 16-19. Barrington, L.J., Hannah, R.R. and 8eirute, Post-Fracturing Temperature Recovery R.: and Its Implication for paper SPE 7560 presented at the 1978 SPE
Stimulation k~i~~s
SPE 13798
C. M. PEARSON ET AL
25.
26.
Annual Technical Conference and Exhibition, Houston, Oct. 1-4. T.A.: Dobkins, Improved Methods to Determine Hydraulic Fracture Height, J. Pet. Tech. (April 1981) 719-26. Wages, P.E.: Interpretation of Post-Fracture Temperature Surveys, paper SPE 11189 presented at the 1982 SPE Annual Technical Conference and Exhibition, New Orleans, Sept. 26-29.
Formation
Data
Ct ,,.
nf S3
Youngs modulus (psi) Poissons ratio Total compressibility (psi- 1) Minimum in-situ stress (psi) Fracture height (ftj
Formation
1.
2. 3. 4. 5.
Reservoir depth greater than 2500 ft. Minimum of natural fractures. Available core and log information. Measurable pre-stimulation production. Well defined shale barriers.
Permeability (red) Porosity (%) ~f Reservoir fluid viscosity (cp) Hp Pay height (ft) Pres Reservoir pressure (psi)
0.5
7 15;
700
Field
Criteria Treatment Data Under active development. Treatments should not be workovers of a previous hydraulic fracture. Fields with drilling limited to infill wells on a close spacing were not suitable. Standard treatment desiqns should reauire fracture length to heig~t ratios greater than two .
1.
2. 3. 4.
Q,
[
C:f sp
Injection rate (bpm) Fluid power index Consistency index (lbf.sn/ft2) Leakoff fluid viscosity (cp) Wallbuilding coef. (ft/mini/2) Spurt loss (gal/ft2)
20
0.52 0.13 0.001: o
TABLE EFFECT
3 ON THE
Fracture (ft) 344-392 376-380 430-352 412-382 590-317 413-360 384-367 362-387 497-336 374-393
Length
v(%)
7 1;
Width v-f%) 12 :
Efficiency v(%) (%) 60-54 56-56 67-51 63-57 29-63 63-53 57-54 53-58 81-48 55-59
3-7X106 .15-.30
40
Ct
S3 Hf k 4 prf Hp Pre.s
:: % 90 50 90 60 56
: 13 3; 9 3 2; 3
3:
7 2 1: 2
.211-.229
.238-.227 .232-.229 .228-.233 .253-.222 .230-.234
:
2 i 7 1
TA8LE 4 EFFECT OF FORMATIONLMTA 80TNMIOLE PRESSURE ON TNE MAXIM Max. Bottomhole Pressure v(%) Jpsi) 2940-3126 3051-3069 3.1 0.3 2::: 10.9
FCimiit.i
Pi3rmm?r
V(%) 40 5 :: 25 60
Length
He;;:
(ft)
Efficiency (z)
S3
l-looxlo1500-2500 100-400
3076-3053
1851-2842 3699-2969
$ ~f
Pres
90 50 90 60 56
71 :: 54 52 50
100
29
17
TA8LE 6
mm
w TREATmiT FYiRAmm
Length v(%) 7 4 3: 1
Average (in)
Width
Efficiency (%)
v(%) 14 13 71 2 12 1
v(%) 21 ; L 2
Q [ ~f Cw sp
60 1:: 95 96 100
.199 -.264 .226 -.295 .056 -.334 .230 -.238 .263 -.205 .231 -.228
base
logs
Bottomhole
~Dsi)
Pressure
v(%) 60
1:: 95
If
core
q
Q 10-40 n 0.5-0.8
K 0.001-0.5 ~f 0.5-20
c .0001-.005 sp 0-0.01
lx
Change completion
Act d cleanup:
q
change method/equipment to allow an ISIP measurement additional flowback tests determine c1 osure stress
Frac
tests evaluation
logs
Transient
q
pressure testing: pre- and post-stimulation testing: test zones individually frequently check rates and calibrate equipment
tests
Production
q q
---- - -
IABLE Y
CALCULATEDLEAKOFFCOEFFICIENTS
Wel1 B-1
ISIP (psi) 870 1230 1090 740 700 810 730 860 920 965 990 880 5255 5303 5230 5704 6463 6900 4450 4530 4650 4400 4600 4600 4415 4720 4800
Comment
(:) B-2
0.;37 0.004
Job C
ISIP decreased
o.ioo7 Job C
B-12 (i) B-18 ; (F) c-1 F([t) c-2 D(~d) c-4 (:) C-62 DF OF
0.005 0.005
0.;001
ISIP decreased
0.0001 ISIP decreased
0.;26 0.002
Job screened
out
0.;003 0.0002
0.003 0.001
Job C
C-63 D D
0.008 0.0003
Job screened
out
F = Solid fluid loss additive D = Diesel Sd = 100 mesh sand St = 100 mesh salt Parenthesis denotes fluid loss additive only pumped for part of stage
(gold)
CO*
CD
T10HVJOl109
COMPUTER
PREDICTED, PREDICTED,
Hf Hf
= =
COMPUTER
.........
.. : .. ..
..
FIELD CALCULATED Hf = DATA 270 ft.
\
\ COMPLJTER PREDICTED,
.= g
w u = u) co u fK n
900 -
. L
.. .
5 2 fc 800 3 co
.. .
0
Figure 5.
... .
10
20
TIME (min.) pressuree for verious
30
height
40
frectures - well A-1278
50
.. . . ...
Bottomhole
700
I 1
I 3
I 5
I 7
1 9
1 11
SHUT-IN Figure
I .. . . . . .. .. ...
TIME Shut.in
8.
..
...
. . .. . .
...
.....
A-
{
(u-l
0
SHUT-IN Figure
0
1 TIME 3 (rein) - SQUARE 5 7 9 11 IROOT SCALE well B30 Figure
0
Bottomhole
20 TIME
40 (rein) leakoff
80
80
7.
Shut-in
pnessure plot 2-
8.
pressures
for vartous
coefficients
- well C-8
290(
300C
6500
6550
3100
= -~ 3200 UJ n
6600
(
I
1
//f l 1 / / //
SHALE BARRIER 1
= 1 + n ~
6650
3300
BASE TEMP. \
6700
3400
SHUT-IN
-~ SHALE BARRIER
~75Q TEMP.
I
! 100
}
85 TEMPERATURE Figure 9, Fracture evaluation (F) log - well A.1274 Figure
95 ( F) B.30
log - well