Академический Документы
Профессиональный Документы
Культура Документы
of the wells. Such a treatment also takes a significant amount calls for a second, more realistic approach, which is
of time due to the low pumping rate through the coil tubing. developed for the first time in the present paper.
Most of the treatments performed nowadays hence make We consider that the horizontal well, in its initial state, is
use of viscous pills as diverter stages. A particular case are composed of two sections: a short, clean one (where damage
foams, which build up their viscosity within the rock itself.8, 9 has been mechanically removed by the pressure drawdown
Treatments diverted with foams are rather difficult to design when the well has been put into production), and the long rest
due to the lack of knowledge about their rheology. On the of the well with variable degrees of damage extent. For the
other hand, pills of gels, made of either polymers or sake of simplicity, we address the case of a well with no
surfactant micelles, have well known properties (unlike stand-off in a large, isotropic reservoir. Its dimensionless
foams, they are not compressible). Intuitively, the diverting pressure when fully open to flow is:13
efficiency should increase with the viscosity of the pill,
exactly as does the efficiency of a fluid in hydraulic pD =ln(4 re / L) + (h / L) ln(h / 2π rw).............................(1)
fracturing. Crosslinked gels would be preferred, however they
are either unstable in presence of acid (borate or zirconate We assume that the damaged section of the well does not
crosslinkers) or activated by temperature, which leaves them communicate at all with the reservoir. The dimensionless
in a linear state outside hydraulic fractures (titanate). For pressure of the damaged well is conveniently obtained (and
treatments in carbonate reservoirs, there is fortunately one slightly overestimated) by replacing (L) in equation (1) by (l),
gelled acid composition which crosslinks by reaction with the the length of the open section. The initial skin of the well,
rock within a limited range of pH and shows superior (Sl), is similar to the Brons and Marting14 pseudoskin for
performance as a diverter.10 partial completion of a vertical well, and is equal to the
The present paper addresses the general case of viscous difference of the dimensionless pressures:
pills made of either linear polymeric gels or micellar systems,
where a design methodology can be developed. Sl= ln(L / l) + [L / l − 1] (h / L) ln(h / 2π rw)..................(2)
Estimating the required pressure drop for efficient The associated pressure drop (∆Pl) is of course:
diversion by a viscous fluid bank around openhole
horizontal wells ∆Pl= [ µo B Q / 2π k h ] Sl............................................(3)
A viscous fluid bank around a section of a well should
manage diversion when the pressure drop through this bank When a viscous fluid bank exists around a wellbore of
exceeds the pressure drop through the damage across another length (l), it induces a pressure drop (∆Pb) equal to:
section. Estimating the initial skin impairing the production
of the well therefore gives direct access to the required ∆Pb=[ µo B Q / 2π k l ] Sb............................................(4)
pressure drop through the bank built by the first diverter
stage. Further estimates when acid dissolves the damage, The bank will efficiently divert the treatment fluids for
making additional diverter stages necessary, would require a any ∆Pb > ∆Pl - for equal values, the fluid is equally shared
perfect knowledge of the distribution of the damage along the between (l) and (L − l) - so its pseudoskin must obey:
well and of acid efficiency. The design methodology is
therefore limited to the definition of the volume of the first Sb ≥ (l / h) Sl =[1 − l/L] ln(h/2π rw) − (l / h) ln(l /L)....(5)
diverter stage. Later stages, which are required to ensure full
coverage of the wellbore by the treatment fluids, are When (l / L) → 1 (clean well), Sb → 0 (there is no need
conservatively assumed to have an equal volume and to for a treatment, hence no need for diversion), and when (l / L)
blanket an equal length of the openhole. → 0 (highly damaged well), Sb → ln(h/2π rw), which
There are two ways of interpreting the skin value of a generally takes values in between 3 and 5. (Sb) is maximum
horizontal wells. The conventional approach considers a for wells where the open length is 1/4 to 1/3 of the total well
homogeneous damage all along the well, as did Hawkins11 in length, and should be larger than 10, 20 or 40 when (L / h) is
the vertical well case: the skin is a simple function of the respectively equal to 20, 50 and 100.
severity (permeability impairment) and extension of the Naturally fissured reservoirs. In the most general case of
damaged zone. Similar functions have been derived for a matrix rock of negligible permeability (e.g. the vast majority
horizontal wells.12, 13 The big drawback of this approach is of naturally fractured carbonates), a system of conjugated
that diversion is not required: the acid should complete the fissures of width (w) and spacing (δ) is at the origin of the
removal of the damage at the same time all along the well, permeability of the reservoir and is equal to w3/12δ, as shown
since this damage is homogeneous. This fundamental by Muskat.15 All the fissures intercepted by a horizontal well
contradiction between the conventional reservoir engineering are initially invaded by the drilling mud. When put into
description of the skin and its highly heterogeneous nature production, the well generally flows from only one of the
SPE 38201 MATRIX TREATMENT IN HORIZONTAL OPENHOLE WELLS: DESIGN OF VISCOUS DIVERTER SLUGS... 3
fissures.16 Its skin is similar to (Sl) in equation (2), with L = N (γp) is the Newtonian shear rate at the wellbore within a
δ and l = n δ (and n most often equal to 1): porous medium.
The fluid bank pseudoskin is estimated from a generalized
Sf =ln(N / n) + [N / n − 1] (h / N δ) ln(h / 2π rw)............(6) form of Hawkins’ equation:11
Viscous fluid flow around openhole horizontal wells Sb, p =1 ⁄ R [αp R (1 − n’) / µo − 1] dR/R............................(13)
and fluid bank pseudoskin estimate
Power-law fluids are considered. The validity of this approach Sb, p =[R (1 − n’) − 1] αp / (1 − n’ )µo − ln(R).................(14)
will have to be checked case by case, depending on the nature
of the diverting gel and on the shear rate. For instance, When n’ → 1, [R (1 − n’) − 1] / (1 − n’ ) → ln(R) and αp →
polyacrylamide solutions show a stable viscosity below ten K’, therefore Sb, p → [K’ / µo − 1] ln(R) which is the correct
and above ten thousands reciprocal seconds,17 with a Power- expression for a Newtonian fluid of viscosity (K’) displacing
law behavior in between. Shear rate equations are provided in the reservoir fluid of viscosity (µo). An other interesting limit
the next paragraph. In the case of flow into a natural fracture, case of equation (14) is for a very shallow invasion with a
the shear rate might be large enough for the gel to be in the bank thickness b << rw:
Cross or Carreau behavior range.18, 19 Rheology tests would
have to be performed and pseudo n’ and K’ coefficients Sb, p (b)≅ (αp / µo − 1) b.............................................(15)
derived within the adapted range of shear rate.
Porous medium case. The apparent viscosity of a Power-law Equation (15) readily shows that maximizing the bank
fluid in a porous medium is derived, as usual, from the skin (and therefore the diverter efficiency) is achieved when
combination of its expression in a pipe of diameter (D)20 and (αp) is made as large as possible. Equation (11) infers that
of Kozeny’s model of a bundle of capillaries:21 this is obtained with a gel of large (n’) and (K’) (both) under
low shear conditions. Equation (12) shows that low shear
µ=(8u / D)n’−1 [(3n’ + 1) / 4 n’]n’ K’............................(7) means low (Q) into large √(kφ) / Cc rocks. Optimum diversion
with viscous fluid banks therefore requires low pumping
D =8 √ (k / 2φ) and u = U /φ...................................(8) rates, which is opposite the usual way acid jobs are performed
in horizontal wells.6
µ=[U √2 / √(kφ)]n’−1 [(3n’ + 1) / 4 n’]n’ K’.........…........(9) Natural fractures case. The apparent viscosity of a Power-
law fluid in a fracture of width (w) is:20
This expression of the apparent viscosity of a Power-law
fluid as a function of its macroscopic (Darcy) velocity (U) in a µ=(6u / w)n’−1 [(2n’ + 1) / 3 n’]n’ K’..........................(16)
porous medium of permeability (k) and porosity (φ) has been
generalized by Cannela et al.22 to take into account tortuosity If (Q) is the constant injection rate into one single
effects23 and the variability of the effective pore radius to the fracture, then u = Q / 2π w r at the distance r < δ of the axis
fluid rheology. 24 The √2 factor next to the velocity in equation of the well. The viscosity at any dimensionless distance R < δ
(9) is replaced by a (Cc) factor called the shear rate / rw is:
conversion constant:
µ(R)= αf R (1−n’) with αf = γf n’−1[(2n’+1)/3n’]n’ K’.......(17)
[ (3n’+ 1) / (1 − n’) ]
Cc=√2 ε λ ...........................................(10)
and γf= 3Q / (π w 2 rw ).............................................(18)
This factor reduces to √2 when both (ε) and (λ) equal 1,
and when the shear rate is large (e.g. more than five hundreds (γf) is the Newtonian shear rate at the wellbore, at the inlet
reciprocal seconds for a solution of xanthan gum25). of a natural fracture.
The velocity (U) of radial flow around an openhole of The fluid bank pseudoskin is:
length (l) is simply U = Q / 2π l r at the distance (r) of the
axis of the well. Assuming a constant injection rate (Q), the Sb, f=[R (1 − n’) − 1] αf / (1 − n’ )µo − ln(R).................(19)
viscosity at any dimensionless distance R = r / rw of the well
is: Conclusions similar to those made in the porous medium
case can be drawn, though (R) will always be much larger
µ(R)= αp R (1−n’) with αp = γpn’−1[(3n’+1)/4n’]n’ K’......(11) than 1. Once again it is recommended to pump at low rate,
and diverter gels will be more efficient in thick fissures where
and γp= Q Cc / [2π l rw √(kφ)]....................................(12) the shear rate is minimized.
4 OLIVIER LIETARD SPE 38201
Equations (20) and (12) giving (γp) are combined and the Gel selection guidelines
shear rate is estimated based on the following considerations: Three factors affect the selection of the type of gel which will
i) µo = 1 cp and B = 1; ii) the maximum ∆P corresponds to be used to mix the diverter slugs:
about 80% of the difference between fracturing pressure and i) Cost, which relates to volume and therefore to the fluid
pore pressure and is in the range of 2,000 psi at a vertical efficiency at building up pseudoskin.
depth of 8,000 feet; iii) rocks of permeability from 1 to 3,000 ii) Potential post-job residual damage for polymer-based gels,
milliDarcy and porosity from 10 to 30% give √(k/φ) from 10-7 a direct function of the polymer content which also affects the
to 3×10-6 meter. The range of shear rates is found to be 500 to efficiency of this type of systems. Despite the incorporation of
50,000 sec-1. live breaker (generally ammonium persulfate), the porous
Natural fractures case. The maximum shear rate is found to medium can be permanently damaged by polymer residues
be about a hundred times larger than in the porous medium (natural fractures are not).
case, for common natural fractures of width in the 100 to iii) Impairment of the post-job production by unbroken gels: it
1,000 microns range.16 These large shear rates are only is critical to ensure that the diverter slug viscosity returns to a
experienced at the inlet of the fracture. Because of the deep value low enough for the slug to be back-produced. A residual
invasion and radial flow they rapidly decrease within the pseudoskin is no better than the initial damage skin.
fracture to about ten times the porous medium case values.
Example: selection of a surfactant-based gel in a
Designing an acid treatment with viscous slugs as porous medium application
diverting agents In this example, a micellar system routinely used as a gravel
The design methodology is composed of the logical pack carrier fluid is studied (assuming it is stable in contact
succession of the following steps: with acid). The gel is a 3-% ammonium chloride solution
i) Test the well and define (kh) and (Sl) or (Sf). Evaluate (k) containing up to 5% of a surfactant forming rod-shape
and (φ) in primary porosity rocks, (δ) from mud loss events16 micelles above 1%. The larger the amount of surfactant, the
and (w) from Muskat’s permeability equation 15 in naturally larger the number of micelles and the larger the viscosity. At
fractured reservoirs. (L) or (N) being known, derive (l) from 100 °F, the rheological parameters of this system are as
(Sl) or (n) from (Sf). Check that (n) is equal to 1; if not, do not follows:
forget to later on divide the surface injection rate by (n).
ii) Evaluate the required pseudoskin for efficient diversion Surfactant % n’ K’ cp @ 170 s-1
(Sb) from equation (5).
iii) Select a treatment rate (Q) and calculate (γp) or (γf) from 1 0.60 0.0024 15
equations (12) or (18) respectively (admit Cc = √2). 2 0.48 0.0079 26
iv) Select a gel as the diverter and derive from its rheological 3 0.40 0.0251 55
parameters the coefficients (αp) or (αf) from equations (11) or 4 0.38 0.0472 94
(17) respectively. Note that (n’) and (K’) should be valid in 5 0.37 0.0695 131
the range of Newtonian shear rates at the wellbore defined
above. Selecting the right composition means maximizing the
v) Calculate the required (R) from equations (14) or (19). pseudoskin at an equal cost, i.e. at an equal total amount of
vi) Determine the fluid bank volume as follows: surfactant. In the following tables, we have summarized for
four different shear rates the fluid bank pseudoskins provided
Vp= π rw2 φ l (R2 − 1)...................................................(21) by the five versions of the considered micellar system at equal
dimensionless surfactant volume (VD,s):
Vf =π rw2 w n (R2 − 1)..................................................(22) VD,s= Vs / π rw2 φ l = (R2 − 1) Cs.................................(23)
SPE 38201 MATRIX TREATMENT IN HORIZONTAL OPENHOLE WELLS: DESIGN OF VISCOUS DIVERTER SLUGS... 5
∆Pl = initial damage pressure drop, Pa 9. Ali, S.A., Bui, H.N., and Edwards, M.B.: “Acid Diversion
ε = Cannella’s pore geometry ratio, dimensionless is Critical in Horizontal Gas Well Treatments,” Petroleum
φ = porosity, dimensionless Engineer International, April 1996, pp. 32-34.
γf = Newtonian wellbore shear rate, natural fissures, s-1 10. Saxon, A., Chariag, B, and Reda Abdel Rahman, M.: “An
γp = Newtonian wellbore shear rate, porous medium, s-1 Effective Matrix Diversion Technique for Carbonate Forma-
λ = Cannella’s pore rheology ratio, dimensionless tions,” paper SPE 37734 presented at the 1997 SPE Middle
µ = apparent viscosity in pipe or slot, Pa.s East Oil Show, Bahrain, March 15-18.
µ(R) = apparent bank viscosity at distance (R), Pa.s 11. Craft, B.C., and Hawkins, M.F., Jr.: “Applied Petroleum
µo = reservoir fluid viscosity, Pa.s Reservoir Engineering,” Prentice-Hall, Inc., Englewoods
Cliffs, New Jersey, 1959.
12. Renard, G., and Dupuy, J.G.: “Influence of
Acknowledgments Formation Damage on the Flow Efficiency of Horizontal
The author thanks the management of Schlumberger Dowell Wells,” paper SPE 19414 presented at the 1990 Formation
for permission to publish. The author’s gratitude goes to John Damage Control Symposium, Lafayette, Louisiana, Feb. 22-
Sudderth and Peter Montague of Dowell in Aberdeen for their 23.
support at the early times of the present study.
13. Lietard, O., Ayoub, J., and Pearson, A.: “Hydraulic
Fractu-ring of Horizontal Wells: An Update of Design and
References
Execution Guidelines,” paper SPE 37122 presented at the
1. Rappuhn, T., Munch, E., and Schuhbauer, W.: “Formation
1996 Hori-zontal Well Technology Conference, Calgary,
Damage Prevention - Completion of the Most Productive
Alberta, Nov. 18-20.
Hori-zontal Well in Germany in the Mittelplate Offshore Oil
14. Brons, F., and Marting, V.E.: “The Effect of
Field,” paper SPE 31138 presented at the 1996 SPE
Restricted Fluid Entry on Well Productivity,” Journal of
Formation Dama-ge Control Symposium, Lafayette,
Petroleum Technology, February 1961, pp. 172-174.
Louisiana, Feb. 14-15.
15. Muskat, M. : “Physical Principles of Oil Production,”
2. Goode, P.A., and Wilkinson, D.J.: “Inflow Performance of
McGraw-Hill Book Co. Inc., New York, 1949, page 246.
Partially Open Horizontal Wells,” paper SPE 19341 presented
16. Lietard, O., Unwin, T., Guillot, D., and Hodder, M.:
at the 1989 SPE Eastern Regional Meeting, Morgantown,
“Fracture Width LWD and Drilling Mud / LCM Selection
West Virginia, Oct. 24-27.
Guidelines in Naturally Fractured Reservoirs,” paper SPE
3. Sognesand, S., Skotner, P., and Hauge, J.: “Use of Partial
36832 presented at the 1996 SPE EUROPEC, Milan, Italy,
Perforations in Oseberg Horizontal Wells,” paper SPE 28569
Oct. 22-24.
presented at the 1994 SPE EUROPEC, London, England,
17. Boger, D.V., Journal of Non-Newtonian Fluid
Oct. 25-27.
Mechanics, volume 3, pp. 87-91 (1977).
4. Gilchrist, J.M., Stephen, A.D., and Lietard, O.: “Use of
18. Cross, M.M., Journal of Colloid Science, volume 20,
High-Angle, Acid-Fractured Wells on the Machar Field
pp. 417-437 (1965).
Development,” paper SPE 28917 presented at the 1994 SPE
19. Carreau, P.J., Transactions of the Society of
EUROPEC, London, England, Oct. 25-27.
Rheology, volume 26, pp. 99-127 (1972).
5. Lietard, O., Bellarby, J., and Holcomb, D.: “Design, Exe-
20. Guillot, D.: “Digest of Rheological Equations,” Appendix
cution and Evaluation of Acid Treatments of Naturally
1 of “Well Cementing,” Nelson E.B. editor, Schlumberger
Fractured Carbonate, Oil Reservoirs of the North Sea,” paper
Educational Services, SMP-7031 (1990).
SPE 30411 presented at the Offshore Europe 1995, Aberdeen,
21. Greenkorn, R.A.: “Flow Phenomena in Porous Media,”
Scotland, Sept. 5-8.
M. Dekker Inc., 1983, pp. 53-55.
6. Tambini, M.: “An Effective Matrix Stimulation Technique
22. Cannella, W.C., Huh, C., and Seright, R.S.: “Prediction
for Horizontal Wells,” paper SPE 24993 presented at the
of Xanthan Rheology in Porous Media,” paper SPE 18089
1992 SPE EUROPEC, Cannes, France, Nov. 16-18.
presented at the 1988 SPE Annual Technical Conference and
7. Economides, M.J., Ben-Naceur, K., and Klem, R.:
Exhibition, Houston, Texas, Oct. 2-5.
“Matrix Stimulation Method for Horizontal Wells,” paper
23. Teeuw, D., and Hesselink, F.: “Power-Law Flow and
SPE 19719 presented at the 1989 SPE Annual Technical
Hydrodynamic Behavior of Polymer Solutions in Porous
Conference and Exhibition, San Antonio, Texas, Oct. 8-11.
Media,” paper SPE 8982 presented at the 1980 Annual
8. Simpson, J.G., Edmunds, A.C., Bittner, T., and Muir,
International Symposium on Oilfield and Geothermal
L.: “Performance Optimization of Horizontal Wells: A
Chemistry, Stanford, California, May 28-30.
Compa-ny’s Experience With Horizontal Well Acid
24. Wreath, D., Pope, G.A., and Sepehrnoori, K.:
Stimulations,” paper SPE 37120 presented at the 1996
“Dependen-ce of Polymer Apparent Viscosity on the
Horizontal Well Technology Conference, Calgary, Alberta,
Permeable Media and Flow Conditions,” In Situ (1990),
Nov. 18-20.
volume 14(3), pp. 263-284.
SPE 38201 MATRIX TREATMENT IN HORIZONTAL OPENHOLE WELLS: DESIGN OF VISCOUS DIVERTER SLUGS... 7