SPE 39477

A Coupled Wellbore-Hydraulic Fracture Simulator for Rigorous Analysis of Frac-Pack Applications

Terri Sherlock-Willis, SPE, Dowell, Jean Romero, SPE, Dowell, and Surya Rajan, SPE, Dowell
Copyright 1998, Society of Petroleum Engineers, Inc. This paper was prepared for presentation at the 1998 SPE International Symposium on Formation Damage Control held in Lafayette, Louisiana, 1819 February 1998. This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

due to a deliberate reduction of the injection rate or the shifting of the tool to the circulating position. This paper shows a field example using this coupled simulator to evaluate a frac pack treatment that was terminated due to a premature screen out. The calibration test is initially analyzed following which the frac pack treatment is simulated and compared with field measurements. Introduction Gravel packing has been the standard completion technique for sand control treatment since the 1930s. A slurry of appropriately sized gravel in a carrier fluid is pumped into the annular space between a centralized screen and a perforated cased hole or the open hole sand face, to reduce the production of failed formation sand particles. The gravel pack treatment can be pumped with the tool in either the squeeze or circulating position. When the tool is in the squeeze position (Figure 1), the fluid leaks off entirely into the formation; however, when the tool is in the circulating positions, the fluid leaks off into the formation and filters through the screen as returns are taken up the washpipe back to the surface. Either mechanism results in slurry dehydration and leads to the gravel packing in the annulus. The gravel pack acts as a granular filter of high permeability that prevents formation sand from entering the well. Once in place, the gravel pack treatment creates a pressure drop (i.e. skin) which may affect productivity. Therefore, to achieve maximum long-term productivity in cased hole completions, gravel must be tightly packed in the perforation tunnels, as well as in the screen- casing annulus. Fracturing has been traditionally used in low permeability reservoirs to create long fractures that provide a high conductivity flow path for the reservoir fluid. Recently, fracturing has been used in conjunction with the gravel pack technique as a means to bypass any near wellbore damage and

Abstract Frac pack stimulations have been used extensively to provide a combined treatment for well stimulation and sand control. It has been observed that a premature wellbore screen out often causes treatment failure in such operations. However, until now the design and evaluation of such treatments have been performed solely using fracture simulators that neglect the effect of the wellbore hardware (crossover ports, blank, screen, washpipe, etc.) on the treatment. In an effort to address this limitation, a wellbore simulator was coupled with a fracture simulator to provide the capabilities of predicting, and subsequently eliminating, such occurrences. The fracture simulator calculates the fracturing parameters such as proppant distribution in the fracture, fracture height, and two-dimensional fluid flow. These parameters provide the boundary conditions for the wellbore simulator. Additionally, the fracture simulator was enhanced to correctly simulate fluid leakoff and tip screen out designs in high permeability formations. The fluid and proppant flow around the screen is calculated using a pseudo three-dimensional wellbore simulator that provides a complete design and evaluation tool of frac pack treatment. The effects of a deviated wellbore, proppant bridging around the screen, proppant settling in the wellbore and fluid flow through the screen are taken into account. Additionally, the coupled simulator models an induced wellbore screenout at the end of the treatment either

TERRI SHERLOCK-WILLIS, JEAN ROMERO, AND SURYA RAJAN

SPE 39477

control sand production. A fracture is created using a tip screenout (TSO) technique1,2; if properly designed and executed, the TSO treatment results in proppant packing from the fracture tip back to the wellbore and from the wellbore into the annulus between the screen and the fracture. Thereby providing a combined well stimulation and sand control completion in a single operation. The first frac pack treatment using the tip screen out technique was pumped in the Gulf of Mexico in the early 1990s3, 4, since then this technique has gained increasing popularity in unconsolidated formations worldwide. The steps in a successful frac pack treatment can be summarized as follows: 1. The fracture is propagated to its desired length just as the slurry begins to dehydrate near the tip of the fracture; 2. Additional slurry is pumped and the net pressure in the fracture increases, causing it to widen; 3. Once the proppant is packed in the fracture, the rate maybe reduced forcing the fracture to close and the wellbore to screen out. A major difference between frac pack and the various types of gravel pack is that a successful frac pack involves the placement of high volume of proppant per unit area in the fracture, thereby significantly reducing the overall completion skin. This allows for linear flow of the hydrocarbons into the wellbore with very little drawdown. The advantage of the frac pack method over gravel pack completion is that hydrocarbon production is enhanced while sand production is controlled. The success of a frac pack treatment depends on candidate selection, treatment design, fluid and proppant selection, the type of downhole tools used, and treatment implementation. A Coupled Simulator In parallel to the development of sand control completion techniques, several numerical simulators have been developed to help design such treatments. The conventional gravel pack completion can be modeled with a wellbore simulator which takes into account fluid leakoff, gravel transport and tool geometry. Several pseudo three-dimensional simulators have been developed based on these assumptions,5,6,7 a fully threedimensional model was introduced by Klein et al.8 However, the design and evaluation of frac pack completions have usually been carried out solely with fracture simulators.8 Such an approach neglects the effect of the tool geometry and setting (Figure 1) during the placement of the proppant in the fracture and around the screen. It has been frequently observed that treatment failure can also be attributed to node buildup or proppant bridging in the wellbore, which eventually leads to a premature screen out. In order to investigate the possible interaction between the fluid displacement in the wellbore and the fluid flow in the fracture, a coupled simulator has been developed. The

coupled simulator has been derived from two existing simulators: a fracture simulator (Appendix A) and a wellbore simulator (Appendix B). Both of these simulators have been widely used in field design and evaluation for several years and their accuracy has been validated against industry benchmark field examples. The coupled wellbore - hydraulic fracture simulator was designed to model all flow paths performed during a frac pack treatment that includes squeeze and circulation. When the tool is set in the squeeze position, the fracturing simulator calculates the bottom-hole pressure and the flow boundary conditions are based on the fracture width at the wellbore. The boundary conditions have been modified in the existing wellbore simulator to handle the high leakoff (i.e. fluid flow) into the fracture. Depending on the fracture width profile at the wellbore, the leakoff flux is calculated based on the fracture width at that position. The fluid flux, q is calculated based on the average velocity which is derived from total flowrate: q= Q ...................................... H

(1) Where, Q is the flowrate in one fracture wing and H is the fracture height at the wellbore. The flux at a depth, z, is related to the average flux and width as follows: w( z ) q(z) = q ............................... w (2) Where, w is the average width and w (z) is the width at depth z. When the tool is set to the circulation position, multiple flow paths exist; the fluid is able to leakoff to the formation and/or through the screens. The leakoff is calculated based on the formation permeability or on the fluid flux into the fracture, in the case where a fracture has been created. For the design and evaluation of a frac pack treatment the simulator models the following: 1. With the tool in the circulating position, the treatment is circulated to spot the pad to the top of the perforations; 2. Once the pad is at the top perforation, the tool is set to the squeeze position and a hydraulic fracture is initiated; 3. The fracture is then propagated until the desired length is obtained and a tip screen out is induced; 4. The proppant is packed from the fracture tip to the wellbore. For the duration of pumping the fracture treatment, the simulator tracks gravel settling in the wellbore. 5. The simulator balances the flow of fluid between the screen- casing annulus and screen- washpipe annulus with respect to the pressure drop; 6. Once the fracture is packed, the wellbore pack can be

A COUPLED WELLBORE-HYDRAULIC SIMULATOR FOR RIGOROUS ANALYSIS OF FRAC-PACK APPLICATIONS

SPE 39477

pumped with the tool in the squeeze or circulating position at any specific rate. A premature screen out can occur during the treatment due to the tool geometry and alternate fluid flow paths, since the fluid has the ability to travel inside or outside the screen. Such a screen out can be modeled by the simulator and will cause the fracture propagation or packing to be stopped. An actual field treatment that was terminated early due to a premature screenout was identified as a case study for validation of the coupled simulator during its field-testing phase. The results from that validation study are presented later in this paper. In addition to this example, a test case was defined to show the ability of the simulator to predict an early screen out as a function of radial annular clearance around the screen and the clearance between the washpipe and the screen. Frac Pack Design and Evaluation One of the most important steps to ensure a successful completion in high permeability formation is the correct evaluation of the calibration treatment. Recent work by Gulrajani et al.9 demonstrated the benefit of the use of a consistent methodology based on pressure history matching. In addition to this study, current design procedures and guidelines for frac pack treatment that have been presented in previous papers2, provide the tools for evaluating the case study to be presented in this paper. A pressure history match was first obtained for the calibration treatment, and subsequently results of the frac pack treatment were predicted using the coupled wellbore-hydraulic fracture simulator. Test Case It has been shown with laboratory experiments that a premature screen out can occur in the wellbore at the top of the screen due to node buildup or proppant bridging. Indeed the fluid will always follow the path of least resistance. If the pressure drop in the screen annulus is higher than the pressure drop in the washpipe annulus the fluid will leakoff through the screen which causes a high concentration of proppant to form nodes on the outside of the screen (Figure 2). This high proppant concentration on the outside of the screen could then lead to subsequent premature termination of the treatment. However, such a flow pattern will only occur when the screen annular clearance is small in comparison to the washpipe annular clearance. Annular clearance plays an important role in the successful execution of the frac pack treatment. With a low annular clearance between the screen and the casing, bridging becomes a risk due to dehydrating slurry forming at the top of the screens. Screen sizing is generally dictated by the casing I.D. and the location of the zone. A radial screen clearance of greater than or equal to 1.00 in. is recommended (based on field practices) to help maintain sand control and provide optimum fishing conditions for subsequent removal. It has

also been proposed that the washpipe O.D. be ideally 80% of the screen base pipe I.D10.; whenever applicable, it is recommended to use the largest size wash pipe that is practical. The reason for the 0.8 ratio is to promote slurry movement on the outside of the screen as it travels down toward the bottom of the wellbore. The path of least resistance must be on the outside of the screen since the fluid has the ability to travel inside or outside the screen; the larger washpipe O.D. minimizes alternate flow path inside the screen. With an undersized washpipe, node buildup of dehydrated slurry will form at the top of the screen causing a proppant bridging risk. Using the recommended dimensions described previously, a test case was defined to confirm that the simulator will predict flow diversion; this would result in proppant bridging as a function of the annular clearance around the screen, in addition to the ratio of the washpipe O.D. to screen I.D. The test case was performed on a deviated wellbore of 60 using a 30-lb/mgal borate cross-linked fluid. Four specific scenarios for the radial clearance around the screen were created; for each of these, the ratio of the washpipe O.D. to screen I.D. was varied twice. It was observed that the simulator predicted treatment termination due to proppant bridging earlier for those cases of smaller radial annular clearances around the screen that had correspondingly smaller washpipe O.D. to screen I.D. ratios (Table 1). This concludes that the smaller the annular clearance around the screen, the greater the pressure drop which causes the fluid to be diverted to the washpipe annulus. It should be noted that these results will vary based on the fluid and proppant type, formation properties, fracture model and wellbore deviation. The results do not serve as a guideline but more so as an indication of the simulators capabilities. Case Study The field example presented here is a frac pack treatment that was performed for Ocean Energy at the Eugene Island field in the Gulf of Mexico. The formation is an oil-bearing reservoir with moderately high permeability through out the zone of interest. The treatment was terminated early due to a premature screen out. A post-job evaluation was performed on this field treatment to validate the newly developed coupled simulator, and simultaneously determine the cause of the early screen out. A brief description of the calibration test is provided first, followed by an evaluation of the calibration treatment. Finally an evaluation of the frac pack treatment by predicting bottom-hole pressures and history matching with retrieved bottom-hole gauge data is presented. Both the calibration test and the propped treatment were injected using a delayed 30-lb/mgal borate cross-linked fluid at a steady rate of 15 bpm. For the propped treatment, 20/40

TERRI SHERLOCK-WILLIS, JEAN ROMERO, AND SURYA RAJAN

SPE 39477

frac sand was injected using a ramped schedule ranging from 0.5 ppa to 12 ppa; the treatment was terminated prematurely while injecting the 6 ppa stage into the perforations. The following post-job evaluation was performed: Closure Determination and Pressure Decline Analysis: The log indicates the perforated interval that ranges from 12,023 ft to 12,087 ft (see Figure 3). The zone of interest was highly laminated with moderate shale barriers at either side of the perforated interval. Closure pressure was determined by analyzing the falloff pressure that was measured during shut-in. The square-root of time plot and G-function plot were both used for comparison to determine closure; closure pressure was determined to be approximately 8000 psi which corresponded to a relatively low net pressure of 500 psi; and a fluid efficiency of 1% was also determined. Calibration Analysis: Based on the closure pressure and fluid efficiency determined from the falloff, a history match was performed. The calibration treatment was simulated using the hydraulic fracture simulator, and a two-dimensional radial fracture geometry model was used based on the slope of the pressure decline observed during the minifrac. The match of the simulated data versus measured data corresponded to the formation characteristics - toughness of 5000 psi.in and a Youngs modulus of 3 X 105 psi. Good agreement is seen between the simulated and measured pressure (Figure 4). Frac Pack Treatment Predictions: Based on the parameters determined from the calibration treatment and using the injected proppant schedule, the coupled simulator was used to predict bottom-hole pressure for comparison with the bottomhole pressure measured from the actual treatment. It was determined that the low efficiency of the treatment fluid was the result of the premature screen out. Upon injecting the first stages of the proppant into the formation, there was a noticeable increase in net pressure with no further apparent propagation of the fracture. The low fluid efficiency resulted in the fracture and subsequently the wellbore, packing off earlier than expected. Less than 65% of the treatment was placed below the crossover of the tool. Good agreement was observed between the simulators predicted versus measured bottom hole pressure (Figure 5 and 6). Conclusion The field case presented in this paper represents the first attempt to evaluate premature screen out in frac pack applications using a coupled wellbore - hydraulic fracture simulator. The coupled simulator provides a better understanding of the fluid flow and mechanics associated with completion tools. The simulator proposed here can be used routinely to help design and evaluate frac pack treatments; it is capable of predicting a premature screen out due to diversion of flow to the screen- washpipe annulus depending on the pressure drop in the wellbore. Furthermore,

the final proppant concentration in the fracture and around the wellbore is calculated by the coupled simulator, this allows for the use of the simulator to determine the effect on proppant packing in the wellbore as a result of reducing the flowrate or shifting the tool to the circulating position at the end of the treatment. Nomenclature A = fracture cross section area, M2 C = proppant volumetric concentration, dimensionless Cv = filtrate leakoff coefficient, L/T Cc = reservoir leakoff coefficient, L/T Cvc = combined filtrate/reservoir leakoff, L/T d = proppant diameter, L E* = plane strain Youngs modulus, M/LT2 f = friction coefficient H = fracture height, L h = zone interval height, L g = gravitational acceleration, M/T2 KI = fracture toughness, M/L1/2T2 k = rock permeability, L2 p = fracture pressure, M/LT2 Q = total injection rate, L3/T q = injection flux, L2/T qleak = leakoff velocity, L/T Re = Reynolds number, dimensionless t = time, T V = fluid velocity, L/T w = fracture width, M = fluid density, M/L3 = viscosity, MT/L2 = minimum in-situ stress, M/LT2 Acknowledgments The authors would like to thank Sunil Gulrajani and Hugo Morales for their input and informative discussions during the course of this project. Appreciation is also expressed to Schlumberger Dowell and Ocean Energy for their permission to publish this paper. References
1. Smith, M.B., Miller, W.K., and Haga, J.: Tip Screenout Fracturing: A Technique for Soft, Unstable Formations, paper SPE 13273 presented at the 1984 SPE Annual Technical Conference and Exhibition, Houston, Sep. 16-19. Morales, R.H., Norman, W.D., Syed, A., Stewart, B.R., WorksMinter, A.M., and Syed, A.A.: Current Practices in the Tip Screenout Fracturing of High Permeability Reservoirs in the Gulf of Mexico, paper SPE 38578 presented at the 1997 SPE Annual Technical Conference and Exhibition, San Antonio, Oct. 5-8. Hannah, R.R., Park, E.I. and Walsh, R. E., Porter, D.A., Black, J.W. and Waters, F.: A Field Study of a Combination Fracturing/Gravel Packing Completion Technique on the

2.

3.

A COUPLED WELLBORE-HYDRAULIC SIMULATOR FOR RIGOROUS ANALYSIS OF FRAC-PACK APPLICATIONS

SPE 39477

4.

5.

6.

7.

8.

9.

10.

11.

12.

13. 14.

15.

16. 17. 18.

Amberjack Mississippi Canyon 109 Field, paper SPE 26562 presented at the 1993 SPE Annual Technical Conference and Exhibition, Houston, TX. Grubert, D.M.: Evolution of a Hybrid Frac-Gravel Pack Completion: Monopod Platform, Trading Bay Field, Cook, Inlet, Alaska, paper SPE 19401 presented at the 1990 SPE Formation Damage Control Symposium, Louisiana, Feb. 22-23, 1990. Winterfeld, P.H., and Schroeder, D.E.: Numerical Simulator of Gravel Packing, SPE 19753, paper SPE 19752 presented at the 1989 SPE Annual Technical Conference and Exhibition, San Antonio, October 8-11. Nguyen, P.D., Fitzpatrick, H.J., Woodbrige, C.A., and Reindenbach, V.G.: Analysis of Gravel Packing Using a 3D Numerical Simulation, paper SPE 23792 presented at the 1992 SPE Formation Damage Control Symposium, Lafayette, Feb. 26-27. Wahlmeier, M A., and Andrews P. W.: Mechanics of Gravel Placement and Packing: A Design and Evaluation Approach, SPEPE, pp 69-82, Feb. 1988. Klein, H. H., Cooper, G.D., and Nelson, S.G..: Design and Evaluation of Gravel and Frac Packs With a Fully Three dimensional Wellbore Simulator, paper SPE 35032 presented at the 1995 SPE Formation Damage Control Symposium, Lafayette, Feb. 14-15. Tibbles, R.J., Gulrajani, S.N., and Nolte, K.G.: Evaluation of Calibration Treatments, for Frac Pack Completion, in Offshore Nigeria, paper SPE 38192 presented at the 1997 SPE European Formation Damage Conference, Hague, June 2-3. Penberthy, Jr. W.L., Shaughnessy, C.M.: Gravel-Pack Placement, Chap. 8, SPE series on Special Topics, Vol. 1, Sand Control, 1992. Mack, M.G., Piggott, A.R., and Elbel, J.L.: Numerical Representation of Multilayer Hydraulic Fracturing, Rock Mechanics, Proceedings of the 33rd U.S. Symposium, pp 335 344, 1992. Fung, R.L., Vijayakumar, S., and Cormack, D.E.: Calculation of Vertical Fracture Containment in Layered Formations, SPEFE, pp 518-522, Dec. 1987. Dodge, D. and Metzner, A.: Turbulent Flow of NonNewtonian Systems, AIChE J. 5, No 2, pp 189-204, 1959. Keck, R.G., Nehmer, W.L., and Strumolo, G.S.: A New Method for Predicting Friction Pressures and Rheology of Proppant-Laden Fracturing Fluids, SPEPE, pp 21-28, Feb. 1992. Swanson, V.F.: The Development of Formula for Direct Determination of Free Settling Velocity of Any Size Particle, Transaction of the American Institute of Mining, Metallurgical and Petroleum Engineering Inc., Vol. 238, pp160-166. Gruesbec, W. M.: Design of gravel packs in Deviated Wellbores, JPT, pp 109-115, Jan. 1979. Williams, B.B.: Acidizing Fundamentals Monograph Series, SPE, Dallas, 1979. Parlar, M., Nelson, E.B., Walton, I.C., Park, E., and DeBonis V.: An Experimental Study on Fluid-Loss Behavior of Fracturing Fluids and Formation Damage in High-Permeability Porous Media, paper SPE 30458 presented at the 1995 SPE Annual Technical Conference and Exhibition, Dallas, Oct. 2225.

19. Settari, A.: A New General Model of Fluid Loss in Hydraulic Fracturing, SPEJ, pp 491-501, Aug. 1995.

Appendix A Hydraulic Fracture Simulator11 In its full generality, the fracturing simulator used in this study incorporates both propped and acid fracturing treatments, and accommodates multiple fracturing fluids with time-varying rheology and temperature variations in the fracturing fluid and rock mass. The model is cell-based and incorporates fracture height-growth using a coupled set of equations implicitly relating the pressure, width and height at each cell. Fluid loss is determined by considering the individual contributions of the fluid filter-cake, filtrate viscosity and the reservoir components to the leakoff behavior. In addition, proppant transport is predicted using a rigorous formulation considering fluid pressure gradients along the fracture length and height. Within this formulation, vertical flow is assumed to occur sufficiently slowly so that pressure variation due to fluid flow in this direction can be neglected. The in-situ stress and toughness profile are assumed to be piece-wise constant, denoted as i and I in the i-th layer. The total fracture height is denoted as h, and the layer geometry is defined by quantities hi , which represent the height from the lower tip to the top of the i-th layer ( Figure A-1). By invoking the previously stated assumptions, fracture height, pressure, pcp, and the fracture cross section area, in a cross-section are related using12; At the upper tip (in the nth layer),
K Iu = + h 3 [ p n + f g ( hcp h)] 2 cp 4 2 h

(
i =1

n 1

i +1

h 2hi h i )[ cos 1 ( ) hi ( h hi ) ] 2 h

(A-1)

and at the lower tip (in the 1st layer),

K Il = h 1 [ pcp n + f g (hcp h)] 2 4

2 + h

(
i =1

n 1

i +1

h 2hi h i )[ cos 1 ( ) + hi ( h hi ) ] 2 h

(A-2)

and the cross-sectional area A,

A= h2 ( p + g(hcp z) n ) 2 E * cp

2 * E

n 1

h h2 h 2hi ( i +1 i )[( hi ) hi (h hi ) cos1 ( )] 2 4 h i =1

(A-3)

Here, KIu and KIl are the fracture stress intensity factor of the layers containing the upper and lower tips of the fracture, the density of the fluid, g gravitational acceleration, and E* the plane-strain Youngs modulus. However, if rapid fracture height growth is expected to occur, an alternative formulation based on the vertical fluid

TERRI SHERLOCK-WILLIS, JEAN ROMERO, AND SURYA RAJAN

SPE 39477

pressure drop in the fracture is used to determine the fracture geometry. This solution is based on vertical pressure drop for the propagation velocity when predicting fracture height growth. The simulator does not incorporate any additional mechanics, such as multiple fractures and tip-dilatancy effects, in order to predict fracture behavior. The continuity equation in the fracture is given by dw + . wV = 2qleak . . . . . . . . . . . . . . . . . . . . . . . . . . . (Adt 4) where w is the fracture width, qleak the leakoff velocity, and V the fluid flow velocity. The pressure drop along the fracture (momentum) is defined by: A V n = n (P / 2 K )( ) 1+n . . . . . . . . . . . . . . . . . . . . . . . . (Ah 5) where n is the fluid power-law, and K the fluid consistency coefficient. The shape factor is defined as: =
1+ n H n 1+ 2 n A n

dC + CV = 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . (Adt
8) where V is the fluid velocity and C the particle concentration. The fluid pressure drop along the wellbore is calculated based either on flowrate-pressure drop table or on correlation the developed by Dodge and Metzner et al.13 An analytical solution of pressure drop for laminar and for turbulent flow is used based on the following expression: 1 4 0.4 = 0.75 Log (Re f 1n '/2 ) 1.2 . . . . . . . . . . . . . (An' n' f 9) where Re is the Reynold number. When gravel is present in the fluid, the friction loss is corrected by a factor proportional to the gravel concentration. The resultant friction factor is:

f s = f f gravel . . . . . . . . . . . . . . . . . . . . . . . . . . . .(A10) where the gravel friction multiplier is: f gravel = r0.55 r0.45 . . . . . . . . . . . . . . . . . . . . . . . . . . (A11) where r is the relative viscosity and r is the relative density. As defined by Keck et al.14 If the wellbore is packed the pressure drop is related to the fluid velocity using the standard Darcys flow equation. The wellbore simulator calculates gravel settling velocity by first finding the settling velocity of a single particle and then multiplying the result by a correction factor that adjust the effects of concentration on settling behavior. Swanson et al.15 define the terminal settling velocity, vs vs = 4 gd ( s f ) 3 e . . . . . . . . . . . . . . . . . . . . . . . .(A-

n 2 + 4n

1+ 2 n w( z ) n

dz . . . . . . . . . . . . . . . . . . . .(A-

6) An implicit scheme is used to solve the conservation equations (mass and momentum) and rock elasticity. In this implicit scheme the fracture tip velocity is guessed and then corrected based on volume balance calculated from the fracture tip to the well. The proppant transport is calculated using the conservation of mass in the fracture:

dCw + CwV = 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . .(Adt

7) Proppant concentration is evaluated explicitly in flux terms and a flux corrected transport scheme is used to limit numerical diffusion. The fracturing simulator is also capable of predicting fracture growth on-site and has real-time pressure matching capabilities. Appendix B Wellbore Simulator The wellbore simulator has been developed based on a pseudo-3D mathematical model. The mathematical model takes into account variations of wellbore deviations, and tool geometry. Independently, the wellbore simulator is capable of handling gravel packing in open hole as well as cased/perforated completions. The current implementation of the coupled simulator, however, does not handle open hole completions. The basic equation for describing particulate transport in the wellbore is:

12) where g is the gravitational acceleration, d is the proppant diameter, g the gravel density, and e the effective viscosity. Gruesbec et al.16 observed that a fluids velocity must be greater than a critical value, the equilibrium velocity, in order for particles to be transported by the fluid. The equilibrium velocity can be defined as the minimum velocity a fluid must have to move suspended particles in the direction of its flow. If the fluid velocity is less than the equilibrium velocity, the gravel velocity is simply the calculated settling velocity. However, if the fluid velocity is greater than the equilibrium velocity, the gravel velocity is the resultant of the settling velocity and the fluid velocity. Appendix C Non-Wall Building Filtrate Controlled Leakoff The reservoir leakoff expression that has been derived by

A COUPLED WELLBORE-HYDRAULIC SIMULATOR FOR RIGOROUS ANALYSIS OF FRAC-PACK APPLICATIONS

SPE 39477

Williams17 for Newtonian filtrates requires an input for the filtrate viscosity. While this parameter can be readily obtained for fracturing fluids used in low permeability reservoirs, it cannot be easily estimated for power law fluids in high permeability formations due to significant gel invasion and the resulting non-unique shear rate. For nonNewtonian filtrate controlled leakoff in high permeability formation, we used an approximate solution derived by for the calculation of Cvc. Since fracturing fluids are shear-thinning with no cake buildup, it is necessary to derive an expression for the combined equivalent leakoff coefficient, Cvc, including the reservoir-fluid compressibility effects. In this derivation, we consider an incompressible, power-law filtrate displacing a slightly compressible, Newtonian reservoir fluid. The method involves consideration of two independent regions: a viscous region and a compressible region. The derivation of the leakoff velocity is based on pressure drop in each region. It involves an approximation of the time dependence, tn/(n+1), of the true leakoff volume for the viscous region with t1/2 using least squares method, but keeping the expression relating velocity and pressure drop as is, and consideration of resistance in series for the total pressure drop. The resulting expression is18: 1
2 CV n+ CVC 1 +

1 C 1 = 0 . . . . . . . . . . . . . . . . . . . . . . (ACC VC

13) which is a nonlinear, algebraic equation that can be solved by standard root-finding algorithm. The coefficients Cc and Cv have been presented in a previous publication19. The total leakoff, Cvc is then a function of time and its variation is shown in Figure A-2 for typical reservoir conditions.

TERRI SHERLOCK-WILLIS, JEAN ROMERO, AND SURYA RAJAN

SPE 39477

Figure 3: Log section for the stimulated interval

9200

Bottom Hole Pressure (psi)

9000

Field Data Simulated

8800

8600

8400

8200

8000 0 2 4 6 8 10 12

Time (min)

Figure 1: Tool set in squeeze position for frac pack treatment.

Figure 4: Minifrac bottom hole pressure comparison.

16000 30.0 BH Prop Conc. (ppa)/ Pump Rate (bpm)

Bottom Hole Pressure (psi)

Node buildup due to wellbore dehydration

14000 25.0 12000 20.0 10000 8000 6000 4000 2000 0 0.0 0 200 400 600 800 1000 1200 1400 Field Data Simulated Series4 Series5 15.0

Tip Screeenout

10.0

5.0

Rw Rs Rc

Time (min)

Figure 2: Wellbore and hydraulic fracture interaction during frac pack.

Figure 5: Frac pack bottom hole pressure comparison.

A COUPLED WELLBORE-HYDRAULIC SIMULATOR FOR RIGOROUS ANALYSIS OF FRAC-PACK APPLICATIONS

SPE 39477

10000

Field Data Simulated

Casing Clearance (in) 3.0

1000

1.5 0.75 0.5

Washpipe ID/ Screen OD 0.82 0.50 0.82 0.50 0.82 0.50 0.82 0.50

Completion of Treatment (%) 94.5 93.0 90.0 88.0 80.0 75.0 65.0 40.0

Net Pressure (psi)

Table 1: Test Case Results.

100 0.1 1 10 100

Time (min)

Figure 6: Frac pack net pressure comparison.

Equivalent Carter Leakoff

Figure A-1: Fracture height schematic.

Figure A-2: Leakoff of a Non-wall building fluid.