R.G. Dusterhoft
Halliburton Energy Services
Duncan, Okla.
B.J. Chapman
Halliburton Energy Services
New Orleans
Hydraulic fracturing of high-permeability reservoirs has increased long-term hydrocarbon production and reduced sand production in many areas of the world.
A key element is the reduction of near well bore drawdown during production.
Drawdown, the difference between reservoir and production pressures, is the driving force for flow into the well bore. As drawdown increases because of higher production rates or depletion, formation instability may cause fines and sand to migrate into the well bore region.
A greater well bore radius reduces both radial velocity and drawdown. Fracturing beyond the well bore region effectively bypasses the damaged zone, increasing the effective radius of the well bore and enabling higher flow rates with lower drawdown pressures. In essence, the reservoir energy is used more efficiently because the conductive proppant bed bypasses the near well bore restrictions.
Except in very unconsolidated formations, maximum fracture conductivity is the most effective means to reduce formation sand production. In very unconsolidated reservoirs, sand production cannot be avoided and conventional gravel packing or packed fracture methods with smaller proppant is recommended for minimizing formation fines migration into the pack.
CANDIDATE SELECTION
Almost all wells that are candidates for gravel packing are also good candidates for high-permeability fracturing. However, wells in which gravel packs might reduce near well bore permeability are especially good candidates. Also good candidates are wells considered for new gravel packs because of mechanical failure or formation problems.
In medium-permeability formations (10-100 md) high-permeability fracturing may be beneficial because of the contrast between the propped fracture and formation. To date, most applications in the Gulf of Mexico have been in "dirty" or less permeable sands or in laminated shale sequences that can be connected with a short propped fracture.
Also, fracturing may have greater benefit in low bottom hole pressure wells in which increased drawdown cannot be tolerated.
PROPPANT SELECTION
In high-permeability reservoirs, experience shows that the most effective means to reduce sand production is to maximize fracture conductivity. Recent successful jobs have run 20/40 mesh or larger sand. Gravel-packing typically employs smaller sand sizes that do not provide the desired fracture conductivity to decrease drawdown.
Fig. 1 compares production from wells in high-permeability reservoirs. One well is gravel packed and the other is fracture-stimulated.
The productive life of the fracture-stimulated well is several times greater and approaches the ideal case. The gravel-packed well lost production earlier, but was recompleted with a fracture treatment.
SAND CONTROL
At first glance, fracturing a high-permeability reservoir to control sand migration may not seem appropriate, But a review of rock mechanics shows how and why fracturing can obtain greater production without sand.
In many high-permeability reservoirs, formation strength and stability is critical for oil and gas production and economics. Highrate production will reduce pressure in the near well bore region. This pressure drop will increase stress on the formation rock as it tries to maintain an equilibrium while other stresses such as overburden are fully exposed to the rock.
If the production rate increases further, additional pressure drawdown will add near well bore stress. If pressure is reduced low enough, the formation may fail or break because of compressive stress.
At failure onset, formation fines and sand will break free and be carried toward the well bore with produced fluids. Often these materials will accumulate in the near well bore region, effectively plugging the well.
If nothing is done to correct this problem, the damage will also increase drawdown pressure near the well bore, causing more damage and decreased production.
Mohr's failure envelope (Fig. 2) can be derived by reviewing rock property logs or by conducting mechanical testing on actual core samples. The envelope shows four failure criteria which can cause fines or sand production. 1-3
- Tensile failure-Tensile failure occurs when the effective stress at the well bore exceeds the rock's tensile strength. Failure may occur under some production conditions, but these conditions are not common. Tensile failure is more likely in shallow reservoirs where normal stresses are small, shifting the initial conditions to the left.
- Shear failure-Shear stress is introduced to the formation while drilling. Drilling mud helps prevent formation fluid production and maintains pressure to prevent formation failure. During production as the formation pressure is reduced near the well bore, additional shear stress is applied to the formation. When shear stress exceeds the rock's yield strength, fines will be generated.
- Cohesive failure-A rock's cohesive strength is best defined as the strength holding a single sand grain to the rock surface. Cohesive failure usually occurs when the drag forces caused by flowing fluids exceed the cohesive strength, breaking the particle free. Cohesive failure is especially important in poorly consolidated formations.
- Pore collapse-Pore collapse occurs when the compressive strength of the formation rock is exceeded. The rock is crushed as it fails in compression. Failure can easily occur if a well is produced at excessively high flow rates, causing extreme pressure drops in the near well bore region.
Fig. 2 shows a Mohr failure envelope for one situation, i.e., a high-permeability, poorly consolidated formation. At initial conditions the formation is stable. If production is at a controlled rate and pressure _P1, production rate can be maintained without significant sand production.
If production rate increases and pressure decreases to _P2, shear failure may begin, causing fines and sand to migrate and damage the well bore. Then, if production continues at the same rate, producing pressure will drop rapidly because of well bore damage. As pressure drops, more fines and sand particles are produced as the formation begins to fail.
This process will continue until pressure drawdown reaches the point of pore collapse. Once pores collapse, the production will drop very rapidly and the well will be effectively lost unless worked over or recompleted.
This shows that sand production can be controlled or prevented by maintaining production rate stresses within the Mohr failure en-velope.
By allowing higher producing rates at reduced drawdown, hydraulic fracturing has proven effective for stimulation and sand control. The fracture creates a high-conductivity path beyond near well bore damage and increases the effective well bore radius.
In a high-permeability well, very high-conductivity fractures are required to significantly increase effective radius. Care must be taken to ensure that drawdown pressures are within the safe regions of Mohr's circle even after the treatment. Good reservoir management is very important for controlling sand production.
MINIFRACS
A detailed minifrac analysis is another key for an effective packed fracture. To obtain maximum fracture conductivity, the design must provide a complete screenout at the end of the job. The minifrac must yield accurate values, for:
- Fracture closure pressure
- Fracture closure time
- Fluid efficiency
- Spurt volume
- Fluid-loss coefficient.
Recent fluid-loss studies show that many assumptions in classical minifrac analysis do not apply for high-permeability wells. Two major deviations are:
- Use of nonwall-building fluids
- Very large volumes of spurt loss with longer spurt times.
Spurt is usually defined as fluid volume that leaks off rapidly as the fracturing fluid first contacts the fracture face, before a filter cake forms to improve fluid-loss control.
Fig. 3 compares and contrasts minifrac results for the following cases:
- Assuming spurt equal to zero (Fig. 3a)
- Estimating fluid-loss coefficient, then iterating the program to determine a spurt value for wall-building fluids (Fig. 3b)
- Comparing response from a nonwall-building fluid with conventional minifrac analysis (Fig. 3c).
Work is in progress to refine minifrac procedures for high-permeability wells. The procedure in the box has yielded excellent results:
SPURT
Both laboratory studies and field results4 5 9 indicate spurt loss becomes a major factor in fracturing high-permeability wells. As much as 60% of the fluid loss has been observed to be due to spurt.
For conventional fracturing in low-permeability reservoirs, spurt loss is minimal because a filter cake is formed quickly. 6 8 10-13 In high-permeability reservoirs, effective filter cakes develop slowly, if at all, so that spurt loss is much greater. 4 5
Table 1 shows that spurt can become very significant when treating high-permeability wells. Spurt volumes are less predictable than the fluid-loss coefficient (Cw).
Spurt loss can be determined using a dual minifrac test as described in the procedure box.9
With most current software, fluid-loss functions capable of simulating hydroxyethylcellulose (HEC) fluids are limited. At present, a workable alternative is to use a model that includes a case for nonwall-building fluids and input a filtrate viscosity corresponding to the base gel viscosity. If such a model is not available, then a standard minifrac analysis assuming zero spurt may be more suitable. In this case, the results will be volume dependent and will only provide an approximate Cw for the greater volumes in the main treatment.
Currently, programs are being developed to include fluid-loss options for nonwall-building, power-law fluids. Such programs would vastly improve analysis of HEC minifrac jobs.
FRACTURE DESIGN
For an optimal design, the same fracture simulator should be used in the design process and the minifrac analysis.
To maximize proppant concentration inside the fracture, jobs should reach a tip screenout after which continued pumping will create additional fracture width as net pressure increases.
The maximum net pressure will vary between wells and formations. Experience in the region is required to optimize the design based on net pressure increases.
A complete screenout at the end of the treatment is desired. If necessary, the pump rate can be reduced slightly to help pack the proppant more completely into the fracture.
FLUID SELECTIONExtensive laboratory testing 4 5 has narrowed the practical choices of high-permeability reservoir water-based fracture fluids to linear gelled HEC and borate-crosslinked fluids.
Linear gelled HEC has been the fluid of choice because of its nondamaging properties. Crosslinked borate fluids are a good choice for high-permeability fracturing because they are reversible and consistently show better regained permeability and fracture conductivity than other metal-crosslinked fluids. Table 1 provides a guide to selecting these two fluid systems.
A dual-fluid system may be effective for very high-permeability wells. A crosslinked gel pad controls fluid loss and establishes the desired fracture geometry. The proppant slurry can then be placed using linear HEC, essentially filling/packing the created fracture with proppant. This method minimizes formation damage and maximizes fracture conductivity.
PROPPANT SELECTION
Experience shows that the most effective means of preventing formation sand production is to maximize fracture conductivity. To enable this, 20/40 mesh sand is the best choice, based on worldwide experience.
Gravel-pack techniques with their small sand sizes do not provide the conductivity necessary to reduce drawdown and consequent sand production. A recent trend has been to use larger proppant sizes to increase fracture conductivity. In most cases, results have been encouraging.
EQUIPMENT
Because of the many conditions encountered, a wide range of equipment is needed for hydraulic fracturing high-permeability reservoirs. In many poorly consolidated formations where sand production is anticipated, fracturing has essentially been combined with gravel packing.
Some situations allow fracturing with gravel-pack screens and tools in place. As the treatment screens out, a gravel pack is placed around the screens and all perforation tunnels connected to the fracture. If a screenout occurs with proppant still in the tubing, the proppant can be reversed out of the well without disturbing the gravel pack.
The mechanical screens stop production of formation fines, and equally important, they help stop proppant flow back from the fracture in case the formation will not prevent flow back.
In wells having multiple producing zones or long production intervals, completion in stages may be most productive. One option is to systematically conduct fracture treatments over the entire zone, maintaining isolation between treatments before completing the gravel pack. If the interval is extremely long, this operation may need to be repeated.
In harder formations gravel-pack screens may not be needed to control production of formation sand and proppant flow back. Use of curable resin-coated proppants or an overflush of resin on formation sand creates an economical, permanent, natural filter to help prevent sand production.
Fig. 4 shows a multipleposition service tool (MPST) for fracture stimulating and gravel packing a high-permeability formation in one operation. During fracturing, the annulus is closed at the surface, preventing circulation. For gravel packing, the annulus can be opened to circulate through the gravel pack screen and place sand into the annular space.
CASE HISTORIES
Three case histories, one off West Aftica and two offshore Louisiana, illustrate the effectiveness of fracturing high-permeability reservoirs.
WEST AFRICA
A major producer in West Africa combined fracture stimulation and sand control techniques to reduce skin and optimize production in a new oil field development.
Two producing interxals were completed to evaluate the packed fracture technique with HEC linear gel and 20/40 synthetic proppant. After both fracture treatments were performed, a gravel pack was completed over the entire interval. Production results were good and the skin factor was below 2.0. These results prompted the remainder of the project to use these techniques.
In the second well, borate-crosslinked fluid was tried. This fluid controlled fluid loss better in the zones with the highest permeability, leading to the decision to use dual fluids in high-permeability zones. In these cases the pad to establish fracture geometry included a borate-crosslinked hydroxypropylguar (HPG). The proppant was then placed with a linear-gelled HEC to ensure maximum fracture conductivity. This procedure was effective to place larger proppant volumes.
The packed fractures reduced the effective skin and increase production while providing excellent sand control.
OFFSHORE LOUISIANA
A well offshore Louisiana was completed at 6,000 ft in a relatively low-permeability (10-50 md) panr zone with about 100 ft of gross pay. After perforating 12 shots/ft, the zone was acidized, and a step rate test determined fluid efficiency of about 45%. Closure pressure was 4,000 psi. With this information, a screenout design was planned using a pseudo 3D simulator.
The job was pumped with a 5,000 psi increase in bottom hole pressure indicating a screenout with proppant packed to the well bore. A volume of 33,000 lb of 20/40 synthetic proppant was placed, ramping sand concentration from 0 to 12 ppg.
The well was gravel packed, and a post-cleanup test showed a negative skin factor. This is unusual for a gravel-packed well in the Gulf area and indicates success in bypassing near well bore damage.
Another offshore Louisiana well was completed at about 10,000 ft in a high-permeability (500-1,000 md) pay zone with about 50 ft of gross pay. The zone was perforated at 12 shots/ft underbalanced followed by a step rate test and a minifrac to provide input for treatment design. The job was designed using a pseudo 3D simulator.
In a borate-crosslinked gel mixed at 40 lb/1,000 gal, 100,000 lb of 20/40 synthetic proppant was placed, ramping from 0 to 12 ppg concentration.
The well produced over 10 MMcfd at very low drawdown, unusual for gravelpacked wells in the area.
REFERENCES
- Abass, H.H., Hedayatti, S., and Meadows, D.L., "Non-Planar Fracture Propagation From a Horizontal Well-Experimental Study," Paper SPE 34823, SPE Annual Tech. Conference, Washington D.C., Oct. 4-7, 1992.
- Terzagbi, K., From Theory to Practice in Soil Mechanics, John Wiley, 1960.
- Biot, M.A., "Theory of Elasticity and Consolidation for a Porous Anisotropic Solid," Journal Applied Physics, Vol. 26, No. 2, February 1955.
- Parker, M.A., Vitthal, S., Rahami, A.B., McGowen, J.M., and Martch, W.E., "Hydraulic Fracturing of High-permeability Formations to Overcome Damage," Paper SPE 27378, SPE Symposium for Formation Damage Control, Lafayette, La., Feb. 7-10.
- McGowen, J.M., Vitthal, S., Parker, M.A., Rahimi, A.B., and Martch, W.E-, "Fluid Selection for Fracturing High Permeability Formations," Paper No. SPE 26599, SPE Annual Technical Conference and Exhibition, Houston, Oct. 4-6, 1993.
- Nolte, K.G., "Determination of Fracture Parameters from Fracturing Pressure Decline," Paper SPE 8341, SPE Annual Technical Conference and Exhibition, Las Vegas, Sept. 23-26, 1979.
- Nolte, K.G., "Determination of Proppant and Fluid Schedules from Fracturing Pressure Decline," Paper No. SPE 13278, SPE Annual Technical Conference and Exhibition, Houston, Sept. 16-19, 1984.
- Tan, H.C., McGowen, J.M., Lee, W.S., and Soliman, M.Y., "Field Application of Minifrac Analysis to Improve Fracture Treatment Design," Paper No. SPE 17463, SPE California Regional Meeting, Long Beach, Mar. 23-25, 1988.
- McMechan, D.E., "Fracturing Fluid Loss at High : Laboratory Tests and Field Applications," Paper No. SPE 21873, SPE Rockv Mountain Regional Meeting and Low-Permeability Reservoirs Symposium, Denver, April 1991.
- Lee, W.S., "Study of the Effects of Fluid Rheology on Minifrac Analysis," Paper No. SPE 16916, SPE Annual Technical Conference and Exhibition, Dallas, Sept. 17-20, 1987.
- Lee, W.S., "Fracture Propagation Theory, and Pressure Decline Analysis with Langrangian Formulation for Penny Shaped and Perkins and Kerns Geometry Models," Paper No. SPE 17151, SPE Formation Damage Control Symposium, Bakersfield, Calif., Feb. 8-9, 1988.
- Nolte, K.C., "Fracture Design Consideration Based on Pressure Analysis," Paper No. SPE 10911, SPE Cotton Valley Symposium, Tyler, Tex., May, 20, 1982.
- Lee, W.S., "New Analysis Technique Provides Effective FluidLoss Coefficients," Paper No. SPE 24062, SPE Western Regional Meeting, Balersfield Calif., Mar. 30-Apr. 1, 1992.
Copyright 1994 Oil & Gas Journal. All Rights Reserved.
