Eduardo R. Blanco
BJ Services
Houston
While stimulation of a horizontal well does not differ significantly from that of its vertical counterpart, the design of a horizontal fracturing operation must be custom developed on a well-to-well basis.
Like a vertical well, a horizontal well can be stimulated either by using matrix treatments for removing damage or by hydraulic fracturing to overcome problems of low permeability.
The characteristics of horizontal wells require careful planning of the design and execution of such stimulations. Initial well planning should consider stimulation because the completion design will need to incorporate the high-pressure pumping required to create usually more than one fracture plane in the horizontal segment.
This conclusion to an eight-part series, begun in OGJ on Sept. 24, presents guidelines that can aid in designing fracturing treatments in horizontal wells. Emphasis is placed on the need for a better interaction among the different disciplines, e.g., rock mechanics, reservoir fluids, pre and postfrac reservoir behavior, etc., to optimize the design and application of this technique in horizontal wells.
Accurate production statistics for horizontal fractured wells will allow for better history matching with mathematical model predictions. This, in turn, will help improve the methodology for determining the optimum number of fracture planes in the horizontal segment needed to economically maximize the production from a multifractured well.
More extensive field applications with correct evaluations to verify laboratory hypotheses will generate a better understanding of the different factors that affect the propagation of fractures in horizontal wells. As a corollary, the mechanics of orthogonal multifracture induction in the horizontal segment could be enhanced.
DESIGN CRITERIA
The design criteria for a horizontal well will differ somewhat from the approach used to fracture a vertical well because the conditions of fluid movement in the reservoir are greatly affected by the geometry of a horizontal segment. Generally, the job design will consider the following aspects:
- Reservoir properties
- Rock mechanics
- Material transport
- Completion and treatment methodology.
RESERVOIR PROPERTIES
The effectiveness of a horizontal well compared to a vertical well is determined from a comparison of productivity and economic concepts. This comparison is based on the theoretical productivity increases that must justify the additional cost of drilling and completing a horizontal well.
From a reservoir engineering standpoint, a horizontal well can best be compared to an infinite conductivity fracture whose vertical development is limited to the open hole diameter of the horizontal segment. This analogy identifies the following parameters to define the success, or failure, of a horizontal well:
- Permeability anisotropies
- Reservoir thickness
- Proximity to a gas or water contact.
PERMEABILITY ANISOTROPIES
A method to calculate the productivity index (PI) of a horizontal well was developed in 1988.1 The equation presented can be combined with the PI of a vertical well to yield Equation 1, shown in the equation box.
An analysis of this relationship shows that a high vertical permeability will enhance the productivity of a horizontal well. Conversely, a low vertical permeability will result in a less attractive PI compared to that of a vertical well (Fig. 1).
The thickness of the formation also has a direct effect on the productivity ratio. The greater the thickness, the lesser the production increase of a horizontal well over a vertical well (Fig. 2).
GAS/WATER CONTACT
Using accepted and conventional approaches to calculate the critical rate to prevent premature gas or Water influx, a maximum permissible drawdown can be calculated for a vertical well.
Applying this permissible drawdown to the aforementioned equation in Reference 1 for the productivity of a horizontal well, it can be determined that a horizontal well bore will allow higher production rates.
However, it must be remembered that the same restrictions and limitations existing in fracturing vertical wells with a gas or water contact also apply to the horizontal case.
PRODUCTIVITY OF FRACTURES
If a horizontal well can be compared to an infinite conductivity fracture, then a reasonable productivity evaluation can be made by comparing it to a vertical well intersected by a finite conductivity fracture.
The PIs of both cases can be correlated to arrive at Equation 6, proposed in 1988.2
Therefore, a horizontal section can be replaced by an hydraulically fractured vertical well depending on the dimensionless conductivity (Fig. 3).2
Analysis of equations representing the PI of horizontal wells reveals that horizontal wells tend to outperform vertical wells that are not fractured when drilled in:
- Reservoirs with limited thicknesses.
- Reservoirs whose anisotrophy values kH/kv are favorable to vertical flow, such as naturally fissured reservoirs whose kH/kv value is close to unity.
- Reservoirs with potential gas or water coning.
Generally, however, it is very rare when an hydraulically fractured vertical well can be replaced by a horizontal well.
Therefore, in those reservoirs that are candidates for hydraulic fracturing, horizontal wells will require some form of stimulation to attain the productivity increases that will justify this type of completion.
FRACTURED HORIZONTAL WELLS
Depending on the orientation of the horizontal segment as related to the azimuth of the minimum principal stress (Sigma 3), the hydraulic fracture created from this segment of the well will either be longitudinal or orthogonal to its axis.
If the fracture is longitudinal, the productivity increase by definition will not be significant.
Using the same correlation previously discussed, it can be determined that longitudinal fractures will increase the productivity of the horizontal well when dimensionless fracture conductivity is low or when the ratio between the horizontal segment and the formation thickness (L/h) is high (Fig. 4).2
When the created fractures are orthogonal, the inefficient contact between the fracture and the well bore will create a "pseudoskin" effect. The behavior of orthogonal fractures need to be simulated as "choked" fracture behavior. Joshi suggests this effect can be represented by Equation 7.
This "pseudoskin" effect (Sch)c results in an additional pressure drop between the fracture and the well bore that directly affects the production rate. Despite this adverse effect, orthogonal fractures have two significant advantages:
- The possibility of creating more than one fracture plane.
- The access to a larger drainage area.
ROCK MECHANICS
An understanding and correct interpretation of the mechanical properties of the rock in question are very important to successfully create hydraulic fractures in horizontal wells. The following factors need to be considered:
- Fracture orientation
- The aspect of the fracture
- Well bore/fracture intersection
- Multiple fracture generation
- Presence of natural fissures.
FRACTURE ORIENTATION
Fracture orientation depends on the orientation of the minimum principal stress. Assuming a simple three-axial component system x, y, and z, where two stresses will be horizontal (x, y) and one stress vertical (z), the fractures will grow perpendicularly to the minimum stress and will have a horizontal orientation if the minimum stress is z.
By the same token, the fractures will be vertical if the orientation of the minimum stress follows the direction of the x or y axis. Generally, fractures created from today's horizontal wells will be vertical since no horizontal fractures are likely below 1,000 ft.
The azimuth of the fractures depends, therefore, on the azimuth of the minimum stress. To orient the horizontal segment of the well favorably, it is necessary to identify the azimuth of the minimum stress.
Specialized rock mechanics studies in cores can be used to identify the orientation, azimuth, and magnitude of the three main stresses. The results, however, will depend on whether the core samples are representative of the reservoir.
For determining the magnitude and the azimuth of the principal stresses, available also are in situ techniques such as microfracturing that can be more trustworthy than laboratory analyses.3
If nondiscriminating perforation patterns exist despite the orientation of the horizontal segment, fractures will tend to grow longitudinally to the fracture's axis. Later the fracture direction can change to create tortuous and narrow paths that make passage of the proppant-laden fluid more difficult.
Therefore, it is important to induce and maintain the orientation of fracture planes from the well bore with a notch or a cluster of perforations.
In vertical wells, the majority of the fractures will orient themselves along the perforations. In horizontal wells, the fractures can be developed both longitudinally or orthogonally to the horizontal segment. If orthogonal, the intersection aspect between the well bore and the fracture becomes extremely important.
WELL/FRACTURE INTERSECTION
Assuming both the fracture and horizontal segment azimuths are known, the intersecting length between the two can be determined. Equation 8 has been proposed.4
From Equation 8, an 8 5/8-in. horizontal well bore and an intersection angle (A) of 45, will have an intersection length between the well bore and the fracture of 12.2 in. However, the effective length over the well bore axis will only be 0.72 ft. This is where the greatest number of perforations should be localized.
Over these small sections, the required high perforation densities can be difficult to achieve with present mechanical means. Unfortunately, practical and conventional perforation densities will result in the introduction of a "pseudoskin" during production, and in high-friction pressures when pumping stimulation treatments.
To avoid these occurrences, one desirable alternative is semicircular notching, since it will increase considerably the contact area between the fracture and the casing (Figs. 5 and 6).
SIMULTANEOUS FRACTURES
The feasibility of creating orthogonal fractures may lead to the consideration of initiating several fracture planes simultaneously. But this practice would generate a temporary alteration in the magnitude of the principal stresses that would limit the number of fractures that could be propagated simultaneously.
Additionally, overpressure in the fractures could create a large enough alteration of the principal stresses to force the development of a longitudinal fracture along the axis of the well bore.
This fracture would continue to grow in this direction and eventually turn or flip where the stresses have not been affected.
The possibility of altering the in situ stresses to modify the azimuth of hydraulic fractures has been shown both analytically and through field experiences.5
The concentration and alteration of stresses can force the fractures to orient themselves in a direction different from the preferential plane, more than likely overcoming a secondary stress. As soon as the cause or the effect of this stress alteration ceases or dissipates, the fracture turns to the direction or the preferred fracture plane.
SELECTIVE FRACTURES
To create multiple fractures in a horizontal well, the horizontal segment must be cased, cemented, and perforated selectively and adequately,
The effect of installing and cementing a production casing will introduce a "pseudoskin" factor. This factor is increased when the perforations in the casing are localized only in the circumference of contact with the fracture plane.
To overcome this "skin" problem, a minimum number of orthogonal fractures will be required to restore the productivity of the open hole.
Mukherjee, et al.2 proposed Equation 9 to calculate the minimum number of infinite conductivity fractures required to replace the lost productivity of the open hole.
Results of the application of similar comparative correlations can be observed in Fig. 7.6 An analysis of these results shows at least three to four fractures are required to restore the productivity of the open hole.
The creation of more than one producing fracture plane will cause interference among fractures. To compensate for this effect, an additional number of fractures should be induced in parallel planes (Fig. 8).7
FRACTURE DENSITY
The number of orthogonal fracture planes that can be independently created in the horizontal segment depends on adequate bonding of the casing to maintain their orientation.
On this basis, the optimum density of fracture planes can be determined by calculating the number of necessary fractures to restore open hole productivity without interference, then optimizing that number by means of economic simulations.
NATURAL FISSURES
Because one of the fundamental reasons for drilling a horizontal well is the presence of a high ratio of vertical-to-horizontal permeability, it must be assumed that in those wells, subject to hydraulic fracturing, some secondary permeability system exists to improve this ratio.
These secondary systems usually consist of lattices of natural fissures whose presence will enhance the productivity of the nonfractured and fractured well. Also, these secondary systems will affect the normal development and growth of hydraulically induced fractures.
When a lattice of natural fissures is intersected by an hydraulically induced fracture, high fluid losses occur. This results in a sharp decrease in the efficiency of the treatment fluid.
This lowered efficiency will affect fracture width and may cause premature bridging of the propping agent. Eventually, this will degenerate in undesirable screen-outs.
MATERIALS TRANSPORT
Because of the limited clearance between the well bore and the tubulars in the horizontal segment, materials transport must be carefully studied when fracturing a horizontal well. For this reason, a strict control of the settling velocity of the solid particles is imperative to assure its successful transportation in the treating fluid.
Settling velocity is directly proportional to particle size and density, and inversely proportional to the viscosity of the carrier fluid and its crosslinked characteristics.
To avoid premature bridging and screen-outs, it is desirable to set a minimum fracture width equal to at least 2.6 times the maximum diameter of the propping agent.
Earlier experiments determined that solids transported in fracturing fluids will bridge in the orifices of the perforations providing their diameter is less than six times the average diameter of the carried particles.7 Bridging in the perforations can cause early screen-outs and also reduce the effectiveness of diverting agents.
PERFORATING
Hydraulic fractures are created to increase the effective radius of the well bore or to overcome a damaged zone in formations where matrix treatments may not be suitable. Perforation of the casing as part of the completion procedures must consider one of these two fracturing criteria.
The creation of a longitudinal or an orthogonal fracture in the horizontal segment also must be taken into account to plan for casing perforation. The common factors for good perforating practices in these cases are high shot density, deep penetration, and maximum perforation diameter.
The initiation of a fracture plane by means of a semicircular notch is especially desirable when it is necessary to create more than one fracture plane in a controlled manner.
For the development of a longitudinal fracture, a helical perforation distribution along the axis of the horizontal segment is preferable.
LIMITED ENTRY
Limited entry has been successfully applied to create more than one simultaneous fracture in vertical wells; however, using this technique in horizontal wells has its disadvantages. These unfavorable conditions are mainly related to frictional effects caused by fluid flow through the reduced number of perforations required to achieve a proportional distribution of rate.
The resulting exaggerated friction pressures, plus the additional pressure differential required to attain limited entry, substantially increase surface pressure. This pressure increase associated with the technique's generally higher pumping rates leads to greater horsepower requirements.
Furthermore, the restricted number of casing perforations causes a higher degradation of the shear-sensitive fracturing fluids whose viscosity loss is undesirable and could ultimately become critical when fracturing horizontal wells.
Another undesirable aspect of the limited-entry method is the creation of a very limited contact area between the borehole and the fracture because of the reduced number of perforations in the casing for each fracture plane
PROPPANT PLACEMENT
Most two-dimensional or three-dimensional models can be used to simulate fracture geometry and proppant placement for job design. When orthogonal fracture planes are developed, the solution becomes very simple as each plane can be treated as an independent fracturing job.
When fracturing longitudinally to the horizontal segment axis, the simulation needs to be approached as if the fracture would originate at the base of the horizontal segment, thereby creating two symmetrical wings.
One of the wings would follow the horizontal segment borehole, while the other would be developed opposite.
The conductivity of both wings would be different and the fracture half-length needs to be limited to the length of the horizontal segment.
FRACTURING FLUIDS
The success of fracturing treatments in horizontal wells depends largely on the quality and characteristics of the fracturing fluids. In horizontal well fracturing, the fluid is subject to greater shear stresses, and its viscosity plays an important role in materials transport, especially along the critical horizontal segment.
These conditions favor fracturing fluids that tend to develop terminal viscosity rapidly as a result of the immediate action of the crosslinking agent. Because some of these fluids present higher shear sensitivity than others, shear conditions should be minimized to preserve their integrity.
Additionally, it is necessary to schedule the gel rupture so that the propping agent can be maintained in suspension up to the moment the fracture is closed. Gel residue should be minimized to preserve fracture conductivity.
FLUID-LOSS ADDITIVES
Keeping in mind that hydraulic fracturing in horizontal wells will occur in the vast majority of cases within reservoirs possessing some type of secondary porosity systems, it is necessary to minimize high fluid losses expected in the presence of such networks.
One of the most common additives to control leakoff in the presence of natural fissures is 100-mesh sand. In horizontal well fracturing, 100-mesh sand can be transported easily, will effectively obstruct the passage to the fissures, and will still retain enough conductivity at the end of the treatment to allow reservoir fluids to flow unrestricted through these sandfilled channels.
The use of 100-mesh sand will increase fluid efficiency. Its use, however, should be limited to the pad stage. This avoids a possible reduction in fracture conductivity caused by intrusion of lesser-diameter particles in the proppant pore throats when the sand is used in the treatment stage.
If necessary, micellar and granular additives can be used to further reduce fluid loss by taking advantage of wall buildup or two-phase-flow mechanics.
Because some of these additives can affect the final conductivity of the fracture, it is advisable to carefully select them.
Ultra-low fluid loss is not desirable in horizontal well fracturing because it will delay the fracture closure time. Delays create the potential for proppant deposition in the critical horizontal segment.
PROPPING AGENTS
The production capacity of fractured wells is in direct relationship to the conductivity of the created fracture and, thus to the permeability and conductivity of the propping agent. While in conventional fracturing 12-20 and 20-40 mesh proppants have been used widely for their conductivity characteristics, smaller-diameter propping agents like 16-30 and 40-70 mesh offer some advantages when fracturing horizontal wells.
Smaller-mesh proppants have lower settling velocities and they better resist the formation of bridges in the primary fracture, and allow higher concentrations in the carrying fluid.
Considering the use of finer-mesh proppants will result in lower fracture conductivity. This could be improved by using low-density artificial ceramic propping agents. Their near perfect roundness and sphericity yield high conductivities, while their low densities grant low settling velocities.
Using a "tail in" of resin-coated proppants whose curing process by temperature and pressure effects will create a cohesive wall at the well bore intersection, will greatly reduce the possibilities of proppants settling during the closure period or migrating to the horizontal borehole during early production.
DIVERTING AGENTS
Diverting agents are used in horizontal well stimulation treatments to "break down" the formation. They also are used when treating long producing intervals to attain complete coverage of the zone.
The most effective diverting agents are those coarse, malleable particulates with a wide range of size distribution and whose softening point is higher than the treating temperatures.
It is desirable that diverting agents be soluble in the reservoir fluid.
It is widely recognized that their effectiveness improves when a minimum density contrast exists between the diverting agent and the carrying fluid.
To assure access to the formation of particulate diverters when pumping treatments through perforated casing or liner, the recommendation is that the perforation orifice be at least six times the average diameter of the diverting agent.
JOB MONITORING
When fracturing in difficult conditions, it is important that job monitoring be carried out with the utmost precision. This will allow immediate corrective actions of any sudden and unpredicted changes in fracturing parameters.
Because horizontal well fracturing jobs present more critical conditions than corresponding treatments in vertical wells, it is obvious that adequate job monitoring is mandatory.
Electronic instrumentation and monitors, and real time software models should be used to allow the precision and speed required in horizontal well job monitoring. Not only should traditional downhole and surface fracture parameters be recorded, but careful real time monitoring should screen fluid conditions, the proportioning of components, and proppant feed.
PRIORITIES
Although the industry is still proceeding up the learning curve, previous experience in the North Sea has enhanced the engineering approach toward fracturing horizontal wells.
Major priorities are the development of fracturing fluids more suitable to horizontal well applications and also research geared to simpler methods of stress azimuth and magnitude determination.
The ultimate success of horizontal well fracturing depends a great deal on the simplicity that could be introduced in the design and execution of these complex jobs.
REFERENCES
- Joshi, S.D., "Augmentation of Well Productivity with Slant and Horizontal Wells," SPE Paper No. 15375, Journal of Petroleum Technology, June 1988, pp. 72939.
- Mukherjee, H., and Economides, M., "A Parametric Comparison of Horizontal and Vertical Well Performance," SPE Paper No. 18303, 63rd Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Houston, Oct. 2-5, 1988.
- Blanco, E., "Micro and Mini-Fracturing for Fracture Job Design Optimization," Joint Colombian Petroleum Congress and Andean Petroleum Congress, Bogota, Colombia, Oct. 28-31, 1986.
- Veeken, C.A.M, Davies, D.R., and Walters, J.V., "Limited Communication Between Hydraulic Fracture and Deviated Well bore," SPE Paper No. 18982, Joint Rocky Mountain Regional/Low Permeability Reservoir Symposium and Exhibition, Denver, Mar. 6-8, 1989.
- Warpinski, N., and Branagan, P., "Altered-Stress Fracturing," SPE Paper No. 17533, SPE Rocky Mountain Regional Meeting, Casper, Wyo., May 11-13, 1988.
- Karcher, B., Giger, F., and Combe, J., "Some Practical Formulas to Predict Horizontal Well Behavior," SPE Paper No. 15430, 61st Annual Technical Conference and Exhibition of the SPE, New Orleans, Oct. 5-8, 1986.
- Anderson, S., Hansen, S., and Fjeldgaard, K., "Horizontal Drilling and Completion: Denmark," SPE Paper No. 18349, SPE European Petroleum Conference, London, Oct. 16-19, 1988.
Copyright 1990 Oil & Gas Journal. All Rights Reserved.
