Ball-activated sliding-sleeve fracturing best practices

ALI DANESHY, Contributing Editor, Shale

Activity in the Bakken formation has seen remarkable growth during the last decade. Innovations in downhole fracturing tools and systems have been very important in this growth, among them systems utilizing ball-activated sliding sleeves. The systems’ popularity stems from several attractive features, including the following:

Speed of operations. Multiple stages of fracturing can easily be completed during a single, continuous pumping session. This translates into earlier production and lower costs.

Production from open hole. The open hole itself—in addition to the hydraulic fracture—contributes to early production.

Improved access. Once the completion assembly has been installed in the horizontal well, assuming no operational problems, the operator doesn’t need to re-enter the well for setting plugs, perforating, milling, etc. This is a particularly attractive feature for many deep Bakken wells, which may have a total reach of close to 20,000 ft.

Despite these advantages, ball-activated sliding-sleeve systems do create issues that must be addressed. For example, cleaning the well can be challenging, should it become necessary during fracturing or production. It is difficult to pass coiled tubing or other tools through the small opening of the ball seats/baffles, which are placed in the toe section of the well. In the case of screen-outs, a common remedy is to open the well and allow the fracturing fluid to flow back, hoping that this will clear the obstruction. If this action is successful, then fracturing operations can resume with the next stages. For fractures close to the heel, coiled tubing can sometimes be inserted into the completion system to clean the well. Extreme situations may require milling the completion and proceeding with a different fracturing scheme.

Sometimes cleanup is required during or just before the well is put on production. This can be triggered by excessive proppant flowback, which may block the flow of reservoir fluid, or by very poor recovery of the balls after the frac job. In these situations, the main option is milling the completion.

Another issue is that openhole packers isolate the different segments of the open hole, and seating them requires adequately gauged hole and proper installation and operation of the packers. There are some field indications of communication between different well segments. It is not clear whether these are caused by fractures crossing the packers, or by improper selection or installation of the packers.

Completion system description. A schematic drawing of a ball-activated sliding sleeve system is presented in Fig. 1. The tool where the ball is seated is called the ball seat or baffle, depending on the service provider. Fracturing begins from the toe of the well and proceeds toward the heel. After the first stage of fracturing, a ball with a specific diameter is dropped into the well. This ball is carried by the fluid through all the earlier, larger-diameter seats/baffles, finally seating in its target seat/baffle. The seating of the ball isolates the downstream part of the wellbore and causes the pressure to rise.

Sliding of each sleeve is triggered by the rupture of an attached pin, which is set in advance to shear at a certain pressure (usually in the range of 1,000–2,000 psi). Setting the pins at a higher pressure makes the ball seating more detectable at the surface. The rupture of the pin releases the sliding sleeve, which exposes the frac ports for the next stage. The fracturing slurry is transmitted through the ports into the liner-openhole annulus, and a new fracture is created within the interval isolated by the openhole packers. This process continues until all of the ports have been opened and all of the well segments have been fractured.

Figure 2 shows an example treatment chart for the first four frac stages in a well using this completion system. Pressure spikes at points A, B, C and D, caused by the seating of the balls, indicate the end of one frac stage and the beginning of the next. As the graph shows, fluid injection is continuous and the separation between stages is triggered by the seating of the ball.

Fig. 1. Schematic of a ball-activated sliding-sleeve fracturing system.

Fig. 2. Example treatment chart for the ball-activated sliding sleeve system.

Pressure computations. The surface fracturing pressure during each stage l of fracturing is:

The term ΔP_sfi may need some explanation. Flow of the slurry through each seat/baffle creates friction pressure, which can be computed by the application of Bernoulli’s equation. For each fracture stage, since the slurry has to travel through all upstream seats/baffles, the surface pressure reflects the sum of all these friction losses. As fracturing progresses from toe to heel, the last seat/baffle is isolated from the flow path after each frac stage, which causes a drop in the total seat friction pressure and consequently the surface pressure.

The magnitude of this friction pressure is dependent on seat/baffle diameter (the smaller the seat diameter, the larger the pressure drop), flowrate and density of the slurry, and, to a much lesser extent, fluid viscosity. Thus, as the seat/baffle diameter increases, the magnitude of this friction pressure becomes smaller and less noticeable. For a small-diameter seat/baffle (usually near the toe), this pressure can be as high as several thousand psi. It drops rapidly to only tens of psi for the larger-diameter seats/baffles near the heel. In the example shown in Fig. 2, this friction pressure is the main reason for the pressure drop after each ball seating.

Premature tool shifting. The cause of pin rupture is the differential pressure between the two sides of each seat/baffle. Although this pressure difference is supposed to be generated by seating of the ball, any other condition that generates a sufficient differential pressure can also rupture the pin and allow the sleeve to slide. One such mechanism is friction pressure, ΔPsf, caused by flow of the slurry through the seat/baffle. If the job is not designed properly, high pressure can be generated while pumping the frac slurry during the frac job. Its avoidance defines the upper limits of rate and proppant concentration that should be attempted during the fracturing of each stage. In practice, these conditions are likely to be met during fracturing of the toe stages, which have smaller-diameter seats/baffles. The larger-diameter seats/baffles easily accommodate the usual rates and concentrations used for industrial treatments. In a properly designed fracture, the maximum allowable rate and proppant concentration are computed for each stage so that the frictional pressure through each upstream seat/baffle does not exceed the safety-adjusted pin-setting pressure. A safety factor of 1.5 is recommended for this computation.

In recent designs, at least one company has incorporated two openings in the ball seats/baffles near the toe stages, and their closure requires the dropping of two balls. The presence of two adjacent seats/baffles causes substantial reduction in ΔP_sf, greatly reducing the risk of accidentally shifting the sleeves.

An important reason to avoid accidental shifting of the sleeves is uncertainty regarding which sleeve is going to be shifted. The actual rupture pressure of the pins is not a precise number, and variations in the sliding pressure between different adjacent sleeves should be expected. This creates an uncertainty as to which of the vulnerable upstream pins is going to fail first and cause its attached port to open. Once that occurs, execution of all the downstream fractures is jeopardized. Thus, determination of the critical rate and proppant concentration to avoid unintended opening of the frac ports is an important part of the fracture design.

In the early application of these systems, when faced with ball malfunction, some operators chose to increase the slurry rate as a way of forcing the sliding of the sleeve. This usually caused a number of subsequent problems, which jeopardized execution of the remaining frac jobs. A better approach for dealing with such problems is to accept the ball malfunction and continue with the frac jobs as planned. Even though this may result in skipping one fracture stage and increasing the size of another, the consequences are not as serious as trying to force ports open by increased slurry rate.

Ball acceleration. Another observation based on actual field data is early ball seating. Fluid flow inside the wellbore and the time of ball seating can easily be tracked by volumetric computations. Actual field data show the balls seating earlier than computed. The condition is more severe for balls seated near the toe. Reasons offered for this include parabolic velocity profile within the wellbore (higher velocity in the center of the well where the ball is traveling), or the ball dropping by gravity through the moving host fluid. A more plausible reason is ball acceleration as it passes through the seats/baffles. When the ball reaches a seat/baffle, the smaller area open to fluid flow creates a force behind the ball that pushes it forward as a projectile. The ball acceleration γ due to this force F can be computed as follows:

where m is the ball’s mass.

The force F is equal to the ball cross-sectional area multiplied by the pressure difference ΔP_i between its upstream and downstream sides:

where r_i is the radius of the ith ball. The ball’s mass m is related to the ball’s density ρ by the equation:

Therefore, solving for the acceleration yields:

As shown by Eq. 5, the acceleration is proportional to ΔP_i. Thus, reducing the flowrate while the ball is moving through the seats/baffles reduces ball acceleration.

The ball acceleration is repeated each time the ball goes through a seat/baffle. One concern with this situation is that acceleration of the ball may cause it to catch up with the proppant pumped ahead of it. If this happens, then some of the proppant can get trapped between the seat/baffle and the ball, and wedge the ball forcefully inside the seat/baffle. This may prevent the wedged-in ball from releasing after the job is complete, permanently isolating part of the wellbore and not allowing production flow through it.

Another concern is that the ball may catch up with proppant in an earlier seat/baffle (before its designated target). If the difference between the ball and seat/baffle diameters is small, then addition of the proppant may narrow the gap to the extent that the ball seats in an earlier seat/baffle. Seating of the ball will apply a force to the trapped proppant. If this force is large enough to cause crushing of the proppant, then the ball may pass through and seat for a second time in its designated target. If not, the targeted stage will not be fractured, and twice the intended volume will be pumped into the next stage.

The author has actually detected double ball-seating events in a few field treatments when the operators were concerned about over-displacement of the fracture and used a very short time gap between the end of the proppant stage (which usually has the highest concentration) and dropping of the ball. The main concern is that absence of proppant may cause closure of the fracture near the wellbore. Actually, the formation rock is quite rigid, and the hydraulic fracture is not going to close at the wellbore as long as there is proppant within a few tens of feet of it. Besides, the negative possible consequences of proppant getting wedged between the ball and seat/baffle are much more serious. This situation is easily avoided by increasing the fluid spacer between the end of the proppant stage and the dropping of the ball. A time gap of 2–3 min. is usually enough to avoid this problem.

A simple way to resolve concerns about ball release after frac jobs is to mix liquid tracers with the frac fluid, especially during the early stages. Return of the tracer is a good indication of the ball release.

Fig. 3. A broken ball recovered after the frac job.

Fig. 4. Example treatment with unclear ball seating action.

Accidental ball breakage. Another complication resulting from ball acceleration through the seats/baffles is that the ball may break if it hits the sides of the liner at high velocity. If the seat/baffle axis is not perfectly concentric with the liner axis, then the ball will be launched off-center and could hit the side of the liner with high velocity, causing it to break. Figure 3 shows a broken ball recovered after an actual job.

There are two concerns here. First, the broken ball may pass through its designated port instead of seating and opening it. This is not a big problem; its effect is the skipping of one frac stage and doubling the size of the previous stage. The signature of this situation is that pressure continues along its previous trend without a spike. The second concern is that the broken ball may still have sufficient intact surface area to seat, but not sufficient to cause the pressure difference necessary to open the port. The much smaller open area at the seat will create high friction pressure and could force reduction of the rate to stay within safe pressure limits. As fluid flows through the gap between the partially broken ball and the seat, it can gradually chip away broken corners of the ball and reduce its size, and eventually the ball may actually go through. This is a more serious operational problem. The best solution is to skip that fracture stage, drop the next ball and continue with the rest of the jobs.

One approach for reducing the probability of ball breakage is to decrease the slurry rate while the ball is moving through the seats/baffles. The rate recommended by the author is around 10 bbl/min. However, the rate should be high enough to avoid proppant settling within the horizontal section. The critical rate for proppant settling depends on fluid viscosity and proppant size and density. Rates higher than 5 bbl/min. are sufficient for most industrial treatments, even those using slick water.

Providers of ball-activated systems are keenly aware of the risks of broken balls. As a result, the material used for manufacturing the balls has changed several times. One recent development is introduction of shatter-proof balls. Some suppliers are exploring use of “disappearing” balls that will be made of a material that dissolves in the produced fluid.

Diagnostic tools. Analyzing frac pressure variations is an excellent tool for diagnosing the performance of these systems. Pressure data can provide answers to many common questions.

The most common way of determining proper ball seating is to look for the pressure spikes, as shown by points A through D in Fig. 2. However, occasionally the surface pressure does not show any spike, or the increased pressure shows unusual behavior. In these cases, there are other tools that help clarify the picture.

An example is presented in Fig. 4. In this case, the pressure behavior at point B is substantially different than at points A and C, and in fact does not show the typical rise and fall. However, if the ball had not properly seated, the slurry would be flowing into the previous stage, and the pressure would have continued its previous trend. The substantially different pressure behavior is a very good indication of a new fracture. In fact, the pressure increases after each ball seating, and its subsequent drop indicates considerable fracture re-orientation following its initiation.1 Comparing this pressure behavior with the behavior shown in Fig. 2 indicates that the latter did not undergo any re-orientation. This raises the possibility of axial fracture initiation and extension. In fact, the pressure data in Fig. 2 shows a high likelihood of each fracture connecting with the previous stage.

In general, the best method for diagnosing the performance of these systems is careful review of surface pressure data. Complicated questions about system performance can often be resolved by logical analysis of pressure behavior during the job. It is important not to avoid addressing tough questions by simply assigning the blame to the formation.

Conclusions. In spite of their simple appearance, effective use of ball-activated sliding-sleeve fracturing systems requires advanced knowledge of their hydrodynamics. Some of the operational problems encountered with these tools can be corrected with very simple preventive steps. The popularity of these systems has prompted new innovations and developments that carry the promise of more robust tools in the future. The goals of all these innovations include the ability to perform more fracturing stages and the ability to re-enter the well.  

LITERATURE CITED
¹ Daneshy, A.A., “Hydraulic fracturing of horizontal wells: Issues and insights,” SPE 140134 presented at the SPE Hydraulic Fracturing Technology Conference and Exhibition, The Woodlands, Texas, Jan. 24–26, 2011.