Multistage fracturing using plug-and-perf systems

ALI DANESHY, Contributing Editor, Shale Technology

The plug-and-perf system creates multiple hydraulic fractures in a horizontal well completed with a cemented casing/liner. This system combines elements of two common fracturing techniques: limited entry and segmented fracturing using bridge plugs. Commercially, these systems are offered and supported by several suppliers.

System description. Fracturing operations begin from the toe of the well and proceed toward the heel, Fig. 1. Each pumping stage is used for fracturing multiple clusters of perforations designed based on the limited-entry technique. The number of clusters is based on the injection rate, with details discussed later. The sequence of operational steps consists of setting the plug in the desired location within the horizontal well, perforating the clusters and pumping the treatment according to the design. After all pumping stages are complete, all the plugs are milled out, and the well is cleaned and put on production. Depending on requirements, there are a variety of plug types used in these completions, but they are all drillable/millable and usually made of special composite material.

Fig. 1. Plug-and-perf completion system.

Mechanics of fracture initiation and extension. Stress distribution around horizontal wells promotes axial (longitudinal) fracture initiation, even when this is not perpendicular to the least in-situ principal stress. The reason is mainly geometrical.¹ While this point can be shown mathematically, a simple explanation for it is that, under high pressure, a cylindrical pipe (such as most oilfield tubulars) always splits along its length (axially) and not perpendicular to it. While perforations can sometimes change this and cause the fracture to initiate perpendicularly, to the least in-situ stress, many field examples show that axial fracture initiation is in fact prevalent in industrial cased-hole fracturing.

In many industrial completions, the objective is to create multiple transverse fractures (perpendicular to the wellbore). The extension of the initial axial fracture will cause it to reorient itself and become perpendicular to the least in-situ principal stress. The reorientation is accompanied by extensive shear fracturing, Fig. 2. In this situation, the transverse fracture may not actually be directly connected to any perforations at all, since the reorientation location is dictated by formation properties and heterogeneiety (planes of weakness). In this case, the fracturing fluid and proppant have to travel through a complicated and tortuous path within the formation to move from the perforations into the fracture. This can complicate the fracture extension process by requiring very high injection pressures even at low rates, and possibly screen-out, as explained later.

Fig. 2. Fracture reorientation process.

Limited-entry fracture design. Each stage of a plug-and-perf process consists of creating multiple fractures from several clusters of perforations. The design principle for creating these fractures is known as the limited-entry technique, where multiple perforation clusters are fractured during a single injection. Required fluid distribution between these fractures is secured by limiting the number of perforations in each cluster, and the total. The engineering basis for this design is as follows.

The friction pressure due to flow through a perforation, ΔPp (psi) is given by the following equation:

where Q is the injection rate (bpm), D is the perforation diameter (in.), n is the number of perforations, C is the perforation coefficient (usually between 0.6 and 1.0), and ρ is the fluid density (lbm/ft3). As an example, injecting water at the rate of 2 bpm through a single perforation with a diameter of 0.4 in. and coefficient C = 0.9 creates around 380 psi of friction pressure. At 3 bpm, this pressure would be around 850 psi, and at 4 bpm it would be over 1,500 psi. This means that if the number of perforations accepting fluid is less than the intended number, the perforation friction will increase and cause an increase in downhole and surface injection pressures.

As an example of limited-entry design, suppose the fracturing treatment is planned to be injected at the rate of 60 bpm with three fractures taking an equal volume. If the well has 30 perforations in three cluster of 10 each, then each perforation will ideally receive fluid at 2 bpm, which creates around 380 psi of friction pressure. Each fracture is created at a total injection rate of 20 bpm through 10 perforations. While it is recognized that the actual rate may vary between different perforations, and that all of them may not be open, nevertheless it is expected that all three clusters of perforations will be getting some fluid volume. The engineering rationale for this expectation is that, if an entire cluster was excluded from the injection process, then the remaining 60 bpm will have to be divided between 20 perforations, or 3 bpm/perforation. This would create around 850 psi of friction pressure. It is reasoned that at 850 psi of pressure difference there is high probability that at least some perforations in the third cluster will also accept fluid, thus creating three separate fractures, albeit with unequal volumes. While this is true for many industrial treatments, there is field evidence suggesting that this is not universally true.

The extent of the axial fracturing near the horizontal well (i.e., the length of the cased-hole interval fractured axially) depends on a number of variables, among them the location of actually fractured perforations and formation heterogeneiety. Figure 3 shows the proppant tracer survey following a three-stage frac job. The noteworthy features of this survey are that the extent of axial fracturing is different for each stage, and that there is no proppant trace in the third perforation cluster, which raises the high probability that this cluster was not fractured at all.

Fig. 3. The fracture trace is a three-stage, limited-entry frac job.

Another interesting observation is that sometimes the axial fracture is not directly opposite the perforated interval, Fig. 4. Obviously, the fracture is connected to the perforations, but the absence of proppant indicates that the reorientation to transverse fracture may have occurred away from the perforated interval.

Fig. 4. Misalignment of fracture and perforations.

Note 1. In general, the extent of axial fracturing in cased and perforated horizontal wells is much less than in open-hole completions.

Note 2. While axial initiation and re-orientation complicates the execution of frac jobs, it can actually help the early production from the well by providing for faster drainage of the near-wellbore region.

Proppant distribution between multiple fracture stages. Another complication with multistage fracturing using the limited-entry technique is the uneven proppant distribution between different perforation clusters.² As stated above, the design of these fractures is based on regulated fluid distribution between multiple clusters of perforations. However, the mechanics of proppant movement inside the wellbore are different from those of fluid movement. While fluid molecules can easily change direction and enter the perforations, the proppant grains with higher density and larger size and mass cannot do this as easily. Furthermore, wellbore flow velocity gradually decreases as more of the fluid is diverted into the perforations. The result is that the early perforations receive less than their allotted share of the proppant.

As the velocity decreases along the horizontal well, it beomes easier for the proppant to change direction, which then gradually increases the amount of proppant entering each perforation. At the end of the perforated interval, the proppant has no option but to enter the last perforations. Figure 5 shows the proppant distribution between four perforation clusters, based on advanced numerical simulations of the process. The multiple scenarios modeled here include variations in fluid viscosity, injection rate, and proppant size and density. Although the data shows some variations between different scenarios, the common feature of all of them is highly uneven proppant distribution.

Fig. 5. Proppant distribution in perforation clusters.

Effect of perforation pattern on fracturing. This is one of the most critical aspects of fracturing cased horizontal wells. Since, by design, the majority of perforations have to be connected to the fracture, it is essential that the perforation pattern accommodate an easy connection. Based on experimental data, Abbas et al recommend a very short perforated interval, on the order of one or two wellbore diameters.3 For wellbore sizes presently used by the industry, this translates into less than one foot. At the present time, the more common perforation pattern is spiral with 60° phasing, Fig. 6a. Even though there is an attempt to keep the length of the perforated interval short, often as low as 6 shots/ft, field data indicate that many cased-hole fractures initiate as axial and then re-orient. In this author’s view, the perforated interval should be kept as short as the existing perforation tools can accommodate. This reduces the fracture tortuosity during reorientation, thus simplifing the connection between the fracture and all perforations, especially for the transmitting proppant. The limited-entry technique, by design, is based on fluid and proppant movement through the vast majority of, if not all, perforations. The manifestation of the ease of slurry movement is the fracturing pressure. High surface pressure is a very good indication of poor communication between the wellbore and the fracture(s). For this reason, one of the recommendations of this author is to delay proppant injection until the surface pressure begins to decrease, indicating adequate connection with the fracture.

Fig. 6. Perforation patterns.

Ideally, for creation of transverse fractures the perforations need to be in a single plane perpendicular to borehole axis, as shown in Fig. 6b. However, ability to implement this pattern is limited by availability of perforating tools.

The manifestation of inadequate connection between the wellbore (perforations) and the fracture is inability to pump at the desired rate because of very high injection pressures. An example of this situation is presented in Fig. 7. During the early phase of this treatment, the injection rate was much less than the design, and had to be adjusted frequently to keep the surface pressure below the maximum allowable limit. The main reason for this behavior was inadequate connection between the fracture and perforations. In fact, the very low rate indicates that only a handful of perforations were taking fluid at the beginning of the job, even though the well had 30 perforations distributed in three clusters. The high pressure persisted for several hours before rates could be increased close to the design values. A more reasonable pressure was reached after more than six hours of pumping, after which the job was completed. However, although the designed fluid volume was eventually pumped into the formation, the injected proppant was much less than planned. The other stages of fracturing in this well went mostly as planned. This situation, while not very frequent, is also not very rare.

Fig. 7. Example frac chart showing inadequate connection between perforations and fracture.

Figure 8 shows another example for a well with 24 perforations distributed in five clusters. The designed injection rate was 36 bpm, providing 1.5 bpm/perf. The fracture reorientation process was completed after about 10 minutes of injection. The pressure increase in the middle of the job was caused by screen-out in one or more of the clusters. The increased pressure caused initiation of more fractures (axial initiation and reorientation). This indicates that all the clusters were not fractured initially, and that screen-out created more fractures. The uneven proppant distribution discussed earlier promotes the partial screen-out. By the way, the net effect of this type of screen-out is actually positive.

Fig. 8. Partial screen-out and new fracture initiation.

Pseudo-openhole condition. If some situations, especially where axial fracturing is extensive, the borehole enlargement is much higher than can be matched by expansion of the steel casing. This results in separation of the formation from the casing, thus creating a pseudo-openhole condition, Fig. 9. Such an occurrence can actually help the execution of the frac job. The gap between the casing and formation provides a reasonable conduit for the fluid and proppant to enter the transverse fracture. This is particularly important in those cases where the transverse fracture is away from the perforated interval.

Fig. 9. Pseudo-openhole condition created by axial fracturing.

Completion system attributes. Briefly, some of the advantages of cased-hole systems are:

Hole stability and access. A cased and cemented well offers a superior environment for movement and deployment of downhole tools and after-frac wellbore clean-out.

Fracture location control. Fractures are created at, or close to, the perforated intervals. Proper perforation planning can provide reasonable control on the location of fractures, and achieve the desired spacing between them. (In openhole completions, the fracture can be anywhere within the isolated openhole interval.)

Number of fractures. Cased horizontal wells are often intentionally fractured with more stages than open holes. This is because the completion system can accommodate more stages (although this advantage is rapidly disappearing with new developments). More fractures allow higher early production rates and faster depletion of the reservoir.

System maturity. Most of the components used in these systems have been in use by the industry for a long time and, therefore, are operationally and technically more mature.

Fullbore access. After fracturing, the downhole plugs are milled out and the entire wellbore area is open to flow. This provides an advantage for future operations and tool deployment, including possible clean-outs in case of proppant flowback.

Re-fracturing options. If needed, cased holes are easier to re-fracture. In fact, some re-fracturing of horizontal wells has already been accomplished in the Barnett and Marcellus shales.

Together with the above benefits, cased-hole completions can be subject to the following limitations:

Delayed production. Completing a cased hole can take longer than an open hole. In fact, the fracturing operations themselves take longer to execute in a cased hole. This can delay the start of production, and generation of revenue.

Duration of fracturing. In these systems, the total time of fracturing includes the pumping time, as well as milling the completions and cleaning the wellbore. Of these, the pumping time can become an issue in locations with shortage of supply in fracturing pumping services.

No wellbore production contribution. At least during the early production, the horizontal well itself contributes to production in open holes. This is not true for cased holes.

Less robust connection with the fracture. Fluid and proppant have to be transmitted into the fracture through the perforations. As we saw earlier, in some cases this can cause a problem.

Lateral access. Deployment of downhole tools, milling and wellbore clean-out require access to the toe of the horizontal well. The total length of the horizontal well is limited by the capabilities of available coiled-tubing or workover systems. In some deep formations, this can limit the length of the horizontal section.

Conclusions. Plug-and-perf completions are an attractive option for creating multiple fractures in horizontal cased and cemented wells. They represent the dominant system used for production from many of the unconventional shale oil and gas reservoirs. In general, a cased-hole completion provides more latitude for experimentation with fracture design, increasing the number of fractures, and options for re-fracturing. However, milling and removal of the plugs at the end of fracturing can sometimes be difficult. 

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.
2 Daneshy, A. A., “Uneven distribution of proppants in perf clusters,” World Oil, April 2011, pp. 75–76.
3 Abass, H. H., Hedayati, S. and D.L. Meadows, “Nonplanar fracture propagation from a horizontal wellbore: Experimental study,” SPE Production & Facilities, pp. 133–137.