Screenout Detection and Avoidance

Abstract

A screenout (SO) event is defined as premature termination of a propped hydraulic fracture (PHF) treatment due to bridging of the proppant at a restriction, usually located at the near-wellbore area or at the perforations—and not at the perimeter tip of the fracture. Numerous Screenout Detection (SD) methodologies have been presented over the past 60 years which are designed to predict the likelihood of a SO event based on analysis of minifrac data, or real-time data—during the mainfrac. Three simple SD and numerous screenout avoidance (SA) methods are presented. The SD methods are: enhanced fracture entry friction (FEF) analysis, median ratio (MR) and inverse slope (IS). Analyses with the first two methods require data from a step-down test (SDT), while the third method uses data which are analyzed in real-time during the mainfrac. By knowing the potential for a SO event or by having advance warning of the onset of a SO event, one is able to apply design modifications for the mainfrac, or is able to initiate abrupt, or incremental step displacement (flush) and achieve SA, or to extend the PHF treatment to improve placement by attaining increased net-pressure gain. The theory and logic of the SDT and of the three methods includes discussion on: the balloon analogy, stagnation pressure, fracture toughness and fracture tip dilatancy. Both the MR and the IS methods do not require a computer or software and all three methods are very easy to use at the well location by the on-site engineer. All three methods are very inexpensive. Numerous publications have dealt with SA over the past 70 years, and have presented design modification procedures, and wellbore intervention procedures, which are presented briefly. These procedures when implemented individually, or in combination, have proven to be very useful in: Preventing SO events, achieving positive SA outcomes, by enabling the safe and effective placement of PHF treatments.

1. Introduction

1.1 Literature review and historical background

Kern et al. [1] were the first to conduct experiments on the movement of proppant in fractures. They determined that proppant drops-out of the carrying fluid and forms a mound immediately after passing through the perforations. These results were confirmed experimentally by Wang et al. [2] and were to be expected as the experiments were conducted with linear fluids that have limited proppant carrying capacity. The first serious real-time methodology of SD was proposed by Nolte and Smith [3] with the logarithmic net-pressure plot (NPP), (a.k.a. Nolte plot or Nolte-Smith plot), plotted at bottomhole conditions. The NPP calculation ignores near-wellbore (NWB) friction and fracture tip dilatancy (FTD), as it assumes linear pressure dissipation along the length of the fracture. See the blue line in Figure 1. The incorrect net-pressure (NP) which is calculated causes the methodology to have very low predictive accuracy.

Figure 1.

Accounting for pressure declines in a wellbore-fracture system. The conventional method ignores near-wellbore friction, and, assumes linear pressure loss along the fracture length (see solid blue line). The true method is more representative of pressure loss along the fracture length as it accounts for the previously mentioned pressure loss components (see solid red line).

The numerous publications on tip-screenout (TSO) will not be discussed or referenced here as TSO occurs only during frackpacks, where SO is the objective rather than an issue, i.e., if TSO is not achieved during a fracpack the operation is repeated. Barree [4, 5] and Barree et al. [6] provide exhaustive discussion and explanations on TSO.

Cleary et al. [7, 8] introduced the step-down test (SDT) and the associated conventional fracture entry friction (FEF) analysis, which, calculates the magnitude of perforation friction and near-wellbore (NWB) friction. They designated NWB friction as the primary cause of screenouts, and perforation friction as a secondary cause of screenouts. Conventional FEF analysis does not provide unique (very accurate) results because of various inherent limitations. These limitations are: (a) use of uncorrected fluid friction factors, (b) use of uniform perforation discharge coefficient, (c) Not using the Maximum Drag Reduction (MDR) asymptote, and (d) not matching Measured Total System Friction (MTSF) with Calculated Total System Friction (CTSF) at all rates of the SDT. Even though Conventional FEF analysis increased the predictive accuracy of SD and enabled the application of appropriate SA procedures, it was not high enough to be satisfactory, so, Massaras et al. [9] introduced enhanced fracture entry friction (FEF) analysis which eliminated and bypassed the above mentioned inherent limitations, thus, the predictive accuracy of SD was significantly improved, along with the outcomes of the SA procedures.

Both Conventional and Enhanced FEF analysis require software, a computer, and a trial-and-error solution procedure, as there are several unknowns which prevent a direct solution. To further simplify the analysis and enable rapid on-location determination of appropriate and useful SO potential, or SO onset information, Massaras et al. [10] developed the post-minifrac Median Ratio (MR) technique, which uses data from the SDT and stipulates that the SDT must be performed according to a prescribed four equal-step procedure. Massaras et al. [11] also developed the proprietary Screenout Index (SOI) technique which is a specialized version of the MR technique used when the SDT is not performed as per the prescribed four equal-steps procedure. Due to the need for a rapid Real-time (RT) in-mainfrac SD technique, Massaras et al. [12] developed the Inverse Slope (IS) technique, which uses smart tangent lines for data trend visualization on the surface pressure plot. The moment the surface pressure starts to deviate from the tangent line on the Inverse Slope it marks the start of a SO downhole. It is an early positive advance warning of a SO occurring downhole, thus, it is imperative to immediately commence the displacement (flush) procedure, as the wellbore travel time can be 5–10 min long, depending on wellbore geometry and injection rate.

Mondal et al. [13, 14] based on bottomhole SDT data—recorded with downhole (DH) gauges—have presented a modified Conventional FEF analysis which stipulates that the use of a near-wellbore friction exponent, β = 0.5, overestimates perforation friction from SDTs, and that the value of β —after proppant slurry has been placed through the perforation and NWB area is between 0.25 and 1.0, as confirmed by El Rabba et al. [15]. Roberts et al. [16, 17] have presented images captured with downhole (DH) video which show rounding of the perforation entrances, and enlargement of the perforation diameter (flow area).

Chipperfield et al. [18] and Roberts et al. [19], proposed an ISIP Gradients methodology, and defined the magnitude of NWB friction as the difference between the pressure prior to shut-in and the instantaneous shut-in pressure (ISIP), at bottomhole conditions, ignoring perforation friction all together. Chipperfield et al. [18] also proposed to use the data from the step-up test (SUT) as a SD methodology. It is well-known and widely accepted that the SUT cannot be used for SD, however, as subsequently discussed the SUT is very useful for other operations: Fracpacks, matrix acidizing, and Closed Fracture Acidizing (CFA). Yang et al. [20] proposed a Pressure Declining Gradient (PDG) methodology which utilizes the leak-off rate calculated from pressure fall-off data recorded during the shut-in interval.

Cai et al. [21, 22] presented a semi-analytical model which has an analytical model for proppant transport and a numerical model for proppant pilling, which predicts the formation of a proppant mound in the fracture and aims to predict SO. Sun et al. [23] presented a data-driven method–like a deep Neural Network (NN)—along with a physics-based approach—like the Inverse Slope Method [12]—to provide advanced warning of a SO in real-time. Hu et al. [24] proposed a Locally Weighted Linear Regression (LWLR) approach, which combines a Particle Filter (PF) algorithm and the Autoregressive Moving Average (ARMA) model together; to provide an early warning of SO events. Yu et al. [25] used Gaussian Hidden Markov Models (GHMMs), to train on simulated data, in order to predict screenouts and provide early warning by learning pre-screenout patterns in simulated surface pressure data. Hou et al. [26] presented a continuous SO evaluation and prediction methodology combining data-driven methods with recorded data. The SO probability is updated, redefined, and used to label the original data using three determining elements of SO, based on which four indicators are generated. The Deep Learning (DL) model is trained with Gated Recurrent Units (GRU), and is tuned by a grid search and walk-forward validation. Barree [5] presented a variable tortuosity screenout model in which the tortuosity and the erosion factor are variable.

Merry et al. [27] presented a novel Tube Wave (TW) methodology which utilizes a low frequency hydraulic tube—or Stoneley-wave signal which is induced by equipment at the surface that is tied into the wellhead or high-pressure treating lines. The tube wave travels rapidly through the fluid within the wellbore and reflects off the bottom of the well, which influences the tube wave form properties—dispersion, and attenuation of the normal modes, depending on the condition and number of perforations, the quality of their connection with the near-field region, and the quality of the near field region itself, i.e., if this region is naturally fractured or not.

Di Vaira et al. [28] investigated screenouts of micro-proppants (400–500 mesh size) in narrow fractures using numerical simulations, where data from numerical test cells are translated to regions of screenout. The technique uses the proppant solid volume fraction, φ, and the ratio of fracture width to proppant diameter, w/d, which is in line with current bridging modeling practice, while taking into consideration collision frequency and bridge stability, and the use of electrostatics, for screenout predictions.

1.2 Simple screenout detection methodologies

Several of the SD methodologies previously summarized were introduced during the past 60 years, and most during the past 5 years, so, they are neither available widely nor used extensively, as they have not been implemented in any of the commercially available Fracture Propagation Simulators (FPS), they are cumbersome to use, as they require a computer, software, and a specially trained operator for the system, and furthermore they do not provide any recommendations for design modifications or wellbore intervention procedures aimed at achieving SA. The TW methodology also calls for equipment to be connected to the wellhead, Thus, these requirements tend to increase the cost of most methodologies, so, this chapter presents in detail only three simple methods, two of which are very simple to use and have low cost or no cost at all. The three methods are: Enhanced FEF analysis, Massaras et al. [9], Median Ratio (MR), Massaras et al. [10] and Inverse Slope (IS), Massaras et al. [12]. The first two methods are classified as post minifrac (pre-frac), and the third as real-time, and are presented in detail in Section 4, entitled “Simple Screenout Detection Methods.”

1.3 Design modification and wellbore intervention procedures

Once it has been determined that there exists potential for a SO event, there are many Design Modification and Wellbore Intervention Procedures available for implementation in order to achieve SA. These options can be implemented in the office during the planning/design phase, on location after the minifrac during preparation of the final design phase, or during the mainfrac execution phase in real-time. All Design Modification and Wellbore Intervention Procedures are discussed briefly in Section 7, entitled “Screenout Avoidance.”

2. Theoretical considerations

As per the balloon analogy—subsequently discussed—the pressure in a fracture like in a balloon is fairly constant. Therefore, in order to account for pressure losses in the fracture one must consider what happens at both the entrance of the fracture, and at the perimeter tip of the fracture. Considering stagnation pressure (SP) and calculating the pressure losses in the conduit (tubing or casing) and at the fracture entrance, due to perforation friction and NWB friction does not account for all the pressure losses in the system. Given that the velocity of the fluid in the fracture is very low (<2 ft./min.), the pressure loss due to friction in the fracture itself is negligible, thus, all pressure losses in the system cannot be accounted-for with friction. Fracture toughness (FT)—a property of the rock formation—dictates that a great amount of pressure must be expended to re-open the small unsupported portion of fracture at the tip which has closed due to fracture tip dilatancy (FTD)—subsequently discussed. Thus, by the process of elimination, the great pressure losses which are unaccounted for must be realized at the perimeter tip of the fracture where the Positive Net Pressure is very large, as shown at the right side of Figure 1 (red line). Attaining complete and thorough understanding of SD and SA requires good knowledge of the previously mentioned theoretical concepts, discussed in the subsequent sections:

1. The balloon analogy

2. Stagnation pressure

3. Fracture toughness

4. Fracture tip dilatancy

2.1 The balloon analogy

Inflating a balloon is analogues to inflating (injecting into) a fracture, and when injection ceases, the pressure in the balloon remains constant everywhere as the entrance is pinched shut and the pressure in the mouth is reduced to atmospheric. See Figure 2. Stated another way, there are no abrupt pressure reductions in the balloon, so, it follows that there are no abrupt pressure reductions in the fracture.

Figure 2.

Inflating a balloon is analogues to inflating (injecting into) a fracture.

The main reason the pressure in a fracture remains fairly constant—even during fluid injection—is because the velocity of the fluid in the fracture is very small (<2 ft./min.), so, pressure loss due to friction in the fracture is negligible, and since prior to commencing the SDT the created fracture is fully inflated, and since there are no abrupt pressure reductions in the fracture, the dynamic pressure response to rate reductions should be related almost completely to the frictional components: wellbore, perforation and near-wellbore.

Since the pressure at the tip is not expended on friction, but rather to reopen a very small fracture, which has closed due to fracture tip dilatancy (FTD) the pressure along the length of the fracture is fairly constant as shown in center section of Figure 1 (red line). FTD helps to account for the pressure reduction to pore pressure. See right side of Figure 1, where the Positive Net Pressure is very large. This is explained in more detail in the subsequent sections entitled “fracture tip dilatancy” and “fracture toughness”. Extensive discussions on both are provided by Massaras et al. [9, 10].

2.2 Stagnation pressure

Fluid flow in a perforation is analogous to a jet nozzle impacting a stationary plate. See Figure 3a and b. Both systems consist of a high velocity jet and both experience very high pressure and very low or zero velocity at the stagnation point. Both configurations satisfy the law of conservation of energy. It is the high stagnation pressure in the perforations that allows rapid pressure reductions to be noted during the abrupt rate reductions that are executed during a SDT.

Figure 3.

Fluid flow in a perforation (a) is analogous to a jet nozzle impacting a stationary plate (b). At the stagnation point of both, the pressure is at a maximum because the velocity of the fluid is zero or close to zero.

As per the balloon analogy—previously discussed—there are no abrupt pressure reductions in the fracture. Therefore, it is clearly evident that the abrupt pressure reductions noted during a SDT, where the pressure decreases to below stabilization level, is due to stagnation pressure as shown in Figure 4a. Conversely, abrupt pressure increases to above the imaginary stabilization level are not noted during a step-up test (SUT) as shown in Figure 4b. The difference in behavior has to do with the very high stagnation pressure experienced at the stagnation point located at the deepest point of the perforations, which allows abrupt pressure reductions—to below stabilization level—to occur and to be noted during the abrupt flow rate reductions of the SDT.

Figure 4.

(a) The SDT shown on the left detects abrupt pressure changes corresponding to abrupt rate changes. (b) The SUT shown on the right cannot detect abrupt pressure increases. Figure 4b adopted and modified from Tinker et al. [29].

The rapid reduction in pressure makes the data from the SDT very useful for analysis as a progressive pressure reduction pattern is considered ideal, and any deviation from this ideal pattern can be utilized to determine which friction component dominates (is larger). It is also possible to separate the pressure loss due to friction into its constituent components of perforation friction, and near-wellbore friction taking advantage of the difference in the plotting pattern of the SDT data, i.e., concave upwards, which is customary for a function of f(Q²), or concave sideways which is customary for a function of f(Q^1/2). The significance of the difference in plotting pattern is as follows:

• The concave upwards f(Q²) plot shown in Figure 5a, results from progressively smaller pressure reductions, and it means that the near-wellbore friction is low, and that the perforation friction dominates (is larger). Low near-wellbore friction means that there are no restrictions in the near-wellbore area.

• The concave sideways f(Q^1/2) plot shown in Figure 5b, results from equal or progressively larger pressure reductions, and it means that the near-wellbore friction dominates (is larger) and that the perforation friction is low. High near-wellbore friction means that there are restrictions in the near-wellbore area.

Figure 5.

(a) Pressure behavior of flowrate squared, Q², and (b) square root of flowrate, $\sqrt{\mathit{Q}}$ or Q^½.

The difference in pressure reduction behavior of the SDT data is explained in more detail in Section 3.4 entitled “Theory of the step-down test.”

2.3 Fracture toughness

Fracture toughness (FT) is a measure of a material’s ability to resist the propagation of cracks or fractures under stress. It represents the amount of energy required to propagate a crack in a material and is an important property used in design and evaluation of materials for structural applications.

In fracturing applications it influences the very first moments of pumping operations, when fracture initiation or fracture reopening are occurring, i.e., when the fracture is very small. The phenomenon is illustrated in Figures 6 and 7, where at the very beginning of the mainfrac portion of the treatment a very large pressure increase (a pressure spike) is noted because the fracture is closed, i.e., it is very small.

Figure 6.

Fracture toughness causes pressure spike as a small closed fracture is being re-opened.

Figure 7.

Hydraulic fracture treatment with sweep stage inserted, which does not have to be very large.

The fact that large net-pressure—the pressure that holds the fracture open—is required to propagate a small fracture in a uniform stress field is proven mathematically for a radial fracture with Eq. (1).

$$P_{\mathit{net}}=\frac{K_{\mathit{IC}}}{\sqrt{\mathit{\pi \alpha }}}$$E1

where, K_IC = Mode I fracture toughness critical stress intensity factor, (psi/in^0.5); α = radial distance to the fracture tip (in).

Since the radial distance term, α, is in the denominator, the magnitude of the net-pressure required to propagate a small fracture is large, while the magnitude of the net-pressure required to propagate a large fracture is small, as shown in Table 1.

KIC (psi/in^0.5)	Radial distance to fracture tip (α)		Net-pressure
			Pnet
	(ft)	(in)	(psi)	(bar)
1000	0.083	1	565.0	38.9
1000	12	144	47.0	3.2
1000	100	1200	16.3	1.1

Table 1.

Large net-pressure is required to propagate a small fracture. Small net-pressure is required to propagate a large fracture.

It is clearly evident from the SUT shown in Figure 4b that the first rate increase is very small while the corresponding pressure increase is very large. This behavior is due to the fact that at the beginning of the SUT the fracture is very small. Conversely, during the second half of the SUT the rate increases are large while the corresponding pressure increases are very small. This difference in behavior is due to the fact that after a short time interval of the SUT has elapsed the fracture is large enough not to be influenced by fracture toughness.

Because FT influences the process of fracturing for only a few minutes at the very beginning of pumping—when a new fracture is initiated, or an existing fracture is being reopened—it is not worth worrying about FT during normal fracturing operations.

FT plays a dominant and important role in the pressure behavior at the Fracture Perimeter Fluid Lag zone at the far end (tip) of the fracture. As the fracture treatment progresses; the leading edge of the fracturing fluid arrives at the dilated and closed fluid lag zone and has to re-open it. See next section entitled fracture tip dilatancy for details. Because of FT very large pressure is needed to re-open the very small fracture at the tip, therefore, a notable amount of pressure is expended at the tip of the fracture as shown at the right side of Figure 1 (red line), where the Positive Net Pressure is very large. Massaras et al. [9] provide detailed discussion on FT.

2.4 Fracture tip dilatancy

Many theories have been developed and presented regarding the mechanics of fracture tip propagation and fracture tip pressure behavior, which have been summarized by Massaras et al. [10]. Fracture tip dilatancy (FTD) is among these theories; however, it is more than a theory—one could say it is a property—as it can be observed in nature (beach), and, modeled, observed, and measured in the lab. See a demo of dilatancy with beach sand at the following URL: http://www.youtube.com/watch?v=EnIXBAmJcZI

FTD is defined as non-linear rock expansion, which manifests itself due to fluid lag near the tip of the fracture. The phenomenon occurs because the tip of the fracture being created normally propagates much faster than the leading edge of the fracturing fluid. This means that the unsupported part of the fracture—the fluid lag zone or dilatancy zone—remains momentarily open and in vacuum, as shown at right side of Figure 1 where the net-pressure is negative. Eventually the dilatancy zone closes as the rock formation dilates because two conditions are met: The rock formation is normally granular (sandstone), and it is confined (in-situ), by the high overburden pressure. As dilation occurs at both faces of the fracture in the fluid lag zone, it causes the small fracture at the dilatancy perimeter to close.

FTD essentially is influenced by fracture toughness, which dictates that a lot of pressure is required to reopen a very small closed fracture as discussed in the previous section entitled “Fracture Toughness.” Massaras et al. [9, 10] provide an extensive and detailed discussion on FTD.

3. Practical considerations for screenout detection

Seven practical considerations regarding SD are listed and discussed:

1. Appropriate test for screenout detection

2. Categories of screenouts.

3. Causes of screenouts

4. Theory of the step-down test

5. Correctly Sizing Minifrac and Diagnostic Fracture Injection Test (DFIT)

6. Correct procedure for the step-down test

7. Performing the step-down test in the Pad

3.1 Appropriate test for screenout detection

Often the question arises as to which test is appropriate for SD. The answer depends on the type of treatment planned to be performed, and on the type of formation being treated. There are two options:

• Option 1: A SDT is recommended when fracturing with proppant conventional sandstone formations, and unconventional shale formations.

• Option 2: A SUT is recommended when performing: (a) Fracpack on unconsolidated high permeability sandstone formations, (b) Closed Fracture Acidizing (CFA) on carbonate formations, and (c) matrix acidizing on either sandstone or carbonate formations.

The reasoning behind these choices depends on the information provided by the analysis of each test, and on how the information is to be utilized in decision making. For option 1 the near-wellbore friction is the primary and the most important parameter to be determined, and perforation friction is a parameter of secondary importance. Only the SDT can be used to calculate the near-wellbore friction, because it is the only test that can detect abrupt pressure reductions corresponding to abrupt rate reductions.

The SUT cannot detect abrupt pressure reductions, as such; the SUT data cannot be utilized to determine friction losses in the perforations or the near-wellbore area. This is evident from the vastly different pressure behavior exhibited by each test: The SDT shown in Figure 4a detects the abrupt pressure reductions, corresponding to the abrupt rate reductions as evidenced by the pressure declines to below the stabilization level. The SUT shown in Figure 4b cannot detect abrupt pressure increases corresponding to abrupt rate increases as evidenced by the fact that the pressure does not increase to above the imaginary stabilization level.

Further evidence that the SUT cannot detect abrupt pressure increases is the fact that large rate increases performed during the second half of the SUT are accompanied by extremely small pressure increases as shown in Figure 4b. This phenomenon is the result of fracture toughness, which has great influence on the pressure only when the fracture is extremely small as evidenced by the very large pressure increase corresponding to the extremely small first rate increase, as shown in Figure 4b. Therefore, the SUT cannot be used to calculate neither the perforation friction nor the NWB friction.

Very useful data are obtained from the SUT, i.e., the Fracture Extension Pressure (FEP), and the Fracture Closure Pressure (Pcl), as shown in Figure 8b. The FEP is not very useful when performing PHF treatments on conventional sandstone formations and unconventional shale formations, as it only serves as check on the upper bound limit for Pcl. See Figure 8b. Essentially the FEP serves as a check on the accuracy of the calculation of the magnitude of the net-pressure (Pnet)—the pressure that holds the fracture open. Because Pcl is critical in calculating net-pressure and fluid/frac efficiency it must be determined with great care, thus, Pcl happens to be the most important calculated value in hydraulic fracturing.

Figure 8.

The SUT is useful to determine the FEP. Adopted and modified from Tinker et al. [29].

The FEP is extremely useful during:

1. Fracpack treatments where it is critical to ensure that injection is occurring in fracture mode (fracture open), and Matrix acidizing treatments, where it is critical to ensure that injection is occurring and matrix mode (no fracture created or fracture is and remains closed).

2. Closed Fracture Acidizing (CFA) treatments, where it is critical to ensure that injection of the 2nd phase of the treatment is performed with the fracture closed on the unetched areas, as the 2nd phase consists of stages comprised of very strong acid, whose purpose is to create a wider fracture by additional etching of the etched areas of both fracture faces, as shown at the bottom of Figure 9b.

Figure 9.

(a) Carbonate formation sample tested with the rotating disk test showing differential etching by acid. (b) Fracture profile showing face of fracture with etched and un-etched areas, and top view showing fracture closed on the un-etched areas.

3.2 Categories of screenouts

Screenouts were classified by Massaras et al. [10] according to pressure behavior into two categories: (1) screenout with gradual pressure behavior, and (2) screenout with abrupt pressure behavior. A third category called tip screenout (TSO) is added here.

Screenouts with gradual pressure behavior are caused by the inability of the proppant to pass through the near wellbore area. The pressure increases gradually in a concave upward fashion. Screenouts with abrupt pressure behavior are caused by a small amount of proppant which does not enter the fracture, but is diverted and falls to the bottom of the wellbore, and slowly fills the sump interval below the perforations. Once the sump interval is filled-up, the perforated interval begins to fill-up slowly, and once the entire perforated interval has been completely blocked by proppant, an abrupt pressure increase is noted. The pressure behavior of a wellbore screenout (WSO) is difficult to miss as the pressure increases in an extremely rapid manner, and in a few seconds it can exceed the safe-operating limits of the wellbore and wellhead equipment. Safety devices are always set to shut down the pumps before the safe-operating limits are reached, Figure 10 illustrates a WSO which caused an abrupt shut-down, making it impossible to execute an incremental step displacement. The pressure decline during the shut-in interval wasn’t gradual but rapid, because both the perforations and a long interval above the perforations were blocked, impeding communication with the fracture (and the reservoir).

Figure 10.

Wellbore screenout on a vertical well. Reducing the rate in steps does not control the screenout, it simply delays the inevitable.

Contrary to popular belief TSO does not occur during fracture treatments on conventional sandstone or unconventional shale formations as the leak-off rate is very low. TSO occurs only when fracturing high permeability unconsolidated sandstone formations, where a fracpack—essentially a planned and designed SO—treatment is standard, as such, it will not be covered here as it is not of interest as far as SD and SA is concerned, Barree [4, 5], and Barree et al. [6] provide extensive discussion and explanation on TSO.

3.3 Causes of screenouts

The causes of screenouts as per Cleary et al. [7] are: (1) NWB friction—a primary cause, and (2) perforation friction—a secondary cause. Massaras et al. [10] provide extensive detailed discussion on additional secondary causes of screenouts, (3) deviatoric stress (high differential stress), (4) non-compliant geologic formations, (5) multiple fractures, (6) segmented en-echelon fractures, (7) backstress due to depletion of reservoir pressure and (8) fracture tip dilatancy.

The advent of fracturing unconventional shale reservoirs, dictates that a ninth cause of SO is added, namely, Proppant Pack Build-up (PPB). This cause was originally investigated by Kern et al. [1], who determined that when proppant is placed in a narrow fracture with linear fluids, a proppant mound is formed near the wellbore with the proppant injected early, as shown in Figure 11a. Subsequently, three vertically stacked zones are formed. Zone 1 consists of an immobile proppant mound at the bottom, zone 2 is a high velocity mobile proppant bed in the middle, and zone 3 consists of clean fluid flowing at the very top, as shown in Figure 11b. The proppant moves in zone 2 at the top of the immobile bed similarly to sand carried by the wind on top of a sand dune, where the sand stays close to the top of the sand dune. The height of the proppant mound can increase as time elapses, to a point where the entire perforation zone could become completely blocked.

Figure 11.

Prop pack build-up in a fracture at equilibrium flow conditions. Adopted and modified from Kern et al. [1].

A PPB obstruction occurs on hydraulic fracturing treatments of unconventional shale formations where the formation permeability is low, and the treatments are placed with linear fluids, very small mesh size proppant, low proppant concentration, and very high injection rate. The linear fluids have low viscosity, as only High Viscosity Friction Reducer (HVFR) is used as gelling agent, so, they do not carry the proppant very far in the fracture. If the proppant mound is allowed to grow, it can block the entire perforated interval.

To avoid complete blockage of the perforated interval and a SO, sweep stages are incorporated in the proppant schedule of the treatment, as shown in Figure 7. The sweep stages do not contain proppant, and do not have be very large as they are very efficient in rapidly reducing the height of the proppant mound, and opening up the mobile bed and clean fluid zones (zones 2 and 3 in Figure 11b).

3.4 Theory of the step-down test

The fluid velocity and pressure profile through the tubing, perforations, and near-wellbore is analogues to that of a Venturi Flow Meter (VFM). See Figure 12. Note that the downstream portion of the VFM has been modified to a parallelogram shape, to represent the near-wellbore area of the fracture. The velocity of the fluid through the throat (narrow section) of the VFM increases, and the pressure decreases, in order to satisfy the law of conservation of energy. Furthermore, the tubing and the perforations have tubular shaped flow streams, which exhibit a flowrate squared f(Q²) behavior, while the near-wellbore area has a between-two-plates shaped flow stream, and exhibits a square root of flowrate (f(√Q) or f(Q^1/2)) behavior.

Figure 12.

Diagram of venturi flow meter with the downstream section modified to represent the NWB area.

Based on the above, the values of the friction components and of the Measured Total System Friction (MTSF) are calculated as follows:

$$\mathit{\text{Perforation friction}}=C_{\mathit{pf}}\ast Q^{\beta 1}$$E2

$$\mathit{NWB}\mathit{\text{friction}}=C_{\mathit{nwb}}\ast Q^{\beta 2}$$E3

$$\mathit{\text{Total}}\mathit{FEF}\mathit{\text{friction}}=C_{\mathit{pf}}\ast Q^{\beta 1}+C_{\mathit{nwb}}\ast Q^{\beta 2}$$E4

$$\mathit{\text{MTSF}}=\mathit{\text{Wellbore Friction}}+C_{\mathit{pf}}\ast Q^{\beta 1}+C_{\mathit{nwb}}\ast Q^{\beta 2}$$E5

Where, C_pf = constant of perforation friction, C_nwb = constant of NWB friction, Q = injection rate, β1 = ideal exponent for flow in a tube usually = 2 but it can vary. β2 = Ideal exponent for flow between two plates, usually = 0.5 but, it can vary. The wellbore friction is calculated as per standard equations found in University level Fluid Mechanics textbooks, and as presented by Massaras et al. [9].

As per Massaras et al. [9], β2 can be less than 0.5 for conditions of high-stress and tough rocks. Enhanced FEF analysis [9] solved the issue of the β exponents long ago, as both are not fixed but variable values, which are allowed to be whatever is required to obtain a match between MTSF and CTSF, as both the perforations and the NWB area are not well-defined geometric shapes, nor do these shapes remain constant as proppant-caused erosion modifies their shape and increases the flow area as increasing amounts of proppant passes through.

Based on bottomhole SDT data—recorded with DH gauges—Mondal et al. [13, 14], have presented a modified FEF analysis which verifies that a near-wellbore friction exponent of β = 0.5, overestimates perforation friction from SDTs, and that the value of β—after proppant slurry has been placed through the perforation and NWB area can vary from 0.25–1.0, due to erosion of the perforations—and also due to erosion of the NWB area. Roberts et al. [16, 17] present images captured with downhole video, which show rounding-off of the perforation entrances, and enlargement of the perforation diameter due to proppant-caused erosion. It has been calculated that the flow area of the perforations can be enlarged by as much as 40%.

It is well-known that the coefficient of discharge of new perforations is 0.65, and that proppant erosion rounds-off the entrances of the perforations and increases the coefficient of discharge to between 0.80 and 0.90, so, all commercially available Fracture Propagation Simulators (FPS) have 0.85 as the default coefficient of discharge.

Flowrate squared f(Q²) plots concave upward as shown in Figure 5a bottom, while square root (f(√Q) or f(Q^1/2) plots concave sideways as shown Figure 5b bottom. Analysis of SDT data takes advantage of the difference in plotting behavior to separate the total fracture entry friction (FEF) (Eq. (4)) into the two constituent components: perforation friction and near-wellbore friction, as shown at the bottom of Figure 5a, and b, and at the right side of Figures 13 and 14.

Figure 13.

Enhanced FEF analysis with low NWB friction and low MR.

Figure 14.

Enhanced FEF analysis with high NWB friction and high MR.

3.5 Sizing minifrac and diagnostic fracture injection tests (DFIT) correctly

It is highly recommended to make the minifrac fairly large in order for the fracture to grow very little during the very short duration of the SDT. This results in introducing negligible internal fracture friction into the equations as shown in Figure 16, interval 2.

Eq. (6) can be used to estimate the injection rate required to initiate and propagate a fracture. This rate is usually much lower than the design injection rate.

$$q_{i,\mathit{max}}=\frac{4.917\ast {10}^{-6}\mathit{kh}\left[\left(\mathit{FG}\ast D\right)-∆P_{\mathit{\text{safe}}-}p\right]}{\mathit{\mu \beta }\left(\mathit{ln}\frac{r_e}{r_w}+s\right)}$$E6

Where, q_i,max = injection rate (bbl/min), k = formation, permeability, h = formation thickness, FG = fracture gradient (psi/ft), D = depth at mid perfs (ft), Δp_safe = pressure safety margin (psi), p = reservoir pressure (psi), μ = viscosity injection fluid (cp), β = formation volume factor (dimensionless), r_e = drainage radius (ft), r_w = wellbore radius (ft), s = skin (dimensionless).

A quick way to estimate the volume of fluid to be injected for a minifrac or a DFIT is to use the simplified fracture dimensions presented in Figure 15.

Figure 15.

Dimensions for estimating minifrac volume.

The following two examples show how to size the minifrac for a conventional sandstone formation, and the DFIT for an unconventional shale formation:

Example 1: Minifrac volume calculations for conventional sandstone formation.

From Eq. (6) it is estimated that the minimum injection rate to initiate fracturing and propagate the fracture is 4 bpm. It has been decided that the design rate for the mainfrac will be 6 bpm. Selecting an aspect ratio of fracture height to fracture half-length equal to 2, we have: fracture height = 75 ft., fracture half-length = 150 ft., out of zone fracture growth = 10 ft., and, an average fracture width = 0.5 in. and, assuming fluid/frac efficiency of 45% (the percentage of injected fluid volume remaining in the fracture at the end of pumping the minifrac). The calculations are as follows:

$$\text{Fracture face area},\mathrm{A}=\left[\left(\mathrm{L}\text{ft}.\times \text{H1 ft}\right)+2\times \left(\left(\mathrm{L}\text{ft}.\times \text{H2 ft}\right)/2\right)\right].$$

$$\text{Fracture face area},\mathrm{A}=\left[\left(150\text{ft}.\times 75\text{ft}\right)+2\times \left(\left(150\text{ft}.\times 10\text{ft}\right)/2\right)\right]=\mathrm{12,750}\text{ft2}.$$

$$\text{Fracture volume},\mathrm{V}=\left(\text{Wavg}/12\right)\times \mathrm{A}.$$

$$\text{Fracture volume},\mathrm{V}=\left(0.5/12\text{ft}\right)\times \mathrm{12,750}\text{ft2}=531.25\text{ft3}.$$

$$\text{Double for fracture wings}=2\times \mathrm{V}=2\times 531.25\text{ft3}=1062.50\text{ft3}.$$

$$\text{Adjust for efficiency}=0.45\times 1062.5\text{ft3}=478.13\text{ft3}.$$

$$\text{Convert to barrels}=478.13\text{ft}3/5.615\text{bbls}/\text{ft3}=85.2\text{bbls}.$$

$$\text{Pumping time}=85.2\text{bbls}/6\text{bbls}/min=14.2min.$$

Unconventional shale reservoirs have very low to ultra low permeability, and require extremely long shut-in intervals to reach pseudoradial flow regime as shown in Table 2. Thus, a DFIT is usually performed only on the very first stage—with minimal equipment on location—usually only one pumping unit and a water tank. For a DFIT, the length of the shut-in interval dependents on the formation permeability and the duration of the injection, two variables that influence to a great extent the time it takes to reach both fracture closure pressure, and pseudoradial flow regime, as shown in Table 2.

Injection duration (min)	Permeability (md)	Time to clossure pressure	Time to pseudoradial flow regime
5	0.1	0.2 h (12 min)	0.6 h (36 min)
20	0.1	10 h	30 h
5	0.01	2–3 days	6–8 days
5	0.001	4–5 days	12–15 days
5	0.0001	10 days	30 days

Table 2.

Estimated time required to reach time to closure and pseudoradial flow as a function of permeability.

It is imperative that pseudoradial flow be reached in order to be able to determine all the reservoir properties: permeability, flow regime, time to closure, closure pressure, fluid leakoff, effective fluid efficiency, instantaneous shut-in pressure (ISIP), fracture gradient, minimum horizontal stress, fracture extension pressure, maximum horizontal stress (approximate), stress anisotropy, pore pressure, and transmissibility.

For a DFIT, the required injection rate to initiate and propagate a fracture is very low as per the Darcy radial flow equation (Eq. (6)). Very low permeability reservoirs have a corresponding very high efficiency and a corresponding very low leak-off rate, thus, the volume to be injected during the DFIT will be very small—10 to 100 barrels—as very thin and very long fractures are created in very low and ultra low permeability reservoirs.

Example 2. DFIT volume calculations for unconventional shale formation.

From Eq. (6) it is estimated that the minimum injection rate to initiate fracturing and propagate the fracture is 2 bpm. It has been decided that the design rate for the DFIT will be 3 bpm. Selecting an aspect ratio of fracture height to fracture half-length equal to 2, we have: fracture height = 50 ft., fracture half-length = 100 ft., out of zone fracture growth = 5 ft., and, an average fracture width = 0.1 in. and, assuming fluid/frac efficiency of 95% (the percentage of injected fluid volume remaining in the fracture at the end of pumping the minifrac). The calculations are as follows:

$$\text{Fracture face area},\mathrm{A}=\left[\left(\mathrm{L}\text{ft}.\times \text{H1 ft}\right)+2\times \left(\left(\mathrm{L}\text{ft}.\times \text{H2 ft}\right)/2\right)\right].$$

$$\text{Fracture face area},\mathrm{A}=\left[\left(100\text{ft}.\times 50\text{ft}\right)+2\times \left(\left(100\text{ft}.\times 5\text{ft}\right)/2\right)\right]=5500\text{ft2}.$$

$$\text{Fracture volume},\mathrm{V}=\left(\text{Wavg}/12\right)\times \mathrm{A}.$$

$$\text{Fracture volume}=\left(0.1/12\text{ft}\right)\times 5550\text{ft2}=45.83\text{ft3}.$$

$$\text{Double for}2\text{fracture wings}=2\times 91.67\text{ft3}=91.67\text{ft3}.$$

$$\text{Adjust for efficiency}=0.95\times 183.33\text{ft3}=87.08\text{ft3}.$$

$$\text{Convert to barrels}=87.08\text{ft}3/5.615\text{bbls}/\text{ft3}=15.17\text{bbls}.$$

$$\text{Pumping time}=15.17\text{bbls}/\text{3bbls}/min=5.17min.$$

3.6 Correct procedure for the step-down test

The recommended and correct procedure for the SDT is to execute four equal steps of 10–15 s. duration each, as shown in Figures 5, 16 and 17. If the pressure does not stabilize in 15 s. most likely it will not do so, thus, there is no point in making the steps any longer. If all four steps cannot be made equal, then the first and fourth steps should be made equal in order to be able to calculate the MR without having to make proration adjustments mathematically.

Figure 16.

Fracture calibration procedures: 1, formation breakdown/calibration; 2, minifrac; 3, SDT; 4, before closure shut-in interval; and 5, after closure shut-in interval.

Figure 17.

The median ratio can be determined when the SDT is performed with a prescribed procedure of four equal and abrupt rate reduction steps. Rate must be brought to zero.

It is best to integrate the SDT at the end of the minifrac which usually follows the formation break down in the fracture calibration procedures, as illustrated in Figure 16. Formation breakdown/calibration consists of injecting very slowly a very small amount of clean linear fluid—usually the same brine that is already in the wellbore—until Formation Breakdown Pressure is noted, which is self-indicated by a small reduction in pressure. The injection rate is then increased to design rate to record the maximum surface pressure that will be experienced during the treatment. A very short shut-in interval follows in order to calibrate the formation. See Interval 1 in Figure 16.

The second injection procedure is called minifrac (interval 2 in Figure 16), and consists of injecting a reasonable amount of fluid so that the fracture will be large enough so that it will grow very little during the very small time interval of the SDT—30–45 s, Thus, small friction from the fracture will be introduced, as shown in Figure 18b. The SDT (interval 3 in Figure 16) is integrated at the end of the minifrac so, no extra fluids or time are required. So, practically there is no cost for the SDT. There is adequate time during pumping of the minifrac to easily configure four or eight pumps to pump at equal rates, then one or two pumps at a time can be taken out abruptly to enable corresponding abrupt pressure reductions. As soon as pressure stabilizes the next rate reduction step is implemented.

Figure 18.

Sizing minifrac and DFIT correctly.

Regarding the choice of four steps, three steps are too few, and five steps are too many. Four steps have distinct advantages as they:

1. Enable making plots with good curvature,

2. Provide the best data possible for calculating the MR,

3. Make it very easy to visually determine if progressively smaller pressure steps are apparent, something that is not clearly evident with a three-step SDT,

4. Eliminate the highly involved mathematical manipulation that would be required to make the rate and pressure reductions equal-rate equivalent, in order to calculate the MR, and,

5. By conducting the test in a consistent way, there is a basis for comparison.

The logic and reasoning of the procedure with four equal-step flowrate reductions is as follows: during the short duration of the SDT (30–45 s) all contributors to friction in the system remain unchanged except the flow rate. The wellbore system parameters that have contributing influence to friction and remain constant are: wellbore diameter, depth to perforations, perforation diameter, and the friction factors, along with fluid type, fluid density, fluid viscosity, etc. If the conditions are met that all contributors to friction in the system remain unchanged—except the flow rate—and that there are no restrictions to flow, then progressively smaller pressure reductions should be noted as shown in Figures 5a, 13, and 16, and the system friction is considered normal.

Since the duration of the SDT is very short, the conditions are always met that all contributors to friction in the system remain unchanged—except the flow rate—however there must exist restriction(s) to flow, if progressively smaller pressure reductions are not noted as shown in Figures 5b, and 14, and the system friction is considered abnormal. By the process of elimination the restriction(s) to flow can be located in either the perforation or the near-wellbore area, or both, with varying severity.

If the restriction severity is in the near-wellbore area, it is a serious issue as the available design modification or wellbore remediation options are limited. This is the reason why near-wellbore friction is the only cause of SO classified as primary. Perforation friction is not such a serious issue as the entrances to the perforations are completely rounded-off after a small amount of proppant (1–2 tons) has passed through each perforation as per El Rabba [15] and others. This is the reason why perforation friction is classified as a secondary cause of SO.

When perforating for fracturing, the most common perforation density is 6 spf, with 60 degree phasing, arranged in a helical configuration around the circumference of the perforating gun. As such, two of the six created perforations, which are opposite each other, are either lining up perfectly with the fracture, or they are offset by at most 30 degrees. This perforation configuration means that at most two of the six perforations will be taking fluid. So, given that during a fracturing operation less than one third of the perforations are open and taking fluid the total amount of proppant required to round-off the entrances of the active perforations is not very large. The main reason the proppant concentration is ramped-up gradually is to round-off the perforation entrances of the few active perforations before placing proppant at the maximum design concentration.

3.7 Performing the SDT in the pad

Most operators that are performing multistage PHF treatments via horizontal wellbores on unconventional shale formation conduct only a DFIT on the very first stage. They do not usually perform DFITs on subsequent stages, so the opportunity to perform a SDT does not arise. That does not mean a SDT cannot be performed, as the opportunity arises always with the placement of the pad stage.

It is the popular belief that reducing the flowrate to zero during the placement of the pad stage or of the proppant laden mainfrac stages will cause the fracture to close abruptly, collapse on itself, and create other problems, which obviously is not true, because, as per the balloon analogy, the internal hydraulic pressure—the net pressure—remains constant and holds the fracture open.

Performing a shut-in during the pad stage has been performed many times, sometimes inadvertently due to equipment failure, and sometimes on purpose to determine the effectiveness of proppant slugs—that sometimes are incorporated in the pad stage—in reducing FEF. No problems have been reported in the literature; therefore, there is no reason not to perform a SDT in the middle of the pad stage. The information to be gained would be very valuable during the subsequent placement of the mainfrac, and also as design consideration for the placement of subsequent PHF stages.

4. Simple screenout detection methods

The logic and reasoning of the three methodologies for SD is presented in the following sections: enhanced fracture entry friction analysis, median ratio methodology, and inverse slope methodology.

The first two methods use data from the step down test (SDT)—performed at the end of the minifrac—so the analyses are categorized as pre-mainfrac. The third method uses data from the mainfrac, so the analysis is categorized as post-minifrac or real-time.

4.1 Enhanced fracture entry friction analysis

Enhanced fracture entry friction (FEF) analysis as presented by Massaras et al. [9] is an improved version of the conventional FEF analysis originally presented by Cleary et al. [7]. Conventional FEF analysis is not as accurate as the three analysis methods presented in this chapter; due to various issues with software implementation, and with the assumptions made. Enhanced FEF analysis is not simple, as it requires a computer and software; however, it is included in this chapter because it has been implemented in all commercially available HFP Simulators with varying degrees of accuracy and success. The most accurate implementation is in the proprietary FEF Analyzer, as illustrated in Figures 13 and 14.

The objective of conventional and enhanced FEF analysis is to accurately determine the magnitude of the perforation friction and of the near-wellbore friction. The analysis calculations are performed three different ways, depending on the pressure gauges configuration:

1. If the FEF analysis is performed with data recorded at the surface, the maximum surface pressure prior to shut-down is selected, and the ISIP is subtracted to determine the MTSF as shown in Figure 5. All contributors to the calculation must be included.

2. If the analysis is performed with reflected bottomhole pressure data (sensed via deadstring) the pressure loss due to friction in the tubing is omitted as the fluid in the annulus is stationary, and vice versa.

3. If the analysis is performed with data recorded with bottomhole gauges run and positioned near the perforated interval either on wireline or on a bundle carrier the hydrostatic pressure is omitted as it already has been captured by the bottomhole gauge pressure measurement, and the wellbore friction loss is omitted, as it is not sensed by the gauges.

In all three cases the objective is to the isolate the fracture entry friction and split it into perforation friction and near-wellbore friction, as shown in the bottom of Figure 5a and b, and at the bottom right side of Figures 13 and 14.

For the first and most widely used case it is crucial to calculate the wellbore friction accurately, as in most cases it is more than 50% of the MTSF. If the wellbore friction is calculated wrong, the magnitude of the perforation friction and of the near-wellbore friction will be incorrect. The calculation procedure is by trial and error until a match is obtained between the measured and the calculated parameters, as there are several unknowns which preclude a direct mathematical solution. Making sure the analysis has been performed correctly is made simple by following the following three rules:

1. The plot of pressure vs. rate must start at coordinates 0, 0, because the rate is brought to zero at the end of the SDT, thus, at the instant of ISIP the friction in the system is zero. See Figures 5, 13 and 14.

2. A summation of the three friction components comprising MTSF and Calculated Total System Friction (CTSF) must be equal; otherwise the law of conservation of energy is being violated, and the calculated friction values are wrong. See Figures 5, 13 and 14.

3. The match of (1) MTSF and CTSF, and (2) measured FEF and calculated FEF must be made at all steps of the SDT, otherwise the distribution of friction into the three constituent components will be incorrect. See Figures 5, 13 and 14.

Enhanced FEF analysis methodology bypasses limitations encountered in Conventional FEF analysis with innovative methods: (a) appropriately corrected fluid friction factors, (b) non-uniform perforation discharge coefficient, (c) application of Maximum Drag Reduction (MDR) asymptote, and (d) matching of the MTSF with the CTSF at all rates of the SDT.

Enhanced FEF analysis uses data from a SDT and calls for a prescribed simple four-step procedure for the SDT. It is very important to visually determine if progressively smaller pressure reductions are noted as the flow rate is reduced. Because all parameters remain constant during the very small duration of the SDT (30–45 s), a deviation from progressively smaller pressure reductions means that there is a restriction in the flow, and by the process of elimination the restriction can only be either in the perforations or in near-wellbore area, or both. Determining which FEF component dominates (is larger) is evident for the way the SDT data plots, i.e., concave upwards or concave sideways, as outlined in Section 2.2, entitled “Stagnation pressure.” Typical output from enhanced FEF analysis is shown in Figures 13 and 14.

The guidelines shown in Table 3 are used to interpret the results of enhanced FEF analysis. For a detailed presentation on enhanced FEF analysis which includes all the mathematical equations the reader is referred to Massaras et al. [9].

General decision control guideline
(Based on ranges of magnitude of FEF components)
Perforation friction		Near-wellbore friction		Remarks
(bar)	(psi)	(bar)	(psi)
0–15	0–218	0–15	0–218	Very low probability of screenout
15–30	218–435	15–30	218–435	Caution required
30–50	435–725	30–50	435–725	Screenout highly probable
>50	>725	>50	>725	Screenout almost certain

Table 3.

Enhanced FEF analysis interpretation guidelines.

4.2 Median ratio methodology

The median ratio (MR) technique was developed empirically with data from about 2500 SDTs carried out at numerous geographically and geologically diverse locations around the world. The methodology stipulates that the SDT must be performed with a specific four equal-step procedure. The MR is defined as:

$$\mathit{MR}=∆P4÷∆P1$$E7

where ∆P4 and ∆P1 are pressure reduction changes as shown in Figure 17.

It is easy to determine which of the FEF components dominates: since the fluid goes through the perforations first and through the near-wellbore area last, if the first pressure reduction, ΔP1, is larger compared to the last pressure reduction, ΔP4, the perforation friction will be larger than near-wellbore friction. Larger means that ΔP1 is larger than two times ΔP4 (ΔP1 > 2*ΔP4). By the same logic, since the fluid goes through the near-wellbore area last, if the last pressure reduction, ΔP4, is fairly large compared to the first pressure reduction, ΔP1, near-wellbore friction is larger. Larger means that ΔP4 is larger than half of ΔP1 (ΔP4 > 0.5*ΔP1).

If the MR is in the low range (MR = 0.2–0.5—see Table 4) a SO is very unlikely. Most screenouts occur when the MR is in the high range (MR = 0.5–1.0). The magnitude of the FEF components is best determined with Enhanced FEF analysis as per discussion in the immediately previous section, and as per Massaras et al. [9].

Median ratio classification by range
MR ranges	Occurrence classification	Comments
0.2 ≤ −0.3	Extremely low	Screenout very unlikely
0.3–0.4	Low
0.4–0.5	Normal
0.5–0.6	Borderline	60% chance of screenout
0.6–0.7	High	70% chance of screenout
0.7–0.8	Very high	80% chance of screenout
0.8–0.9	Extremely high	90% chance of screenout
09–1.0	Ultra high	100% chance of screenout
1.0–1.1	Mega high	Screenout certain to occur; Extremely rare to manifest
1.1–1.2	Giga high
1.2–≥1.3	Tera high

Table 4.

Median ratio classification by range.

Massaras et al. [10] stipulate that that SOs occur when the MR is equal-to or greater-than 0.5 (MR ≥ 0.5) while simultaneously, the magnitude of NWB friction—determined with enhanced FEF analysis—is equal-to or greater-than 30 bar (≥435 psi). See Table 3. However, it is not an absolute requirement, to use the MR and NWB friction concurrently, because if the MR is ≥0.5, the magnitude of NWB friction is almost always greater that 30 bar as shown in Table 1 of Massaras et al. [10]. Thus, the MR value by itself will be sufficiently satisfactory for analysis and correct interpretation. The MR is a powerful diagnostic criterion of Proppant Admittance (PA), and with a prediction accuracy of 95% it is very useful for the successful design and placement of safe and effective PHF treatments.

4.3 The inverse slope methodology

The Inverse Slope methodology introduced by Massaras et al. [12] was developed empirically by applying the methodology in real-time on about 1000 mainfrac treatments which were carried out in diverse geographic and geologic conditions all around the world. The method uses smart tangent lines on the surface pressure plot for data visualization, and, is capable of providing advance warning of imminent SO events in real-time, and facilitates the ability to exercise effective decision control for early termination or extension of a PHF treatment. After the pad stage—which consists of clean fluid (no proppant is included)—has been pumped, proppant addition begins, which causes the hydrostatic pressure in the wellbore to gradually increase, and the surface pressure to gradually decrease at a Negative Slope. See Figures 19–21. When the entire wellbore is full with slurry which is at maximum proppant concentration (plateau)—the hydrostatic pressure has reached maximum magnitude—the surface pressure stops decreasing and begins to flatten out, and may remain constant for a while.

Figure 19.

Typical inverse slope analysis.

Figure 20.

Extending a hydraulic fracture treatment.

Figure 21.

Fully worked out MR analysis calculations and inverse slope smart tangent line analysis.

Eventually as propped stages—which are laden at maximum proppant concentration and proppant of larger mesh size—travel via the wellbore and arrive at the near wellbore area, the surface pressure will commence to increase gradually—due to increased friction—at an Inverse Slope which has inclination equal with the Negative Slope, but on a positive slope. See Figures 19–21. The rate of inclination of the Inverse Slope depends greatly on the presence, size, and concentration of proppant in the near-wellbore area of the fracture, and to a much lesser extent to the presence of proppant deep (far-field) in the fracture. This is evident by the pressure increase noted when a higher proppant concentration or larger mesh proppant arrives at the near-wellbore area.

As long as the surface pressure plots parallel with the Inverse Slope there are no issues, and the treatment can continue to be placed as per design or it can even be extended. The start of the deviation of the surface pressure from the Inverse Slope tangent line and onto a Deviated Slope marks the start of a SO Advance Warning. Plotting of the surface pressure along the Deviated Slope does not last very long, so it is of critical importance that the displacement (flush) operation be started immediately. It is not apparent from the surface pressure plotting on a Deviated Slope that a SO is occurring downhole, however, when the surface pressure plotting deviates from the Deviated Slope a SO behavior is noted in the bottomhole pressure, which is not visible at the surface (unless a bottomhole pressure gauge is being used to monitor the pressure).

If displacement starts at the Start of SO Advance Warning, a planned SO with the desired proppant amount left in the wellbore is ideal and possible, as great net-pressure gain will be noted and the fracture at the near-wellbore area will be very wide. Net-pressure gain is defined as the difference between the ISIP at the end of the minifrac and the ISIP at the end of the mainfrac. By universal agreement placement with a planned SO would be considered perfect placement, and is associated with better that expected long-term production rates. Achieving a planned SO is extremely difficult to achieve but has been done.

5. Extending a hydraulic fracture treatment

Ideally, the design of PHF treatment should be made with an option to extend it, if there are enough materials available on location—fluids and proppants. Extending a fracture treatment is not an issue when multiple stages are planned to be placed in the well, as extra materials are usually stocked on location for use in the subsequent stages.

Most fracture treatments are placed with deficient (not aggressive) designs, and do not develop enough net-pressure gain—the difference between the ISIP at the end of the mainfrac and the ISIP at the end of the minifrac. In fact some treatments are completed with a net-pressure loss, because the fracture grew vertically—up or down—instead of horizontally which is preferred. Vertical growth without net-pressure gain means that the proppant may have settled in the fracture which has grown into lower bounding formation and will not be placed opposite and all across the target production zone. This will prevent a PHF from being considered a complete success, as the production rate on a long-term basis will not be as expected. At the end of many PHF treatments it becomes obvious that extension of the treatment is required in order to achieve sufficient net-pressure gain, but it is not possible to extend as the extra materials required are not available or loaded.

The Inverse Slope method can be used in real-time to extend a hydraulic fracture treatment without risking a SO event. The extent to which a treatment can be extended is determined by utilizing the stabilized pressure reached during pumping of the pad stage—when only clean fluid is being pumped. A smart line, termed Pressure Stabilization Line is extended horizontally from the stabilized pressure as required, and serves as delineation of the upper limit the surface pressure is allowed to reach while placing the proppant laden stages.

The treatment can be extended for as long as the surface treating pressure is not deviating from the Inverse Slope without risking a SO. See dashed black line at the top right side of Figure 20. The reason that this is the maximum allowable pressure that can be reached is to prevent using-up all or part of the Hydrostatic Clearance. Note that in the case shown the maximum allowable pressure was 11,000 psi which was easy to reach. See dashed line at the top right side of Figure 20.

Hydrostatic Clearance is experienced during the displacement (flush) stage, and is the result of using fluids with mach lower specific gravity than the fluids used for the proppant laden stages. This causes the hydrostatic pressure in the wellbore to decrease rapidly and corresponding rapidly increasing surface pressure is noted. The consequence of using-up the Hydrostatic Clearance is that maximum allowable pressure of the well equipment will be reached and an unplanned shut-down will become necessary. Well equipment that have pressure limitations are: wellhead, wellhead isolation tool (tree saver), tubing, casing, packers, etc. An unplanned shut-down also means that excess proppant will be left in the wellbore, which will necessitate a wellbore cleanout operations with CTU, a time consuming and costly operation. Extending a PHF treatment results in a much better placement of the proppant pack, i.e., opposite and all across the target production zone. It results in a wider fracture as evidenced by, the net-pressure gain, and the higher than expected long-term sustained production rates.

6. Analysis of example case

The PHF treatment presented in Figure 21 is analyzed fully with both MR analysis and Inverse Slope analysis. The SDT was performed at the end of the Formation Breakdown procedure, (see Figure 16) and the Minifrac was omitted. If the Minifrac is to be omitted, a larger fluid volume should be injected prior to performing the SDT as per logic presented in Section 3.6 entitled “Correct procedure for the SDT.” The SDT itself was performed correctly with four equal rate reductions to zero flow rate.

The magnitude of the MR is calculated to be 1.03 as shown in Figure 21a, which is classified as mega high. See Table 2. A magnitude of the MR above 1.0 is not noted very frequently, and when it does manifest itself it means that the tortuous pathways that connect the perforations to the fracture—at the near-wellbore area—are very narrow. Since there have been SO events with 100 mesh sand when the MR was greater than 1.00 it is postulated that the width of these pathways may be less than 0.0177 in. or 450 microns which is equal to three diameters of 100 mesh sand.

Inverse Slope analysis of this PHF treatment in real-time would have given an advance warning of about 2.6 min of the impeding SO event. See lower right side of Figure 21b. This is significant, as the treatment was being placed down a 5.5 in. OD casing, (no tubing) and the depth to the perforations was 4585 m. The displacement volume was 49.0 m³, and the rate 3.0 m³/min. This gives a wellbore travel time of 16.33 min. So, time was of the essence. At the time of the SO only 30 m³ of the 49.0 m³ had been displaced, leaving in the wellbore 19.0 m³ of proppant laden slurry. During the 2.6 min prior to displacement the average BH proppant concentration of the slurry was 690.0 kg/m³ and the volume pumped during 2.6 min was 7.8 m³ (3.0 m³/min _* 2.6 min). This resulted in leaving an extra 5.4 tons of proppant in the wellbore (7.8 m³_* 690.0 kg/ m³) which could have been avoided. Since the wellbore travel time was 16.33 min (49 m³/3.0 m³/min), the advance warning of 2.6 min was not adequate. In the end a total of 13.5 tons of proppant were left in the wellbore, and it took at least 24 h to lift it out with a pre-spotted standby CTU before proceeding to fracture the next stage.

Since the MR was in the mega high MR range, it caused the SO pressure to increase very rapidly and there was not adequate time to react. When the MR is Borderline (0.5–0.6) and above there is adequate time to react to the onset of a SO event, as plotting on the Deviated Slope will be longer, since it takes longer for the SO event to manifest itself at the surface. There are occasions, when a Deviated Slope does not develop at all. In such cases there is not much the location engineer can do.

So what is the location engineer supposed to do when faced with high and ultra high MR? The answer is the choices are limited, as the well is completed, all the equipment and materials are ready and waiting on location, and the service company charges are running. The best choice is to re-perforate a small section (≤1 m) in the zone with the highest permeability. Never re-perforate the entire zone, as the objective is to create a path of least resistance. Other choices would be to insert a proppant slug in the pad, or re-design with a lower proppant concentration and hope for the best. Many other choices are listed in Section 7.

It is very rare that PHF treatments are placed via casing, as most of the time placement is via tubing which has smaller diameter than casing, thus, the displacement volume and the wellbore travel time are much smaller—in the range of 5 min. Thus, the Inverse slope real-time methodology provides adequate advance warning to place safe and effective PHF treatments. Safe means to avoid a SO, effective means to increase the net-pressure gain or even SO with the designed amount proppant left in the WB—an extremely difficult feat to accomplish, but it has been done.

7. Screenout avoidance

7.1 Design modification and wellbore intervention procedures

Numerous publications over the past 5 years have presented SD methodologies; however, none has presented any SA procedures. Instead, many other publications have presented numerous SA procedures over the past 70 years, which have been categorized into two groups: 1. design modification procedures, and 2. wellbore intervention procedures. The former are incorporated into the preliminary design—made in the office, or into the final design—made on-location after the minifrac, and do not require a trip into the wellbore prior to or after the minifrac, while the later require a trip into the wellbore with various tool assemblies. These procedures can be implemented individually or in combination and have proven effective in preventing SO events, achieving effective SA; and placement of safe and effective PHF treatments.

7.2 Design modification procedures

There are many design modification procedures which can be used to minimize the chances of a SO occurrence, or to eliminate those chances altogether. They are briefly discussed as follows:

1. Proppant slug technique involves placing small proppant laden stage (usually 2000 lbs., or 1000 kg) at low 2.0 ppg maximum proppant concentration and displacing it to the perforations. Cleary et al. [7] call for shutting-in on the proppant at the perforations in order to trap proppant in both narrow and wide fractures in the near-wellbore area. The concept is when pumping restarts the fluid will take the path of least resistance through the widest fracture and pinch the other narrow fractures shut, thus, maintaining open a dominant wide fracture. Most operators simply over-displace the proppant slug, as they are afraid that the proppant left in the wellbore during the shut-in interval may settle across the perforations, block them, and prevent restarting of pumping operations.

2. Proppant schedule redesign involves using proppant of smaller mesh size, ramping-up the proppant at a gentler slope, lowering the maximum proppant concentration, and shortening the proppant plateau (the time duration at maximum proppant concentration).

3. High energy-large diameter perforations can have density of 4–8 spf in specific circumferential orientations around the wellbore—with phasing options of 0, 60, 90, 120, or 180 degrees (resulting in perforations that are pointing in one to six different directions). This low perforation density allows for usage of large high energy perforating charges that create perforations with large Entrance Hole Diameter (EHD)—ranging from 0.22 to 0.49 in., that also provide deep penetration into the formation ranging from 14.4 to 59.2 in. Cleary et al. [7], advocate usage of high energy perforating systems, which are beneficial on one hand with the large perforation diameter and detrimental on the other hand as the High Energy from the explosives favors creation of complex and narrow fracture networks in the near-wellbore area.

4. Extreme overbalance perforating can be performed in two ways. The first is used if there are existing perforations, and involves setting a rupture disk at tail-end of the tubing. The disk is set to rupture at a specific differential pressure. The wellbore fluid is lifted out of the wellbore or displaced into the formation with nitrogen gas, which lowers the hydrostatic pressure. Subsequently, the tubing pressure is increased by injecting nitrogen gas until the disk ruptures. The second way is for wells without perforations. The well is lifted dry with nitrogen, and the perforating guns are run. The perforating guns are set across the zone, and the wellbore is pressurized with nitrogen gas. When the desired pressure is reached the perforating guns are fired. These methods are discussed in detail by Petitjean et al. [30].

5. High viscosity—high rate fracture initiation involves spotting a high viscosity fluid across and above the perforations, usually with a CTU, and initiating fracturing at a very high rate. As fractures initiate along the wellbore and turn towards the preferred fracture plane the radius of curvature may be small and the fracture may be pinched shut. Initiating pumping with high viscosity fluid and high rate, the radius of curvature is large and the fracture wider as described by El Rabaa [31], and Yew et al. [32].

6. Hydra-jet fracturing (HJF) is a type of hydraulic fracturing that uses high-pressure water jets to create fractures in the formation. The jet nozzles are conveyed either on tubing or CTU. The jet is at very low pressure due to the very high velocity, thus, fluids are drawn into the fracture due the differential pressure. This technique is discussed by Surjaatmadja et al. [33] and by many other authors.

7. Jetted slotted perforations are created by hydrajetting with sand-laden fluid which has a concentration of 2 ppg. The jets are configured 180 degrees apart to enable centralization in the casing, and the tubing or CTU is reciprocated 1 ft. to create slotted perforations. The tubing is rotated 30–45 degrees to cover the circumference of the casing. Subsequently the tubing is pulled up to leave 1 ft. gap, and start creating a 2nd set of slotted perforations. It is preferred and highly recommended to rotate the tubing so that the 2nd set of slotted perforations is staggered with respect to the 1st set, in order not to cause excessive weakness in the casing. Chernyshov et al. [34] describe the development of this technology.

8. Limited entry (LE) perforating calls for perforating a very small section of the production interval say 5 ft. out of 100 ft. The objective is to perforate for fracturing and not for production, the crucial aspect of which is to connect the wellbore to the fracture. Perforating a small interval minimizes the creation of multiple fractures and SO events. Lestz et al. [35] give more details.

9. eXtreme Limited Entry (XLE) perforating calls for limiting the number of perforation and sizing the perforation diameter in a way that perforation friction exceeds 2000 psi. This procedure results in a very small amount of perforations. For example in a horizontal well where many stages are placed, there could be up to 12–15 perforation clusters with one perforation per cluster. The sizing procedure takes into account the fact that the perforation edges will be rounded-off, and the perforation diameter will enlarge as increasing amounts of proppant passes through them. Weddle et al. [36] and many other authors discuss XLE in great detail.

10. Waterfrac (WF) treatment involves fracturing with large amounts of fluids pumped at very high rates. If any proppant is used the mesh size is very small and the concentration very low, conditions that ensure that no SO event can occur, as outlined by Woodworth et al. [37] and many other authors.

11. Zero-degree phased perforating (ZDP) refers to a pattern of perforations that are not oriented but all are pointing in the same direction, i.e. they have a zero-degree phasing. Usually a small interval is perforated in order to originate as few fractures as possible in an attempt to minimize fluid loss and near-wellbore friction. Even though the perforations may not align with the fracture, the technique has shown to be effective in reducing SO events as documented by Stadulis [38].

12. Mini-wellbore (MW) consists of jetting with sand-laden fluid a mini hole in a horizontal wellbore which is similar to a perforation of very large diameter. The mini hole is usually oriented vertically upwards thus it originates from the uppermost section of the wellbore. This configuration allows the fracture originating from the mini hole to align itself perfectly with the maximum horizontal stress and be perpendicular to the minimum horizontal stress. This configuration minimizes both near-wellbore friction and perforation friction as presented by Abass et al. [39] and Surjaatmadja [40].

13. Oriented perforating (OP) is applied when the orientations of the minimum and maximum horizontal stresses are known. The perforations are oriented with the maximum horizontal stress, and the phasing is 180 degrees. This ensures that perforations are pointed at both wings of the fracture. This configuration minimizes both near-wellbore friction and perforation friction as per Bou et al. [41], and Abass et al. [42].

14. Downhole mixing fracturing (DMF) involves pumping slurry of extremely high proppant concentration (18–22 ppg) down the tubing, while clean fluid is pumped down the annulus, where the two flow steams meet, and mix together near the perforated zone. This allows for rapid adjustment of the downhole proppant concentration based on real-time downhole conditions. This enables effective fracture placement at lower costs—due to low horsepower requirements, and better well performance—due to extremely good connection between the wellbore and the fracture, resulting from the intentional SO at the end of each stage. Massaras et al. [43], van Gijtenbeek et al. [44], and many other authors describe this technology in great detail.

15. Incremental step displacement starts at the onset of a SO event, and as the name implies involves reducing the displacement rate in steps, as shown in Figure 22. The reduced injection rate decreases the MTSF, which in turn lowers the surface pressure, and prevents it from reaching the maximum allowable pressure (MAP). This allows for a portion, or for the entire displacement fluid volume to be pumped into wellbore. In the best case 90–100% of the displacement fluid volume is pumped. If less than 90% of the displacement fluid volume is pumped, most likely a CO operation with CTU will be required.

Figure 22.

Incremental step displacement extends the length of the displacement stage.

7.3 Wellbore intervention procedures

Wellbore Intervention procedures require a trip with tool assemblies into the wellbore, after the minifrac: wireline or CTU conveyed perforating guns (for re-perforating), CTU conveyed tool assemblies (for wellbore cleanout, jetting, etc.), or slickline conveyed tools (for tagging, drift diameter determination, callipering, etc.). They are briefly discussed as follows:

1. Re-perforating (RP) (a small interval) is the most used wellbore intervention procedure due to the wide availability and low cost of wireline perforation units on short notice. The procedure calls for re-perforating a small interval (1–3 ft. or 0.3–1.0 m) over the highest permeability zone in the target formation. The recommended phasing for both perforating and re-perforating is 60 degrees. Because the re-perforations are bound to be slightly offset circumferentially from the original perforations, a better line-up of the perforations with the fracture is possible. In addition, the fluid is most likely to take the path of least resistance offered by the re-perforated high permeability zone, thus, assisting in initiating a single dominant fracture, which is bound to be wider than many narrow fractures.

   The geometry of inclined wellbores in combination with long perforated intervals favors the initiation and creation of multiple fractures. Thus, re-perforating only a very small interval is highly recommended. Re-perforating a long interval or the entire interval is strongly discouraged.

2. Wellbore cleanout (WCO) after a SO event is often required. The SO event can occur either when a proppant slug has been inserted in the pad stage or during the mainfrac. A SO event with a proppant slug is equivalent to applying the proppant slug technique previously mentioned. The result of SO event is that when pumping restarts the fluid will take the path of least resistance through the widest fracture and pinch the other narrow fractures shut, thus, maintaining open a dominant wide fracture. WCO is performed almost exclusively with CTU, requires a large amount of gel for lifting the proppant incrementally in small amounts; which is a costly and time consuming.

8. Conclusions

1. Most of the SD methods that have been developed are briefly presented. They are neither widely available nor widely used.

2. The three very simple SD and SA methods presented are: Enhanced FEF analysis, Median Ratio, and Inverse Slope.

3. The three very simple SD methods presented are: readily available, simple to use, and inexpensive.

4. Enhanced FEF analysis and Median Ratio use data from the SDT which is performed during the minifrac, so, they are categorized as pre-mainfrac. The Inverse Slope uses data from the mainfrac, so, it is categorized as real-time.

5. Median Ratio provides adequate predictive accuracy by itself, so, it is not absolutely necessary to use it concurrently with enhanced FEF analysis.

6. Median Ratio has a predictive accuracy of 95%.

7. The Inverse Slope method can be used to detect an imminent screenout and provides adequate warning in order to initiate displacement (flush) and avoid it.

8. The Inverse Slope method can be used to extend a PHF treatment.

9. Inverse Slope has a predictive accuracy of 95%.

10. There are many design modification procedures and wellbore remediation procedures that can be implemented in order to achieve SA. They can be applied either during the planning and design phases, or post minifrac during the execution phases.

11. The design modification procedures can be incorporated into the design or in real-time during the mainfrac individually or in combination and have proven to be effective in SA by reducing SO events, and achieving safe and effective placement of PHF treatments.

12. The wellbore modification procedures, although expensive and time consuming, can be used in difficult situations to modify the completion configuration in order to minimize NWB friction and make placement of PHF treatments to completion possible.

Nomenclature

CFA

Closed Fracture Acidizing

CTSF

Calculated Total System Friction

CTU

Coil Tubing Unit

CO

cleanout

DH

downhole

DMF

downhole mixing fracturing

DFIT

Diagnostic Fracture Injection Test

XLE

eXtreme Limited Entry

EHD

entrance hole diameter

FEF

fracture entry friction

FPS

fracture propagation simulators

FT

fracture toughness

Frac

fracture

HFP

hydraulic fracture propagation

HJF

hydra-jet fracturing

HVFR

high viscosity friction reducer

ISIP

instantaneous shut-in pressure

LE

limited entry

MTSF

measured total system friction

MAP

maximum allowable pressure

NN

neural network

NP

net-pressure

NPP

net-pressure plot

NWB

near-wellbore

PA

proppant admittance

PDG

pressure declining gradient

PF

particle filter

PHF

propped hydraulic fracture

Pnet

net pressure

PPB

proppant pack build-up

ppg

pounds per gallon

prop

proppant

RP

re-perforating

SD

screenout detection

SDT

step-down test

SUT

step-up test

SO

screenout

SOI

screenout index

spf

shots per foot

TW

tube wave

WF

weterfrac

WSO

wellbore screenout