You are on page 1of 9

SPE 151979 Investigation of Horizontal Well Fracture Extension Pattern Based on Downhole Pressure Data

Ali Daneshy, Daneshy Consultants Int'l Daniel Herman, PennWest Exploration Dean Tymko, PennWest Exploration

Copyright 2012, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Hydraulic Fracturing Technology Conference held in The Woodlands, Texas, USA, 68 February 2012. This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract. This paper presents the results of an investigation into fracture growth pattern in three horizontal wells, each fractured multiple times. The completion system was selected such that it allowed recording of the bottom-hole pressure in two critical locations; at the frac port which was being fractured, and, within the previously fractured part of the same wellbore. Downhole cups isolated the two locations from each other. The data show remarkable results. All fractures initiated axially and re-oriented to become perpendicular to the minimum principal stress (MPS). The re-orientation details varied widely between different fractures in the same well, and also between wells. In-spite of these variations, there was no communication within the formation between the multiple fractures. All fractures in the same well had the same value of MPS, and in fact nearly the same in all three wells. In one of the wells there was obstruction to proppant movement inside the fracture which caused increasing pressures during fracture extension. In one instance this resulted in screen-out very near the wellbore early in the treatment, and in another case inside the fracture and close to the end of the stage. Still, the high pressures encountered during these fractures did not cause communication with the previous fractures. The growth pattern in all fractures can best be described as off-balance, with no evidence of complexity. Introduction Fracturing of horizontal wells is the foundation for production from tight oil and gas reservoirs. Each year thousands of fractures are created in horizontal wells at a substantial cost to the operators. There is considerable difference of opinion about the growth pattern of these fractures, and ways of optimizing their outcome. Some of the ambiguity is caused by conflicting attempts to explain highly variable fracturing treatment pressure data. Some others are created by sometimes large differences in the results of various surveys conducted to determine the growth pattern. The catch-all term used to explain these divergences is the word complexity which by its nature means something that is not explainable to the layman. Yet, this word is often used by the experts themselves1,2,3. The problem with assuming a complex fracture growth pattern is that it discourages application of engineering and scientific principles, and indeed solving problems. The other term used to describe fracture behavior is off-balance4. In this case fracture growth includes both tensile (dominant) and shear fracturing modes, and various degrees of branching. Much of the shear fracturing and branching occurs near the wellbore during the reorientation process. Additional shear fracturing and branching occur at intersections between the created hydraulic and existing natural fractures. The overall fracture grows in a narrow band perpendicular to the minimum in-situ principal stress and therefore has a very clear orientation. Multiple fractures created from adjacent perforation clusters will therefore be parallel with each other. Any connection between these fractures is likely to occur near the wellbore and as a result of initial axial fracture growth. The purpose of these field measurements was to do a detailed analysis of the pressure data with the aim of gaining insight into the growth pattern of hydraulic fractures created in horizontal wells. To do this, pressures in two specific critical wellbore locations were recorded while fracturing. These were; 1. At the port where fracturing slurry was being injected 2. In the well segment downstream of this port where previous fractures had been created and now were connected to the wellbore.

SPE 151979

The objective in selecting the completion system was to achieve wellbore isolation between the two gauges during the fracturing operations. Therefore, any pressure communication between the two gauges during the fracturing operations would have to be through the fracture/formation combination. As it turned out, the system did perform as expected. The above measurements were repeated in three separate wells in the same field. Below we present a brief description of the critical aspects of these experiments. Formation and reservoir description. The experiments were performed in the Esther Field, in southeastern Alberta, Canada. The target formation is Viking Gas. It consists of interbedded, bioturbated sandy and silty shales to shaly and silty sandstones. Its mineralogy consists of 15 60% sandstone. The permeability and porosity values range from 0.1 to 60 md, and 15 25%. Water saturation range is 40 60%. Completion Information. The completion system used in these treatments consisted of multiple frac ports pre-installed in the 4 casing cemented in the well. Each cluster of ports, when pressurized, would burst and provide access to the formation. The length of the collar containing the ported assembly was 1.25 and it contained 15 ports, each with a diameter of 0.438 inches. Fracturing operations started at the toe of the well and progressed towards the heel. Locations of ports were selected before casing installation. Spacing between the ports varied along the length of the well. Some of the details of these completions are listed in Table 1. Fracturing slurry was injected through a 3 OD coil tubing (CT) with 2.84 ID. The fluid was an emulsified CO2. The liquid phase contained 20% methanol. The type and weights of proppant are also listed in Table 1. The three wells under study have almost the same orientation; Well #1 is drilled SE to NW while wells #2 and # 3 are in the opposite direction of NW to SE. All three wells have approximately the same vertical depth; 2,241 2,559 ft (Table 1). Well # 3 is about 4 miles east and 14 miles north of Well #1. Well #2 is 0.6 miles west and 1.6 miles south of Well #1. Data Acquisition Set-up. The coil tubing used for these fracturing treatments has two cups at its end, as shown in Fig. 1. The cups straddle and isolate the target frac ports from the previous treatments, as well as the annulus between the casing and coil tubing. During each fracturing stage the gauge at point B recorded the shut-in pressure in the previous stages, while the gauge at point A recorded the bottom-hole fracturing pressure. Both gauges A and B also recorded the variations of the temperature in their respective positions. However, this data was found very difficult to analyze because of the presence of CO2, and also because the coil tubing and casing were both conducting temperature. During fracturing of the first stage (at the toe of the well) both gauges record the bottomhole fracturing pressure (there is no Figure 1. Bottom-hole completion schematic previous fracture below the cups at this stage) and the corresponding curves are identical. After that the two gauges recorded different pressures. At the end of each stage, pulling on the coil tubing released a valve at the bottom cups, thus allowing the CT to move up. This also created a temporary hydraulic link between the two sides of the bottom cup and caused the pressure below the CT to increase a small amount. Once the CT was located in its proper position the valve was closed and the hydraulic link between the two sides of the bottom cup was broken. The completion set-up presented above provides a unique way of diagnosing fracture behavior. During each stage of the operation, if the extending fracture connects to the previous fracture(s), this will cause an increase in the pressure below the cups (Gauge B). Two situations can cause linkage between these fractures; a nearly axial (longitudinal) fracture where the fracture runs along the axis of the borehole and connects with the previous axial fracture, or, a so-called complex fracture where fracture orientation is random and the two fractures connect somewhere within the formation. It is expected that there would be more fracture connections in the case of axial fractures, while complex fractures may connect only occasionally. Pressure values at gauge B also provide a unique way of observing variations of MPS within the different fracture stages, as well as the pressure fall-off due to leak-off after shut-in.

SPE 151979

Data Review and Analysis The results and the interpretation of data from each of the three wells are presented below. The order of well naming selection is simplicity of fracture behavior, not chronological. Well #1. Some of the completion and fracturing data are listed in Table 1. The average spacing between the stages was 200 ft. In order to explain the type and details of data collected during all fracturing stages, the data for the first two stages of the fracture will be discussed in more detail. This data is presented in Fig. 2. The following color and axis convention will be used in this and all subsequent graphs presented in this paper; All pressures are plotted against the left-hand side y-axis. Their color schemes are; Surface fracturing pressure in dark blue Bottom-hole fracturing pressure (gauge A) in purple Shut-in pressure in previous stages (gauge B) in blue MPS (minimum principal stress) line in dashed green. This line was determined after review of all the data and is plotted in all the graphs.

The injection rate is plotted against the right-hand y-axis. It is plotted in color pink. This is the total injection rate and includes the CO2 and 20% methanol in the liquid phase. Proppant concentration at the wellhead (WH Prop Conc.) is plotted in dark brown while the bottom-hole proppant concentration (BH Prop Conc.) is in light brown. Both are also plotted against the right-hand side y-axis (together with rate). The x-axis denotes either clock or elapsed time. During stage 1 of the treatment the curves representing pressures at gauges A and B overlap. The data in both stages shows large pressure fluctuations at the surface and bottom-hole. These fluctuations are caused by the tortuosity created during re-orientation of the initial axial fracture to a final transverse fracture5,6. In fact the first fluctuation occurs before proppant reaches the ports. This also indicates an off-balance growth pattern, with severe initial near wellbore tortuosity. The tortuosity also interfered with proppant movement inside the fracture at a later time. Once a fracture growth pattern was established, the fluid pressure gradually declined and followed a more familiar pattern. The shut-in pressure at the end of this stage shows very little drop, which indicates low leak-off. Pressure variations during stage 2 follow a very similar pattern, except that gauges A and B are recording different pressures. They show no communication between the new fracture in stage 2 and the existing fracture below the bottom cup created during stage 1. The difference during fracturing between pressures recorded at the surface and with gauge A is caused by hydrostatic pressure as well as friction due to flow of slurry inside the coil tubing. The difference between surface and bottom-hole pressures during shut-in is due to hydrostatic pressure.

SPE 151979

At the end of stage 2, pressure at gauge B increased a small amount while the cups were moving to the next stage, thus opening of the valve at the bottom cup. Since the volume of fluid moving through the bottom cup was small, this did not cause a significant pressure increase in gauge B. In fact the pressure quickly stabilized at the same MPS value and decreased very slowly during the shut-in. One undesirable feature of Fig. 2 is the clutter of data because it includes too much detail. From this point on, the graphs for all treatments will be simplified to show only those variables that are critical for this review. Specifically, the surface pressure data will not be plotted anymore, since a more accurate bottom-hole pressure is available. Similarly, the wellhead proppant concentration data will be removed in favor of more appropriate BH data. Multiple injection cycles precede fracturing at each stage. These are used for bursting the ports, and sometimes for moving the cups. The injection rate data during these operations will also be removed so that the only injection data displayed on the graphs is that of the fracturing treatment itself. However, the pressure curve for gauge A provides an indirect indication of these injections, and the pressures they have caused. To reduce the number of plots, the data from multiple stages will also be combined into a single graph. The chart for stages 3 8 is presented in Fig.3. The data shows large variations in the fracturing pressures of different stages. This is very typical of all horizontal well fractures that undergo re-orientation from axial to transverse6. High pressures during early times indicate strong near wellbore tortuosity. In this example, most of these changes occur before proppant enters the fracture and are therefore unrelated to proppant. The data also does not show any scouring effect from the proppant. The very small pressure hikes in gauge B occurring while moving the cups is dissipated quickly in the large fluid volume below the cups (fracture plus wellbore volumes). The most remarkable feature of this data is the constant value of shut-in pressure and how closely it matches the MPS line. This indicates a very consistent in-situ stress regime. The data also shows very low leak-off. The graph for the remaining stages 9 14 is presented in Fig. 4. Once again, the data shows very large pressure variation between the different stages. These fluctuations are all associated with fracture re-orientation. Proppant does not seem to have any positive (scouring) or detrimental (blockage inside the fracture) effect. Again, the most remarkable feature of this data is its consistency during the shut-in period, which indicates a constant value of MPS. In summary, all the data recorded in this well show; 1. Axial fracture initiation and transverse extension. The re-orientation process caused large variations in pressures between different stages. 2. Constant stress regime and magnitude of MPS.

SPE 151979

3.

In-spite of large pressure variations in the early times of each stage there was no communication between the various stages. This indicates that fracture growth occurred in a narrow band around the ports. This is indicative of an off-balance growth, instead of the random complex pattern. Very low pressure drop during the shut-in periods indicates that fracture had not completely closed and some of the loss of fluid due to leak-off was compensated by reduction of fracture volume due to closure. MPS value was 2,190 psi.

4.

5.

Well #2. This well had the largest spacing between fracture stages; 312 ft average. There were several operational breaks during fracturing operations. Fluid leak-off during these breaks caused pressure drops below the bottom cup (in the previously fractured part of the reservoir). There is no bottom-hole pressure data available for the first six stages. The data from stages 7 to 9 are presented in Fig. 5. Before stage 7 pressure below the cups was declining due to leak-off. But once stage 7 was completed and the cups were pulled up, exposing gauge B to the pressure of the new fracture in Stage 7, the pressure below the cups stabilized at around MPS. Similar to Well #1, there is no communication between the newly created and previous fractures. All fractures initiated as axial and re-oriented to become transverse, although the reorientation process was smoother than Well #1. After stage 9 there was another break in fracturing operations. Stage 10 of these fractures started the next day, almost 11 hours later. Fluid leak-off during this period caused pressure drop in the previous fractures. As a result gauge B pressure was lower than the MPS value. Even after stage 10, the added contribution of the new fracture was not sufficient to raise the low pressure of the previous stages. This was further compounded by the hour long delay between stages 10 and 11. But after stage 11 the pressure in the previous fractures (gauge B) stabilized at around the MSP value. This trend continued until the end of fracturing. Fracture behavior in this well was quite similar to Well #1. Main areas of similarity include; Axial fracture initiation and re-orientation to a transverse fracture. However, the re-orientation process was smoother in Well #2, as evidenced by shorter times and less variations in pressures.

SPE 151979

Proppant did not have a major effect on fracture propagation. There was no sign of scouring or proppant blockage during these jobs. There is no evidence of communication between the various fracture stages outside of the wellbore, indicating continued offbalance growth pattern. The magnitude of MPS was 2,050 psi. Well #3. The treatment details are listed in Table 1. Average spacing between these fractures was 184 ft. The data for the first four stages of the treatment is presented in Fig. 7. As before, during stage 1 both gauges A and B are recording the bottomhole pressure and their graphs are identical. Figure 7 shows that the fracture in all these stages initiated as axial and then re-oriented to become transverse. However, the re-orientation process was not smooth and there was blockage to proppant movement inside most of the fractures, as evidenced by the increase in pressure when proppant entered the fracture. In all these stages the obstruction to proppant movement continued for some time, and even when pressure finally started to decline it still had a relatively high value. The net effect of this obstruction was considerable variation in the fluid pressure inside the fracture, which then resulted in higher wellbore pressure in order to compensate for lower pressures further along the fracture7. The shut-in data at the end of each stage does not show the typical sharp drop and break in the slope of the pressure curve. This is caused by poor hydraulic communication between the wellbore and the fracture (which is exacerbated by proppant blockage inside the fracture). The net effect is a large difference between the pressure at the wellbore and deeper inside the fracture, which then also caused a more rapid pressure decline during the shut-in period while pressure was trying to equalize inside the fracture. This is the reason why pressure at gauge B was higher than the MPS line pressure. There was a disruption in the fracturing operations after stage 4. Consequently stage 5 did not start until about 45 minutes later during which pressure had time to equalize in the existing fractures. The graph for stages 5 8 is presented in Fig. 8. This data shows the continuation of obstruction to initial proppant entry inside the fracture in stage 7, and more strongly in stage 8. While this is likely to have caused large pressure variations inside fractures in stages 7 and 8, when combined with the large volume of previous fractures the wellbore pressure was dominated by the overall behavior of all fractures. The pressure fall-off pattern at the end of stage 8, and specifically the absence of a crisp shut-in pressure indicate that this stage of the fracture was at the verge of screen-out at the end. There was another long break after the stage 8 fracture and operations did not resume until the next day. This caused a drop in the pressure of existing fractures. The graph for fractures created in stages 9 13 is presented in Fig. 9. It shows a much lower pressure at gauge B (nearly hydrostatic).

SPE 151979

Connecting the large number of existing low pressure fractures with stage 9 caused only a small increase in the pressure at gauge B. However, opening of the valve at the bottom cup at the end of stage 10 caused the pressure to equalize at MPS value. There was another break in fracturing operations after stage 13 and the next fracture stage was executed almost 19 hours later, which caused a drop in the pressure at gauge B due to leak-off. The data for stages 14 to 16 is presented in Fig. 10. This graph shows the long delays between these stages. It also shows the absence of any communication between each of these stages and the previous existing fractures. The next stage of fracturing screened out as soon as proppant reached the ports, Fig. 11. But even the high pressure at screen-out did not cause communication between this stage and the previous fractures. Following the screen-out the coil tubing had to be moved several times and this caused communication between gauges A and B and reduction in pressure at gauge B. Fractures in stages 18 and 19 did not have sufficient volume to raise the pressure at gauge B to the value of MPS. Opening of the bottom cup valve after stage 19 caused sufficient pressure communication to increase the pressure to MPS value. The data from stage 21 of this well is not presented here since it does not add to the content. The magnitude of the minimum principal stress in this well was 2,220 psi.

The data from this well was very useful in clarifying several issues long debated in the fracturing technical community. Among these are; Screen-out is a near wellbore event. Its cause is very poor communication within the fracture, between the near wellbore region and deeper inside the fracture. As a result the very high pressures at the wellbore are not transmitted inside the fracture to cause it to widen and enhance its conductivity to fracturing fluid. The same situation, but with less drastic outcome, occurs when there is obstruction to proppant movement inside the fracture. This also causes a rise in the wellbore pressure. Poor hydraulic communication within the fracture causes the pressure to increase upstream of the obstruction in order to compensate for the lower than required pressure downstream7. This poor communication extends to the shut-in period too and causes delays in pressure stabilizing at the value of MPS.

SPE 151979

High wellbore pressures while fracturing are indicative of poor conductivity within the fracture (which is required to stabilize the pressure). This poor conductivity manifests itself by slow reduction of pressure after the job.

Conclusions Combining all the results of these experiments leads us to the following conclusions; 1. In-spite of variations from one stage to the next, and one well to the next, the fracturing behavior was very consistent throughout these treatments and followed expected behavior. 2. and then re-oriented to become perpendicular to the minimum principal stress. 3. In-spite of its initial axial growth and near wellbore tortuosity during the re-orientation process, there was no indication of any communication between the different fracture stages in the same well. The absence of communication between these fractures supports their growth within a narrow band and generally perpendicular to MPS. This growth pattern is known as off-balance. The data shows no support for a complex growth pattern, where the fracture does not have a clear orientation and can grow anywhere in the formation. The coalescence of pressure data in the same well towards a single value of MPS means a consistent stress regime. The observed variations between MPS in different wells are caused mostly by depth. In fact, the variations in fracture gradient are very small and within expectation. Screen-outs are caused by near wellbore obstructions to proppant movement inside the fracture. The stress regime is regional and does not change from one well to the next. All fractures initiated axially

4. 5.

6. 7.

Acknowledgement The authors thank the management of PennWest Exploration for permission to publish this paper and their enthusiastic support. References T. N. Olsen, T. R. Bratton, M. J. Thiercelin, Quantifying Proppant Transport for Complex Fractures in Unconventional Formations SPE Hydraulic Fracturing Conf., 19 21 Jan 2009, The Woodlands, TX G. E. King, L. Haile, J. Shuss, T. A. Dobkins, Increasing Fracture Path Complexity and Controlling Downward Fracture Growth in Barnett Shale Shale Gas Production Conf., 16 18 Nov. 2008, Fort Worth TX C. L. Cipolla, N. R. Warpinski, M. J. Mayerhofer, Hydraulic Fracture Complexity: Diagnosis, Remediation and Exploitation SPE APOGCE, 2022, Oct. 2008, Perth Australia Daneshy, A. A., Off-Balance Growth: A New Concept in Hydraulic Fracturing JPT, April 2003, 78-85 Daneshy, A. A,; Hydraulic Fracturing of Horizontal Wells: Issues and Insights SPE 140134 presented at the SPE Hydraulic Fracturing Technology Conference and Exhibition held in The Woodlands, Texas, USA, 2426 January 2011

1.

2.

3.

4. 5.

SPE 151979

6. 7.

Daneshy, A. A., Horizontal Well Fracturing: Why Is it so Different JPT, Sept. 2009, 28-35 Daneshy, A. A., Pressure Variation Inside the Hydraulic Fracture and its Impact on Fracture Propagation, Conductivity, and Screen-out" SPE Production and Operations, Feb. 2007, 107-111

Table 1. Summary of Well Data Well #1 6.26 4.511.6# L-80 3.25, 2.84 SE to NW 7758 2343 Emulsified CO2 with 20% MEOH in the liquid phase 426,000 20/40 + 231,000 16/30 Well #2 6.26 4.511.6# L-80 LT&C 3.25, 2.84 NW to SE 7216 2241.5 Emulsified CO2 with 20% MEOH in the liquid phase 308,000 20/40 + 154,000 16/30 Well #3 6.26 4.5 11.6# L-80 3.25, 2.84 NW to SE 7036 2559 Emulsified CO2 with 20% MEOH in the liquid phase 426,000 20/40 + 231,000 16/30

Open-hole size (in) Casing size Coil Tubing size (in) (OD, ID) Well orientation TMD (ft) TVD (ft) Frac fluid Proppant weight(lbs)

You might also like