Impact of water blocking in fractures on the performance of hydraulically fractured horizontal wells in tight gas reservoir

Accepted Manuscript Impact of water blocking in fractures on the performance of hydraulically fractured horizontal wells in tight gas reservoir Fengpeng Lai, Zhiping Li, Yining Wang PII:

S0920-4105(17)30457-6

DOI:

10.1016/j.petrol.2017.05.002

Reference:

PETROL 3986

To appear in:

Journal of Petroleum Science and Engineering

Received Date: 12 October 2015 Revised Date:

5 January 2017

Accepted Date: 6 May 2017

Please cite this article as: Lai, F., Li, Z., Wang, Y., Impact of water blocking in fractures on the performance of hydraulically fractured horizontal wells in tight gas reservoir, Journal of Petroleum Science and Engineering (2017), doi: 10.1016/j.petrol.2017.05.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Impact of water blocking in fractures on the performance of

2

hydraulically fractured horizontal wells in tight gas reservoir

3

Fengpeng Lai 1*, Zhiping Li 1, Yining Wang 2

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

1. School of Energy Resources, China University of Geosciences, Beijing 100083, China 2. China Zhenhua Oil Co. ltd, Beijing 100031, China

Abstract: Tight gas is a major gas resource which accounts for 14% of the total gas resources and 29% of unconventional gas in the world. The natural production of tight reservoir is very low, and multi-stage fracturing technology is widely used in horizontal wells. Multi-stage hydraulic fracturing for stimulating tight gas reservoirs requires a significant amount of fracturing fluid, which is usually water-based. Water blocking is considered as a potential type of damage in tight gas reservoirs. There are many factors affecting the production decline caused by the water blocking in fractures. However, the water blocking and fluid segregation in fractures are ignored. This study investigates the characteristics of water blocking in fractures and the effects of wettability, fracture fluid viscosity and fracture fluid filtration on water blocking and gas production. Numerical simulation is used to show that three stages can be used to describe the change of water saturation at the top and bottom of fractures, where the second stage indicates water blocking in hydraulic fractures. The results demonstrate that the stronger the interfacial tension, the more obvious the water wettability is, and the more water traps in the reservoir matrix. Decreasing interfacial tension improves the load recovery, and reduces the formation damage, and further enhances the gas production. A higher fracturing fluid viscosity causes more water traps at the bottom of fractures and intensifies the formation damage on tight gas wells. Results indicate that the fracturing fluid filtration has a great impact on the flowback performance and later gas production. Results suggest that increasing the filter cake thickness decreases the gas production peak value, and delays the time showing the peak of gas production, and further reduces the cumulative gas production.

29 30

Keywords: Water Blocking; Hydraulic Fracture; Wettability; Fracture Fluid Viscosity; Fracture Fluid Filtration

31

1 Introduction

32 33 34 35 36 37 38 39

Tight gas reservoirs have been the subject of many studies over the past three decades. In the absence of open natural fractures, economic development of tight gas reservoirs is possible only through hydraulic fracturing in vertical or horizontal wells (Shaoul et al., 2011). Hydraulic fracturing of horizontal wells is instrumental in providing large reservoir contact area so as to achieve commercial gas recovery. The advantages of using water as the fracturing fluid include low cost and efficient proppant transportation (Agrawal et al., 2013). The issue of fracturing fluid loss/invasion after a stimulation operation has been widely studied in the past. Tannich (1975) studied the removal of liquid from hydraulically fractured gas wells

AC C

EP

TE D

M AN U

SC

*Corresponding author (Fengpeng Lai, [email protected])

RI PT

1

1

ACCEPTED MANUSCRIPT

EP

TE D

M AN U

SC

RI PT

using known well testing solutions. Holditch (1979) studied water blocking issues in hydraulically fractured tight gas wells and demonstrated the role of capillary pressure, matrix permeability damage and relative permeability in fluid recovery. Solimon et al. (1985) present a numerical simulation based analysis of fracture fluid cleanup and its impact on gas production, and pointed out that the presence of mobile water in the fracture has a significant impact on buildup pressure response after flow back. Montgomery et al. (1990) noted that in case of severe damage around the fracture, the pressure derivative in the buildup test will have an identifiable shape. Settari et al. (2002) have investigated the water blocking and fracturing associated geomechnical issues in water fracs in the Bossier play. Mahadevan et al. (2003) revealed that there are two primary mechanisms for water blocking removal, and provided an excellent way of computing water cleanup from wells. Parekh and Sharma (2004) analyzed the effect of various factors governing the cleanup of water blockings in fractured and unfractured wells for both gas and oil reservoirs. Gdanski et al. (2005) revisited the concept of fracture face damage and studied the skin evolution (multiphase skin calculation) during cleanup (Gdanski et al., 2006) using a two-phase, two-dimensional fully implicit simulator. Friehauf et al. (2009) modeled the productivity of hydraulically fractured gas wells including the effects of fracture fluid leak off into the matrix. Wills et al. (2009) used a two-phase 3-D numerical simulation model to investigated hydraulic fracture cleanup for both slick water and gelled fluids, and showed the pressure drawdown, shut-in times, and perforation placement all had significant effect on fracture face damage and clean-up potential. Gdanski et al. (2010) looked at the impact of different relative permeability curves, drawdowns, drawdown rate, shut-in times and fracture conductivities on the load recovery. A large amount of water is produced back after fracturing treatment. This produced water are primarily recovered fracturing fluid. The amount of fracturing fluid produced back is called load recovery, and a wide range of load recoveries (15%-80%) are reported in the literature (Zanganeh et al., 2015). King (2010) indicated that the modest load recovery may be related to relative permeability in the natural fractures, to wetting phenomena, and to the tortuous path far from the hydraulic fractures. Shaoul et al. (2011) investigated the importance of formation damage in unconventional reservoirs, and noted that there was still a lot of speculation in the industry about the dominant mechanisms and their impact on the gas productivity. Wang et al. (2012) analyzed the various damage mechanisms in unconventional reservoirs and attempted to quantify their relative impact on the gas productivity. Agrawal and Sharma (2013) presented the results of 3D simulations of liquid loading in hydraulic fractures for horizontal wells. The effect of gravity and capillary pressure on the liquid loading are also shown in their study. Alkouh et al. (2014) developed a simulation model to investigate how water traps in the natural fractures. They pointed out the phenomenon of liquid loading in hydraulic fractures. As shown in the above studies, many different factors may influence the water invasion in the formation, which further cause the reduction of production. However, the water blocking in the fracture itself, the effect of gravity, and fluid segregation in fractures are generally ignored. The gas flow is impaired when water blocking takes place in a certain section of hydraulic fractures. In this study, a 3D tight gas reservoir simulator is used to investigate water blocking in the fractures for a horizontal well. The main objective of this study is to show the characteristics of water blocking in fractures through the changes of water saturation, and illustrate the effects of wettability, fracture fluid viscosity, and fracture fluid filtration on the water blocking and gas production.

AC C

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

2

ACCEPTED MANUSCRIPT 1

2 Water blocking in hydraulic fractures Experience from the field seems to be contradictory (Cimolai et al., 1993; Bennion et al., 1994; Settari et al., 2002; Bang et al., 2010; Shaoul et al., 2011). Sometimes no water is produced back, but gas production does not appear to suffer. In other cases, the gas rate is significantly lower than expected, but significant fractured fluid is recovered. The natural production of tight gas reservoir is very low, and the production will be improved by multi-stage fracturing technology. The absolute open flow (AOF) potential of a well is the rate at which the well would produce against zero sandface back pressure. It is used as a measure of gas well performance because it quantifies the ability of a reservoir to deliver gas to the wellbore. Figure 1 shows the load recovery and the AOF of multistage fractured horizontal wells in Daniudi tight gas reservoir. It indicates that the average load recovery is 50.47%, and the load recovery of half wells are lower than 50%. Figure 1 also shows that the high load recovery does not mean the great AOF. AOF describes the potential of gas productivity, and load recovery describes the production of fracturing water. For tight reservoirs, load recovery may not correlate to well performance. AOF primarily depends on the reservoir properties. However, the water production after fracturing treatment is influenced by many factors including fracturing treatment design, imbibition, and reservoir properties. In another field example (Shaoul et al., 2009), a total of around 2000 barrels of water was leaked off into the formation during a fracturing operation, and around 700 barrels of water was produced back during the clean-up period of 35 days. In this period, gas flow rate reduced from 3.5 MMSCFD to 1.5 MMSCFD.

21 22 23 24

Figure 1. The load recovery of horizontal wells in Daniudi gas field: The left figure shows 37 wells (74 wells in total) have a load recovery less than 50%; The right figure shows scattered data of load recovery and AOF.

25 26 27 28 29 30 31 32 33 34 35 36

During the fracturing process, water leaks off into the reservoir pores and remains there until production starts and the cleanup commences. Then, if pressure drawdown overcomes capillary pressure within the reservoir, some of the water is produced back through the fractures. This water can also remain trapped in the pores of the reservoir by capillary forces and it can affect well productivity negatively (Le et al., 2009). This process is called water blocking. The loss of fracturing fluid during multi-stage fracturing of horizontal wells can be divided into three types. The first type is leak-off, where some of the fracturing fluid filtrates in the formation. The second type is liquid-loading, which distributes at the bottom of fractures. The last one is due to a residual fluid, which attaches to the fracture surface because of water-wet conditions. Gas flow can be affected by these three types of fracturing fluid loss. Here, we use IMEX simulator to study the effect of water blocking in fractures on tight gas production. IMEX is CMG’s implicit-explicit black oil reservoir simulator, and it is a mature

AC C

EP

TE D

M AN U

SC

RI PT

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

3

ACCEPTED MANUSCRIPT

Parameters

Value

Matrix porosity

10 %

Fracture porosity Matrix permeability Fracture permeability Initial reservoir pressure

Parameters

Value

Fracture half length

53 m

60 %

Irreducible water saturation

20 %

0.1 md

Bottom hole pressure

31 MPa

2000 md

Injection time

1 day

44.8 MPa

Shut-in time

30 days

17 m

Initial water saturation

20%

Residual water saturation

20%

EP

Fracture height

24 25 26

3.05 cm

AC C

Fracture thickness

23

M AN U

SC

RI PT

commercial software. It can provide local grid refinement, comprehensive well management, horizontal wells, dual porosity/permeability reservoir models and many more. The model is based on a 46×16×20 Cartesian grid in the X, Y, and Z direction, respectively. The total volume of the reservoir simulated here is 122 m×122 m×61 m. Modeling of capillary pressure is very important in the study. Capillary trapping within the matrix, near the fracture face, is a dominant mechanism for fracturing fluid retention. It is also important for the gas flow through the proppant pack. Agrawal (2013) used a Leverett J-function to scale the capillary pressures based on permeability and porosity. In this section, the static parameters for the simulation are from Mullen (2010) and Agrawal (2013). The gas reservoir is a type of dry gas reservoir. The initial water saturation is irreducible. Table 1 shows the static parameters used in the simulations. Figure 2 shows the relative permeability curves for matrix and fracture, respectively. Matrix relative permeability characteristics have been modeled using a Corey type formulation (Corey, 1954). A straight line of rel-perms have been used for fractures considering the extreme low capillary pressures (Agrawal, 2013). Figure 3 shows the relationship between water saturation and J-function value. Figure 4 shows a schematic of a horizontal well multi-stage fracturing. The top and bottom of fracture represent different locations of fracture. This study uses a section of one stage fracture system for simulation. In order to simulate the flow back of fracturing fluid, it is firstly injected into the gas well at a certain injection rate. The well begins to produce after a shut-in over a certain period of time. We then compare the water saturation at different part of fractures, and the water and gas production. Table 1. The static parameters for the simulation (Mullen, 2010)

TE D

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Figure 2. Relative permeability curves used for simulation (Agrawal, 2013): The left figure is gas-water relative permeability curve of matrix; The right figure is gas-water relative permeability curve of fracture.

4

RI PT

ACCEPTED MANUSCRIPT

1

Figure 3. The relationship between water saturation and J-function value (Agrawal, 2013)

SC

2

M AN U

3

Figure 4. A schematic of grids in simulation for multi-fractured horizontal well: The red line represents the horizontal section of a horizontal well; The five vertical black lines represent five hydraulic fractures.

7 8 9 10

Figure 5 shows the changes of water saturation at the top and bottom of a fracture. The water saturation at the top of the fracture at different times is obtained through the simulation, which is based on fine-scale gridding. Three stages can be used to describe the changes of water saturation at the top and bottom of the fracture.

11 12 13 14

Figure 5. The water saturation at the top and bottom of fracture against time: The red line shows the dynamic water saturation at the top of fracture; The green line represents the dynamic water saturation at the bottom of fracture.

15 16 17 18

First stage: The water enters into the matrix under the pressure difference between the fracture and matrix during the injection progress. The water saturation in the fracture decreases with time. Second stage: The water saturation at the top of the fracture initially increases after opening well for flowback, and then decreases after reaching to a maximum value. The water in the matrix

AC C

EP

TE D

4 5 6

5

ACCEPTED MANUSCRIPT

TE D

M AN U

SC

RI PT

enters into the fracture under the pressure difference. Because of the wellbore storage effect, the water which is stored in wellbore during shut-in time first flows to the wellhead. During this period, the rate of water flows from fracture to wellbore is negligible. Under the action of pressure difference after opening, water seepage from matrix to fracture. However, the flow of water from fracture to wellbore is restricted by wellbore storage effect. The combined effect of pressure difference and wellbore storage effect is that the rate of the water flow from the fracture to the well bore is lower than that from the matrix to the fracture. This results in a temporary increase of the fracture water saturation, after which it gradually decreases when the water flow rate is in equilibrium. Effusion occurs at the bottom of the fracture, and the buoyancy is lower than gravity. This affects the gas flow by causing a decrease of the effective flow conductivity and water blocking in the hydraulic fracture. Third stage: With the progress of production, the reservoir pressure gradually approaches the flowing bottom-hole pressure. The fracture bottom starts cleaning up when the gas velocities is large enough to lift the water out of the fracture. The water saturation at the bottom of the fracture gradually decreases to the irreducible state, while the water saturation at the top of the fracture is irreducible during the whole stage. The gravity and capillary force affect the water distribution at the horizontal and vertical directions. After one day of injection, the water saturation in the fracture is high. The water in the fracture gets infiltrated into the matrix by capillary force. The matrix that near fracture face does not start cleaning up until the fracture is cleaned up completely. The reasons are that these two zones are in equilibrium and imbibition. The water saturation in the fracture decreases significantly after 30 days of shut-in. The water at the top of the fracture flows faster to the wellbore assisted by gravity. After 210 days, there is water retention at the bottom of the fracture, while the water at the top of fracture starts flowing back completely. The residual liquid is still at the bottom of the fracture, causing a reduction of the fracture effective length. Figure 6 shows the water saturation in the fracture at different times.

27

AC C

EP

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

28 29 30 31

Figure 6. A cross section view shows the water saturation distribution in the fractures at different time scales: (1) at the first day after fracturing treatment; (2) 30 days after fracturing treatment, which is at the onset of flowback; (3) 210 days after fracturing treatment, representing early gas production; (4) 500 days after fracturing treatment, representing long-term gas production.

32

3 Effect of wettability on water blocking

33 34 35

Wills et al. (2009) studied the relationship between the well performance and wettability alteration. The change in interfacial tension only impacted the capillary pressure value directly. Their results showed that reducing the interfacial tension from 72 Dyne/cm to 35 Dyne/cm had no 6

ACCEPTED MANUSCRIPT effect on the production. The interfacial tension in this study was set at 20, 25, 30, 40, 60, 70, 80, 100 and 120 Dyne/cm. Well performance was analyzed to show the effect of interfacial tension on water blocking. Table 2 shows the results of well performance under different interfacial tensions. The load recovery gradually decreases with the increasing interfacial tension. The stronger the interfacial tension, the more obvious the water wettability is, and the more water is left in the reservoir matrix. Table 2. The data of well performance under different interfacial tension

(Dyne/cm)

(%)

Water saturation at bottom th

Maximum

of fracture on the 500 day

production

(%)

(m3/d)

Cumulative production （104m3）

37.04

61.84

9460.70

205.70

25

34.49

28.76

9230.04

203.14

30

32.21

21.33

9007.18

201.49

40

27.95

20.06

60

22.20

20.02

70

19.37

20.01

80

16.79

20.01

100

11.45

120

6.58

SC

20

8510.99

200.12

6727.68

178.63

6037.71

170.70

5439.34

165.20

20.00

4469.03

157.17

20.00

3636.57

151.84

EP

TE D

Figure 7 shows the change of water saturation at the bottom of the fracture under different interfacial tensions. The change of water saturation at the top of the fracture shows the same trend, becoming irreducible after 200 days. The water saturation at the bottom of the fracture becomes irreducible earlier and the effect of interfacial tension on the water saturation at the top of fracture is smaller. Figure 8 shows the change of water saturation in the fracture at different times for 5 cases of interfacial tensions. In the case of low interfacial tension, the water saturation in the fracture is high, leading to more liquid in the matrix returning to the ground. Under these conditions, the velocity of the water flow from the matrix to the fracture is greater than that from the fracture to the wellbore. The average water saturation decreases in the matrix and increases in the fracture. The smaller the interfacial tension, the larger the load recovery and gas production. Therefore, the effect of water blocking in the fracture on well performance is smaller when the interfacial tension is low.

AC C

8 9 10 11 12 13 14 15 16 17 18 19 20

Load recovery

RI PT

Interfacial tension

M AN U

1 2 3 4 5 6 7

7

RI PT

ACCEPTED MANUSCRIPT

1

Figure 7. The water saturation at the bottom of fracture against time for 5 cases of interfacial tension: 20 Dyne/cm, 30 Dyne/cm, 40 Dyne/cm, 70 Dyne/cm, and 120 Dyne/cm.

4 5 6

Figure 8. Cross section views of water saturation distribution in fractures at different times for 3 cases of interfacial tension: 20 Dyne/cm, 70 Dyne/cm, and 120 Dyne/cm.

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Figure 9 illustrates the dynamical change of gas production. Within a certain range, after reaching a peak at about 40 days, as the interfacial tension increases, the gas production starts decreasing. However, when the interfacial tension exceeds 80 Dyne/cm, the production peak of the gas wells decreases with the increasing interfacial tension, and the time to production emergence is increases. In the case of a higher interfacial tension, although the gas production rebounds later, the recovery is low and the cumulative gas production is also remains low. When interfacial tension is 120 Dyne/cm, there is an upward trend in gas production from 104 days to 180 days. The possible reason is that the decreases of water saturation at the bottom of fracture relieves the water blocking effect in the fracture. These results show that the interfacial tension has a great impact on the flowback and gas production. The load recovery can be improved by lowering the interfacial tension, which can also lead to reducing the damage caused by the fracturing fluid and increasing the gas production. Therefore, the effect of water blocking in the fracture on gas production can be reduced or relieved through decreasing the interfacial tension. Such a reduction can be realized by adding an appropriate amount of a wetting reversal agent to the fracturing fluid.

AC C

EP

TE D

M AN U

SC

2 3

8

RI PT

ACCEPTED MANUSCRIPT

1

Figure 9. Gas production rate against time for 5 cases of interfacial tension: 20 Dyne/cm, 30 Dyne/cm, 40 Dyne/cm, 70 Dyne/cm, and 120 Dyne/cm.

4

4 Effect of fracture fluid viscosity on water blocking

Viscosity

Water saturation at

Maximum production

Cumulative production in

3

(m /d)

500 days（104m3）

24.6

7384.85

186.95

bottom of fracture

(cp)

(%)

0.6

29.14

0.8

29.23

54.7

6869.71

181.89

1.0

29.60

64.9

6539.63

178.68

1.5

29.35

65.6

5901.94

172.13

(%)

AC C

14 15 16 17 18 19 20 21 22 23 24 25

Load recovery

TE D

M AN U

When fracturing, viscosity play a major role in providing sufficient fracture width to ensure proppant entrance into the fracture, carrying the proppant from the wellbore to the fracture tip, generating a desired net pressure to control height growth and providing fluid loss control (Montgomery, 2013). We use 4 cases of different fracturing fluid viscosity (0.6, 0.8, 1.0, and 1.5 cp) to investigate the influence of the fracturing fluid viscosity on the fluid distribution in the matrix and fractures, and on the productivity of the gas well. In this section, the reservoir is divided into two parts: fractured zone and formation. The relative permeability curve and capillary pressure are the same as basic model (Section 2). Table 3 shows the results of simulation. Table 3. The simulation results under different fracturing fluid viscosity

EP

5 6 7 8 9 10 11 12 13

SC

2 3

With the increasing viscosity of fracturing fluid, the load recovery firstly increases and then decreases. The water saturation at the bottom of the fracture firstly increases and then remains unchanged, with the third stage of water saturation distribution at the bottom of the fracture not occurring. For example, the load recovery after 500 days is 29.14% and 29.35% for the case of viscosities with 0.6 cp and 1.5 cp, respectively. The corresponding water saturation at the bottom of the fracture are 24.6% and 65.6%. This suggests that a significant amount of water still remains in the fracture. Figure 10 shows the change of water saturation at the top of fracture. In the progress of production, more water from fracture flow back to wellhead surface, and the water saturation in the fracture is changing during the process. Eventually, the water saturation at the top of fracture becomes the same as the irreducible water saturation. The higher fracturing fluid viscosity, the larger water saturation at the top of fracture. This indicates the presence of more residual fluid in the fractures. 9

ACCEPTED MANUSCRIPT The relationship between the load recovery and fracturing fluid viscosity indicates a positive correlation when the viscosity is between 0.6 cp and 1 cp. In that range, higher fracturing fluid viscosity contributes to a higher load recovery, leading to less fracturing fluid remained in the reservoir. The greater the viscosity of fracturing fluid is, the more difficult for the fracturing fluid invades into the matrix, resulting a higher water saturation in the fracture than that in the matrix. In the progress of production, more water from fracture flow back to wellhead surface, and the load recovery is greater. Figure 11 shows the change of water saturation in the fractures for 3 cases of viscosities: 0.6 cp, 0.8 cp, and 1.5 cp. The water saturation in the reservoir matrix decreases, while the water saturation in the fracture increases when the viscosity is lower than 1 cp. This result suggests the rate of the fluid flow from the fracture to wellbore is lower than that from the matrix to the fracture. The imbibition is more obvious under the smaller viscosity, and it results more gas production. The presence of the gas reduces the viscosity of the fluid which seepage from matrix to fracture, and increases the mobility of the fluid. It is the reason that the rate of the fluid flow from the matrix to the fracture is greater than that from the fracture to the wellbore.

15 16 17

Figure 10. The water saturation at the top of fracture against time for 4 cases of fracturing fluid viscosity: 0.6 cp, 0.8 cp, 1.0 cp, and 1.5 cp.

18 19 20

Figure 11. Cross section views of water saturation distribution in fractures at different times for 3 cases of viscosities: 0.6 cp, 0.8 cp, and 1.5 cp.

AC C

EP

TE D

M AN U

SC

RI PT

1 2 3 4 5 6 7 8 9 10 11 12 13 14

10

RI PT

ACCEPTED MANUSCRIPT

1

Figure 12. Gas production rate against time for 4 cases of viscosities, which is 0.6, 0.8, 1.0 and 1.5 cp, respectively.

4 5 6 7 8 9 10 11 12 13 14 15 16

As soon as the production is initiated, water comes out from the fractures into wellhead surface. However, the load recovery is limited. A high fracturing fluid viscosity in the fracture and the formation impedes the gas flow in the production process, resulting a lower gas production rate. Figure 12 shows gas production rate for different values of the fracturing fluid viscosity. The greater the viscosity of the fracturing fluid, the lower the maximum gas production rate. The maximum production rate is 7,384.85 m3/d for viscosity 0.6 cp, decreasing by 20.08% for viscosity 1.5 cp. This illustrates that an increase of the fracturing fluid viscosity has a negative effect on the gas production. These results show that the viscosity of fracturing fluid has a significant impact on the load recovery and gas production for the tight wells studied in this research. The greater fracturing fluid viscosity, the worse water blocking at the bottom of the fracture, and the greater the damage on the gas production. Therefore, the effect of water blocking on gas production can be reduced through adding an appropriate amount of viscosity depressant into the fracturing fluid.

17

5 Effect of fracture fluid filtration on water blocking

18 19 20 21 22 23 24 25 26 27 28 29 30

Holditch (1979) pointed out that the injection of liquid made the clay minerals swell, aggregate and migrate in the invaded zone, resulting a decrease of reservoir permeability. Gruber (1996) asserted that the effects of liquid invasion on formation permeability in reservoirs with ultra-low water saturations had been over-dramatized, perhaps because the mechanisms of permeability reduction as a result of fluid invasion were not well understood. We use numerical simulation to study the damage caused by fracturing fluid filtration. For this purpose, the reservoir is divided into three parts: fractured zone, filter cake and formation. The fracturing fluid flow back and gas well production is compared for different filtration conditions. The fluid relative permeability in the filter cake is corrected using a method introduced by Cheng (2010). Figure 13 shows curves illustrating the relationship between relative permeability and capillary force. Krw1 and Krg1 indicate the relative permeability of water and gas in the reservoir, outside of the filter cake. Krw2 and Krg2 denote the same in the filter cake. Pc1 and Pc2 mark the capillary pressure in the reservoir outside the filter cake and in the filter cake, respectively.

AC C

EP

TE D

M AN U

SC

2 3

11

RI PT

ACCEPTED MANUSCRIPT

1

Figure 13. Relative permeability and capillary pressure used for different regions in the simulation: The left one shows the capillary pressure curves for the reservoir matrix and filter cake; The right one shows the relative permeability curves for the reservoir matrix and filter cake.

5 6 7 8 9 10 11 12 13

Four cases are compared by setting the filter cake thickness at 0 mm, 27.45 mm, 63.95 mm and 115.95 mm. The second thickness is 27.45 mm, which is the size of first grid near the fracture. The size of second grid near the fracture is 36.50 mm. We set the third case of filter cake thickness as is 63.95 mm, which is the value of adding up the size of the first and second grid near the fracture. The size of third grid near the fracture is 52 mm. Similarly, we set the forth case of filter cake thickness as 115.95 mm by adding up the size of three grids near the fracture. Relative permeability and capillary pressure are shown in Figure 13. Other parameters used are in agreement with the basic model, and the numerical simulation results are shown in Table 4. Table 4. The simulation results under different filter cake thickness

M AN U

Filter cake

Load

Water saturation at

thickness

recovery

bottom of fracture

(mm)

(%)

0

27.91

27.45

30.76

63.95

31.35

115.95

30.47

Cumulative production in

3

(m /d)

500 days（104m3）

20

8547.23

200.32

24

13490.49

178.21

34

11119.66

161.24

32

10467.65

155.74

EP

TE D

Maximum production

(%)

Figure 14 shows the water saturation distribution adhere to fracture for 3 cases of different filter cake thickness. As the filter cake thickness increases, the load recovery of the fracturing fluid, water saturation at the bottom of the fracture, and the flow in the fracture during the third stage firstly increase and then decrease. Figure 15 shows the change of water saturation at the bottom of the fracture. The water saturation at the top of the fracture is similar for 4 cases. The load recovery for the case of 63.95 mm increases by 0.6% compared with that for 27.45 mm. The start time of fluid flow at the bottom of the fracture is delayed for about 50 days. The load recovery for the case of 115.95 mm is reduced by 0.88% comparing the case of 63.95 mm. The flow at the bottom of the fracture emerges a little earlier in the third stage. These results suggest that the load recovery increases with the increasing filter cake thickness when the filter cake thickness is small. When the filter cake thickness exceeds a certain threshold, the load recovery starts decreasing instead.

AC C

14 15 16 17 18 19 20 21 22 23 24

SC

2 3 4

12

ACCEPTED MANUSCRIPT

0

1

30

210

500

0

27.45 mm

1

30

210

500

0

1

63.95 mm

30

210

500

115.95 mm

RI PT

1

Figure 14. A section view of water saturation distribution at the grids adheres to the hydraulic fracture for 3 cases of filter cake thickness at different times. The 3 cases of filter cake thickness are 27.45, 63.95, and 115.95 mm, respectively.

5 6 7

Figure 15. The water saturation at the bottom of fracture against time for 4 cases of filter cake thicknesses: 0 mm, 27.45 mm, 63.95 mm, and 115.95 mm.

8 9 10 11 12 13 14 15 16 17

Figure 16 shows the change of gas production rate for 4 cases of filter cake thickness. The larger the filter cake thickness, the lower peak value of the gas production rate, the later the peak appears, and the lower the cumulative gas production. For instance, the maximum gas production rate for the case of 63.95 mm is 2,370.77 m3 lower than that for case of 27.45 mm. Also, the time is delayed by 1.44 days and the cumulative gas production is decreased by 5,5011.63 m3. The load recovery of the fracturing fluid for the case of 63.95 mm is higher than that for the case of 27.45 mm. Both cases show that the water saturation in the matrix and the gas production decrease with time. But the case of 27.45 mm leaks off more water than that for the case of 63.95 mm. This illustrates the existence of filtration is the reason why the water blocking damage is larger in the fracture than that in the matrix, resulting in a decrease of production.

AC C

EP

TE D

M AN U

SC

2 3 4

13

RI PT

ACCEPTED MANUSCRIPT

Figure 16. Gas production rate against time for 4 cases of filter cake thicknesses: 0 mm, 27.45 mm, 63.95 mm, and 115.95 mm.

4 5 6 7 8 9

Our results demonstrate that the fracturing fluid filtration has a great impact on both the fracturing fluid flow back and the gas production. The smaller the filter cake thickness, the less damage to the gas production well caused by the fracturing fluid, and the better production capacity. Therefore, it is important to decrease the thickness of filtration cake by adding an appropriate amount of filtrate reducer to the fracturing fluid. The damage in the gas production wells caused by the water blocking effect in the matrix and the fractures can thus be minimized.

M AN U

SC

1 2 3

Conclusions

11 12 13 14 15 16 17 18 19 20 21 22 23

Water blocking is considered as a cause for potential damage in tight gas reservoirs. In this paper, numerical simulation is used to study the features of water saturation change. It is suggested to use three stages to describe the phenomena of changing water saturation at the top and bottom of hydraulic fractures. At the second stage, the rate of water flow from the fracture to the well bore is lower than that from the matrix to the fracture, resulting a temporary increase of the fracture water saturation. This affects the gas flow and causes a reduction of the effective flow conductivity. As the interfacial tension increases from 20 to 120 Dyne/cm, the load recovery gradually decreases. The effect of water blocking in the fractures on well performance is smaller when the interfacial tension is lower. With the increase of the fracturing fluid viscosity from 0.6 to 1.5 cp, the water blocking effect at the bottom of the fracture becomes more significant, and the damage to the gas production is greater. As the filter cake thickness increase from 0 to 115.95 mm, the load recovery firstly increases, and then decreases after reaching a certain filter cake thickness.

24

Acknowledgments

25 26 27

We acknowledge financial support from the Fundamental Research Funds of the Central Universities (2-9-2015-144). The authors gratefully thank valuable manuscript modification suggestions from Yingkun Fu, who is a PhD student at University of Alberta in Canada.

28

References

29 30 31

Agrawal, S. (2013) Investigation of Liquid Loading Mechanism Within Hydraulic Fractures in Unconventional/Tight Gas Reservoirs and Its Impact on Productivity. MS Thesis, The University of Texas at Austin, May 2013.

AC C

EP

TE D

10

14

ACCEPTED MANUSCRIPT Agrawal, S. and Sharma, M.M. (2013) Impact of Liquid Loading in Hydraulic Fractures on Well Productivity. SPE Hydraulic Fracturing Technology Conference, 4-6 February, Texas, USA.

3 4 5

Agrawal, S. and Sharma, M.M. (2013) Liquid Loading Within Hydraulic Fractures and Its Impact on Unconventional Reservoir Productivity. Unconventional Resources Technology Conference held in Denver, 12-14 August, Colorado, USA.

6 7 8

Alkouh, A., McKetta, S. and Wattenbarger, R.A. (2014) Estimation of Effective-Fracture Volume Using Water-Flowback and Production Data for Shale-Gas Wells. Journal of Canadian Petroleum Technology, 53(5), 290-303.

9 10 11

Bang, V., Pope, G.A., Sharma, M.M., Baran, J.R.Jr. and Ahmadi, M. (2010) A New Solution to Restore Productivity of Gas Wells with Condensate and Water Blocks. SPE Reservoir Evaluation & Engineering, 13(2), 323-331.

12 13 14

Bennion, D.B., Bietz, R.F., Thomas, F.B. and Cimolai, M.P. (1994) Reductions in the Productivity of Oil and Low Permeability Gas Reservoirs Due to Aqueous Phase Trapping. Journal of Canadian Petroleum Technology, 33(9), 45-54.

15 16 17

Cheng, Y.M. (2010) Impact of Water Dynamics in Fractures on the Performance of Hydraulically Fractured Wells in Gas Shale Reservoirs. SPE International Symposium and Exhibition on Formation Damage Control, 10-12 February, Louisiana, USA.

18 19 20

Cimolai, M.P., Gies, R.M., Bennion, D.B. and Myers, D.L. (1993) Mitigating Horizontal Well Formation Damage in a Low-Permeability Conglomerate Gas Reservoir. SPE Gas Technology Symposium, 28-30 June, Alberta, Canada.

21 22

Corey, A.T. (1954) The interrelation between gas and oil relative permeabilities. Producers Monthly, 19(1), 38-41.

23 24 25

Friehauf, K.E., Suri, A. and Sharma, M.M. (2009) A Simple and Accurate Model for Well Productivity for Hydraulic Fractured Wells. SPE Hydraulic Fracturing Technology Conference, 19-21 January, Texas, USA.

26 27

Gruber, N.G. (1996) Water Block Effects In Low Permeability Gas Reservoirs. 47th Annual Technical Meeting, 10-12 June, Calgary, Alberta, Canada.

28 29 30

Gdanski, R.D., Weaver, J.D., Slabaugh, B.F., Walters, H.G. and Parker, M.A. (2005) Fracture Face Damage-It Matters. SPE European Formation Damage Conference, 25-27 May, Sheveningen, The Netherlands.

31 32

Gdanski, R.D., Fulton, D.D. and Shen, C. (2006) Fracture Face Skin Evolution During Cleanup. SPE Annual Technical Conference and Exhibition, 24-27 September, Texas, USA.

33 34 35

Gdanski, R. and Walters, H. (2010) Impact of Fracture Conductivity and Matrix Relative Permeability on Load Recovery. SPE Annual Technical Conference and Exhibition, 19-22 September, Florence, Italy.

36 37

Holditch, S.A. (1979) Factors Affecting Water Blocking and Gas Flow From Hydraulically Fractured Gas Wells. Journal of Petroleum Technology, 31(12), 1515-1524.

AC C

EP

TE D

M AN U

SC

RI PT

1 2

15

ACCEPTED MANUSCRIPT King, G.E. (2010) Thirty Years of Gas Shale Fracturing: What Have We Learned? SPE Annual Technical Conference and Exhibition, 19-22 September, Florence, Italy.

3 4 5

Le, D.H., Hoang, H.N. and Mahadevan, J. (2009) Impact of Capillary Suction on Fracture Face Skin Evolution in Waterblocked Wells. SPE Hydraulic Fracturing Technology Conference, 19-21 January, Texas, USA.

6 7

Mahadevan, J. and Sharma, M.M. (2003) Clean-up of Water blocks in Low Permeability Formations. Annual Technical Conference and Exhibition, 5-8 October, Colorado, USA.

8 9 10

Montgomery, C. (2013) Fracturing Fluids, Effective and Sustainable Hydraulic Fracturing, Dr. Rob Jeffrey (Ed.), ISBN: 978-953-51-1137-5, InTech, DOI: 10.5772/56192. Available from: http://www.intechopen.com/books/effective-and-sustainable-hydraulic-fracturing/fracturing-fluids.

11 12 13

Montgomery, K.T., Holditch, S.A. and Berthelot, J.M. (1990) Effects of Fracture Fluid Invasion on Cleanup Behavior and Pressure Buildup Analysis. SPE Annual Technical Conference and Exhibition, 23-26 September, Louisiana, USA.

14 15 16

Mullen, J. (2010) Petrophysical Characterization of the Eagle Ford Shale in South Texas. Canadian Unconventional Resources and International Petroleum Conference, 19-21 October, Alberta, Canada.

17 18

Parekh, B. and Sharma, M.M. (2004) Cleanup of Water blocks in Depleted Low-Permeability Reservoirs. SPE Annual Technical Conference and Exhibition, 26-29 September, Texas, USA.

19 20 21

Settari, A., Sullivan, R.B. and Bachman, R.C. (2002) The Modeling of the Effect of Water blockage and Geomechanics in Waterfracs. SPE Annual technical conference and exhibition, 29 September-2 October, Texas, USA.

22 23 24

Shaoul, J.R., Koning, J., Chapuis, C. and Rochon, J.P. (2009) Successful Modelling of Post-Fracture Cleanup in a Layered Tight Gas Reservoir. 8th European Formation Damage Conference, 27-29 May, Scheveningen, The Netherlands.

25 26

Shaoul, J., Zelm, L.V. and Pater, C.J. (2011) Damage Mechanisms in Unconventional-Gas-Well Stimulation-A New Look at an Old Problem. SPE Production & Operations, 26(4), 388-400.

27 28

Solimon, M.Y. and Hunt, J.L. (1985) Effect of Fracturing Fluid and Its Cleanup on Well Performance. SPE Eastern Regional Meeting, 6-8 November, West Virginia, USA.

29 30

Tannich, J.D. (1975) Liquid Removal from Hydraulically Fractured Wells. Journal of Petroleum Technology, 27(11), 1309-1317.

31 32 33

Wang, Q., Guo, B. and Gao, D. (2012) Is Formation Damage an Issue in Shale Gas Development? SPE International Symposium and Exhibition on Formation Damage Control, 15-17 February, Louisiana, USA.

34 35 36

Wills, H.A., Miskimins, J.L. and Kazemi, H. (2009) Coupled 3D Numerical Investigation of Hydraulic Fracture Cleanup for Both Slickwater and Gelled Fluids. SPE Annual Technical Conference and Exhibition, 4-7 October, Louisiana, USA.

37

Zanganeh, B., Soroush, M., Williams-Kovacs, J.D. and Clarkson, C.R. (2015) Parameters

AC C

EP

TE D

M AN U

SC

RI PT

1 2

16

ACCEPTED MANUSCRIPT

EP

TE D

M AN U

SC

RI PT

Affecting Load Recovery and Oil Breakthrough Time after Hydraulic Fracturing in Tight Oil Wells. SPE/CSUR Unconventional Resources Conference, 20-22 October, Alberta, Canada.

AC C

1 2

17

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

1. This study investigates the characteristics of water blocking in fractures and the effects of different factors. 2. Three stages are used to describe the change of water saturation, where the second stage indicates water blocking in hydraulic fractures. 3. The load recovery and gas production can be improved by the low interfacial tension. 4. The greater the water viscosity, the more significant the water blocking at the bottom of the fractures. 5. The smaller the filter cake thickness, the less damage to the well caused by the fracturing fluid, and the better production capacity.