Professional Documents
Culture Documents
1
State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation, Southwest Petroleum University
2
Petroleum Engineering School, Southwest Petroleum University
3
CCDC Downhole Operation Company
Summary
Ensuring the sealing integrity of cement sheaths used for shale gas wells during hydraulic fracturing processes has become a major
challenge. The sustained casing pressure, strong temperature difference stress, and cyclic loading and unloading affect the safe long-term
production of shale gas wells. In this paper, we propose a new hydraulic sealing capacity (HSC) evaluation system for cementing inter-
Introduction
Hydraulic fracturing is widely used in shale gas development. However, it is not conducive to the efficient and safe production of shale
gas. Particularly, it can cause annular pressure buildup (APB) problems (Liu et al. 2021). According to statistical data, the number of wells
with APB problems constitutes more than 25% of the shale gas wells in the Marcellus Shale region in the US and the southwestern shale
region in Sichuan, China (Williams et al. 2011; Yan et al. 2018). For instance, by the end of 2015, more than half of the producing wells
in the Fu-Ling and Wei-Yuan gas fields experienced APB problems (Yan et al. 2018). The sealing failure of cement sheaths is the main
reason for APB problems (Liu et al. 2020).
In China, the shale gas exploration and development time is short, and the related technology is not sufficiently mature. In the process
of exploration and development, cementing quality cannot be easily guaranteed owing to deep reservoir burial, high bottomhole tempera-
tures, complex formation pressure systems, and stage and multistage fracturing. Wellbore integrity is one of the important conditions for
ensuring cementing quality. Notably, wellbore failure is very likely to result in accidents such as blowout, interlayer channeling, APB, and
casing corrosion damage (Yousuf et al. 2021). Moreover, it can affect the improvement of the oil and gas recovery coefficient and reduce
the benefits of exploration and development. More seriously, it can cause interlayer channeling, shale gas leakages, high acid corrosion
gas leakages (CO2, H2S), and environmental pollution problems (Roy et al. 2018). In recent years, the frequent occurrence of wellbore
integrity problems has brought new technical challenges to the safe development of shale gas. Cement annulus integrity failures mainly
include cement annulus plastic deformations, tensile failures, interface stripping, and section stratifications (Jian 2013; Zhen 2011). Thus
far, most research has focused mainly on the sealing integrity of cement sheath bodies. By establishing a casing-cement-formation com-
bination model, the characteristics of the cement sheath under the influence of various factors were determined based on a failure criterion
(Jun et al. 2005; Liu et al. 2021). In addition, some scholars have simulated the loading-unloading process of cement annuli by establish-
ing cement annulus integrity evaluation devices and using numerical analysis software to conduct numerical simulations and analyses of
cement annulus integrity (Lin et al. 2020; Xi et al. 2020). The sealing integrity of cement sheath interfaces must be ensured to prevent
APB. Currently, the analysis of the sealing failure of cement sheath interfaces is based on the ideal cementation condition; however,
cement sheath interface cementation is also affected by formation factors, borehole conditions, mudcakes, and cement properties (Yang
et al. 2021). Furthermore, the bonding performances corresponding to the first interface (cement-casing interface) and the second inter-
face (cement-formation interface) are key to ensuring the long-term sealing ability of a cement sheath (Yijin et al. 2021). After cementing,
the risk of sealing failure for a cement-formation interface is considerable because of complex geological and operational conditions (such
as the unstable mudcakes and mudcakes of different thicknesses formed in boreholes, the unstable physical and chemical properties of
borehole rocks, and the high temperature and high pressure of the oil/gas/water layer).
However, few papers have reported quantitative evaluation systems for the HSC of cementing interfaces. Bosma et al. (1999) proposed
a finite element model to study the effects of various loads on the sealing integrity of a cement sheath under different initial stress condi-
tions. Xinyang et al. (2020) designed an interface bonding strength tester to characterize the sealing ability of a cement sheath interface
Apparent Plastic
Viscosity Viscosity
3
(g/cm ) T (s) HTHP (B) Yield Value (Pa) Shearing Force (Pa) (mPa·s) (mPa·s)
1.97 80 2.0 10 3/5.5 72 62
600
ˆ 300
ˆ 200
ˆ 100
ˆ 6
ˆ 3
ˆ Oil/water ratio
145 83 60 35 6 5 85:15
Mudcake Flushing. To simulate the actual mudcake cleaning and scouring process of the well wall, we made appropriate improve-
ments based on the principle of a conventional viscometer. Only the effect of temperature was considered in the flushing process. The
working principle of the equipment is shown in Fig. 2. The diameter and height of the slurry cup are 10 and 10 cm, respectively. The
equipment includes the following modules:
• Cores with mudcakes are fixed and kept in a simulated circulation system by a removable fixture.
• A variable-speed stirrer provides the scouring dynamic for the borehole wall.
• The rotation of the outer cylinder drives the preflushing liquid to scour the mudcake on the borehole wall. The flushing efficiency
is calculated by the formula:
M1 M2
A= 100%, (1)
M1 M0
where Ais the mudcake flushing efficiency; M0 is the quality of rock core, which is drilled from a block of shale (g); M1 is
the quality of core that the mudcake is formed (g); M2 is the quality of core that the mudcake is removed (g).
• A suitable preflushing liquid system is used to simulate the hydrodynamic effect.
• The heating jacket keeps the circulation system at a predefined temperature.
Interface Curing. (1) The self-developed cement sheath interface forming curing mold was assembled, and the mold structure is
shown in Fig. 3. (2) The core after the mudcake washing is placed in the center of the curing mold, and the configured cement slurry is
poured and filled in the annulus. (3) The mold was placed in a pressurized curing kettle after covering the top cover for curing for 7 days.
According to the formation static temperature and pressure curing, curing temperature is 194°F, curing pressure is 20.7 MPa, respectively.
The diameter and height of the simulated cement-formation interface specimens are 50 and 50 mm, respectively, the photo of the physical
object is shown in Fig. 3.
HSC Test. The wellbore simulator is shown in Fig. 4. First, the formed cement-formation interface specimen was placed into a well-
bore simulator. The composite specimen is sealed by wrapping rubber sleeve and applying annular pressure. Subsequently, considering
the actual shallow and deep downhole temperature environment, the annular pressure outside the cement annulus only acts as a seal,
allowing the fluid only to pass through the interface. Finally, by injecting water or inert gas (N2) into one end of the specimen, the break-
through pressure value was gradually increased, and water/gas migration was quantitatively detected. If there is any such migration, it
indicates damage to the structure of the cement-formation interface. The HSC of the cement sheath interface is comprehensively eval-
uated by three indexes: breakthrough pressure (the pressure value corresponding to the end face liquid outlet time), maximum sealing
pressure (the maximum pressure value that the end face can stabilize), and channeling pressure (the pressure value that the end face can
maintain after forming the crossflow channel).
cohesive bonds of the cement-formation interface improved, and the breakthrough pressure increased to 26.8 psi (Fig. 5b). When the flushing
time was increased to 15 minutes, the HSC of the cement-formation interface improved further, with a breakthrough pressure of 164.6 psi and
a maximum sealing pressure of 181.5 psi (Fig. 5c). When the flushing time was increased to 20 minutes, the interface breakthrough pressure
reached 388.5 psi (i.e., 136% higher than that of the 15-minute flushing sample) and the maximum sealing pressure reached 404.5 psi (i.e.,
123% higher than that of the 15-minute flushing sample). After 150 seconds, the channeling pressure of the cement-formation interface
stabilized at 95.9 psi (Fig. 5d). As shown, a fixed time is needed for the prefluid to destroy the dense mudcake grid structure. Fig. 5e shows
that the cement-formation interface breakthrough pressure increased to 618.6 psi after 25 minutes of flushing (i.e., 59% higher than that of the
sample with 20-minute flushing) and the maximum sealing pressure increased to 1,061.8 psi (i.e., 162% higher than that of the sample with
20-minute flushing). When the flushing time increased to 30 minutes, the pressure at the water inlet increased to 2,029.8 psi, while the HSC of
the cement-formation interface did not fail (Fig. 5f). The corresponding relationship between the interface breakthrough pressure, maximum
sealing pressure, channeling pressure, flushing efficiency, and flushing time is summarized herein; the influence laws of the flushing time on
the HSC of the cementing cement-formation interface are shown in Fig. 5. There is a positive correlation between the HSC of the cementing
cement-formation interface and the flushing time. For example, when the flushing time is 15–20 minutes, the HSC of the cementing cement-
formation interface significantly improved, and the effect was enhanced when the flushing time was 25 minutes (Fig. 6). Overall, there was a
positive correlation between the flushing efficiency and flushing time. When the flushing time exceeded 15 minutes, the flushing efficiency of
the cement-formation interface mudcake significantly improved (Fig. 6).
Effects of Flushing Velocity on HSC of Interface. Based on the field construction design data of the shale gas field in southwest Sichuan,
we investigated the impact of the flushing velocity on the HSC of the cementing cement-formation interface; the experimental results
are shown in Fig. 7. We conducted a flushing experiment at 458 rev/min [458 rev/min (0.6 m/s)—10 minutes], 611 rev/min [611 rev/
min (0.8 m/s)—10 minutes], 764 rev/min [764 rev/min (1.0 m/s)—10 minutes], 917 rev/min [917 rev/min (1.2 m/s)—10 minutes], and
1,146 rev/min [1,146 rev/min (1.5 m/s)—10 minutes] for 10 minutes. Fig. 7a shows that when the flushing velocity was less than
0.6 m/s, the breakthrough pressure of the cementing cement-formation interface was only 10.1 psi, and the maximum sealing pressure
and channeling pressure were only 12.2 psi. When the flushing velocity was too low, the cutting pressure of the preflush on the borehole
wall was insufficient to destroy the mudcake structure, leading to the existence of natural cementation cracks between the cement and
the borehole wall. Therefore, the HSC of the cement-formation interface was weak. When the flushing velocity was increased to 0.8 m/s,
the HSC of the cement-formation interface was significantly improved, and the breakthrough pressure was 415.9 psi. In addition, the
maximum sealing pressure was 498.7 psi, and the channeling pressure was 397.8 psi (Fig. 7b). When the flushing velocity reached
1.0 m/s, the HSC of the cement-formation interface was almost unchanged, and the breakthrough pressure became 324 psi. Moreover,
the maximum sealing pressure was 476.3 psi, and the channeling pressure was 420.8 psi (Fig. 7c). Fig. 7d shows that with a flushing
velocity of 1.2 m/s, the cement-formation interface breakthrough pressure increased to 467.2 psi, which is 12% higher than that of the
sample with a 0.8 m/s flushing velocity; moreover, the maximum sealing pressure increased to 564.4 psi, which is 13% higher than that of
the sample with the 0.8 m/s flushing velocity. In addition, the channeling pressure increased to 511.6 psi, which is 29% higher than that
of the sample with the 0.8 m/s flushing velocity. When the flushing velocity increased to 1.5 m/s, the pressure at the water inlet increased
to 2,322 psi, while the HSC of the cement-formation interface did not fail (Fig. 7e). The influence laws of flushing velocity on the HSC
of the cementing cement-formation interface are shown in Fig. 8. When the flushing velocity was less than 600 rev/min (0.8 m/s), the
cement-formation interface was almost unable to form an effective HSC. When the flushing velocity was 600–900 rev/min (0.8–1.2
m/s), the maximum sealing pressure at the cement-formation interface was almost constant, at approximately 500 psi. When the flushing
velocity exceeded 900 rev/min (1.2 m/s), the HSC of the cement-formation interface was significantly improved. Generally, a positive
Fig. 7—Effects of flushing velocity on HSC: (a) 458 rev/min—10 minutes; (b) 611 rev/min—10 minutes; (c) 764 rev/min—10 minutes;
(d) 917 rev/min—10 minutes; and (e) 1,146 rev/min—10 minutes.
correlation was observed between the flushing efficiency of the cement-formation interface mudcake and the flushing velocity. When the
displacement exceeded 0.8 m/s, the flushing efficiency of the cement-formation interface mudcake exceeded 70% (Fig. 8).
Effects of Annulus Cement Contamination on HSC of Interface. To study the influence of the spacer, OBM, and cement slurry
contamination on the HSC of the cement-formation interface, we selected the classic ratio (the recommended ratios are 7:2:1, 7:3, and 1:1:1
for each fluid combination mixture of cement slurry/spacer/OBM) of a cement slurry contamination experiment; the experimental results
are shown in Fig. 9. When the cement slurry contamination ratio was 7:2:1, the breakthrough pressure of the cement-formation interface
was 155.2 psi, the maximum sealing pressure was 328.9 psi, and the channeling pressure was stabilized at 277.6 psi after approximately
6 minutes (Fig. 9a). When the cement slurry contamination ratio was 7:3, the breakthrough pressure of the cement-formation interface
was 64.4 psi, and the maximum sealing pressure was 69.2 psi. Moreover, the channeling pressure stabilized at 54.1 psi after approximately
3 minutes (Fig. 9b). Comparing Fig. 9a with Fig. 9b, when the cement slurry was mixed with a 30 vol% spacer, the breakthrough pressure
decreased by 58.5%, the maximum sealing pressure decreased by 79.0%, and the channeling pressure decreased by 80.5%. When the
cement slurry was mixed with more than 20 vol% of the spacer, the HSC of the cement-formation interface decreased sharply. When the
cement slurry contamination ratio was 1:1:1, the breakthrough pressure of the cement-formation interface was 724.7 psi, but the hydraulic
sealing pressure of the cement-formation interface could always be increased (Fig. 9c). The cement sheath body structure was dissolved,
and the volume was reduced significantly. Moreover, the liquid did not flow through the cement-formation interface after the sample
was cut (Fig. 9d). When the cement slurry was mixed with >10 vol% OBM, the cement sheath structure became loose, its strength was
reduced, and its permeability increased; therefore, the liquid was easily channeled from the cement sheath body (weak channel), while
the cement sheath body structure was dissolved continuously. As the confining pressure increased, the structure of the cement sheath
Fig. 9—Effects of cement slurry contamination on HSC: (a) 7:2:1; (b) 7:3; (c) 1:1:1; and (d) physical photography.
became more compact; therefore, the hydraulic sealing pressure of the cement-formation interface could always increase. The influence
laws of the cement slurry contamination on the HSC of the cementing cement-formation interface are shown in Fig. 10. When the cement
slurry was mixed with a spacer with >20 vol%, or with an OBM with >10 vol%, the cement stone and the cement-formation interface
became loose and porous; hence, the fluid flowed directly from the cement sheath body, which inhibited effective sealing. The greater the
proportion of the cement slurry mixed with other slurries, the worse the sealing performance.
Effects of Preflush Injection Sequence on HSC of Interface. According to the injection sequence and flushing time of the shale gas
field preflush, four types of preflush injection sequences (10-minute spacer; 10-minute spacer + 2-minute flushing fluid; 2-minute flushing
fluid + 10-minute spacer; and 2-minute flushing fluid + 10-minute flushing fluid + 2-minute flushing fluid) were selected for the flushing
experiment. All the experiments were conducted with a return speed of 0.6 m/s; the experimental results are shown in Fig. 11. Fig. 11a
shows that the HSC of the cement-formation interface was extremely weak after only 10 minutes of flushing with a spacer fluid, and the
maximum sealing pressure was only 10.1 psi. When the preflush injection sequence was “10-minute spacer + 2-minute flushing fluid,”
the HSC of the cement-formation interface was considerably improved (Fig. 11b). Fig. 11b shows that the breakthrough pressure of the
cement-formation interface was 225 psi, and the maximum sealing pressure was 233.2 psi. Moreover, the channeling pressure became
stabilized at 151.9 psi after approximately 2 minutes. When the injection sequence of the spacer and flushing fluid was switched, the
HSC of the cement-formation interface was further improved (Fig. 11c). Fig. 11c shows that the breakthrough pressure of the cement-
formation interface was 569.4 psi and that the maximum sealing pressure was 603.4 psi. Moreover, the channeling pressure stabilized
at 552.1 psi after approximately 4 minutes. When the preflush injection sequence was “2-minute flushing fluid + 10-minute spacer +
2-minute flushing fluid,” the HSC of the cement-formation interface was significantly improved (Fig. 11d). Fig. 11d shows that the
breakthrough pressure of the cement-formation interface was 1,850 psi and that the maximum sealing pressure was 2,968.5 psi. Moreover,
the channeling pressure stabilized at 2,458.4 psi. The HSC was the best when the same amount of flushing fluid was used before and after
the spacer flushing. However, the cement slurry was at risk of being contaminated; the flushing fluid could dilute the cement slurry and
considerably reduce its density. When the preflush injection sequence of “flushing fluid + spacer” was adopted, the HSC of the cement-
formation interface was significantly improved. Several researchers have demonstrated that the surfactant can penetrate, adsorb, wet,
and emulsify the mudcake of the OBM, making it loose, such that it can be easily stripped and removed (Van Zanten and Ezzat 2010;
Zhou et al. 2018). The experimental results of this study corroborate the aforementioned reports. The influence laws of the preflush slurry
column structure on the HSC of the cementing cement-formation interface are shown in Fig. 12. A positive correlation existed between
Fig. 11—Effects of preflush injection sequence on HSC: (a) 10-minute spacer (SF); (b) 10-minute SF + 2-minute flushing fluid (FF);
(c) 2-minute FF + 10-minute SF; and (d) 2-minute FF + 10-minute SF + 2-minute FF.
the HSC of the cementing cement-formation interface and the flushing efficiency. Preflushing with flushing fluids in advance could
notably improve the flushing efficiency of the cement-formation interface mudcake (Fig. 12). The HSC of the cement-formation interface
could be significantly improved using a surfactant to emulsify and dissolve the mudcake of the OBM in advance.
Based on the CT scan results in Fig. 13, the pores of the weak interface between the cement sheath and the clean formation were uni-
formly distributed and loosened (Fig. 13a). The number of pores and gaps at the oil-contaminated cement-formation interface were sig-
nificantly increased, and several natural channels are visible (Fig. 13b). Therefore, the HSC of the cement-formation interface was
severely affected by the OBM (Huang et al. 2021). Before the initial cementing operation, the borehole was cleaned with the flushing and
spacer fluids, and the cement-formation interface was significantly improved. Moreover, the number and size of pores significantly
decreased, and numerous micropores were visible (Fig. 13c). Furthermore, we quantitatively analyzed the relationship between the pore
distribution, pore size, pore connectivity, flushing time, flushing speed, and preflush injection sequence and conducted a 3D reconstruc-
tion analysis of the CT images, as shown in Fig. 14. The clean cement-formation interface exhibited a few uniformly distributed pores
with a radius of approximately 0.65 mm (Fig. 14a). Approximately 84% of the pores in the clean cement-formation interface had radii of
300–700 μm (Fig. 14f). When the interface was polluted by the OBM, the number of pores increased; specifically, the number of large
pores (R > 1 mm) ncreased significantly (Fig. 14b). The pore radius of the oil-contaminated cement-formation interface was 600–900 μm,
and the radius of 8.2% of the pores exceeded 1 mm (Fig. 14g). Figs. 14c and 14d show that increasing the flushing time (300 rev/20 min)
and flushing speed (917 rev/10 min) of the spacer appropriately can significantly reduce the number and diameter of pores at the oil-
contaminated cement-formation interface. More than 75% of the pores had radii less than 600 μm (Figs. 14h and 14i). When the preflush
injection sequence was “2-minute flushing fluid + 10-minute spacer + 2-minute flushing fluid,” the number of oil-contaminated cement-
formation interfacial pores were further reduced (Fig. 14e). The radii of more than 90% of the pores were less than 500 μm, and almost
no pores had radii exceeding 900 μm (Fig. 14j).
Fig. 14—3D reconstruction of CT images of cement-formation interface specimens after 7 days of curing: (a, f) blank control group;
(b, g) oil-contaminated control group; (c, h) 300 rev/20 min; (d, i) 917 rev/10 min; (e, j) 2-minute FF + 10-minute SF + 2-minute FF.
Acknowledgments
This work has been supported by the Science and Technology Cooperation Project of the CNPC-SWPU Joint Research Institute (CQXN-
2020-03), the Science and Technology Cooperation Project of the CNPC-SWPU Innovation Alliance (2020CX040000). This support is
gratefully acknowledged.
References
API RP 10B-2. 2013. Recommended Practice for Testing Well Cements, 2nd edition. Washington, DC, USA: American Petroleum Institute.
Bosma, M., Ravi, K., Van Driel, W. et al. 1999. Design Approach to Sealant Selection for the Life of the Well. Paper presented at the SPE Annual Technical
Conference and Exhibition, Houston, Texas, USA, 3–6 October. SPE-56536-MS. https://doi.org/10.2118/56536-MS.
Gray, K. E., Podnos, E., and Becker, E. 2009. Finite-Element Studies of Near-Wellbore Region During Cementing Operations: Part I. SPE Drill & Compl
24 (1): 127–136. SPE-106998-PA. https://doi.org/10.2118/106998-PA.
Huang, S., Feng, B., Li, Z. et al. 2021. Remediation of Oil-Based Mud Contaminated Cement with Talcum in Shale Gas Well Primary Cementing:
Mechanical Properties, Microstructure, and Hydration. Constr Build Mater 300. https://doi.org/10.1016/j.conbuildmat.2021.124047.
Jian, L. 2013. Research on Constitutive Equations and Evaluation Models of Mechanical Integrity of Cement Stone. Chengdu, China: Southwest Petroleum
University.
Jun, L., Mian, C., Gong-hui, L. et al. 2005. Elastic-Plastic Analysis of Casing-Concrete Sheath-Rock Combination. Acta Petrol Sin 26 (6). https://doi.org/
99-103.10.3321/j.issn:0253-2697.2005.06.023.
Jun, G., Weihua, Y., Wenzheng, Q. et al. 2008. Evaluation Method for Isolation Ability of Cement-Formation Interface. Acta Petrol Sin 29 (3): 451–454.
Lin, Y., Deng, K., Yi, H. et al. 2020. Integrity Tests of Cement Sheath for Shale Gas Wells under Strong Alternating Thermal Loads. Nat Gas Ind B 7 (6):
671–679. https://doi.org/10.1016/j.ngib.2020.05.006.
Liu, K., Ding, S., Du, X. et al. 2020. Study on How the Hydraulic Fracturing Affects Integrity of Cement Sheath at Inclined Section of Shale Gas Wells.
Paper presented at the 54th US Rock Mechanics / Geomechanics Symposium, Virtual, 28 June–1 July. ARMA-2020-2095.
Liu, K., Ding, S., Zhou, S. et al. 2021. Study on Preapplied Annulus Backpressure Increasing the Sealing Ability of Cement Sheath in Shale Gas Wells.
SPE J. 26 (6): 3544–3560. SPE-205360-PA. https://doi.org/10.2118/205360-PA.
Ravi, K., Bosma, M., and Gastebled, O. 2002. Improve the Economics of Oil and Gas Wells by Reducing the Risk of Cement Failure. Paper presented at
the IADC/SPE Drilling Conference, Dallas, Texas, USA, 26–28 February. SPE-74497-MS. https://doi.org/10.2118/74497-MS.
Roy, P., Morris, J. P., Walsh, S. D. C. et al. 2018. Effect of Thermal Stress on Wellbore Integrity during CO2 Injection. Int J Greenh Gas Control 77: 14–26.
https://doi.org/10.1016/j.ijggc.2018.07.012.
Van Zanten, R. and Ezzat, D. 2010. Surfactant Nanotechnology Offers New Method for Removing Oil-Based Mud Residue to Achieve Fast, Effective
Wellbore Cleaning and Remediation. Paper presented at the SPE International Symposium and Exhibition on Formation Damage Control, Lafayette,
Louisiana, USA, 10–12 February. SPE-127884-MS. https://doi.org/10.2118/127884-MS.
Williams, H., Khatri, D., Keese, R. et al. 2011. Flexible, Expanding Cement System (FECS) Successfully Provides Zonal Isolation Across Marcellus Shale
Gas Trends. Paper presented at the Canadian Unconventional Resources Conference, Calgary, Alberta, Canada, 15–17 November. SPE-149440-MS.
https://doi.org/10.2118/149440-MS.