Sie sind auf Seite 1von 7

ARMA 18–047

Laboratory Study of Hydraulic Fracturing in Cyclic Injection


Liang Tiancheng, Guan Baoshan, Yan Yuzhong, Fu Haifeng, Xu Yun, Liu Yunzhi
Petrochina Research Institute of Petroleum Exploration and Development, Langfang, China

Copyright 2018 ARMA, American Rock Mechanics Association


This paper was prepared for presentation at the 52nd US Rock Mechanics / Geomechanics Symposium held in Seattle, Washington, USA, 17–20
June 2018. This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and
critical review of the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any position of
ARMA, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the
written consent of ARMA is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 200 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgement of where and by whom the paper was presented.

ABSTRACT: In order to study the influence of cyclic injection on the initiation and extension of hydraulic fracturing, Two large-
scale tri-axial laboratory experiments, with the sample size of 30in*30in*36in, are conducted with spiral perforation completion on
isotropic cement samples by two different ways of cycle and general injection. After the experiment, the pressure data, the actual
crack morphology and acoustic emission (AE) data are analyzed. The results show that: (1) Compared with general injection,
hydraulic fracturing with cyclic injection can effectively reduce the initiation pressure, which is like the behavior of rock under
uniaxial and tri-axial cyclic loading, cyclic injection can bring about fatigue. (2) For spiral perforation completion wellbore, the
hydraulic fracture initiation only starts at the weakest perforation. Once a fracture is open, other perforation holes are hard to
initiate. The stage spacing and cluster spacing should be chosen reasonably in hydraulic fracturing in order to optimize the
efficiency of reservoir stimulation. (3) The occurrence of the Kaiser effect before fracture initiation is the result of pore fluid
diffuse to rock and cause the local rise in pore pressure, the failure modes can still be represented by Mohr circle. The research is of
guiding significance to choose the proper stage spacing and cluster spacing in hydraulic fracturing. Meanwhile, the exploration to
the law of fracture propagation in cyclic injection is of great importance for the development of new fracturing technology.

loading and repeated loading, and used it to evaluate the


1. INTRODUCTION
rock damage process. Patel[6] studied the breakdown
With the large-scale development of unconventional pressure and AE characteristics by using pre-breakdown
reservoirs such as tight oil and gas, coal-bed gas and cyclic injection during hydraulic fracturing. Laboratory
shale gas, the hydraulic fracturing is increasingly results indicated that the SRV generated by cyclic
important. The ultra-deep reservoir strata in Tarim Basin injection hydraulic fracturing is approximately twice
and Sichuan Basin have high temperature and stress, than that generated by conventional injection. Also, the
which causes the high initiation pressure in the hydraulic breakdown pressure recorded during cyclic injection is
fracturing that severely restricted by pumping equipment. 16% lower than that in the conventional injection. Bai[7]
The cyclic injection technology is used to reduce the studied the characteristics of the closed and reopened of
initiation pressure. The mechanics research of the cyclic the fractures in cyclic pumping through numerical
pumping mechanism is fundamental important, but is simulation to determine whether the original fractures
understood unclearly. The large-scale hydraulic reopened or new fractures were formed. The geometry of
fracturing experiment, with the sample size of hydraulic fractures were compared with two methods of
30in*30in*36in, can simulate reservoir stimulation conventional and cyclic pumping injection under
under the in-situ stress with the acoustic emission (AE) different stress conditions and pump injection conditions
monitoring. AE is used to monitor fracture initiation and (injection rate, fluid viscosity, filtration and sand
propagation in real-time and analyze the source screening). Brenne[8] studied AE characteristics by
mechanism of rocks. The large-scale hydraulic fracturing constant pressure and cyclic pump injection. The Kaiser
experiment is an important means to research and effect is defined as the AE will disappear when the load
demonstrate hydraulic fracturing theory[1-3]. is less than the maximum value of the previous cycle,
Nowadays AE/microseismic and initiation pressure which indicates the memory of the material that has
under cyclic loading are mostly studied. AE is an elastic suffered the maximal stress, and it has been used for core
wave excited by the generation of micro-cracks in the stress measurements. Dinske[9] showed that the Kaiser
material[4], which is directly related to the dynamic effect is absent in phases of reopening of the hydraulic
fracture in the rock. Liang[5] adopted a combination of fracture while well observable after termination of fluid
active and passive methods to synchronously monitor injection. After stop of an injection cycle the fluid-
the AE and acoustic wave velocity under uniaxial pressure load on the fracture walls gradually vanishes (a
quasi-diffusional process, which is supported by validity
of the back front of microseismic) and, therefore, the
fracture surface gradually folds up. Falser[10] studied the
effect of different perforation methods on the initiation
pressure, and showed that the plain perforation is
conducive to reduce the initiation pressure. Stoeckhert [11]
studied the characteristics of pumping pressure, strain
and AE of sandstone and shale under the condition of
cyclic injection. Experiments show that the fracture
types of the two lithologic samples are different during
the extension and closure process. Tuokko[12] showed the
Kaiser and the Felicity effect exist in uniaxial
compression multi-stage tests. Wang[13] proposed a Fig.1 Experimental system block diagram
method to improve the identification accuracy of Kaiser
effect points by preloading samples to remove the 2.2.Experimental design
recognition error caused by particle friction and micro- In the experiment, two isotropic cement samples were
fracture closure, and proposed both the frictional AE and selected. According to the similarity between laboratory
rupture AE. Rozhko[14] showed the effect of local pore experiment and oilfield, the loading stress, fluid property
fluid pressure rise on the failure mode by numerical and injection rates used for the four experiments are
simulation. specified in the Table 1. Fig.2 shows well the design and
loading scheme. Two tests with spiral perforation
Most of the studies above have used small-sized samples completion on isotropic cement samples showed the
that are less than 30 cm, whose boundary effect is effect of breakdown pressure by two different ways of
obvious. All the experiments adopting naked-hole cycle and general injection, and verified whether the
completion, but there are usually spiral perforations in Kaiser effect exists for AE induced by hydraulic
oilfields. The Kaiser effect is usually studied under fracturing. The AE characteristic during hydraulic
uniaxial or tri-axial compression test, rather than in fracturing was evaluated. The effect of multistage cyclic
hydraulic fracturing. pump on hydraulic fracturing is studied.
This paper carried out two large-scale (cement sample Table 1 Design parameters of hydraulic fracturing experiment
size of 762mm×762mm×914mm) physical simulation for rock samples
experiments, adopted 24-channel AE sensors. Both
samples were made of isotropic cement samples, which Stress[MPa] Injection
Injection Frac
were not subjected to pressure, and spiral perforation No. σ V, σ H, σ h σmax- rate
mode fluid
σmin [mL/min]
completion is adopted, they were tested by cyclic and
normal injection respectively. We observed the initiation Guar
A Normal 14, 25, 22 11 60-180
pressure and AE changes with time. Besides, after the gum
experiment, we can see the morphology of fractures in Guar 0.137MPa/s
B Cyclic 14, 25, 22 11
the form of cutting blocks. gum 60

2. EXPERIMENTAL PROCEDURE
2.1.Experiment system
Experimental equipment is called large-scale tri-axial
test system for hydraulic fracturing, which mainly
includes the following parts: polyaxial confing system,
borehole injection system, polyaxial load frame, AE
monitoring system and real-time controller (fig.1). The
tri-axial system is capable of independently controlling (a) wellbore (b) loading stress[MPa]
three principle stresses up to 69MPa. The borehole Fig.2 Wellbore and loading stress design
injection system is used to inject frac fluid into the
borehole with flow rate from 0 to over 10L/min and Table 2 Fundamental parameter of sample
pressures up to 82MPa. AE monitoring systems can Young's Poisson Density Compressive Permeability
detect and record AE events during the entire testing. modulus 's ratio ρ[g/cm3] strength [10–3 μm2]
For this purpose a multi-channel Vallen system (AMSY- E[GPa] ν [MPa]
6) with 24 sensors and AEP4 preamplifiers has been 12.5 0.20 2.07 46 0.40
used.
To ensure consistency and comparability with oilfield, 3. EXPERIMENTAL RESULTS
two artificial cement samples made of pour-on-grade G-
grade(HSR) cement were used in the experiment 3.1.Sample A
(cement --quartz powder and water ratio--25:10:14). The The hydraulic fracturing experiment was carried out on
wellbore adopts a standard tubing with an outside the sample A with the conventional pumping method.
diameter of 73 mm and an inside diameter of 62 mm. The pressure curve and displacement are shown in Fig. 4,
After being poured, the cement samples were maintained the total volume of the injection was 2743mL.
in standard environment for 28 days to ensure the Fracturing was carried out with injection rate of
strength requirement. 60mL/min. The breakdown pressure and extension
pressure were observed. The breakdown pressure was
Using prefabricated special solid materials to simulate the maximum of the pressure curve. The breakdown
perforation, we can design the parameters of different pressure was 18.6MPa and the propagation pressure was
perforation (hole depth, hole diameter and method). 14.18-14.59MPa. If the fracturing fluid was kept
After pouring, we poured a special liquid into holes to pumping after fracture initiation, the pressure did not
dissolve the solid without residue to ensure the good significantly change. Then the injection rate was raised
communication between the perforation hole and the twice at 120,180mL/min until the fracturing fluid
sample. In this way the purpose of simulating the true volume went up to the limiting value of 3000mL. The
perforation morphology is achieved, which can simulate final total injected volume was 2743 ml.
the completion methods of horizontal perforation, spiral
perforation and casing sleeve perforation as well.
2.3.Real-time AE monitoring and positioning
AE device, gain of preamplifier, sampling rate,
thresholds and other parameters can be adjusted
according to the noise level, lithology and other factors.
In all the experiments in this paper, we set the gain of
preamplifier to 30dB and the sampling rate to 5MHz. All
sensors are calibrated by active ultrasonic test to make
the same sensitivity. The pumping rate and amount can
be identified according to the pressure curve and real-
time monitoring results of AE. During the whole
experiment, the pressure of each system and the changes Fig.4 Injection rate and injection pressure with time
of the total injection volume with time were recorded
and the experimental data were output. In addition to Fig.5 shows the AE amplitude, the hits number and the
recording the pump pressure and real-time AE event pumping pressure changes with time. The AE amplitude
location, the samples can be cut into pieces to observe is positively correlated with the pump pressure. With the
the fracture morphology after the experiment. Fig.3 pumping pressure increasing, the amplitude and AE rate
shows a series of AE sensors position, the AIC algorithm become larger.
is used to automatically pick up P-wave arrival time,
After the experiment, the samples were cut into slices
which has greatly improved the pick-up accuracy
according to the liquid loss on the surface of the sample
comparing with the threshold method.
and the AE results. The fracture morphology is shown in
fig.6. The result shows that the fractures only open from
No.2 perforation hole, form horizontal fracture
perpendicular in the direction of the minimum horizontal
principal stress. According to the pressure curve and AE
results, it can be preliminarily determined that the
fracture has reached beyond the boundary after the
initiation of fracture.
3.2.Sample B
The hydraulic fracturing experiment was carried out on
the sample B with multi-stage cyclic pumping method.
The pressure curve and injection volume are shown in
Fig.7-8, the total volume of the injection fracturing fluid
Fig.3 AE sensor positions with label was 1816mL. The total number of loading cycles was 5
and the loading rate was controlled by force mode with a
loading rate of 0.137MPa/s. The sample started initiation
at the 5th cycle, the breakdown pressure was 13.8MPa. reached the maximum indicates that the force transfer
The reopening is controlled by injection rate at 60 distance increases with the increase of pump pressure,
mL/min, which is the same as sample A. The pressure of which leads to more AE events.
initiation and propagation pressure were observed. The
There are AE events both in the loading and unloading
reopening pressure was 12.17MPa and the propagation
stages, but the AE amplitude is lower in the unloading
pressure was 9.43-9.5 MPa.
stage. The AE number accurately located throughout the
experiment was 990, but the number of AE monitored
was 10,156.

(a) Amplitude

(a) Amplitude

(b) AE rate
Fig.5 AE amplitude, rate and injection pressure with time
(b) AE rate
Fig.7 Injection pressure, AE rate and AE amplitude with time

Fig.6 Fracture morphologies of hydraulic fracturing


(a) Cumulative frac fluid volume and wellbore pressure curve
Fig.7 shows the AE results for sample B. The AE
characteristic parameters are analyzed and the curve is
drawn. Because cement sample has never been subjected
to pressure, AE signal exists throughout the whole
experiment. At different stages of the cycle, there
indicate the significant difference in AE rate and
amplitude. With the increase of the times of cyclic
loading, the AE events and amplitude gradually
increased. The fact that the AE rate and amplitude (b) Located AE event distance from wellbore with time
Fig.8 Experiment data of hydraulic fracturing for sample B

Due to the high injection rate and the brittle samples, the
boundary has been reached at the 5th stage, which was
based on AE spatial location and distance analysis
between the incident points and the wellbore (see Fig.8,
9). After the initiation at cycle 5, the AE number
remained stable except at the point of re-initiation of
cycle 6.
After the experiment, the sample was cut into pieces, and (b) Surface fracture morphology
the morphology of the fracture was shown in Fig.10. The Fig.10 Fracture morphologies for sample B
experimental results show that the fractures initiation is
at No.2 perforation hole, which forms a single horizontal
fracture, whereas the other five perforation holes do not 4. EXPERIMENTAL ANALYSIS AND
product any obvious fracture. The fracture propagates DISCUSSION
asymmetrically and expands to the southwest. 4.1.Breakdown pressure
The injection pressure curve of two experiments is
shown in Fig.11. The dashed line is a conventional
injection pressure curve, and the solid line is a multi-
stage cyclic injection pressure curve. The other three
lines are three direction loading stress. The two
experiments adopt the same injection rate. The
breakdown pressure is 13.8MPa at the fifth cycle, which
is close to the minimum principal stress and is lower
than 18.6 MPa of the conventional pump injection mode.
Compared with conventional pump injection, multi-stage
(a) Top view cyclic injection can effectively reduce the breakdown
pressure, which is similar to the behavior of rock under
uniaxial and tri-axial cyclic loading. The fatigue damage
is caused by cyclic loading.

(b) lateral view


Fig.9 AE location results of cyclic injection

Fig.11 The pressure curves of conventional injection and


cyclic injection

4.2.Initiation position
In both experiments, the spiral perforation completion
mode is adopted with six holes and a plain angle of 60°,
which is the same as the hydraulic fracture on the
oilfield spot. After the experiment, the sample is cut
(a) Fracture morphology near wellbore along the wellbore. The initiation position of hydraulic
fracture was observed (see fig.6, 10). The two
experiments were initiated from the position of No. 2
hole, and the other five perforated holes had no obvious
initiation. The fracture morphology of the two
experiments was the same, which only formed a single When the pressure of fracturing fluid in wellbore
fracture. It shows that hydraulic fracturing fracture only increases, the circumferential stress of the sample is
initiates at the weakest point. Once there exists the larger than the tensile strength of the rock, the sample
initiation, the other perforation holes are difficult to will initiate along the direction of parallel maximum
initiate. The conclusion is of great significance for principal stress, which can be expressed by the Mohr
cluster perforation in oilfield. The optimization of circle. As is shown in Fig. 13, high pressure fracturing
reservoir stimulation need reasonable stage spacing and fluid injection causes local rise of wellbore and pore
cluster spacing. pressure, which results in fracture initiation. It is the
same as the failure mode of tri-axial rock mechanics
4.3.Kaiser effect experiment. However, fracture’s re-opening and
AE has a memory to its stress history during the cyclic fracture’s propagation mainly affected by fracture
loading and unloading of rock materials. The Kaiser toughness and confining pressure.
effect is used to measure the in-situ stress under uniaxial
loading. Because cement samples have never been
subjected to loading, AE events develop through the
whole experiment, and the AE events mainly generate on
the loading stage before the initiation of fractures. There
are only a few events in the unloading stage. As can be
seen from the fig.12, in the elastic deformation stage of
the sample, the Felicity ratio is larger than 1.0, such a
material memory to its stress history is named the Kaiser
effect. Fig.13 Change of Mohr circle with wellbore pressure

Hydraulic fracturing by cyclic injection can effectively


reduce the breakdown pressure. The results are similar to
uniaxial and tri-axial cyclic loading rock mechanical
experiment, which due to fatigue damage caused by
cyclic loading. At the same time, this conclusion can
also be verified by AE monitoring results: 1) the
distance from located AE events to wellbore becomes
longer with the increase of injection series (see Fig. 8). 2)
Kaiser effect exists by cycle injection (see Fig. 12).

5. CONCLUSION
Fig.12 Relationships among of amplitude of AE event, pump
pressure and time by cyclic injection (1) Compared with general injection, hydraulic
fracturing with cyclic injection can effectively reduce the
4.4.Analysis on mechanism of cyclic injection initiation pressure, which is like the behavior of rock
In both hydraulic fracturing experiments, vertical wells under uniaxial and tri-axial cyclic loading, cyclic
are used. The samples are tight cement, and the injection injection can bring about fatigue.
time is short. The stress distribution around the wellbore (2) For spiral perforation completion wellbore, the
is as follows: hydraulic fracture initiation only starts at the weakest
perforation. Once a fracture is open, other perforation
 r  PW  P0 (1) holes are hard to initiate. The stage spacing and cluster
  3Sh  SH  PW  P0 (2) spacing should be chosen reasonably in hydraulic
fracturing in order to optimize the efficiency of reservoir
As is shown,  r is the radial stress around the borehole,
stimulation.
  is the circumferential stress around the borehole, PW (3) The occurrence of the Kaiser effect before fracture
initiation is the result of pore fluid diffuse to rock and
is the pressure of fracturing fluid in wells, P0 is pore cause the local rise in pore pressure, the failure modes
pressure in the formation. The dry sample is used in the can still be represented by Mohr circle.
experiment, and its value is 0, S H is maximum principal As is presented above, the research is of guiding
significance to choose the proper stage spacing and
stress, S h is the minimum principal stress.
cluster spacing in hydraulic fracturing. Meanwhile, the
exploration to the law of fracture propagation in cyclic
injection is of great importance for the development of 12. Tuokko T, & Poutanen M S. Acoustic emission based
new fracturing technology. rock stress measurement method[C]//ISRM Regional
Symposium-EUROCK. Cavtat: [s. n.], 2009.

REFERENCES 13. WANG Xiao-qiong, GE Hong-kui, SONG Li-li, et al.


Experimental study of two types of rock sample
1. FU Hai-feng, CUI Ming-yue, ZOU Jing, et al. acoustic emission events and Kaiser effect point
Experimental study of Changqing sandstone fracture recognition approach[J]. Chinese Journal of Rock
propagation based on acoustic monitoring[J]. Journal of Mechanics and Engineering, 2011, 30(3): 580-588.
Northeast Petroleum University, 2013, 37(2): 96-101.
14. Rozhko A Y, Podladchikov Y Y, Renard F. Failure
2. LIANG Tian-cheng, FU Hai-feng, LU Yong-jun, et al. patterns caused by localized rise in pore-fluid
Source mechanism studies of acoustic emission in overpressure and effective strength of rocks[J].
large-scale hydraulic fracturing experiment[C]//SPE Geophysical Research Letters. 2007, 34(22): 304.
Asia Pacific Hydraulic Fracturing Conference. Beijing:
[s. n.]. 2016.
3. LIANG Tian-cheng, FU Hai-feng, LIU Yun-zhi, et al.
The property study of acoustic emission in hydraulic
fracturing simulation experiment of different
rocks[C]//SEG/SPG International Geophysical
Conference. Beijing: [s. n.], 2016.
4. ASTM International. ASTM E1316-10 Standard
terminology for nondestructive examinations[S]. West
Conshohocken, Pennsylvania: ASTM International,
2003.
5. LIANG Tian-cheng, GE Hong-kui, GUO Zhi-wei, et al.
Evaluation of rock damage state with acoustic emission
and velocity variation[J]. Earthquake Research in China.
2012, 28(2): 154-166.
6. Patel S, Sondergeld C, Rai C. Laboratory studies of
cyclic injection hydraulic fracturing[C]//SEG
International Exposition and 86th Annual Meeting,
Okla: [s. n.], 2016.
7. Bai M, McLennan J, Guo Q, et al. Cyclic Injection
Modeling of Cutting Re-Injection[C]//ARMA/USRMS
41st U.S. Symposium on Rock Mechanics. Colorado: [s.
n.], 2006.
8. Brenne S, Alber M. Interpretation of laboratory
hydraulic fracturing expriments with emphasis on
acoustic emissions[C]// The ISRM International
Symposium: Stockholm: [s. n.], 2012.
9. Dinske C, Shapiro S. Microseismicity induced by
hydraulic fracturing: Evaluation and interpretation in
terms of the Kaiser effect[C]//SEG 2007 Annual
Meeting. San Antonio: [s. n.], 2007: 1594-1598.
10. Falser S, Mo W. Reducing breakdown pressure and
fracture tortuosity by in-plane perforations and cyclic
pressure ramping[C]//ARMA the 50th US Rock
Mechanics/ Geomechanics Symposium. Houston: [s.
n.], 2016.
11. Stoeckhert F, Molenda M, Brenne S, et al. Fracture
propagation in sandstone and slate-Laboratory
experiments, acoustic emissions and fracture
mechanics[J]. Journal of Rock Mechanics and
Geotechnical Engineering.2015, 7(3): 237-249.

Das könnte Ihnen auch gefallen