Content uploaded by Yue Xiao
Author content
All content in this area was uploaded by Yue Xiao on Sep 29, 2017
Content may be subject to copyright.
67www.petrodomain.com
Pulsed Power Plasma Stimulation Technique - Experimental Study
on Single Pulse Test for Fracture Initiation
Yue Xiao, Waylon House, Texas Tech University, Ebru Unal, M. Y. Soliman, University of Houston
Promotion of Innovation
Abstract
is paper discusses an experimental study of single pulse PPPS (Pulsed Power Plasma Stimulation) with 2 kJ stored electrical energy. e
primary goal of the study is to investigate the potential of PPPS as a cost-eective near wellbore stimulation technique that does not require
the injection of water or the use of chemical additives. e technique envisions the stimulation of the wellbore by creating self-propped short
fractures and/or by changing the near wellbore permeability by more than one order of magnitude.
During these experiments, a pulsed power plasma was created by electrical discharge that generates a pressure shock wave and other forms of
energy. is paper will focus on the created shock wave. e presented experiments show that PPPS creates multiple fractures emanating from
the wellbore. e use of ionizable material in the plasma discharge introduces an exothermic chemical reaction that enhances the produced
energy by up to two orders of magnitude. e level of the exothermic energy depends on several parameters including the type and amount of
the elements used in creating the plasma.
Two groups of cylindrical concrete test samples were built under dierent conditions: water-to-cement ratio, wellbore size, and completion
method. e aim of this work was to test the performance of PPPS with current completion methods at various underground conditions. is
experimental work focused on study of fracture initiation during single-pulse tests.
Experiments using 2 kJ of stored energy achieved peak pressures of up 200 MPa (29,000 psi). ese shockwave pressures were generated by
typical plasma durations of 6 microseconds. e produced eective mechanical energy far exceeded electrical energy input indicating that a
thermite reaction contributed about 90% of the generated energy. However, no separate pressure pulse due to a thermite component was detected
suggesting that the two events happened almost simultaneously. Small-bore vertical models with perforations that do not penetrate the rock
produced the most signicant eects. In all cases, signicant multiple micro-fractures were created and propagated.
Introduction
Recently the possibility of electrical-pulse induced fractures has drawn industry attention especially in ultra-tight sandstone and shale reservoirs.
“During this process, electrical energy is used to introduce mechanical loads into the rock. If high enough, this loading will fracture rock.”
(Gandossi 2013). In one technique, PAED (Pulsed Arc Electrohydraulic Discharge), an intense compressive shock wave is produced. Unlike
quasi-static hydraulic fracturing, PAED exerts a sharp dynamic load on the formation that creates distributed micro-cracks (Reess et al. 2009;
Maurel et al. 2010). A series of studies of shock waves generated by underwater PAED have been published. e results of those experiments (Chen
et al. 2012; Chen et al. 2014b; Cabot et al. 2016) showed that the intrinsic permeability of rock is increased when input electrical energy reaches a
certain critical level. Concrete behavior under dynamic loading induced by shock wave was characterized using computational simulation (Chen
et al. 2011; Chen et al. 2014a; Cabot et al. 2016).
Experiments using the PAED method found no more than 25% of the stored electrical energy is converted into mechanical energy in the form
of shock waves, while most of the energy is lost to heating (Zhou et al. 2015). e intrinsic nature of PAED limits its ability to increase rock
permeability around the wellbore and it may only work as partial replacement for small to medium hydraulic fracturing treatments. As an
extension of conventional underwater PAED, we have identied a new stimulation method named PPPS (Pulsed Power Plasma Stimulation)
which takes advantage of an exothermal thermite reaction. In PAED the electrical discharge occurs between electrodes in air or water. In PPPS,
the electrical discharge occurs in a fusible link in water. Depending on the link composition, an additional electrochemical reaction can occur
which supplements the energy transfer of PAED. Using aluminum in the fusible link that triggers the plasma introduces a second source of energy
from the recombination of aluminum and water ions in a thermite reaction. When aluminum is ionized into the plasma, it reacts with water ions
producing a strong exothermic reaction. Lee (1993) found that this thermite reaction could produce 40 – 100 times the electrical energy that
started the reaction. Given the same level of input electrical energy, PPPS outperforms PAED with higher energy transfer eciency and better
stimulation performance.
e primary aim of the study is to investigate the potentials of PPPS as a cost-eective near wellbore stimulation technique that does not require
the use of water and/or additives, and the possibility of creating self-propped short fractures. Instead of creating bi-wing fracture as conventional
hydraulic fracturing does, pulsed power plasma fracturing has the potential of creating a multi-fracture system with micro fractures connected
with pre-existing fractures; or to increase the reservoir permeability near the wellbore by two to three orders of magnitude. In tight reservoirs,
multiple fractures are benecial to reservoir stimulation when drawdown is limited to a narrow zone around the fractures. Multiple fractures are
also preferred when pre-existing natural fractures are aligned with potential fracture planes (Mao et al. 2012).
is paper discusses an experimental program of single pulse PPPS with 2 kJ stored electrical energy. e practical performance of PPPS depends
highly on the electrical energy transfer eciency and properties of target samples. Two groups of cylindrical test samples were built under
dierent conditions: water-to-cement ratio, wellbore size, and completion method. e aim of this work was to test the combined performance
of PPPS with current completion methods at various underground conditions.
Processes and Mechanisms
e process described in this paper involves complex interrelated physical and chemical phenomena. First, electrical energy is stored slowly into a
capacitor. is energy is then released in a very short time (microseconds) through a spark gap or fusible link creating an intense plasma arc and
attendant intense shockwave. e ensuing processes depend on the media in which the plasma is created, and the materials used in the creation
of the plasma. e processes involve three types of new energy forms:
1) Mechanical energy in the form of compressive shock waves.
2) Electrochemical energy released by chemical reactions of plasma ions.
3) Electromagnetic energy in the form of near and far zone elds created by the plasma current.
ere will be some loss of energy in the form of heat. Mechanisms of other losses of energy are Ohmic losses in conductors and Ohmic losses in
matrix materials. Figure 1 summarizes the discussion above.
Figure 1. Analysis of Energy Transformation and Dissipation.
e output mechanical energy in the form of compressive shock wave is the major cause of deformation in test samples, whether it comes from
the initial plasma or consequent thermite reaction.
As shown in Figure 2, the process of energy release consists of four regimes. Ohmic heating by the electric current vaporizes the aluminum
link and provides a path for the sudden electrical discharge. e discharge is accomplished through rapid ionization of molecules between the
electrodes. is ionization is crucial in overcoming the impedance to the reaction caused by the protective oxide layer on the metal surface. Fast
electrical ionization of the reactive metal and water ignites and sustains the metal/water reaction that in turn generates hydrogen gas and thermal
energy.
2Al +3H2O => Al2O3 + 3H2
Figure 2. Energy Processes.
Conductor
Heating
Ionization
(Plasma)
Exothermic
Recombination
Conductor
Melting &
Vaporization
Mechanical Energy Release
69www.petrodomain.com
Lee (1993) proposed that “Due to the inherent time lapse for transport of the water molecules and for the chemical interaction of the reacting
molecules, the completion of chemical reaction lags behind the electrical explosion of the metal conductor triggered by the pulse” .
Figure 3. Time Correlation of Electrical Excitation and Chemical Reaction (Aer Lee 1993).
e electrical energy input vaporizes/ionizes the aluminum link and surrounding media explosively, releasing mechanical energy. Chemical
exothermic reactions, amongst the ions in the plasma, release additional energy creating a further source of mechanical energy and potentially
heating the plasma as well.
Optimizing the whole process is complex. For example, the mass and shape of the link are critical. e eective conductance of the aluminum
link must be low enough to ohmically absorb enough of the input current pulse to heat the link to ionization temperatures. At the same time,
the amount of aluminum consumed in the process should be as large as possible to maximize the number of aluminum ions in the plasma for
chemical recombination.
Experimental Program
A series of experiments on concrete samples were performed utilizing a 10 µf capacitor charged to 20 kV. Discharge time of the stored energy into
the samples was typically 6 µs. e power delivered during a pulse was measured as well as the resulting shockwave pressures. Various test samples
were built under dierent conditions: water-to-cement ratio, wellbore size, and orientation and completion method. Both current (I) and voltage
(V) of the electrical discharge were measured as a function of time. e resulting pressure pulse was measured with PVDC pressure transducers.
Figure 4 is a schematic cross-sectional view of a vertical experimental arrangement. e borehole is tted with a PVC sleeve to keep the water in
the hole from penetrating into the concrete. e electrical pulse is delivered through the top plate of the sample xture. No conning pressure is
applied during the single-pulse test for current experiments.
During each test the electrical parameters are measured as a function of time. e voltage is measured directly across the cable delivering the
pulse and a Rogowski coil measures the current through the cable. A PVDC high speed pressure transducer measures the shockwave pressure at
one end of the bore. Figure 5 shows the typical voltage and current output into the sparker link. e circuitry involved includes a capacitor with
inherent inductance, a short length of cable and the link itself. us, the electrical system has resistive and reactive components and the complex
power delivered to the load, the product of current and voltage (IV), has reactive and real components. Only the real resistive components (I and V
both in phase) represent real absorbed energy by the load. Out of phase components of the power (e product of voltage and current is negative)
represent reactance – energy reecting between inductive and capacitive components. In Figure 5 the positive current region corresponds to
current owing out of the capacitor into the load and the negative current is current owing back into the capacitor due to stored energy in the
reactive inductance. e negative voltage corresponds to this same source. Figure 6 shows the real power delivered to the link, as a function of
time. Figure 7 shows the pressure pulse of the produced shockwave.
e rst pressure peak in Figure 7 at about 6 µs corresponds to the rst power peak in Figure 6. e vaporization/ionization of the aluminum
link produces a sudden change in the load impedance and, if all the capacitors’ energy has not been dissipated, the reactive components will
reect this energy. Figure 6 shows a secondary discharge at 16.5 µs corresponding to this reected energy. e second pressure peak in Figure 7
corresponds to the second power peak in Figure 6 produced by this ringing. Figure 8 illustrates the pressure oscillation during the rst 8 micro
seconds. Tab le 1 summarizes pressure pulses of the various experiments that were run.
e shockwave pressure is measured by the solid state transducer mounted on the solid surface at the bottom of the bore. e transducer captures
the stress wave prole as it impacts the surface of the detector and passes through. e front of the shockwave compresses the molecular matrix
in the detector whereas the back of the shockwave expands the molecular matrix of the detector. e matrix structure oscillates aer the wave
has passed through. e transducer converts the compression and expansion to voltage, mirroring the forces and stresses experienced by the
detector lm. ese waveforms are typical of shockwaves in solids –compression and expansion followed by transient ringing characteristic of the
molecular binding in the solid through which the shockwave passes. e peak value is the initial impact value. e pressure in this context is a
vectorial quantity where the positive and negative signs indicate direction. It should not be confused with the common understanding of pressure
as a scalar quantity where absolute pressure has to be positive value.
70 Hydraulic Fracturing Journal | July 2017Volume 4 - Number 3
Figure 4. Schematic Cross-sectional View of Sample Arrangement.
Figure 5. Voltage –Current Output. Figure 6. Real Power Input (Product of Current
and Voltage) of the primary pulse.
Figure 7. Pressure-time prole showing
Main Pulse and Two additional pulses.
Figure 8. Pressure-time Prole showing the 1st Primary Pulse.
Table 1 . Summary of pressure pulses.
Pulse Number t0, μs Duration, µs Peak Pressure, MPa
1 6 2.5 210
2 16.5 1.5 75
3 47 1 23
71www.petrodomain.com
Samples and preparation
Cylindrical test samples were made with Portland Type I cement. PVC pipe placed at the center of the sample served as casing. Aluminum foil
with coiled length of 50.8 cm (20 in.) was used as conductor and ignitor of the thermite reaction between electrodes (Lati 2010).
Two dierent groups of test cement samples were made to simulate reservoir conditions. e rst group was created using water-to-cement ratio
of 0.7 with minimal mixing. is was to create a heterogeneous system with pre-existing natural fractures. e second group was made using
water-to-cement ratio of 0.6 with higher mechanical properties and was mixed appropriately to create a homogenous medium without pre-
existing fractures. Geometric and experimental set-ups of each test sample are summarized in Tab le 2 .
e orientation of the “perforations” slits is shown in Figure 9
Figure 9. 120⁰ Perforation on the Middle of PVC Pipe.
Results of Single Pulse Test
Test Sample 1: Mimicking Vertical Well with 180 ° Slit on 2.54 cm (1 in.) Casing
Test Sample #1 was built with 180° slit casing. Slits were wrapped by PVC tape beforehand, as shown in rst picture of Figure 10. e results show
that single pulse test creates vertical multi-fractures with two of them initiated from slits and the rest of them occurring randomly. e pulse
created hairline fractures and the sample was torn apart very easily with minimum force applied to the core. Since no pressure data was recorded
during the rst test, shock wave could not be characterized.
Figure 10. Test cement sample 1.
Test Sample 2: Mimicking Vertical Well with no perforation/slit on 2.54 cm (1 in.) casing
Test Sample # 2 was set in comparison with test sample # 1 in order to see the performance of fracture initiation without slits. Four fractures were
Table 2 . Summary of test cement samples.
Test Cement
Sample
Dimension D×H
(cm) Borehole size (cm) Slit/Perforation Orientation Number of
Fractures
1 28×25 2.54 180° slits Vertical 4
2 29×25 2.54 None Vertical 4
3 28×25 1.91 None Vertical 3
4 28×25 2.54 3 perfs at 120° Vertical 5
72 Hydraulic Fracturing Journal | July 2017Volume 4 - Number 3
initiated from the PVC pipe, Figure 11. It was found that in the case of unperforated or non-slit casing, even though the propagation of fractures
were dominated by existing natural fractures and four macro-fractures were created, the test sample was not fragmented into pieces and the
performance of the shock wave in dynamic fracturing was weakened. e east fracture did not go through the whole sample and other fractures
were thinner than those in sample #1.
Figure 11. Test cement sample 2.
Test Sample 3: Mimicking Vertical Well with no perforation/slit on 1.91 cm (3/4 in.) casing
Test sample # 3 was built with smaller casing of ¾” without any perforation or slit. Single pulse test exerts immediate fragmentation and the
sample was broken up into three parts, see Figure 12 for details.
Figure 12. Test Cement sample 3.
Test Sample 4: Test Sample with Partial Perforation with 120° Phase Angle 2.54 cm (1 in.) Casing Mimicking completion of a horizontal well
by creating three perforation in one plane perpendicular to the wellbore.
An important observation is that the single burst was able to create hairline fractures at the supercial layer of the cement samples. e formation
of hairline fractures is probably because the output mechanical energy of the current single pulse test is not high enough to create micro fractures
in a large sample. e pressure pulses are not sucient and focused enough to create micro fractures. However, with shock wave reection at
the air/cement boundary (Figure 13), the supercial layer of test sample 4 exhibited immediate cracking aer the rst blast (Figures 14 and 15).
Concluding Remarks
Test results from this study showed that pressure as high as 200 MPa (29,000 psi) was reached (Figure 7 and Figure 8) with input electrical energy
of 2 KJ within milliseconds. e aluminum-water reaction improved the eective output mechanical energy by 1600 % on the basis of input
electrical energy despite the fact that 31% of the input energy was lost due to electrical circuit oscillations. e pressure measurements did not
show the time dependence expected based on Figure 3. Performance of PPPS on test samples demonstrated that pre-existing natural fractures
and completion methods like perforation or slits have an important impact on fracture propagation direction. It is important to notice that
Figure 13. Test Cement Sample 4. Figure 14. Test Cement
Sample 4 aer the 1st blast.
Figure 15. Supercial Map Cracking of
Test Sample 4 aer the 1st blast.
73www.petrodomain.com
reections of shock waves could also cause immediate cracking around dierent medium boundaries. e fracture initiation pattern of single-
pulse PPPS will provide guidance for next-step multi-pulse tests and characterization of shock waves.
e performance of PPPS is highly dependent on the mechanical properties of reservoir rock and completion methods. e threshold energy
in order to create micro fractures increases with the mechanical strength of the rock and completion technique. Future work will lay more
emphasis on multiple-pulse tests, as well as various types of completion and stimulation methods, and wellbore size. Understanding of the
thermite reaction coupled with the electrical pulse will also need deeper investigation before nal eld trials. Furthermore, the characterization
and simulation of shock waves are needed for more accurate prediction of fracture initiation patterns under dynamic loads generated by pulsed
power plasma techniques.
Compared with conventional quasi-static fracturing, pulse fracturing is expected to have the advantage of inducing multiple fractures. Conventional
quasi-static hydraulic fracturing normally induces bi-wing fractures only, but multiple fractures are benecial for reservoir stimulation in very
tight reservoirs when the drawdown is limited to a narrow zone around the fractures. Multiple fractures are also needed when the reservoir has
natural fractures aligned with preferred fracture plane. In that case, the o-plane hydraulic fractures will intersect natural fractures so that the
well is better connected to the natural fracture system.
Current stage of experimental study on single pulse test does not involve conning pressure on test sample. e experimental results on creating
the multi-fractures system need further verication for both vertical and horizontal congurations before eld-scale applications.
Nomenclature
D = diameter of test sample, cm
H = Height of test samle, cm
I = Current, A
IU = Product of voltage and current, A∙V
t0 = time of puls intiation, μs
U = Voltage, V
Reference
Chen, W., Maurel, O., Reess, T. (2011). “Modelling Anisotropic Damage and Permeability of Mortar under Dynamic Loads”. European Journal of
Environmental and Civil Engineering (15):727-742. DOI: 10.3166/EJECE. 15. 727-724
Chen, W., Maurel, O., Reess, T. (2012). “Experimental Study on an Alternative Oil Stimulation Technique for Tight Gas Reservoirs based on
Dynamic Shock Waves generated by Pulsed Arc Electrohydraulic Discharges”. Journal of Petroleum Science and Engineering (88-89):67-74.
DOI: 10.1016/j.petrol.2012.01.009.
Chen, W., Borderie, C.L., Maurel, O. (2014a). “Simulation of Damage-permeability Coupling for Mortar under Dynamic Loads”. Int. J. Numer.
Anal. Meth. Geomech (38):457-474. DOI: 10.1002/nag.2212.
Chen, W., Maurel, O., Borderie, C.L. (2014b). “Experimental and Numerical Study of Shock Wave Propagation in Water Generated by Pulsed Arc
Electrohydraulic Discharges”. Journal of Heat Mass Transfer (50): 673-684. DOI: 10.1007/s00231-013-1262-4.
Gandossi, L. Report EUR 26437 EN (2013). “An Overview of Hydraulic Fracturing and Other Formation Stimulation Technologies for Shale Gas
Production”. 2013. Petten. e Netherlands. DOI: 10.2790/99937.
Jizba, D.L. (1991). “Mechanical and Acoustical Properties of Sandstones and Shales”, Doctoral Dissertation, Stanford University, Stanford, CA
(March 1991)
Lati, S. (2010). “An Experimental Investigation to Study the Eect of Pulsed-Power Discharge on Rock Fracturing for Well Stimulation”. MS
esis, Texas Tech University, Lubbock, TX (August 2010)
Lee, W. M. and R. D. Ford. (1988). “Pressure Measurements correlated with Electrical Explosion of Metals in Water”. Journal of Applied Physics
64(8): 3851-3854.
Lee, W. M. and R. D. Ford. (1989). “Eective coupling of Pulsed-Power to Chemically Reactive Metallic Systems”. Pulsed Power Conference, 1989.
7th IEEE: 981 – 984
Lee, W. M. (1993). “Aluminum Powder/water reaction ignited by Electrical Pulsed- Power”. NAVAL SURFACE WAREFARE CENTER, Final
report, October 1990-April 1992: Medium: X; Size: Pages: (33 p).
Mao, R., De Pater, H. J., Leon, J. F., Fram, J. H., Ewy, R. T., Storslett, S., & Stefani, J. (2012). “Experiments on Pulse Power Fracturing”. Presented
at the SPE Western Regional, Bakerseld, California, 19-23 March 2012. SPE 153805.
Maurel, O., Reess, T., Matallah, T. (2010). “Electrohydraulic Shock Wave Generation as a Mean to Increase Intrinsic Permeability of Mortar”.
Cement and Concrete Research 40(12):1631-1638
Pijaudier-Cabot, G., Borderie, C.L., Reess, T., Chen, W. (2016). “Electrohydraulic Fracturing of Rocks”, 1st Edn. John Wiley and Sons Inc, New
Jersey, USA, ISBN: 978-1-84821-710-2.
Reess, T., De Ferron, A., Maurel, O. (2009). “Electrohydraulic Shock Wave Generation as A Mean to Increase Intrinsic Permeability of Concrete”.
IEEE Pulsed Power Conference pp.1079-1084, June 29-July 2, 2009.
Zhou, H.B., Zhang, Y.M., Li, H.L. (2015). “Generation of Electrohydraulic Shock Waves by Plasma-Ignited Energetic Materials: I. Fundamental
Mechanisms and Processes”. IEEE TRANSACTIONS ON PLASMA SCIENCE 43 (12): 3999-4008.
74 Hydraulic Fracturing Journal | July 2017Volume 4 - Number 3
Biographies
Yue Xiao studied as an undergraduate
student majoring in petroleum engineering
in Yangtze University in Jinzhou, China
from 2007-2011. Aer graduation, Yue went
directly to New Mexico Institute of Mining
and Technology and enrolled in the Petroleum
Engineering master program. During her stay
in New Mexico Tech, she was working as a
research assistant in the simulation group led by Dr. Robert Balch
for the Petroleum Recovery Research Center (PRRC). In 2013, she
got admitted by Petroleum Engineering department of Texas Tech
University and began to study as a Ph.D. student. Finishing most of
the course work, she joined the Pulsed Power Plasma Stimulation
(PPPS) research group in 2015 spring and started her doctoral
research under the advisement of Dr. Mohamed Soliman.
Dr. Waylon House obtained his degrees in Physics at M.I.T and the
University of Pittsburgh. In the early 70’s, as a post-doc and research
associate with recent Nobel Laureate Paul Lauterbur, he became one
of the pioneers of MRI. A co-founder of the Society of Magnetic
Resonance Imaging, he played signicant roles in the scientic
and commercial development of MRI and has been a principal or
consultant with a number of high technology businesses in several
industries. As an adjunct professor in Rice’s Dept. of Chem. Eng. for
30 years and as a faculty member of TTU’s Dept. of Petr. Eng. for
12, he has spent nearly 40 years investigating the fundamentals of
uid behavior, phase transitions and transport in liquids and porous
media. He has authored or co-authored more than 60 articles.
Ebru Unal is currently a PhD candidate
in Petroleum Engineering Department at
University of Houston. Prior to joining the
University of Houston, she was conducting
her studies under the supervision of Dr.
Mohamed Soliman at Texas Tech University.
Her research interests include well testing,
innovative completion techniques, and
analysis. She has been working as a Research and Teaching Assistant
in various drilling and completions courses, including Mud
Rheology Lab, Well Control Simulation Lab, Core Lab etc. She has
various industry internship experience in drilling and completions
on both conventional and unconventional reservoirs. She holds an
M.Sc degree in Petroleum Engineering from New Mexico Tech,
Socorro, NM (2011), and a B.Sc degree in Geological Engineering
from Middle East Technical University, Turkey (2009).
M.Y. Soliman is the Chairman of the
Petroleum Engineering Department at the
University of Houston and the holder of
the Miller endowed chair. Prior to joining
the University of Houston, Soliman was the
George P. Livermore Professor of Petroleum
Engineering at Texas Tech University for
more than 5 years. Professor Soliman had also
worked for Halliburton Energy for more than 30 years in several
technical and managerial positions. He is a distinguished member of
SPE and a licensed professional engineer by the State of Texas. He
is also a fellow of the National Academy of Inventors (NAI). He has
authored and co-authored more 200 technical papers and holds 30
US patents. He is also the editor of “Fracturing Horizontal Wells”
published by McGraw Hill in July 2016. M. Soliman received a BS
in Petroleum Engineering with top honors from Cairo University
in 1971. He earned MS and Ph.D. degrees from Stanford University
in 1975 and 1979 respectively. His areas of interest include well test
analysis, diagnostic testing, fracturing and numerical simulation.
