Numerical Modeling of the Propellant Assisted Extreme Overbalanced Perforation

Process
Sina Mohajeri, Ali Alizadeh, Mostafa Zeinali, Hamed Rajabi
Engineering Support & Technology Development Company (ESTD Co.)

Abstract
A new method is developed to numerically model the propellant assisted extreme overbalanced
perforating process and evaluate its effect on the well skin factor and productivity/injectivity. The
model consists of cylindrical propellant combustion, fluid drainage through perforations, well fluid
compression and movement, fractures propagation and flow through them. Using a suitable algorithm,
the pressure profile is calculated in time. We also derive the formulae to estimate the productivity
increase due to multiple fractures. The models have been verified by comparison with field
measurements. Also, we discuss on method of estimating the productivity increase made by the
propellant assisted perforating and the main factors of influence.
The model is capable to design propellant assisted perforation. Also it allows us to make a parametric
sensitivity study on propellant charge, well completion, perforation and formation properties, and so
on.

Introduction
Propellant Assisted Extreme Overbalanced Perforation process (PAEOP) is a well completion
technique that uses quick burning propellant materials which are capable to release a huge volume of
gas in the well, causing fractures in the near wellbore zone through perforation tunnels which
improves the well productivity/injectivity. In this technique the pressure pulse which is generated by
the released propellant gas can be controlled efficiently. A series of experiments was conducted by
Cuderman (1986) in the 1980s, in Sandia National Labs in the USA for understanding the propellant
assisted perforating mechanism. In the same decade, two researchers studied simulation of the
propellant assisted perforating and fracturing (Nilson, 1984; Schaltz, 1989). For quantitatively
designing and applying PAEOP in filed jobs, a precise model is required which includes cylindrical
propellant combustion, fluid drainage through perforations, well fluid compression and movement,
fracture propagation and flow through them. Using a suitable algorithm, the pressure profile (well
transient pressure) is calculated over time. We establish a numerical model for predicting pressure
pulses and transient pressure for designing propellant assisted perforation. The developed method and
formulae help estimate the productivity increase achieved by multiple radial fractures and formulate
the principles of designing pressure pulses. The model also enables us to investigate the sensitivity of
design parameters to pressure pulses and the potential productivity increase.
The model is implemented as a module in an advanced Computer program. This module complements
the conventional perforation design program for designing optimal propellant assisted extreme
overbalanced perforating systems. The downhole pressure gauges with extremely fast sampling rate
(15000 samples per second) are used to validate the peak pressure and real skin of the developed
model.

Method
We use the geometric combustion law for flame propagation on the solid propellant. Based on this
law, the burning film gradually develops from the entire exposed surface into the propellant solid in
which the newly created surface at each time step is parallel to the original one. The propellant
burning rate (propagating velocity of the burning surface normal to it) is expressed as:
u  de dt (Eq.1)
Where, e is the thickness of burnt layer. The affecting factors on the propellant burning rate are
propellant composition, propellant shape, ambient pressure and temperature, and so on, which are
obtained in lab (Nilson, 1981). For propellants which are used in the perforating process the following
simplified expression is used (Nilson, 1984):
u  u 0 Pwn (Eq.2)
Where, u0 is burning rate coefficient and n is the pressure index. The value of the pressure index is in
range of 0.2-1 which is obtained from lab experiments for each propellant. For must propellants the

80th EAGE Conference & Exhibition 2018

11-14 June 2018, Copenhagen, Denmark
pressure index is 1 for pressures above 20Mpa (3000 psi). The shape of propellant directly affects the
propellant burning process. The most common shape for propellant in perforation job is a hollow
cylinder column. The shape function can be determined assuming the law of geometric combustion.
In cases where canisters are present for the propellants, the flame is quickly propagated from the
ignition position to all the exposed surfaces; the shape function is expressed as:

 p   p    p 
 D  2e 2  d  2e 2  l  2e
f 1  (Eq.3)
 
D p2  d p2 l p
If there are no canisters present for the propellant the shape function will be:
 
2
D p2  d p  2e
f 1 (Eq.4)
D p2  d p2
Where the Dp and dp are propellant outer and inner diameters, respectively. lp is the propellant length
and e is the thickness of burnt layer.
There must be a fluid compression sub-model to determine the space freed by fluid compression and
the upward movement. The fluid volume reduction due to compression opens a volume for the gas
which will be calculated by the (Eq.5):
V l  C l Pw Aw  al t (Eq.5)
Where, Cl is fluid compressibility, Aw is the well cross sectional area, Pw is the difference between
released gas pressure and initial well hydrostatic pressure, val is the sound velocity in the well fluid,
and t is time.
After breaking down the formation, released gas starts to discharge into the formation which creates
and propagates fractures. The pressure of the transient wellbore during the propellant assisted
perforation is driving force for the fractures and also the outcome of the discharged gas into
formation. The accumulated gas and liquid mixture discharged mass to fractures, is calculated by the
(Eq.6):
mdisch arge  C1Ah  g  C 2 l  Ph  g (Eq.6)
Where, Ah is the total area of perforation,  is density, Ph is pressure drop along perforation, C1 and
C2 are experimental coefficients, and subscripts g and l refer to gas and liquid, respectively. The gas
density will be calculated using Noble-Abel equation of state (Kulkami et. al.2000; Oliveira et. al.
2005) due to extremely high pressure and temperature condition while propellant combustion. The
value of Ph may be calculated by assuming a fully turbulent flow in the perforation tunnels.
The temperature variation in the well is calculated by thermodynamic relations and Nobel-Abel
equation of state (Oliveira et. al. 2005).
We modify the ballistic model of the closed volumes to reach a prediction for transient well pressure
generated by propellant released gas (Eq.7):
FM gw
Pw  Pw ,0  (Eq.7)
V gw   M gw
Where Pw,0 is the well initial pressure, Mgw is the mass of gas inside the well generated by propellant
combustion, Vgw is the occupied volume of the propellant gas,  is the propellant co-volume and F is
the propellant force constant. The Mgw can be calculate by the (Eq.8)
M gw  fM 0  M gd (Eq.8)
Where, f is the shape function, M0 is the propellant initial mass, Mgd is the propellant mass drained to
the fracture. For solving the ballistic equation, the pressure derivate equation must be solved
numerically:

80th EAGE Conference & Exhibition 2018

11-14 June 2018, Copenhagen, Denmark
 dM gw dV gw
V gw  M gw
dPw dt dt
F
V 
2
dt
gw   M gw
dM gw df (Eq.9)
M0  m gd
dt dt
dV gw M 0 df m dV
  ld  l
dt s dt l dt
Where, mgd and mld are the gas and liquid mass discharge rates.
For the prediction of fracture geometry the bi-wing model of the PKN is used due to varying fracture
width around the perforation tunnels (Economides 2007).
For estimating the productivity/injectivity of the well, the skin factor must be calculated. There are
two mechanisms of the shear failure and tensile failure in the propellant perforation technique. The
equivalent skin factor of the fracturing (always negative value) due to propellant gas fracturing, can
be estimate by (Cinco Ley et. al., 1981)
In the (Eq.10) rd is the radius of damaged zone, and k and kd are permeability of formation and
damaged zone, respectively. FCD is fracture conductivity. Note that this calculated skin must be added
to the other skin components in order to obtain the total skin. The other components of skin are
calculated using the known methods (Bolchover et.al. 2006)

 rd 
 r  N f FCD 
 r k r  1  N f FCD k rf
   ln e  ln d   ln  ln
e
 ln 
 rd k d rw   rf rd
 N f FCD
k d rw
 N f FCD 
 
 rf rf  (Eq.10)
 k  r 
  1 ln  d 
s f   1  1   ln e   d   rw 
r k
rw 
Validating Model with a Real Field Case
As a real case, the Figure 1a compares the peak pressure for a deviated well in a tight carbonate
reservoir and the predicted pressure value by the model (blue curve). Figure 1b displays the three
dimensional view of the perforation tunnels for the same job. The minimum total skin values for the
conventional perforation, and PAEOP operation for the same well is 0.2 and -3.58, respectively; the
input data of this case is displayed in table 1. Figure 2 displays the skin versus propellant length.

Table 1 The input data for validating the model

Bore hole diameter 8.375 in Porosity 0.5 %
Hydrostatic Pressure 10856 psi Formation top depth 14291 ft
Deviation angle 70 Deg Formation thickness 1083 ft
Casing burst pressure 9410 psi Formation density 2.75 g/cc
Casing grade T95 Rock compressive Strength 10000 psi
Casing Size 7 in Effective stress 7500 psi
Gun type 22.5T Zhoobin Rock compressibility 3.2e-6 psi-1
Gun phasing & size 60, 3-3/8, 6spf Reservoir pressure 9000 psi
Average permeability 0.2 md Oil Gravity 35 oAPI
Young modulus 5.55e6 psi Initial temperature 275 F
Poisson ratio 0.32 Formation Pressure gradient 0.69 psi/ft

80th EAGE Conference & Exhibition 2018

11-14 June 2018, Copenhagen, Denmark
Figure 1 a) The well transient pressure response of the model and real fast gauge record, b) the three
dimensional result view of the model

According to Figure 2, increasing either the outer diameter or length of the propellant will reduce the
skin factor due to high peak pressure and long created fractures. But note that this high peak pressure
sometimes causes bursting in casing or crushing the formation. Pressure must not be allowed reach
the threshold value for bursting, so stability estimation on formation and casing are made in our
model. For this case the optimum length and diameter for the used propellant to preventing casing and
formation failure is 300 ft and 4.075 in, respectively.

Figure 2 The total skin versus propellant length

Conclusions
A novel method developed for modeling the PAEOP process and analyzing the transient pressure
during propellant assisted perforating. A mathematical model has been established, which includes the
propellant combustion, well fluid compression and movement, fluid discharge to perforations, and
fracture propagation. The model results were validated with a real case and it was concluded that the
propellant used in this technique has always a specific optimum length and diameter to prevent casing
and formation failure. In this optimum configuration the skin and well productivity/injectivity are
minimum and maximum achievable values, respectively.

Acknowledgement
We are thankful to Petro Danial Kish Co. for funding the project and providing expertise that greatly
assisted the study.

References
Cinco-Ley, H., and Samaniego, E [1981] Transient Pressure Analysis for Fractured Wells. JPT, 1749-
1766.

80th EAGE Conference & Exhibition 2018

11-14 June 2018, Copenhagen, Denmark
Cuderman J F. [1986a] Effects of Well-bore Liquids in Propellant-based Fracturing. 27th Society
Mining Engineering Rock Mechanics US Symposium, Tuscaloosa, ALA, 23-25/6.

Economides M. J. , Martin T. [2007], Modern Fracturing, ET publishing

Kulkami U .P, Smita D. [2000] Modelling of Heat Loss in Closed Vessels during Propellant Burning.
Defence Science Journal, V.oI50, No 4, pp. 401-409.

Nilson R H. [1984] Engineering Formulas for Fractures Emanating from Cylindril and Soherical
Holes. J. Applied Mechanics, Vol. 51, pp929-933.

Nilson R H. [1981] Gas-Driven Fracture Propagation. J. Applied Mechanics, Vol. 48, pp757-762.

Oliveira, A. A. M. F. Filho, G. M. Platt, and F. C. Peixoto [2005] Estimation of Ballistic Parameters

of Gun Propellants through Closed Vessel Experiment Modeling. Engenharia Térmica (Thermal
Engineering), Vol. 4 · No. 1 · p. 50-55

Schaltz J F. [1989] Laboratory, Computer Modelling and Field Studies of the Pulse Fracturing
Process. SPE 18866.

80th EAGE Conference & Exhibition 2018

11-14 June 2018, Copenhagen, Denmark