Underbalance or Extreme Overbalance

L.A. Behrmann,* SPE, and Bryan McDonald, Schlumberger Wireline and Testing

Summary Underbalance Perforating

Application of an extreme overbalance, sufficient to fracture the Perforation Damage. The primary purpose of underbalance per-
reservoir formation, applied either during or after perforating was forating is to minimize perforation damage. However, there are
first presented as an alternate completion technique during the two different types of perforation damage. The primary one is a
1993 Fall SPE Conference.1 During the ensuing two years, a num- damage region that surrounds the perforation tunnel having re-
ber of jobs have been run by different operators in a variety of duced permeability. This damaged region has historically been
formations with varying applications. The primary question asked called the crushed zone. This damage is a result of fractured sand
by an operator relates to the benefit of extreme overbalance 共EOB兲 grains reducing the size of the pore throats. The other, and nor-
versus current perforation techniques of underbalance 共UB兲 or mally ignored, perforation damage is caused by residual commi-
modest overbalance for well control. Of course this question nuted sand not removed from the perforation tunnel. This mate-
spawns a whole set of new questions about when and how to use rial, except at the very end of the perforation tunnel, typically has
and the benefits and risks of the EOB. EOB, as used in this paper, high permeability and porosities and does not impede production.
includes extreme overbalance perforating and rapid overpressured
However, it can restrict injectivity and require higher injection
perforation extension. This paper compares and contrasts UB and
pressures. Removal of this residual comminuted sand is a function
EOB strengths and weaknesses and provides recommendations of
of underbalance, charge type and rock strength. Low underbal-
when and why to use the different techniques. Field examples are
ances, big hole charges, weak rock and downside oriented perfo-
provided of successful and unsuccessful application of EOB
rations enhance this residual comminuted sand damage.
completions.
Removal of Crushed Zone Damage. Determination of sufficient
underbalance to remove the perforation crushed zone permeability
damage is still subject to a number of uncertainties. King, Ander-
son, and Bingham2 published field data of underbalance versus
permeability with acid improvement/no improvement as a criteria
Introduction for sufficient underbalance without the need for acidization. Using
Perforating, fracturing and acidizing are simply a means of pro- laboratory data, Behrmann3 published equations for optimum un-
viding and enhancing flow communication between the reservoir derbalance versus permeability, porosity and perforation hole size
and the wellbore. The objective of these different completion op- to obtain nondamaged perforations. These underbalances are typi-
erations is to cost effectively maximize the producibility of the cally three to five times greater than the King data which suggests
well. Perforating is used to provide the basic fluid communication that even acidization of sandstones does not remove all the perfo-
between the reservoir and the wellbore. Underbalance perforating ration damage. Curve fit of permeability-porosity data from
and acidizing are typically used to bypass or eliminate near well- Alaska and North Sea Brent reservoirs,4 Fig. 1, applied to the
bore damage. Fracturing is generally used to expose a larger res- equations in Ref. 3, provides a simple underbalance-permeability
ervoir flow area to the wellbore. Use of propellant tools and now relationship, Fig. 2 and Eq. 共1兲,
extreme overbalance 共EOB兲 produce short fractures and thus p⫽d 0.3关 90.4⫺7.86* ln共 k 兲兴 2 , 共1兲
complements underbalance 共UB兲 perforating and acidizing to by-
pass wellbore damage. However, the process of perforating also where d⫽perforation tunnel diameter in the rock, inches; k
creates a permeability damaged region around the perforation ⫽reservoir permeability, md; and p⫽optimum underbalance, psi.
commonly called the ‘‘crushed zone.’’ Underbalance perforating The perforation tunnel diameter is not recorded on any data
has been used to remove the perforating damage zone in reser- sheets and the casing entrance hole 共EH兲 diameter has been used
voirs with sufficient pressure and permeability. in the past as a default value. However, the hole diameter in the
Since perforating is the primary means to provide the initial rock is a function of the charge size and the unconfined compres-
fluid communication with the reservoir and where applicable is sive rock strength 共UCS兲. An empirical relationship has been de-
performed underbalance, the relevant issue is when-where-how to veloped for average tunnel diameter as a function of casing hole
use EOB as an addition to perforating. size and rock unconfined compressive strength,3 Eq. 共2兲
We will first review where UB perforating is applicable, then
discuss the interaction between the perforation and the fractures n 关 3.27⫺0.61 ln共 S uc 兲兴
d⫽e * 共2兲
from an EOB process and the implications/questions with regard for deep penetrating charges, where e h is the casing EH in N80
to the fracture flow path. Several field examples will then be dis- casing, and S uc is the UCS in kpsi.
cussed. Finally, based on the physics of EOB, recommendations Applying the same analysis to the quartz arenites, Fig. 1, gives
of applicability will be made. One word of caution: the following the following underbalance-permeability relationship, Fig. 3 and
discussion on underbalance perforating does not apply to uncon- Eq. 共3兲:
solidated or very weak rock because of the potential failure of the
rock from a high underbalance. p⫽d 0.3关 50.6⫹3.44* ln共 k 兲兴 2 . 共3兲
Eqs. 共1兲 and 共3兲 should be used with caution since they are ex-
*SPE Member trapolations of laboratory data on outcrop rocks to reservoir rock
Copyright © 1999 Society of Petroleum Engineers using permeability-porosity relations with large data scatter. How-
This paper (SPE 57390) was revised for publication from paper SPE 31083, first presented ever, limited laboratory data on low permeability reservoir quartz
at the 1996 SPE International Symposium on Formation Damage Control held in Lafayette,
Louisiana, 14–15 February. Original manuscript received for review 15 May 1996. Revised
arenites5 and a clean low permeability outcrop rock at 3,000 psi
manuscript received 8 April 1999. Paper peer approved 14 April 1999. underbalance gave zero to negative perforation damage skins,

SPE Prod. & Facilities 14 共3兲, August 1999 1064-668X/99/14共3兲/187/10/$3.50⫹0.15 187

 Fig. 1–Permeability-porosity relationships.

Table 1, which suggests that clean 共clay free兲 sands can use lower
underbalances as proposed in the above analysis.

Removal of Residual Comminuted Sand. During underbalance

operations, varying amounts of the communited sand in the per- Fig. 3–Optimum underbalance vs. permeability „quartz
foration tunnel are removed. Although an extensive study of this arenites….
‘‘injectivity’’ damage has not been made, limited laboratory ex-
periments have shown that it is more difficult to remove this re-
sidual sand under the following conditions: perforation cavity can substantially reduce injectivity without af-
䊉 Low underbalance, fecting producibility. A specific laboratory experiment is dis-
䊉 use of big hole charges versus deep penetrator charges, cussed in Appendix A.
䊉 weak rock, When To Perforate Underbalanced. To minimize the crushed
䊉 downside perforations, and zone perforation damage in naturally completed producers and
䊉 single phase oil flow versus two phase oil/brine flow. injectors, perforating should be underbalanced when there is suf-
ficient reservoir pressure and no major operational constraints.
Except for very hard rock, the tunnel diameter in the rock is When less than the optimum underbalance is used, the equations
larger than the casing entrance hole. As a result, the hole through given in Ref. 3 can be used to estimate the residual perforation
the casing/cement provides a flow restriction during the underbal- damage and decide if a stimulation operation is required/justified.
ance surge and may prevent all the comminuted sand being swept It is also suggested that for all injectors and producers that will
from the perforation tunnel. Laboratory experiments have shown be hydraulic or dynamic fracture stimulated, that underbalance
that the presence of comminuted sand filling at least 50% of the perforating be performed prior to a fracture operation if excessive
breakdown/injection pressures are deemed to be a potential prob-
lem. This recommendation assumes there is sufficient reservoir
pressure and permeability to elute most of the comminuted sand
from the perforation tunnel. The recommended underbalance
given in Eqs. 共1兲 and 共3兲 is generally sufficient to remove most of
the comminuted sand debris except for weak rocks.

Extreme Overbalanced Perforating

The Concept. Work by Oryx Energy and Arco has resulted in a
new completion technique called extreme overbalance 共EOB兲. Ex-
treme overbalance has been defined as either the application of a
very high overbalance pressure during the perforating process
共EOP兲 or very high pressure ‘‘surging’’ of existing perforations.
The overbalance, which is significantly above the formation
breakdown pressure, creates short fractures in the formation and
improves productivity in some wells. EOB has been used success-
fully in a variety of well environments by a number of operators.
More than 800 EOB jobs have been reported, most of these in the
U.S. and Canada. While there have been notable successes, EOB
is not a global replacement for underbalanced perforating. Rather,
it is a complementary process for specific applications.
The EOB approach, developed independently by Oryx and
Arco, uses the wellbore tubulars to contain the energy source. The
basic technique involves pressuring a large portion of the tubing
Fig. 2–Optimum underbalance vs. permeability „Alaska and with gas over a modest amount of fluid. The column of pressur-
North Sea Brent…. ized gas and the liquid beneath it create a high pressure opposite

188 L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999
 TABLE 1– PERFORATION DAMAGE IN CLEAN SANDSTONES
„3,000 psi UNDERBALANCE, 5,500 psi EFFECTIVE STRESS…

Average Unconfined
Permeability Charge Hole Single Compressive
to Kerosene Porosity Weight Penetration Diameter Shot Strength
Test (md) (%) (g) (in.) (in.) CFE Skin (kpsi)

BPC-1 0.96 8.6 23 6.0 0.62 1.67 ⫺0.75 25.1

BPC-3 17.75 8.28 14 5.13 0.62 0.84 0.34 13.58
BPC-5 7.66 12.95 3.5 1.9 0.29 5.82 ⫺2.1 16.06
BPC-5 ⬘ 9.17 12.95 6.5 3.0 0.36 6.74 ⫺2.0 16.06
BPC-8 4.72 7.9 6.5 2.6 0.25 0.016 170 20.88
CT-1 10.4 14.78 3.5 3.0 0.28 1.08 ⫺0.23 8.8
CT-2 10.4 14.78 3.5 3.15 0.31 1.47 ⫺1.0 8.8

the perforated zone. Field experience, supported by laboratory test fractures. Bi-wing or multiple fractures will work as long as the
data, indicates that the overbalance pressure in the perforation fracture conductivity is not compromised by competing multiple
zone should be in the range of 1.4–2.0 psi/ft of well depth. fractures. EOB as a matrix acid displacement technique relies on
For wells without existing perforations, the high pressure is the acid providing fracture conductivity, therefore multiple frac-
applied instantaneously to the formation upon firing the gun, and tures are preferred. When used as a prehydraulic fracture treat-
the wellbore fluid is driven into the perforations. The expanding ment, initiation of only a bi-wing fracture is desired to minimize
gas maintains the wellbore pressure in the perforated zone to en- near wellbore competition from multiple fractures. For a naturally
sure sufficient loading time to create short fractures 共in the order fractured reservoir, one wants conductive multiple fractures inter-
of 10 ft兲 into the formation. The high downhole flow rates during secting the natural fractures. Proper application of EOB to all of
EOB are not typically attainable in conventional hydraulic frac- the above treatments requires an understanding of the fracture
turing operations. The EOB event lasts tens of seconds. mechanics of EOB.
The basic technique lends itself to a number of interesting
variations. Perforating guns may be either tubing or wireline con- EOB Fracture Mechanics. The fracture mechanisms involved in
veyed. Liquids in the well can range from clear brines to acid, to EOB are similar to conventional hydraulic fracturing in some re-
fracturing gels, to gels with suspended proppant, to resin for sand spects and yet significantly different in others. Among the factors
control. Alternatively, the wellbore can be filled entirely with gas. influencing fracture development in EOB are in-situ stress, well-
Previously perforated wells may be treated using either a shear head pressure, volume of gas, formation permeability, shot den-
disk or a shearable ball seat or similar device placed at the bottom sity, casing entrance hole diameter, shot phasing, orientation of
of the tubing. The shear disk may be run in on wireline and set in perforations with respect to the preferred fracture plane 共PFP兲 and
a profile. The disk is designed to rupture at a predetermined pres-
the number of fractures created.
sure. With the disk in place the nitrogen pressure is increased until
Hydraulic fracturing is generally characterized by a single large
the disk ruptures and the gas pressure drives the fluid into the
bi-wing fracture along the formation PFP. However, prior to the
perforations and the formation. The procedure is similar when
extension of this bi-wing fracture, there can be many multiple
using a shearable ball seat. When the packer is set, a ball is
fracture sites depending on the perforation phasing,6 Fig. 4. These
dropped from the surface to complete the seal. Nitrogen pressure
multiple fracture initiation sites may contribute to near wellbore
is increased until shear pins fail, the ball seal is released and the
tortuosity. On the other hand, multiple fractures from EOB typi-
well is surged. The pumpout shear disk is also run on tubing and
functions on the same principle. The viability of surging existing cally occur only where perforations are located and on the PFP if
perforations is demonstrated by the preference of some operators a perforation is not near. EOB also allows a larger angle between
for separate perforating and surging operations even in the case of the PFP and a perforation before the fracture ignores the perfora-
initial completions. tion and initiates a fracture at the wellbore along the PFP, Fig. 5.
The following discussion of EOB fracture mechanics is con-
One Perspective of EOB. EOB may be viewed as a means to fined to vertical or near vertical wells and uses EOB laboratory
obtain short fractures without the need for large amounts of sur- data, one example of which is given in Appendix A.
face equipment. With multiphased guns, it can also be viewed as Multiple Fracture. The dynamics of EOB enhance the creation
a dynamic diversion fluid placement technique. Assuming the of multiple fractures from the phased perforations while tending
EOB concept works, then it might have application for the fol- to limit fracture length and width. The multiple fracture geometry
lowing situations: may be detrimental to overall fracture treatment if one wants a
䊉 Fracture past extensive wellbore damage from either new or single bi-wing fracture and care must be exercised in the perfora-
producing wells, tion phasing and orientation to obtain the most favorable fracture
䊉 fracture past the perforation damage when there is insufficient
arrangement.
underbalance, There will always be a bi-wing fracture turning into the PFP. In
䊉 provide a matrix acid diversion/placement in place of balls or
the case of multiple fractures, it is not clear how far the fractures
other diversion techniques, not in the PFP will propagate nor what the fracture conductivity
䊉 prehydraulic fracture treatment to break down the formation,
will be. The extension of fractures originally misaligned with the
and PFP is dependent on reservoir stresses, stored energy and pressure
䊉 enhance communication with natural fractures.
of the gas in the tubing and competition from the fracture in the
The first two applications require sufficient fracture conductiv- PFP. The misaligned fractures are shorter and do not intersect the
ity to be successful since the primary flow path is now through the fracture in the PFP.

L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999 189
 Fracture Initiation Pressure. Field and laboratory data show
that the EOB fracture initiation pressure is always higher than the
fracture initiation pressure observed in conventional hydraulic
fracturing. EOB also traps residual crushed sand particles in the
perforation tunnels 共underbalanced perforating tends to flush the
particles into the wellbore兲. The particles may act as a filter cake
during injection resulting in a further increase in the dynamic
共EOB兲 fracture initiation pressure.

EOB Field Examples. The following examples refer to Table 2.

Each EOB job in the table was chosen to show a wide assortment
of well and reservoir conditions and give a broad perspective of
where EOB has been attempted and the results of these attempts.
In most cases EOB was attempted because conventional comple-
tion techniques failed to achieve desired results, or EOB has been
successful in a particular surrounding area. The examples and
reasons are expressed below.
The Conoco well is considered a success as the production rate
achieved from EOB was 50% greater than the production rate of
wells in the immediate area. The formation is a very stringy car-
bonate in which the permeability range is large. The procedure
calls for 500 gal of 15% acid to be placed in the bullplugged
tubing before the guns are fired.
This procedure takes advantage of the explosively initiated pro-
Fig. 4–EOB multiple initiation for 120° shot phasing for two duction valve 共SXPV兲 which due to its design opens as the deto-
different gun alignments with PFP. Arrows indicate the direc- nation process to the perforating guns is activated. Thus pressure
tion of the PFP. is kept off the casing until the firinghead functions. The operation
of the SXPV also ensures that the valve is fully open as the com-
pressed energy in the tubing is released. Therefore, the frictional
Fracture Length and Width. The integrated length of all frac- losses, as the acid is being driven through the tubing and out the
tures is essentially constant and is not dependent on the number of SXPV, are held at a minimum. This in turn prolongs the EOB
fractures but rather on the stored energy available before the pres- event allowing more energy into the formation which can possibly
sure drops below some minimum value. Fracture width is a func- increase the fracture length. This acid is injected into the reservoir
tion of the individual fracture length and the injection pressure ahead of the N2. As the table shows, this is a high permeability
differential above the local formation stress at the perforation. carbonate reservoir which yielded very good results.
Fracture conductivity is generally considered to be a function of The Aramco well was a poor performing injector in a friable
the fracture width, placed proppant, acid or gas erosion of the sandstone. This injector’s perforation tunnels were assumed to
fracture faces. have sanded up in a short amount of time as the original injection
rates dropped substantially in a few weeks, which is a good ex-
ample of what comminuted sand does to injectors if not properly
expelled. Due to poor sand consolidation of the reservoir the
needed underbalance pressure to properly expel the sand from the
perforation tunnels was not possible without sanding up the string.
It was suggested to perforate overbalanced, to stop sand influx
into the wellbore, and pump a prepack immediately after perforat-
ing. It was believed that the EOB would create propped fractures
that would not be as susceptible to sanding due to the larger sur-
face area open to flow. The results of this injector are very good as
seen from the table.
The Operator C well is unique due to its high permeability
共2,000 md兲. The EOB job was run with a DST string to be used
for fluid loss control and was run in this reservoir as an attempt to
improve cleaning of perforation tunnels prior to gravel packing. It
was also hoped that the EOB would blow the sand from the per-
foration tunnels into the formation prior to the trip in and out of
the well for gravel pack. This technique had worked well in an
offset field with much lower permeability. The operator was
warned that the results may not be as successful for this high
permeability application, but EOB was decided on to evaluate
technique for this field.
The skin is usually quite high in this reservoir due to the partial
penetration from the drilling process, and it is difficult to deter-
Fig. 5–Hydraulic multiple initiation for 120° shot phasing for mine what the skin was before the EOB job. An injectivity test
two different gun alignments with PFP. Arrows indicate the di- afterward was poor at 0.4 bbl/min with 700 psi pump pressure. It
rection of the PFP. was then decided to flow the well with a high drawdown. The

190 L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999
 TABLE 2– EXTREME OVERBALANCE JOB DATA

Completion Producer/ Reservoir Depth EOB Acid Permeability

Client Country Location Well Name Type Injector Type (ft) Type (HCl%) (md)

Conoco USA North Dakota Well A New Well Producer Carbonate 9,666 TCP 15 10 to 1,000
Aramco Saudi Arabia Saudi Arabia Hawtah No. 42 Recompletion Injector Sandstone 6,500 TCP ¯ 250
Operator ¯ ¯ Well C New Well Producer Sandstone 9,500 TCP ¯ 2,000
Petrobras Brazil Brazil Well D Recompletion Producer Sandstone 2,100 Surge ¯ 20
Petrobras Brazil Brazil Well E Recompletion Injector Sandstone 2,150 Surge 5 20
Operator P USA Colorado Well F New Well Producer Carbonate 1,300 TCP ¯
**Arco USA Alaska Well G Recompletion Producer Sandstone 8,500⫾ Surge ¯ 20
Marathon USA Texas Indian Basin No. 9 New Well Producer Carbonate 7,837 TCP 20 39

*Offset wells in area typically produce at 2,000 BOPD.

**Pretreatment done prior to hydraulic fracturing.
†
In the Petrobras example the existing perforations were deep penetrating charges .39 entry hole with a shot density of 12 SPF at 120° phasing.
‡
In the Arco example the existing perforations were deep penetrating charges .45 entry hole with a shot density of 4 SPF at 180° phasing.

injectivity after the drawdown yielded 5 bpm at 800 psi pump duction. These are just two examples of several wells completed
pressure. This was considered good. The well was then gravel with EOB in this area. Results of EOB jobs in this area are con-
packed and a production test performed. The production test sidered very good by Petrobras.
yielded a low productivity index 共PI兲 with a damage ratio of 7.9. The operator F well is a very shallow EOB example where the
The well was stimulated with a mud acid and the PI was improved psi/ft gradient was quite high at 2.1. This well was drilled through
by 100%. The damage ratio also was reduced to 3.7. The fact that
a one of a kind formation that showed up on the logs while evalu-
the well had to be backflowed after the EOB operation convinced
the operator that EOB is not applicable in this field. ating a deeper reservoir. The operator in this case has had very
The examples for the Petrobras wells incorporate the surge good results in microdarcy permeabilities with EOB in an adja-
technique in which the perforations are already existing across the cent area and, although the true permeability here was not known,
desired completion interval. The reason EOB was used in this area EOB was tried. The reservoir did not respond as well as expected
was due to past problems with fracturing and acidizing. It is be- and flowed only 3 BOPD. Acid was then applied as a way to
lieved the EOB technique allows the reduction of skin without stimulate the reservoir with no increase in production. This is a
attacking primary cementation and the creation of microfractures good case showing where the well was evaluated using a low cost
which do not propagate far enough to allow communication be- technique instead of using a full Frac job for evaluation which
tween zones. would have been more costly for the operator.
A gel-sand cushion is run in the tubing and released through a
The Arco example7 uses their rapid overpressured perforation
shear disk at the time the desired nitrogen pressure is achieved.
extension 共ROPE兲 and high energy ROPE 共HEROPE兲 techniques
EOB was used on the injector well 共well E兲 because previous
completion techniques with acid had broken down the primary as a pretreatment to hydraulic fracturing. In Arco’s case there are
cementation allowing undesired communication between zones. several well and reservoir factors they are trying to overcome with
As Table 2 shows, the desired injection rate was increased five- EOB. These include deviation, large tubulars, low permeability,
fold. The producer, a pumped well, used acid in the tubing at time small intervals and, due to a water oil contact 共WOC兲, a desire to
of isolation valve release and showed a four-fold increase in pro- stay in the oil zone.

TABLE 2– EXTREME OVERBALANCE JOB DATA „continued…

Gun Shot
psi/ft Interval Gun Size Phasing Density/ Prejob Postjob
Client Country Location Well Name Gradien (ft) (in.) (degrees) Type Production Production

Conoco USA North Dakota Well A 1.3 16 3 1/2 60 4 SPF/D.P. * 3,000 BOPD
Aramco Saudi Arabia Saudi Arabia Hawtah No. 42 1.4 72 4 1/2 45/135 12 SPF/B.H. 2,700 BWPD 10,635 BWPD
Operator ¯ ¯ Well C 1.7 40 7 45/135 12 SPF/B.H. ¯ ?
Petrobras Brazil Brazil Well D 1.4 10 5 120 12 SPF/D.P. 11 BOPD 44 BOPD
Petrobras Brazil Brazil Well E 1.4 10 ¯ ¯ ¯ 180 BWPD 1,800 BWPD
Operator P USA Colorado Well F 2.1 90 3 3/8 60 6 SPF/D.P. ¯ 3 BOPD
**Arco USA Alaska Well G 1.1 20⫾ ¯ ¯ ¯ 0 BOPD 2,300 BOPD
Marathon USA Texas Indian Basin No. 9 1.2 70 3 1/2 60 4 SPF/D.P. ¯ 5,500 MCFD

*Offset wells in area typically produce at 2,000 BOPD.

**Pretreatment done prior to hydraulic fracturing.
†
In the Petrobras example the existing perforations were deep penetrating charges .39 entry hole with a shot density of 12 SPF at 120° phasing.
‡
In the Arco example the existing perforations were deep penetrating charges .45 entry hole with a shot density of 4 SPF at 180° phasing.

L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999 191
 The ROPE technique has been improved by Arco since its in- EOB in Carbonate Reservoirs. EOB has been successful in car-
ception to the HEROPE. This technique uses the large tubulars to bonate reservoirs when the wellbore fluid is acid. EOB acts as a
store large volumes of nitrogen and very little fluid if any in the dynamic diversion for the acid. HSD guns with many shot phas-
wellbore at the perforations. Due to the large volume of nitrogen ings are recommended.
and little fluid in the wellbore, it is believed that the gas flowing
through the perforation tunnels and into the formation erodes EOB in Naturally Fractured Reservoirs. Creation of multiple
away any tortuous flow paths, reducing the near wellbore pressure fractures that intersect the natural fractures provides another EOB
losses 共NWPL兲. These reductions of NWPL have ranged up to application providing there is sufficient fracture conductivity as
1,350 psi. already discussed. One issue is how far the fractures travel before
The amount of proppant placement seen by Arco since the turning into the PFP, generally the natural fracture orientation.
ROPE inception has gone from just over 50% in the early stages Experiments suggest the fracture starts turning after one wellbore
of ROPE to over 95% in 1994. The example well Table 2 well G diameter. If the EOB fracture is reoriented in the PFP after two to
designed for 34,000 pounds of proppant placed with the actual three wellbore diameters, then the natural fracture spacing should
pounds of prop placement was 39,000 lbm or 115%! Arco’s be less than two to three wellbore diameters. High shot density
evaluation of the ROPE technique shows that the success rate and guns with many shot phasings are recommended.
the efficiency of the hydraulic fracturing operation and the pro-
duction rate of the wells are far better than pre-ROPEd wells in
Conclusions
the area. The example given in the table is one of the 80 plus
wells that have been ROPEd for Arco. EOB is a completion process that may be used during or after
The Marathon example is from a tight carbonate reservoir. The perforating. It is not a replacement for perforating. For new
reason EOB was used as a completion technique here was due to completions 共no existing perforations兲, one can choose between
the success of another operator in the area and due to lower than perforating alone, either underbalance or modest overbalance, or
expected production rate of the wells in general in this reservoir. EOB during the perforating process. EOB while perforating re-
Therefore, a new completion technique was considered. The EOB quires a consideration of completion objectives, reservoir proper-
completion technique allowed the well to be surged leaving the ties, maximum allowable surface pressure, economics, etc., to de-
well full of nitrogen which could easily be bled off allowing the termine if an integrated approach is better than perforating
well to flow. This allowed the well to be put on production much followed by EOB.
sooner with less cost. Another benefit seen from EOB on this Proper application of EOB, either integrated or segregated from
reservoir was a far lower skin than other wells in the area. perforating, has been successfully applied in a variety of comple-
tions. Laboratory tests and field observations have provided a bet-
ter understanding of the EOB physics and provided recommenda-
Where to Use EOB tions on choices of perforating systems and application of the
process.
EOB Followed by Hydraulic Fracture Stimulation, Modest to
Low Permeability. If EOB is to be followed by conventional
Nomenclature
hydraulic fracture stimulation, the initiation and propagation of
multiple fractures should be discouraged. A long bi-wing fracture p ⫽ underbalance pressure, psi
is needed. Therefore, either a 180° phased gun oriented in the k ⫽ permeability, md
PFP, or a 120° phased gun with random orientation is suggested. d ⫽ tunnel diameter, in.
Shot density and required casing hole diameter will dictate the S uc ⫽ unconfined compressive rock strength, kpsi
charge attributes; generally big-hole charges will work well ex- e h ⫽ casing entrance hole in N80 casing, in.
cept where large gun-to-casing clearance exists. Centralization, or
partial centralization in the casing is desirable. Shot density will
be a function of the fracing hydraulic horsepower and will gener- Acknowledgments
ally be modest to low. Direct communication between the perfo- Thanks to Schlumberger Well Services for allowing publication.
rations and the primary fracture with minimum near wellbore Also, thanks for the operators listed in Table 2 for allowing pub-
pressure losses is desired. Acid may be used to enlarge the flow lication of field results.
path around the casing/cement annulus when the perforations can-
not be closely aligned with the PFP. Arco uses a high energy
References
ROPE technique where the amount of liquid is minimized and the 1. Handren, P.J., Jupp, T.B., and Dees, J.M.: ‘‘Overbalance Perforating
amount of pressurized gas is maximized. The assumption is that and Stimulation Methods for Wells,’’ paper SPE 26515 presented at
fracture extension is gas driven and gas erosion of near wellbore the 1993 SPE Annual Technical Conference and Exhibition Meeting,
torturous paths eliminates near wellbore pressure losses. Houston, 3–6 October.
2. King, G.E., Anderson, A., and Bingham, M.: ‘‘A Field Study of Un-
derbalance Pressures Necessary to Obtain Clean Perforation Using
EOB to Overcome Near Wellbore Damage, Modest to High Tubing-Conveyed Perforating,’’ JPT 共June 1986兲 662.
Permeability. Even though the produced fractures are relatively 3. Behrmann, L.A.: ‘‘Underbalance Criteria for Minimum Perforation
short. EOB treatment can be beneficial in reservoirs where well- Damage,’’ paper SPE 30081 presented at the 1995 European Forma-
bore damage depth is greater than about two thirds of the average tion Damage Conference held in the Hague, the Netherlands, 15–16
perforation depth. High shot density 共HSD兲 guns loaded at high May.
4. Evans, I.J., Bryant, S.L., and Cade, C.A.: ‘‘Modeling the Effect of
shot phasings 共say 45° or 60°兲 should provide good results taking
Diagentic Cements on Sandstone Permeability,’’ in Geofluids ’93,
advantage of multiple drainage paths into the wellbore. The key ‘Contributions to an International Conference on Fluid Evolution, Mi-
for this application is for the fractures to have some fluid conduc- gration and Interaction in Rocks’, Torquay, England, 4–7 May 1993,
tivity. Some operators use proppant filled canisters which open J. Parnell, A.H. Ruffell, and N.R. Moles 共eds兲. Also BP Columbia.
during the EOB process. The theory is that the proppant is en- 5. Blosser, W.R., Behrmann, L.A., Venkitaraman, A., and Kapacki, B.:
trained in the high velocity flow and erodes the fracture face. ‘‘An Assessment of Perforating Performance for High Compressive

192 L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999
 Fig. 6–Gun orientation, Test One. Fig. 7–EOB surge test.

Strength Non-Homogeneous Sandstones,’’ paper SPE 30082 pre- of 2.75 in. inside diameter 共ID兲 casing兲 was placed on top of the
sented at the 1995 SPE European Formation Damage Conference held
existing 2.75 in. ID casing with a break seal separating the two.
in the Hauge, the Netherlands, 15–16 May.
The top stack was pressurized with argon to 7880 psi 共total gas
6. Berhmann, L.A. and Elbel, J.L.: ‘‘Effect of Perforations on Fracture
Initiation,’’ JPT 共May 1991兲 608. volume at 7880 psi was 1.51 gal兲. The fluid in the casing across
7. Petitjean, L., Couet, B., Abel, J.C., Schmidt, J.H., and Ferguson, the perforations was 1.51 gal of brine at 2110 psi. The seal was
K.R.: ‘‘Well Productivity Improvement Using Extreme Overbalanced explosively removed and the perforations surged at a differential
Perforating and Surging—Case History,’’ paper SPE 30527 presented pressure of 5770 psi. The time history of the wellbore and pore
at the 1995 SPE Annual Technical Conference & Exhibition, Dallas, pressures are shown in Fig. 7. The rapid decrease in wellbore
22–25 October. pressure is interpreted as fracture initiation. The pore pressure can
leaked after the surge and it was not possible to produce the rock
Appendix A: EOB Laboratory Experiments under a back pressure to remove the injected Argon gas. The
Two large Berea block EOB experiments were conducted in the block was removed and mounted in another pore-pressure can and
TerraTek polyaxial stress frame to: 共1兲 evaluate the performance placed back in the stress frame. Brine was flowed for 4 hours with
of EOB during the perforating event and after perforating under- increasing drawdown to 300 psi. Substantial gas was being eluted.
balanced, 共2兲 evaluate fracture conductivity for very short frac- The final PI was 0.12 cc/s/psi indicating a two phase liquid/gas
tures, and 共3兲 determine if multiple fractures are initiated. Six
shot/ft guns with 3.5 g charges with ⫾45° phasing were used. The
guns were oriented with one plane of three charges 30° from the
preferred fracture plane 共PFP兲, Fig. 6. For both experiments, the
horizontal stress differential was 1,000 psi and the differential
wellbore-pore pressure was 5,750 psi. Details of the use of this
experimental facility can be found in Ref. 6.

Block Preparation. Two in-situ pore pressure probes were

mounted in the block, one 3 in. from the wellbore and a second
one 3.88 in. from the block outer flow boundary. The block was
vacuum brine saturated by flowing 3% KCl from an 18 in. section
of the open wellbore to the block boundaries. The block perme-
ability between these two in-situ probes was about 245 md. Al-
though the wellbore was drilled with a diamond coring bit, there
was considerable wellbore damage with a permeability of 28 md
assuming a damage depth of 1 in. Observation of the split block
after dye injection through the perforations shows the damage
zone to be about 1 in. giving a wellbore damage zone permeability
reduction of about 0.1. The casing was cemented using oilfield
cement and allowed to cure with the block under stress.

Test One. Test one was shot with an underbalance of 500 psi,
sufficient to elute the comminuted sand from the perforation tun-
nel but not sufficient to produce a nondamaged perforation. The
PI increased from 0.08 to 0.22 cc/s/psi as the total brine flow rate
was increased from 3.7 to 60 cc/s. A mini stack 共five foot length Fig. 8–Photo 1.

L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999 193
 Fig. 11–Photo 4.

TABLE A-1– PRODUCTION-INJECTION FLOW RESULTS,

TEST 2

Differential
Time Pressure Production Index Injection Index
Fig. 9–Photo 2. (min) (psi) (cc/s/psi) (cc/s/psi)

0 142 0.255
41 148 0.243
relative permeability problem. Red dye was injected at 50 psi for
0 142 0.165
5 minutes followed by injection of RTV. 10 150 0.095
Fractures were initiated from perforations 1, 3, and 5 which 0 148 0.307
were 30° from the PFP, between perforations 4 and 6 which were 5 146 0.296
60° from the PFP and opposite perforations 1, 3, and 5 共called the 10 299 0.275
180° fracture兲. Fig. 8 shows the RTV in the six perforations, the 20 300 0.267
fracture crack between perforations 1, 3, and 5, and the red dye 0 143 0.146
halo around perforations 2, 4, and 6. Fig. 9 is a close-up of per- 6 148 0.091
forations 4 and 6. The fracture extended the length of the perfo- 12 313 0.08
rations, 4 and 6, and stopped. The lack of dye around the first inch 16 280 0.057
20 464 0.053
of the wellbore suggests wellbore damage. The minimal dye ad-
28 436 0.038
jacent the perforation suggests the extent of perforation damage. 33 589 0.04
Fig. 10 is a close-up of the fracture initiation from perforations 1, 41 617 0.03
3, and 5. Fig. 11 shows the exposed fracture from perforations 1, 0 295 0.257
3, and 5. Note that after about one wellbore diameter, the fracture 8 309 0.25

Fig. 10–Photo 3. Fig. 12–Photo 5.

194 L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999
 Fig. 13–Photo 6.
Fig. 15–Photo 7.

turned into the PFP. Fig. 12 shows the 180° fraction initiation.
The perforations seen in Fig. 12 are from Test Two. Fig. 13 shows
the rock being opened along the 180° fracture.
with debris and formed an impermeable barrier for injection. The
Test Two. The perforations from Test One were squeezed and stack was removed and brine produced through the block with a
sealed from the wellbore. The same gun as used in Test One was PI of 0.243 cc/s/psi, about 10% greater than for Test One. Al-
also used in this test with perforations 1, 3, and 5 located 30° west though the sand debris formed an external filter cake which pre-
of south. The PFP was in the North–South direction. A small vented injection, it was easily removed upon production. At this
stack, 44.5 in. of 9 5/8 in. casing, was placed on the casing. One point, a series of injection-production tests were performed in
gallon of red dyed fluid was placed in the stack followed by a gas
hopes of achieving good injection for a subsequent overbalance
volume of 11 gal at 7,180 psi. There was no communication with
surge test similar to Test One. Table A-1 summarizes the flow
the rock indicating the squeezed perforations were sealed. Pore
results. Note that injection continued to decline with time and
pressure was increased to 1,440 psi and the gun fired. This was the
pressure without degrading production. The final injection was
same differential as for Test One but with seven times more vol-
only 12% of the production. Fig. 15 shows perforations 3 and 5
ume of pressurized gas. The rock did not fracture. Fig. 14 shows
from Test Two. The lack of open tunnels is obvious and gives
a modest drop in wellbore pressure and a small increase in pore
validity to the hypothesis that the comminuted sand debris from
pressure. Since postshot evaluation showed that perforations were
the perforation process can form an effective external filter cake
made in the rock, it is clear that the perforations were plugged
resulting in an increased fracture gradient.

Conclusions
These tests demonstrate the following:
䊉 Fractures can be initiated from phased perforations.

䊉 Perforations nearest the PFP will have the dominant fracture,

i.e., longest.
䊉 The dominant fracture will create a second fracture opposite it

to have a bi-wing fracture.

䊉 Sand debris in the perforation tunnel acts both as an imper-

meable filter cake and as a pressure barrier resulting in higher

fracture initiation pressure.
䊉 Additional multiple fractures, as seen in hydraulic fracture

tests, have not been observed.

䊉 A fracture will initiate from a perforation that is at least 30°

from the PFP without initiating a fracture from the sand face
along the PFP. This is in contrast to hydraulic fractures.
䊉 For these very short fractures, there was zero fracture conduc-

tivity. The limited block size restricts the study of fracture con-
Fig. 14–EOB perforating test. ductivity.

L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999 195
 Lawrence A. Behrmann is Scientific Advisor and Manager
SI Metric Conversion Factors of Perforating Research at the Schlumberger Perforating
ft ⫻ 3.048* E⫺01 ⫽ m and Testing Center in Rosharon, Texas. e-mail:
gal ⫻ 3.785 412 E⫺03 ⫽ m3 lbehrmann@slb.com. With Schlumberger for 17 years, he has
in. ⫻ 2.54* E⫹00 ⫽ cm been involved in all aspects of perforating. Behrmann holds a
psi ⫻ 6.894 757 E⫹00 ⫽ kPa BS degree from the U. of California, Berkeley, and an MS de-
gree from Lehigh U. He was a 1993-94 member of the Forum
*Conversion factors are exact SPEPF Series North America Steering Committee. Bryan McDonald
is Operations Manager Wireline and Testing with Schlumberger
in Alaska. He has held various positions with Schlumberger in-
cluding Laredo District Manager. McDonald holds a degree in
petroleum engineering technology from Oklahoma State U.

196 L.A. Behrmann and B. McDonald: Underbalance or Extreme Overbalance SPE Prod. & Facilities, Vol. 14, No. 3, August 1999