Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

.

Effect of Perforation Damage on


Well Productivity
J. A. Klotz, SP13-AIME,
UnionOilCo, of California
?. F. Krueger, SPE-AIME,UnionOilCo. of California
D. $. Pye, 5PE-AIME,Union nil Co.of California

Introduction

Gun perforation has been used for more than 40 years


for generating a controlled flow channel between oil
and gas reservoirs and the bore of an injection or
production well. The first well reported to be gun
perforated was a Union Oil Co. of California well in
the MontebeHo field, Los Angeles County, Calif., in
1932. Since that time many types of special bullets
and jets have been int~oduced to improve the perforating process, and we include both devices in the
term gun perforating,
Although gun perforating became widely accepted
as a practical completion method, engineers long suspected that well productivities should be better than
observed. Experimental and field studies,:- 7 ranging
over more than a 20-year period, exposed deficiencies
in perforator design and perforating procedures that
accounted for the reduced productivities or injectivities and lead to improved field results. Laboratory
observations on linear, perforated cores demonstrated
that crushing and compaction of rock during perforating substantially impaired the flow capacity of the
hole.
Although this experimental work provided a better
understanding of the physical effects of perforating
into rock, our ability to estimate how much oil, gas,
or water should flow into a wellbore through perforations formed under down-hole conditions has been
Iimited by the simplifying assumptions required to
handle this problem. Because of the difficulty of the

problem, the early work- on the productivity of


perforated welis assumed flow through clean, undamaged perforations. Recently, Bell et al, in
analytical and experimental studies related ~erforation
efficiencyfor a single damaged perforation in a radial
system to that observed in a linear core, but no attempt
was made to estimate over-all well productivity.
The purpose of our paper is to extend the earlier
work on productivity of perforated completions by
taking into account the depth and severity of permeability damage caused by both the dril;ing and the
perforating processes. We are now able to relate flow
efficiencies and permeability damage in a laboratoryperforated linear core system to that in a practical
radial well system. To limit the length of the text and
yet illustrate adequately the general combined effect
of perforating and drilling damage, we have modeled
a radial system with an optimum perforating condition
of 4 holes/ft. At this short density and a penetration
of 6 in., perforated well productivity approximates
open-hole productivity. The actual numerical effects
on well productivity will, of course, be different for
other perforation patterns but the general effects of
perforation damage on productivity will be similar.
Thus, regardless of perforation pattern, the trends
indicated in this work should be helpful in designing
perforating jobs.
We accomplished this objective with a computerized finite element method; and, although our com-

To maximize pr~ductivity, perforations must penetrate substantially beyond the zone of


drilling damage, and they must be of the highest possible quality. In a well with drilling
damage, a \ew deep perforations are more effective than many shallow ones; but within
the limits of current technology and economics, severe perforation damage cannot be
entirely overcome by increasing either shot density or penetration.
IOVEMBER,1974

1303

puter model is not perfect, we beliele it represents


a realistic step forward in estimating well productivity
when the formation is damaged from the drilling and
perforating processes.

Evaluation of Perforation Damage


Publications of several investigators- have shown
that conventional perforating practices impair the
productivity and injectivity of perforations. Perforations are never clean, at best; the act of perforating
crushes the rock and forces the particles from the hole
area into the surrounding formation. Under adverse
conditions of too great wellbore pressure and presence
of drilling mud or dirty completion fluid, severe additional darrage may result,
As a result of this work, the industry adopted in
1962 a standard test procedure, API Recommended
Practice No. 43 (RI? 43), for comparing the flow
properties of perforations. We shall describe briefly
the process and philosophy of this test because it forms
a foundation for our present work.
The 1962 edition of RP 43 describes perforation
effectiveness in terms of 2 Well Flow Index (WFI)
determined for flow through a perforation made into
a standard cylindrical Berea sandstone core under
simulated wellbore conditions. WFI was defined as
the ratio of the apparent permeability of the perforated
core (kP)to the pemleability of the unpe.rfoiated core
with both ends open (kO),The WFI so defined was
misleading in that the effect of perforation damage
was masked by the effect of penetration depth. The
index name was a misnomer because it had no relation
to down-hole weHproductivity in a radial system.
In recognition of these problems, API RP 43 was
revised in 1971to provide a measure of permeability
damage in the perforation. In the new procedure, the
apparent permeability measured in a core with a real,

damaged perforation (Q is compared mathematically


with the permeability of a core with an ideal, clean
perforation of the same depth (ki). This new measure
of perforator performance is called Core Flow Efficiency and is defined in RP 43 (197 1) as

For a clean, undamaged perforation, ki/ko can


be calculated by the finite element method to be
described later, as it depends upon the dimensions of
the cylinder and of the perforation. RF 43 tabulates
the theoretical permeability ratios, ki/kO, for several
test core lengths, and Fig. 1 presents a graph of typical
values for a 12-in.-long core. Thus R~ 43 now provides a measure of perforation qual:ty. CFE near 1,0
indicates a relatively clean perforation; CFE < 1.0,
a dirty, or damaged, perforation,
But how does CFE for a linear core relate to well
productivity in a radial system? To answer this question a modification of our finite element model was
used to estimate flow through the same perforation in
a radial well system.
The relationship between CFE and well productivity depends upon a large number of parameters:
well diameter, perforation diameter, perforation depth,
shot density, severity and depth of perforation damage, severity zmd depth of formation damage from
drilling or workover fluids, and well drainage radius.
Because of length limitations for this paper, we will
not attempt to discuss all combinations of these factors, but will present the results involving variations
of those we consider most influential: perforation
depth, formation damage from drilling or workover
as well as from perforating, and shot density. I-Iowever, before discussing the results, we shall describe
how the calculations are made.

The Finite Element Method

1!
OL---J
02

Perforation
Depth, inches
Fig. l-Calculated
permeability ratio for ideal,
clean perforation in a 12-in. core
cylinder (data from Ref. 13).

1304

The heart of our work involves calculation of flow


through a perforation in either the RP 43 core cylinders or in a well. The calculations were made using
a finite element method for estimating pressures and
flow in permeab!e porous media.
The finite element method was first proposed for
use in problems involving analjjsis of stress and strain
in structures such as steel frame buildings, aircraft
frames, and offshore platforms. For such structures
each frame member constitutes an element and each
joint between members a node. Equations can be
written in matrix form for forces at each node and
for stress and strain in each element. Then, the
problem of determining the forces, stresses, and
strains can be solved by well known methods of matrix
algebra, This application is described by Willems and
Lucas.4
After considerable success in structure analysis,
the finite element method was extended to continuous
materials, These applications and the theory behind
them are described by Zienkiewicz and Cheung.15
Chapter 10 of their book outlines the finite element
theory for F]eld Problems Heat Conduction,
Seepage Flow, Etc., which has been used in our
JOURNALOF PETROLEUMTECHNOLOGY

work.
For this paper the finite element method consists
of the following procedure:
A plane section of the reservoir rock or test
cylinder is difided into a number of four-sided finite
elemems. Th: elements can have almost any shape;
but for this work, we use rectangular elements for
line~r flow through a test cylinder and nearly parallelogram elements for radial flow into a well, Figs, 2
and 3 illustrate the finite element grids used for the
cylindrical core and the radial well system.
Junctions of the lines surrounding neighboring
elements are serially numbered nodes, We used 600
nodes and eIements for this work; but for clarity of
illustration, the figures are drawn for grids containing
only 200 nodes and elements. The geometry of the
system is described by the position, i.e., the x and y
coordinate of each node. Each element is identified
by means of the serial numbers of the nodes at its
corners. The program can handle a large number of
nodes and elements, limited only by dimension statement in the computer code. We used approximately
600 nodes and 600 elements to describe each flow
system.
Flow resistance through the permeable media is
described by assigning a permeability to each element;
we used four different permeabilities:
1. Virgin undamaged reservoir rock permeability,
k, = 1.0.
2. Permeability of the region around the wellbore
damaged by drilling, 0.0< k, <1.0.
3. Permeability of the region around the perforation damaged by perforating, 0.0< kv <1,0.
4. Permeability of the region damaged by both
drilling and perforating, k,. For most of our work,
k, = k, X k,,lG but for the work summarized in Figs.
11 and 12, k, was varied independently.
The computer code works in the following reamer,
First, it calculates coefficients describing the flow conductivity of each element as it depends upon the shape
of the element (location of the corner nodes), and on
the permeability of the element. Then, these coefEcients are stored in a 30 X 600 array that eventually
will be a banded matrix describing the conductivity of
the entire system,
Next, boundary conditions are used to modtiy a
600-point vector that also will be part of the final
solution. Our boundary conditions consisted of a
fixed pressure at the nodes along the input face of
the test cylinder, or along the outer, input boundary
of the radial system, and a different fixed pressure at
the nodes lying on the surface of the perforation.
Next, the computer code solves the banded 30 X
600 matrix along with the boundary condition vector
by means of a direct Gaussian solution to determine
the pressure at each of the nudes corresponding to its
position, conductivity of each element, and specified
boundary conditions. Finally, after pressurt% at each
node have been determined, the computer code calculates velocity and the direction of flow through each
element in the system.
The finite element computer code that forms the
heart of our calculation was written in the Civil
NOVEMBER,1974

Engineering Dept. at the U, of California at Berkeley


and is described in a paper by Taylor and Brown.17
The basic mathematical technology in the code* has
not been changed, but we adapted it to our particular
problem.
The arrangement of nodes and elements shown on
Figs. 2 and 3 is calculated automatically by an auxiliary gridding subroutine,
For the linear, cylindrical system, the gridding
program accepts cylinder diameter, cylinder length,
perforation diameter, perforation length, and damaged zone thicknesses as input data, Then it automatically draws a grid such as shown in Fig, 2,
Elements and nodes are clustered near the tip of the
perforation, where flow directions change rapidly and
where pressure gradients are greatest.
Although the grid appears here as a two-dimensional, plane m-ray, the computer code considers it
to be a two-dimensional, axisymmetric array with the
axis of the perforation, coincidental with the axis of
the cylinder, as the axis of symmetry, Thus, each
element is a three-dimensional ring whose rectangular
cross-sectiori shows in Fig. 2. For the linear, cylindrical system, this is an accurate representation.
For the radial system, the grid shown in Fig. 3
is developed in a similar manner but an approximation of well geometry is involved. Input data to the
*A Iktinz

of

Tavlors

code can be obtained

Rcoion
of
Pw-foration

5arnage-(1/2

E$#av
;~

inch)

from the amhor% of

k=kl=l.-O

Typical
length
15 in
E&n

Pressure at Inlet
Nodes = + 10

Ill

Fig, 2Finite.eiement
grid for core cyiinder. Grid is a
haif section of cylinder with center
iine at perforation axis.
1305

auxiliary gridding code for this system include well


diameter, vertical and angular distance in the well
between perforations, and outer radius of the system
as well as perforation depth and diameter, and ciamaged region dimensions.
The computer code considers that each perforation
is centered on a portion of wellbore face whose area
depends upon well diameter and perforation spacing.
For example, in a 6-in.-diameter wellbore, with 4
shots/ft, spaced at 90 around the circumference of
the well, each perforation is at the center of a bore
face area 4.71 in. wide and 12 in. tall (56.5 sq in.),
Ihe computer code makes the approximation that
this area is a plane circle instead of a bent rectangle.
The plane circle has the same area as the bent rectangle and is centered at the perforation axis. For
this example, the circle has a radius of 4.24 in,
Similarly at the outer edge of a 5-ft radial system,
the cross-sectional area available for radial flow
would be a bent rectangle, 94.25 in. wide and 12 in,
tall (1,131 sq in. total area). The computer code
maintains the assumption that this area is circular,
Thus, the bent rectangle at 5-ft radius in a system with
four perforations per foot is represented by a circle
18.98 in. in radius.
The net result is an axisymmetric flow system that

d
PERFORATIONS

Region

of

Drilling

d WELL

Radiu IS from
Well
t

is shaped somewhat like a truncated circular paraboloid, Fig. 3 shows a section of this paraboloid cut
through the axis of symmetry, the axis of the perforation. The elements in this system are rings of nearly
parallelogram cross-section, except that elements in
the damaged zone around the perforation remain
rectangular in order to maintain a damaged zone of
uniform thickness,
To obtain good flow definition with small elements
around the perforation within the 600 node and
element limit imposed by our current computer code,
we limited the radius of the radial finite element
system shown in Fig. 3 to 5 ft. For each perforation
configuration, we then were able to calculate an effective flow resistance for the 5-ft radial finite element
system and to translate this to a 660-ft radial well
system by means of the usual fcrrnulaS for flow
through porous media with an effective discontinuity
in permeability. The translation assur,les uniform
pressure at the outer nodes of our radial system and
uniform flow rates in the outer ring of elements,
Uniform pressure was specified as a boundary condition, and calculated flow rates in the outer elements,
although not absolutely uniform, deviated by only
about 1 percent from an integrated average. Thus, we
felt secure in the translation from a 5-ft outer radius
to a 660-ft radius.
The computer code calculated the velocity and
direction of flow through each element. We considered
that the summed flow through the single row of
elements at the input end of the cylinder, or the
single row of elements at the outer input edge of the
radial system, represented flow through the entire
system. Our calcu-latedflow rate was then the sum of
the products: (flow velocity in each input element)
X (annular area availab!e for flow in this element).
Figs. 4 and 5 show the flow distribution into a
perforation in a linear core and in a radial system,
respectively. A half-section of an 8-in.-long perforation is shown schematically, and arrows indicate the
.~ercenta~e of flow into each of four 2-in.-long segments. In the linear system when the perforation is
undamaged, almost 80 percent of the flOWis near the
tip of the hole, However, when the permeability of
the rock surrounding the perforation is damaged by
the perforating process, the added resistance forces
flow more toward the base of the perforation, and
it is more evenly distributed over the entire surface
of the perforation. Of course, total flow is less.
In the radial system, on the other hand, flow into
the perforation is distributed fairly uniformly over
the entire length of the perforation for both the undamaged and damaged cases, and total flow at the
tip of the undamaged perforation is only 9 percent
greater than at the tip of the damaged one.

Results

Fig, 3Finite.element
grid for radial system around a
well. The grid is a half section of a paraboloid with
center iine at perforation axis,
1306

We investigated three relationships in this work:


1. The relationship between Core Flow Efficiency
(CFE), as defined in API 22P 43, and a similar term,
Well Flow Efficiency (WFE), that we defined for flow
into a perforated well, WFE is defined as the ratio of
flow rate into a cased well through a real perforation
JOURNALOF PETROLEUMTECHNOLOGY

in a zone that has been damaged by perforating,


drilling, or workover to the flOWrate into the same
well through a clean, ideal perforation of the same
depth in an undamaged zone. The term WFE for a
radial system should not be confused with WFI from
RP 43, 1962 edition, which related to :; linear core.
2. The effect on WFE of a permeability-damaged
region around the wellbore, such as might be caused
by the drilling or workover process when the perforations are ideal and undamaged.
3. The effect on WFE of a permeability-damaged
region around the perforation, such as might be
caused by the perforating process, in a formation without formation damage from drilling or workover.
Finally, we combined all these. effects to determine
how WFE is affected when both drilling damage and
perforation damage are present at the same time.
Our calculations assume steady-state flow of an
incompressible fluia and all single-phase perrneabilities. Thus, all permeabilities reported in this paper
are, in effect, the permeability to whatever phase is
flowing; all other phases are immobiie and permeability to the flowing phase does not change from place
to place in the system because of saturation changes.
Examples of the effects of various parameters on
WFE in the following sections are given for a well
with a 6-in.-diameter wellbore, perforated with four
0.4-in.-diameter holes per foot, and with a drainage
radius of 660 ft. This shot density was selected
because it is representative of common field practice;
and, for comparative purposes, if data for a 6-in.
perforation are selected, our results can be related to
open-hole well productivity in a formation without
permeability damage. However, we shall compare
also the effects of shot density in a damaged formation
with that in an undamaged formation.

Correlation of CFF, With WFE


Fig. 6 presents thv results of our calculations to
relate CFE for a 15-in. test core to WFE at a shot
density of 4/ft. Correlations in this paper relate to
CFES for a 15-in. test core because most published
RP 43 data for commercial perforators are given for
this length, We should point out, however, that our
studies showed that for a given penetration and damaged zone permeflbility CFE varies with test core
length because of geometrical effects; and, therefore,
the numerical results would be somewhat different for
other core lengths. However, these differences are
small within the range of perforation depths specified
in RP 43 for the various target Iengths and for the
observed CFE range for commercial perforators.
Both CFE and WFE are plotted against the pemleability, ks, of the damaged zone around the perforation and for perforation depths ranging from 2 to 10
in. The permeability, kt, in the damaged zone is
normalized and defined as the ratio of the permeability in the damaged region to the permeability of the
virgin rock. In this normalized system a damaged
zone permeability of 1,0 is equivalent to no damage,
For our studies, we assumed a ?4-in.-thick damaged
zone. The assumed thickness is consistent with experimental observations that indicate values ranging from
1/4 to s% in. in Berea cores, depending upon type of
NOVEMBE&1974

With

No Perforation
Domage
ks = 1.0

Perforation
ks = 0.05

% Flow Through Each


2 inch Segment

% Flow ?hrough Each


2 inch Segment

2 v.

Damoge

18 %

T
4 %

20%

P,,l.a, otzon

15 %

25%

\
37%

79%

Fig. 4-Distribution
of flow into perforations in core
cylinder for clean, ideal perforation and for
damaged perforation. Perforation diameter,
0.4 in.; damaged region, 1/2 in. thick.
Outline is half section of a cylinder,

No Per foratmn
k3=

Perforot,cm Domage

With

Oamoge

k3=O05

1.0

% flow ThroughEoch

% Flow Through Each


2 snch Segment
I 9 1.
*..

P*, Io#. t#o.

20%
\

2 Inch Se9ment
23 1.
\
8
P., fo..,,m

23%
\
\
24 1.

22%

\
\
39%

30%

\\

T)

Fig. &Distribution
of flow into perforations in a radial
well system for clean, ideal perforation and for damaged
perforation. Well diameter, 6 in.; drainage radius, 660 ft;
damaged zone, 1/2 in. thick; perforation diameter, 0.4 in.
4 shots/ft spaced at 90. Outline is half
section of a paraboloid.

1.0
.8 g:
~:.6
e!
* Cl
.o~

.4 -

VLU
P, 10,0 I,o
DeDlh

.-

---

2
4 .,

,.
%>
~u
c. 4
~ .%
= .:

;%

..

8
.-

10

6
8I
o 1

I I I I I 1 I
.03 .05.07 .1

1
.3

.5

.7

1.0

Permeability

Fig. 6-Relationship
between well tlow efficiency and
core flow efficiency. Well diameter, 6 in.; drainage radius,
660 ft 4 shot/ft spaced at 90.
1307

gun. However, investigation of this parameter has


shown that the assumed thickness does not critically
affect the results.
Although for convenience we have related CFE
values for Berea cores to WFE, the effects of different
types of rock material can be readily determined from
the curves if experimental results of perforating in
linear cores are available. From experimentally determined CFE values for particular rock types, the
damaged zone permeability can be estimated from
Fig, 6. These values can then be used to interpolate
WFE values in subsequent figures.
With the upper portion of Fig. 6 we can compare
the permeability damage caused by different perforators. Published CFE values for most modem gun
perforators range between 0.65 and 0.85, and from
the figure we see that the damaged zone permeability
can range from about 7 to 35 percent of the undamaged formation permeability. Our results compare
well with the calculated value of 10 to 20 percent
of the undamaged zone permeability for a CFE of
0.75 reported by Bell et al.; at this same CFE, Fig.
6 indicates values ranging from 12 to 19 percent.
WFE and CFE are related as illustrated by the
arrows on Fig. 6, For each CFE we determine a
damaged permeability for the perforation and then
relate this damage to a corresponding WFE. For
example, a measured CFE of 0.7 in a 15-in. test
cylinder with a 6-in. pel foration is the result of a
normalized permeability of 0.085 in the assumed 1Ain.-thick damaged zone. In the example radial well
system with 4 holes/ft, the same damaged zone will,
in turn, cause a WFE of 0.58, These results could be
typical for some perforated completions shot under
favorable conditions (salt water in hole, pressure drop
into wellbore),
For CFE = 0,3, corresponding to a perforation
made under mud with pressure drop into the formation, the normalized damaged zone permeability is
0,013 and the corresponding WFE is 0.16.
The relationship between CFE and WFE is not
sensitive to our assumption of a Y2 -in.-thick damaged
region. But if we had assumed a different thickness, 3
the calculated intermediate values, the permeabilities
of the d maged region, would be changed; for a given
CFE the permeabilities would be lower for a region
less than 1A in. thick and greater for a region more
than 1/2 in. thick. But the WFE/CFE relationship is
essentially unchanged for a damaged zone thickness
ranging between 0.2 and 0.7 in.
In the radial system, then, a WFE range of 0.50
to 0.90 corresponds to the CFE range of 0.65 to 0.85
observed for conventional perforators under the conditions set in RP 43. These WFE values are substantially higher than the values of 0.25 to 0.35 calculated
by Bell et al. The major difference is probably associated with our shot density of 4/ft, compared with
Bells analysis for a single shot in a semi-inthite
medium.
From the correlations given in Fjg. 6 and reference
to relationships between open-hole productivity and
perforation depth and density, we shall be able to
estimate in subsequent sections how laboratory1308

measured perforator performance (CFE values) relates to well productivity when perforating and drilling conditions are known.

Effect of Perforation Depth in a Zone


With Permeability Damage
In earlier work by others,- the effects of perforation depth were investigaled for a virgin, undamaged
formation, Now with our finite element model we are
able to illustrate the effe,:ts of perforation depth cm
the productivity of a well that has been completed in
a zone in which the permeability has been damaged
by drilling, In the examp!e shown in Fig. 7, the perforations are assumed tc be ideal and undamaged,
and drilling damage extends for a radial distance of
4 in. from the wellbore. Permeability of the damaged
zone ranges between 5 and 100 percent of the virgin
reservoir permeability,
As mjght be expected, WFE is low when drilling
damage is seveie and the perforation does not penetrate through the dama,ged zone, No appreciable
improvement occurs unti I the perforation penetrates
through the damaged zone. However, the productivity
is significantly reduced until the perforation extends
40 or 50 percent bevond the region of driliing damage.
Inasmuch as simdar relationships will hold for perforations in a deeply penetrating damaged zone, a
strong effort should be made to avoid drilling or workover damage that cannot be penetrated substantially
by commercial gun perforators.

Combined Effects of Perforation


Damage and Drilling Damage
Application of our radial model has been extended
a step further in Figs. 8 through 10 to include a study
of perforation effectiveness in a damaged radial
system when the perforations are also damaged, In
these figures it is assumed that perforating damage
was superimposed on the drilling damage; that is,
k, = k, X k,. Inasmuch as the effect of damage during perforating on already damaged permeability is
not known, we shall also show similar results when
perforating damage is assumed to be independent of
drilling damage.
Figs. 8,9, and 10 are the inverse of Fig, 7. Whereas
Fig. 7 demonstrates the effect of changes in perforation depth in and through a 4-in. damaged zone, in
these figures the perforation depth is held constant at
8 in, and the effect of formation permeability damage
is shown for damage depths ranging from O to 24 in.
The family of curves again shows the result~ for
damaged permeability ranging from 5 to 100 percent
of virgin rock permeability,
The information given in Fig, 8 is derived for a
completion in which there is no perforation damage
and therefore is comparable with the results in Fig. 7
for the same depth of drilling damage and perforation
depth. As in Fig. 7, WFE is depressed stuix!antially
when a moderate to severe amount of drilling damage
extends to, or beyond, the tip of the perforation; and
it is not until penetraticm of the drilling damage is
limited to about 50 percent of the perforation depth
that a major improvement is noted in WFE.
JOURNALC)FPETROLEUMTECHNOLOGY

Figs. 9 and 10 include the effects of perforation


damage. In Fig. 9 we assume that the perforation
process reduces the permeability in the V2-in.-thick
region around the perforation to 20 percent of the
virgin rock perrneability; in Fig, 10 we assume that
the damaged permeability in this region is 5 percent
of the original value. As shown in these figures, perforation damage prevents attainment of undamaged
productivity values. For k, = 0.2, a common ckgree
of perforation damage, maximum WFE is only 80
percent of the undamaged value; and for k:, = 0.05.
severe perforation damage, maximum WFE is only
about 50 percent of the undamaged value, There is
little change in WFE as long as the drilling damage
penetrates beyond the perforation depth. As in the
preceding figures, WFE does not approach the maximum value unless the depth of drilling damage can
be restricted to less than half the perforation depth.
The following example can be instructive, Fig. 10
shows the WFE in a zone with drilling damage when
the permeability of the perforation-damaged region
is 5 percent of the virgin permeability. Consider a
well where drilling damage reduces permeability to

10 percent. Then, an S-in. perforation with 4 in. of


drilling damage results in a WFE of about 34 percent.
If the perforating process car, be improved so that
permeability in the perforation-damaged zone is 20
percent, then as shown on Fig. 9, the well will tolerate
a drilling damaged zone 8 in. deep (equal to the
perforation depth) while maintaining a 34-percent
productivity ratio, If perforation damage can be eliminated, then as shown on Fig, 8, the depth of drilling
damage can be twice the depth of the 8-in, perforation
while maintaining a 34-percent productivity.
A major point of interest is the importance of
perforating performance on well productivity. We
conclude from this study thrit when there is no perforation dan,age, it is possible to overcome the effect
of drilling damage by substantial penetration of the
perforation beyond the zone of drilling damage; hex<lever, when the perforation is damaged by the perforating process, VVFEis limited to a maximum of
0.8 when k. = 0.2 and to about 0.5 when k, = 0,05,
even when the perforation extends well beyond the
drilling damage. On the other hand, when the perforation does not adequately penetrate the drilling damage,

,8 inch
,@rotion

kz

depth, k3 e 0.2

1.0
.7

.4

Oo&_i-~
2

10

12

14

16

18

Depth of Penetration,

inches

Fig. 7Effect of ~erforation depth on WFE in well with


dr~ling damage (k; < 1,0) when perforations are urrdam.
aged (k, = 1,0). Well diameter, 6 in.; drainage radius
660 ft; 4 shots/ft,

8 inch perforation

1.0

depth,

ka s 1.0

>
v

Depth

Drilling

Damage,

inches

kz

~o~

1.0

of

Fig. 9Effect of drilling damage on WFE when perforations


are damaged (average performance, k:] = 0.2);
k4 = k3 X kz.

,7

.8 -

j
u

.4

:
ii

,4 -

1.0

.7

IL

.1

<

3.2 _

.05

($~

kz

~ .6

.4

g ,2 -

Depth of Drilling
Damage,
inches
Ffg. 8-Effect
of drilling darnage on WFE when perforations
are undamaged,

NOVEMBER,1974

k
Depth

of

Drilling

Damage,

inches

Fig. lCzEffect
of drilling damage on WFE when perfora.
tiers are severely damaged (k3 = 0.05); k4 = ks x kz.
1309

8 inch

1.0 -

~
.-al
U

.8

.6 -

depth, k3 = 0,2
I

kz
1.0
.7
.4

3
L0

,4 .1

.2 -

.05

oo~ilg.

Depth of Drilling
Damage, inches
n-Effect
of drilling damage on WFE when perfora.
tions are damaged (average performance, ks = 0.2):
I(4 = k3; k4 # kz x k3.

~ .8 -

8 inch

.-aJ

perforation
depth, ka = 0.05

~ .6 u-!

5
0 .4 c

5
~ .2 -

j~,),\;o;4
Depth

of

Drilling

Damage,

inches

Fig. 12Effect of drililng damage on WFE when perfora.


tions are severely damaged (ks = 0.05); ks = ks;
kd # k2 x k3.

Thickness
of Damage
Around the Perforation,
inches
Fig. 13Effect of thickness of perforation-damaged
on WFE in well with no drilling damage.
1310

zone

WFE declines rapidly with increasing damage to as


low as 5 to 20 percent of the undamaged values.
Experience has shown that in many formations moderate permeability damage ratios of 0.4 to 0.7 are
common after drilling, and the perforation damage
values used above are common for typical perforators
and typical shooting conditions. Our curves indicate
that for this range of values WFES would range from
0.3 to 0.7.
Because of the conditions selected for our examples,
WFE values in our examples should approximate well
productivity ratios relative to open hole. Thus our
analysis indicates that commonly used drilling and
perforating practices could result in restrictions in well
productivities on the order of 30 to 70 percent of
undamaged open-hole productivities.
To relate these results tc observed perforator performance in laboratory, linear core systems, note that
the 20-percent permeability ratio assumed in Fig. 9
for the damaged zone around an 8-in. deep perforation is associated with CFE = 0.75 to 0.85 (Fig. 6);
a 5-percent permeability ratio is associated with CFE
= 0.5 to 0.6, Both values can be easily obtained with
commercial perforators, depending upon perforating
conditions used.
As noted earlier in discussing Figs. 9 and 10, we
assumed that when perforation damage occurs in a
zone previously damaged by drilling, the effective
damaged permeability around the perforation, k,, is
equal to k, X k,. In Figs. 11 and 12, both drilling
damage and perforation damage are again included;
but k, is assumed to vary independently from k,, the
permeability after drilling damage, That is, k, = k,
and k, is not a function of k,.
In view of the current lack of information on
second-order damage, the reader may select the
family of curves tha: best fits his own experience or
inclination.

Effect of Thickness of
Pwforation-Damaged Zone
In the discussion of Fig. 6, we stated that the final
relationship between CFE and WFE would be relatively independent of the thickness of perforation
damage. However, for a specific well, the well production efficiency obviously will depend upon the
thickness and depth of the perforation damage. This
relationship is shown in Fig. 13, where the depth of
damage around an 8-in, perforation is allowed to
vary up to 3 in, Permeability in the damaged zone
ranges between 5 and 100 percent of virgin rock
permeability.
These curves provide an interesting insight into
productivity damage that may occur during workover. The most severe damage to WFE occurs for
the first 1/2 in. of perforation damage. Beyond that
point WFE does not change much. Thus, if a dim
workover fluid is injected without fluid-loss control
through an undamaged perforation into a nondamaged formation, the permeability damage produced
around the perforations can readily reduce well productivity as much as 50 percent, even with a limited
degree of invasion. However, if reasonable precauJOURNALOF PETROLEUMTECHNOLOG-f

tion fluids are used that is, with high-fluid loss and
formation damaging characteristicsinability to
penetrate through the darnaged zone with existing
perforators can result in extreme loss in well productivity. On the other hand, even completions that are
producing in an undamaged manner because of
effective, deep perforations, wi!l be easily damaged
by invasion of a damaging fluid during workover,
because after the job the perforation will be inside
the damaged region (see Fig. 7).
To illustrate more graphically the practical implications of our studies, we have used our model to
estimate Ihe productivity that could be expected from
a typical well, completed with cemented casing and
perforated with 4 shots/ft, compared with an assumed
potential productivity of 800 B/D in an undamaged
open hole. The results are given in Tables 1 and 2.
We have assumed two different sets of drilling conditions: one with an ideal fluid that causes no formation
damage, the other with a fluid that damages formation
permeability, In the second case, we investigated the
effectsof both a moderately damaging fluid (k, = 0.7)
and a severely damaging fluid (k~ = 0.1) with three
different invasion depths, 4, 8, and 12 in. In practicq
of course, fluid invasion and the associated damage
may penetrate even deeper, depending upon drilling
and completion conditions.
Results are given for two different perforation
depths, 4 and 8 in., and a range of CFE values from
0.3 to 1.0. The 4-in. penetration corresponds to iesults
with certain small through-tubing guns, poor shot
phasing, or shooting through multiple strings of pipe,

tions are taken to use a relatively nondamaging fluid


with adequate fluid-loss control, the productivity reduction can be limited to a minor amount.
If the perforation is already seriously ddmaged
from the completion process, for example as indicated
by the curve labeled 0.1, fluid invasion that adds to
the depth of damage say from 1/2 to 3 in. or more
reduces WFE a relatively small additional amount
if reasonably good workover fluid is used.
From an over-all viewpoint we should keep in mind
that a workover treatment can create effectively a
condition similar to perforating with insufficient penetration into a previously damaged zone. Therefore,
previous relationships should be remembered regarding the productivity of wells in which perforations are
damaged in a zone with drilling damage also.

Some Examples of the Effect of Drilling and


Completion Conditions on Well Pro: ~~
.ivity
In the Introduction, we pointed out that CFE values
were, up to now, suitable only for comparisons as to
the quality of perforations produced by different guns.
However, with the foregoing analysis, we are now
able to show how CFE values can be used practically
to estimate how different completion practices can
affect well productivity. Although the values obtained
may not be precisely accurate because of simplifying
assumptions, we believe the effects of the various
parameters are qualitatively correct and therefore our
analysis provides a tool for critically evaluating the
importance of various operating practices. For example, we have seen that if poor drilling and comple-

TABLE

lEFFECT

CFE
0.3

0.5
0.7
0.8
0.9
1.0

OF PERFORATING CONDITIONS AND PERFORATION DEPTH ON PRODUCTIVITY OF A WELL


WITH AN OPEN-HOLE POTENTIAL OF 800 B/D
(No drilling damage, perforated with 4 holes/ft.)

Perforator
Average Poor
Average
Average
Best
Ideal

Perforating Conditions
Fluid

-1-Ap = Wellbore pressure > formation


- Ap = Wellbore pressure < formation
TABLE 2EFFECT

CFE

Perforator

0,3

Average
Poor
Average
Ideal

05
0.8
1,0

Pressu rc

High solids, mud in hole


Unfiltered salt water
Filtered salt water
FiItered salt water
Clean, nondamaging fluid,
best techniques available
Clean, nondamaging

Well Productivity (B/D)


at Perforation Depth of
4 in.
8 in.

+ AP
+ Ap
+ Ap
AP
Ap

115
253
429
538
653

154
330
569
689
792

AP

768

856

pressure.
pressure.

OF PERFORATING CONDITIONS AND DRILLING DAMAGE ON PRODUCTIVITY OF A WELL WITH


AN OPEN-HOLE POTENTIAL OF 800 B;D
(S-in,.deep perfomtion, 4 holes/ft, formation damaged during drilling.)

Perforatimz Conditions
~luid

High solids, mud


Unfiltered s?lt water
FiItered salt water
Clean, nondamaging

4 in.
kD/k.
Pressur=
+
+

Ap
AI)
Ap
AP

0,1
Q~

219
576
803

Well Productivity (B/D)


Formation Damaged to Depth of
8 in.

12 in.

ko/k,

kD/k.

_0.7
136
297
661
843

_0.1
15
56
247
530

_0,7
114
259
615
813

_0.1
9
36
162
331

0.7

112
254
601
794

Note: k.
permeability of damzged zone
= kz.
-K
permeability of undamaged zone
+ Ap = wellbore pressure > formation pressure,
Ap = wellbore pressure < formation pressure.

NOVEMBER,1974

1311

The 8-ii. depth is fairly typical for many well designed


perforating jobs. Certain special guns can provide
even deeper penetrations; but with the conditions
selected for our example, the penetration values are
adequate to illustrate the effects of perforating conditions and amount of formation damage.
The CFE values used cover common completion
practices ranging from poor to ideal, When a producing formation is perforated under drifling fluid with
a large pressure drop into the formation, earlier publications- give permeability damage ratios of 0.5
or less, which correspond to CFE values of about 0.3
or less. The larger CFE values used in our example
correspond to improved perforators and perforating
conditions as described in the tables.
As shown in the tables, depending upon perforating
conditions and type of perforator, the productivity of
this hypothetical 800 B/D well could range from as
low as 115 B/D to 856 B/D when no drilling damage
is present, and from 9 to 843 B/D when drilling
damage is present.

1.4
1.2

K1.mz,
K,v, o,,
-

.2tiu_lA4~
O

S P,.

H.,,,,

. . ..-..

Perforation

-. M. DL.w. !I 6 Mvsk.t
rb..d
& W.t,.a.

10
12
14
Penetration,
inches

16

of shot density and penetration on productivity ratio. Well diameter, 6 in.; drainage radius, 660 ft:
perforation diameter, 0.1 i%

o
..

FJo. of Shots Per Foot


FI . 15-Effect
of perforation damage, as characterized by
la f oratoty CFE values or calculated k3 values, on WFE for
given shot density Perforation depth, 6 in.
1312

Relative Effects of Shot Density


And Penetration
Previous investigators 9,10 concluded that in ideal,
undamaged radial flow systems shot density primarily
governs the productivity of the well and is more important than penetration. Four perforations 6 in. deep
were shown to provide well productivity equivalent
to open hole, and several shallow perforations were
concluded to be more effective than a single, deeply
penetrating one.
To compare the results from our finite element
model with the work of these earlier investigators, we
calculated the well productivity ratio
productivity of a perforated well
productivity of anopenhole )

Fig. 14-Effect

00+

This example illustrates the importance of careful


engineering of the completion to minimize formation
damage from drilling and to optimize perforator
selection. The prociuctivity values show clearly that
when drilling damage cannot be avoided, it is extremely important to select a gun perforator with both
a high CFE value and a penetration potential that will
substantially exceed the depth of formation damage,
particularly when the permeability reduction is severe.
If this is done, the effect of formation damage can
be almost negated if the CFE value for the perforator
is 1.0; that is, if there is no damage in the perforation
itself. However, with a CFE value of 0.8, typical of
many commercial guns, the perforation damage prevents effective bypassing of the drilling damage,
On the other hand, if the well can be drilled with
little formation damage, the effect of perforation
depth is less critical, although it is still important to
perforate in a manner that will provide high CFE
values.

for the same ideal, radial flow system (the perforated


well is assumed io have no drilling or perforation
damage). The denominator of this ratio can be calculated in our perforation model by assuming (1) a
perforation cross-sectional area almost equal t the
surface area of the wellbore segment and (2) an irdinitesimal perforation depth. In Fig. 14 the results of
this calculation are plotted along with published
results of Harrisll and of McDowell and Muskat.10
The shape of our curves agrees well with the curves
of McDowell and Muskat, and the variation in actual
values from point to point probably results from their
inability to exactly scale perforation diameter in thGh
electrolytic model. On the other hand, our curves
agree well with I-Iarns results at low penetration
values and at high-shot density but diverge at low-shot
density. The simplifying assumptions that we make
are somewhat &Terent from his and therefore we cannot expect perfect agreemtint, For example, Harris
assumes a radially expanding perforation to conform
with his radial segment, whereas we assume a cylindrical perforation in a truncated circular paraboloid
model with cross-sectional areas equivalent to a radial
segment. Our truncated circular paraboloid model
cannot amountaccurately for flow around the casing.
Because these effects are most important at ve~ low
JOURNALOF PETROLEUMTECHNOLOGY

perforation densities, our results are less accurate at


a density of 1 shot/ft.
We carried these studies one step further and investigated the effect of shot density in a perforated
radial system in which the perforations are damaged.
As a measure of perforation damage, we used
laboratory -measured CFE values for 15-in, cores
and assumed again a 1/2-in.-deep damaged zone. Results are given in Fig. 15 for 6-in,-deep perforations.
The top curve labeled CFE = 1.0 (k, = 1.0) corresponds to the results of previous investigations of an
ideal, undamaged completion. As previously, at a
density of 4 shots/ft the productivity ratio is 1,0.
We note now that, with the exception of perforations with only moderate permeability damage (ks =
0.46), corresponding to CFE = 0.9, it is not possible
to achieve open-hole productivity with a reasonable
shot density, With a perforator and conditions characterized h, CFE = 0.9, about the best system presently available, shot density must be increased to
about 7/ft to achieve open-hole productivity.
Corollary to the above, even neglecting the effects
of drilling damage, it is not possible within practical
limits to overcome poor perforating procedures by
increasing shot density. We have pointed out earlier
that use of over-balanced pressures and high-solidscontcnt fluids during perforating corresponds to
CFE x 0.3 for a 6-in. perforation. At a perforation
density of 4 holes/ft, this practice would result in a
well productivityy ratio of about 0.35, according to
these curves. Doubling the shot density to 8/ ft increases the productivityy ratio only to about 0.5, compared with a value of 0,7 to 0.9 that could be attained
with better perforating practice.
Fig. 16 illustrates the relative effects of shot density and penetration depth on well productivity ratio
when perforations are damaged in an otherwise undamaged formation. Productivity ratios are plotted
for penetrations up to 18 in. and for shot densities of
4/ft and 8/ft. The perforation-damaged zone is again
assumed to be 1%-in.thick. Perforation damage values
(k,) are indicated on the curves.
The curves in Fig, 16 show that an increase in shot
density is more beneficial when the perforation is
damaged than when it is undamaged and that doubling perforation depth has about the same effect as
doubling shot density. However, for severely damaged perforations (ks = 0.05) a deeply penetrating
perforation is more effective than an increase in shot
density. For example, an increase in shot density from
4/ft to 8/ft i~i a 4-in.-deep perforation raises the
well productivity ratio from 0.35 to 0.52, which still
corresponds to a large restriction in well productivity;
but an increase in perforation depth from 4 in. to 18
in. at a shot density of 4/ft raises this ratio from 0.35
to 0.79, thus overcoming a substantial portion of the
effects of perforation damage. With the severely damaged perforations in this example, open-hole productivity can be attained by increasing the combhtdd
effects of shot density to 8/ft and perforation depth
to 18 in. or, as indicated by extrapolation of the
curve for k~ = 0.05, by increasing perforation depth
to about 26 in. at 4 shots/ft.
In Fig. 17 we investigate the effect of shot density
NOVEMEE~

1974

and penetration when both drilling damage and perforation damage are present in. the completed well.
Commonly experienced damage factors are assumed:
k, = 0.4 and k, = 0.2
The importance of perforation depth is highlighted
again, Two shots per foot 18 in. deep is more effective
than 8 shots/ft 4 in. deep, In general, within practical
limits increasing shot density alone cannot overcome
the combined effects of permeability damage from
perforating and drilling or workover, Therefore,
regardless of shot density, deeply penetrating perforations that extend substantially beyond the permeability damage from drilling (or workover) are necessary
if the productivity of a damaged well is to approach
that of an undamaged, open-hole completion.
Our conclusions on the importance of per.-:ration
in wells with permeability damage are contrary to
the results of previous work by other investigatorse-
who concluded that in ideal, undamaged completions
shot density is more important than perforation depth.
This work strongly emphasizes the importance of
careful attention to completion practices to minimize
the depth and severity of formation damage from
drilling, perforating, and workover,

Conclusions
Through use of a finite element model, we have extended previous studies of the productivity of a well

00

6
8
Perforation

IO
12
Penetration.

14
inches

14

18

Fig. 16-Effect
of perforation parameters on well productivity in well with no drilling damage.

:*6
/
,.,

,0

.,

,/

.,... -

..:1

/ ..... ..
/ /..,..

,..

Perforation

Penetration,

inches

Fig. 17Effect of perforation parameter


on well produce
tivity in well with moderate drillin[; damage and normal
perforation damage.

1313

in art ideal radial system to include the effects of


permeability damage from drilling, workover, and
perforating. For illustrative purposes, this paper has
dealt specifically with results for a perforation density
of 4 hcJes/ft in a symmetrical pattern, although the
method can be applied to other densities and patterns.
Our analysis enables us to draw a number of general
conclusions about the effectivenessof gun perforating,
but in the summary below it should be remembered
that the numerical results apply specificallyto the shot
density and pattern used.
1, Damage factors indicated in linear cores from
Core Flow Efficiencies, as determined according to
API RP 43, second edition, can be related to Well
Flow Efficiencies fcr the model described.
2. Permeability in a 95-in.-thick damaged zone resulting from the perforating process ranges from about
0.3 of the undamaged-formation permeability for
good perforating conditions to about 0.01 for adverse
perforating conditions,
3, In a radial system with formation damageand
no perforation damage, Well Flow Efficiency is substantially reduced until the perforation penetrates substantially beyond the damaged zone.
4. In a radial system with both formation damage
and perforation damage, Well Flow Efficiency remains considerably below that for an undamaged system even when the perforation penetrates substantially
through the zone of formation damage.
5. The major effect of permeability damage around
a perforation occurs from the damage within the first
%2 in, of the perforation,
6. Application of our study to a hypothetical perforated completion indicates that productivity may
range fronl as low as 5 to 90 percent of undemaged,
open-hole productivity, depending upon the nature of
drilling and perforating opemtions. Therefore, every
precaution should be taken to avoid permeability
damage to the formation during drilling, workover,
and perforating.
7. If formation damage is avoided during the drilling process, a perforation depth of 12 in. or more is
required to overcome the loss in productivity from
damaged perforations that is indicated for many commercial guns by standard API RP 43 tests (CFE 0.7-0.8). Increasing shot density from 4/ft to 8/ft has
about the same effect as doubling penetration from 6
in. to 12 in.
8. Perforation quality is more important than
either shot density or penetration. The effect of severe
Original manuscrifM received in Society of Petroleum Engineers
office Aug. 6, 1973. Revised manuscript
received July 29, 1974.
Paper (SPE 4654) was first presented
at the SPE.AIME
48th
Annual Fall Meeting, held in Las Vegas, Nev., Sept. 30-Ott, 3,
1973. @) Copyright
1974 American
Institute
of Mining,
Metal.
Iurgical, and Petroleum Engineers, Inc.
This paper will be printed In Transactions
cover 1974.

1314

volume 257, which will

damage in the perforations cannot be overcome by


increasing either shot density or depth of penetration
within the limits of present-day technology and economics.
9. In completions with drilling (or workover) and
perforating damage, a few deeply penetrating perforations are more effective than many shallow perforations.

Acknowledgment
We are grateful to the Union Oil Co. of California for
permission to publish this paper. We acknowledge
with many thanks the efforts of M. L. Garrett who
performed most of the computing work.
References
1. History of Petrolewn Engineering, API Div. of Production, Dallas, ( 1961).
2. Oliphant, S. C., and Farris. R. F.: A Study of Some
Factors Affecting Gun Perforating, Trans.,
Al ME
( 1947) 170. 22 S-237.
3. Lewelling, ~1.: Experimental
Evaluation of Well Perforation Methods as Applied
to Hard Limestone;
Trans., AIME ( 1952) 195, 163-168.
4. Allen, T. O., and Atterbury,
J. H,, Jr.: Effectiveness
of Gun Perforating,
Trans., AIME ( 1954) 201, 8-14,
5. Allen. T. O..,. and WorzeL.> C.
, H.: Productivity Method
of Evaluating Gun Perforating, Drill. and Prod. Prac.,
API (1956) 112.
6. Krue~er, R. F.: Join: Bullet and Jet Perforation Tests,
Progre>s Report, Dr;ll. and Prod. Pruc., API ( 1956)
126.
7. Suman, G, O., Jr.: Perforations
A Prime Source
of Well Performance
Problems, J. Per. Tech. (April
1972) 399-4i I.
8. Muskat, M.: The Effect of Casing Perforation on Well
Productivity,
TM/ns., AIPIE (1943) 151, 175-184.
9. Howard, R. A., and Watson, M, S., Jr.: Relative Productivity of Perforated Casingl, Trans., AIME ( 1950)
189, 179-182.
10. McDowell, J. M., and Muskat, M,: The Effect on Well
Productivity
of Formation
Penetratimi
Beyond Perforated Casing, Trans., Al ME ( 1950) 189, 309-312.
11. Harris, M, H.: The Effect of Perforating on Well Productively, J. Per. Tee/I. (April 1966) 518-528; Tram.,
AIME, 237.
12. Bell, W. T., Briege~, E. F., and Harrigan, J. W., Jr.:
Laboratory
Flow Characteristics
of Gun Perforations;
J. Pet. TCC}I, (Sept. 1972) 1095-1103.
Practice: Standard Procedure for
13. API Recommended
Evaluation of Well Perforators:
API Div. of Production, RP 43, 2nd ed. (Nov. 1971).
]4. Willems, N., and Lucas, W. M,, Jr.: A4ufrix Analysis
/or Srructmaf En~ineers, Prentice-Hall
Inc., Englewood
Cliffs. N. J. (1968).
15. Zienkicwicz, O. C., and Cheun.g, Y. K.: The Finite Element Method in Structural ancf Continnam Mechunics,
McGraw-Hill Publishing Co., Ltd., London ( 1967).
16. Krueger, R, F.: An Evaluation of Well Completion
EtTectiveness, API Paper 801-38P (May 9-10, 1962).
17. Taylor, R. L, and Brown, C. B.: Darcy Flow SoluDiv. Proc.,
tions with a Free Surface, J. Hydrmlics
ASCE ( March 1967) 93, I HYDJ 25.
18. Muskat, M.: Physical Principles oj Oil Production,
McGraw-Hilt
Book Co., Inc., N,Y. (1945) Eq. 6, 244.

tJPT

JOURNAL

OF PETROLEUM

TECHNOLOGY

You might also like