Copyright 2002, SPE/PS-CIM/CHOA International Thermal Operations and Heavy Oil

Symposium and International Horizontal Well Technology Conference

This paper was prepared for presentation at the 2002 SPE International Thermal Operations
and Heavy Oil Symposium and International Horizontal Well Technology Conference held in
Calgary, Alberta, Canada, 47 November 2002.

This paper was selected for presentation by the ITOHOS/ICHWT Program Committee
following review of information contained in an abstract submitted by the author(s). Contents of
the paper, as presented, have not been reviewed by the Society of Petroleum Engineers, the
Petroleum Society of CIM, or CHOA and are subject to correction by the author(s). The
material, as presented, does not necessarily reflect any position of the Society of Petroleum
Engineers, the Petroleum Society of CIM, or CHOA, its officers, or members. Electronic
reproduction, distribution, or storage of any part of this paper for commercial purposes without
the written consent of the Society of Petroleum Engineers, Petroleum Society of CIM, or CHOA
is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE,
P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

Abstract
Approximately 100 pressure pulse workovers have been
performed on heavy oil wells in Canada since 1998. Analysis
of a number of the wells shows varying degrees of success,
but in general, the workovers have paid for themselves in
almost all recent cases. Although there is insufficient
information to fully prove hypotheses using statistical criteria,
it appears that pressure pulse workovers have particular
application in three situations:
In cases where the CHOPS well did not initially start
producing sand during the primary completion attempts;
In cases where the well became disconnected from the
far-field pressure and gravity driving forces; and,
In cases when it is desired to place a workover chemical
in a well-dispersed fashion around a well.
This article will review the physical mechanisms and drive
energy sources that are believed to be responsible for CHOPS
production; it will also present the physical processes that we
believe occur during aggressive pressure pulse workovers
around a heavy oil well.

Introduction
The Canadian heavy oil and oil sand deposits (Fig 1) comprise
the largest known petroleum accumulation in the world, with
over 40010
9
m
3
OOIP (2.510
12
b). If only 30% of this
resource can be economically extracted in the foreseeable
future, 100% of Canadian and US oil needs at current rates of
consumption (about 2010
6
b/d) could be met for the next 100
years. Given recent events, pursuit of this resource more
aggressively seems eminently reasonable. Furthermore,
existing technologies, Steam-Assisted Gravity Drainage
(SAGD), Vapor Assisted Petroleum Extraction (VAPEX),
Cold Heavy Oil Production with Sand (CHOPS) and pressure
pulse flow enhancement can likely achieve this recovery ratio.

Figure 1: Disposition of Major Heavy Oil and Oil Sands Deposits
in Canada (Alberta and Saskatchewan)

CHOPS is now responsible for over 550,000 b/d oil
production in the Alberta and Saskatchewan Heavy Oil Belt,
over 20% of total Canadian production. The limits on
production rate are not related to lack of suitable fields, lack of
profitability, or lack of technological capability: limits are
because of a lack of heavy oil upgrading capacity in Canada
and the mid-west United States. Approximately 850,000 b/d
is produced from all sources (thermal, non-thermal, CHOPS,
SAGD), and only about 150,000 b/d can be upgraded in
Canada: the rest is shipped to upgrading facilities in the
United States (Minneapolis, Chicago, Billings). Because of
the shortage of upgrading capacity, the difference between the
price offered for heavy oil and that offered for light oil in

SPE/Petroleum Society of CIM/CHOA 79033
Pressure Pulse Workovers in Heavy Oil
Maurice Dusseault, University of Waterloo, Darrell Shand, Brett Davidson, PRISM Production Technologies Inc.
Edmonton
Calgary
Saskatoon
Regina
Alberta Saskatchewan
Lloydminster
Athabasca
deposit
Cold Lake
deposit
Peace
River
deposit
Heavy
Oil
Belt
Ft McMurray
Major cities
Heavy oil areas,
~500-20,000 cP
Extra-heavy oil,
> ~20,000 cP
Carbonate triangle
W
a
b
is
k
a
w

d
e
p
o
s
it
Edmonton
Calgary
Saskatoon
Regina
Alberta Saskatchewan
Lloydminster
Athabasca
deposit
Cold Lake
deposit
Peace
River
deposit
Heavy
Oil
Belt
Ft McMurray
Major cities Major cities
Heavy oil areas,
~500-20,000 cP
Heavy oil areas,
~500-20,000 cP
Extra-heavy oil,
> ~20,000 cP
Extra-heavy oil,
> ~20,000 cP
Carbonate triangle Carbonate triangle
W
a
b
is
k
a
w

d
e
p
o
s
it
2 SPE/PS-CIM/CHOA 79033
Edmonton averaged over US$9.00/b in 2001, rather than
US$4.50-5.00/b, which is probably the level at which
upgrading costs can be met, along with a profit.
In the right reservoirs, CHOPS has become the favored
approach for production of heavy oil
1
because it gives a period
of reasonable primary production rate with a modest to low
investment. This is because exploitation is through low-cost
vertical or slightly inclined wells, and operating expenses have
been reduced to about US$4.50/b through the use of
progressing cavity pumps and better approaches to
sand management.
To execute rational screening for different workover
technologies, it is necessary to have a good understanding of
the physical mechanisms that may have led to the need for a
workover. This requires that a basic understanding of the
mechanisms and the sources of drive energy in CHOPS be
achieved. This is not straightforward because of the distinctly
non-conventional aspects of CHOPS processes and because of
the massive changes in reservoir properties that occur over the
life of CHOPS wells.
Workovers in CHOPS wells cannot be carried out like
workovers in conventional wells; because the reservoir is
continuously evolving in terms of rock properties. The
reasons for this will be reviewed, and a brief outline of some
recent developments will be presented to allow pressure pulse
workovers to be assessed in a broader sense.

CHOPS Production Mechanisms
CHOPS entails continuous sand influx into the wellbore along
with heavy oil.
2
The sand consists of the entire formation
material, not just fine-grained particles, although only one
zone may contribute the majority of the sand being produced.
Clearly, this flux of sand has to be brought to the wellbore
through some form of drive energy, and the production of
hundreds of cubic meters of sand per year from a single well
must alter the physical characteristics of the formation
surrounding that well.
It has been unequivocally proven, for the Canadian heavy
oil deposits and undoubtedly will be for other similar
reservoirs,
3,4
that sand influx permits an increase in the oil
production rates of vertical wells by a factor up to twenty in
some cases.
5
A vertical heavy oil well that produces only 20
b/d of 5000 cP oil through use of a sand exclusion screen may
produce from 100 to 400 b/d once the screen is removed and
sand encouraged. This oil will be produced along with about
3-5% sand and an amount of solution gas that is proportional
to the volume of oil produced, with a constant GOR over time.
There appear to be five direct production mechanisms
associated with oil production using massive sand influx:
6

Increased fluid flux associated with the sand flux;
Growth of a zone of enhanced permeability because of
sand withdrawal;
Accelerated flow toward the wellbore because of
expansion of a dispersed bubble phase and greatly
retarded solution gas depletion because a continuous gas
phase is not formed in the intact reservoir portion;
Compaction of the weakened and dilated formation
formed by the sand production; and,
Elimination of the formation of near-wellbore mechanical
skin through asphaltene precipitation and fines blockage.

General Yield or Piping Channels?
There appear to be two limiting physical structures that can
be created through the continuous long-term production of
sand: a general dilated zone of higher porosity and
permeability but roughly homogeneous; or, a region of
intact matrix transected by open piping channels from which
the sand has been removed by liquefaction at the tip of the
channel. Interestingly, the effects on flow rate to the well are
similar, and it is not possible to distinguish based on well tests,
nor has any technology been proposed that could identify the
nature of the altered zone, except perhaps in the immediate
wellbore vicinity (i.e. 10-30 cm from the casing).
If the system is a piping network,
7
there are two limiting
structures: a self-similar dendritic net with repeated channel
bifurcations with distance from the well; or, a number of
channels that is fixed or that increases only slowly with
distance from the well.
These three structures are shown in Fig 2. It is likely
that the reality in the ground is a combination of these
structures at different locations, and perhaps also an evolution
of the dominant structures over the life of the well. The
writers do not believe that it is possible to propagate small
piping channels (20-40 mm diameter) in dense sands ( 28-
31%) because of effective sand grain arching in dense sand;
rather, we believe that the large overburden stresses that are
redistributed during sand production lead to lateral unloading,
yield and dilation. This process destroys any fragile cohesion
and makes it possible for channels to develop.
Note that in the case of the piping channel hypotheses, oil
is oozing into the flow system not only from the propagating
tips of the channels, but also from the flanks of the channels,
so that the high sand content slurry becomes progressively
more diluted during transit to the wellbore.

Increased Darcy Flux
If some portion of the solid matrix can flow at velocity v
f
,
the Darcy velocity, v
f
v
s
, changes, and the fluid velocity
relative to a fixed reference frame increases. During early
time, immediately after a CHOPS well is placed on
production, the sand cut can be as high as 30-40% sand by
volume of the total dead (gas-free) produced materials. At this
stage, there is an appreciable enhancement of fluid flux, by as
much as a factor of two compared to a static matrix case.
If the nature of the sand producing mechanism is mainly
the propagation of piping channels into an intact matrix, it is
likely that at the tip of the channel the flux of the sand grains
reduces flow impedance to the oil. Also, in the channels, the
hydraulic diameter is large enough that there is a greatly
reduced resistance to flow of the sand-oil-gas slurry toward
the wellbore.

SPE/PS-CIM/CHOA 79033 3

Figure 2: Possible Macroscopic Reservoir fabrics Generated
through Massive Sand Production:
a. General shear, dilation and yield (homogeneous)
b. Dendritic (self-similar) network of piping channels
c. Constant number of piping channels

Zone of Enhanced Permeability
Whether the reservoir structure is a network of channels, a
disturbed zone of higher porosity and permeability, or some
combination of the two, the effect of continuous sand influx is
to generate an ever-larger effective wellbore radius, which is
beneficial to flow. Only when the effective wellbore radius
approaches the interwell scale (i.e. approaches the pattern-
fixed drainage radius) will the beneficial effect on flow rates
taper off. Various models can be postulated to simulate the
enhanced flow rates, and the results depend strongly on the
assumptions associated with the nature of the reservoir
structures. Testing which permeability enhancement model is
likely to be correct through well flow testing and type curve
analysis has proven impossible. No well flow software
addresses the mechanisms discussed in this section, therefore
software packages can give results that are simply irrelevant,
but may be seriously misleading to those seeking to
understand the process.

Solution Gas Drive
In conventional oil wells in reservoirs drawn down below
the gas bubble point (p
w
< p
b
), it is supposed that there is a
zone, called the induction zone, where gas bubbles come out
of solution, coalesce and form a continuous phase at saturation
values of about S
g
~ 15-25%. (S
g
+ S
w
+ S
o
= 1.0). It is
further assumed that the width of the induction zone is small
with respect to the characteristic flow path length of the fluids,
so that the induction zone can be assumed to have little effect
on overall flow behavior. The permeability of the three-phase
region is treated using multiphase permeability concepts, and
the induction zone width and effects are ignored. Except in
the early well life, this gives little error in conventional oil.
In most heavy oil deposits in Canada, the bubble point is
fairly close to the pore pressure (p
b
= 0.7 1.0p
o
), and the
Henry solution gas constant appears to be the same for most
heavy oils (0.20/bar or 0.20 vol/vol/100 kPa), although of
course the total gas content is a function of the initial pressure,
and shallower heavy oil deposits thus have less gas in solution.
Similar figures for Henrys constant have been determined for
the Venezuelan Faja del Orinoco, and in several Chinese
heavy oil deposits (Liaohe, Shengli, Jilin).
When the well pressure drops below p
b
in heavy oil,
bubbles form,
8,9,10,11,12
but the low diffusion rate of CH
4
in the
viscous oil,
13
the laminar flow (reduced Brownian motion),
and other factors result in a low bubble capture rate or
diffusive growth rate, so that the induction zone remains at a
scale equal to the characteristic flow path length throughout
the life of the well, and the gas phase flows to the well bore as
bubbles along with the sand and liquids. Occasionally, a small
gas cap will form near the wellbore, but no continuous gas
phase develops in the reservoir over the life of the well.
Furthermore, the identification of virgin reservoir pressures in
the interwell region of 10-acre spaced wells, plus a constant
GOR over time, both prove that a critical gas saturation is not
reached in the high porosity sands over the life of the well.
This means that as long as the liquefaction front (channels
or otherwise) is propagating into new reservoir zones, there
will be virgin pressures to help destabilize the sand and cause
flow to the well. Because gas bubbles remain discrete, they
expand during transit to the wellbore (where p
w
is invariably
very low, 200-500 kPa) and provide an internal drive force
that is absent in conventional oil. This is often referred to as
foamy oil drive, and the lower the p
w
, the better it should be,
liquefied
yielded
r
intact
region
intact
region
a)
b)
c)
shearing and weakening
liquefied
yielded
r
intact
region
intact
region
a)
b)
c)
shearing and weakening
4 SPE/PS-CIM/CHOA 79033
giving an incentive to operate wells with as low an annular
fluid level as possible (BHP gauges are recommended).

Compaction Drive
Fig 3 shows vertical stress distributions in a CHOPS
reservoir at early, intermediate and late time. In early time
stresses are changing because of near-wellbore sand
production, but the strata are successfully carrying the loads.
In intermediate time, large stresses are imposed on the intact
reservoir, causing it to shear and dilate (if it is a dense sand),
but perhaps to compact somewhat (if it is a high porosity
sand), generating some induced compaction.
Figure 3: Stress Redistributions around a CHOPS well
a. Early time, stress arching, little shear and dilation
b. Intermediate time, shear and dilation
c. Late time, compaction of dilated material
Frame c of Fig 3 shows the situation after the disturbed
zones have grown to the reservoir scale: stresses are re-
imposed on the dilated and loose zones, compacting them
from their high porosity to a lower value, and contributing
some late-time compaction drive to the process. It is believed
that in the Canadian 30% porosity reservoirs, this happens at
late time only, and requires that the interwell pillars be
successfully destabilized.

Removal of Blockages
If sand is free to move, particularly in the near-wellbore
environment, blockages such as precipitated or agglomerated
asphaltenes and fine-grained mineral cannot accumulate to
block flow paths, and there is no stable substrate for the
precipitation of minerals to block pore throats. Thus, as long
as sand is moving, the well behaves as if it has a massively
negative skin.
Empirically, it is well known that use of any technology
that seeks to exclude sand results in massively lowered
production that continues to diminish with time. It is believed
that a filtration process takes place, with smaller flow channels
being progressively more and more blocked with flowing
solid materials.

Spatiotemporal Changes of Mechanisms
With the possible exception of the blockage removal, these
five mechanisms do not act equally throughout the life of a
CHOPS well. For example, at early time when sand cuts
approach 30-40%, enhanced Darcy flux is clearly much
stronger that late in the well life when the sand flux is only a
few percent. On the other hand, compaction is a late-time
effect in the quartzose and dense Canadian reservoirs, Fig 4 is
an attempt to qualitatively show the relative effect of the
five mechanisms.
The relative weights of the five processes are as yet ill
understood, largely because of a lack of detailed monitoring
data on CHOPS field from the beginning of production. This
is due to the low profit margins traditionally associated with
heavy oil, resulting in aggressive cost-cutting to reduce
operational costs, which in turn means that careful monitoring
and quantification of processes over time is sacrificed.

Sources of Drive Energy in CHOPS

Solution Gas Drive
CHOPS is widely viewed as a solution gas drive process.
Empirical evidence in Canada shows that if the reservoir has a
low gas content relative to its virgin pore pressure, it is more
difficult to sustain CHOPS processes. The nature of this drive
energy is worth exploring in a bit more detail.
The low diffusivity of gas in solution in viscous heavy oil
means that the mean-free-path length for gas molecule
Brownian motion is small, compared to lower molecular
weight, less viscous liquids. It appears that there is retardation
in gas nucleation so that the formation of bubbles is a slow
process and the liquid can remain supersaturated with no free
gas bubbles for a long time after the pore pressure is reduced
disturbed
zones
sanding wells
Note:
v
p
stresses now flow without arching
Reloading of
reservoir
compaction
triggered
a.
interwell zones
are shearing
and dilating
stress paths
b.
c.
arching
disturbed
zones
sanding wells
Note:
v
p
disturbed
zones
sanding wells
Note:
v
p
stresses now flow without arching
Reloading of
reservoir
compaction
triggered
a.
interwell zones
are shearing
and dilating
stress paths
b.
interwell zones
are shearing
and dilating
stress paths
b.
c.
arching
SPE/PS-CIM/CHOA 79033 5
below the bubble point. This can be explained in a
straightforward manner.
Figure 4: Production Enhancement Mechanisms over Time

Approximately 40-60 CH
4
molecules are required to form
a stable bubble that does not spontaneously collapse because
of the small radius of curvature. Statistically, in a low
viscosity fluid, this number of molecules can be assembled
through random motion quite easily, so that bubbles nucleate
rapidly and then grow by further gas diffusion into the
bubbles. In heavy oil, the statistical probability is so low that
bubble nucleation is improbable, and once bubbles form,
diffusive growth is slow. Similarly, workover methods
attempting to recharge the near-wellbore through the
injection of CH
4
are doomed to failure because of the
extremely slow process of dissolution into the oil.
Bubbles also grow by diffusional capture of smaller
bubbles and by the coalescence of two small bubbles into a
larger single bubble. These processes are also retarded by the
low diffusivity of the gas in the heavy oil, and by the slow
laminar flow rates characteristic of CHOPS production.
There has been speculation that bubbles are further
stabilized by a coating of polar molecules so that the outside
of the bubbles are similarly charged, leading to repulsion, but
this has not been demonstrated at this time.
An important aspect of the solution gas drive in CHOPS is
that the GOR values tend to be stable over time. This means
that only the gas in the oil being mobilized is available, and
gas in immobile oil remains undepleted. It appears that the
shear distortion of the reservoir leads to dilation, which
provides sufficient free volume for bubbles to form, and the
bubble growth and pore throat blockage help liquefy the sand.
Thus, in the absence of sand flux and continued shearing and
destabilizing of the sand structure, the solution gas drive is not
highly effective. This seems to be confirmed in many cases
by the low production rates per metre of well length for long
horizontal wells, by the rapid drop in productivity of such
wells (limited true radius of influence of the drawdown), and
the low recovery factors (generally less than 10% OOIP
overall, albeit higher in high k zones).
The nature of the solution gas drive mechanism is thought
to be the reason why thermal stimulation, such as injection of
a large slug of steam, has only temporary beneficial effects
and often results in a cessation of CHOPS after some time.
Also, experience has shown that wells that have previously
been steamed in an attempt to generate thermal production are
extremely poor candidates for conversion to CHOPS wells.
The high temperatures allow the gas to come out of solution
rapidly, form a continuous gas phase, and bleed off the gas in
adjacent heated zones, to the detriment of the bubble
drive mechanism.
Finally, based on field evidence of virgin pressures
remaining for long times in the interwell regions, it is believed
that for the more viscous cases (~10,000 cP), the oil in situ is
genuinely immobile in response to the modest gradients that
can be generated in a typical shallow heavy oil reservoir. This
could be evidence of some gelation structure in the virgin oil
because of the high content of polar molecules, and that it
takes some initial perturbation (shear) for the Bingham-type
gel structure to be broken down so that liquid flow can take
place. This in situ gelation is generated over geological time;
it is disrupted by core damage, and thus will prove extremely
difficult to confirm or disprove unequivocally. It is likely not
possible to reconstitute such a structure in the laboratory over
typically short experimental times, and in situ probing using
seismic or magnetic resonance methods remains
problematical.

Gravitational Drive
The vital role of gravitational forces in sustaining CHOPS
has not been widely addressed because it is difficult to
simulate in any conventional manner, and because it is not
understood. However, understanding this drive force is vital
to planning successful workovers in CHOPS wells.
As sand is produced from the near-wellbore region during
early CHOPS, the lateral stress that naturally holds back the
sand is removed, but the vertical stress is not.
14
This results in
shearing of the sand, and because the lateral stress is low, this
leads to dilation, and goes from 30% to 35-40% in the
sheared region. This destroys any natural gelation in the oil
and provides volume for bubbles to be generated. The sheared
and dilated sand is under very low confining stress, and can be
more easily diluted, liquefied and carried toward the wellbore.
However, the weakened zone is essentially incapable of
supporting vertical stress, so the stress redistribution that must
occur (Fig 3b) leads to stress concentrations, more shearing,
and a self-sustaining process that leads to outward propagation
of the dilated and high permeability region.
Is there independent evidence that this takes place? There
are three lines of evidence that, taken together, seem to be
0
1.0
Enhanced Darcy flux
Increased permeability zone
Solution gas drive
Compaction drive
Near-well blockage elimination
early time intermediate late time
M
E
C
H
A
N
I
S
M
0
1.0
Enhanced Darcy flux
Increased permeability zone
Solution gas drive
Compaction drive
Near-well blockage elimination
Enhanced Darcy flux
Increased permeability zone
Solution gas drive
Compaction drive
Near-well blockage elimination
early time intermediate late time
M
E
C
H
A
N
I
S
M
6 SPE/PS-CIM/CHOA 79033
convincing. First, the few surface subsidence measurements
that have been taken show a small (10 cm scale)
downwarping, confirming that downward overburden
movement is taking place. Second, in an instrumented test in
the Morgan Field in Alberta, interwell microseismic shear
events were identified, implying episodic (stick-slip) shearing
of the strata. Third, surface seismic surveys have identified
large low velocity, high impedance zones around good
CHOPS wells,
15
and no zones around wells that failed to
produce large quantities of sand. This information confirms
the existence of shearing and dilation, even in the absence of
other mechanisms such as the possible formation of piping
channels (which may nonetheless be developing locally).
The energy that this gravitational force can provide is quite
large, on the order of several teraJoules for one metre of
overburden motion for a 1000 m deep well on 250 m spacing.
Thus, in addition to the more classically understood
compaction drive that may occur later in the well life when
the high porosity zone is fully reloaded (Fig 3c), gravity forces
appear to be a vital factor in the continued destabilization of
the sand, which is essential for successful CHOPS. However,
successfully incorporating this drive mechanism into reservoir
simulation remains problematic, which is one of the basic
reasons that physics-based simulation of CHOPS remains a
challenging task, and massive model calibration on empirical
data remains the favored means of extrapolation of production
histories and prediction of new well performance.

Workover Approaches
This area is in a state of rapid evolution. Some articles
have been published,
16,17
but history seems to be continuously
overtaken by new methods.
Two general methods are of interest in this article: those
that seek to perturb the near-wellbore region (out to perhaps 4-
6 m), and those that seek to perturb the reservoir at a great
distance (> 10 m), perhaps approaching the interwell scale in
the case of pressure pulsing. A brief review is given here.

Why does Production Stop?
Leaving aside mechanical problems, pump wear, and so
on, the major reservoir reasons for production losses appear
to be the following:
Well perforations become blocked by sand and gravel
grains as well as precipitated asphaltenes.
The near-wellbore region becomes re-packed with
sand that cannot be destabilized and liquefied by well
drawdown during continuous production.
Access to distant virgin solution gas drive forces is
stopped, a process we call well disconnection from
the far-field.
The interwell regions (Fig 3b) become stabilized, and
shearing ceases to be active, eliminating the beneficial
effects of gravitational drive forces on the sand.
Depletion of solution gas at the reservoir scale
gradually reduces and eventually eliminates gas
expansion drive forces.

Near-Wellbore Methods
These include all methods of injecting limited quantities of
fluids (5-20 m
3
), aggressive swabbing, chemical placements,
foam clean-outs, mechanical sand bailers, and the like. They
are intended to unblock the perforations and near-wellbore
regions to re-establish sand flow.
Vibrational workovers using frequencies of 10-50 Hz have
proven effective in loosening blockages near the well,
allowing particles to be re-mobilized and flushed out, but this
effect is local because of the high frequencies and the low
amplitudes that can be achieved. Furthermore, loose sand
with gas bubbles is a highly adsorptive medium.
An issue with all workover methods in CHOPS wells is
that, without pump-to-surface methods or foams (low density),
it is impossible to obtain returns at the surface. This occurs
because sand production reduces the fracture pressure gradient
to as low as 8-10 kPa, below the weight of a column of pure
water. Thus, liquid addition or flushing attempts generally
push sand and other near-wellbore blocking agents farther out
into the formation, thereby allowing better local flow.

Reservoir Perturbation Methods
Methods that shock or perturb the reservoir at a larger
scale have proven to be effective in CHOPS wells, and
perhaps could have a much larger application in conventional
wells experiencing flow impairment.
Reperforating serves to install larger diameter flow ports
that are less prone to blockage, and the shock associated with
explosive discharge creates a shock wave that travels some
distance out from the wellbore. Thus, perforation unblocking
and near-wellbore perturbation can help re-initiate sand flux to
the wellbore. However, the region around the wellbore is
extremely soft and there is usually a high gas content, albeit in
the form of bubbles. This serves to severely attenuate any
shock wave and reduces the efficacy of reperforation.
Apparently, this method is best in initiating initial sand
production in zones where it proved impossible to generate
stable sand influx during the initial completion, or in
rehabilitating old wells that were never very productive of
sand and hence oil because of small perforations.
Rocket propellant discharge down-hole with the wellhead
shut-in gives a large impulse to the formation. Because a
large amount of gas is generated and liquid is forced through
the perforations very aggressively, the magnitude of the shock
wave is large enough to propagate effectively into the more
remote reservoir regions, although the effect is probably
inconsequential beyond perhaps 15-25 m.
Pressure pulsing is now discussed in greater detail.

Pressure Pulse Workovers
The concept of prolonged aggressive liquid phase pulsing
18

arose in part through a gradual appreciation of the production
loss mechanisms typical of CHOPS wells, combined with an
improved theoretical understanding of incompressible
behavior of liquids that lead to the possibility of generating
displacement waves that are highly conservative and of low
frequency (hence low attenuation). Pulsing from the surface is
SPE/PS-CIM/CHOA 79033 7
not possible for several reasons, the major ones being the
impossibility of filling the wells with liquid and the issue of
the impulse magnitude required to suddenly accelerate a 500-
800 m column of liquid. Bottom-hole pulsing is necessary.
There are two basic options for pulse generation: use of the
entire tubing string as a drop weight, forcing all the casing
liquid below the PPT tool through the perforations; and, use of
a bottom hole independent piston driven by mechanical-
hydraulic means. The former will be discussed in greater
detail (Fig 5).
Figure 5: Schematic of a Pulse Workover Tool

In a typical PPT workover, all hardware is removed from
the hole, the tool on tubing is located above the perforations,
and stroking is carried out by the service rig draw-works. The
tool is lifted slowly (15-60 s) to allow fluid to enter the
wellbore through the perforations and from the annulus, then
dropped and allowed to free-fall for a distance of 4-16 m.
Typically, the tool reaches terminal velocity in less than two
seconds, and the downstroke takes 2-6 seconds. The process
is repeated for 45-90 min, then paused for 15 min, at which
point a liquid level in the annulus is taken to guide decisions
as to liquid input.
Liquid recharge can be entirely from backflow into the
chamber, or the annulus or tubing can be filled or pressurized
so that any portion or even 100% of the fluid impelled into the
reservoir is fed from surface. This allows slow trickling in of
well-mixed workover fluids (e.g. 5% surface feed), or even
operation of the PPT tool as a positive displacement impulse
pump (100% surface feed).
Pulsing for 4-5 hours was typical for initially workovers,
but more recent workovers have generally been for 8 hours,
even 24 hours in some cases. At the end of the workover, the
tool is withdrawn slowly from the hole, and the well placed
back on production.
Bottom-hole pressure responses (Figs 6, 7) show high p
during active pulsing. In the first example the impulse was
300-350 psi for most of the workover in this case. The build-
up of p in the third pulsing cycle (each cycle consisted of ~
60 strokes) is thought to be the driving of free gas bubbles
back into solution, allowing a higher impulse. The casing was
recharged with 2-3 m
3
of somewhat lighter oil at the end of
each station stop.
Figure 6: BHP Response for a Pulse Workover (Initial p = 200 psi,
final static p = 238 psi, peak p = 400 psi)

Figure 7: BHP Response (Initial p 145 psi, final static p = 465 psi,
peak p ~900 psi)

In the second example, an increase in the static pressure of
over 300 psi (2 MPa) was achieved, although the impulse size
remained the same, indicating that the resistant to the
excitation of the formation remained constant. Both of these
wells were economically successful workovers.

f
o
r
m
a
t
i
o
n
liquid
bypass
valve
steel casing
perforations
tool body
open base
tubing
centralizers
seals
f
o
r
m
a
t
i
o
n
liquid
bypass
valve
steel casing
perforations
tool body
open base
tubing
centralizers
seals
steel casing
perforations
tool body
open base
tubing
centralizers
seals
0 1 2 3 4 5 6 7 8 9
0
100
200
300
400
station stops
Pressure (psi)
Time - hours
Well XXX-63-5-W4
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
0
100
200
300
400
0
100
200
300
400
station stops
Pressure (psi)
Time - hours
Well XXX-63-5-W4
Pressure (psi)
8 m stroke each 50 sec,
station stops every 1.2 hr
Time - hours
200
400
600
800
0 1 2 3 4 5 6 7 8 9
0
Pressure (psi)
8 m stroke each 50 sec,
station stops every 1.2 hr
Time - hours
200
400
600
800
200
400
600
800
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
0
8 SPE/PS-CIM/CHOA 79033
Mechanisms in Pressure Pulsing
If used entirely as a pump, liquid is displaced into the
reservoir, and the gas saturation is reduced as pressure
increases. This causes a displacement front to propagate from
the wellbore, and the impulse is less and less attenuated as the
gas saturation drops.

Initiation of Sanding Through Sand Disaggregation
In a well that did not respond to conventional completion
approaches to initiate sand influx (for reasons largely
unknown this happens in ~10-12% of cases), pulsing develops
very sharp pressure spikes at the sand face, enough to reduce
the matrix stress to zero, and allow local complete remolding
of the sand. The extent of such a remolded zone after 6-20 hr
of pulsing is unknown, but probably it is several meters radius
around the well.

Blockage Removal
Repeated perturbations in an old well that is blocked have
a cumulative effect on the near-wellbore, opening up more and
more perforations with time (as shown by the BHP response in
many cases), and also remolding and liquefying any sand that
has become recompacted over time around the wellbore.
Typically, these cases show partially rejuvenated sand (and
oil) flux when production is recommenced.
The volume of liquid injected in a single pulse stroke
depends on the tool design and the approach taken; for the
full-casing displacement option in standard 7 casing, the
volume is 25 liters per metre, or approximately 200 l (~1.3 b)
for a stroke of 8 m. Assuming 100 open 25 mm perforations
and constant displacement in 4 seconds, the velocities
approach 100 m/s, far lower than a rocket propellant surge, but
substantially larger than any reasonable workover injection
rate from surface that is displacing heavy oil through
the perforations.

Workover Chemical Mixing and Dispersion
A serious problem with all attempts to place treatment
liquids (surfactants, diluents) around heavy oil wells is
channeling, resulting often in almost pure treatment liquid
flowback when the well is placed on production. Channeling
can occur because fracturing as well as because of poor
mobility ratios (workover fluids are usually far less viscous
that the formation oil). Because the sharp pulses tend to
suppress fingering and partially overcome capillary blockage
at pore throats, treatment chemicals are well-dispersed around
the wellbore.
In addition to superior dispersal, if chemical placement is
through partial recharge from the surface, and the great
majority of the recharge is through backflow through the
perforations, chemicals are intimately mixed with formation
fluids through the high shear rates of the fluids through the
perforation ports. This means that the chemicals are being
mixed directly downhole with the virgin reservoir fluids
themselves, and potential fluid incompatibility problems and
the need for aggressive surface mixing are thereby diminished.
Far-Field Impulse Propagation
Apparently, pulsing is the only way to affect the far-field,
inter-well region. At the characteristic frequencies of the
impulse phase (< 1 Hz), liquids behave incompressibly, and
this leads to the propagation of a displacement wave at a low
velocity (~50 m/s).
19,20,21
The incompressible behavior causes
short-term but large magnitude liquid accelerations at the pore
throats, which serve several purposes, including break-down
of any inherent gelation, overcoming of capillary blockages,
and triggering of formation shearing in the interwell region.
22

If the formation is fully liquid saturated (no free gas), the
displacement wave attenuates very little (except for
geometrical spreading of course), and this allows the effect to
propagate to the inter-well scale with time. Observations on
offset wells 200-300 m distant during pulse workovers
occasionally show associated liquid level responses and even
some increases in sand and oil flux, proving that the effects
propagate far. If there is a lot of free gas, it has proven
expedient to feed in oil during the workover to help resaturate
the well region so that the pulse propagates farther with lower
amplitude loss. In some cases 25-30 m
3
of oil has been used.
The triggering of shearing in the interwell region,
confirmed directly in one case only through microseismic
monitoring, can be likened to the cyclic pore pressure
perturbations during earthquakes that lead to the liquefaction
of shallow, high porosity water-saturated sand. In the case of
a CHOPS well, it is re-triggering of the gravitational drive
effects that takes place.

Comparative Study of Pulse Workovers
To date, a comprehensive assessment of the efficacy of the
various heavy oil well workover alternatives has not been
carried out because of the vast data collection effort from a
number of companies that would be required. Detailed sand
cut analyses and day-to-day production records are seldom
kept, and different companies use different metrics to define
success. Furthermore, because workover techniques (time of
treatment, pump rates, swabbing methods) are evolving so
rapidly, it is almost impossible to collect a sufficient data base
on one static method among many variants to allow
systematic and meaningful comparisons to be made.
Using the standard industry measure of success, 90-day
payback of workover costs in terms of increased oil
production, the great majority of pulse workovers in the last
three years have been successful. However, it should also be
pointed out that the workovers are usually executed on
problem wells after other approaches have failed, and it is
impossible to work miracles, particularly in cases of poor
geology and substantial general pressure depletion.
A typical well performance before and after a pulse
workover is shown (Fig 8). This well had actually been on
production for over 20 years before 1998, and had been
providing about 50 m
3
/mo (~10 b/d) for many years. In the
two years before the workover, oil production had declined
and water/oil ratios had risen. After the workover, an average
production of 50 m
3
/mo was re-established for at least a year.
SPE/PS-CIM/CHOA 79033 9
Fig 8: Fluid Production Before and After a Pulse Workover

One approach that serves to illustrate the progress being
made is to compare a recent randomly selected pulse workover
set to an earlier set. There is only a four-year history, and
approximately 100 workovers to date. The first 20 workovers
from Jan 1998 were grouped, and the last 20 workovers to
date were also grouped. Monthly production of water and oil
are plotted (left axis), and the oil production as a ratio of all
produced fluids is plotted on the right axis.


Figure 9: First 20 Workovers (1998)

The first 20 workovers were performed mainly on wells that
had been better producers than the last 20 workovers. This
makes an absolute comparison impossible. Rather, the
relative trends must be examined. In the first set, the dropping
oil rates were arrested, and the rising water rate was also
arrested, although at a somewhat higher rate than before the
workovers. Assuming that the wells would have been all shut
in permanently, the pulse workovers resulted in an aggregate
35,000 m
3
additional oil.


Figure 10: Production From the Last 20 Workovers

For the case of the last 20 workovers, largely attempts to re-
establish some production from extremely poor and depleted
wells in thinner zones than the first 20 workovers, a higher
post-workover oil rate is apparent, but the oil rate declined
more rapidly than in the first 20 wells. Incremental oil is
about 10,000 m
3
.

Discussion
Better screening criteria have evolved for pulse workovers, but
the wells that were treated recently were of lower quality.
Based on the economic success criterion, a higher ratio of
successes now take place, even though the overall oil return
rate is not as high for the recent wells. It is difficult to squeeze
a little bit of extra value from poor wells that have produced in
some cases since 1942 (with long inactive periods), but it
appears to be possible to do this profitably.
Pulse workovers are not recommended in cases of massive
depletion, nor in cases where it is apparent that a gas cap has
developed or there is high gas saturation near the well.
Several additional cases deserve mention. First, in this
data base, a few cases where pulsing succeeding in bringing a
well on production after the primary completion failed are not
included, and these were all successes. Second, a series of
chemical placements incorporating pulsing achieved
reasonable success,
23
indicating that a combination of the
various effects (perforation unblocking, far-field excitation
and good placement chemical dispersion) may be particularly
useful. Third, only one workover has been executed to date on
a consolidated reservoir, including placement of chemicals,
and it was an economic success.
Extensive application of pulsing on various reservoir types
has not yet happened. Optimistically, we believe that it will
prove to be effective in loosening blockages in the pore throats
and achieving greater uniform dispersion of chemicals, but
01-007 Fluid Production
0
50
100
150
200
250
300
J
a
n
-
9
8
J
u
l
-
9
8
F
e
b
-
9
9
A
u
g
-
9
9
M
a
r
-
0
0
S
e
p
-
0
0
A
p
r
-
0
1
N
o
v
-
0
1
M
a
y
-
0
2
F
l
u
i
d

P
r
o
d
u
c
t
i
o
n

m
3
/
m
o
Oil
Water
Workover
01-007 Fluid Production
0
50
100
150
200
250
300
J
a
n
-
9
8
J
u
l
-
9
8
F
e
b
-
9
9
A
u
g
-
9
9
M
a
r
-
0
0
S
e
p
-
0
0
A
p
r
-
0
1
N
o
v
-
0
1
M
a
y
-
0
2
F
l
u
i
d

P
r
o
d
u
c
t
i
o
n

m
3
/
m
o
Oil
Water
Workover
01-007 Fluid Production
0
50
100
150
200
250
300
J
a
n
-
9
8
J
u
l
-
9
8
F
e
b
-
9
9
A
u
g
-
9
9
M
a
r
-
0
0
S
e
p
-
0
0
A
p
r
-
0
1
N
o
v
-
0
1
M
a
y
-
0
2
F
l
u
i
d

P
r
o
d
u
c
t
i
o
n

m
3
/
m
o
Oil
Water
Workover
F
l
u
i
d

p
r
o
d
.
(
m
3
/
m
o
)
-15 -10 -5 0 5 10 15 20
0.2
0.4
0.6
0.8
1.0
0
2000
4000
6000
8000
10000
12000
F
l
u
i
d

C
u
t
s

(
O
i
l

/

F
l
u
i
d
)
After workover Before workover
oil
water
oil
ratio
Months
F
l
u
i
d

p
r
o
d
.
(
m
3
/
m
o
)
-15 -10 -5 0 5 10 15 20
0.2
0.4
0.6
0.8
1.0
0.2
0.4
0.6
0.8
1.0
0
2000
4000
6000
8000
10000
12000
F
l
u
i
d

C
u
t
s

(
O
i
l

/

F
l
u
i
d
)
After workover Before workover
oil
water
oil
ratio
Months
0
500
1000
1500
2000
2500
3000
-15 -10 -5 0 5 10 15 20
Months
0.2
0.4
0.6
0.8
After workover Before workover
F
l
u
i
d

p
r
o
d
u
c
t
i
o
n
(
m
3
/
m
o
)
F
l
u
i
d

C
u
t
s

(
O
i
l

/

F
l
u
i
d
)
oil
water
oil
ratio
0
500
1000
1500
2000
2500
3000
0
500
1000
1500
2000
2500
3000
-15 -10 -5 0 5 10 15 20
Months
0.2
0.4
0.6
0.8
After workover Before workover
F
l
u
i
d

p
r
o
d
u
c
t
i
o
n
(
m
3
/
m
o
)
F
l
u
i
d

p
r
o
d
u
c
t
i
o
n
(
m
3
/
m
o
)
F
l
u
i
d

C
u
t
s

(
O
i
l

/

F
l
u
i
d
)
oil
water
oil
ratio
10 SPE/PS-CIM/CHOA 79033
this hope will have to wait for the results from appropriate
opportunities.

Conclusions
Pulse workovers in low production heavy oil wells have
proven economically successful.
The evolution of better screening criteria has helped the
success ratio.
Pulsing can bring a great deal of energy to the reservoir
through the strong, repeated impulses used; this has
benefits in both the near-wellbore and the far-field.
Pulsing remains a new technology, tried only in Canadian
heavy oil, and much remains to be learned.

Acknowledgements
The companies who tried the pulse workovers deserve a great
deal of thanks; we hope their attempts result in the
development of a new and useful tool for the business of
oil extraction.

References

1
Adamson, M.F. and Yeung, K.C., Managing cold flow
production at Cold Lake, AOSTRA/CANMET/ACR Oil
Sands Our Petroleum Future Conference, Edmonton,
AB, 1993
2
Chalaturnyk, R.J., Wagg, B.T. and Dusseault M.B., The
mechanism of solids production in unconsolidated heavy-
oil reservoirs, Proc. SPE Int. Symposium on Formation
Damage Control, Lafayette, LA, pp 137-150, SPE
#
23780,
1992
3
Hu, C., Liu, X., Wang, J., Song, Z., Fan, Z. and Yang, F.,
Cold production of thin-bedded heavy oil reservoir in
Hunan Oilfield, Proc. 1998 SPE International
Conference and Exhibition in China, Beijing, China, SPE
#
50885, 1998
4
Dusseault, M.B., Liang, C.X., Gu, W., Ma, Y.Q., Xu, B.C.
CHOPS in Jilin Province, China. Proc SPE Int Therm
Oper Conf and Heavy Oil Symp, Calgary, Nov., SPE
#
79032 (this conference), 2002
5
Dusseault, M.B. and El-Sayed, S. Heavy oil well
production enhancement by encouraging sand influx.
Proc SPE/DOE Improved Oil Recovery Symposium, Tulsa,
SPE
#
59276, 2000
6
Geilikman, M.B. and Dusseault, M.B. Fluid-rate
enhancement from massive sand production in heavy oil
reservoirs. J. of Petr. Science & Engineering, 17, 5-18.
Special Issue: Near Wellbore Formation Damage and
Remediation, 1997
7
Tremblay, B., Sedgwick, G. and Forshner, K., Modelling of
sand production from wells on primary recovery, Jour
Can Petrol Tech, Vol. 37, 3, 41-50, 1998
8
Maini, B.B., Foamy oil flow in heavy oil production, Jour
Can Petrol Tech, 35, 6, pp 21-24, 1996
9
Smith, G.E., Fluid flow and sand production in heavy-oil
reservoirs under solution-gas drive, SPE J. of Production
Engineering, 169-180, 1988

10
Solanki, S. and Metwally, M., Heavy oil reservoir
mechanism, Lindbergh and Frog Lake fields, Alberta, Part
II: Geomechanical Evaluation, Proc. SPE International
Heavy Oil Symposium, Calgary, pp. 87-102, SPE
#
30249,
1995
11
Shen, C. and Batycky, J., Some observations of mobility
enhancement of heavy oils flowing through sand pack
under solution gas drive, Proc. CIM Petroleum Society
47
th
Annual Technical Meeting, Calgary, Paper 96-27,
1996
12
Turta, A.T., Maini, B.B. and Jackson, C., Mobility of gas-
in-oil dispersions in enhanced solution gas drive (foamy
oil) exploitation of heavy oil reservoirs, Proc. CIM
Petroleum Society 52
nd
Annual Technical Meeting,
Calgary, Paper 2001-149, 2001
13
Sheng, J.J., Maini, B.B. Hayes R.E. and Tortike, W.S., A
non-equilibrium model to calculate foamy oil properties,
Jour Can Petrol Tech, Vol. 38, 4, pp 38-45, 1999
14
Dusseault, M.B., Geilikman, M.B. and Spanos, T.J.T.
Mechanisms of massive sand production in heavy oils,
Proc. 7th UNITAR Int. Conf. Heavy Oils and Tar Sands,
Beijing, 1998
15
Mayo, L. Seismic monitoring of foamy heavy oil,
Lloydminster, western Canada, 66
th
Ann. Internat. Mtg:
Soc. of Expl. Geophys., 2091-2094, 1996
16
Dusseault, M.B., Hayes, K.C., Kremer, M., Wallin, C.
Workover strategies in CHOP wells. Proc. CIM PetSoc
Annual Technical Meeting, Calgary, 2000
17
Dusseault, M.B., Hayes, K.C., Kremer, M., Wallin, C.
Workover strategies in CHOPS wells. Jour Can Petr
Tech, 41(4), 2002
18
Dusseault, M.B., Spanos, T.J.T. Davidson, B.C., A new
workover tool - applications in CHOP wells, Proc. CIM
Petroleum Society 50
th
Annual Tech. Meeting, Calgary,
1999
19
Geilikman, M.B., Spanos, T.J.T. and E. Nyland E. Porosity
diffusion in fluid-saturated media, Tectonoph.,217, 11-113.
1993.
20
Spanos, T.J.T. The Thermophysics of Porous Media, Boca
Raton, Chapman & Hall/CRC,
#
126 in Monographs and
Surveys in Pure and Applied Mathematics, 220 p, 2002.
21
Spanos, T.J.T., Udey, N. and Dusseault, M.B. Completing
Biot Theory. Proc. 2
nd
Biot Conf. on Poromechanics,
Grenoble France, Aug., Balkema, p 819-826, 2002.
22
Dusseault, M.B., Shand, D., Meling, T., Spanos, T. and
Davidson, BC. Field applications of pressure pulsing in
porous media, Proc. 2
nd
Biot Conf on Poromechanics
(Grenoble France), Balkema, Rotterdam, 639-645, 2002.
23
Dusseault, M.B., Gall, C., Shand, D., Davidson, B.C.,
Hayes, K.C. Rehabilitating heavy oil wells using pulsing
workovers to place treatment chemicals, Proc. CIM
PetSoc Annual Technical Meeting, Calgary, Paper 2001-
57, 2001